Agglomeration Processes: Phenomena, Technologies, Equipment

Wolfgang Pietsch Agglomeration Processes Phenomena, Technologies, Equipment Wolfgang Pietsch Agglomeration Processes...

Author: Wolfgang Pietsch

161 downloads 1748 Views 56MB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form

DOWNLOAD PDF

Wolfgang Pietsch Agglomeration Processes Phenomena, Technologies, Equipment

Wolfgang Pietsch

Agglomeration Processes Phe no mena, Tech no Iogi e s, Eq ui p me nt

8WILEY-VCH

Dr.-lng. Wolkang Pietsch, EUR INC COMPACTCONSULT, INC. 2614 N. Tamiami Trail, #520 Naples, Florida 34103-4409 USA

In Europe: Holzweg 127 67098 Bad Durkheim, Germany Cover Illustration Like an agglomerate, the picture on the cover is composed of many disparate components, all of which relate to the topics discussed in this book. The panels on the left and right are microphotographs of naturally agglomerated nano-particles. The top and the bottom panels depict different products from spray drying and fluid bed agglomeration. The four sectors (between the panels and the circle) represent Scanning Electron Micrographs (SEMs) of agglomerate structures as well as photographs of coated agglomerates and of granules. The top half of the circle shows products from tumble/growth agglomeration and the lower half are briquettes from roller presses as well as product from compaction/granulation. The center square includes tablets from punch and in die presses. The originals of the individual pictures from which sections are reproduced were supplied by (in alphabetical order): Albemarle Corp., Baton Rouge, LA, USA; Cabot Corp., Tuscola, IL, USA: Eirich, Hardheim, Germany: Euragglo, Qievrechain, France; Niro A/S, Soeborg, Denmark Norchem Concrete Products, Inc., Fort Pierce, FL, USA; Koppern GmbH & Co, KG, Hattingen, Germany. Their support is appreciated and acknowledged.

This book was carefully produced. Nevertheless, author and publisher do not warrant the information contained therein to be free of errors. Readers are advised to keep in mind that statements, data, illustrations, procedural details or other items may inadvertently be inaccurate. Library of Congress Card No.: Applied for. British Library Cataloguingin-Publication Data: A catalogue record for this book is available from the British Library.

-

Die Deutsche Bibliothek CIP Cataloguingin.Pub. lication Data: A catalogue record for this publication is available from Die Deutsche Bibliothek.

Wiley-VCH Verlag GmbH, Weinheim, 2002 Printed on acid-free paper. All rights reserved (including those of translation in other languages). No part of this book may be reproduced in any form - by photoprinting, microfilm, or any other means - nor transmitted or translated into machine language without written permission from the publishers. In this publication, even without specific indication, use of registered names, trademarks, etc., and reference to patents or utility models does not imply that such names or any such information are exempt from the relevant protective laws and r e g ulations and, therefore, free for general use, nor does mention of suppliers or of palticular commercial products constitute endorsement or recommendation for use. Mittenveger & Partner Kommunikationsgesellschaft mbH, Plankstadt Printing betz.druck GmbH, Darmstadt Bookbinding GroBbuchbinderei J. Schaffer GmbH & Co. KG, Griinstadt

Composition

Printed in the Federal Republic of Germany. ISBN 3-527-30369.3

I”

Contents Dedication, Acknowledgements and References

VII

1

Introduction

2

A Short History o f Agglomeration

3

Agglomeration as a Generic, Independent, and Interdisciplinary Field of Science

4

Glossary of Agglomeration Terms

5

Agglomeration Theories

5.1 5.1.1 5.1.2 5.2 5.2.1 5.2.2 5.3 5.3.1 5.3.2 5.4 5.5

The Development of Strength of Agglomerates 32 Binding Mechanisms 35 Binders, Lubricants, and Other Additives 42 Estimation of Agglomerate Strength 55 Theoretical Considerations 55 Laboratory and Industrial Evaluations 62 Structure of Agglomerates 76 General Considerations 78 Porosity and Techniques That Influence Porosity 89 Other Characteristics of Agglomerates 100 Undesired and Desired Agglomeration 109

6

Agglomeration Technologies

7

Tu mble/G rowth Agglomeration

7.1 7.2 7.3 7.4 7.4.1 7.4.2

Mechanisms of TumblejGrowth Agglomeration 140 Kinetics of Tumble/Growth Agglomeration 144 Post-treatment Methods 150 Tumble/Growth Agglomeration Technologies 151 Disc and Drum Agglomerators 153 Mixer Agglomerators 164

1

3

11

29

133 13 9

5

VI

I

Contents

7.4.3 7.4.4 7.4.5 7.4.6

Spray Dryers 187 Fluidized Bed Agglomerators 196 Other Low Density Tumble/Growth Agglomerators Agglomeration in Liquid Suspensions 221

8 8.1 8.2 8.3 8.4 8.4.1 8.4.2 8.4.3 8.4.4

Mechanisms of Pressure Agglomeration 23 1 Structure of Pressure Agglomerates 236 Post-treatment Methods 241 Pressure Agglomeration Technologies 252 Low-Pressure Agglomeration 253 Medium-Pressure Agglomeration/Pelleting 266 High-pressure Agglomeration 300 Isostatic Pressing 373

9

Agglomeration by HeatlSintering 385

9.1 9.2 9.2.1 9.2.2

Mechanisms of Sintering 385 Sintering Technologies 389 Batch Sintering 390 Continuous Sintering 397

10

Special Technologies Using the Binding Mechanisms of Agglomeration

10.1 10.2 10.2.1 10.2.2 10.3

Coating 415 Separation Technologies 440 Gas/Solid Separation 440 Liquid/Solid Separation 442 Fiber Technologies 447

11

11.2 11.3

Engineering Criteria, Development, and Plant Design 453 Preselection of the Most Suitable Agglomeration Process for a Specific Task 462 Laboratory Equipment, Testing, and Scale-Up 468 Peripheral Equipment 492

12

Outlook

13

Bibliography 525 List of Books or Major Chapters on Agglomeration and Related Subject References 530 Author’s Biography, Patents, and Publications 53 1 Tables of Contents of Related Books by the Author 541

11.1

13.1 13.2 13.3 13.4 14

14.1 14.2 14.3

Pressure Agglomeration

212

229

409

507

Indexes 543 List of Vendors 543 Wordfinder Index 580 Subject Index 591

526

Dedication, Acknowledgements and References When this book was first planned, the idea was to combine in one volume concise descriptions of the agglomeration phenomena, technologies, equipment, and systems as well as a compilation of the applications of agglomeration techniques in industry. The latter was intended to demonstrate the widespread natural, mostly undesired occurrences of the phenomena, including possibilities to avoid them, and discuss the varied old, conventional, and new beneficial uses of the technologies. However, it soon became obvious that, in its entirety, this project became too voluminous and required much more time than anticipated. Therefore, it was decided to split the subject’s presentation into two volumes whereby both books will be “stand alone” publications that are also complementary. The first volume, available here, covers the fundamental phenomena that define agglomeration as well as the industrial technologies and equipment for the size enlargement by agglomeration. Applications are mentioned in a general way throughout the text of this presentation but without going into details. These applications will be presented in a separate book entitled “Agglomeration Technologies - Industrial Applications” that is scheduled for publication in 2003. A preliminary table of contents is given in Section 13.4. Many persons, institutions, and companies have contributed to the two volumes of this book. First and foremost, I wish to thank my wife Hannelore for her support and understanding while, thomghout my professional career, I was compiling various papers and books (see Section 13.3). All are dedicated to her. Without my wife’s active participation in preparing almost all publications, in elaborating the textbook entitled “Size Enlargement by Agglomeration” [B.42],which is a major reference for this publication (see also below), and her, if sometimes reluctant, acceptance that I was not available for long hours on many days during almost four decades, the books, in particular, could not have been completed. It is impossible to acknowledge all the help, extensive and small, that was provided by a large number of individuals and companies. In Section 14.1,a list ofvendors and other organizations is compiled which mentions those who have, in one way or another, contributed as well as some others who may be of interest as potential contacts for the readers of this book. While I have decided not to clutter the text with references, sources have been acknowledged if figures or tables were provided by or are based

Vlll

I

Dedication, Acknowledgements and References

on information from particular companies. The Discialmer at the beginning of this book (see page IV) should be referred to when using such cross-references. Regarding references to literature, Chapter 13 should be consulted. The earlier textbook “Size Enlargement by Agglomeration” [B.42]contains treatments as well as many references relation to the developing science of the unit operation and covers in some detail the sizing of and scale-up methods for agglomeration equipment. Since the emphasis of the new book is on practical considerations and industrial applications, not theory, the earlier book “Size Enlargements by Agglomeration” (Wiley, 1991) should always also be referred to. Information on the availability of reprints is available at the beginning of Section 13.1 and as a footnote later in the same Section. Since Size Enlargement by Agglomeration is one of the unit operations of Mechanical Process Technology (see Chapter 1) and, for the design and construction of agglomeration systems of any kind, many or all of the other unit operations are required, together with the associated transport and storage technologies, often even in multiplicity, and the analytical methods are applied for process evaluation and control, the reader who is interested in the topic of this book should also learn about or have access to information on the other fields of Mechanical Process Technology. This is also emphasized in Chapter 13. At this point I wish to acknowledge two books of general importance to which I have contributed chapters on agglomeration and ofwhich major portions were included in this book. They are: “Handbook of Powder Science and Technology” M. E. Fayed, L. Otten (eds,), 1st ed., Van Nostrand Reinhold Co., New York, NY (1983) and 2nd ed., Chapman & Hall, New York, NY (1997). Source references can be found in [B.21] and [B.56], Section 13.1. Finally, I like to mention with gratitude the following individuals who, as professionals and experts in their own fields, are or have been colleagues and/or partners in several continuing education courses over many years in the USA as well as in Europe. They have agreed that statements during their presentations and the elaborations for their course notes can be used directly, adopted, or modified for this book. They are, in alphabetical order: T. van Doorslaer, W. E. Engelleitner, M. E. Fayed, M. Gursch, D. C. Hicks, S. Jagnow, R. H. Leaver, R. Lobe, K. Masters, S. Mortensen, H. B. Ries, F. V. Shaw, J. Storm, R. Wicke, and R. Zisselmar. For additional references and acknowledgements please refer to Sections 13.1 and 13.2. Naples, November 2001

Wolfgang Pietsch

Agglomeration Processes Wolfgang Pietsch Copyright 0Wiley-VCH Verlag GmbH, Weinheim, 2002

I’

1

Introduction In 1957, under the leadership of Professor Dr.-Ing. Hans Rumpf at the Technical University (TH) of Karlsruhe, Germany, Mechanical Process Technology or Particle Technology IB.111 was first introduced as a field of science in its own right. It comprises the interdisciplinary treatment of all activities for the investigation, processing, and handling of solid particles as well as the interactions of such particulate solids. Four unit operations and associated techniques were defined (Fig. 1.1).Other common English names for this field of science, which was quickly adopted around the world, are Mechanical Process Engineering, Powder Technology, or Powder & Bulk Solids Technology. Size enlargement by agglomeration is the generic term for that unit operation of mechanical process engineering which is characterized by “combination with change in particle size” (Fig. 1.1).The author of this book had the privilege to become one of the first assistants of Professor Rumpf. For several years he was responsible for the research and development of size enlargement by agglomeration at the Institute of

Fig. 1.1

Unit operations and associated techniques

of Mechanical Process Technology

2

I

1 Introduction

Mechanical Process Engineering and earned his PhD with a doctoral thesis on specific aspects of a binding mechanism [1.1]of agglomeration. Webster’s Unabridged Third New International Dictionary [1.2] defines the verb agglomerate as: “to gather into a mass or cluster; to collect or come together in a mass; to collect into a ball, heap, or mass, specifically: clustered or growing together but not coherent”, and the noun agglomerate as: “a cluster of disparate elements; an indiscriminately (= randomly) formed mass”. A technical dictionary [ 1.31 defines agglomeration as: “sticking or balling of (often very fine) powder particles due to short range physical forces. Therefore these forces become active only if the individual particles (forming the agglomerate) are brought closely together by external effects”. These definitions distinguish the term size enlargement by agglomeration from the more general size enlargement such that particle growth occurring, for example, during crystallization or the production of particulate solids by melt solidification are not part of this unit operation of Mechanical Process Technology.


2

A Short History of Agglomeration As a basic physical efect, agglomeration has existed since particulate solids were first formed on Earth. Binding mechanisms between solid particles cause the stability of wet and dry soil and (often under the influence ofheat and pressure) participate in the development of rock formations. Sandstone is the most easily recognized “agglomerate”. Agglomeration as a phenomenon, e.g. the natural caking and build-up of particulate solids, must have been observed and has been used by higher developed organisms and later by humans since prehistoric times. Sea creatures covered themselves with protective coats, birds as well as other animals built nests, and humanoids formed artificial stones, all from various solids, sand, clay and different binders that were often secreted by the creature itself. As a “tool” to improve powder characteristics, agglomeration was used by ancient “doctors” in producing pills from medicinal powders and a binder (e.g.honey) or by food preparers during the making of bread from flour whereby the inherent starchy components act as binder. In spite of this long “history”,agglomeration as a technology is only about 150 years old today (excluding small scale pharmaceutical and some little-known ancient, mostly Chinese applications as well as brick and bread making). Agglomeration as a unit operation, defined within solids processing, started around the middle of the nineteenth century as a method to recover and use coal fines. Agglomeration as a science is very young. It began in the 1950s with the formal definition of the binding mechanisms of agglomeration, interdisciplinary collection of knowledge relating to all aspects of agglomeration, and fundamental research which was no longer application oriented [B.42].At approximately that time, the first recurring series of professional meetings were organized which were exclusively devoted to the science and technology of agglomeration (International Briquetting Association (IBA),- today Institute for Briquetting and Agglomeration (IBA) -, beginning in 1949 with biennial meetings and proceedings: International Symposia on Agglomeration, initiated in 1962 with proceedings, (see also Section 13.1)). Since that time, agglomeration science, technology, and use have experienced rapid growth but still without finding a corresponding awareness at institutions of higher learning and in the technical or process engineering communities. This book is the second by the author on the general subject of size enlargement by agglomeration. While frequently referring to fundamentals and specifics which are

4

I

2 A Short History ofAgglomeration

covered in more detail in the first book [B.42], this new text tries to provide an updated, comprehensive summary of the state-of-the-art of agglomeration, its basics, technologies, and applications, at the beginning of the 21st century.


3

Agglomeration as a Generic, Independent, and Interdisciplinary Field of Science As mentioned in the previous chapter, size enlargement of particulate solids by agglomeration is as old as the existence of solids themselves. Originally, agglomeration happened naturally during the development of soil, stone, and rock formations. Later, unwanted agglomeration occurred during handling and storage of particulate matter particularly when hygroscopic and/or soluble materials (such as salt) “set-up’’ into lumps or large, more or less solidified masses. In the animal world agglomeration was used to develop protective coatings (e.g. many marine worms, Fig. 3.1), to build nests (e.g. swallows, termites, Fig. 3.2), and to provide a nourishing and protective environment for the offsprings (e.g. dung beetles, Fig. 3.3). Humans most probably first used agglomeration during the making of bread by taking flour (= particulate solids including an inherent binder, starch) and liquid additives (= additional binder for plasticity and “green”bonding), mixing and forming the mass, and, finally, “curing”the product, the removal of much of the moisture that was added during the mixing and agglomeration steps, to obtain structure and permanent bonding during baking. The technology of bread making combines all com-

Fig. 3.1 Protective agglomerated coating of a Rhizopod, a creping marine Protozoan (Difflugia urceolata)

6

I

3 Agglomeration as a Generic, Independent, and Interdisciplinary Field of Science

Fig. 3.2a Nest of swallows made by agglomeration from mud, the bird's saliva as a binder, and organic fibers for strengthening

Fig. 3.2b Nest o f termites made from earth as well as the animal's excrements andlor secretions as binder

3 Agglomeration as a Generic, Independent, and lnterdisciplinary Field of5cience 17

Fig. 3.3

Dung beetle, Scarabaeus Sacer, “pelletizing” dung

ponents of a complex agglomeration process including preparation of solid feed particles by milling (= adjustment of particle size and activation of the inherent binder, starch),mixing of particulate solids with additional binder@),forming the mass into a “green”agglomerate, and a “post-treatment” (curing =baking = heating and cooling) to provide strength and texture. Very early it was also found that the porosity of the final product could be modified (= increased) by making use of gases that are produced during fermentation (initiated by sour dough or yeast) and result in bubbles in the green mass. These voids are stabilized by “strengthening”the bread during post-treatment (baking). For the construction of permanent shelter, humans may have observed the activities of animals that formed nests and protective “walls” from wet clay which hardened during drying (Fig. 3.2). By copying this behavior, wet clay, which was soon reinforced and made more water resistant by mixingin straw or other fibrous material, was filled into a framework of wood branches and let harden during natural drying. To make building activities independent of the location of clay “mines”, during prehistoric times bricks were already produced from clay and sand and, after hardening, transported to building sites. Since fire was known for providing heat, the accidental “firing” of a piece of clay most probably resulted in the adaptation of an improved posttreatment that yielded waterproof bricks for areas where rock was not easily available, thus allowing the development of villages and, during the 4th millenium B.C. in Mesopotamia, cities with large brick structures. By experience, humans learned that certain natural materials helped cure specific illnesses. Minerals as well as dried animal and plant matter were ground to powder and “formulated”to yield medicines. Since powders cannot be easily consumed orally, natural binders, such as honey which, incidentally, also masked the unpleasant taste of

8

I some of the medicinal components, were mixed with the powder and the resulting 3 Agglomeration as a Generic, Independent, and Interdisciplinary Field of Science

plastic mass was rolled by hand into little balls (= pills). The sticky binder(s) caused pills to adhere to each other; therefore, it was soon found that coating the pills by rolling them in flour or pollen solved this problem (see also Section 10.1). These three, well known ancient agglomeration techniques were used with little change through the ages of human history. Many other, lesser known and somewhat more recent processes could be added. However, it is not the objective here to produce a history book. Rather, these examples relating to three major modern “industries”, food, building materials, and health products, were selected to show that humans always lived with and used agglomeration. As a result, agglomeration technologies as all the other unit operations and associated techniques of Mechanical Process Technology (see Fig. 1.1)were considered to be “normal activities” which, with the beginning ofindustrialization in the 18th and 19th centuries A.D., were merely mechanized by simulating what was done manually before. During these early modernization efforts it was not considered necessary to question the fundamentals of the processes and “improvements” were based on empirical developments. Until very recently, agglomeration technologies had been developed independently in the particular industries in which they were applied. Because the process requirements are fundamentally different in such unlike industries handling, for example, coal and ores on one hand or food and pharmaceuticals on the other, no interdisciplinary contact and exchange of information took place. In fact, although agglomeration techniques developed along similar lines, application related “theories” were defined which were derived from investigations of specific requirements and their solutions together with a terminology that was often incomprehensible and, therefore, not useable by the “agglomeration expert” of another industry (see Chapter 4). Agglomeration as a science began when an effort was made to interdisciplinarily combine the extensive knowledge that had been accumulated during sometimes hundreds of years in specific fields of human activities. This approach showed that (in alphabetical order, not indicating importance): Baking: A thermal post-treatment process, does not only induce the development of final bonding, structure, and consistency in bread but produces similar characteristics during the heat curing of any “green” agglomerate. Briquetting: Is not predominantly a technique for the enlargement of coal fines for beneficial use but equipment which was specifically developed for that application is also suitable for such diverse uses as, for example, the briquetting of salt for the regeneration of water softeners, the briquetting of flaked DMT to decrease the bulk volume and improve handling and shipping, the briquetting of frozen vegetable pulps to be used as rations for field kitchens, the hot briquetting of sponge iron to reduce this commodity’s reactivity and allow open handling and storage, or the production of fertilizer spikes and the manufacturing of artificial fireplace logs. Coating: Is not only suitable for the modification of surface characteristics or the control of dispersion and dissolution of medicinal specialties but also to achieve similar properties in agrochemicals as well as human and animal foods, among others.

3 Agglomeration as a Generic, Independent, and Interdisciplinaty Field of Science 19

Compacting: Is not only applicable for the production of “green” bricks or other ceramic bodies prior to firing but finds many uses in powder metallurgy or for the production of battery cathodes, etc. Granulating: Is not primarily a method to improve flowability of powders and formulations in the pharmaceutical industry but also in the fertilizer and bulk chemical industries as well as for carbon black, silica fume, and many other materials. Instantizing: As an example of a relatively new agglomertion process, is not limited to applications in the food industry for easily dissolvable drink and soup mixtures but is equally important for pigments, insectizides, fungizides, and many more. Pelleting: Originally developed for the shaping of animal feed formulations by extrusion, is also applicable for the production of catalyst carriers and other materials requiring uniform size and shape together with relatively high porosity. Sinteuing: When going back to the fundamentals of this process, was found to be not only a high temperature process for the agglomeration of ores but, at much lower temperatures, also for plastics and other man made powders with low melting points or softening ranges, and, quite obviously, for powder metallurgy, mechanical alloying, or the like at many different temperature levels including extremely high ones for refractory metals. The above is only a small selection of the many diverse applications of particular agglomeration methods which, in all the different environments, follow the same fundamentals, apply the same rules, and use essentially the same equipment and systems if looked at from an interdisciplinary point of view. Although these facts become more and more known, there is still the understandable preconceived notion of, for example, somebody working in an ultraclean environment, such as the pharmaceutical, food, or electronic industries, that developments, expertise, and know-how gained in the “dirty”plants of, for example, minerals or metals production and processing, can not be considered as valid information that may be applied for the solution of a “clean”problem - and vice versa. In the case of “dirty”industries, a typical concern is that the often more deeply and completely investigated technologies originating in “clean” industries can not be applied because the production capacities are too small, the process may be batch, the equipment too complex, the execution and the materials of construction too expensive, etc., etc. However, as will be shown among other topics in this book, methods for the selection of the most suitable agglomeration process for a specific application (see Section 11.1)are the same for all projects. While some requirements, for example in regard to equipment or system capacity, or on the shape, size, and special properties of the products, may result in the definition of “cleaner” or “more heavy-duty, rugged” processes already in the preselection phase, the normal approach is to determine the preferred method and/or technique by considering the fundamentals as well as an interdisciplinary pool of expertise and know-how first. Conditions of the particular application such as, for example, “hot and dusty large volume processing”, or the opposite, “clean, small capacity operation with cGMP (= current Good Manufacturing Practice) and CIP (= Cleaning In Place) capabilities” are special design criteria that can be added to most of the systems later during the engineering phase.

10

I

3 Agglomeration as a Generic, Independent, and interdisciplinary Field of Science

Nevertheless, for manufacturing reasons and sometimes also because of special requirements on the company’s test facilities (see Section 114, some vendors specialize in equipment for one or the other industry. This is a decision of convenience by the individual supplier and does not indicate the existence of a fundamentally different technology. In fact, techniques or apparatus that were developed for a specific industry can be adopted for use in areas with different environment and requirements while still maintaining the fundamental underlying principle as well as the general machinery and process. Examples are flaking (see Section 8.4.3), instantizing (see Section 5.4), spheronizing (see Section 8.3), and spray dryer agglomerators (see Sections 7.4.3 and 7.4.4).


4

Glossary o f Agglomeration Terms Newly developing fields of science are organized according to universally recognized classifications using well-defined terms to describe the fundamentals, correlations, equipment, procedures, and processes. This is not the case for those technologies that were known for centuries and have been developed empirically and independently for different applications (see also Chapters 2 and 3). In such cases the same process, procedure, activity or piece of equipment may have different names in different industries or the same term may have different meanings in different fields of application. The earlier book “Size Enlargement by Agglomeration” [ B.421 contained already a glossary of agglomeration terms. In the following this glossary is repeated and updated. Although the author and many others that are active in the promotion of “agglomeration” are trying to use scientific and technical terms consistently in an interdisciplinary manner (terms shown bold), it is still helpful to also explain some of the more common names and expressions including a few historical ones. In the following, crossreferences are indicated by italic letters. The same and many more “agglomeration terms”, the latter mostly descriptive and/or trade names, are mentioned and used in the text of the book. Sections 14.2 and 14.3 help locate these words and expressions. Abrasion [n.]

Abrasion resistance Accretion [n.]

Accumulate [vb.] Accumulation [n.] Adhesion [n.]

Removal of solid matter from the surface or edges of an agglomerate. The matter removed is much smaller than the agglomerate itself. (See also attrition, erosion.) Measure for the ability of a body, for example an agglomerate, to withstand abrasion. The process of growth or enlargement by a gradual buildup, such as: increase by external addition or accumulation, for example by adhesion of external parts or particles. (See also agglomeration, aggregation, build-up.) To heap up into a mass; pile up. The action or process of accumulating; an accumulated mass, quantity, or number. A sticking together of solids. The molecular attraction exerted between the surfaces of solids. Distinguished from cohesion.

12

I Agglomerate [vb.]

4 Glossary of Agglomeration Terms

Agglomerate [n.]

Agglomeration [n.] Agglomerator [n.] Aggregate b.1

Aggregate [vb.] Aggregation [n.]

Agitation [n.] Agitator [n.] Ammoniation [n.] Angle of repose Angle o f compaction Anticaking agent

Apparent density Atomizer [n.] Atomizing [vb.] Attrition [n.] Auger [n.] Axial extruder

To gather (particulate solids) into a ball, mass, or cluster. (See also aggregate.) An assemblage of particles which is either loosely or rigidly joint together. Particles adhering to each other. (See also conglomerate.) The action or process o f gathering particulate solids into a conglomerate. Specific equipment in which agglomeration is accomplished. Any o f several hard, inert materials (as sand, gravel, rock, slag) used for mixing with a binding material to form concrete, mortar, plaster or, for example, road surfacing products. Also: A mass or body o f units or parts somewhat associated with one another. To collect or gather into a mass. (See also agglomerate.) A group, body, or mass composed o f many distinct parts or individuals; the collection of units or parts into a mass or whole; the condition o f so collected. (See also agglomerate, aggregate, cluster, agglomeration, accretion, build-up.) Changes in characteristics ofparticulate solids or agglomerates that occur naturally with time. (See also post-treatment.) A state of movement of particulate solids and/or fluids induced by external effects or forces. See mixing tool, intensijer bar. The formation o f fertilizer granulates using ammonia to obtain chemical modification and bonding. The basal angle of a pile o f powder that has been freely poured onto a horizontal surface. See nip angle. Liquid or solid matter applied to the surface of, for example, agglomerates that prohibits sticking or growing together. (See also caking.) The weight of the unit volume of a porous mass, for example, an agglomerate. See nozzle. Finely dispersing liquids. The unwanted break-down of agglomerates. (See also abrasion, erosion.) See screw. Low, medium, or high pressure extruder with a flat die plate at the end of a barrel; the material is extruded in the same direction as it is transported by the screw(s).

4 Clossarj of Agglomeration Terms

Backmixing [n.]

Bag set

Ball [n.] Ballability [n.]

Balling [n.]

Barrel [n.] Basket extruder Beading [n.]

Bin [n.] Binder [n.] Binding mechanism

Biomass [n.] Blade [n.] Blunger [n.] Boiling Bed Bonding [n.] Bowl [n.]

Bridging [n.]

Briquette [n.] Briquetter [n.]

During the flow of particulate solids, reverse movement of some particles due to their stochastic motion caused by turbulence or special equipment design. Typical in the fertilizer industry; unwanted agglomeration of particulate solids in a closed bag during storage. Mostly caused by recrystallization of dissolved materials. Synonymous with spherical agglomerate. (See also pellet.) Typical in the iron ore industry; the capability of particulate solids to form more or less spherical agglomerates during growth agglomeration. Originally in the iron ore industry; any method producing spherical agglomerates by tumble or growth agglomeration. (See also pelletizing.) Cylindrical (or sometimes tapered) housing for screws, e.g. offeeders or extruders. Low pressure extruder in which the die plate resembles a basket, using rotating or oszillating extrusion blades. Formation of bead-like particles; typical in solidification of melt droplets. (See also prilling, pastillation, melt solidijcation.) A container, box, frame, crib, or enclosed volume used for storage. (See also hopper, silo.) An inherent component of or an additive to particulate matter providing bonding between the disparate particles. Physical and chemical effects that cause adhesion and bonding between solid surfaces. See Section 5.1, Tab. 5.1 and Fig. 5.8. Organic plant and animal residuals. Often organic waste material that is especially used as a source for fuel. See extrusion blade. Typical in the ceramic and fertilizer industries; double shafted pug mill. SeeJuid bed. The process of binding particles together by the action of binding mechanisms. A vertical or inclined, cylindrical, conical or convex vessel enclosing and defining the operating volume of some coaters, mixers, spheronizers, etc. Unwanted arching of solid matter in a converging discharge chute or cone. Prohibits discharge of particulate solids from containers or chutes. Also briquet; agglomerate produced and shaped by highpressure agglomeration. (See also compact, tablette.) Also briquetting machine; equipment that produces briquettes.

14

I


Briquetting [n.] Brittleness [n.] Build-up [n.]

Bulk density

Capillary [adj.] Capping [n.]

Cake [n.] Caking [n.] Cement [n.]

Cement [vb.] Cementitious [adj.] Channel [n.]

Chelate [adj.]

Chelate [n.] Chopper [n.] Clam shelling

Closed pore Cluster [n.] Clustering [n.] Coalesce [vb.]

The process of forming briquettes or compacts. The tendency of particles or agglomerates to break down in size easily. (See also friability.) The unwanted coating of surfaces with particles which adhere naturally due to their fineness and/or inherent binding mechanisms. The weight of the unit volume of a particulate mass under non-specific condition, e.g. in storage or in a shipping container. (See also density.) Describing full liquid saturation. Separation of a thin layer from the face(s) of compacts during decompression. Defect in tablettes caused by the recovery of elastic deformation and/or expansion of compressed air. See sheet; typical in fertilizer applications. Unwanted agglomeration during storage mostly by recrystallization of dissolved materials. (See also bag set.) A powder of alumina, silica, lime, iron oxide, and magnesium oxide burned together in a kiln, finely pulverized, and used as an ingredient of mortar and concrete. Also any mixture used for a similar purpose. (See also pozzolan.) To unite or make firm by or as if by cement. Having the properties of cement. Open ended compacting tool set for high pressure extrusion in a ram press; also any elongated opening through which material is extruded. (See also pressway.) Relating to, producing, or characterized by a cyclic structure usually containing five or six atoms in a ring in which a central metallic ion is held in a coordination complex by one or more groups each of which can attach itself to the central ion by at least two bonds. To combine with a metal to form a chelate ring or rings. See knive head. Opening of the leading or trailing edge of briquettes discharging from roller presses; one-sided splitting along the web. Also duck billing, oyster mouthing. A pore not communicating with or connected to the surface of a porous body. A number of similar individual entities that occur together. (See also accretion, agglomerate, aggregation.) The growing together of primary agglomerates to form larger entities. (See also satellites formation.) To unite by growth.

4 Glossary of Agglomeration Terms 115

Coalescence [n.] Cohesion [n.]

Coating [n.]

Coating pan

Cold bonding

Compact [n.]

Compact disperse Compactibility [n.] Compacting [n] Compacting tool set Compaction/granulation

Composite [adj.] Compressibility [n.]

Compression ratio

Conditioning [n.]

Cone agglomerator

A growing together or union in one body, form, or group. (See also growth agglomeration.) Molecular attraction by which the particles of a body (e.g. agglomerate) are united throughout the mass whether like or unlike. Distinguished from adhesion. Applying a layer of material, a film, or a finish to a substrate; in agglomeration, application of a layer of solids to a particulate unit. Specially shaped p a n in which a material layer is applied on agglomerates (such as tablettes) usually in the presence of liquid, heat, or both. Typical in the pharmaceutical and food industries. A binding process that occurs at ambient or low temperatures and uses the cementitous or pozzolanic reactions of many hydroxides; often assisted by pressure. An object of specific size and shape produced by the compression of particulate matter. Synonymous with briquette. A state of particulate solids in which individual particles are closely packed. Distinguished from discrete dispers. See compressibility. Also compaction. The method of producing sheet. The part or parts making up the confining form in which a powder is pressed. Synonymous with die. The normally dry methods of obtaining granular products by crushing and screening compacts and/or sheet into granulate. Consisting of two or more separate materials whereby each retains its own identity. The capacity of a particulate matter to be compacted. Compressibility may be expressed as the pressure or force to reach a required density or, alternately, the density at a given pressure or force. Synonymous with compactibility. The ratio of the volume of loose particulate matter in a die to the volume of the compact made from it. Synonymous with fill ratio. In low and medium pressure extruders, the total thickness of material that is under compression in a die (including any inlet chamfer) divided by the nominal hole diameter. Development of special characteristics of particulate solids by, for example, treatment with steam, kneading, heating, etc., or surface treatment by, for example, anticaking agents. Pan with relatively high conical rim.

16

I Conglomerate [n.]

4 Glossary ofAgglomeration Terms

Contact point Coordination number

Core rod Couffinhal press

CUP Ln.1 Curing [n.] Cut size Decrepitation [n.] Densification [n.] Density [n.] Deposit [n.] Die [n.]

Die plate Disc [n.] Discrete dispers

Dispers [adj.] Dispersibility [n.]

Distribution plate

Doctor blades Dome extruder Double action pressing

An adhering mass of particles made up of parts from different sources or of various kinds. (See also agglomerate.) Area at which two particles touch each other. Sum of all near and contact points of a particle with surrounding particles in a structure made up of particulate solids, for example an agglomerate. Member of the compacting tool set that forms a through hole in the compact. (See also mandrel.) Punch-and-die press with multiple die sets on an indexed table for making large (e.g. coal) briquettes. (No longer used.) See pocket. h d u r a t i o n of green agglomerates by any method. (See also post-treatment.) The actual value at which separation of a particle size distribution into “coarse” and “fines” has taken place. Breakdown in the size of particles or agglomerates due to internal forces, generally induced by heat. The act or process of making dense. Mass per unit volume of matter at specific conditions. For example: apparent, bulk, or true densities. A (natural) accumulation of particles. Member of the compacting tool set that forms the periphery of the part being produced. Also open ended channels for extrusion. Plates, rings, or other machine parts with perforations for extrusion. (See also die.) See pan. A state ofparticulate solids in which individual elements can be clearly distinguished. Distinguished from compact dispers. See particulate. Measure for the ease with which, under specific conditions (e.g. in liquids), an agglomerate breaks down into primary particles. Perforated plate at the bottom of a j u i d bed through which fluidizing gas enters from the plenum. (See also gil plate.) See scraper. Axial, low pressure extruder, most often with two screws, in which the die plates resemble domes. A method by which particulate solids are pressed between opposing punches which are both moving relative to the die.

4 Glossary of Agglomeration Tirms

Downdraft [n.] Drum agglomerator Dry granulation Duck billing Dwell time

Encapsulation [n.] Erosion [n.]

Equivalent diameter

Expansion [n.]

Exter press Extrudate [n.] Extruder [n.] Extrusion [n.]

Extrusion blade

Feeder [n.] Feed screw Fill ratio

Flake [n.]

Flake breaker Flashing [n.] Flight [n.]

Downward flow of gas, for example through a particle bed. Slowly rotating, slightly inclined drum for growth a g glomeration. See compaction/granulation. See clam shelling, oyster mouthing. In compacting, time during which certain process conditions, for example pressure, persist or are held constant. Typically used as microencapsulation. The gradual wearing away of an agglomerate by the progressive removal of small pieces of material. (See also abrasion.) Diameter of immaginary monosized spherical particles which feature the same property as the particulate mass to be characterized. For example: surface equivalent diameter. Increase in volume of, for example, an agglomerate after production or during post-treatment. Converse of shrinkage. See ram extruder. Product of extrusion. (See also pellet.) Machinery for the production of extrudates. (See also screw and ram extruder.) The formation of (often cylindrical) agglomerates by forcing a “plastic” mass through open ended channels or holes in (perforated) dies. In low pressure extruders, the flat, curved, or engineered blade that pushes material through the openings of a die plate; it is the part closest to the die plate. Device to deliver feed material to a processing unit. (see also force feeder.) Element(s) providing forces onto particulate solids in a feeder. (See also screw.) Typically used in tabletting or other confined volume compression equipment. Synonymous with compression ratio. See sheet. Also: 1. Grains or other malleable particles flattened between smooth rollers. 2. Material solidified from a melt on a rotating, cooled drum (flaker) and removed by scrapers. A primary crusher (often two rollers with teeth) used to reduce the size of sheet. See web. A continuous or semi-continuous spiral flat plate that is attached to the shaft of a screw.

18

I

4 Glossmy ofAgglomeration Terms

Floc [n.] Flocculant [n.] Flocculate [vb.] Flocculation [n.] Flocculent [adj.] Fluid bed

Fluid bed agglomeration Force feeder Fraction [n.]

Fragmentation [n.]

Friability [n.] Friction plate

Funicular [adj.] Gap w.1

Gear pelleter

Gil plate Globulation [n.] Granular [adj.] Granulate [n.]

Aflocculent mass formed by the aggregation of a number of fine suspended particles. A flocculating agent. To aggregate or coalesce into small lumps or loose clusters or into aflocculent mass or deposit. The act or process offlocculating. Containing, consisting of, or occurring in the form of loosely aggregated particles or soft flocs. Also fluidized bed. A bed of particles in which the particulate solids are kept in suspension by forces caused by an upward flowing fluid. Growth agglomeration in a fluid bed. A feeder that provides forces onto particulate matter by, for example, the action offeed screws. That portion of a sample of particulate solids which is between two particle sizes (see cut) or in a stated range (e.g. fine, coarse, etc.). The process whereby aparticle (or agglomerate) splits into usually a large number of smaller parts with a range of sizes. The tendency of particles to break down in size during storage and handling. (See also brittleness.) In spheronizers, a circular flat disc with a rough surface or uniformly spaced grooves which rotates inside a cylindrical bowl. Describing the transitional liquid saturation. In pressure agglomeration,the distance between the surfaces of compacting tool sets; specifically: in extrusion, the distance between the pressure generating device and the die plate, in roller presses, the closest distance between the rollers. Double-rollpellet mill in which the rollers are in the shape of coarse, intermeshing gears with bores at the root sections between the gear teeth. (Also gear pelletizer.) Distribution plate in which the perforations are manufactured such that they produce a directional flow of gas. See melt solidijcation. Present as particles in “grain” shape and size. Coarsely particulate. Also Granule. From Latin granula = grain, particle. Any kind of relatively coarse particulate matter. In size enlargement, synonymous with agglomeration to a size range of between approx. 0.1 and 10 mm. In size reduction, synonymous with crushing into approx. the same size range. Granulate is normally considered dustfree, free flowing, and non-segregating.


Granulate [vb.]

Granulation [n.] Green [adj.] Grid [n.] Growth [n.] Growth agglomeration Heat bonding Heel [n.]

Hopper [n.] Hot melt agglomeration Hot pressing Immiscible binder agglomeration Induration [n.] Inkbottle pore

Instant [adj.] Instantizing [n.]

Intensifier bar

Interconnected porosity

Producing a granular solid matter; possible by size enlargement (agglomeration, melt solidijhtion [pastillation, prilling], and crystallization) or by size reduction (crushing). (See also compaction/granulation.) A general term for the production of solids in granular form by either size reduction or size enlargement. As in “green agglomerate”, “green pellet”, etc., means fresh, moist, uncured, etc. In spheronizers, the design (size and shape) of the grooves on the friction plate surface. An increase in dimension by for example agglomeration or crystallization. (See also coalescence.) See coalescence, tumble agglomeration. See sintering. In batch processing, for example agglomeration, a percentage of the previous batch retained in or returned to the processing vessel. The funnel or chute that stores material and/or directs it into equipment. (See also bin, silo.) Granulation of a hot melt of e.g. urea or ammonium nitrate in a pan. The simultaneous heating and molding of a compact or briquetting of hot material. Selective agglomeration of particles suspended in a liquid by adding an immiscible binder during agitation. (See also oil agglomeration.) Strengthening of green agglomerates, mostly by heat. Non-cylindrical pore with varying diameter; particularly a pore with narrow entrance followed by a large, internal volume. Quickly soluble. Characteristic as, for example, in “instant coffee”. Producing agglomerated products with instant characteristics, i.e. material exhibiting, as compared with the untreated powder, particularly high solubility, even in cold liquids. In high shear mixers and agglomerators, an independently driven bar, rotating with high speed, usually carrying mixing tools and, sometimes means for atomizing liquid binder, that extends into the particulate mass to be mixed and causes an additional turbulent motion of the particles. (See also knive heads.) A network of contiguous pores in and extending to the surface of a porous body, e.g. agglomerate.

20

I Interface [n.]


A plane or other surface forming a common boundary of two bodies or spaces. The densijcation of a particulate mass by subjecting it to Isostatic pressing nominally equal pressure from every direction. In high shear mixers and agglomerators, independently h i v e head driven high speed rotating tools which extend into the particulate mass and cause additional turbulent motion of the particles as well as desagglomeration in mixing and controlled destruction of agglomerates in agglomeration. (See also intensijer bar.) The area surrounding briquette pockets on the roller surLand area face of briquetters. (See also flashing, web.) During briquetting in roller presses the forward edge of a Leading edge discharging briquette. A member of the compacting tool set that determines the Lower punch powder fill level and forms the bottom of the part in a punch-and-die press. Extruder in which the die plates consist of screens or thin, Low pressure extruder perforated sheets and exert small frictional resistance during extrusion. An agent mixed with or incorporated into particulate Lubricant [n.] matter or applied to the tooling to facilitate pressing and ejection of a compact, tablette, or extrudate. See second meaning of aggregate. Lump [n.] Also mandril. A metal bar that serves as a core around Mandrel [n.] which material may be bent, cast, forged, molded, or otherwise shaped. (See also core rod.) Sometimes used to describe a particle which has been Marum [n.] spheronized. See spheronizer. Original (Japanese)name. Marumerizer [n.] A technology of powder metallurgy by which powders of Mechanical alloying metals, that cannot be combined in molten stage, are mixed and compacted to form the alloy. Medium pressure extruder See pellet mill. A method by which molten substances are converted Melt solidification into particulate solids by cooling droplets of the melt. (See also beading, pastillation, prilling.) A method by which small portions of liquids, particulate Microencapsulation [n.] solids, or gases are enclosed by a shell (membrane, capsule) to form a dry, free flowing product often with spherical particle shape. The capsule shell may provide specific product characteristics (e.g. dispersibility, solubility). Trle formation of small agglomerates, usually not larger Micropelletizing in.] than 3 mm, by growth agglomeration. (See also pelletizing.)

4 C/ossay of Agglomeration Terms

Mixer agglomeration Mixing tool

Mix-Muller [n.] Mold [n.] Muller [n.]

Multiple pressing Near point

Nip [n.]

Nip angle

Nodulizing [n.]

Nozzle [n.] Nucleus [n.], Nuclei [pl.]

Oil agglomeration

One pot processor

Open pore Orifice [n.]

&tation and growth agglomeration in powder mixers. Any of a large number of differently shaped tools that are attached to a rotating shaft and cause irregular movement in a particle bed. See Muller. See die. Also Mix-Muller. Originally, a device that used a heavy stone roller to grind and/or mix particulate solids. Today, a blender with one, two or four large metal rollers that mix and knead (densify) material. Often used prior to pressure agglomeration. (See also pan grinder.) A method of pressing whereby two or more compacts are produced simultaneously in separate die cavities. Area at which two particles approach each other closely enough for a binding mechanism to become effective. (See also coordination number.) In roller presses and pellet mills, converging space (volume) between two counter-rotating rollers and, respectively, the pressure generating device and the extrusion surface. (See also nip angle.) In rollerpresses, radial angle defining the line on the roller surface at which the speed of the particulate mass is identical with that of the roller; in extruders, the angle between the extrusion surface (e.g. dieplate) and the pressure generating device (e.g. extrusion blade, screw, roller). Formation of nearly spherical lumps (agglomerates)from a wet mixture of particulate solids by either drying or chemical reaction during tumbling; typically accomplished in dryers or rotary kilns. Also atomizer. Means for atomizing liquids. Primary agglomerate(s) consisting of only a few particles on which further growth occurs. (See also seed.) Also spherical agglomeration. Selective agglomeration of suspended particles in water by adding a bonding oil during agitation; typical in coal preparation. (See also immiscible binder agglomeration.) A batch processing vessel in which several process steps, for example mixing, agglomeration, post-treatment, and finishing, are carried out without opening the vessel during the entire processing sequence. A pore communicating with or connected to the surface of a porous body. (See also inkbottle pore.) The mouth or opening of something, for example an extrusion channel, that forms material into defined shape.

I

21

22

I Oyster mouthing

4 Clossaty of Agglomeration Terms

Pan [n.] Pan grinder Particle [n.] Particle size Particulate [adj.] Pastillation [n.]

Pastille [n.] Pellet [n.]

Pellet mill Pelleting [n.]

Pelletizing [n.]

Pelletization [n.]

Pelletizer [n.] Pendular [adj.] Penetrating pore Pin mixer Piston press Plenum [n.]

Plow [n.] Plug flow

See clam shelling, duck billing. Also disc. An inclined rotating circular plate with low cylindrical rim for growth agglomeration. See Muller. A piece of solid material that is an entity in itself. The controlling dimension of an individual particle as determined by analysis. Of or relating to separate particles. A method of melt solidijcation by which droplets of a molten material are solidified on a cooled, moving stainless steel belt. Product of pastillation. Name for many different types of agglomerates. Most commonly used in the iron ore industry for nearly spherical agglomerates formed by growth agglomeration in pans, cones, or drums and in the animal feed industry for extrudates produced by pelleting. Often synonymous with agglomerate. Equipment for extrusion through perforated dies. Agglomeration by extrusion of plastic material or of particulate matter containing binders through bores of dies in “pelleting machines” or pellet mills. Originally, production of pellets by growth agglomeration. Today typically agglomeration by balling. Often also used as synonym for agglomeration. Typical in the (iron) ore industry; any agglomeration method involving growth agglomeration with subsequent heat induration. (See also sintering.) Usually rotating pan, drum, cone, or the like for growth agglomeration. (See also “gear pelletizer”.) Describing the liquid bridge model. A pore that connects opposite sides of a porous body, for example, an agglomerate. (See also through pore.) A stationary, cylindrical mixer using a single shaft agitator with pins. See punch-and-die press. Specially designed chamber at the bottom of afluid bed from which fluidizing gas enters the apparatus through the openings of a distribution plate. Plow shaped mixing tool. Forward movement of particulate solids to the discharge end of tumbling drums orfluid beds, caused by a continuous particle feed and optionally assisted by downsloping the drum or the application of gil plates influid beds.


Pocket [n] Pore [n.] Pore volume Porosity [n.] Porous [adj.] Post-treatment

Pozzolan [n.]

Pozzolanic [adj.] Powder [n.] Powder metallurgy

Powder rolling Pressway [n.]

Pressure agglomeration

Prill [n.] Prilling [n.] Pug mill Punch [n.] Punch-and-die press

Radial extruder

Indentation on the surface of rollers, normally forming one half of a briquette shape. (See also cup.) An inherent or induced cavity in a particle or void space between particles within an object e.g. agglomerate. Void space (volume) in porous objects. (See also porosity.) The amount of pores (voids) in an object expressed as percentage of the object’s total volume. Possessing or “full of” pores. Any treatment of green agglomerates to modify moisture content, strength, structure, etc., by, for example, aging, drying, heating, sintering, etc. (See also curing.) Also Pozzolana. Finely divided siliceous or siliceous and aluminous material that reacts chemically with slaked lime at ordinary temperature and in the presence of moisture to form a strong, slow hardening cement. Having the properties of pozzolan. Particles of dry matter typically with a maximum dimension of less than approx. 1,000 pm. The art of producing metal powders and of their utilization for the production of massive materials and shaped objects as well as for mechanical alloying. See roll compacting. Also used in powder metallurgy for direct rolling of sheet from metal powders. In extruders, the (length of the) channel in which frictional resistance causes the extrusion pressure; the total distance material is compressed inside a die. Also press agglomeration. Agglomeration technique during which agglomerates are formed by pressure. Distinguished from tumble agglomeration. Product of prilling. In the fertilizer industry often (incorrectly!!) synonymous with agglomerate. The formation of spherical particles by solidification of melt droplets. (See also melt solidijcation, shot forming.) A paddle type mixer usually with open top, single or double shafts, and trough shaped chamber. Part of a compacting tool set which transmits pressure to the particulate matter in the die cavity. A mechanically or hydraulically actuated press in which a reciprocating piston compacts particulate matter in a die. Low pressure extruder in which part ofthe barrel consists of a screen or perforated thin sheet through which moist, plastic material is passed by extrusion blades to form extrudates; the material is extruded radially to the direction in which it is transported.

I

23

24

I

4 Glossary ofAgglomeration Terms

Ram [n.] Ram extruder

Ram press Rim [n.] Ring die

Ring die extruder Ring roll press Roller [n.]

Roll(er) compacting

Roll(er) press Roll(er) pressing

Rope [n.] Rotary press Satellites formation

Saturation [n.]

Schugi flexowall

Scraper [n.]

Synonymous with punch. Press in which a fly-wheel powered reciprocating ram densifies and extrudes particulate solids through a long extrusion channel. Particularly suitable for elastic materials (such as peat, lignite, biomass, etc.). Also ram press, exter press. See ram extruder. Cylindrical or conical wall surrounding the circular plate of pan, disc, or cone agglomerators. A usually narrow hollow cylinder that is equipped with perforations for extrusion. See pellet mill. Special roller press with one press roller within a large ring-shaped die. (No longer used.) Also Roll. Cylindrical rotating body that is: 1.Paired with an identical, counter-rotating one in a suitable frame. This arrangement is used for briquetting, compacting, pelleting, densijication, Jaking, and granulating particulate solids. 2. Rolling close to a die plate and forces material to flow through openings, for example, in flat die pellet mills. 3. Mixing and kneading material in a cylindrical or “figure eight”-shapedbowl. (See also Muller.) Also powder rolling. The (progressive) compacting of (metal) powders in roller presses (often called “rolling mills”). (See also roll pressing.) Equipment for pressure agglomeration between two rollers. Densification between two counter-rotating rollers. (See also compacting.) In spheronization, referring to the rotating particulate material. Tabletting machine in which compacting tool sets are arranged on a rotating table (= turret). In agglomeration, the attachment of smaller solid entities, often agglomerates, to other agglomerates by binding mechanisms. (See also clustering.) Relative amount of pores in an agglomerate filled with a liquid or solid substance, as in “liquid saturation”, “binder saturation”. High speed, high shear mixer and/or agglomerator with vertical axis, adjustable mixing tools, flexible shell, flexing roller cage, and short residence time. A tool for removing build-up in agglomeration equipment. Also doctor blades.


Screen [n.]

Screw [n.]

Screw extruder Seed [n.] Segregation [n.]

Selective agglomeration

Sheet [n.]

Shot forming Shrinkage [n.] Silo [n.]

Single action pressing Sinter [n.] Sintering [n.]

Slug [n.]

Slugging [adj.]

Slugging press

A (usually mounted) perforated thin plate or cylinder or a meshed wire or cloth fabric used to: 1. Separate coarser from finer particles or 2. Form extrudates. A mechanical device spiral in form or appearance; a conveyor working on the principle of a screw; a conveying tool in afeeder, mixer, or extruder. Also auger, worm. Extruder in which screw(s) produce the extrusion pressure. See nucleus. The desirable or undesirable separation (according to mass, shape, size, etc.) of one or more components of a particulate mass. Agglomerationof only one component of a powder mixture controlled by, for example, binding mechanism,binder, particle size. (See also immiscible binder agglomeration.) A more or less continuous band of compacted material produced in roller presses featuring smooth or shallowly profiled rollers and a gap between those rollers. Also, anything that is thin in comparison to its length and/or breadth. The solidification of a melt into little spheres in a tall form tower. (See also prilling.) A decrease in dimension. In agglomeration, usually of a compact during sintering. Converse of expansion. A trench, pit, or especially a tall cylinder (as of wood, metal, or concrete) often sealed and used for storing particulate solids. (See also hopper, bin.) A method by which a particulate mass is pressed in a stationary die between one moving and one fixed punch. Agglomerated product of sintering. Technique involving induration of green agglomerates by heat. Generally, bonding at a temperature below the melting or softening points of the main constituent of a mixture by the application of heat. (See also heat bonding.) Large, flat faced compressed disk prepared for the purpose of stabilizing the mixture of ingredients in the pharmaceutical industry. 1. Producing slugs in a sluggingpress. 2. Influid bed technology, the slow, upward movement of large, somewhat cohesive masses of particulate solids. Punch-and-die press for the production of large tablettes or slugs which are crushed to obtain granulate. Mostly in the pharmaceutical industry. (See also tabletting machine.)

I

25

26

I


Spherical agglomeration Spheronizing [n.]

See oil agglomeration. Rounding of soft, plastic (usually green) agglomerates (usually extrudates) in a spheronizer. Vertical drum with rotating bottom for spheronizing. Spheronizer [n.] (See also Marumerizer.) Characteristic parameter for roller presses; defined as Specific force pressing forcelactive roller width. The formation of granular solids or small spherical agSpray drying glomerates by dispersing a liquid or slurry in droplet form at the top of a tower and evaporating the liquid in the presence of drying gases. The formation of small, spheroidal agglomerates in a Spray granulation &id, circulating, or spouted bed by spraying a solution, slurry or melt onto the particles; often combined with drying. Stabilize [vb.] Avoid segregation by agglomerating a powder mixture. An elongated body; synonymous with uncut extrudate. Strand [n.] Strip [n.] See sheet. Surface equivalent diameter The diameter of immaginary monosized spherical particles, calculated from the mass related specific surface area, in m2/g,of a particle size distribution, that produce the same specific surface area as the powder. The state ofparticulate solids which are uniformly mixed Suspension [n.] with but undissolved in a fluid. Also Tablet. A compressed agglomerate made of particuTablette [n.] late solids, specifically, in pharmacy, a small compact of a medicated particulate formulation usually in the shape of a disc or a flat polyhedral body. (See also briquette.) The process of forming tablettes. Tabletting [n.] Tabletting machine Compaction press for the manufacture of tablettes. A tall tower with enlarged conical bottom. Tall form tower To drive in or down by a succession of light or medium Tamp [vb.] blows; predensify. See tamp. Tamper [n.] A receptacle for holding, storing, or transporting liTank [n.] quids. Using heat to fuse particulate solids into agglomerates. Thermal agglomeration (See also heat bonding, sintering.) The property of various materials to become fluid when Thixotropy [n.] disturbed (as by shaking, vibration, pressure, etc.). Materials tending to exhibit Thixotropy. Thixotropic [adj.] See penetrating pore. Through pore Parts making up the compacting tool set of a tabletting Tooling [n.] machine.


Tower [n.]

Trailing edge True density Tumble agglomeration

Turret [n.] Updraft [n.] Upper punch Wear [n.] WDG Web [n.] Wet agglomeration Withdrawal process

Worm [n.] WSG

In spray drying or prilling, a cylindrical structure in which liquid droplets that were formed at the top solidify during their descend in a gas atmosphere with suitable temperature. During briquetting in roller presses the back edge of a discharging briquette. The mass of the unit volume of a solid material that is free of pores. Agglomeration technique during which agglomerates are formed by growth during tumbling; synonymous with growth agglomeration. (See also coalescence.) Rotating table carrying the compacting tool set of some tabletting machines. Upward flow of gas, for example through a particle bed. Member of the compacting tool set that closes the die and forms the top of the part being produced. Similar to erosion, but usually refers to the surface of a solid body such as a part of machinery. (Easily) Water Dispersible Granulate. Thin Jlashing surrounding briquettes made in roller presses; caused by the land area. Tumble and growth agglomeration in which the major binder is a liquid. Operation of some tablettingpresses by which the die descends over a fixed lower punch to reduce density variation in the tablette and facilitate removal of the compact. See screw. (Easily) Water Soluble Granulate. (See also instant.)


5

Agglomeration Theories The distinguishing characteristic of size enlargement by agglomeration is the formation of larger entities from particulate solids by sticking particles together by short range physical forces between the particles themselves or through binders, substances that adhere chemically or physically to the solid surfaces and form a material bridge between the particles. The components of an agglomerate are often widely disparate and, except if matrix binders are applied (see Section 5.1.2) or after shrinkage during sintering (see Sections 5.3.2 and 9.1), void spaces are present between the particles forming an agglomerate. The above definition of size enlargement by agglomeration sets this unit operation of Mechanical Process Technology apart from other grain growth techniques, particularly crystallization whereby a uniform solid body grows from a mother liquor by forming a structure in which the same atoms and/or molecules have a regularly repeating internal arrangement. As will be shown later (see Section 7.4.6), agglomeration may also play a role during crystallization if nuclei or crystallites adhere to each other in the mother liquor and form macroscopically amorphous, porous structures. Size enlargement by agglomeration is also distinguished from another particle forming technique, melt solidification. In this process a molten material is divided into droplets or extruded through die plates and cut into cylindrical pellets. The product is then solidified by cooling. The melt may be directly synthesized, as in the case of urea prilling, or obtained by heating the solid. In the latter case, similar to the meaning of the term granulation, melt solidification can be a particle size reduction, if large chunks of a solid are melted and then divided into small droplets or extrudates that are solidified, or a particle size enlargement, if a powder is melted, divided into relatively larger droplets or extrudates, and solidified. Droplet formation can be by spraying through a number of differently designed nozzles (see also Section 7.4.3)or by dividing a liquid stream either naturally, by mechanical means, or by gas or liquid impingement. Solidification is accomplished during the free fall in a cooling tower (Fig. 5.la) which results in spherical “prills” (Fig. 5.2), on a cooled stainless steel belt (Fig. 5.lb) yielding flattened “pastilles” (Fig. 5.3), or in water (Fig. 5 . 1 ~ producing ) cylindrical extrudates (underwater granulation/peIleting).

30

I

5 Agglomeration Theories

Fig. 5.la

Fig. 5.1 b

Fig. 5.1 Schematic representations o f the three most c o m m o n melt solidification processes. (a) Prilling [B.42], (b) pastillation (courtesy Sandvik, Totowa, NJ, USA), (c) underwater granulation/ pelletizing (courtesy Gala, Eagle Rock, VA, USA).

Agglomeration Theories

Fig. 5.2 Photograph o f urea prills (courtesy KaltenbachThuring, Beauvais, France).

I

31

32

I


Fig. 5.3 Photograph of the discharge end o f a pastillator also showing pastillated product (courtesy Berndorf Band, Berndorf, Austria).

Since most of the commercially produced urea for fertilizer applications is prilled by the tower melt solidification process and urea is one of the most important nitrogen providing fertilizers, farmers and suppliers often wrongly name all spheroidal agrochemicals “prills” even if they were produced by true agglomeration processes, for instance on discs or in drums (see also Section 7.4.1). In the following only size enlargement by agglomeration will be covered.

5.1 The Development of Strength of Agglomerates

Fig. 5.4 is the random cut through part of an agglomerate. Obviously, in reality, the structure is three dimensional. In such a body strength can be caused in several ways. In Fig. 5.4a the entire pore space is filled with a matrix binder. Typical examples of agglomerates held together in this manner are concrete, where the matrix between the aggregate particles consists of hardened cement (Fig. 5.5), or road surfaces, in which bitumen occupies the volume between crushed stone (Fig. 5.6). Fig. 5.413 generally looks very similar to 5.4a but shows an agglomerate structure in which the entire void volume is filled with a liquid that wets the solid particles. If concave menisci form at the pore ends on the surface of the agglomerate, a (negative) capillary pressure develops within the pores which affords strength to the body. As explained in Fig. 5.7, depicting a series of situations representing different liquid saturations in particulate bulk solids or of agglomerates, distinct distribution models exist which depend on the amount ofliquid in the structure. The term liquid saturation is defined as the percentage of total void space that is filled with the liquid. A precondition for cohesiveness of particulate solids due to the presence ofliquid is that the liquid wets the solids. Although, depending on the application, other liquids may be used to totally or partially fill the voids between particulate solids, in agglomeration water is most commonly used. Referring to Fig. 5.7, absolutely dry particulate bulk solids (Fig. 5.7a) are non existent under normal atmospheric conditions. The water molecules of adsorption layers (Fig.


Fig. 5.4 Random cut through part o f an agglomerate or a particulate bulk solid mass and explanations o f how strength may be caused. (a) Pore volume filled with a matrix binder. (b) Pore volume filled with a wetting liquid. (c) Liquid bridges at the coordination points. (d) Adhesion forces at the coordination points.

I

33

34

I


Fig. 5.7 Schematic representations o f different liquid saturations in particulate bulk solids o r o f agglomerates. (a) Dry, (b) adsorption layers, (c) liquid bridges (“pendular” state), (d) transitional (“funicular” state), (e) fully saturated (“capillary” state), (f) droplet.

5.7b) that quickly form on the solid surfaces are bonded so strongly that they are not mobile and, therefore, do not cause “liquid saturation” or moisture content which can be measured with “normal”laboratory equipment. However, as will be shown later (see Section 5.2.1), adsorption layers can participate in the development of strength by enhancing molecular (van-der-Waals) forces. With small amounts of “free” water, i.e. producing moisture contents of little more than a few tenth of a percent and, correspondingly, very small “saturation”, liquid bridges begin to form at the contact points between particles. With increasing moisture content or saturation liquid bridges form at all coordination points (see below) in the structure (Figs 5 . 4 ~and 5 . 7 ~ )Further . increase in liquid saturation produces a transitional situation in which liquid bridges and void spaces that are filled with liquid coexist (Fig. 5.7d). The theoretically highest saturation (100 %) is reached, when all voids within a bulk mass or an agglomerate are filled (Fig. 5.4b) and concave menisci are formed at the pore ends (Fig. 5.7e). Beyond complete saturation, liquid droplets shaped by the surface tension may enclose solid particles (Fig. 5.7f). Slurries, bulk particulate solids containing an excess amount of water, are shapeless. All above models exist in wet agglomeration, methods that are based on the processing of slurries, suspensions, or solutions (see Sections 7.4.3 and 7.4.6) or the presence of liquids as binders (see Section 7.4).

5.1 The Development of5trength of Agglomerates

Fig. 5.4d depicts the action of solid bridges or forces at the coordination points of a particle with other particles surrounding it in the agglomerate structure. Coordination points are points of contact with other particles and near points, areas of the particle surface which are so close to a neighboring particle surface that significant adhesion forces act or bridges can form. The coordination number is the average of the sum of all contact and near points of each particle with others surrounding it in a particular agglomerate structure (see also Section 5.3.1). Typical examples of agglomerates bonded in this manner are “natural” aggregates of very fine particles which are held together by molecular forces or agglomerates with solid bridges at the coordination points which have formed during drying of originally wet agglomerates by recrystallizing materials which had been dissolved in the liquid. 5.1.1

Binding Mechanisms

The binding mechanisms of agglomeration were first defined and classified by H. Rumpf and his co-workers (see Chapter 1).According to Tab. 5.1 they are divided into five major groups, I to V, and several subgroups (see also Fig. 5.8).

Tab. 5.1

Binding mechanisms of agglomeration

I. Solid bridges 1. Sintering 2. Partial melting 3. Chemical reaction 4. Hardening binders 5. Recrystallization G. During drying: a) Recrystallization (dissolved substances) b) Deposition (colloidal particles) II. Adhesion and cohesion forces 1. Highly viscous binders 2. Adsorption layers ( < 3 nm thickness) I I I . S u r f c e tension and capillary pressure 1. Liquid bridges 2. Capillary pressure IV. Attraction forces between solids 1. Molecular forces

a) Van-der-Waals forces b) Free chemical bonds (Valence forces) c) Associations (nonvalence);hydrogen bridges 2. Electric forces (electrostatic, electrical double layers, excess charges) 3. Magnetic forces V. Interlocking bonds

I

35

36

I


(4 Molecular Forces (E.1) Electrostatic Forces ( E . 2 ) Magnetic Forces (E.3)

(e) Interlocking(Y)

(f) Matrix Binder (1.3,1.4,E.l) Capillary Forces (Conglomerates Saturated with Liquid ) (El2 )

Fig. 5.8 Pictorial representation of the binding mechanisms o f agglomeration.

I. Solid Bridges If the temperature in a disperse system rises above approximately two- thirds of 1. the melting temperature or softening range of the solids, diffusion of atoms or molecules from one particle to an other one occurs at the points of contact. The solid bridges that develop with time are called sinter bridges. The velocity of diffusion depends on temperature, size of contact area, and contact pressure. It increases with rising temperature, larger contact area, and higher pressure. Heat can be introduced from an external source or created during agglomeration by friction and/or energy conversion (see also Section 9.1). 2. At the contact points of particles, roughness peaks may melt due to heat caused by friction and/or pressure. In such cases, liquid bridges develop which solidify quickly due to the large heat sink provided by the solids themselves. This mechanism, called partial melting, is often responsible for unwanted agglomeration and caking of substances with low melting point or softening temperature. 3./4. The formation of solid bridges by chemical reaction or hardening binders depends only on the participating materials, their reactivity, and their tendency to harden. Elevated temperature and/or pressure may improve the reaction and result in a modified, potentially stronger bridge structure. These binding mechanisms are often activated by moisture.

5.I The Development of Strength of Agglomerates

5. Temperature fluctuations can result in recrystallization and bridge formation within otherwise stable or sealed bulk particulate solids. The temperature induced physical recrystallization of some substances may extend through the interface at contact points causing solid particles to grow together. Salts or mixtures of salts that contain some free moisture may cake when exposed to varying temperatures, even if the amount of moisture is very small and the material is packed in airtight enclosures. This is because, often, more salt dissolves at higher temperature which recrystallizes if the temperature drops, forming crystal bridges between the solid particles in the bulk. During temperature fluctuations caused, for example, by day and night or seasonal differences, this is a continuing process that will, with time, result in more and stronger caking (see also Section 5.5). 6. The more common method of forming solid bridges by recrystallization of dissolved substances or deposition of suspended colloidal particles is to evaporate the liquid. The strength of crystal bridges depends not only on the amount of the dissolved and recrystallizing material but also on the speed of crystallization. At higher crystallization rates a finer bridge structure is formed which results in higher strength (see also Section 5.2.2). Colloidal particles form solid bridges if the liquid between the macroscopic particles of a disperse system consists of a colloidal suspension. During drying the colloidal particles concentrate in diminishing liquid bridges and the pressure caused by the liquid’s surface tension compacts the colloidal particles. After complete evaporation of the liquid, solid bridges remain which are made up of colloidal particles. Adhesion in the bridges is mostly caused by molecular forces which may be enhanced by electrical and magnetic effects (see Group IV below). 11. Adhesion and Cohesion Forces 1. If highly viscous binders, such as bitumen, honey, pitch, tar, etc., are applied, adhesion forces at the solid-binder interface and cohesion forces within the viscous material can be fully exploited for agglomerate strength until the weaker of the two fails. Highly viscous binders are often used as matrix binders (see also Sections 5.1 and 5.2.1). 2. Most finely divided solids easily attract free atoms or molecules from the surrounding atmosphere. The thin adsorption layers thus formed are not mobile. However, they can contact and penetrate each other. It can be assumed that molecular forces can be fully transmitted if the adsorption layer is thinner than 3 nm. Such forces are often high enough to cause deformation of solid particles at the contact points (Fig. 5.9) thus increasing the contact area and, therefore, strength of the bond between adhering partners. The application of external forces and/or elevated temperatures may increase the contact area and strength further [B.14, pp 97-1291.

I

37

38

I


Fig. 5.9 Viscoelastic deformation at the contact point between two glass spheres due to molecular attraction.

Adsorption layers may also increase adhesion forces if the layers do not contact or touch each other (see Section 5.2.1). 111. Suface Tension and Capillary Forces One of the most common binding mechanism of wet agglomeration is liquid bridges at the coordination points between the particles forming the agglomerate. Liquid bridges can develop from free water or by capillary condensation. They are often the precondition for the formation of solid bridges (see above, 1.G). If the entire pore volume between the particles of a disperse system is filled with a liquid and concave menisci form at the pore ends on the surface of the system, a negative capillary pressure exists in the interior causing strength. Wet agglomerates are very often bonded by a combination of the above two mechanisms. In that case partial volumes exist which are completely filled with the liquid while in others liquid bridges prevail. Technically it is almost impossible to attain 100 % saturation because there is a high probability that during the agglomeration process air is trapped in some pores. IV. Attraction Forces Between Solid Particles Attraction forces between solid particles are often the cause for unwanted agglomeration: bridging, caking, coating, and build-up. The most important binding mechanisms in this category are molecular, electric, and magnetic forces (Fig. 5.10). At extremely small distances between the adhesion partners these forces can be very high but, due to their short range effect, they diminish quickly with increasing distance at the coordination points. Since particles approach each other with roughness peaks (Fig. 5.11) and the absolute roughness of smaller particles is less than that of larger ones, the adhesion probability, i.e. the chance of such particles moving closer together, increases as powders become finer. High adhesion forces are obtained if fine and ultrafine or nano-sized particles are involved.


Molecular Forces van-der-Waals Forces

1.a.

/

\ Valence Forces at newly created surfaces (Recombination Bonding)

1. b.

Nonvalence Association e.g.. Hydrogen bridges betweenoxygenand hydroxyl radicals a: Association - H interacts with nonbinding electron pair of oxygen b: Water molecules intensify association c: Bridging by nonvalence association of bipolar (water) molecules

Electrostatic Forces

(-I---

3. I S

N y - - - [ N ] ,

Fig. 5.10

Magnetic Forces

-_-

I

Attraction forces between solid surfaces o r particles.

1.a) Van-der-Waals forces are naturally occurring forces at the surfaces of all solid materials. The molecules, atoms, or ions in the interior of a solid interact with each other such that they retain their relative, equilibrium positions. At the surface of, for example, a particle, the molecular forces that are directed to the outside are not satisfied and produce a force field that interacts with that of other particles.

I

39

40

I


Fig. 5.11 Model depicting the true situation at a coordination point between two particles. Roughness exists o n all real particles. 09; is the representative distance between the particles.

Then, van-der-Waals forces arise because of the electric polarization induced in each of the particles by the presence of the other ones. Forces are in the order of 0.1 eV and decrease with the sixth power of the distance between the partners. The maximum range of the van-der-Waalsinteraction is in the order of 100 nm which, compared to chemical bonds (valence forces), is large. 1.b) During size reduction (comminution) bonds between the atoms and molecules of a solid are stressed and ultimately part creating new surfaces. Immediately after separation, unsatisfied valences exist on these newly created surfaces. Normally, the free radicals quickly combine with atoms and molecules from the surrounding atmosphere, thereby becoming neutralized. However, conditions exist where either the newly created surface area is so large at any given moment that the number of atoms and molecules that are available in the immediate vicinity is too small to satisfy all the available valences or mobile, reactive atoms and molecules that could neutralize the free radicals are not present. In those cases, the valences themselves may recombine if newly created surfaces come close to each other. Such recombination bonding occurs during fine grinding due to the first mechanism, eventually resulting in an equilibrium between size reduction by comminution and size enlargement by agglomeration (“grindinglimit”) (see also Section 5.5). Recombination bonding also occurs during high-pressure agglomeration (see Section 8.1). If brittle particles break in the compact under the influence of high forces, new surfaces are created within a densifying mass of particulate solids where the possibilities are limited to satisfy the exposed free valences with gaseous atoms or molecules. At the same time, high compaction forces cause particle surfaces, including the newly created, reactive ones, to approach each other so closely that, after some lateral movement of the fractured pieces, free valences recombine, forming strong, permanent bonds. 1.c) Nonvalence associations of certain molecular groups can also cause bonding and provide strength to a particulate bulk solid. One important phenomenon, hydrogen bridges, is, for example, the prevailing, naturally occurring binding mechanisms between organic macromolecules in coal. Hydrogen bridges form if a hydrogen atom is bonded to a strongly electronegative atom, such as oxygen in a typical OH group, and the hydrogen atom interacts with the non-binding electron pair of another electronegative atom, e.g. oxygen of a COOH group. Water

5.1 The Development ofstrength of Agglomerates

(H - 0 - H) intensifies this association and the bipolar molecules can also form nonvalence association bridges which participate in the development of strength (see also Fig. 5.10, l.c, b and c). 2. At their surfaces, ionic solids possess an unsatisfied electrostatic field which is superimposed on that produced by the van-der-Waals forces. The strength of this field diminishes rapidly with distance from the surface and is soon negligible. However, this external electrical field can induce a dipole or a higher order moment in the charge distribution of the molecules in an adsorbed layer thus participating in adhesion. When two solid surfaces come in contact with each other, electrostatic forces of attraction arise as a result of the contact potential, forming electrical double layers. The physical cause for the transfer of electrons when two solid bodies come into contact is the difference between their electron work functions. Electrons migrate from the body with the smaller work function to the one with the larger one until equilibrium is reached (double layer). The action of this mechanism is permanent. Particles also can be charged by providing electrons from external sources (e.g. spray electrodes). Such excess charges can also cause attraction (or repulsion). Because of the field character of this binding mechanism, strength is independent of particle size. Also, the strength due to excess charges is very small and the charges tend to equalize (disappear) with time. Therefore, this mechanism is, in most cases, only significant for initial, temporary bonding (typical application: electrostatic precipitators/filters). As mentioned before, it is also possible that bonding between two oppositely charged solid surfaces is caused by the nonvalence association of bi- or multipolar molecules or radicals. Hydrogen bonding is a well known example. 3. The attraction mechanism caused by magnetic forces is similar to that of electrostatic forces. The presence of magnetic forces is limited to ferromagnetic particles although, recently, based on the understanding of the nature of magnetism, it was reported that it is now possible to engineer completely man-made plastic materials with magnetic properties. The latter may enlarge the applicability of this mechanism in the future. V. Interlocking Bonds Normally, interlocking bonds occur if the particulate solids have the shape of, for example, fibers, threads, or lamellae that twist, weave, and bend about each other or entangle during agglomeration. Sometimes interlocking bonds of elongated, fibrous additives are used to strengthen agglomerates which are otherwise too weak (see also Section 5.3.1). In high-pressure agglomeration, another interlocking mechanism may occur if a mixture of rigid and plastic materials is compacted. In this situation, the plastic component flows into recesses and, more generally, envelopes the exterior structure of harder particles, thus producing a strong structural bond that resembles the effect of a matrix binder (see also Section 8.1).

I

41

42

I


Fig.s 5.8 and 5.10 describe pictorially the binding mechanisms that were reviewed above. It should be pointed out that only the two-dimensional situation at one coordination point between two particles or solid surfaces is shown. In reality, each particle has many interaction sites (coordination points) with other particles in the three-dimensional structure. It should be further understood that in typical particulate bulk solids and agglomerates large numbers of particles are present per unit volume (see also Section 5.3.1) and participate in bonding due to the binding mechanisms presented above. With exception of capillary and matrix bonded structures of particulate solids, it is unlikely that only one binding mechanism acts on all the coordination points within a mass. If molecular and electric forces as well as liquid bridges and the solid bridges, resulting from the latter by one or the other of the mechanisms that were discussed above, are considered, it must be assumed that the effect of each binding mechanism is different at essentially every coordination point due to varying microscopic surface structures and distances at each interaction point (see also Section 5.2.1). 5.1.2

Binders, Lubricants, and Other Additives If size enlargement by agglomeration is desired and the correct agglomeration technique is selected, many of the binding mechanisms described in the previous Section 5.1.1 are inherently available or can be activated. Under certain conditions, some binding mechanisms also act naturally to produce undesirable agglomeration phenomena. Generally speaking, if agglomeration is wanted, means to enhance the available binding mechanisms must be developed and applied, while the effect of binding mechanisms must be eliminated or reduced to avoid unwanted agglomeration. Both aspects will be covered in much more detail in Section 5.5. As will be shown in Sections 5.2 and 5.3, particle size of the particulate solids plays an important role in agglomeration. While the surface area of particles, the interface at which all binding mechanisms act, decreases with the second power of particle size, volume and, therefore also, mass, the most important particle properties which result in forces that challenge adhesion and cause separation of bonds, diminish with the third power of the particle size. If the particle size reaches a few pm or is in the n m range, the natural adhesion forces dominate and particles which contact each other or come into close proximity adhere to one another. This phenomenon can not be economically eliminated so that very fine particles always adhere and form loose agglomerates which may be desirable or undesirable (see Section 5.5). Naturally available adhesion tendencies can be considerably increased if moisture is added during the agglomeration process. Application of external forces can contribute to the enhancement of inherently present binding mechanisms. Depending on the magnitude and nature of these forces, improved structure (by shear and low to medium compression) or plastic deformation and brittle breakage (due to high external forces) can occur. Plasticity, an often preferred response to external forces that results in high agglomerate strength (see Section 8.1), increases with many solids if the temperature of the material rises. Therefore, hot densification is often a desirable agglomeration technology, particularly for minerals and metal bearing materials.

5.I The Development of Strength ofAgglomerates

Since all binding mechanisms rely on molecular interactions on and between surfaces or interfaces, the structure and distance at these points is of great importance for the ability of powders to agglomerate. Often, the presence of ultra fine particles facilitates size enlargement of coarser particulate matter. Fines that are suspended in a liquid accumulate during drying at coordination points and form solid bridges which are bonded by molecular forces. Dry fines may fill areas with high surface energy, such as holes and depressions, thus reducing the effective distance between larger particles and increasing the attraction force (similar to the influence of adsorption layers; see Section 5.2.1). With other materials, e.g. certain coals and chemicals with low softening or melting points or containing such components, mechanical energy, introduced by dynamic forces, compression, or shear and converted into thermal energy, activates the inherently available binding properties. Under this influence, momentary softening and melting can occur upon contact at minute roughness peaks which, after almost instantaneous solidification, produce a small solid bridge between the powder particles. Similar mechanisms are responsible for the bonding of soluble materials in the presence of moisture. Mechanical energy converted into heat or the direct external supply of thermal energy result first in dissolution and then in recrystallization at the coordination points. The larger the number of coordination points in a unit volume (increasing with decreasing size of the agglomerate forming particles), the higher will be the strength of the agglomerated part. In spite of the availability of all these “natural” binding mechanisms and the various possibilities to enhance them for the desirable production of agglomerates, sometimes no economic method can be found to process a specific material and form a product with sufficient strength. Grinding the particulate solid to a sufficient fineness for strong molecular bonding and/or heating it to high enough temperatures that result in either sufficient dissolution for recrystallization, plasticity for large area contact and bonding, or sintering and melt solidification, would be too expensive and, therefore, prohibit economic processing. In those cases where no bonding can be achieved, particle size is relatively large, or specific product characteristics must be obtained, binders, mostly for higher strength, lubricants, mostly for improved density and structure, and other additives, which produce special properties, can or must be used. Binders are components which are added prior or during agglomeration to increase the strength of the agglomerated product at otherwise unchanged processing conditions. They can affect strength directly or after a curing step. Binder selection depends on many considerations which are specific for the particular application. They must be compatible with the materials to be agglomerated and the proposed uses of the product. For example, for pharmaceutical and food applications only officially approved materials may be used and for the agglomeration of metal bearing dusts which are intended for recirculation into steel mills, sulfur containing binders are normally prohibited. Many such limitations can be defined for specific materials and applications.

I

43

44

I


For those reasons, binder development can not be generally treated. Rather, each individual case must be evaluated separately. However, a few common characteristics can be considered before starting a specific development program. Binders can be divided into inorganic or organic components and their distribution in the agglomerate structure may be in the form of films and bridges or a matrix. Film or bridge type additives are normally fluids which coat particles or are drawn to the coordination points where they form bridges. If applicable, only relatively small additions are required; porosity of the agglomerates as well as their freely accessible surface area (including internal surfaces, see also Section 5 . 3 . 2 ) are only insignificantly changed. Water is the most well known film and bridge forming binder. Matrix forming binder components, on the other hand, more or less fill the entire pore space and, therefore drastically reduce porosity and accessible surface area. Cement is a typical matrix forming additive. Water or other liquids may act as matrix binder in fully saturated wet agglomerates (capillarystate, see Section 5.1). However, this is only a temporary binding mechanism and the liquid will disappear naturally or during a posttreatment step (see Section 7 . 3 ) so that pores open up and surfaces become accessible again. Still other binders will react chemically with different components of the additive mixture or with some or all of the materials to be processed. Such reactions can result in high strength products with, for example, waterproof bonds. Tab. 5 . 2 lists examples of organic and inorganic binders that were previously employed in agglomeration. It shows that many different substances and materials have already been used. Commonly available and applied binders are printed in italic letters. Investigation of by-products or wastes as binders may result in the discovery of cheap and very acceptable additives. For example, molasses, a by-product of sugar making, is an excellent and nutritionally beneficial binder for animal feed and organic wastes can be incorporated in fertilizers as nutrient and binder. Binder development must take into consideration the availability of the substance at the point of ultimate use and over time. Normally, evaluations begin at a vendor facility with traditional and/or new materials that are available at that time and location. Often such developments become unacceptable when during the final cost analysis the binder turns out to be excessively expensive due to the need for its transportation to the location of the planned industrial agglomeration facility. A recent example for the “drying out” of a binder source with time is Brewex, the somewhat modified by-product of a specific beer brewing technology. The material, a liquid starch material, which was available at reasonable cost in the USA and quickly enjoyed a relatively widespread use, had to be taken off the market when the beer brewing technology changed and the by-product source disappeared. In such a situation, the operator of an already established agglomeration process has to search for a replacement binder with similar properties, acceptable price, and good availability to be able to remain in business and continue to be profitable. Therefore, unless there is a safe and unlimited binder supply for a particular application, it is prudent to continuously observe the market, evaluate new developments, and be ready for change.

5.1 The Development of Strength of Agglomerates Examples o f organic and inorganic binders that were previously employed in agglomeration (in alphabetical order).

Tab. 5.2

Organic binders

Inorganic binders

Albumates (Albuminates) Alcohols Alcotac" Alginates AsphaltlAsphalt EmulsionslReJned Asphalts Brewex Carnauba Wax Caseins CAFA (Chemically Activated FlyAsh) Cellulose Compounds Chicken Manure C M C (Carbo-Methyl Cellulose) Coal Tar, Pitch, and Creosote Coke Oven Tar Covol Crude Oil Dextnne Drying Oils Elveron" Fir Tar (Pine Wood Tar) Fish Waste Gelatine Gilsonite@ (Natural Asphalt) Glues Gums (e.g. Arabic) Humates (Humic Acid) Lignins (Liquor and Powder) Lignite Lignite Tar Lignosulfonates Maltose Molasses Orimulsion" Paper Pulp (from secondary paper making) Paraffin Peat Petroleum Pitch Peridura Pittsburgh Flux Polyvinyl Alcohol ( P V A ) Resins (Natural and Synthetic) Rosin Sawdust Seaweed Slaughterhouse Refuse Starches, pregelatinited (e.g. Corn, Potato, Tapioca, Wheat) Straw (Ground or Pulped) Sucrose

Alkali Silicates (e.g. Sodium, Potassium) Alum Alumina (see Colloidal ..) Attapulgite (Clay) Bentonite (Montmorillonite Clay) Caustic Soda Cements (e.g. Portland, Slag) Clays Colloidal Alumina, Silica, etc. Dolomite Fuller's Earth Gypsum Lime Lime Hydrate (often as hardener) MagnesialMagnesium Oxide Magnesium Chloride Metal Swarf Metal Fibers Plaster of Paris Salts Silica (see Colloidal ..) Silicates (see Alkali Silicates) Sodium Borate Sulfates (e.g. Copper) Water

I

45

46

I

5 Agglomeration Theories Organic binders (cont’d)

Sugars Tanning Liquors (Tannic Acid) Terravest (Liquid Polybutadiene Emulsion) Thermoplastic Powders Tree Sap Vegetable Pulp Waxes and Wax Tailings Wood Pulp

Lubricants may be either liquid or solid additives (Tab. 5 . 3 ) . They reduce the coefficient of friction between the particles of a bulk mass and, therefore, result in a somewhat higher agglomerate density or lower porosity, E . According to the relationship k E = j~ (see Section 5.2.1) additional adhesion sites (characterized by the coordination number, k ) are activated by which increased agglomerate strength is expected. Tab. 5.3

Examples of some typical liquid and solid lubricants

Liquids

Solids

Glycerine Oil/Water Emulsions Water Dry Starch Molybdenum Disulfide Stearates (Metallic, e.g. Magnesium Stearate) Talc Etylene Glycol Oils Silicones

Graphite Paraffin Stearic Acid Waxes

In pressure agglomeration, lubricants also reduce the coefficient of friction between the material to be compacted and the tooling. This results in a more uniform structure of the compact and in less density variation (see also Section 8.2). During ejection from a die or release from a mold lower forces are required for separation and, therefore, higher survival rates are obtained. Development and selection of lubricants must apply the same considerations as discussed for binders above. While, in some cases, binders may be valuable ingredients of the final product or disappear during post-treatment, lubricants are almost always contaminants. For this reason and to keep costs down, the most acceptable lubricants are those that are effective in very small amounts. In former times lubricants were mixed into the formulation prior to, for example, tabletting, even if the lubricant was only meant to reduce the friction between the solids and the tooling. Newer developments came-up with applicators that deposit the lubricant on the surfaces of the tooling thus decreasing amount, cost, and product contamination considerably (see also Section 8.4.3).

5. I The Development of Strength of Agglomerates

With the growing importance of size enlargement by agglomeration for the manufacturing of engineered products (see also Chapter 12), many other additives are used as “functional” components. Particularly in the food industry (Fig. 5.12), but also in other, by the public less well known applications, materials with specific, predetermined, and controlled properties are formulated from particulate ingredients and then agglomerated to yield consumer products that feature desirable characteristics. For example, convenience foods can be easily and quickly used such as “instant” soups, sauces, and drinks or products that were recombined from fine, ground food stuffs, contain already the correct amount of spices as well as other aromas, and, after preparation, feature a texture and taste that pleases the palate. Functional foods, also called designer foods on the other hand, have been treated to eliminate unhealthy ingredients, such as fat. They are then recombined with additives that replace the removed components without sacrificing the “mouthfeel” that is expected from the untreated food. Functional foods may also contain dietary additives that make a product particularly acceptable for a special group of often chronically sick people, such as, for example, diabetics. For those reasons, the market for food additives is growing overproportionately, largely due to the increasing production of more nutritious and better balanced designer foods whereby calorie reduction agents are the largest segment. Fun foods are the wide range of modern sweets and snacks where mostly sugar and fat based binders are applied to obtain agglomerates or, for example, bar shaped products from a multitude of ingredients for the consumer martket. A more complete coverage of these fast growing technologies is far beyond the scope of this book. They are mentioned to demonstrate the wide range of applications of agglomeration in areas that are not immediately recognized as common uses of the unit operation. Still other additives are more generally introduced to overcome problems caused by the need to obtain sufficient strength for packaging, handling, and storage. Special components may have to be added to the formulation which assist in the break-up of the agglomerated product when it comes into contact with water or other liquids. Such materials are commonly starches or their derivatives and other compounds that swell when absorbing liquids. Fibers may be added for a number of reasons, for example, as a dry binder, a structural component, a moisture absorbent, and a conduit for liquid. Mixtures of carbonates will produce carbon dioxide with water and result in the well known effect of effervescence. Fig. 5.13 shows schematically the influence of wicking by fibers or swelling of suitable components on the dispersion of agglomerates in liquids. Both effects may be also used together. Produced from renewable resources, organic fibers and their derivatives have a wide range of functional applications. In the pharmaceutical and food industries, the presently best known cellulosic additive is microcrystalline cellulose (MCC). It is obtained from wood cellulose by acidic hydrolysis. The product does no longer contain lignins, hemicelluloses, or other impurities and is bleached to produce a high degree of brightness. In a cellulose molecule, approx. 15,000 D-glucose units are connected in a 1.4-pglucosidic linear arrangement to form a filamentary molecule. Individual molecules of Other Additives

I

47

48

I


Fig. 5.12

A few examples o f modern food products that were manufactured using agglomeration technologies. (a.1-a.4) Cereals and cereal bars (courtesy Kellog Co., Battle Creek, MI, USA); (b) cubed and granulated beef bouillon, both with "instant" (see Section 5.4) characteristics (courtesy Borden Foods/Wyler's, Columbus, OH/Chicago, IL, USA); (c.1-c.3) various snack bars from cereals, whole grains, nuts, dried fruit, and processed food materials (courtesy Hosokawa Bepex/Hutt, Leingarten, Germany); (c.4) various dumplings (courtesy Hosokawa Bepex/Hutt, Leingarten, Germany).


WlCKlNG

4

4 8

SWELLING

0

Fig. 5.13 Schematic representation o f the influence o f wicking and swelling on the dispersion o f agglomerates in liquids.

cellulose are bonded together by hydrogen bridges yielding pseudo crystalline stmctures. Although the hydrogen bonding is not destroyed by acidic hydrolysis, the cellulose chains are depolymerized and form “microcrystallites”. Fig. 5.14 depicts the structural and molecular formulas. n is 500 and 1,000, respectively. MCC is insoluble, physiologically inert, has high microbial purity, and is no substrate for microorganisms.

Structural formula:

Molecular formula: (C,H,,O,), Fig. 5.14

Chemical composition o f cellulose.

I

49

50

I


Fig. 5.15 Examples of PC cellulosic fibers (Vivapur 101, courtesy J. Rettenmaier & Sohne, Rosenberg, Germany).

Powdered cellulose (PC) is also prepared from plant material by chemical digestion and purification processes. Further mechanical processing, without the use of chemical additives, yields high purity fibers. They are chemically inert and insoluble in water, organic solvents, and dilute acids or alkaline liquors. Depending on the specific technical requirement of a customer, different qualities of PC fibers can be developed and manufactured (Fig. 5.15). Today, both MCC and PC fiber grades are widely used in tabletting. Depending on the composition of the formulation, one or the other cellulose product results in better hardness, friability, and disintegration values. However, the quantity of MCC required to yield comparable tablet properties is normally at least one-third higher than that of PC fibers. Since, because of a more economical production process, the cost of PC fibers is also lower than that of MCC, monetary advantages can be derived from using powdered cellulose. Other organic fiber products which are mostly used in foods as “dietary” ballast additives are made from wheat, oats, tomato, apples, and citrus. Such dietary fibers are “non-starch” polysaccharides obtained from cell walls only, which can not be broken down by the digestive enzymes of the human organism and, therefore, constitute inert ballast materials. Color, taste, and odor relate to the fiber source. Unlike cereal brans or dietary fibers derived from, for example, sugar beets, which are often rejected by consumers because of their specific taste, wheat, oat, tomato, apple, and citrus fibers offer physiological properties that are much more readily accepted. Although these fibers are primarily used in foods, there are also applications in other industries. Functional characteristics of the fibers include high water binding and retention capacity (as a rule: the longer the fibre, the more water it retains),no synergy effect with thickening agents in the normal dosage range (for example, up to 10 % of wheat fiber can be added to absorb and bind liquids and oils before a thickening effect can be detected), improvement of the rheological properties of various thickening agents (e.g., improvement of the thixotropic qualities of carbomethyl cellulose (CMC)),and free flowlanticaking agent (with very low dust content). Starches and compounds derived from starches have long been known as additives in many industry. These materials improve flowability and act as binders as well as disintegrants.


A particularly interesting newer starch derivative is sodium starch glycolate (SSG). It is the sodium salt of the carboxymethylether of potato starch or other starches (e.g. wheat, maize (corn), rice, etc.) and is a fine, almost white, odorless and tasteless, free flowing powder. Because of the low degree of substitution (see Fig. 5.16), the form and particle size of the original starch remains almost unchanged. SSG is practically insoluble in organic solvents and forms translucent suspensions or clear gels with water. Until recently, sodium starch glycolate has been used exclusively as a disintegrant in pharmaceutical solid dosage forms. Since it was found that the manufacturing process can be modified, specific SSG grades are produced for different new applications (Tab. 5.4). Finally, as further examples in the context ofthis chapter, the beneficial use of totally different fibers than discussed above shall be mentioned and reviewed. Metal swarf, fine, elongated grindings and turnings which are fibers in a generic sense, may be applied to “mechanically reinforce” briquettes made from metal bearing dusts for recycling into metal making processes. For this application, it is important to produce high strength of which at least a certain part is retained at high temperatures, until melting occurs, so that secondary contamination due to premature release ofdust is avoided. The influence of these fibers on briquette strength is demonstrated in Fig. 5.17 which depicts that the strength of briquettes increases with growing addition of swarf while the necessary amount of chemical binder, constituting contamination and non temperature resistant bonding, decreases [B.42]. Fig. 5.17 also shows a broken cylindrical compact (a)that was manufactured with a laboratory piston press during process development (see also Section 11.2) and actual

g... Fig. 5.16

Structural formula o f sodium starch glycolate (SSC).

I

51

52

I

5 Agglomeration Theories Tab. 5.4

New applications of sodium starch glycolate (SSC), according to J. Rettenmaier & Sohne (JRS). Application as

Disintegrant

Quasi soluble disintegrant

Wet granulation Taste masking binder

Thickening agent Gel former

Grade P ")

Grade P5000

Grades PlSOO, PSOOO

Grade P5000

Grades P1000, P3500, P5000

Guarantees excellent disintegration times of tablets, (film-)coated tablets, capsules, and granules.

Swells very much in water and forms translucent gels. These are particularly suitable to improve disintegration of tablets, effetvescent tablets, soluble tablets, granules, etc.

Have very good adhesion properties and are sui. table as binding agents with disintegration properties in wet granulation. They can be added in powder form and granulated with water.

Is used because of its gel forming properties to mask the taste in lozenges and chewable tablets.

Are used as Forms clear gels thickening and which are stable stabilizing agents within a wide in juices, suspen- range of sions, emulsions, temperatures. ointments, creams, etc.

Grade PO100

Typical dosage

1-5 %

2-20 %

up to 5 %

5-20 %

2-5 % in special 5 - 2 0 % cases also higher

")"P" in the grade designation refers to potato starch as origin. All grades can also be made from other starches ("M" = maize (corn), "R" = rice, " W =wheat, etc.). For formulations that are incompatible with alcohol the grade designation "SF" guarantees an alcohol content below 1 %.

briquettes (b) obtained in the industrial plant. Concerns that the swarf "fibers" would prohibit separation of briquettes that are produced with a roller press did not turn out to be a problem. Similar to concrete, refractory linings and components are agglomerates in which highly temperature resistant aggregates and mortars represent a system which is shaped and fired to yield bricks or other components that are then set into mortar for mounting, or is applied by casting or gunning. Today's high temperature processing industries demand high performance and predictable service life from the refractory. The latest generation of low cement, ultra-low cement, and self-flow castables, which are resistant to high temperatures, continue to be weak in tension and offer minimal resistance to damage from sudden changes in stress. Thermal cycling or shock as well as mechanical impact or vibration can all cause cracking, which, in turn, may lead to premature failure and substantial costs. Because the development of cracks can not be avoided in the rough environments of the typical applications of refractories, the probability must be reduced that such cracks result in failure. This is possible by reinforcement with fibers. Some materials that have been added to accomplish this are stiff, needle-like chopped wire or slit sheet fibers which are sometimes even supplied with, for example, hooks on their ends to increase anchorage. As schematically shown in Fig. 5.18 steel fibers in the refractory structure arrest the cracks and prohibit their propagation. Newer reinforcement


2001

I

I

I

1

I

I

----I

I

Z

o 180-

V

O 0-,

I

2 c

% content of ?:inforcing swarf and 3 p a r t s sulfite-waste powder as a binder

160-

01

I)

(J L

140-

a

0-0-

'ii 120-

reinforcing swarf and 5 parts sulf I t e -waste powder as a binder

E

7, 80 0,

i

-

4 - . 0

0-O

vt

5

40-

z! 0

O-

;I

I

I

1

2

-

waste powder as a binder I

3 L Hardening duration (days)

Fig. 5.17 Cold crushing strength o f briquettes from metal bearing dust, containing different amounts o f dry lignosulfonate binder (called "sulfite waste"), with and without swarf reinforcement, as a function o f (natural) curing time. Photographs o f "reinforced" metal bearing dust briquettes (scales not identical). (a) Cylindrical test briquette and fracture surface, (b) commercial pillow-shaped briquettes.

I

5

_

I

53

54

I

5 Agglomeration Theon‘es

materials use direct spun stainless fibers that are rapidly cooled. The resulting products are fully annealed and, therefore, more pliable and ductile, feature better flow characteristics and an improved aspect ratio that results in optimum dispersion due to easy disentanglement of the fibers during mixing with the wet refractory system, and offer exceptional resistance to high temperature corrosion since a beneficial metallurgical structure is “frozen” during the ultra-rapid cooling process. The photograph in Fig. 5.18 shows a representative selection of some of these stainless steel fibers for the refractory industry.

Fig. 5.18 Schematic depiction o f the crack stopping mechanism of steel fibers in a refractory. Photographs o f some typical stainless steel fibers for the reinforcement o f refractories (courtesy RIBTEC, Gahanna, OH, USA).

5.2 Estimation of Agglomerate Strength


The most important property of all agglomerates, desired or undesired, is their strength. For the practical and industrial investigation of agglomerate strength, stresses that occur in reality during storage and handling are experimentally simulated (see Section 5.2.2). In addition to the frequently used crushing, drop, and abrasion tests, methods for the determination of impact, bending, cutting or shear strength are employed. All values obtained by these methods are strictly empirical and cannot be predicted by theory because it is not known which component of the applied stresses causes the agglomerate to fail. For the same reason, the experimental results from different methods can not be compared with each other. Therefore, Rumpf (see Chapter 1) proposed to determine the tensile strength of agglomerates. It is defined by the tensile force at failure divided by the cross section or, if the test body has no uniform shape, the area of the failure plane(s) of the agglomerate@) (see Section 5.2.2). Because failure occurs in all stressing situations with great probability under the influence of the highest tensile force, this proposal is justified. Moreover, tensile force and strength can be approximated by models and theoretical calculations. 5.2.1 Theoretical Considerations

All binding mechanisms of agglomeration (see Section 5.1.1) can be described by one of three models (see Section 5.1, Fig. 5.4): 1. The entire pore volume of the agglomerate is filled with a substance that can 2. 3.

transmit forces and, thereby, causes strength (matrix binder, Fig. 5.4a). The pore volume of the agglomerate is entirely filled with a liquid (Fig. 5.4b). Binding forces are transmitted at the coordination points of the primary particles forming the agglomerate (Fig. 5.4d).

Liquid bridges at the coordination points (Fig.s 5 . 4 ~and 5 . 7 ~are ) described by model ( 3 ) while the transitional state (Fig. 5.7d) is connected with model (2) through the liquid saturation, S (see Section 5.1).

ad 1 ) Maximum tensile strength the pore volume is filled with a strength-transmitting substance If the pore volume of the agglomerate is completely filled with a stress transmitting substance, e.g. a hardened binder, three strength components can define agglomerate strength: (a) ota(pore volume strength) = tensile strength of the binder substance, (b) ota(grain boundary strength) = tensile strength caused by the adhesion between binder and particulate solids forming the agglomerate,

I

55

56

I


Fig. 5.19 Two-dimensional schematic representation o f the failure lines derived from the three models describing strength o f agglomerates with a matrix binder.

(c) o ~ ( ~ = strength -~) of the particulate solids forming the agglomerate. The relatively lowest component determines the agglomerate strength. Fig. 5.19 depicts schematically the expected failure lines in a two dimensional schematic representation.

ad 2 ) Maximum tensile strength ifthe pore volume i s filled with a liquid If a liquid that wets the solid(s)fills the entire pore volume of an agglomerate to such a degree that concave menisci are formed at the pore ends on the surface, a negative capillary pressure pc develops in the interior of the agglomerate. Because the membrane forces at the surface are negligibly small in relation to the capillary pressure, the tensile strength otcof agglomerates that are completely filled with a liquid can be approximated by the capillary pressure: Otc

Pc

(Eq. 5.1)

Assuming that the pore diameter is characterized by the mean half hydraulic radius of the pore system, further assuming perfect wetting and spherical monosized particles, the following formula is obtained: Otc N

pc = a’ ( 1 - & ) / & ax

(Eq. 5.2)

The maximum tensile strength of agglomerates that are completely filled with a perfectly wetting liquid depends on the porosity of the agglomerate, characterized by the strong term (1- & ) / e ,the surface tension a of the liquid, and the size 3c of the particles forming the agglomerate. The empirical correction factor a’ has values between G and 8. An approximation of the agglomerate strength 5‘ , in the transitional (“funicular”) state, in which a certain percentage S (= saturation, see Section 5.1) of the pore volume is filled with liquid, is possible by multiplying the maximum strength otcwith the appropriate saturation S: Ott

s (Jtc

(Eq. 5 . 3 )


ad 3 ) Maximum tensile strength ifforces are transmitted at the coordination points ofthe particles forming the agglomerate Estimation of the strength of agglomerates which is caused by solid bridges at the coordination points assumes that the entire solid binder material is uniformly distributed at all coordination points and forms bridges with constant strength oB.If, in addition, failure only occurs through solid bridges, the relative cross section of that material defines the agglomerate strength: (3tB

A,

(MB

p ~ / PB) ~ p -

OB =

VB

oB

(Eq. 5.4)

MB is the mass of the bridge building material and Mp the mass of the agglomerate building particulate solids, pBand pp are the densities of the respective solid materials, 1 - E is the relative volume of the particulate solids building the agglomerate, E is the specific void volume (porosity) of the agglomerate, and vBis the fraction of voids in the agglomerate that is filled with the bridge building material. Strength may be also caused by adhesion forces A acting at the coordination points of the particles forming the agglomerate. Based on statistical considerations and a simple model, Rumpf [5.1] developed a general formula that is often used to describe agglomerate strength: ot = (1-

E)/E

k A/d

(Eq. 5.5)

E is the specific void volume (porosity) of the agglomerate and (1- E) the respective volume of the particulate solids, TI = 3.14...., k the average coordination number, and x the representative size of the particulate solids forming the agglomerate. For k an empirical approximation exists:

kE

N

(Eq. 5.6)

TC

with which Equation 5.5 is simplified to: O~ =

(1-

E)/E

A/x'

(Eq. 5.7)

Theoretical Approximation o f Adhesion Forces The still unknown term in Equation 5.7 is the adhesion force A. Firstly, it must be recognized that, normally, more than one

binding mechanism participates in the production of agglomerate strength. Secondly, due to differences in micro conditions, it must be expected that the adhesion force A, at each coordination point is different. Therefore, Equation 5.7 becomes in its most general form: ot = (1-

E)/E

CAi/x2

(Eq. 5.8)

Work of many researchers concentrates on modelling and calculating adhesion forces that are caused by the different binding mechanisms [B.42].So far, all models are based on simplified conditions at the coordination points. For example, modelling of the adhesion force of a liquid bridge is based on two monosized spherical particles with a distance a from each other (Fig. 5.20). Adding the two adhesion force components, one caused by the negative capillary pressure in the bridge and the other by the boundary force at the solid/liquid/gas

I

57

58

I


Fig. 5.20 I

Liquid bridge between two monosized spherical

particles

contact line, a general formula for the adhesion force of a liquid bridge AiLcan be derived: Ai,

=

u x f(P,G,a/x)

(Eq. 5.9)

The adhesion force of a liquid bridge between two monosized spherical particles depends on the surface tension of the liquid a, the particle diameter x, and a function of the angle P which defines the size of the bridge, the angle of contact or “wetting angle” 6, and a dimensionless term a / x which represents the distance at the coordination point. Obviously, a large number of different partner shapes, other than sphere to sphere, are possible and, normally, the size of the partners will be different and can vary infinitely. As already mentioned in Section 5.1.1 and shown schematically in Fig. 5.11 all particles also feature rough surfaces. Proving roughness, even on the macroscopically smoothest surfaces, depends only on the magnification. Therefore, when modelling surface interactions, this can be done macroscopically,disregarding surface roughness (for example the liquid bridge model above), or microscopically. In the case of liquid bridges the latter means that the distance a is an average as depicted in Fig. 5.11 and the angle of contact depends on the microscopic topography and, therefore, results in very complicated bridge geometries which can not be modelled. Generally, the description of the true shape of a particle, including surface roughness can not yet be described unequivocally. New techniques, such as fractal dimensions [B.37], may be applied in the future to solve this problem. As another example of modelling efforts, the estimation of the van-der-Waals adhesion force will be discussed. Three different situations at the coordination point, two flat surfaces, a spherical particle opposing a flat surface, and two spherical particles, are presented. Because van-der-Waals forces are field forces, models take into consideration the atomic and molecular interactions between the two entities. A microscopic


theorie (Hamaker [5.2])assumes that all interactions may be added up and obtains the van-der-Waals adhesion force AivdWby integrating over all pairs of atoms and molecules. The characteristic term His the “Hamakerconstant” with a value ofapprox. lo-” to J. The macroscopic theorie (Lifshitz [5.3], Krupp [5.4]) calculates the interaction force from the energy dissipation of the electromagnetic fields that emanate from the bodies and obtains a similar van-der-Waals adhesion force. In this case hw is the Lifshitz-van-der-Waals constant with a value of approx. 1.6.10-20 to 1.6.10-18 J. Fig. 5.21 summarizes the model conditions and the results. 0 is the respective unit area on the opposing flat surfaces. The equations in Fig. 5.21 are only valid for distances a that are less than 150 nm. However, because already at much smaller distances at the coordination points the contribution of van-der-Waals adhesion to the strength of agglomerates becomes insignificant, this limitation is of no concern. It should be also noted that for very small distances, the “Born repulsion” is predominant as shown in Fig. 5.22. In addition to the already discussed influence of the actual micro topography at the coordination point, other conditions may influence the true adhesion forces that act between the solid partners. In the case of van-der-Waals forces the average distance a as shown in Section 5.1.1, Fig. 5.11, may be changed by the presence of adsorption layers (Fig. 5.23). From an adhesion physics point of view, adsorption layers with a thickness of less than 3 nm are so strongly bonded that they are immobile and can be considered as part of the solid. Because adsorption occurs primarily at energetically favorable locations, such as in depressions or valleys, it tends to smooth-out the surface roughness resulting in a reduction of the actual distance between the particles at the coordination point

Hamaker microscopic

Lifshitz macroscopic

Andw/O= H/6na3

A,,d,,/O

Fig. 5.21 Three model conditions for the estimation ofthe van-derWaals adhesion force and the results o f two theories.

=

hw/8n2a3

(Eq. 5.10)

I

59

60

I

5 Agglomeration Theories ARepulsion

A

A vdW. man

van- der-Waals Adhesion

AAdhesion Fig. 5.22 Relationship of Born repulsion and van-der-Waals adhesion as a function o f the distance a at the coordination point

(Fig. 5.23) and an increased adhesion force. At ambient conditions the adsorption of atoms and/or molecules from the atmosphere is a natural phenomenon. Therefore, it can also happen during storage in bulk solids and even within agglomerates. While in the latter case agglomerate strength is enhanced which, in most cases, is not detrimental, the development of adsorption layers in bulk solids can lead to difficulties during discharge, feeding, and metering.

Fig. 5.23 Model explaining the increase in strength due to adsorption layers during van-der-Waals bonding [5.1].

5.2 Estimation ofAgglomerate Strength

5.2.2 Laboratory and Industrial Evaluations

Major parameters determining the properties of agglomerates are: The primary particle size, x , distribution,&), surface area, s(x), and shape. The agglomerate size, d, distribution, Ad), and shape. The apparent and bulk densities as well as the porosity E (= voids between the primary particles), also the pore sizes and their distribution in the agglomerate. The strength of the agglomerate. Primary particle size, distribution, surface area as well as micro (= surface structure) and macro shape, define the agglomerative behavior of a given type of particulate solids. The agglomerate (used as a generic term) size, distribution, and shape together with the characteristics discussed in Section 5 . 3 determine most of the advantages of agglomerated materials. The apparent density describes the mass of the agglomerates themselves, and the bulk density delineates the space filling behavior (e.g.the packing volume) of an agglomerated product. The porosity of agglomerates (see Section 5.3.2) is another method of describing their apparent density; it is the void volume between the primary particles forming the agglomerate and defines the accessibility of the internal surface area while the pore sizes and their distribution regulate the capillary suction which is responsible for “takingup” liquids (as in absorbents). The strength of agglomerates is one of their most important properties and may have many different meanings. In most cases the attribute “strength” defines a survival characteristic and may be defined as crushing, bending, cutting, shear, or tensile strength, as tolerance to one or several drops from a specific height, thereby reproducing stresses experienced at transfer points, or as resistance to attrition and the formation of dust [B.42]. For special applications still other measures of “strength” may be elaborated that simulate the real handling or processing conditions. Scientifically the only unequivocally defined and reproducible strength, that is ultimately and with a high degree of probability responsible for all failure modes and can be also approximated by theoretical calculations, is the tensile strength. A general formula describing the tensile strength ot of agglomerates, which are held together by binding mechanisms acting at the coordination points, was given in Section 5.2.1, Equation 5.8. The equation shows, that the porosity of agglomerates plays the most important role for their strength. The lower the porosity or, in other words, the higher the apparent density of the agglomerate, the stronger is the agglomerate. Since many of the desirable characteristics of agglomerated products require high porosity, sufficient strength is obtained in such cases by selecting a suitable binding mechanism featuring high adhesion or binding forces, using a powder with a small representative particle size, applying suitable curing techniques that produce permanent bonds with high strength (e.g. by sintering), and/or incorporating temporary additives in the feed. During or after the curing step such components are removed by melting, evaporation, or combustion (see Section 5.3.2).

I

62

I


Results of experimental determinations of agglomerate strength have been published in many scientific works and were summarized in numerous specific books on agglomeration or in major chapters of more general handbooks (see Section 13.1). In the following, a few examples will be presented to describe generally important trends. For a more detailed coverage, the literature, particularly also the proceedings of the International Symposia on Agglomeration [B.4, B.14, B.18, B.23, B.35, B.48, B.701 should be consulted. Because strength depends critically on porosity, this property should be always measured first. To allow a comparison of individual strength values which were determined on different agglomerates they must be adjusted to fit a representative porosity. Then a larger number of results should be averaged and presented together with the statistical standard deviation or the minimum and maximum deviation of single values. If the density (specific mass p,) of of the solid particles forming the agglomerate (composite density if more than one material participates) is known and the volume of the agglomerate can be accurately determined, the porosity can be calculated as:

Some Results of Laboratory Determinations o f Agglomerate Strength

Agglomerates often contain moisture. If this is the case, they must be dried prior to the determination of the solid mass, M,.Also, with the exception of flat, cylindrical tablettes (see Section 8.4.3) and similarly well defined shapes, agglomerate volume can not be easily calculated. In those cases, the buoyancy of the agglomerate in a liquid is often measured. Since, according to the principle of Archimedes, the buoyancy is equal to the mass of the displaced liquid (under the assumption that the liquid does not penetrate into the agglomerate) the volume can be calculated as: Vagglomerate =

ML/PL

(Eq. 5.14)

MLis the mass of the liquid which is displaced by the agglomerate during the buoyancy test and pL is the liquid’s specific mass. The requirement that the liquid must not penetrate into the liquid can be met by using of a non wetting liquid (mercury was applied widely, also because of its high specific mass, by coating the surface of the agglomerate with a liquid repellant (e.g. oil)),or by painting a thin film of lacquer onto the agglomerate. The error caused by any of the protective measures is insignificant. If the binding mechanism between the agglomerate forming particles is not destroyed by the liquid, it is also possible to totally saturate the porous body and then reimmerse it to determine the buoyancy. By inserting Equation 5.14 into 5.13 another formula for determining porosity is obtained: E =

1-

(PL/Ps)

(M,/ML)

(Eq. 5.15)

Porosity can be also measured by pressure permeation methods [B.GO] ifthe agglomerate can not be treated and submerged in a liquid without losing its integrity. Fig. 5.24 depicts some laboratory methods for the determination of the strength of agglomerates and cohesive powders. Often, even in a scientific environment, the transversal crushing force is measured (Fig. 5.24a). This method is quite acceptable for


I bl

la1

)”

IP Agglomerate

Adhesive

If I

R

6

Fig. 5.24 Laboratory methods for the determination o f t h e strength o f agglomerates or cohesive powders [B.42].

1

I

63

64

I


perfect cylinders, such as tablettes and some extrudates. However, any spheroidal agglomerate is so irregular that a perfect diametral loading is impossible. This results in undefined stressing with compression, shear, and tensile forces acting in unknown ratios so that a wide scatter of data is obtained from undefined sources. Also, the definition of a compression “strength”by dividing the force at failure by the projection area ofthe agglomerate is, from a scientific point ofview, not acceptable. Normally, the statistical mean force at failure of testing a large number of agglomerates is reported. Crushing a sheet or cylindrical agglomerate by loading parallel opposite flat surfaces between plates is even more problematic because in very few cases the faces of the agglomerate are truly parallel resulting in uneven loading or the lateral expansion is blocked by friction between the agglomerate and the plates so that uncontrolled stress concentrations build up which may be the true cause for failure. As a consequence, data obtained from transversal crushing tests are seldom comparable. More reproducible results are obtained if the shear strength of a well defined, often specially prepared agglomerate is measured (Fig. 5.24b). This method was adapted from the well known shear cell for the evaluation of cohesive particulate solids (Jenike shear cell and derivatives [5.5]). During fundamental work on the binding mechanisms and strength of model agglomerates or cohesive powders, most of the laboratory evaluations determine tensile strength (Fig.s 5 . 2 4 ~- g). Machinable agglomerates are converted into cylinders that are glued between two adapters (Fig. 5.24~)and torn apart in a standard tensile test machine (e.g. Frank, Instron, see also Section 11.2). Other methods use “split” dies, with or without mandrils, for the manufacturing of a compacted agglomerate which is then pulled apart at a cross section which is defined by the split mold. Low strength caused by various binding mechanisms with or without prior densification is measured in flat split containers of which one part is fued and the other part is movable with insignificant frictional resistance. The load can be applied by slowly lifting up the support table (Fig. 5.24e) or providing an incrementally increasing horizontal force (Fig. 5.24f). Often, it is desirable or necessary to measure the strength of agglomerates which are partially or completely filled with a liquid. Particularly, in the high range of saturation the correlation of strength with the capillary pressure and their change during wetting or drying of the bed (hysteresis effect) is of interest. For this purpose, the simple method shown in Fig. 5.24fwas modified as shown in Fig. 5.248 [B.42]. The most reliable results of tests determining agglomerate or cohesive powder strength that can be also interpreted best are those in which the binding mechanism is caused by the surface tension of liquids and/or the resulting capillary forces (see Section 5.1.1, I11 in Tab. 5.1). With a high degree of probability the influence of other binding mechanisms can be excluded in agglomerates or powders that are bonded by a liquid. As shown in Equations 5.2, 5.3, and 5.7 together with 5.9 (see Section 5.2.1), this binding mechanism in defined by the surface tension a as well as other characteristics of the liquid and the solid, such as the wetting angle 6, the porosity E , and the representative size of the particles forming the agglomerate. Fig. 5.25 depicts the tensile strength, determined in the laboratory according to the method shown in Fig. 5.24c, of nearly saturated agglomerates made from narrowly distributed quartz and limestone powders as a function of the size 3c of the particles


d

5 m C

E

c

m

Fig. 5.25 Tensile strength crt o f nearly saturated agglomerates as a function o f the size x o f the particles forming the agglomerate. Porosity adjusted to E =

0.0 0 6 1 o,ooLI

I

2

0.35. otcaccording t o Eq. 5.2.

1 1

1

1

1

I

I

L 6 8 10 20 LO P a r t i c l e s i z e x Ipm)

I

60

/

100

forming the agglomerate. The porosities of the individual agglomerates were adjusted arithmetically to E = 35 %. The diagonal lines represent the theoretical tensile strengths according to Equation 5.2 with a’ = 6 and a’ = 8, respectively. The diagram shows that the relationship o, l / x is fulfilled. The actual values are lower than theoretically predicted because the agglomerates which were produced in a pan (see Section 7.4.1) are not fully saturated with water and the structure of technically manufactured agglomerates is not perfect. Although not unequivocally visible in Fig. 5.25, regression analyses of these and many more sets of data revealed that the representative particle size for agglomeration processes is the surface equivalent diameter, x,. The importance of this representative diameter for the unit operation is not surprising as structure and bonding of the products critically depend on the surfaces of the particles forming an agglomerate as well as on the surfaces’ microscopic and macroscopic conditions. Of course, only the exterior particle surface is responsible for the effects; potentially internal surface area of the agglomerate forming particles must not be included when calculating the surface equivalent diameter of a particulate mass. Therefore, experimentally, surface area should be determined by permeametry, for example the well known and in the cement industry universally applied Blaine method [B.60]. The data in Fig. 5.26 confirm that the relationship between tensile strength o,,agglomerate forming particle size x , and surface tension of the binder liquid a and the porosity function (1- E)/E as per Equation 5.2 is correct and Fig. 5.27 proves that the (compression) strength of agglomerates increases linearly with the surface tension of the binder liquid as indicated by Eq. 5.2. Finally, Fig. 5.28 presents the tensile strength o,of moist and wet agglomerates as a function of liquid saturation S. At the two extremes S = 0 % and S = 100 % the strength

-

I

65

66

I


3.0

I

I

I

al

T

1 0.L

LT

I

I

I

1

/ Y

I

I

0.6

I

I

1.2 0.8 1.0 Porosity function ( 1 - E ) / E

l.L

Fig. 5.26 Relative tensile strength % / a o f agglomerates made from sperical glass powder related to the porosity function (1 - E)/E and compared with the theory (Eq. 5.2) [B.42].

N -

1.0

I

I

I

I

I

I

I

E E

-2 0.8b

0.6 5 Gl C

-

c

m

710 (Nlcm)

$0 0 0

Surface tension a

x

lo5

8b

Fig. 5.27 Compression strength 0 ofspherical wet agglomerates as a function o f the surface tension a o f the liquid [8.42]. Surface tensions are that of pure alcohol and water and o f 30/70 and 10/90 vol.% mixtures o f alcohol and water.

is close to zero. “Bone dry” powders feature very low tensile strength unless they are compacted or the representative size of the agglomerate forming particles is 90 %. At approximately that point the maximum tensile strength of wet agglomerates exists. The circular and square open symbols represent average values of experimental results of measurements of tensile strength with their standard deviation. In the bridge model (“pendular”)state the results seem to fit the curve for a / x = 0.02 best and for high saturations (“capillary” state) they approach pe. In the transition (“funicular”)range between 30 % < S < 95 % both binding mechanisms, liquid bridges and saturated pores, contribute to the development of strength. The fact that in the transition range a difference exists between the strength of agglomerates to which liquid was added (e.g. during agglomerate growth, circular open symbols) and agglomerates from which liquid was drained (square open symbols) confirms that liquid can only be drained from saturated pores and liquid bridges are not influenced. The mechanism of capillary flow in wet agglomerates is an important factor if the liquid is a solution or becomes one (because all or some of the agglomerate forming particles are soluble) and the dissolved material recrystallizes during drying [1.1,B.421. If the agglomerate is highly saturated, drying takes place only on the surface. Liquid moves by capillary flow to the surface where evaporation occurs and recrystallizing substances deposit. The formation of a crust may influence further drying of the porous body considerably. The developing crust reduces the drying rate and may, after

I

67

68

I


forming a dense crust, stop drying altogether. Since the crystal structure is influenced by the drying rate, the strength of recrystallizing substances in agglomerates during drying will be controlled by either the drying temperature, the crust, or both. Fig. 5.29 presents the tensile strength q of the core of dry agglomerates with recrystallized salt bridges which was obtained after removing a surface layer (including a crust, if applicable). The diagram shows that for agglomerates with very low initial moisture contents (curves 1 and 2) the strength increases as expected, almost linearly with an increasing amount of available salt (rising saturation) and with the drying temperature. At higher drying rate, finer and stronger crystallites grow at the coordination points in the agglomerate and, because the liquid (solution) was concentrated in discrete, immobile bridges, no crust had developed. At an initial saturation of 20 % (curve 3) the formation of a crust begins to influence strength at high drying temperatures while for the highest liquid saturations (S = 45 % and 60 %) the dense crust, formed at all temperatures, is the deciding factor for drying and development of strength (curves 5 and 6). Above 175- 200 'C the temperature within the porous body rises so quickly that the vapor pressure building up below the dense crust causes the agglomerate to burst (Fig. 5.30). The unexpectedly high tensile strength obtained at a liquid saturation of 30 % and a drying temperature of 350 "C is due to the formation of a network of small cracks in the crust that did not cause the agglomerate to fracture but increased the drying rate and, thus, the tensile strength of the dry agglomerate core.

Drying temperature t d ('C) Fig. 5.29 Tensile strength ct o f the core o f agglomerates with salt bridges as a function o f t h e drying temperature td for different liquid saturations 5 prior to drying [1.1, 8.421.


Fig. 5.30 Photographs ofcylindrical agglomerates which contained a high amount o f a nearly saturated salt solution and burst during drying.

The above mentioned incrustation may be positive or negative. On the positive side, the phenomenon can be used as a method to achieve encapsulation of agglomerates if a film forming, easily soluble polymer is dissolved in the liquid phase. On the other hand, if a dryer is controlled by sensing the moisture content in the off-gas,the process instrumentation may mistakenly identify a heavily encrusted product as being dry when, underneath of the crust, moisture still remains. Such a product can, of course, cause a whole host of problems, such as caking during storage when the liquid slowly redistributes and problems during a secondary process, for example tabletting of a still partially moist granulated pharmaceutical formulation, as well as many more difficulties. During the initial phase of drying, when all evaporation occurs on the surface of the porous bodies, the temperature of the material to be dried stays at or below 100 “C. For highly temperature sensitive materials this temperature can be lowered by the application of vacuum. However, if incrustation occurs, the temperature of the mass to be dried increases to the temperature of the drying gas and can cause damage to the material. A considerable amount of fundamental research is going on in many places of the world trying to increase knowledge of all binding mechanisms and develop numerical methods to calculate or at least estimate binding forces as well as agglomerate strength. In addition to the “classic” standard methods discussed above many novel technologies, such as, for example, application of the atomic force microscope (AFM) (also called lateral force microscope (LFM) or scanning probe microscope (SPM)) for the measurement of adhesion in the micron and submicron particle range, and new theories, for instance, Fractals [B.37] and the Chaos Theory, are applied to agglomeration research. However, as mentioned earlier (see Section 5.1.1), it is unlikely that only

I

69

70


one binding mechanism acts on all coordination points within even a single agglomerate. Moreover, the microscopic conditions at each coordination point are so diverse that bonding at virtually each individual coordination point is different. Therefore, although big advances are being made, the science of agglomeration is still far away from formulating a useful general theory. Furthermore, this book is devoted to a more practical coverage of agglomeration. Therefore, the reader is encouraged to search for and study the increasing number of publications that report on the advances in this area (see Sections 13.1 and 13.2). Industrial Evalutations of Agglomerate Strength Determination of agglomerate strength in industry is much more pragmatic [B.42].Although knowledge and understanding of the fundamentals of agglomeration, particularly the nature and effect of the binding mechanisms and how they can be influenced, become more and more important during the development of new and for the optimization of existing agglomeration processes, agglomeration as a unit operation is still more an art than a science. While an increasing number of criteria are known for the preselection of the most suitable agglomeration process for a specific application (see Chapter 11), it is still necessary to test the selected equipment in the laboratories of vendors or development organizations (see Section 11.2). Often, if the process is a new one, it is even desirable to operate a smaller pilot plant or to involve a “toller”,an outside processor for hire, prior to an investment decision for a large scale plant (see Section 11.2). Agglomerate strength in industry is defined as a commercial or process characteristic of the particular intermediate or final product. For example, if the agglomerated material is a final product, strength may be defined as resistance to breakage, chipping, or abrasion. The definition of this property and of other strength related requirements will differ whether it is an industrial bulk material or a consumer product. While the former may break down to a certain extent, as long as it remains free flowing and dust free, a consumer product must have perfect and pleasing appearance where even minimal chipping or breakage into large chunks must be avoided. Intermediate products must have characteristics that are suitable for the intended further processing. For example, a material may have to be strong enough and abrasion resistant for storage and handling to avoid bridging, flow problems, dusting and segregation of components. If it is a feed material for tabletting or other pressure agglomeration methods, the agglomerates must break down totally under pressure and produce a uniform final product structure. Other agglomerated intermediates may have to feature the opposite property, i.e. to yield a filter with bimodal pore size distribution it must retain its shape and structure during pressing (see also Section 5.3.2). For those reasons, “strength”means many different things in industry. Typically, measurement of strength is based on a simulation of the stresses which a particular agglomerated product must withstand. Very few industrial methods for the determination of this property are standardized or even known. In a competitive environment it is of less interest to compare quality between rivals than to make sure internally that the product properties that are expected by the industrial or public consumer are maintained. Therefore, most measurements of strength are undertaken as quality assurance. A few will be described below as examples.


A general problem associated with the determination of product properties in industry is sampling [B.24, B.271. Particularly the measurement of strength is in most cases based on totally or partially destructive methods. If taken during production, these “lost” samples are extracted from the product stream in a random but representative manner and either tested directly “in-line”or, sometimes after again sampling the sample, in a quality assurance laboratory which is associated with production. Afterwards they are discarded. During initial and occasionally repeated process optimization, the influence of different process parameters on agglomerate strength is determined. Results of the measurement of strength are often used to adjust process parameters as required. If there is a difference between “green”and “cured”or final strength, both strength values may have to be evaluated to allow adjustment of the respective process steps. Even bigger problems exist iflarge bulk masses (e.g.stock piles, silos, ship loads, rail cars, trucks, etc.) must be sampled. This is done to guarantee product quality prior to or after shipment and at the point of consumption. Results of those tests are only of commercial value because, typically, they can no longer be corrected but may influence acceptability or price of the commodity. Often, if quality is below standard but does not meet the guarantee, the price will have to be adjusted by offering discounts or rebates. Among the few standardized methods for determining agglomerate strength are the compression strength ( I S 0 TC 102/Sc 3 DP 4700 and ASTM E 382-97) and tumble ( I S 0 3271 1975 E and ASTM E 279-97)tests for iron ore pellets as well as the “tumbler test” for coke (ASTM D 294-72). In this case, a group of consumers (steel companies) forced a growing number of independent suppliers to test and guarantee agglomerated bulk commodities by formulating the standards. For iron ore the tests are on finished pellets, either prior to shipment or at the consumer’s facility and, therefore, are not intended or even suitable for process control. It will be shown later that iron ore pellets are first produced as “green” agglomerates and then indurated by sintering. For the determination of compression strength, a bulk sample is first screened and at least GOO pellets are taken from the size range in which the maximum is found. In a “Riffle” splitter [B.24, B.271 four samples, containing at least 100 pellets each are prepared. From two of the samples, individual pellets are placed between the parallel, surface-hardened platens of a compression testing machine, loaded with a constant speed, and crushed. The maximum force at which each pellet breaks is determined and recorded. After testing 100 pellets of each of the two samples, the arithmetic averages for the batches are calculated. If they deviate by more than a predetermined amount, another 100 pellets are tested to confirm one or the other value. In the tumble test the abrasion resistance of the pellets is measured. The “ASTM drum” is a cylindrical container with specific dimensions which is rotated around its horizontal axis at a predetermined speed and for a defined number of revolutions. A given mass containing a representative sample of clean pellets is filled into the drum, tumbled for so many revolutions at the constant speed, removed and screened at GOO m. The “strength”of the pellets is defined by the amount of “fines”smaller than GOO m that was abraded during the test. For coke, the “tumbler test” is carried out correspondingly in a similar drum (Fig. 5.31).

I

71

72

I


SIOL LLLVATIOY. M U f I N S L E W ON

Fig. 5.31

*I

LIm C L C V A I I W

,W

f I*

Itnlol4

om

A&

Sketch o f the ASTM “tumbler test apparatus” for coke

Both compression and tumble tests have been widely used and modified for other applications. The crushing test can be utilized for any agglomerate that is large enough for individual testing and, sometimes, a single layer of many narrowly sized granules is crushed by this method. However, as discussed above, in most cases, due to a more or less irregular macroscopic and microscopic shape of the agglomerates, stressing is not uniform or reproducible and, therefore, the results can not be used for scientific or general purposes. If large enough numbers are crushed and the results are statistically treated and evaluated, the average values are good enough for quality control in a specific plant. It has been repeatedly shown, however, that a comparison of data between different plants, laboratories, and even between technicians in the same laboratory (often referred to as the “human effect”) is not possible. Many publications also report on the fact, that most agglomerates do not break under the influence of a single, well defined force. Rather, because agglomerates are porous bodies, which are made up from particulate solids with binding mechanisms acting between them, and often feature irregularities in their structure, they will disintegrate in steps. It is possible, that several small pieces break from the agglomerate before it finally fails catastrophically. Other products, particularly wet agglomerates, deform plastically before failure occurs. Some binding mechanism, for example those caused by highly viscous binders or capillary forces (see Section 5.1.1), can also produce a “selfhealing” effect after a first, smaller crack has developed. Therefore, even an unequivocal definition of the crushing strength is problematic. For the testing of other agglomerates by tumbling, the drum shape and execution is often modified. To avoid a sliding motion and produce cascading during the test, square drums have been designed or varying numbers of differently designed lifters have been built into cylindrical drums. The composition and mass of the sample to be tested, the rotational speed, the duration of the test, and the screen size defining “fines” are varied to fit particular needs.


If the abrasion resistance of smaller granules, for example of fertilizers, agrochemicals, intermediate products, etc., must be tested, specific, often smaller drums can be used as described above. Recently, based on this technique, again influenced by pressure from consumers and the desire to develop a quality assurance plan, the Saskatchewan Potash Producers Association has defined a standard procedure for the determination of degradation characteristics (= “strength”)of this granulated bulk fertilizer [5.6]. More often however, a representative sample is placed on the particular test screen that defines the “fines” and vibrated or shaken in a laboratory screening machine (e.g. Rotap, Fritsch, etc.) for a predetermined time [5.7].To produce a sufficiently significant amount of abrasion for quality control, “grinding media”, such as a specific number of steel bearing balls of a particular size or other pieces with the same purpose, are added to the granular sample. If the separation size defining “fines”is very small and, therefore, the screen is delicate, the test can be carried out in the pan. The amount of fines is then determined in a separate screening step. Because of the well defined shape of tablets, crushing tests are regularly and with great success used in the pharmaceutical industry in-line or off-line and often automatically, in combination with an automatic sampler, for monitoring tablet strength. Other modern, fully automated equipment measures tablet weight, thickness, diameter, and hardness for quality control and validation (Fig. 5.32). For some agglomerated products it is important to make sure that they meet certain strength related characteristics. For example, many animal feeds are pelleted by extruding mixtures of conditioned components through cylindrical bores in flat or cylindrical dies (see also Section 8.4.2).While pelleted food for fowl or fish is swallowed whole, products for feeding mammals need to be chewable. A compromise must be found between high strength and abrasion resistance, which allows storage, transportation, and handling without breakdown andlor the production of fines, and the requirement that pellets must be safely crushed between the teeth of the animal. A crushing test to measure this type of crushing strength was developed (Fig. 5.33) and is used for quality control in feed mills.

Fig. 5.32 Photograph o f the “Schleuniger Autotest 4” tablet testing system for the quick and automatic measurement of tablet weight, thickness, diameter, and hardness (courtesy Dr. Schleuniger Pharmatron, Manchester, NH,USA).

I

73

74

I


Fig. 5.33 Handheld (a) and automated (b) crushingtest equipment for the determination o f strength o f pelleted animal feed (courtesy Amandus Kahl, Reinbek, Germany).

In many industries, the requirements on product quality are not very stringent. The agglomerates must withstand handling, including the loading of silos and transport vessels, as well as transfers. Generally speaking, they must survive several drops with only limited breakage and the creation of a minimum of fines. For the measurement of this characteristic “drop tests” are carried out. The actual equipment and procedure may vary widely and is normally a simulation of the expected “abuse” which the material will encounter on its way from production to consumption or use. Fig. 5.34 is the sketch of a typical arrangement. A drop test arrangement can be easily built and applied in the field. The “equipment” consists of a heavy plate (1)made from steel with at least 20 m m thickness, a concrete slab, or - in the most simple case - a stone laboratory floor on which a


Fig. 5.34

Sketch o f a drop test arrangement (explanations see text).

tube or some kind of collar (2) is placed to contain the sample after the drop and avoid material losses. Diameter, height, and material of construction of the retaining wall must be such that, after impact, pieces can dissipate without experiencing secondary breakage. A vertical pipe (3), the upper end of which is at a distance h from the impact plate on the floor, extends into the retaining container. Length and diameter of the pipe depend on the size of the agglomerates to be tested. The diameter should be at least 5times or, even better, lotimes greater than the largest agglomerate dimension. The length is simulating the expected drops during further handling of the product. The pipe must end at a sufficient distance from the impact plate to allow free lateral movement of the mass upon impacting the plate. The test itself can be carried out in different ways. One method is to drop batches, each, for example, consisting of five large agglomerates (in most cases briquettes), one after the other, from different, increasing heights. “Strength” is defined as that height from which all five agglomerates still survive the drops without damage. This test determines the maximum drop height that can be tolerated in a plant which must produce whole agglomerates and handle them without breakage. Such a requirement may exist if products are manufactured that must have a certain appeal such as charcoal briquettes for barbecueing, salt briquettes for the regeneration of home water softeners, or, generally, consumer products. The test as described before is carriedout during system development, prior to plant design: later, for quality control during operation, representative samples are extracted in regular intervals from the product stream and dropped from the predetermined height to recheck and confirm their survival. Often it is not necessary to produce industrial agglomerates that must survive all handling completely intact. In this case, a relatively great drop height is selected

I

75

76


and individual agglomerates are tested. The particle size distribution of the broken pieces is determined by screening and the result of the drop test is judged based on the amount of “fines”that is produced during impact. The definition of what constitutes fines and their permissible amount depend on the application. In still other cases, the collective behavior of a large number of agglomerates is of interest. Then, a sample, sometimes weighing several kilograms, is dropped, all at once, from a bucket through the pipe. This setup simulates the “cushioning” effect of a bed of material at the impact point. The evaluation is again carried out by screening and determining the amount of “fines”.During an alternative procedure the entire sample or only pieces larger than “fines”are dropped again or repeatedly for a specific number of times. The evaluation of the test is done in the same way as before whereby data are determined either after each drop or after a certain number of drops. These few examples of industrial methods for the determination of agglomerate strength and descriptions of how these test may be carried out shall suffice. The technical literature ofthe past century is full ofreports that cover industrial plants producing agglomerates and evaluations oftheir “strength”,whatever that characteristic may mean in a particular case. Readers who wish to know more are encouraged to seek out these publications whereby it should be recognized that, in the past, most papers have appeared in application oriented journals and proceedings of conferences. The above descriptions of alternatives that are possible in the execution of the same test to obtain specific information for agglomerate handling or use should also help to understand how existing methods, which may have been used in a completely different context for other agglomerates, can be adapted interdisciplinarily to fit new requirements.

5.3 Structure of Agglomerates

Agglomerates are bodies that are, often artificially and with purpose, produced from individual “small” particles. The term “small” is to be understood in relation to the agglomerate. Although there are agglomerates, for example in the food industry (see Section 5.1.2 and Fig. 5.12), or natural, often undesired agglomerates (see Section 5.5), which consist of only a few particles, typical agglomerates contain very large numbers of particles (see Section 5.3.1, Table 5.6) with sizes that are orders of magnitude smaller than that of the agglomerate. Binding mechanisms (see Section 5.1.1) cause these particles to temporarily or permanently stick together and form a lose or porous entity (see Section 5.3.2). Since binding mechanisms act in different ways (see Section 5.1), the structure of agglomerates is of great importance for all properties of agglomerates. The sketch in Fig. 5.35 depicts a random cut through an agglomerate. The area within the heavy solid lines is arbitrarily defined as “one”. Fig. 5.35 seems to show particles and their distribution. In reality, what is visible are cross sections through particles at a random level. If another random cut through the same agglomerate is made, a totally different picture is obtained. Moreover, particles that seem to float in space are in contact with other particles at some level. For example, the shaded cross section maybe the result of cutting the particle, shown in elevation on


Examples of:

@

Contact points

0 Nearpoints

Elevation (Side View)

Fig. 5.35 Sketch o f a random cut through an agglomerate.

the side of Fig. 5.35, at the indicated line. Obviously this particle will have a completely different outline at another level. The same observation is true for the void spaces (= porosity) that are visible between the particle cross sections. If the heavily bordered square in Fig. 5.35, which represents the area “one”,is large enough and contains a great number of the two significant structural characteristics, i.e. outlines of cross sections through particles and of pores between the particles, a statistical evaluation of any random cut will produce generally valid results with an accuracy that can be described by the standard deviation which is associated with that statistical treatment. Therefore, for example, scanning the picture of the cut will produce information on particle size and distribution, porosity, E, solids content, 1 - E, and, with the appropriate software, a shape factor and the specific surface area of the particles [B.GO]. Accuracy can be increased by investigating multiple cuts through the same agglomerate and determining the statistical averages for all of them. A visual evaluation of the enlarged picture of the cut through an agglomerate also reveals certain other features, although the observations can only be used to explain phenomena and do not serve any scientific purpose. The shaded circles in Fig. 5.35 indicate, for example, some of the contact points between particles in this particular cross section and the open circles depict some of the “near points” at which a binding mechanism, such as liquid bridges or one ofthe field forces (see Section 5.1.1), could develop. The average of the sum of both types of interaction points for one particle defines the coordination number k. Taking into consideration the statements made above in regard to random cuts, it is of course possible that “near points” in a particular cut are actually contact points in a level slightly above or below and it is impossible to determine all the interaction points which are distributed three-dimensionally around a particle.

I

77

78

I

5 Agglomeration Theories 5.3.1

General Considerations

The structure of agglomerates depends on many different parameters. Generally, they are parameters related to the particles building and all the processes involved in forming the green and final agglomerate. Particularly during high-pressure agglomeration (see Section 8.2) and various post-treatment (curing)processes (see Sections 5.3.2,7.3, and 8.3) parameters that relate to the original feed particles may change. Parameters Related to Particles Building an Agglomerate. The most important particle related parameters that influence agglomerate structure are:

Particle size, Particle size distribution, Macroscopic particle shape, Microscopic particle shape (surface configuration, e.g. roughness). With the exception of geometrically well defined particles, particularly spheres and cubes, it is difficult to describe with one dimension and measure particle size (Fig. 5.36). During particle size analysis [B.60],the response of each particle to a physical effect is determined; for example, whether a particle will pass a defined opening (screening),how fast a particle will settle in a stationary fluid under the influence of gravity or centrifugal force (sedimentation),at what speed of a gas flow a particle will be entrained (sifting), how much extinction will be caused by a particle passing through a sensing zone (sensor output),what is the outline of a picture or projection of a particle (scanning),how much energy is reflected from a particle at a particular angle (scattering), etc., etc. Only absolutely spherical, microscopically smooth particles will produce results with which particle size (diameter of the sphere) is determined unequivocally and by modifying the effect which determines size, the distribution of the sizes of spherical particles can be determined, if dilute samples are analyzed where the particles do not influence each other during the test. For all other situations, particle shape has an overwhelming effect on how they behave during a test or in any process. Shape is characterized by form and proportions. Form refers to the degree to which a particle approaches a definite form, such as a sphere, cube, tetrahedron (Fig. 5.36), or higher order polyhedron. The relative proportions distinguish one spheroid, cuboid, tetrahedron, or polyhedron from another of the same class. Macroscopically, shape may be described rather subjectively by comparison with “standard shapes” or defined by coefficients (Fig. 5.37). The major problem of characterizing the three-dimensional shape of a particle by its size is that size is one-dimensional and coefficients of “standard shapes” are two-dimensional. To overcome this problem, particles may be described by polar coordinates, for example, radius vectors from the center of gravity extending to any point of the surface. By using the radius and the two polar coordinate angles, the shape of the particle surface can be described to any desired degree of accuracy. Obviously, for the time being, this technique is limited to scientific work.


Microscopically, particle shape, particularly surface texture, may be defined by fractals or Fourier functions. It must be realized that in nature no absolutely smooth surfaces exist. With increasing magnification macroscopically smooth, e.g. polished, surfaces first reveal scratches, caused by the polishing media, and later “natural”roughness with peaks and valleys. Since the size of small particles which are interacting with other, larger particles extends into the nano range and, therefore, such particles are themselves similar to or potentially smaller than many surface features on other particles, it is understandable that knowledge of the microscopic particle texture is of great importance for agglomeration. However, even with the quickly growing technological advances in the nano scale it is still impossible in practice to apply the information for general theories with which agglomerate characteristics can be predicted. As will be shown below, the extremely large number of particles that are involved and their variability (it can be assumed that no two particles are exactly alike) is another reason for today’s inability to generally and unequivocally describe the interactions between particles in an agglomerate.

0 97

00000

0 95

00000

0 93

00000 00000

0 91

0 89 1 .

0 87

+.

p L oi p Ln

Fig. 5.37 “Standard set o f shapes” for the determination of particle sphericity according to Rittenhouse [8.42].

~

00000 00000

085

00000

083

00000

0 81

oQDO0

0 79

00000

0 77

00000

075

00000

I

79

80

I


All that is normally known about a particle is its silhouette, projection, or profile or, in those cases where size is derived from other physical effects, such as, for example, the settling velocity, a dimension related to volume, mass, or surface texture. Therefore, methods must be found that interpret information from cuts through the particle, scans of portions of the surface area, or information from particle behavior in, for example, fluids and connect it with overall shape. Unless the measured outline of the particle misses a unique, dominant feature of the particle shape, the result will be representative of the particle. The methods are still very complicated and require a large number of discrete items of information to describe a particle signature reasonably well. Shape influences particle behavior in powder packings and the representative (particle) equivalent diameter changes with the physical situation. In agglomeration, in addition to the surface equivalent diameter of an entire particle size distribution, which is the representative value for estimating the strength of agglomerates (see Section 5.2.2), for the packing structure, the diameter of an inscribed average circle representing the particle projection is less significant than that of the circle enveloping all peaks and protrusions (Fig. 5.38). This is particularly true for loose packings (Fig. 5.38, top). A different equivalent diameter would be representative for closely packed particles (Fig. 5.38, bottom). It appears possible that, in the future, equivalent particle diameters can be computed for loose and dense packings. Furthermore, it should become feasible to calculate the work that is required to go from one packing structure to another, the resistance of a powder to penetration, and its angle of repose. It is already possible to characterize the pore structure of a particle system by using automatic scans of a cross section and employing fractals to analyze the data. Nevertheless, the characterization of particulate matter and the structure of particle systems is still at the beginning of becoming an exact and widely used science.

P a r t i c l e size versus packing r a d i u s Size Packing radius-

Close packed p a r t i c l e s

Fig. 5.38

Effect ofparticle shape on its packing behavior.


Much of the industrial research of packing structures is still based on spherical particles. The most fundamental information is obtained if the regular packings of monosized spheres are evaluated. Fig. 5.39 shows the six regular packing structures of monosized spheres. For the packings depicted in Fig. 5.39 porosity E , the void volume between the monosized spherical particles, and the coordination number k, the number of interaction points of a sphere in the structure with neighboring spheres, can be exactly determined (Tab. 5.5). All coordination points in these structures are contact points.

Fig. 5.39 Systematic arrangements o f spherical particles ("regular packings"). For explanations see text and Table 5 . 5 .

I

82

I

5 Agglomeration Theories Porosities, coordination numbers, and approximations according to Eq. 5.6 for the “regular packings” shown in Figure 5.39.

Tab. 5.5

-

Coordination number, k

k

E

Cubic (Figure 5.39a)

0.476

6

6.599

Orthorhombic, two alternatives (Figure 5.3%)

0.395

8

7.953

(Figure 5.3%)

0.395

8

7.953

Tetragonal-spheroidal [Figure 5.39d)

0.302

10

10.40

Rhombohedral (pyramidal) [Fig. 5.39e)

0.260

12

12.08

Rhombohedral (hexagonal) (Fig. 5.3%)

0.260

12

12.08

Geometric arrangement ~~~

Porosity

n/c

(Eq. 5.6)

~

Tab. 5.5 shows that the porosity of regular packings ranges from 47.6 % for the most open structure to 26 % for the densest packing. Real packings of monosized spheres are called irregular packings. Unlike the specified location of each sphere in a regular packing (deterministic system),the position of any sphere in a randomly packed bed can only be described by a probability distribution (stochastic system). Moreover, the density (or porosity) of a randomly packed bed depends on the mode of packing. Normally, in freely developing, infinite beds two structures are distinguished: a very loose random packing with a typical porosity of 40-43 % and a lose random packing with 39 -41 % porosity. If packings are produced in a container they are influenced by the “wall effect” (Fig. 5.40). On and near rigid walls the positioning of the spherical particles can not occur freely and this disturbance is continuing into the packing, creating voids and other irregularities. Nevertheless, poured random packings in containers may attain 37 - 39 % porosity, depending on the dimensions of the container in relation to the size of the spheres and, if the container is vibrated or vigorously shaken, a porosity of approx. 36 % may be obtained. Tab. 5.5 also shows that the coordination numbers for regular packings of monosized spheres are 6, 8, 10, and 12 and that even for these unique conditions the approximation of Equation 5.6 is rather good. Therefore, it can be assumed that Equation 5.6 results in a close approximation of k which indicates that, based on a purely mathematical estimation, high densities of 3O, explaining, among other reasons, the immediate high strength of agglomerates produced by these methods. Even though many studies have been and are being carried out to characterize packings derived from two or more sphere sizes, there is still no theory that satisfactorily describes the structure and allows an universally valid prediction of density or porosity, pore sizes and distribution as well as the coordination number of specific packings. For particles with irregular shape and a particle size distribution, the typical case in industry, a general understanding of structure and its characteristics is still remote. If


Fig. 5.40

Examples o f packings demonstrating the "wall effect"

[5.8].

packing parameters need to be known they must be determined experimentally. Nevertheless, some interesting information has evolved from the many tests that were carried out over time. It relates to so called optimum packings. The most important optimum packing is the densest regular packing of spherical particles. It can be most easily derived from the two loosest regular packings (a and c in Fig. 5.39) of the largest spherical particles in a mixture (Fig. 5.41). It is obtained by inscribing the largest possible spherical particle into the void between the four or three larger spheres (depending on the model used) and adding the appropriate amount of these smaller spherical particles to fill all the voids between the larger ones. This method is then continued as shown in Fig. 5.41.

I

83

84

I


--r-- I-----

I

Fig. 5.41 Sketches o f the systematic arrangements of differently sized spherical particles to obtain the densest possible packing.

Obviously, the resulting particle sizes and their relative masses do not fit a continuous distribution. The mixture consists of discrete classes of particles. In reality, where such densest packings are desired for a number of reasons, for example in the production of high quality concrete to obtain high strength and water impermeability by minimizing porosity and the presence of interconnected pores (see Section 5.3.2), narrow size ranges of aggregate particles are mixed in the appropriate amounts as


prescribed by the model. Although neither particle size nor shape of the components correspond with the assumptions of the model (spherical monosized particles in different classes) the real random packing produces the desired high density and strength. To obtain the best possible impermeability in concrete, the smallest particles that are added today feature particle sizes GO % of theoretical. Yet, by liquid phase sintering a perfect pore-free compact can be obtained. The detailed metallurgical description of the sintering operation is highly complex because of the complicated nature of the phase equilibria that are involved. Very simply explained, sintering begins by solid diffusion, frequently aided by a carburizing atmosphere, and results in the formation of a tungsten-cobalt-carbon eutectic with a relatively low melting point at suitable points of contact between particles. Once small amounts of liquid have formed, further liquefaction proceeds rapidly and surface tension forces cause rapid flow and wetting of all the other solid surfaces. This is accompanied by considerable contraction (shrinkage), often 40 % by volume. The whole process occurs with great speed at only a few degrees above the eutectic temperature. Very small quantities of liquid phase, approx. 1 % by volume, which is much less than one might assume necessary, are sufficient to result in rapid and near complete densification. The high packing density must be caused by a considerable rearrangement of the positions of carbide particles. However, because of the small volume of liquid that is involved it is not possible that this densification is obtained by particle rearrangement alone. Even with the best packing of the solid particles, the liquid phase would be insufficient to completely fill the interstices. Therefore, some change in particle shape must also occur and there must be considerable material transfer at the contact points to bring about this shape change. For this, the only possible phenomenon is the solution-reprecipitation process. Ceramic parts are often made from finely powdered components which are shaped by a pressure agglomeration technique and then sintered by the application of heat. In some cases this simple technique is not applicable because the powders may not sinter together, the sintering temperature or atmosphere requirements may be impractical or uneconomic, or a base material may not be readily available or may decompose under normal sintering compositions. In some of these cases, reaction sintering may offer the possibility of making sound ceramic bodies, often also with high density, which are impossible or at least difficult to produce by other methods [B.l2c]. The term reaction sintering is not very well defined. In the simplest case, the powder that decomposes during heating is substituted although, in the true sense of the word, this is not actually reaction sintering. In true reaction sintering processes, two (or more) components of the desired ceramic compound are selected such that they react with each other during sintering. In a first method, the powder components are mechanically mixed, shaped, and reaction-sintered. It is particularly suited for those materials which are solids at room temperature and feature relatively high melting points. Another process is applicable when one of the constituents is gaseous at room temperature (e.g. oxides, nitrides) or oflow melting point relative to the other components (e.g. phosphides, sulphides and some silicides or aluminides). Then the more refractory powder components are shaped and afterwards reacted hot with the other constituent in gaseous or liquid form. The preshaped body must be porous to allow for entry of the reactant(s) and for extra volume (if any) of the product of reaction. As mentioned before, sintering may be beneficially carried out under pressure (see also Section 8.4.4, HIP). During pressure sintering of a powder, an external pressure is applied and the initial stage of compaction, probably up to a relative density of approx.

9.2 Sintering Technologies I 3 8 9

85 %, includes the complex mechanisms of pressure agglomeration, i.e. particle packing, sliding, fragmentation, and deformation (see Section 8.1).The subsequent intermediate (featuring connected porosity) and final (with closed porosity) stages both involve a solid matrix and a definite pore system. A theory of pressure sintering [B.l2c]is valuable because it allows to extrapolate experimental data for a given material, for predicting performance under changed conditions, and also allows the calculation of viscosity or diffusion data, thus affording a means of assessing possible results when changing the composition of the material. In pressure assisted sintering, the more accurate name for pressure sintering, pressure and heat are applied simultaneously to a powder that is enclosed in a die. Generally, it permits the use of lower temperatures and pressures and shorter processing times than those required for cold pressing and subsequent sintering. It can also assist in the production of parts with finer grain size, lower porosity, and higher purity. As mentioned in Chapter 9, sintering is a binding mechanism and the different technologies may be used for size enlargement by agglomeration and as methods for the creation of final characteristics of various products. Depending on their applications, the required properties of pieces or parts after the sintering process may vary widely. Correspondingly, different methods for powder preparation, the manufacturing of preforms, and the application of heat must be chosen. For example, iron ore pellets, which must feature high strength to guarantee the excellent transportation and handling characteristics required for a bulk commodity, uniform size and shape for good and reproducible packing in shaft and blast furnaces, and a large percentage of open porosity for optimum reducibility, are made by tumblelgrowth agglomeration in discs and drums (see Section 7.4.1), dried and sintered to produce necks between the ore particles but retain the high porosity of the agglomerates. Powder metal or most ceramic parts, on the other hand are formed into dense compacts by pressure agglomeration (see Sections 8.4.1 through 8.4.4) and then sintered, possibly with the assistance of pressure (HIP, see Section 8.4.4) to yield well (sometimes near net) shaped parts with nearly theoretical density and high strength. Other powder metal or ceramic parts may have to become filters or catalyst carriers requiring large numbers of penetrating pores (see Section 5.3.2), uniform structure, and high strength. In those cases sintering of preforms is carried out such that no densification occurs and porosity remains unchanged.

9.2

Sintering Technologies

To remain within the scope of this book, the description of industrial agglomeration processes, only those sintering technologies will be reviewed which are used either directly for the size enlargement of particulate solids or for the post-treatment of agglomerates to gain final product properties. The many highly sophisticated sintering techniques that have been developed during the past few decades for the production of new, engineered, often composite materials with novel characteristics, will be covered in a future book [B.71].

390

I

9 Agglomeration by HeatlSintering

9.2.1

Batch Sintering

Batch sintering in stationary furnaces of different kinds is used primarily for the posttreatment of agglomerated products to gain final strength and properties but also for laboratory work in connection with the development of continuous sintering technologies and for pilot plants [B.12c, B.16, B.251. The two most important applications of batch sintering are in the ceramic and powder metal industries. Muffle, bell, elevator, or pit type furnaces are used. As shown in Tab. 9.1, the processes occurring during sintering are somewhat different for ceramic and powder metal parts. During sintering, ceramic ware is heated to a temperature between 700 and 2,00OCC.Because raw ceramic parts are almost always “green” (= moist), removal of water is the first post-treatment step. Following or simultaneously and immediately preceeding firing, binders and plasticizers, which have provided the properties needed for forming, are also removed. The amount of residual moisture and/or additives that can be tolerated in the part when firing begins depends on its shape and structure as well as on the heating rate of the furnace. If the part is made by dry pressing, removal of additives can be incorporated into the heat-up stage of the sintering furnace if the time for this process is not too long. However, since additives are often cellulose, wax, or starch type products, they can be conveniently decomposed by oxidation in air at low temperature prior to loading the parts into the furnace. In the furnace, clay minerals usually dehydroxylize between 500 and 600 ”Cwhereby steam is produced. The loss of strength that occurs at this stage may result in cracking. Often, carbon and sulphur compounds are present in unfired ceramics. They must be oxidized before sintering densification has advanced too far to avoid black cores. Oxidation can be accomplished by holding the temperature at a certain level which varies with the manufacturing method that was used for and the type of the body but is often in the range of 300°C. A decomposition of carbonates and sulphates may produce bloating in vitrified parts. Silica, which exists in many different crystalline forms, is an important constituent of most ceramics. The conversion from one form into another is accompanied by sometimes large volume changes. Because this occurs during heating and cooling, the rate of temperature change must be considered and may have to be adjusted to avoid deformation and/or cracking. Processes occurring during the sintering of ceramic and powder metal parts [B.l2c].

Tab. 9.1:

Ceramic Parts

Powder Metal Parts

Removal of water Removal of binder and organic media Dehydroxylation Oxidation Decomposition Phase transformation Cooling

nia Burn-off of pressing lubricant n/a Heating to sintering temperature Soaking nia Cooling


Fig. 9.3: Photographs ofsimple atmospheric muffle furnaces (courtesy Gasbarre, Sinterite Furnace Div., St. Marys, PA, USA).

Most ceramic products are fired in air, i.e. under oxidizing conditions. The ideal kiln for the firing of ceramics is capable of heating and cooling the parts uniformly at the maximum rate of temperature change for each of the stages mentioned in Tab. 9.1. Simple muffle furnaces are typically used for batch sintering in the ceramics industry (Fig. 9.3). For high quality wares, temperature control is very important to avoid the previously mentioned potential problems in different processing stages. It can be accomplished easiest and most accurately in batch furnaces although many bulk ceramic products must be of such low cost that continuous furnaces are used which operate more economically (see Section 9.2.2). Some materials must be, at least during certain stages, fired in reducing atmosphere which can be also easily provided in batch kilns. In powder metallurgy, sintering requirements are different. Although the volatilization of pressing lubricant from the compact prior to sintering is sometimes carried out separately, it is more typically an integral part of the process. During sintering itself, temperature is held constant so that no distortion takes place and full bonding is obtained. Therefore, temperature control and soaking periods are most important. Additionally, it is necessary to retain the composition of the atmosphere to ensure reproducibility of strength, carbon content, dimensional stability, etc. of the final part. Therefore, ingress of air into the furnace during sintering must be avoided. This is achieved by using either a gas tight furnace shell or a muffle or retort which is usually manufactured from a nickel-chromium alloy. Batch sintering furnaces are employed for: 1. Low output production, 2. special duties, (because there are no moving parts, batch furnaces can be designed for higher temperatures; furthermore, since it is possible to seal the interior more effectively, purer atmospheres can be realized and maintained) and 3. experimental work.

For (1)and/or (2), a small manual pusher furnace can be applied in which parts on a tray are moved through a furnace, one tray at a time (Fig. 9.4). If it features gas tight

392

I


UNLOAD

COOL

I

l

l

I

I EXIT DOOR

I

I

I I

/I

i/(

TRAVEL

ELEMENTS

CURTAIN

Fig. 9.4 Sketch o f a longitudinal section-box or manual pusher furnace [B.25].

interlocks and/or doors for charging and discharging, it can also be used for sintering processes in which the atmosphere must be well controlled. The bell type furnace (Fig. 9.5) is widely used for P/M parts requiring long sintering cycles. Typical equipment consists of one or more supporting bases with removable sealed retorts, to cover the loads and to retain the protective atmosphere around them throughout the entire heating and cooling cycles, a portable heating bell, and a standby (idling) base, a hoist, and an optional (not shown) cooling bell. The elevator type furnace (Fig. 9.6) is useful for sintering heavy and/or bulky loads. It has an elevated heating chamber with open bottom in a fixed position, a mechanism for raising and lowering load supporting cars into and out of the furnace, a stand-by car to plug the kiln opening during idling periods, and optional cooling chambers. It is also applicable if protective atmospheres of exceptionally high purity are required. Flexible hoses carry atmosphere gas and cooling water to and from the cars.

IDLING BASE

HEAT ING BELL ON RETORT

RETORT ON LOAD

Fig. 9.5: Schematic representation of a bell furnace [8.25]

LOAD ON BASE

9.2 Sintering Technologies

I

I 1

DUCT

ON 4 SIDES-

LOAD- SUPPORTING CAR

+

Fig. 9.6: Schematic representation o f an elevator furnace [B.25].

Batch kilns can be operated under vacuum and as direct-resistance furnaces for the sintering of refractory metals (e.g. for tungsten at 3,000 "C). For hardmetal processing, lower temperatures are used. Fig. 9.7 shows typical pressure and temperature profiles. A complete sintering cycle may take from G to 14 h or longer for thick parts. Therefore, such furnaces are often connected in pairs to a common vacuum pumping system and power supply as well as a single set of controls because approx. 50 % of the cycle is required for heating under vacuum and the balance for cooling in inert gas. During various phases of the heating cycle, inert or active gases may be injected into the vacuum system, thereby changing the partial pressure in the sintering chamber, to

4-

!

700 i-

500

T i m e (Hours) Fig. 9.7: Typical pressure and temperature profiles of a vacuum sintering cycle in a batch furnace [B.25].

393

394

I


Pot grate Insulating material : fired pellets

3

Pot grate Insulating material : retroctory lining

/-

1.) Green pellets

1.) Green pellets

2 . ) hsulating side walls : pellet fragments 3.) Grate bars :cast alloys 4.) Insulating hearth layer

2 . ) hsulating side w o k : bricks or rammed mass

3.) Grate bars: usually silicium carbide

Fig. 9.8: Two-pot grate furnaces for the heat treatment of (iron ore) pellets [B.l6].

achieve special effects which influence the structure and properties of the sintered part. Since batch furnaces are normally relatively small and can be controlled easily, they are also commonly used for development work. For the sintering of ceramic or powder metal parts the results from small scale testing can be directly transferred to larger or continuously operating kilns. As will be shown in Section 9.2.3, large continuous sintering facilities are used in the minerals industry for the size enlargement of fine ores and the induration of “green balls”, spherical agglomerates made by tumble/growth agglomeration from fine ore concentrates (see Section 7.4.1).Since, in the development phase, the heating and gas flow conditions in industrial plants must be simulated, batch pot grate sintering equipment (Fig. 9.8) is being used [B.lG]. Representative tests for determining the performance of travelling grate, grate-kiln, and shaft furnace processes (see Section 9.2.3) can be carried out in these laboratory and pilot facilities. According to the actual conditions in industry, pot grates are operated either with insulating side walls of fired pellet fragments and a hearth layer (Fig. 9.8, left) or are equipped with a corrugated refractory brick lining (Fig. 9.8, right). In the first case, the pot grate itself is a metal container with a bottom of metallic grate bars. By placing indurated pellets between the hot wall and, respectively,the grate and the green pellets which are to be sintered, the charge is protected by a refractory envelope and overheating of the metallic parts is avoided. Other pot grates feature a refractory lining and a high temperature resistant grate (Fig. 9.8, right). To eliminate the wall effect


in these relatively small furnaces and to be able to test samples of different ores during a single test run, stainless steel baskets, which are filled with the appropriate ores or pellets, are embedded in the charge. After the test, they are retrieved and the characteristics of the fired materials are determined. During a pot grate test the following parameters are determined and may be varied: Direction of the air flow, gas volume, suction, and pressure in the wind box, preheating rate and temperature profile, fuel type (gas or oil), gas atmosphere (using additional oxygen, if necessary). Fig. 9.9 shows the operation of a pot grate and the locations of thermoelements. In this case, drying is carried out first with updraft warm air (flowing up through the pellet bed), followed by downdraft sintering with hot air from the burner above the bed, and, finally, cooling is accomplished by an updraft flow of ambient air. It is also possible to design the system such that during sintering separate up- and downdraft stages can be used. During the entire process, the temperatures are continuously monitored as they are decisive for the quality of the fired product. In a pot grate as shown in Fig. 9.9, the different process stages occur intermittently, one after the other. When the flow of gas is reversed, the temperature of the peripheral

Burnet hood mvable

TZ M i l e ot pellet bed

16 In wirdbox Wasle gas M Drying gas or Cwling oir

Fig. 9.9 Operation and temperature control of a pot grate [B.16].

I

395

396

I


V - 1

Fig. 9.10:

iv

U

n

I

Pol grote positions _______(

Movable pot grate system for sintering tests lB.16)

parts, which are associated with the pot grate itself, must first change to the new condition which can critically modify and impair the test conditions. To avoid this, a movable pot grate system as shown in Fig. 9.10 was developed [B.lG] in which all equipment for the different steps is kept at process conditions and the pot grate is moved as necessary into the various positions. It should be mentioned that pot grate sintering machines may be also used for the industrial sintering of small amounts of metal ores. Fig. 9.11 shows the flow diagram of a pan sintering plant [B.23]. In such sintering systems, heat is provided by the burning of carbonaceous solid fuels that were mixed with and are uniformly distributed in the charge. After ignition of the bed surface, for example by depositing red-hot CYCLONE SEPARATOR

RETURN S I N T E R C O K E ORE

WET SCRU 8 B € A

SlNTERlNG

Fig. 9.11:

Flow diagram of a pan sintering plant p . 2 3 , p. 2121.


charcoal or coke, oxygen from the air, which is pulled through by a suction fan, produces the intense heat that is necessary for sintering. During sintering, the entire charge becomes one large cake which, after cooling, is broken and screened into the desired sinter size. Fines are recirculated and blended with fresh fine ore and fuel. Such plants typically produce 30-50 t/day of sized sinter for use in reduction furnaces. 9.2.2 Continuous Sintering

As mentioned in Section 9.2.1, sintering of ceramic wares normally occurs in oxidizing atmosphere and without a special gas environment. Therefore, continuous sintering furnaces are often directly flame heated. Fig. 9.12 is the side elevation of a tunnel furnace for the firing of ceramic parts, indicating the direction of material (car)and gas movement as well as the process zones. The diagram below depicts the temperature profile over the length of the furnace and shows that temperature control is normally quite unpretentious. Most tunnel kilns for ceramics are of the car type. Cars are more rugged and reliable than belts and other continuous methods of movement. Fig. 9.13 is a schematic cross section through the sintering zone of a directly fired, atmospheric tunnel furnace. The tunnel is enclosed by refractory walls and a simple sand seal prohibits the exit of hot combustion gases at the car base and wheels. Operation of modern furnaces is computer controlled and continuous movement is accomplished with automatic loading and unloading systems. Fig. 9.14 is the partial view of a state-of-the-art tunnel kiln for the firing of table ware and also shows an automated car handling system. Direction of cars

aL

Direction of goses *

W Reheat zone--

f,

h

h6t:*Y+

*

Y

Firing z m e

0. 4

A

--

4

I200 r

3L

Air blower

I

s-0

300

200 I00

Fig. 9.12: Side elevation (schematic) o f a directly flame heated tunnel kiln for the firing o f ceramic parts and typical temperature profile [B.lZc].

W

Cooling z m -

ia

I

397

398

I


wheel

base

Fig. 9.13: Cross section through the sintering zone of a tunnel kiln [B.l2c].

In comparison, continuous furnaces for the sintering of P/M parts are more complicated because of the needs for a more differentiated temperature profile (Fig. 9.15) as well as for controlled gas environments (see Section 9.2.1). The latter requires some sort of separation of the atmospheres in different sections along the kiln, either by oscillating doors or by gas curtains.

Fig. 9.14

Photograph o f a modern tunnel kiln for the tiring o f table ware with a fully automated car handling system (courtesy Eisenmann, Boblingen, Germany).


+ BURNOFF -+-SINTER -+-? SLOW 1 HEAT 1 SOAK I HEAT I SOAK ]COOL(

TI ME Fig. 9.15: Typical temperature profile o f a continuous furnace for the sintering o f powder metallurgical parts [B.25].

Fig. 9.17: Schematic and photograph o f a horizontal mesh-belt sintering furnace including an optional "accelerated delube system'' (ADS) (courtesy Gasbarre, Sinterite Furnace Div., St. Marys, PA, USA).

COOL

-- 7

!

I

399

400

I


The most commonly used conveyor in continuous furnaces for the sintering of small, light parts is the mesh-belt. Fig. 9.16 is a schematic longitudinal section through a mesh-belt sintering furnace and includes an indication of the different zones. The doors, which can be manually operated, are usually left open and the atmospheric conditions within the furnace are created by and between the gas inlets. This puts a certain strain on the atmosphere generating equipment as ample capacity must be available. Fig. 9.17 shows both the schematic and a photograph of a horizontal mesh-belt sintering furnace which includes an optional "accelerated delube system" (ADS) for the rapid removal of powder metal lubricants. A variation ofthe straight mesh-belt furnace is the humpback kiln (Fig. 9.18),which is used where high purity of the atmosphere in the sintering zone is required. The belt in a long, inclined, gas tight purge chamber carries the work from the charge area up to the elevated hot zone. This design is particularly advantageous iflight gases are used in the sintering zone because these gases tend to naturally rise to the highest point of the furnace. Fig. 9.19 is a schematic representation of a roller hearth furnace. Loaded trays are conveyed through the kiln by riding on individually driven rolls (Fig. 9.20). Depending on roll spacing, a section is capable of holding a substantially greater load than an equivalent length of mesh-belt. The grade of alloy used for the rolls limits furnace temperature to between 1,150 and 1,260 "C. The charge and discharge doors are automatically opened or closed and are interlocked with the tray handling system. Because they are only opened when a tray is charged or discharged, the amount of atmosphere gas is optimized and heat losses are minimized. HEATING CHAMBER

COOLING CHAMBER EXIT INCLINE

EN1. R A K E INCLINE

MUFFLE ~. .............

WIRE MESH BELT

Fig. 9.18:

Fig. 9.19

Schematic representation o f a hump-back furnace [B.25]

Schematic longitudinal section through a continuous roller hearth sintering furnace [B.25].

IDLING

PULLEY-J


Fig. 9.20: Roller drive system using cogged belts which offer durability, low maintenance, and quiet operation Roller hearth sintering furnaces with temperatures of up to 1,450 ‘ C can be equipped with this design (courtesy Eisenmann, Boblingen, Germany)

MECHANICAL PUSHER

PURGE 8 PRCHEAT CMPIYDER

k

HIGH H E A T CHAMBER

WATERCOOLEOCHbYBER

UNLOADING P L A T F O R Y

m a w ) . . r i ~ e sY I O Y H e i r

Fig. 9.21: Longitudinal section through a continuous pusher furnace lB.251.

The pusher type furnace that was shown schematically as manually operated equipment in Fig. 9.4 (Section 9.2.1) can be mechanized and then becomes a continuous kiln. Fig. 9.21 is the longitudinal section through a continuous mechanical pusher furnace. It is suitable for sintering metal parts which are too heavy for the meshbelt and production rates do not warrant the roller hearth. It can also be used for sintering temperatures of up to 1,GSO”C which are too high for the mesh-belt and the roller hearth furnaces. Mechanically or hydraulically operated, intermittent or continuous, stoker type pushers are available. Fig. 9.22 is the photograph of a pusher furnace for the high temperature sintering of powder metal compacts. The final typical design of continuous sintering furnaces for powder metallurgy and similar applications is the walking beam furnace (Fig. 9.23). With this furnace the weight of the work that can be conveyed safely is practically unlimited and the m a imum continuous operating temperature is only limited by the refractory material used to line the hot zone chamber and by the compatibility of the atmosphere with the heating element. The sintering temperature may be as high as 1,800”C. Fig. 9.24 is a schematic cross section through the hot zone of a walking beam furnace. A comparison with Fig. 9.13, above, shows the hermetically closed furnace housing which is typical for all metal sintering furnaces and the vertical (left) or horizontal (right) noncontaminating electrical resistance heating elements. Both are required to

I

401

402

I


Fig. 9.22: Photograph o f a high temperature pusher furnace (courtesy Casbarre, Sinterite Furnace Div., St. Marys, PA, USA).

I

W A L K I N G BEAM'

HORIZONTAL MOTION CYLINDER

8 Fig. 9.23: Longitudinal section through a walking beam sintering furnace [B.25].

maintain a particular composition of the atmosphere in the kiln. Movement of the charge in a walking beam furnace is accomplished by a mechanism that lifts and pushes forward, by only a few centimeters each, a bottom tray with the beams below (see hydraulic cylinders (A) and (B) in Fig. 9.23). The two, timed displacements produce a rectangular motion which conveys the work through the furnace at the required speed. The use of sintering for the induration of ceramic products (bricks, pots, vases, etc., see also Chapters 2 and 3) is quite old but through the centuries, even though empirically improved, was exclusively carried out in batch kilns. Continuous heat treatment of ceramic, powder metal, and other pre-agglomerated parts is less than 150 years old. Another original, also quite old application of sintering is found in the minerals industry that is associated with technologies for the production of metals. There, agglomeration by heat was introduced many centuries ago for the size enlargement of fine ores. Primitive versions of the pan sintering process (see Section 9.2.1, Fig. 9.11) were used to produce sinter from fine ores which had been mixed with a particulate solid fuel. The necessary heat was produced by blowing air through a bed of ore particles with bellows and burning charcoal that was uniformly distributed within (Fig. 9.25). Continuous sinter plants for ores were developed at the beginning of the 21st century for the size enlargement of fine ores, flue dust, mill scale and other fine metal bearing materials [B.42].At the beginning, metal cars with perforated or slotted bottom


IGHT S H E L L

-

Y

Fig. 9.24: Sketch showing a cross section through the hot zone o f a walking beam furnace (left: vertical heating element; right: horizontal heating element) [B.25].

were pushed through a directly fired tunnel furnace. As discussed in Section 9.2.1, Fig. 9.8, left, screened, fired fines were placed between the metal walls and bottom and the feed containing the solid fuel to avoid overheating of the mechanical parts. Later, the carts were connected to form a continuous grate belt which was moved continuously through a tunnel furnace by a motorized drive. Fig. 9.26 shows schematically the operating principle of such a travelling grate sintering machine. First, recirculated, fired (= sintered) fines from the sinter screens are deposited on the grate as a hearth layer. Then, feed, consisting of a blend of fine ore

404

I


Fig. 9.25: Georgius Agricola's (De Re Metallica, 1556) representation o f a small ore sintering furnace.

and solid fuel, is placed onto the insulating hearth layer with a swinging conveyor (for uniform bed depth across the wide grate carts) and leveled with a roll feeder. Next, the charge thus built passes through an ignition furnace which is a short hood with burners inside. The flames impinge the surface of the bed and ignite solid fuel that is close to the hot interface. Beginning at this point and after leaving the hood, air is pulled through the completely open bed in a downdraft fashion; combustion of the solid fuel is sustained and enhanced by providing excess oxygen and the burning hot zone travels downward through the bed. The amount of air which is pulled through the entire length of the bed and the speed of the grate are adjusted such that, at the end of the machine, the fuel has disappeared and the entire charge has sintered together. The porous solidified mass is broken in a (hot) sinter breaker before the pieces are screened into the desired particle size distribution and externally cooled. Fines are recirculated to provide the hearth layer and potentially oversized pieces are recrushed in closed loop with a screen. Solid particles that are entrained in the combustion air, settle in bins which are part of the main suction duct and fine dust is removed in a dust collector. These solids are recirculated to the burden preparation plant and ultimately fed back to the sintering machine.

Feed i f r a m burden preparot!onl

Swinging conveyor

Roll f e e d e r sin:^: b r e a k e r

c o i l e c t o r ond

Counterweighted dust valves

Fig. 9.26

Schematic of an early travelling grate sintering machine.

'


Although, after the development of pelletizing (see below), sinter has lost some ofits importance as a sized feed for reduction furnaces, the technology is still used, particularly in the iron and steel industry, and occasionally new sinter plants are installed. However, overall, since the 19GOs, worldwide production of sinter is decreasing in favor of pelletizing (Fig. 9.27).Nevertheless, even today reports dealing with improved sinter machine designs are published [B.48].While the basic process is still the same, new developments are directed towards better burden preparation, particularly in connection with unusual ore compositions, improved temperature control, minimization of pollution, general process optimization, and reduced energy consumption (Fig. 9.28). During the middle of the 20th century in several locations, particularly in the USA, development work started to render the large reserves of Taconite and Itabirite, low grade iron ores, useable for iron and steel making. In the early 1960s iron ore pelletizing was developed. The iron ore concentrates which, after upgrading, have a particle size xmin, curve (d). If agglomeration occurs, the finest particles may form larger entities or attach themselves to larger particles, thereby changing the separation curve in Fig. 2.1 to (e). The new cut size xt2’ is still somewhat larger than the desired xmin, at which all particles would have been removed from the fluid, but agglomeration definitely helps to move the actual cut size closer to the ideal one. Agglomeration in Industry. Wolfgang Pietsch Copyright ª 2005 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN 3-527-30582-3

24

4 Undesired Agglomeration: Methods of Avoiding or Lessening its Effect Tab. 4.1 The occurrence of undesired and desired agglomeration in mechanical and related process technologies Unit operation

Process

Separation

Mixing

Comminution Particle size enlargement

Conveying

Storage Batching, metering Drying Explanations

Agglomeration

Screening, Sieving Classifying, Sorting Flotation Dust Precipitation Clarification, Thickening Particle Size Analysis Dry Mixing Wet Mixing Stirring Suspending, Dispersing Fluidized Bed Dry Grinding Wet Grinding Agglomeration } Briquetting } Granulating } Pelletizing } Pelleting } Sintering Tabletting Mechanical Conveying Vibratory Conveying Pneumatic Conveying Silos, Hoppers, Stockpiles

(+) sometimes YES ( ) sometimes NO

+ YES NO

Undesirable

Desirable

+ + + (-) (-) ++ + + + + + + +

(+) (+) + + -(+) + (+) + -

(+)

++

+ + + + + +

++ ++ (+) (+) --+

++ decisively YES - - decisively NO

In general, whenever the task is to remove all particulate solids from a fluid, agglomeration will be advantageous. Since the smallest, low-mass particles are, on one hand, the most difficult to remove and, on the other hand, have the highest natural adhesion tendency, the chance agglomeration of these particles improves separation efficiency. Therefore, techniques for enhancing the natural agglomeration tendency of very fine particles are often applied during gas and liquid cleaning. Even relatively loose conglomerates behave according to the combined weight of all adhering particles during, for example, settling or in the centrifugal fields of cyclone separators. For all those separation cases that attempt to separate a particle collective according to certain properties of the particulate solids, agglomeration is most often undesired. Techniques for which this statement is true include screening, sifting, classification, sorting, flotation, and, as a general analytical method, particle size characterization. It should also be realized that the respective separation property does not have to be size; it could be density, shape, color, chemical composition, and others.

4.1 Separation Fig. 4.1 curves

Examples of separation

During screening, unwanted agglomeration is often facilitated by the motion of the material on the screen; spheroidal agglomerates are frequently formed from material containing fines or featuring other binding characteristics, for example if it is moist. Among others, binding mechanisms are: for finely divided solids, molecular and electrical forces and/or adsorption layers; for plastics, electrostatic forces; for ores, magnetic forces; for moist powders, liquid bridges and capillary forces; for fibers, interlocking; and for materials with low melting points, partial melting and solidification. With some substances several bonding mechanisms may occur simultaneously. In all cases, the result of screening is distorted because agglomerated fines are classified as coarse. The immediate and complete removal by dedusting or “scalping” of the finest fraction prior to screening into the desired property classes is one practical method of avoiding selective agglomeration of the fines or adhesion of fines to larger particles. During screening itself, the effect of adhesion is reduced by mechanical destruction of agglomerates with, for example, rubber cubes or balls placed on or under the screen decks, the application of brushes, air jets, or ultrasound, and the modification of screen amplitude or frequency (e.g., ultrasonic screen excitation). During the screening of moist bulk materials, difficulties increase with moisture content, but agglomeration tendencies are almost completely eliminated during wet screening when the particles are suspended in a liquid. Since, in moist screening, particles are often held in the mesh openings by liquid bridges, the separation of such materials is facilitated by direct resistance heating, inductive heating, or by modifying the wetting angle and/or the surface tension with surfactants, and blinding of the screen is avoided. In air classification, products from, for example, dry fine grinding are separated. Particular problems arise if the material to be separated contains agglomerates that were formed during comminution. Such conglomerates would be recirculated into the mill and “overgrinding” occurs. Therefore, attempts are made to destroy them by special feeder designs. Destructive forces are caused by, for example, sudden

25

26

4 Undesired Agglomeration: Methods of Avoiding or Lessening its Effect

changes in speed or direction of flow and by installing air jet mills in front of the classifier. In the classifier itself, agglomerates are formed by molecular forces that may be reinforced by adsorption layers if separation is carried out with ambient air, by liquid bridges if moist materials are processed, and by electrostatic forces in a dry environment. As an example, Fig. 4.2 depicts separation curves of various air classifiers. With decreasing particle size, due to agglomeration, the amount found in the coarse fraction increases, because fine particles adhere to larger ones and conglomerates of fines behave as if they were coarser particles. Both effects reduce the separation efficiency and can be avoided only if the causes for adhesion are removed, mostly by eliminating moisture in the material and humidity in the air. Sorting processes that separate materials according to particle characteristics other than size are often carried out in liquids. Agglomeration can again reduce the separation efficiency when smaller particles of other components stick to larger ones. On the other hand, agglomeration can be also beneficial in dense media sink/swim separation, centrifugal separation, or during jigging if particles of a particular ingredient can be made to selectively adhere to each other and form larger, heavier agglomerates. During particle size analysis, in addition to screening, air sifting, and counting, sedimentation in liquids is often applied, which produces unequivocal results only if the individual particles can move without influencing each other. For that reason, very dilute suspensions are used. Nevertheless, it is possible that agglomerates form or conglomerates that are already present do not disintegrate completely. Therefore, dispersion aids are often added, which reduce particle affinity. The molecules of dispersion aids attach to the particles, eliminating polarities and/or reducing interfacial tensions. Separation forces, such as ultrasonic vibrations, can be also introduced for improved disagglomeration and dispersion. In connection with particle size analysis, the importance of correct sample preparation must be stressed. Because natural, unintentionally formed agglomerates always incorporate a larger than average number of the finest particles, the result of particle size analysis will be incorrect if pre-existing agglomerates are not destroyed or conditions prevail during testing that promote agglomeration.

Fig. 4.2 Separation curves of different air classifiers [B.71, B.97]

4.3 Comminution

4.2

Mixing

Many of the previously mentioned considerations apply to the formation and prevention of undesired agglomerates during mixing. Little needs to be added concerning mixing in liquids by stirring or methods for the production of suspensions and dispersions. The addition of dispersion agents is always recommended if the tendency of the solid particles to agglomerate is high. Agglomerates or flocs that are already intentionally, for example to improve handling characteristics of fine powders, or unintentionally present prior to mixing can be destroyed by shear forces in the liquid. Consequently, the generation of the highest possible shear gradient is often considered advantageous when selecting agitators. During extended storage, the particles in pharmaceutical suspensions often form agglomerates that can no longer be destroyed by shaking the preparation. This is of particular concern in, for example, eye drops. The problem can be avoided by controlled flocculation of the solids. After the addition of an electrolyte, the fine particles aggregate to loose flocs that can be easily redispersed by shaking the dispenser prior to application. When mixing dry or moist bulk solids, agglomerates may form, which originate from the finest components of the mixture. They are held together by molecular and electrostatic as well as capillary forces. These undesired agglomerates should be broken up by shear or frictional stresses, generated by the motion of the bulk mass, or by special disintegration devices that are built into the blender [B.97].

4.3

Comminution

During fine grinding in roller crushers and tube mills containing grinding media, with all materials problems begin to develop at a certain fineness of the solids to be milled. Two types of phenomena can be distinguished. In the first case, the finest particles start to adhere to walls and the grinding media in the mill. On this first coating, even coarser particles find excellent conditions for adhesion and massive deposits form rapidly. These layers adhering to the inside walls and the grinding media produce a cushioning effect, which lowers the intensity of stressing and, therefore, increase the duration of grinding and decrease efficiency. The second phenomenon during dry fine grinding is the occurrence of agglomerates in the freely moving charge itself. Formation of such agglomerates is associated with the so called “limit of grinding”. For each material a fineness exists at which, in spite of continued consumption of energy, the finest particles in the charge do not seem to become finer. Agglomeration and adhesion in mills can be attributed to various binding mechanisms. Since the mill housing may become highly charged by the friction between its contents and the walls, electric forces are often the cause for the initial build-up. This effect can be eliminated quite easily by grounding the mill. In other cases, wall deposits

27

28


will begin with particles of a size that generally corresponds to that of the wall roughness. The strength of the layer depends on the contact pressure, which is magnified by the mill charge consisting of grinding media and material to be crushed. Adhesion is largely affected by molecular forces; however, partial melting and sintering are also possible. Agglomerates are formed in the freely moving charge of a tube mill by the compaction of fine particles between the grinding media and by recombination bonding (Chapter 5). Adhesion is affected by van-der-Waals forces between the particles that have been compressed very tightly and by the recombination of free valence forces at newly created surfaces. Since these agglomerates are very strong, a “grinding equilibrium” exists, which has been observed and described by many. It means that in fine grinding, after a certain time, a state of equilibrium between size reduction and size enlargement by agglomeration occurs. From that point on, agglomerates are formed, crushed, and re-formed so that the apparent particle size does not change. However, since destruction of particles continues to occur, a growing amorphization of the material can be observed, which also results in increasing specific surface area and is often called “mechanical activation” as, in many cases, higher reactivity is obtained. Every form of agglomeration during size reduction reduces the efficiency of grinding and the fineness obtained at the limit of grinding is often insufficient for many

Fig. 4.3 Agglomerates produced during grinding in a roller mill with a high reduction ratio

4.3 Comminution

tasks, even though the agglomerates actually contain much smaller particles. Therefore, it is desirable to prevent or, at least, reduce these effects. In milling, one possibility to achieve less unwanted agglomeration is to add surface-active substances, so called surfactants or “grinding aids”. It has long been known that small amounts of such additives may reduce the grinding time required to reach a particular fineness by 20–30 %. Molecules of these substances, which are present in a gas or vapor phase, quickly saturate free valences at the newly created surfaces and avoid recombination bonding. As a rule, grinding aids also reduce caking. The formation of lamellar flakes or flat agglomerates in tube mills has been attributed to compaction between the grinding media during impact. The same mechanism occurs in all comminution processes in which the material to be crushed is subjected to stresses between two surfaces, for example in roller mills. Since the second condition for the formation of agglomerates is a sufficient fineness of the particles that are involved, the occurrence of flat flakes is mostly observed during fine grinding. One measure for the fineness, the intensity of stressing, and the unintentional formation of agglomerates is the so called “reduction ratio” that is, the quotient of maximum feed particle size and gap between the rollers. Fig. 4.3 depicts typical agglomerates produced in a roller mill with a high reduction ratio. Since the fine material is immediately compacted, almost all free valences at the newly created surfaces participate in recombination bonding. Consequently, the formation of agglomerates in roller mills can be avoided only if a smaller reduction ratio is chosen or by applying friction between the rollers. Agglomerates can be also formed during impact grinding. Fig. 4.4a shows the fracture lines that develop during impact stressing of a glass sphere. A cone of fine material is created at the impact point and is compacted into an agglomerated mass by the pressure resulting from the kinetic energy of the system (Fig. 4.4b and c). Here too, the effect of free valence forces on newly created surfaces is used to its almost complete extent yielding a quite strong agglomerate. It is very difficult to avoid this type of agglomeration; it can be affected only by reducing the impact velocity which, in turn, results in a lower degree of comminution. For glass spheres, for example, the formation of agglomerates was observed only at impact speeds exceeding 80 m/s.

Fig. 4.4 a) Schematic representation of the fracture lines caused in a glass sphere by impact stress. b) Agglomerated cone of fines created during the impact stressing of a glass sphere (sphere diameter 8 mm, impact velocity 150 m/s). c) Agglomerated cone of fines created during impact stressing of a sugar crystal (shown on the left)

29

30


During the impact crushing of thermoplastic materials or inorganic substances with low melting points, solidified bridges of partially melted material may further increase adhesion and the strength of agglomerated fines. Since the increase in temperature depends on the energy input and is constant for a given impact velocity, cryogenic milling, whereby the particles to be crushed are cooled prior to feeding into the mill, not only results in an increased brittleness but may also avoid partial melting and unwanted agglomeration. In wet grinding, as a rule, agglomeration is avoided by suspending the particles in liquid. Sometimes, the product of dry fine grinding is subjected to a brief final wet grinding to destroy the previously formed agglomerates. Nevertheless, some materials also tend to flocculate in wet grinding. Since the adhesion forces causing flocs are mostly electrical in nature, the addition of a small amount of electrolyte to the suspending liquid nearly always suffices to prevent flocculation.

4.4

Agglomeration

By definition, the size enlargement by agglomeration operation makes use of all binding mechanisms and often enhances them in suitable environments and equipment. All agglomerates that are manufactured are made intentionally and are desired. Nevertheless, there are instances where adhesion and agglomeration are unwanted and undesired. Because, particularly in tumble/growth agglomeration, binders are added, the effect of these binders is still present on the surface of green agglomerates and during post-treatment new binding mechanisms between the agglomerates may develop, which result in the formation of clusters. Of course, because agglomerates are larger bodies and only a few interaction points (coordination points) are present in a unit volume of the cluster, even relatively strong solid bridges, which may have developed by recrystallization, chemical reaction, or sintering during post-treatment can be broken relatively easily. Nevertheless, such clusters could be detrimental during storage, feeding, or metering and, therefore, should be avoided or broken up prior to a following process step. More information on potential problems is given in the section 4.6 on storage.

4.5

Transportation

During the conveying of particulate solids, especially of finely dispersed powders, the unintentional formation of agglomerates and (sometimes thick) coatings on walls is often observed. Whereas agglomerates occur mostly on vibrating or shaking conveyors and inclined conveyors or chutes, build-up on walls is more common in pneumatic conveyors. The main causes of agglomeration during the conveying of fine particulate solids are molecular and electric forces as well as binding mechanisms related

4.6 Storage

to moisture and, as a result of mechanical or thermal energy input, binding mechanisms such as partial melting and solidification can occur, too. Although it is very difficult to avoid agglomeration on vibrating or shaking conveyors, several possibilities exist for the prevention of wall build-up and deposits during pneumatic transport. Since the adhesion of the finest particles always begins in the roughness depressions, one of the most important conditions for avoiding a common reason for initial build-up is to provide smooth inner wall surfaces of pneumatic conveying lines. Because high drag forces tend to remove particles that have already adhered to the walls, high transport velocities also reduce the danger of build-up. Deposits preferably start in dead or calm flow zones; therefore, when designing such systems, low speed areas must be avoided. On the other hand, sudden changes in the direction of flow will cause high energy impacts of particles with the wall, causing build-up. Friction between particulate solids and pneumatic conveyor walls may result in high electrostatic charges on both partners. They depend to a large extent on whether the particles and/or the walls are electrically conductive and if the lines are grounded or not. System design must take these conditions into consideration.

4.6

Storage

Adhesion phenomena are involved in, for example, the bridging of particulate solids in hoppers. In the case of relatively coarse materials, bridge formation is caused by domelike structures, which are supported on the inclined walls in the lower conical part of bins. With decreasing particle size, the participation of adhesion forces in bridging and agglomeration increases. Binding mechanisms are molecular forces and adsorption layers or liquid bridges. The latter often play an important role whereby liquid bridges form by capillary condensation at the coordination points. Feeding warm and moist material into silos must be avoided, even if the moisture content is very low. Evaporating moisture may condense on the cooler silo walls and drip into the charge, forming wet agglomerates and causing strong capillary adhesion bonding of particles on the walls. Insulation of the silo and/or forced large volume venting can be employed to avoid condensation and agglomeration problems. Bridging can totally block the discharge from silos, thus causing severe operating problems. Because adhesion even of finely dispersed dry solids cannot be avoided, agglomerates and bridges must be destroyed by special devices. For this purpose, inflatable cushions are mounted to the inside walls of silos or the material is momentarily fluidized by the (pulsed) injection of air. In the case of coarser solids, which tend to form domes, it is often sufficient to select a cone with steeper sides (“mass flow” design). Small remaining flow problems due to adhesion can be overcome by installing vibrators or “hammers” on the outside silo walls. Unwanted agglomeration is often observed if the particulate materials are soluble or if chemical reactions can occur, particularly in the presence of moisture. These phenomena are very common and are called caking if they occur in bulk masses or bag-set if material solidifies in bags.

31

32


Different materials become caked during various storage and handling procedures but caking itself is almost exclusively by solid bridges or, more specifically, by chemical reaction and crystallization of dissolved substances. Other binding mechanisms contribute only slightly to caking. The rate and extent to which caking takes place depends on the moisture content, the particle size or specific surface area, the pressure under which the material is stored (e.g., top or bottom of the pile), the temperature and its variation during storage, as well as the time. The influence of temperature and temperature variations depends on the solubility of the solids. Fig. 4.5 shows four different temperature–solubility curves. Whereas the solubility of sodium chloride changes little with temperature, this is not true for potassium chloride (or potash) and potassium nitrate, for example. The latter features a very steep curve. Some salts, such as sodium sulfate, exhibit various temperature-dependent solubility ranges. Salts or mixtures of different salts containing a small amount of moisture, may cake during storage and/or transport if exposed to changing temperatures even if the moisture content is very small and the material is packed in airtight containers. In many cases (Fig. 4.5), more salt will be dissolved at higher temperatures, which recrystallizes and forms solid bridges between the particles when the temperature drops again. Repeated cycling, for instance due to climatic changes or differences in day and night temperatures, reinforces this bonding and causes bag-set.

Fig. 4.5 Solubility curves of four different salts

4.6 Storage

The crushing strength of caked materials depends on the number of bridges formed per unit volume and, therefore, decreases with increasing particle size. As mentioned earlier, mixtures of powdered soluble materials that were granulated by agglomeration may still set up somewhat due to the mechanism described above. However normally, the granulated material can be broken and disagglomerated easily. The answer to what can be done to avoid or, at least, lessen caking is complex but generally the same as in all other cases where unwanted adhesion or agglomeration occurs: Detect the binding mechanism that is responsible and the parameters that influence the process. Then try to reduce their effect. In the following some examples will be discussed briefly. If (unobjectional) chemical reactions between components of a mixture do occur, these components should be mixed separately until the reaction is completed. The resulting intermediate product can then be blended with the other components and no longer induces caking. An almost trivial precaution is very often to lower the moisture content. However, this is not always necessary. Different maximum moisture levels exist that depend on the material. It was found during microscopic studies that caking usually resulted from bonding by the crystals of soluble salts. These crystals often covered the entire granule surface in the form of a veneer or hull. Those salts migrated to the surface of the granule as a solution, leaving numerous small cavities within. Since this mechanism requires water, drying should reduce caking. Some materials respond favorably to several days of curing prior to bagging. Such products cake in a few days to their final strength and the resulting lumps are broken up before the cured materials are bagged and put into long term storage. Curing can even accelerate hull formation owing to heat and moisture retention. In products that respond well to curing, additional development of crystals on the granule surfaces during subsequent storage is not sufficient to cause significant caking. However, many products do not improve during this type of curing. Another curing method will be described below. The oldest method of conditioning granular materials is coating with a parting agent. Storage properties are improved after the addition of up to 3 % of an extremely fine particulate solid such as pollen, diatomaceous earth, kaolin, vermiculite, pulverized limestone, magnesium oxide, and a variety of other inexpensive, very fine powders. Again, microscopic studies revealed the fundamental properties of a conditioner, which are threefold. 1. The powder coating acts as a separator and prevents crystal bridge growth between the individual granules during and after drying. 2. The hull crystals from beneath the coating rarely project beyond the layer of conditioner. 3. The moisture is distributed uniformly over the surface of the granules due to the high sorptive capacity of the finely porous coating. Thus the localized growth of crystals at the coordination points is prevented and the surface hulls are much finer grained, more intergrown, and more densely packed than those covering un-

33

34


conditioned products. Such anticaking conditioning agents are usually applied by mixing them with the granular material in a rotary tumbler (typically a drum) prior to bagging. A modern variation of the above mentioned conditioning process is the coating with surface-active organic chemicals. It was found, however, that not all surfactants improve the physical conditions of granules. It was reported that the caking tendencies can be reduced by as much as 45 % if non-ionic chemicals were used but increased by as much as 37 % with the use of anionic materials. Where in the process the surfaceactive agents were applied was also found to be of decisive importance. Typical cationic anticaking agents are fatty amines with a general formula R-NH2 with R representing C16 and C18 chains. they are believed to attach directly to the particles with their surface-active amine group. The fatty, hydrophobic part of the molecule extends outward, thus preventing hygroscopic products from attracting moisture. Of course, this is only true if a monomolecular layer covers the granules and all amine molecules extend their hydrophobic portion outward. Therefore, too much conditioner may cause rather than prevent caking. Multiple layers are alternately hydrophobic and hydrophilic. The above makes an alternative curing process advantageous: The molecules of a second molecular layer, if attached, would position themselves with the amine group extending outward. These amine groups are free to interact with other particles. Pressure intensifies this effect. The resulting chemical “bridge” is not as strong as a recrystallized salt bridge and the “set” can be broken easily. Sometimes a combination of the two types of conditioners is used. An example of this approach is finely divided kaolin or talcum powder treated with surfactant. A last but not least method is granulation by agglomeration. Today, for very fine solids, this technology is almost obligatory, particularly for powder mixtures. Size-enlarged, granular products offer fewer coordination points per unit volume where solid bridges can develop. If the strength of the bridges is low anyway, granulation alone is sufficient to prevent severe caking. Agglomeration also prevents segregation during storage. The above examples were selected to demonstrate how unwanted agglomeration problems can be studied and how possible remediation methods and techniques

Tab. 4.2

Some dvantages of uncontrolled, natural agglomeration

Separation

Mixing Comminution Agglomeration Transportation Storage

After agglomeration, size enlarged ultrafine particles can be removed by conventional separation equipment. Selective agglomeration of specific particles in liquid suspensions (immiscible binder agglomeration. Selective agglomeration (powder coating). Mechanical activation (amorphization by crushing and recombination bonding). See Table 6.1 Larger, heavier particles. Smaller number of coordination points per unit volume.

4.6 Storage

can be determined. They date back to a point in time when the fundamentals of unwanted agglomeration in different industries were first investigated and means were developed to avoid some of these phenomena. Tab. 4.2 and 4.3 summarize what has been discussed. Tab. 4.2 shows that some of the uncontrolled, natural agglomeration phenomena are not necessarily detrimental. Tab. 4.3 lists a few of the possibilities to avoid or at least lessen the effect of unwanted agglomeration. While undesired agglomeration is often very important, because its effects may result in considerable losses of production and profit, most of the past and present Tab. 4.3 Summary of some of the possibilities to avoid or at least lessen the effect of unwanted agglomeration Separation

Mixing

Dry/moist Wet

Comminution

Agglomeration Transportation

Storage

Remove or “scalp-off” fines (immediately) at the source. Destroy agglomerates mechanically (rubber cubes/balls, shear, impact, brushes, gas (air) jets, ultrasound, etc.) Modify amplitude and/or frequency. Remove moisture from material and/or gas (air) environment. Heat (direct resistance/inductive). Modify wetting angle and/or surface tension of liquid (surfactants). Disperse particulate solids in suitable liquid (wet separation). Ground equipment (removal of electric charges). Use dispersion aids (chemicals, mechanical {e.g. stirring, ultrasound}). Apply shear or frictional stresses (tumbling mass, mixing tools, special accessories (e.g. shredders, baffles) Use shear (stirring) or dispersion agents. Utilize controlled flocculation (addition of electrolyte) that allows repeatable redispersion by shaking. Remove fines and/or moisture. Ground equipment (removal of electric charges). Coating of the inner walls of mills and crushers (elastic or non-stick). Add grinding aids (steam, vapors, surface-active substances). Lower reduction ratio (potentially multiple grinding steps). Use cryogenic milling (making use of material brittleness at low temperatures). Apply wet grinding (possibly with the addition of a small amount of electrolyte). Use drying and other post treatment measures. Modify surfaces by coating or encapsulation. Remove fines and/or moisture. Ground all lines (removal of electric charges). Make equipment, chutes, and pipes from (grounded) electrically conducting materials (avoid plastics). Finish inner walls smooth or with coating (e.g. Teflon). Remove fines and/or moisture. Cool materials to ambient conditions prior to storage. Insulate and/or heat walls. Ground (electrically conductive) hoppers, silos, and storage containers. Smooth and/or coat (e.g. with Teflon) all inner walls. Use steep (as close to vertical) walls, especially towards discharge openings. Employ mechanical bridge breakers (vibration, shock, {pulsed} air jets, inflatable inserts, etc.). Apply load (overburden) relieving means (e.g. “chinese hats”, baffles, etc.). Carry-out conditioning/curing prior to storage.

35

36


major publications deal only with methods, equipment, and systems for the production of agglomerates with beneficial properties. Therefore, it is an important achievement that a book authored by Griffith and entitled Cake Formation in Particulate Systems [B.50] was published that discusses especially the unwanted adhesion and agglomeration phenomena. The author distinguishes four major classes of caking in particulate systems: * * * *

mechanical caking, plastic-flow caking, chemical caking, and electrical caking.

In addition, several sub-classes are defined whereby specific properties of components, either pure substances or part(s) of a formulation, can be expected to cause caking under certain conditions. After describing the above, considerable emphasis is given in the book to laboratory techniques and test procedures that need to be considered by those engaged in solving caking problems. Griffith [B.50] states that one of the greatest difficulties in classifying caking (undesired agglomeration) problems comes from the very large number of variables that can contribute to its occurrence. They include material type and composition, including modifications by trace elements, which may be desired ingredients or contaminants, moisture, temperature, pressure, motion, and others. Therefore, in the section that deals with the experimental design of laboratory studies a few of the statistical methods of identifying and ranking variables are also discussed along with an evaluation of possible interactions.

Further Reading

The book [B.50] is recommended for further reading.

37

5

Beneficial Uses of Agglomeration: Agglomeration Technologies and Methods Three technologies are available for the desired size enlargement of small particulate solids by agglomeration: * * *

tumble/growth agglomeration, pressure agglomeration, and agglomeration by heat/sintering.

They can be divided into the following sub-groups [B.97]. A 1. 2. 3. 4. 5. 6. 7.

Tumble or growth agglomeration: high-density tumbling bed, high-shear tumbling bed, high-density/high-shear with abrasion or crushing transfer, low-density fluidized bed, low-density particle clouds, agglomeration in stirred suspensions, immiscible liquid agglomeration.

B Pressure agglomeration: 1. low-pressure agglomeration, extrusion through screens, 2. medium-pressure agglomeration, pelleting, extrusion through perforates die plates, 3. high-pressure extrusion, ram presses, 4. high-pressure agglomeration, – in confined spaces, punch-and-die pressing, tabletting, – in confined spaces, isostatic pressing, – in semi-confined spaces, roller presses. C Agglomeration by heat/sintering: 1. agglomeration of stationary particle beds by sintering, 2. bonding of pre-agglomerated bodies or parts during post-treatment to obtain final product properties, 3. agglomeration and bonding during special pressure agglomeration processes (i.e., hot isostatic pressing). Agglomeration in Industry. Wolfgang Pietsch Copyright ª 2005 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN 3-527-30582-3

38

5 Beneficial Uses of Agglomeration: Agglomeration Technologies and Methods

In addition, existing and new innovative technologies use the phenomena and fundamentals of agglomeration for purposes other than size enlargement. Specifically, they produce changes or improvements of the properties of particulate solids or achieve modifications of the surfaces of solids. Others manipulate individual particles or deposit ultrafine particles onto substrates in a controlled manner and subsequently bond them with the base material and/or with each other. D Technologies using the phenomena and fundamentals of agglomeration for purposes other than size enlargement: 1. coating, 2. hybridization, mechanofusion, 3. deposition and/or manipulation of individual particles. Most technologies in categories (A) to (D) can be subdivided into two techniques: * *

those utilizing no binder, and those requiring a binder.

It should be pointed out that the binding mechanism in binderless agglomeration often resembles that of bonding with a binder. This is due to the fact that binders are sometimes inherently available and act during agglomeration and/or post-treatment. A typical example of this process is the wet agglomeration (Fig. 5.2, left) of materials that are easily soluble in the liquid. The modified surface tension of the solution may already influence the strength of the green agglomerate and during drying (the necessary post-treatment to convert the intermediate wet agglomerate into the dry, final product) solid bridges develop by recrystallization of the dissolved substance(s). Other inherently available binders have to be activated by a so-called conditioning process prior to agglomeration. A typical example of this technique is in the pelleting (Fig. 5.10, b1–b6) of animal feed where the starchy component of feed grains becomes plastic and sticky during moistening and heating with steam while mixing the formulation in a kneader. After this conditioning, the starch provides plasticity that is required for extrusion through bores in the dies of pelleters as well as green and dry product strengths (see Section 6.5.2).

5.1

Tumble or Growth Agglomeration

The basic mechanism of tumble/growth agglomeration is shown in Fig. 5.1. Adhesion of individual particles to each other or to solid surfaces is controlled by the competition between volume-related separation and surface-related adhesion forces [B.48, B.97]. To cause permanent adhesion, certain criteria must be fulfilled. The most important is that the sum of all separation forces in the system (e.g., caused by gravity, inertia, drag, etc.) must be smaller than the attraction forces that act between the adhering

5.1 Tumble or Growth Agglomeration

partners. According to Fig. 3.6 and Equation 5.1, the ratio between the binding forces Bi(x) and the sum of the active components of all ambient forces Fjy(x) is a measure of the adhesion tendency Ta Ta = RBi(x)/RFjy(x) > 1

(5.1)

Both the attraction and the ambient forces are mainly dependent on the size x of the powder particle(s). To cause adhesion, Ta must be larger than one. In most cases, to keep the particle(s) adhering, the sum of all moments Qj(x) must be zero, too: Qj(x) = x/2 RFjx(x) = 0

(5.2)

Most of the attraction forces (Tab. 3.1) have only a short range; their magnitude and strength decreases quickly with increasing distance. Therefore, because the surfaces of all particulates are, at least microscopically, rough, and the mass of the particles decreases with the third power of the particle size, the adhesion tendency increases with decreasing particle dimensions. The mechanism as depicted in Fig. 5.1 occurs naturally if the agglomerate forming particles are micron and submicron or nanosized solids (beginning at about 300 %) must be again activated for agglomeration by rewetting and needs to pass once more through the entire process, including heating, drying, and cooling, which, in the final analysis, may render the technology uneconomical [B.97]. To arrive at the methods that achieve size enlargement by agglomeration in a desired and controlled manner, both a movement of particles and binding mechanisms must be created and enhanced. As the solids move in relation to each other, for example in the relatively dense bed of a rotating or otherwise actuated containment of some sort or in a low-density suspension, particles of any size and kind will collide from time to time and, if the attraction force at the collision site is high enough, coalesce.

5.1 Tumble or Growth Agglomeration

Theoretically, for this phenomenon to occur, no specific piece of equipment is necessary. As long as the solid particles are kept in irregular, stochastic motion, the probability of collision and coalescence exists. If, additionally, the binding force that has developed upon impact is strong enough to withstand the separating effects of all system forces (above and Fig. 3.6, Chapter 3) and does not disappear with time without being replaced by some other binding mechanism, the “seed agglomerate” will survive and eventually collide with other particles or agglomerates (Fig. 5.1). At each instance of collision the bonding criterion as defined in Equations 5.1 and 5.2 will be tested, leading to either growth, indifference, that is, the colliding partners will separate again and remain single, or the destruction of weaker agglomerates. To achieve growth, the individual mass of adhering particles must be small and their surface large. This is equivalent to the requirement that the size of agglomerating particles must be small. Typically, the surface equivalent diameter [B.48, B.75, B.97] should be in a range below about 100–200 lm. The limitation to small dimensions of the particles forming the agglomerate and the fact that, in most cases, only temporary bonds are formed constitute major drawbacks of all tumble/growth agglomeration methods. If particles are larger than required, crushing to achieve the necessary fineness is normally uneconomical. Immediately after growth agglomeration, in the green (moist or wet) stage, the main binding mechanisms (Chapter 3) are caused by bridges of freely movable liquids, capillary pressure at the surface of particle conglomerates that are filled with a freely movable liquid, or adhesion caused by viscous binders and slurries. To a lesser degree, other binding mechanisms, such as van-der-Waals, electric, and magnetic forces, may also participate. After curing, which often results also in a considerable strengthening of the agglomerates, bonding is achieved by solid bridges resulting form sintering, chemical reactions, partial melting and solidification, or recrystallization of dissolved substances. Some tumble/growth agglomeration equipment can handle large volumes effectively if the above requirements (small primary particle size and instantaneous bonding with high strength) are fulfilled. The apparatus is simple and the design is unsophisticated but control depends largely on operator experience. Curing is normally the expensive part of plant investment and also contributes to a large extent to operating costs, both of which may render an otherwise perfect technology uneconomical. However, if very large amounts of solids must be agglomerated and the finely divided particulate form of the primary particles is required for other reasons, for example, the concentration of valuable components of ores (Section 6.8), tumble agglomeration is the preferred technology. In those cases the main binder is water. At production capacities exceeding 1 million t per year, the curing facilities become cheaper and more economical and methods for, for example, the recuperation of heat to make the process more efficient and reduce operating costs become feasible. Other reasons for the application of tumble/growth agglomeration, even at small capacities, may be the high porosity of the agglomerates with other attendant beneficial product characteristics, such as high surface area, e.g., for catalyst carriers, and easy solubility, e.g., for food (drink) and pharmaceutical products (Sections 6.2.1, 6.3.1, and 6.4.1). These advantages may be so valuable that additional costs for grinding to

41

42


obtain the necessary small particle size for agglomeration will be acceptable and high operating costs can be absorbed. In these cases, even the agglomeration liquids (binders for the formation of green agglomerates) may be so costly that they are condensed from the dryer off-gas and recirculated. In tumble/growth agglomeration distinct process steps can be defined in which (Fig. 5.2): 1. green agglomerates are formed from solid particles and binder, 2. green agglomerates are cured, 3. if necessary, the cured agglomerates are sized (undersized material is recirculated and oversized agglomerates are crushed and rescreened or recirculated), 4. if desired, post-treatment takes place, for example, the application of anticaking agents, coating, etc. Steps 3 and 4 may sometimes move ahead of step 2 to avoid the energy cost of repeated drying and rewetting of large circulating streams of material. However, since sticking and other unwanted agglomeration problems may be encountered during sizing and oversize crushing, application of this alternative may not always be feasible. In a broad sense, equipment for the tumble/growth agglomeration process itself may be divided into the following. I. Dense Phase Tumble/Growth Agglomeration: apparatus producing movement of a densely dispersed mass of particulate solids. II. Suspended Solids Agglomeration: apparatus producing movement while keeping solid particulate matter suspended or dispersed in a fluid. In both cases, finely divided binder is added in a suitable manner to the turbulently agitated mass of particles. If solid particles are suspended in a liquid, agglomerates may be formed after adding a second, immiscible bridging (binder) liquid. In the widest sense, this technology (called Immiscible Binder Agglomeration) belongs to the type II processes. Generally, the basic process, described in Fig. 5.1, continues, causing size enlargement by agglomerate growth. However, as it proceeds, somewhat more complicated mechanisms evolve. Fig. 5.3 and 5.4 [B.48, B.97] present almost identical explanations of what is happening. While Fig. 5.3 is the more easily understandable series of sketches defining nucleation, random coalescence, abrasion transfer, as well as crushing and layering (preferential coalescence), Fig. 5.4 distinguishes between size enlargement and size reduction phenomena, both of which take place simultaneously. Nucleation, the production of primary agglomerates or “seeds”, occurs when several individual particles adhere to each other. Nucleation is the most difficult and time consuming part of any tumble/growth agglomeration process. Since only a small number of nuclei survives at any given time, this initial part of the growth process is time consuming. As long as individual particles are available they tend to adhere, trying to form nuclei or attach themselves to larger agglomerates. The latter becomes the preferential pro-

5.1 Tumble or Growth Agglomeration Fig. 5.3 Sketches explaining the different processes taking place during tumble/ growth agglomeration [B.48, B.97]

Fig. 5.4 Diagram of the mechanisms involved in size changes during tumble/growth agglomeration [B.48, B.97]

43

44


cess because the larger entities with more mass and higher kinetic energy easily “pickup” individual particles and incorporate them into their surface structure. Therefore, to accelerate the tumble/growth agglomeration process specific operating strategies that influence the nucleation stage are commonly applied [B.97]. Most of the industrial tumble/growth agglomeration methods require liquid binders to achieve satisfactory bonding for nucleation and continuing size enlargement. Although these liquids are suitably dispersed, for example by atomization [B.97], the size of the droplets produced by even the finest liquid spray nozzles is often much larger than that of the agglomerate forming primary particles. Therefore, as described by Ennis [5.1], nucleus formation in a dense tumbling particle bed (type I process) occurs as depicted in Fig. 5.5. As a droplet contacts fine particulate solids, the initial distribution of binding liquid, which depends on its wetting characteristics, influences the sizes of the resulting nuclei. Perfect wetting results in relatively strong large primary agglomerates that are saturated with liquid and are consolidated and held together by capillary forces, while imperfect wetting produces moistened particles, which may coalesce later and produce much weaker nuclei with different sizes which, in addition, are prone to attrition. If, on the other hand, a droplet makes contact with a larger entity, i.e., an agglomerate, wetting either results in spreading of the liquid on the surface or, at least partially, penetration into empty pore spaces below (Fig. 5.6). Both result in a moistening of the surface, which then attracts and captures other units, single particles or agglomerates, causing growth. Fig. 5.7 explains schematically the processes [5.1]. As the mass of the growing agglomerates increases they may break apart at structurally weaker areas or as a result of the force of impact. Abrasion will also take place resulting in newly liberated primary particles or small conglomerates, which then try

Fig. 5.5 Nucleus formation in a dense tumbling bed of fine particles [5.1]

5.1 Tumble or Growth Agglomeration Fig. 5.6 Spreading or penetration of a droplet into a powder bed [5.1]

to attach themselves to entities offering better binding conditions (Fig. 5.3 and 5.4). Particularly in batch operations, both mechanisms help to prevent the growth of a few agglomerates to excessively large sizes. To make sure that the production of oversized agglomerates is avoided or, at least, reduced, individually controlled cutting or shredding devices are often installed, which will continuously or intermittently operate and mechanically assist the breakdown of agglomerates (high-density/high-shear methods with abrasion or crushing transfer [B.97]). Depending on the density of the tumbling material, the (changing) mass of the individual agglomerates, and the type of equipment causing agitation, the growth phenomena and, herewith, the agglomerate properties will differ. One reason for change is the varying extent of the previously mentioned naturally occurring or mechanically induced abrasion, break-down, and reagglomeration. Another is how new particles are attached and incorporated into the structure. It is obvious that particle beds, tumbling in rotating equipment or agitated by mixing tools, will produce denser agglomerates than obtained in the low-density particle clouds of fluidized beds.

Fig. 5.7 Processes occurring between particles of a tumbling powder bed after wetting with droplets of a binder liquid [5.1]

45

46

5 Beneficial Uses of Agglomeration: Agglomeration Technologies and Methods Fig. 5.8 Diagram of particle and droplet distribution in a low-density fluidized bed and of particle penetration into a liquid [5.1]

Referring to Fig. 5.8 and comparing it with Fig. 5.6, in low-density fluidized particle beds (type II process), particles and droplets coexist and, in most cases, agglomerates form by particles penetrating into the liquid. This is the prevailing growth phenomenon in re-wet fluidized bed agglomerators, were growth is limited by the requirements of particle fluidization. A more basic growth mechanism in fluidized beds, which resembles that of “natural” agglomeration is obtained if the particle surfaces are moistened by condensing steam in “steam jet agglomeration” [B.97] or by ultrafine liquid droplets (including solutions and suspensions) produced with the “RESS process” [5.2].

Tab. 5.1 Some conditions, reasons, and requirements for the selection of tumble/growth agglomeration. Process conditions Small feed particle size (xo < 100–200 lm) Wet processing Normally, (liquid) binder(s) is (are) required Green (wet) agglomerates with low initial strength Post-treatment for final strength and properties Simple equipment design Control depends largely on operator experience Difficult to clean (danger of cross-contamination) Not amenable to quick turn-over. Product characteristics Irregular (overall spheroidal) shape Final strength depends on binder(s) and post-treatment Wide distribution of agglomerate sizes and weights High porosity High solubility, dispersibility, reactivity, etc. Shelf-life often limited

Cold processing

System “footprint” is large

Small pieces

Low density

5.2 Pressure Agglomeration

During the agglomeration by growth in tumbling particle beds, “natural” adhesion forces, which are either totally inherent, enhanced by suitable methods, introduced by binders, or acting as a combination of two or all of these effects, cause particles to stick together when they collide in a stochastically moving mass of particulate solids. With the exception of forces that are exerted during the interaction between the particles, the environment and equipment walls as well as, in some cases, various mixing and/or shredding tools, no externally induced directional forces or pressures act on the growing agglomerates and no shaping, other than caused by attrition, occurs. As a result, depending on the level of interaction, which is largely influenced by the tumbling bed density, more or less spherical agglomerates are grown featuring a size distribution that depends on the system properties. Because of the relatively small forces caused by interactions in and with the tumbling charge, porosity of the agglomerates is high and increases as bed density decreases. Also, since the adhesion forces are small, too, and separation forces, which try to destroy the growing agglomerates, are mass related, the primary particles forming the agglomerate must be small. Typically, with a few exceptions, temporarily bonded “green” agglomerates are produced, which require posttreatment to achieve permanent, final strength. Narrowly sized products are obtained by additional processing such as sizing, crushing, and shaping or attrition. In summary, Tab. 5.1 compiles some of the conditions, reasons, and requirements that are typical of tumble/growth agglomeration. Of course, this table does not claim to be complete but is meant to provide some guidelines that can be also gleaned in more detail from the discussion above and from other sources (Section 13.1).

5.2

Pressure Agglomeration

In pressure agglomeration, new, enlarged entities are formed by applying external forces to particulate solids in more or less closed dies that define the shape of the agglomerated product. In contrast to tumble/growth agglomeration, pressure agglomeration is used to achieve one or more and sometimes all process conditions and product characteristics summarized in Tab. 5.2. The level of force that is applied during densification and shaping is the most distinguishing factor in pressure agglomeration. Therefore, the technology is subdivided into low-, medium-, and high-pressure techniques. Of course, as will be shown later, certain conditions and characteristics are better obtained with one or the other pressure agglomeration process and, sometimes, one or more of the parameters of Tab. 5.2 can not be met with a specific technique and/or equipment, system, or plant. Each of the great variety of pressure agglomeration methods corresponds to one or more of the binding mechanisms of agglomeration (Chapter 3, Tab. 3.1, Fig. 3.4 and 3.5). While in low- and medium-pressure agglomeration all binding mechanisms are equally possible, in high-pressure agglomeration attraction forces (Fig. 3.5) provide the most common bonding. According to the mechanisms involved, the processes can be again further categorized as those using binders and those without binders. However, all have in common a basic compaction mechanism.

47

48

5 Beneficial Uses of Agglomeration: Agglomeration Technologies and Methods Tab. 5.2 Some conditions, reasons, and requirements for the selection of pressure agglomeration. Process conditions Larger feed particle size Dry processing Hot processing Processing of elastic materials Easy clean-out Product characteristics Specific shape Specific mass (or weight) Low porosity Long shelf-life Amenable to the production of near-net-shape parts

High initial strength No or little binder No post-treatment Automatic operation Quick turn-over Large pieces High density High final strength

The upper part of Fig. 5.9 shows with four model sketches the structural change of a bulk mass of particulate matter during densification in a die, the attendant change in volume, and an indication of the modifications of particle shape and size that occur at high pressure. The lower part of Fig. 5.9 depicts the build-up of pressing force with time under the assumption that the forward movement of the punch occurs with a constant rate until pmax is reached, at which point the direction of movement reverses and the punch retracts with the same or a higher speed. Referring to Fig. 5.9, as a first step, pressure agglomeration achieves a rearrangement of particles that requires little force and does not change particle shape and size. This is followed by a steep rise of pressing force during which brittle particles break and malleable particles deform. Sketches 3 (brittle) and 4 (plastic) occur either/or and often simultaneously if both brittle and malleable particles are present in the mix.

Fig. 5.9

Sketches explaining the mechanism of pressure agglomeration


Two important phenomena limit the speed of compaction and, therefore, the capacity of any pressure agglomeration equipment: compressed residual gas (in most cases air) in the pores and elastic springback. Both cause different, equipment specific cracking and a weakening or, sometimes, total destruction of the products [B.13(b), B.48, B.97]. Low- and medium-pressure agglomeration apply small forces (up to about the beginning of the steep slope of the curve in Fig. 5.9) but, nevertheless, the removal of a relatively large amount of gas must be guaranteed (the time axis is directly proportional to the volume change since the punch advances with constant speed). Development of compressed gas pockets within the densifying product can be avoided if compaction occurs slowly enough for all gas to escape from the diminishing pore space. High-pressure agglomeration extends into the steep increase of the pressing force. In this range, particle size and shape change by breakage and or deformation and porosity is further reduced. The maximum pressing force is normally defined by an overload feature of the equipment. Since a predetermined final strength and structure must be reached, equipment selection must take into consideration that a sufficiently high maximum pressing force can be attained. After arriving at the maximum pressing force, pressure is released. If, as shown in Fig. 5.9, compaction is performed by a punch in a die, the direction of travel of the piston reverses and, when no expansion of the densified body occurs, the pressing force should drop to zero immediately. In reality, there is always a more or less pronounced rebound, which is caused by the expansion of compressed gas and the relaxation of elastic deformation. This effect becomes more pronounced with increasing speed of densification until, at a certain compression rate, the compacted body disintegrates partially or totally upon de-pressurization. Therefore, it is often necessary to find an optimal compromise between densification speed (capacity) and product integrity (quality). The problem becomes greater with finer particles because such materials are naturally more cohesive and, therefore, in the feed state, feature lower bulk density or higher bulk volume. In these cases, cohesive arches will collapse at low pressure whereby large amounts of gas are driven out. At the same time, pores between fine particles are small, which results in low diffusivity so it takes a relatively long time for the large amount of displaced gas to escape. To help overcome problems associated with degassing or deaeration, special design features, such as force feeders and/or various provisions for venting, are applied with all pressure agglomeration methods, particularly if fine powders must be processed [B.97]. If the mechanism of densification is considered (refer to the sketches in the upper part of Fig. 5.9), it becomes clear that the pores in the feed to a pressure agglomeration process of any kind must not be filled completely (saturated) with a liquid. An example of such a material would be a normal filter cake: one that has not been blown dry or otherwise further de-watered. Since liquids are incompressible, the pressing force would increase quickly and mechanical de-watering would have to occur, which further reduces the speed of densification. It would also require an effective separation of solids and liquid during the densification process; this is a problem that has not yet been solved. Therefore, with increasing pressure applied to the particulate solids, which typically results in higher densification or lower porosity, the moisture content

49

50

5 Beneficial Uses of Agglomeration: Agglomeration Technologies and Methods Fig. 5.10 Diagram of equipment for: a) low-; b) medium-pressure agglomeration: a1, screen; a2, basket; a3, radial; a4, dome; a5, axial; b1, screw; b2, flat die; b3–b5, different designs of cylindrical dies; b6, gear

of the feed must diminish. In high-pressure agglomeration the feed must be essentially dry. The equipment in which pressure agglomeration occurs is a machine that operates with well-defined mechanical parameters that are independent of the performance and characteristics of the particulate solids to be processed. Therefore, pressure agglomeration techniques lend themselves readily to automation and remote control and are essentially independent of operator presence and/or skill. Because the equipment is relatively complex and the throughput per unit is often limited, this technology finds its greatest usage in low- to medium-sized applications (about 0.1–50 t/h). Of course, this statement is relative. Specialty products, such as those in the pharmaceutical industry (Section 6.2.2), may be processed in very small and sophisticated machinery, handling only a few kilograms per hour, while certain high-tonnage materials, for example some fertilizer, refractory, and mineral materials, are briquetted or compacted in large facilities employing multiple units (Sections 6.6.2 to 6.10.2). Relatively uniformly shaped and sized agglomerates can be obtained with low- and medium-pressure agglomeration. For these processes, the feed mixture must still be made-up of relatively small particles and inherently available, activated, or externally added binders (above). The moist, often sticky mass of particulate solids as well as plastic and liquid binders is extruded through holes in differently shaped screens or perforated dies (Fig. 5.10). Agglomeration and shaping are caused by the pressure


forcing the mass through the holes and by the frictional forces developing during the material’s passage. Depending on the plasticity of the feed mix and the dimensions of the holes, short “crumbly”, elongated “spaghetti-like”, or cylindrical green extrudates are produced. Particularly the thin, string shaped agglomerates that are obtained from low-pressure agglomeration (Fig. 5.10, a1–a5) are often spheronized, that is, rolled into small spherical particles while the product is still plastic [B.48, B.97]. In most cases a post-treatment (typically drying and cooling) is required to yield final, permanent strength. As far as applicability is concerned, high-pressure agglomeration (Fig. 5.11) is the most versatile technique for the size enlargement of particulate solids by agglomeration. If certain characteristics of the feed materials and conditions occurring during densification are considered during equipment selection as well as plant design and operation, dry particulate solids of any kind and size, from nanometers to centimeters, and at any condition, for example with temperatures from below freezing to 1000 8C, can be successfully processed. Typically, the products from high-pressure agglomeration feature high strength immediately after discharge from the equipment. Nevertheless, to further increase strength, addition of a small amount of binder and/or post-treatment methods are possible. An advantage of high-pressure agglomeration is that, as discussed above, in most cases, essentially dry solids are processed, which do not tend to set, so that the process can be stopped at almost any time and re-started easily; also, the amount of material in the system is relatively small. Therefore, pressure agglomeration methods, specifically those applying high pressure, lend themselves particularly well to batch or shift operation and to applications in which several products must be manufactured from different feed mixtures in the same unit. At the end of a campaign, the system can be easily and completely emptied in a relatively short time. If the danger for crosscontamination is unimportant, for example in the fertilizer industry, a new campaign

Fig. 5.11 Diagram of equipment for high-pressure agglomeration. Ram press (upper left), punch and die press (upper right), roller presses for compaction (lower left) and briquetting (lower right)

51

52


Fig. 5.12 Cycles of force build-up in the three different high-pressure agglomeration techniques

with a different feed can be started immediately (Section 6.6.2). Another possibility for fast and quick change-over is to install different feed and product discharge/handling systems as, in most cases (for example the pharmaceutical industry, Section 6.2.2), the expensive pressure densification/shaping equipment itself can be easily cleaned or adapted to the manufacturing of a new product. The above-mentioned destructive effects of expanding compressed air and relaxation of elastic deformation can be reduced if the maximum pressure is held for some time, called dwell time, before it is released. Fig. 5.12 shows that, without special

5.3 Agglomeration by Heat/Sintering

technical provisions, this is only achieved in ram extruders (Fig. 5.11, upper left). In such equipment, a number of briquettes are retained in the long pressing channel and are redensified during each stroke. After the wall friction is overcome and the entire line of briquettes moves forward, the pressing force remains almost constant (Fig. 5.12b). A similar, but much smaller effect is obtained in pellet presses (Fig. 5.10, b1–b6). Since a dwell time and, particularly, the application of several densification cycles also helps to convert temporary elastic deformation into permanent plastic deformation, these techniques are especially suitable for the densification of elastic materials such as, for example, biomass. If a dwell time is required or desired in punch-and-die presses (Fig. 5.11, upper right), special drive systems must be used [B.48, B.97]. It is obvious from Fig. 5.11, lower left and right, that no such possibility exists in roller presses where a continuous rolling action densifies the material between approaching surfaces until, immediately after passing the point of closest approximation, the relative motion is reversed, the surfaces retract, and the pressing force drops, ideally to zero if no expansion due to compressed gas and/or stored elastic energy takes place.

5.3

Agglomeration by Heat/Sintering

At a certain elevated temperature, which is different for various materials and can be quite low for organic products or very high for minerals, atoms and molecules begin to migrate across the interface where particles touch each other. While still in solid state, depending on temperature, time, and intensity of contact (caused by pressure during the manufacturing of a pre-form or the sintering process itself), diffusion of matter forms bridge-like structures between the surfaces, which solidify upon cooling. The process may also result in a densification of the compact, which is due to an elimination of pores and associated shrinkage. The entire group of phenomena is called sintering [B.13(c), B.97]. Agglomeration by heat or sintering has been developed and is applied mostly in industries processing minerals and ores for the size enlargement of fines before further use (Section 6.8.3). Because the technology requires large amounts of thermal energy, special efforts are made to recover heat or use sources of waste heat. The resulting agglomerates are crude but meet the requirements of the industry. Another large application of sintering in agglomeration is in post-treatment where the phenomenon is used to produce strong permanent bonds and/or specific final properties in many parts that may have been manufactured by virtually any one of the other agglomeration techniques. Particularly in powder metallurgy, sintering is the most important finishing process for the achievement of final strength and structure (Chapter 7). More recently, in this field and for other post-treatment processes, it has been found that several mechanisms exist that cause material transport at elevated temperatures for bridging and, thus, strengthening of the preform or the agglomerate, but do not result in shrinkage. Therefore, sintering can also yield strong final products with high porosity [B.78, B.97].

53

54


5.4

Technologies Using the Phenomena and Fundamentals of Agglomeration for Purposes other than Size Enlargement

Among the technologies that use the phenomena and fundamentals of agglomeration for purposes other than size enlargement the oldest is coating, which was first used to avoid the sticking together of pills and/or to mask the taste of the often unpleasantly tasting medicinal remedies. The coating material was either applied as a fine naturally occurring powder (such as pollen) together with some moisture or as a natural (e.g., nectar) or man-made sugar solution. The coating was made permanent by drying (originally by the heat of the sun). Today this technology has evolved into an important field of solids processing [B.48, B.97]. It is used in almost all industries that are specifically covered in Chapter 6 and has further potentials in many other applications. The most common modern methods employ equipment similar to that utilized in tumble agglomeration [B.97]. Inclined pans, often pear shaped, are applied for simple coatings (for example in the food industry for sugar or chocolate covers, Section 6.4.3) and sophisticated drum designs are used widely if coatings with uniform distribution and well-controlled thickness are required (for example in the pharmaceutical industry for functional layers, Section 6.2.3). Fine solid particles and a liquid binder, solutions, or suspensions are sprayed onto or into the tumbling bed of solids that are to be coated and form a temporary wet layer on their surface. Gas (normally dry hot air) is simultaneously passed through the bed, evaporating the liquid and leaving a solid coating,

Fig. 5.13 Sketch of the material processing section of a bottom spray fluidized bed coater (Wurster coating system) [B.97]

5.4 Technologies Using the Phenomena and Fundamentals of Agglomeration for Purposes

Fig. 5.14

Diagram of possible structures of microcapsules [B.97]

the thickness of which grows as the process continues. Unique design features have been developed with regard to the shape of the container used for tumbling, the feeding of solids (if any), the spraying of the liquid, and the input as well as exhaust of the drying gas [B.97]. Another technique sprays melts onto the moving bed of particles, which form a coat that solidifies upon cooling with a flow of ambient air (Section 6.6.3). The stochastic movement in low-density fluidized beds is used for coating relatively small particles with narrow particle size distribution. Many of the newer designs use some modification of the Wurster coater [B.48, B.97] (Fig. 5.13).

55

56


Fig. 5.15 Diagram depicting the process of hybridization and microphotographs showing particles with intermediate as well as final coating [B.97]

A further coating method is encapsulation, most commonly called microencapsulation [B.48, B.97]. It is a packaging technique that enrobes powders, particles, liquids, or even gases and forms free flowing, dustfree particulate products. The most basic method makes use of the capillary flow of liquid and the recrystallization of dissolved (or deposition of colloidal) solids on the surface of saturated wet agglomerates during drying. More recently, technologies are being developed in which the capsule features a specific, well-defined functionality. The capsule material may offer controlled-release characteristics (e.g., pharmaceutical products, Section 6.2.3), insolubility and well-defined bursting strength (e.g., “dry” inks and toners, Section 6.11.3), delayed availability (e.g., agrochemicals, Section 6.6.3), and other properties. Fig. 5.14 describes schematically the possible structures of microcapsules. As shown in Fig. 5.14d, bottom, heterogeneous coatings on a core can be also fixed by embedding. The processes that accomplish this are called mechanofusion or hybridization [B.97]. Mechanical forces act either on a previously deposited layer or on systems consisting of larger core and ultrafine coating particles. Often the core particles are relatively soft and inert with the coating providing functionality (for example

5.4 Technologies Using the Phenomena and Fundamentals of Agglomeration for Purposes

Fig. 5.16 Overview diagram describing manipulation techniques for small solid particles [5.3]

Fig. 5.17 Precise arrangement of SiO2 spheres on pinpointed locations: a) voltage contrast image of a dotted line; b) silica spheres arranged on the dotted line [5.3]

57

58


electrical conductivity, Section 6.11.3) by embedding in the surface (Fig. 5.15). Because the coating particles are often so small that no dislocations are present in their structure, they behave as very hard entities. Therefore, it is, for example, possible to embed ultrafine titanium particles in the surface of glass spheres. A very new group of methods for the deposition and bonding of particulate solids onto surfaces assembles nano- to micrometer sized particles in a predetermined and orderly fashion onto substrates [5.3]. Fig. 5.16 gives an overview of the two groups of manipulation techniques that are available. While with the upper methods particles can be deposited with great accuracy but low rate, with the lower ones larger numbers of particles can be delivered with less accuracy. The limitations of both can be overcome if the intended locations of the particles are predetermined by electrification with a focused ion beam (FIB). Fig. 5.17a is the voltage contrast image of a line in which the dots where formed by the impingement of a Ga+-FIB. The spacing of the dots is 5 lm. As shown in Fig. 5.17b, 5 lm silica spheres are attracted to the dotted electrified line and form an almost perfect array of particles. Microdevices or microstructures with multi-functions (Chapter 11) can be obtained after permanently fastening and interconnecting the particles with the substrate and with each other, for example by sintering.

Further Reading

With exception of the following: B.14, B.20, B.23, B.27, B.30, B.42, B.75, B.81, B.91, B.103, B.104, all publications listed in Chapter 13.1 are recommended for further reading.

59

6

Industrial Applications of Size Enlargement by Agglomeration All techniques of beneficial size enlargement by agglomeration yield products that feature some or all of the advantages listed in Tab. 6.1. In certain cases many of the characteristics are desired, but sometimes one specific requirement may be the sole reason for adopting a suitable agglomeration method in a manufacturing line. Agglomerated materials may constitute final products, such as tabletted pharmaceutical specialties (Section 6.2), pelleted animal feed (Section 6.5), instant drink concentrates (Section 6.4), briquetted solid fuels (Section 6.10) and so on, or intermediate products, such as granulated pharmaceutical formulations as feed for tabletting

Tab. 6.1

General advantages of agglomerated products

Enlarged apparent size. No or low content of dust. Increased safety during the handling of, for example, toxic or explosive (highly reactive) materials. Fewer losses (by dusting or elutriation). Less primary and/or secondary pollution. Freely flowing. Improved storage and handling characteristics. Better metering and dosing capabilities. Reduced tendency of unwanted agglomeration. Increased bulk density and reduced bulk volume. Smaller size of packages and storage or transportation volume. No segregation of co-agglomerated materials. More uniform supply of multi-component materials to users. Defined size and shape. Sometimes defined weight of each agglomerate (dosage form). Distribution may be narrowed by post treatment (sizing). Shape may be modified by post treatment (e.g. spheronizing). Within limits, porosity or density can be controlled. Possibility to influence dispersibility, solubility, reactivity, heat conductivity, or other related properties. Improved product appeal Material properties can be adjusted to meet government, local, or industry standards or requirements. Higher sales value and increased profit potential. Agglomeration in Industry. Wolfgang Pietsch Copyright ª 2005 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN 3-527-30582-3

60

6 Industrial Applications of Size Enlargement by Agglomeration

presses (Section 6.2), briquetted DMT (di-methyl therephtalate) flakes for safe and economic shipment to the fiber production plants (Section 6.3), agglomerated silica fume to reduce its volume and improve the handling characteristics during transportation (Section 6.7), and so on. The most common reason for size enlargement by agglomeration is, as the term implies, the desire or need to obtain a product with larger particle dimensions. As described in Chapter 3, the resulting entity is only apparently a new unit. The original solid particles are still present in the structure [B.97], often with a completely unaltered shape and size, and are held together by binding mechanisms. Furthermore, the new entity features porosity: voids between the agglomerate-forming particles (Chapter 3). This is beneficial in some cases, for example, for easily dispersible food (Section 6.4) and pigment (Section 6.3) granules, for “designer” foods (Section 6.4), and for catalyst carriers (Section 6.11), or undesirable in others, as in hot densification of sponge iron (DRI) to reduce reactivity (Section 6.9) and high-density ceramic (Section 6.7) and powder metallurgical (Chapter 7) parts. As a further result of this modification of the original particulate solids, many bulk characteristics of the new product are changed too, mostly for the better. For example, larger particles have less dust, exhibit improved flow behavior, and feature much-reduced sticking tendencies. Therefore, storage, handling, and feeding is less risky, even for “difficult” materials. After production, many modern particulate solid materials are very aerated and remain loose, largely due to their small particle size and high surface area, which cause instant, natural adhesion and prohibit consolidation by settling. If packaged, loaded, or stored, such products occupy large volumes and require excessively large containers for packaging or big silos. During agglomeration the natural adhesive bonds are broken and the individual particles are bonded more closely together, resulting in higher density, so that the aforementioned problems are eliminated. If mixtures are agglomerated after blending, the distribution of the different components is fixed in the agglomerates and thereafter segregation is avoided. Agglomerated products are better sized and shaped and often can be used as solid dosage forms. Their properties may be adjusted to meet specific requirements by influencing the agglomerate structure, either directly or by post-treatments. Generally, many agglomerated materials have better appeal and command higher prices and profits. After agglomeration some particulate solids exhibit properties that are undesirable for their use or application. In most cases, these negative characteristics relate to the new porosity of the agglomerate, which may be, for example, too low for easy dispersibility, or too high, resulting in excessive reactivity. Bad product properties can also be caused by unsuitable binding mechanisms, for example absorbents for liquids may disintegrate when wetted or, vice versa, a binder in water-dispersible granules may not dissolve when the product is submerged in the liquid. More recently, agglomeration has entered new fields in which individual particles, mostly in the nanometer size range, are attached to each other or to substrates in a controlled fashion (Chapter 11). Areas of application include the manufacturing of new composite materials with novel characteristics and of microscopic structures for electrical circuits or material associations on a molecular level. The latter result

6.1 General Applications

in the evolution of unique drug-delivery systems or life-science products, while the former is increasingly used for electronics and in the communication industries. In the following sections examples of specific applications of beneficial agglomeration by size enlargement in different industries are presented. According to the commonly used classification of the available methods (Chapter 5) they are subdivided into tumble/growth, pressure, and other agglomeration technologies. The latter include agglomeration by heat/sintering and technologies using the phenomena and fundamentals of agglomeration for purposes other than size enlargement, such as coating and some of the already industrially applied technologies involving nanoparticles for the building of new engineered materials and novel particle-modification technologies.

6.1

General Applications The most general application of beneficial size enlargement by agglomeration is the granulation of powders. Basically, the terms “granules” or “granulate” mean relatively coarse (0.1–10 mm) particulate solids (Chapter 14). Contrary to the often performed granulation of solids by crushing (the size reduction of larger solid pieces), granulation of powders uses the technologies of size enlargement to achieve the same goal, the production of a “granular” product. Although with regard to the overall granulometry both products are similar or even identical, physically they are different. Granular material obtained by the crushing of solids consists of particles that have similar structure to the solid from which they were produced. Because failure during crushing begins preferentially at imperfections (dislocations, faults, cracks, inclusions, and other defects), the smaller particles become increasingly more homogeneous and, with decreasing size, approach the structure of the ideal solid. In contrast, the granulation of powders involves binding mechanisms and suitable processing of fine and/or ultrafine particulate solids. The individual particles are brought closely together during impacts or by applying external forces. At that instant, binding forces that are larger than the separating forces caused by the ambient conditions (Fig. 3.6 and [B.48, B.97]), develop between the solid surfaces, resulting in particles sticking together. By these mechanisms, larger units develop by growth or as defined by a die in or by which they are formed. However, the new entities are only seemingly solid. In reality, unless modified by a post-treatment procedure (Chapter 7), the resulting agglomerates are made-up from the largely unaltered powder particles, retaining most of these particle characteristics, and feature porosity, that is interconnected voids, which may represent a volume percentage as high as 95 % or more and as low as 5 % or less. Granular products from powders may be produced by tumble/growth agglomeration (Section 6.1.1); while some are immediately in the correct size range, others re-

61

62


quire post-treatment and/or sizing to remove over- and under-sized materials. In the latter case, it is necessary to consider what to do with the rejects. In many cases, large agglomerates are produced first, to make use of economical tumble/growth and pressure agglomeration or sintering methods (Sections 6.1.1, 6.1.2 and 6.1.3); the pieces resulting from these procedures are then crushed to form the granular material. Effective size reduction into a defined size range (granulation), whereby over- and undersized particles must be kept within certain narrow limits, is a fundamental problem of fracture mechanics. Moreover the failure mode of solids is different from that of agglomerates. Therefore, these phenomena and how they influence granulation will be described in more detail below. Granulation of solids by crushing, results in a product distribution that typically includes oversized and undersized particles. The size distribution of the discharge from a crusher depends (among many other parameters) on the stressing mechanism (Figs. 4.3 and 4.4, Chapter 4). Other important influences are the feed size and shape, the energy input, the crushing tool and chamber configuration, and the reduction ratio, which is a characteristic derived from the previous variables. The larger the reduction ratio (the difference between feed and product size, Fig. 4.3), the higher the energy input must be and the more fines are produced. When a solid body is stressed, which requires the addition of energy from the outside, the tensions that are produced within stretch all bonds between the molecules. Theoretically, if the solid features an ideal structure with no imperfections and only tensile forces are applied, the bonds separate in a structural plane in which the load exceeds the bond strength and the solid splits cleanly into two parts. Then all other areas return to their equilibrium structure, and the unused energy is freed and converted into other forms: kinetic and thermal energy and/or sound. However, real solids contain many flaws, which become stress raisers and initiate cracks at loads much below those required for the separation of an ideal body, even if only tensile forces are applied. The stressed system is unstable and cracks expand rapidly, accelerating to high propagation velocities: only the bonds at the crack tips are breaking at any instant. Under combined stresses, the normal condition in an industrial crushing situation, many cracks will be initiated and propagate in a direction perpendicular to the local tensile stress but, sooner or later, they run into a region of compression, which stops further crack growth [B.12]. Therefore, the highest number of cracks will be initiated and cause failure near the point(s) of energy input. Also, if primary fractures relieve the

Fig. 6.1-1 Diagram of failure lines in single spherical and irregular particles during: left) compression. right) impact crushing [B.12]


Fig. 6.1-2 Photographs of: a) the fines cone; b) coarse residual pieces obtained during impact crushing of a single glass sphere [6.1.1]

stress sufficiently rapidly, vibrations are induced that result in new tensile stresses and may cause secondary fractures. As shown in Fig. 6.1-1, mainly fine material is produced when a high energy density is released at or near the point(s) of energy input. The volume of the fines cones and, therefore, the mass of fines grow with increasing energy input or, in other words, with larger reduction ratio. Fig. 6.1-2 shows the fines cone and coarse residual pieces obtained during the impact-crushing of a single glass sphere [6.1.1]. Since the main objective of the crushing of solids is to achieve a certain fineness and avoid oversized particles, many impact crushers and mills (e.g., hammer mills) feature some sort of a restriction (bar cage, screen, or perforated plate) at the discharge from the chamber in which the material is stressed. This exit grate, screen, or perforated plate retains material in the chamber until it is small enough to pass through the openings. Although this eliminates the presence of oversized particles in the product, a number of problems can or will arise. *

*

*

*

Such a measure will increase the amount of fines, as each stressing event will produce fines as discussed above. A finer than desired product may be produced because particles that are just below the size of the openings have a low probability of exiting it and, therefore, are retained, stressed again, and broken to finer sizes. To reduce this effect, the exit grid may be increased which, of course, increases the danger of oversized particles in the product. Depending on the requirements on size distribution this measure may not be acceptable. If crushers operate near their upper capacity limit, temporary or permanent choking of the grinding chamber may occur, requiring downtimes for clean-out.

63

64

6 Industrial Applications of Size Enlargement by Agglomeration *

*

If the solid feed contains hard inclusions, selective crushing may occur. Hard, difficult-to-crush particles may remain larger than the openings of the exit grid and accumulate within the chamber, eventually also causing choking. Finally, but not least important, the exit restriction is a highly stressed part. It needs frequent maintenance (exchange or rebuilding) and, in some cases, for example in the ultraclean food and pharmaceutical industries but also in other applications where minute contaminations with iron or alloying components cause problems (for example, discoloration of glass when small amounts of iron are present), are not acceptable.

Large amounts of fines are produced by different mechanisms and increase with energy input or larger reduction ratio. They are often undesirable in the final product, so most of them must be removed by sizing, for example by screening or air classification. Although these fines normally have the same composition as the solid and, therefore are potentially valuable, they were traditionally discarded as waste. Size enlargement by agglomeration makes it possible to produce materials from fines for recycling or for use as undesirableecondary raw materials (Section 8.2). Nevertheless, since the conversion of fines into products for beneficial use is costly, it is more economical to minimize the production of fines during crushing. When the mechanisms that occur during size reduction of solids were first described by Rittinger [6.1.2] and later modified by Griffith [6.1.3] it became clear that to minimize fines production, a low energy input must be used. While this measure reduces fines, a substantial amount of oversized material is also obtained, which must be restressed. As shown in Fig. 6.1-3aa the simplest solution to this problem is to separate the larger pieces from the discharge and recirculate them to the mill, which must be oversized to be capable of handling the considerably greater throughput. As mentioned before, particles (of any size) in the milled product tend to be stronger because imperfections have been removed. Therefore, the optimal energy input for crushing oversized particles that resulted from a primary size-reduction event is increased. The arrangement shown in Fig. 6.1-3ab, whereby the second mill operates at a higher energy input than the first one, would seem to be appropriate. Although a somewhat improved result (narrower product size distribution with less fines) is obtained from both mills, they do not represent an optimized solution, because system ab still produces some oversized particles, which may have to be removed and otherwise used or (Fig. 6.1-3ac), possibly recirculated to the second mill (“closed loop” handling of larger pieces). Today it is commonly understood [6.1.4] that, from an energetic point of view, sizereduction processes are considered optimal when: *

* *

the particles to be crushed obtain a specific stressing energy directly from the milling tools (which may include grinding media and specially designed chamber walls) an undefined interaction between particles does not occur energy is not wasted on particles that are already fine enough – such materials are removed from the chamber


Fig. 6.1-3 Diagrams of a crusher with oversize recycling and various two- or three-stage crushing circuits [6.1.4]. Circles represent crushers or mills; horizontal lines flanked by + and – represent classifiers (screens) that split the crushed material into over- and undersized particles. “A” is an air-classifier that removes fines by entrainment

65

66


fines that tend to interfere with the controlled energy input are eliminated, both prior to and during size reduction.

With this in mind, Fig. 6.1-3ba depicts a much better system, which can be further improved (if all particles above a certain size must be eliminated) by recirculating the secondary crusher discharge into the classifier (Fig. 6.1-3bb). Alternatively, a second classifier separates oversized particles from the secondary crusher and only recirculates this portion to the secondary crusher (Fig. 6.1-3bc). If the emphasis is on reducing fines in the product and optimizing performance of the secondary crusher, an air classifier to remove fines by entrainment could be installed between the primary and secondary crushers (Fig. 6.1-3bd). Figs. 6.1-3ca, cb, and cc show the next step of optimization, which is three-stage crushing. Depending on the quality requirements on the final product, more grinding steps and/or additional closed loops can be added, on the understanding that crushing circuits become more expensive with each step. Therefore, a compromise between best result and cost must be found. In the sketches of Fig. 6.1-3 it has been assumed that by optimizing crushing the amount of fines has been reduced to such an extent that they can remain in the product, but fines elimination at the end of the circuit, possibly by installing a screen or similar separator for fines, may be necessary as a final quality adjustment step. As defined by the method of energy input, excluding shear, there are two main crushing mechanisms: compression and impact (Fig. 6.1-1). Although the above comments are valid for both mechanisms, the explanations referred more directly to impact stressing. As shown in Fig. 4.3 (Chapter 4), the importance of the reduction ratio for the outcome of milling is even more important when compression is applied. If the reduction ratio is too high, flat agglomerates are formed in which coarser pieces that are in at least one dimension smaller than the narrowest final distance or gap, are embedded in fines. Since the fine particles are immediately compacted after formation, almost all free valences at the newly created surfaces participate in recombination bonding [Chapter 3, B.48]. Recently it was found that compared with the stochastic milling process in a tube mill with grinding media, the combination of well-defined compression stressing in a high-pressure roller mill with large reduction ratio (Fig. 4.3) and the ensuing disagglomeration of the compacted flakes that were produced result in a significantly lower overall energy consumption during the fine grinding of brittle materials, such as cement clinker and many ores. Nevertheless, during “normal” crushing of solids by compression, either between two approaching flat (jaw crusher) or curved (roller mill) surfaces, the formation of compacted flakes or similar agglomerates must be avoided. This can be only achieved if relatively small reduction ratios are applied. In Fig. 6.1-3 it can be seen that a certain amount of oversized material is produced after a particular crushing step and the corresponding system sketches of Fig. 6.1-3 can be applied whereby the second, third, or potentially higher-stage compression crushers feature decreasing gap sizes (as compared with higher energy input, mostly defined by speed, in the case of impact crushers).


Fig. 6.1-4 Diagram of multi-stage crushing between three sets of rollers and photographs of two-stage and three-stage Gran-U-Lizer roller mills (courtesy Modern Process Equipment, Inc. (MPE), Chicago, IL, USA)

If rollers are used for crushing, several sets with decreasing gaps can be installed in the same housing (Fig. 6.1-4). Because the material flows through the machine under the influence of gravity, and if the throughput is such that the nip area is not overfed (no pile-up), fines produced in the first stage that are smaller than the gap width of the second stage pass it without additional stressing while larger particles are crushed. This is demonstrated in Fig. 6.1-5, which shows the comparison of typical cumulative particle size distributions after crushing solids in a single stage roller mill and an MPE

Fig. 6.1-5 Comparison of typical cumulative particle size distributions after crushing solids in a singlestage roller mill and an MPE Gran-U-Lizer with three sets of rollers. Note the much reduced amount of oversize (shaded area) and the almost unchanged amount of fines (courtesy Modern Process Equipment, Inc. (MPE), Chicago, IL, USA)

67

68


Fig. 6.1-6 Diagrams of various arrangements of profiled rolls for roller mills: a) peak-to-valley arrangement showing the “cracking” of a particle; b) parameters determining control of the particle size in a peak-to-valley arrangement; c) peak-to-peak arrangement; d) example of a peak-to-valley arrangement with modified pitch and gap (courtesy Modern Process Equipment, Inc. (MPE), Chicago, IL, USA)

Gran-U-Lizer with three sets of rollers. The shaded area, upper right, depicts the much-reduced amount of oversize material after three-stage roller mill crushing, while the amount of fines (lower left) remains almost unchanged. For certain materials and applications, the result of crushing in roller mills can be improved by using profiled roller surfaces. Fig. 6.1-6a shows a circumferentially grooved pair of rollers with a “peak-to-valley” arrangement and how a particle is crushed between such rollers. The mating peaks and valleys on the rollers act as fulcrum points that actually “crack” a particle that is bridged as shown. Variables that influence the result of crushing are: the pitch and gap adjustment (Fig. 6.1-6b) and the positioning of the matching rollers. Fig. 6.1-6a and b describe the staggered arrangement (peak-to-valley); another possibility is a directly opposed arrangement (peak-to-peak, Fig. 6.1-6c. Fig. 6.1-6d shows, as an example, another pitch and gap in a peak-to-valley arrangement. The granulation of powders by size enlargement, whereby agglomerates are first formed, then crushed into the desired particle size range, and finally classified to obtain the desired product size distribution, is even more difficult to analyze than the complex size reduction of solids. The reason for this is twofold. 1. During the crushing of solids the production of fines is expected, often with a certain amount desired, and in other cases the production of fine particles is the sole objective of the operation, but the optimal result of the granulation of powders by agglomeration and crushing would be to produce no fines at all. This is understandable as powders are, by definition, fines and granulation is performed to produce larger particles. However, since the formation of fines can not be totally avoided, it is most important to select methods that minimize this effect.


2. Agglomerates are assemblages of powder particles that are joined together by binding mechanisms. If compared with the (ideal) structure of the solid itself, where atoms and molecules are held together by intermolecular attraction forces (valence forces) and form regular arrangements, which depend on the type of the atoms and molecules and on the formation mechanism, the structure of agglomerates is made up of irregularly arranged small particles with different size and shape and features void spaces (porosity). Because small particles contain few structural irregularities and flaws, their strength is high and always exceeds that of the binding mechanism acting between the solids. As a result of this characteristic property of all agglomerates, it is easy to break them into the powder particles from which they were originally made. Therefore, breaking down larger agglomerates into smaller, but still agglomerated granules (featuring sizes that are greater than those of the agglomerate-forming primary particles) must be done with gentle, low-energy methods to avoid the formation of excessive amounts of fines. Harsh methods of stressing and high energy input may

Fig. 6.1-7 Diagram of the structure of different agglomerates: a) tumble/growth agglomeration; b) low- and medium-pressure agglomeration (extrusion); c) high-pressure agglomeration

69

70


easily produce solids that are, at least partially, finer than the particles from which the agglomerates were originally made. Generally, the same measures must be applied for agglomerates as described above for solids to obtain a good result from granulation without the production of excessive amounts of fines. This means that the energy input should be selected so that the agglomerate to be granulated breaks into as many large fragments as possible with minimum fines production at the points of stress initiation. Because of the porous structure and the method of creating strength (bonding at coordination points, Chapter 3), application of shear is particularly detrimental to avoiding fines. Powder particles can be easily removed from agglomerates by abrasion, especially from the surfaces of freshly produced fragments. Therefore, the use of exit grates or screens in crushers to retain larger material in the milling chamber and avoid oversized particles in the product is a sure way of creating additional fines when granulating agglomerates. Referring back to Fig. 6.1-3, single-stage impact milling of agglomerates should employ the system depicted in aa. In single-stage compression crushing, recirculation of oversized particles is ineffective. Both types of crushing can be optimized by applying two or more crushing stages. Systems bc and cd are particularly well-suited if impact milling is used, while compression stressing is best carried out in roller mills with multiple sets of rolls (Fig. 6.1-4) and profiled roll surface (Fig. 6.1-6). An additional peculiarity of the granulation of powders by agglomeration and crushing is due to the fact that most agglomerates to be reduced in size feature different structure, strength, and/or density over the cross section of the body. Large agglomerates resulting from tumble/growth agglomeration are produced in dense moving particle beds (in drum or mixer agglomerators [B.48, B.97]). According to the agglomeration mechanism in such equipment (Chapter 5), powder particles adhere to the outside surface of nuclei or intermediate aggregates and, if the bond is stronger than the sum of all separating system forces (Fig. 3.6) become an integral part of the growing entity. As the mass of the agglomerate becomes larger, under the influence of kinetic and reaction forces from the surrounding material and equipment walls, the newly attached particles may be more tightly embedded in the current surface structure of the agglomerate. Therefore, it is possible that the density and strength are increasing towards the surface of large agglomerates that have been produced by tumble/growth mechanisms in dense particle beds, particularly if these bodies feature onion-skin structure (Fig. 6.1-7a1) [B.48]. As described in Chapter 5, tumble/growth agglomeration occurs almost always in the presence of a liquid binder, which wets the solid particles and derives strength from surface tension and capillary forces. Such bonding is only temporary because if the liquid evaporates and no other binding mechanism takes over, the agglomerate looses its entire strength. The most common secondary binding mechanisms that develop during drying of wet or moist agglomerates are the recrystallization of dissolved substances or the deposition of colloidal particles at coordination points within the agglomerate (Chapter 3, [B.48, B.71]). If the pores within an agglomerate are completely or partly filled with liquid (Fig. 3-1b and Fig. 3-2d and e, Chapter 3), drying


begins on the surface of the body and the drying zone remains there as long as liquid is replenished from the interior by capillary flow. Only when pores cease to be filled continuously, will the drying zone move into the agglomerate. During the first phase of drying, which takes place only on the entity surface, dissolved substances or suspended particles will deposit on or near the surface, resulting in a strong and dense crust [B.48, B.71]. Since large agglomerates made by tumble/growth methods require a high moisture content to reach bigger sizes and crushing for the production of a granular product always occurs after drying, when the permanent binding mechanism has developed, it is common that such agglomerates feature a strong surface layer and a looser core (Fig. 6.1-7a2). In extrusion (low- and medium-pressure agglomeration, Fig. 5-10, Chapter 5) the force causing densification results from friction between the material and the die channel. In addition to a somewhat higher density close to the die wall, as compared with the center of the extrudate, due to force dissipation, shear from the sliding motion of the material in the channel produces a “skin” on the surface (Fig. 6.1-7b). The extent of this effect depends on the forces applied, which are a function of the equipment and the material characteristics, and on the length of the extrusion mold. It is more pronounced in products from medium-pressure agglomeration and almost not noticeable if wet powder masses are passed through screens (low-pressure agglomeration). In high-pressure agglomeration, particulate solids are compressed and densified in confined-volume, open-ended, or converging dies (Fig. 5.12, Chapter 5). When forces (or pressures) are applied onto particulate solids they dissipate at the contact points of one particle with other particles surrounding it. Because each particle contacts several others and the number of contact points increases during densification, the forces decrease quickly towards the center of the compact. This dissipation becomes larger with smaller particles, and if forces are introduced from opposite sides, a neutral plane exists where the directional sign of the acting forces changes. The location of the neutral plane depends on the size and direction of the opposing forces [B.28, B.48]. As a result of this dissipation of forces, particulate solids are more densified near the points or planes where the external forces are exerted onto the particulate mass and less dense in more distant locations. While, normally, there is a rather well-defined highly densified surface layer, overall a density gradient exists towards the center or the agglomerate. Because strength is strongly dependent on porosity (the opposite of density) and decreases with increasing porosity (or lower density), a strength gradient also exists across the agglomerated body. Fig. 6.1-7 shows the situations; the shaded areas feature high density and strength: c1, briquette from high-pressure ram extrusion or cylindrical compact from punch-and-die pressing; c2, briquettes from roller pressing; c3, flat sheet from roller pressing; c4, corrugated sheet from roller pressing. With this understanding of how agglomerates are structured and considering the laws of fracture mechanics it is now possible to define phenomenologically how the granulation of powders by crushing agglomerates should be best carried out. By now it should be obvious that a direct, single-stage reduction of the particle size from the agglomerated feed to the granulated product will normally result in an excessive amount of fines. Therefore, in a first step, agglomerates from any process source must be cracked, taking into consideration the surface layer with higher

71

72


strength (Fig. 6.1-7), to form as many particles as possible, including oversized pieces. During this crushing step the energy input must be selected to minimize the amount of fines. The discharge must be separated to yield fines, correctly sized product, and oversized particles. Most of the oversized pieces are the strongest parts of the agglomerate. They must be recirculated to the crushing equipment or fed to another sizereduction process. If, depending on the system used, oversized particles are again obtained after restressing, they will be even stronger and require additional size reduction. This can be accomplished with various systems using different methods of disintegration. Beginning with compression stressing, the aforementioned multiple-pass roller mills (Fig. 6.1-4) promise best results. These mills are designed so that the gap between the rollers can be easily adjusted (Fig. 6.1-8), even during operation, to allow optimization of the product particle size distribution. The gap of the last set of rollers defines the maximum particle size and is normally adjusted such that no oversized material is obtained. Depending on requirements on the product particle size distribution, fines may have to be removed and discarded or otherwise used, including recirculation into the agglomeration process. Differently profiled rollers (Fig. 6.1-6) may be advantageous in certain cases. Today, by far the largest number of granulation systems that crush agglomerates into a specific particle size range use some kind of impact mill. Hammer mills are most commonly applied but other types, notably cage mills with one static bar cage and one or more rotating one(s), may be selected. As mentioned before, to avoid the production of excessive amounts of fines, if this is a requirement, agglomerated solids should be stressed once, thereafter immediately removed from the grinding chamber, and separated into at least three fractions containing undersized (fines), correctly sized (product), and oversized particles. Because larger than desired pieces are always unacceptable, this fraction is either recirculated or fed to a secondary crusher. Fig. 6.1-9 is a simplified comparison of the results of granulation by crushing agglomerates, using three different milling circuits. In all cases hammermills are utilized. All diagrams depict the agglomerated feed with 100 % in a relatively narrow, large particle size range and the desired (shaded) particle size of the granular product. For the result shown in Fig. 6.1-9a, a hammermill with discharge restriction in the form of a bar cage or a screen was used (Fig. 6.1-10). The restriction is meant to avoid oversized pieces in the discharge of the mill and the opening size of this feature is selected such that nothing leaves the milling chamber with a size larger than the

Fig. 6.1-8 Automatic adjustment with pneumatic servomotors and micrometer or digital readout in multi-stage roller mills (courtesy Modern Process Equipment, Inc. (MPE), Chicago, IL, USA)


Fig. 6.1-9 Comparison of the results of granulation by crushing agglomerates: a) a mill with exit grate or screen, b) an impact (hammer) mill with unobstructed discharge and recirculation of oversized particles (Fig. 6.1-3aa), and c) two-stage crushing (Fig. 6.1-3bb) applying two individually controlled impact mills with unobstructed discharge (for explanations see text)

upper limit of the granular range. Oversized material, including already sufficiently reduced pieces and fines, which can not immediately clear the discharge openings, remain in the milling zone and experience turbulent movement and a multitude of stressing events, including shear. In addition to clean impacts from the hammers, largely undefined energy input occurs, mostly between the hammers, the chamber walls, and the discharge restriction that leads to the production of fines, particularly since the material to be crushed consists of agglomerates. The example in Fig. 6.19a assumes that the acceptable granular size range is relatively narrow. Therefore, only 40 % product and 60 % fines (undersized particles) are obtained. This represents an ineffective method of granulation.

73

74

6 Industrial Applications of Size Enlargement by Agglomeration Fig. 6.1-10 Diagram of an impact mill with hinged or fixed hammers on a horizontal rotor and discharge restriction (bar cage or screen) to avoid oversized particles in the product: a) grate bars, b) cage with axially elongated holes, c) cage with round holes

It should be mentioned, however, that in some applications, especially those in which the granular material represents an intermediate product, manufactured as a freely flowing, essentially dustfree, non-segregating, directly compressible feed for punch-and-die presses (for example in the pharmaceutical industry, Section 6.2), the entire output from such a mill, including all of the fines, may be adequate for further use. Also if, in other cases, the granular size range can be wider, the yield of acceptable product becomes larger so that the correspondingly smaller amount of fines may be recirculated or otherwise used without sacrificing the economics of the process. In Fig. 6.1-9b it may be assumed that the same mill as used in example A is employed but without discharge restriction. The tip speed of the hammers (defining the energy input) is selected such that during crushing a wide size range of broken pieces, including oversized ones, is produced. After separation into undersized (fines, 25 %), correctly sized (product, 40 %), and oversized (35 %) particles, the large pieces are recirculated to the mill (Fig. 6.1-3aa) where it is assumed that they will disintegrate into the same particle size distribution as experienced before with the whole agglomerates. From the discussion above, which stated that oversized particles after the crushing of agglomerates represent the stronger parts of the structure, this assumption is not totally correct. If the oversized particles are recirculated into the same mill, exerting an identical energy input as during the first crushing event, the stronger pieces should produce a somewhat different size distribution. For the sake of this presentation the difference is disregarded. Since the oversized fraction, which, in a closed loop recirculation layout, results from a multitude of cycles, produces material fitting the three size fractions during every stressing event, the final yield of product and fines, the only two streams discharging from the system, comes from an infinite accumulation of iterative amounts whereby all the coarse particles are ultimately converted such that they fall into the two size ranges. As compared with Fig. 6.1-9a, 21.5 % more product (61.5 %) and, correspondingly, 21.5 % less fines (38.5 %) are produced with the granulation method shown in Fig. 6.1-9b. As discussed before, this result can be further improved if the granular size range is wider and, depending on the requirements on the granulated material, the now much-smaller percentage of fines may actually be an acceptable component of the product.


Because after the first crushing step the oversized particles represent the strongest portions of the agglomerates, a further improvement of granular yield and reduction of fines can be obtained if, instead of returning this size fraction back to the same mill (Fig. 6.1-3aa) another similar or identical mill, the latter employing modified energy input, is installed to crush the oversized portion. This second-stage mill could be in a straight through arrangement (Fig. 6.1-3ba), employ an internal closed loop (Fig. 6.1-3bb), or feature a separate closed loop for the still-remaining oversized pieces (Fig.s 6.1-3 bc and bd). By being able to control the energy input of the second mill, which now receives only the oversized particles from previous crushing events, it can be assumed that the particle size distribution of the discharge is improved in regard to product and fines (for example: undersized (fines, 25 %, unchanged), correctly sized (product, 50 % instead of 40 %), and oversized (25 % instead of 35 %) particles). With this and using crushing systems according to Fig. 6.1-3bb or bc, Fig. 6.1-9c shows that a further small increase in product yield of 1.9 % to 63.4 % and the corresponding reduction in fines (to 36.6 %) can be achieved. Instead of recirculating the coarse fraction from the second mill to that same one (Fig. 6.1-3bb, bc, or bd), a third crushing stage could be added (Fig. 6.1-3ca, cb, and cc) and even higher-order multi-stage systems are conceivable. Based on the small improvement of granular yield by adding a second mill (comparison between Fig. 6.19b and Fig. 6.1-9c) it may seem uneconomical to use two or higher order granulating circuits. However, with different agglomerates, employing other types of crushers, and by optimizing the results of crushing in each milling stage through modified energy input, it is possible to achieve dramatic improvements of the yield from the granulation of powders via agglomeration and controlled size reduction. It should be pointed out that the maximization of granular product yield may not always be the preferred solution. As will be described later (Section 6.6.2), granular yield can be maximized by employing gentle crushing methods, low energy input, and multi-stage milling but, at the same time, weaker parts of the original agglomerates may survive and become product. Such granules feature lower abrasion resistance and produce dust during bulk storage, transshipment, and distribution, which is objectionable for some materials, leading to complaints from the customers. Although specific applications of granulation by agglomeration and crushing will be described in many of the sections covering different industries, in Sections 6.1.1, 6.1.2, and 6.1.3 some generic flowcharts will be introduced and discussed. They are mostly applied to generally improve the handling characteristics of finely divided particulate solids in mechanical process technologies without trying to modify other properties of the materials or obtain special, beneficial product qualities. In most cases, the reason for increasing the size of fine or ultrafine particles by agglomeration is to avoid undesired adhesion or agglomeration (Chapter 4). To achieve this, the particle size of the material is commonly adjusted to 0.1–1.0 mm. In this size range solids, whether naturally compact or agglomerated, have insufficient mass for the always-present adhesion forces to cause permanent bonding between the granular particles or with surfaces of any kind. If bonding occurs due to the presence of binders (moisture) or extraneous binding mechanisms (e.g., electrostatic charges), the strength of the agglomerate is low and the aggregation can be easily destroyed or build-ups and coatings can be removed without difficulties.

75

76


To explain this in the context of common incidents occurring in the process industry, two examples are discussed in more detail. The first example refers to the adhesion of fine particles to walls of storage, transfer, or processing contrivances. Although ultrafine particles may permanently stick to technically smooth surfaces due to natural adhesion forces, an even stronger bond is obtained if the size of the solids is in the range of or smaller than the roughness of the substrate. Because all technical surfaces feature a microscopic texture that results from the machining or polishing technology used, powders will always contain some particles, which, in addition to the action of molecular forces, will interact mechanically with the surface topography. On such a first layer of adhering particles, more of the very small particles will attach themselves. Because a large number of bonding sites exist per unit volume and weakly bonded particles are removed by external forces so that only strong bonds survive, such a build-up will continue to grow and will be difficult to remove by methods other than mechanical cleaning and, potentially, washing.

Fig. 6.1-11 Glass cylinders filled with the same gravimetric amount of a powder having different particle sizes (a) from left to right: 4 mm, 1.5 mm, 100 lm, 12 lm, 5 lm, 28 nm, 18 nm, 16 nm, 12 nm and (b) left 4 mm, right 12 nm


On the other hand, when the powder particles have been agglomerated into granules with somewhat larger size, these new entities may still adhere to walls if the conditions are favorable. However, because the number of bonds between the particles is smaller and since the agglomerates have greater mass, the cohesion of the build-up is so small that it will break-off under its own weight when it reaches a certain thickness. The second example relates to agglomerates that form in a particulate mass. Apart from the fact that the previously discussed build-up grows on a substrate (wall), similar conditions apply. If the particles are very small, a large number of bonds exist per unit volume rendering the resulting agglomerate strong and difficult to destroy. Secondary agglomerates from larger sized granulated powders are weaker and can be easily dispersed by the forces prevailing, for example, in a moving mass of particulate solids. At the same time, it is a precondition that the agglomerated granules, making-up the easily dispersible secondary agglomerate, are so strong that they withstand those same forces without breaking-up. A further problem that is associated with fine and ultrafine solids and can be remedied by size enlargement is that the entire mass of powder particles responds to the presence of the naturally existing molecular adhesion. This means that, in bulk, they adhere to each other and do not settle under the influence of gravity. This effect increases with diminishing particle size and is particularly pronounced when the solids are in the nanometer range. Primarily this results in a large, often excessive bulk volume (low bulk density) (Fig. 6.1-11) but also in bad flowability and in discharge difficulties. As shown in Tab. 6.1, size enlargement by agglomeration can overcome these handling and process problems.

6.1.1

Tumble/Growth Technologies

As discussed before, size enlargement by agglomeration for general applications is typically used to improve the handling characteristics of finely divided particulate solids in mechanical process technologies without trying to modify other properties of the materials or obtain special, beneficial product qualities. In most cases, the reason for increasing the size of fine or ultrafine particles by agglomeration is to avoid undesired adhesion or agglomeration (Chapter 4). To achieve this, the particle size of the material is commonly adjusted to 0.1–1.0 mm. In Chapter 5 the following methods of tumble/growth agglomeration were listed. 1. 2. 3. 4. 5. 6. 7.

high-density tumbling bed high-shear tumbling bed high density/high shear with abrasion or crushing transfer low-density fluidized bed low-density particle clouds agglomeration in stirred suspensions immiscible liquid agglomeration

77

78


A detailed discussion of these methods and of the equipment used to accomplish them can be found in the author’s earlier books on the subject [B.48, B.71, B.97]. Although in Section 6.1 it was indicated that, theoretically, granular products can be made by crushing cured (dried) large agglomerates from tumble/growth technologies, the problems inherent with these procedures were also mentioned which, in most cases, prohibit the use of such a technique. Therefore, the granulation of powders by tumble/growth agglomeration for the improvement of general handling properties in the process industry normally tries to manufacture directly the desired particle size. Owing to this requirement some of the above seven methods are preferred for this task while others are normally not used. At “normal” operating conditions, the high density tumbling bed obtained in discs, cones, and drums best produces agglomerates in the range 10–15 mm. In addition, particularly from drums, a wide agglomerate size distribution is obtained. Therefore drums are not used for the granulation of powders by agglomeration. Under certain conditions and with continuous, extensive observation and control by experienced operators, discs with low rim height [B.48] can be applied for “micro-agglomeration” (often called “micropelletizing”) yielding particles in the range 1–3 mm. However, because of the costly operator involvement, this technique is only used for materials with high value, for example the granulation of certain agrochemicals (Section 6.6.1), or if a worker is required at the pan for other reasons (for example, packing the discharge into drying trays, see Section 6.11.1) who will observe and control the operation. High shear tumbling beds are produced in mixer agglomerators with rotating tools [B.48, B.71, B.97]. The small size and relatively narrow distribution requested for the granulation of powders can be produced best if agglomeration occurs in high density/ high shear particle beds with abrasion or crushing transfer (Figs. 5-3 and 5-4, Chapter 5). This procedure applies shear plates between the mixing tools or independently driven, high speed knife heads (also called shredders or chopper elements), which are periodically operated [B.48, B.71, B.97]. These provisions are used to destroy oversized agglomerates and make the fragments available for new growth within the desired size range. Although, with good preparation and control, batch or continuous mixer agglomerators can produce rather well-defined granular materials from powders they are infrequently used for general applications. Since in this rather unpretentious field the emphasis is on large throughput capacity and low cost processing, mixer agglomerators do not normally meet these conditions. Because of the mechanisms defining agglomeration in low-density fluidized beds or particle clouds [B.48, B.71, B.97], where only relatively small, narrowly sized granules can be manufactured and kept suspended without either settling or becoming entrained, this group of processes seems to be well suited for converting fine or ultrafine particles into granular products. However the motions, forces, and growth mechanisms that are at work result also in relatively weak agglomerates with low density, which are not suitable to withstand the stresses encountered during regular handling in mechanical processing plants. The structure of agglomerates obtained in low density fluidized beds or particle clouds yield other beneficial product qualities, which make them more suitable for specific applications, for example in the food (Section 6.4.1)


Fig. 6.1-12 Block diagram of a powder granulation system by tumble/growth agglomeration for general applications also showing various optional features

or pharmaceutical industries (Section 6.2.1). Nevertheless, certain solids, which are prepared by spray drying of solutions, suspensions, or slurries are granulated by simultaneously agglomerating them in a fluidized bed that is associated with the spray dryer (Section 6.7.1) to make them more easily handleable. Agglomeration in stirred suspensions and immiscible liquid agglomeration are always used to produce special, beneficial material qualities and, therefore are not of interest for general applications. With exception of the dry granulation of carbon black (Section 6.11.1) and silica fume (Section 6.7.1) all tumble/growth processes that are employed for granulation just to obtain improved handling properties of powders make use of wet agglomeration. Therefore, they include drying as a post-treatment process because the temporary bonding provided by liquids must be converted into a permanent strength producing mechanism. As a result, a block diagram for these general applications is as depicted in Fig. 6.1-12. If different powders are to be granulated they are proportioned (metered) from day bins into a blender, which may operate in batches or continuously. It is also possible to install a mill instead, especially if the various raw feeds have different particle size distributions that might lead to segregation and selective agglomeration in the size enlargement step. Particularly if a blender is applied, it is feasible to introduce additives and/or some of the liquid binder(s) at this point (preconditioning). However, care must be taken that not all of the binder liquid is added, which would eliminate an important control factor for size adjustment in the tumble/growth agglomerator.

79

80


A single component or homogeneous material feed is directly metered into the unit accomplishing size enlargement by tumble/growth agglomeration. The remaining binder liquid, normally atomized by suitable means (spray nozzles), is added and causes the tumbling particle mass to coalesce and form the desired granules. In dense beds of particulate solids a wide distribution of agglomerates, including oversized ones, is developing. The application of high shear or the operation of intensifier bars and intermittent use of knife heads [B.48, B.97] specifically results in the destruction of large agglomerates and of new growth from the fragments. In size enlargement by tumble/growth in any of the processes, green (moist or wet) agglomerates are formed, which are held together by temporary binding mechanisms. If natural curing, for example the hydration of cementitious components, does not take place, some post-treatment is required, almost always encompassing the elimination of the binder liquid and, at the same time, the development of a new, permanent binding mechanism (e.g., the recrystallization of a dissolved solid). Particularly if the granules are highly saturated, which is typical if size enlargement has occurred in a dense tumbling bed, liquid travels to the surface of the agglomerates during drying by capillary flow. Since, on the other hand, the green granules are weak and often become weaker during drying before a new bond takes over, curing takes often place in a stationary bed in which, caused by the previously described mechanisms, the aggregates may stick together and form oversized lumps. This is normally not observed with granules that were obtained in fluidized particle beds or clouds as drying takes place simultaneously or in another fluidized bed that is closely associated with the agglomerator. In such an arrangement, the particles are in constant motion and do not stick together. While for general applications the presence of some over and under sized particulates is often not objectionable, as an option the discharge from post-treatment may be separated into two or three size fractions. Fines (undersized particles) are discarded, used elsewhere, or recirculated to the agglomerator and coarse pieces (oversized particles and lumps) may be crushed in a closed loop whereby the fragments, which typically include again all three fractions, are recirculated to the sizing operation. Selection criteria for the crusher have been discussed in Section 6.1.

6.1.2

Pressure Agglomeration Technologies

Pressure agglomeration is also frequently used to produce granular materials for general applications. The same improvements of the powder properties as discussed in Section 6.1.1 are sought, which means that the particle size of the material is typically adjusted to 0.1–1.0 mm by granulation. In Chapter 5 the following methods of pressure agglomeration were listed. 1. low-pressure agglomeration: extrusion through screens 2. medium-pressure agglomeration: pelleting, extrusion through perforates die plates 3. high-pressure extrusion: ram presses


4. high-pressure agglomeration – in confined spaces: punch-and-die pressing, tabletting – in confined spaces: isostatic pressing – in semi-confined spaces: roller presses A detailed discussion of these methods and of the equipment used to accomplish them can be found in the author’s earlier books on the subject [B.13b, B.48, B.71, B.97]. Only with low-pressure agglomeration can small extrudates with diameters < 1 mm be manufactured directly by passing formable (normally moist) particulate masses through screens. Because these machines are delicate and feature low capacity they are not used for general applications. All other pressure-agglomeration equipment produces relatively large, more-or-less densified pieces called briquettes, compacts, extrudates, pellets, sheets, slabs, and similar names. To meet the requirement that most of the granular materials, manufactured solely to improve the handling behavior of finely divided particular solids (general applications), should be in a particle size range of < 1 mm, the regular products from medium- and high-pressure agglomeration must be crushed. A plant embodying such a process, irrespective of which type of pressure agglomeration equipment is used, is called a compaction/granulation system. Fig. 6.1-13 is the block diagram of a typical compaction/granulation system. Right away it should be mentioned that the advantage of this technique of granulating powders is that any product size and distribution can be obtained. The final quality only depends on the crusher(s) and the sizing method(s) used. In this connection the discussion of Section 6.1 must be considered. A disadvantage is that always fines are also

Fig. 6.1-13 Block diagram of a powder granulation system by pressure agglomeration (compaction/granulation) for general applications also showing various optional features

81

82


produced, the amount of which depends on the cut size (Chapter 4) that is associated with this property and whether or not undersized particles are detrimental and must be removed. If the granular material must be free of fines, the rejected percentage increases with narrower product size distribution. As discussed in Section 6.1, the granulation of agglomerates by crushing is a difficult undertaking. It is, of course, always the primary requirement of a compaction/ granulation plant to produce the highest possible yield of high-quality granules in the desired particle size range. To accomplish this it is necessary to define and obtain the granule quality and size range. Both properties may differ widely. The quality of granules that are produced solely to improve the handling characteristics of fine particulate solids may be defined such that they just survive the stresses, for example at transfer points or in feeders, but easily disperse when finally processed or, in the other extreme, they must be hard and abrasion resistant if they are, for example, produced to facilitate safe disposal. The size range of the granulated solids may be wide and consist of the entire output from the granulator, even including the unavoidable fine particles, or narrowed down to a small fraction by sizing. This can be accomplished by merely removing the undersized portion from a granulator discharge that contains no coarse particles because, for example, a screen or bar cage are installed in the crusher exit. As discussed in Section 6.1, when agglomerates are stressed in such a mill, excessive amounts of fines must be expected, which have to be used or discarded elsewhere or recirculated. The latter increases the equipment sizes in the system considerably. Therefore, if a narrow size range of the granular product is required, in keeping with the explanations of Section 6.1, the optional closed crushing loop for oversized particles featuring a secondary optimized crusher and twin separation steps (with a double deck screen) should be considered. It has been mentioned in Section 6.1 that the products of medium- and high-pressure agglomeration feature a well defined density and, therefore, strength gradient. Both density and strength decrease from the surface of the compact to its center. The effect depends on the size (thickness) and the shape of the pressure agglomerate (Fig. 6.1-7). As will be shown in more detail in Section 6.6.2, where it has been a common problem, crushing the compacts too gently results in relatively soft parts of the agglomerate becoming part of the final product. This may not be objectionable if granulation is solely performed to improve, often only temporarily, the storage and handling problems of fine particulate solids. However, the fact that efforts to maximize the yield of granular material may lead to the presence of soft, low-strength granules should be acknowledged and remedied if the production of (additional) fines during handling becomes a problem. Even though general applications of size enlargement by agglomeration normally do not require special or high granule quality and are often only produced as an intermediate product with temporary life, sometimes the irregular shape featuring corners and edges (Fig. 6.1-14) is undesirable, mostly because these parts rub-off easily and create particles that are so small that they become airborne. In such cases conditioning may be added. Conditioning typically happens in rotating drums where the granules are rounded during tumbling and the abraded fines are separated by entrainment in an


Fig. 6.1-14 Typical granules from compaction/granulation showing the irregular shape of the unconditioned particles

air stream or afterwards mechanically on a screen; alternatively, a small amount of mostly liquid coating may be added to provide either adhesion properties for fines, which are then picked-up by the moistened agglomerates, or to smoothen and/or strengthen the surface of the granules.

6.1.3

Other Technologies

“Other technologies” comprise agglomeration by heat, also called sintering, and methods that manipulate fine particulate solids such that specific structures are produced. They are not normally used to produce granular materials for general applications

83

84


because the first, requiring a large amount of thermal energy, is too expensive and the second does not manufacture agglomerates solely to improve the handling behavior; in fact, particle manipulation procedures and the use of binding mechanisms of agglomeration are always applied in specific industries to engineer products with defined new properties. One technique that is being described in more detail in Section 6.9.2 as a preparatory step for the hot briquetting of steel mill dusts for recycling shall be mentioned here, also to introduce the idea for future use with a variety of materials. The process could be conceived simply for the conversion of fine, reactive (e.g., metal-bearing) dusts, which are not valuable enough for reuse as secondary raw material (Chapter 8), into dustfree, easily handleable granular aggregate for safe disposal. In this respect it meets the requirement of “general applications” because no specific characteristics in regard to size, structure, or properties are necessary other than the need to be dust free, of sufficient strength, and easily handleable. During the production of, for example, metals by the liquid route, particularly in gasblown converters, ultrafine particles are produced that are so small that they are difficult to capture with conventional methods, become airborne, and, for example, were responsible for the brownish atmosphere in and around steel mills during the late 19th and the early part of the 20th centuries. Although most of these pollutants are oxidized impurities, tiny droplets of metal are also produced, which quickly solidify and become part of the dust. Because these particles are very small their specific surface area is so large that the smallest ones feature self-ignition characteristics. This means that, in the presence of oxygen, they are liable to oxidize, even at ambient temperatures. This chemical reaction is exothermic (producing heat). In an effort to clean the air in and around metal producing centers, improved dry dust collection systems were and are being developed, which remove the particulate contaminants but, at the same time, result in large volumes of difficult to handle dust. It is possible to use the auto-oxidation behavior of such materials beneficially for the production of agglomerates by sintering. For this purpose, the dust is tumbled in an inclined steel drum and air (or oxygen) is introduced into the bed by a suitable sparger pipe. The reoxidation reaction produces heat, the level of which is controlled by the supply of oxygen, and sintering into agglomerates of various sizes occurs. While the binding mechanism is that of sintering, the agglomeration mechanism is that of tumble/growth. Typically, this process operates continuously and fines discharging with the agglomerates may be separated and reintroduced into the feed end of the drum. Because more and more of the particulate solids that are removed from plant effluents are or contain ultrafine particles and legislation requires the safe disposal of these materials to avoid recontamination of air and/or water (Chapter 8), the high reactivity of these particles and the exothermic production of heat during oxidation can be generally used to achieve low cost granulation by heat. In this context it should be remembered that, while for the sintering of metal and mineral particles very high temperatures (typically exceeding 1200 8C) are required, the sintering mechanism begins at about two thirds of the melting or softening temperature of a particular solid. Therefore, it is feasible to apply the technology for the granulation of other reactive ultrafine particles at much lower temperatures.

6.2 Pharmaceutical Applications

6.2

Pharmaceutical Applications

By trial and error, humans found that certain herbs and other organic materials and some minerals can be used to alleviate or cure ailments. Early physicians dried the materials, ground them to yield fine powders, proportioned and blended different ingredients, and finally administered the remedy in a traditional way. Since some of these medications were given orally, “solid dosage forms” were already prepared early in human history by the tribal medicine people, because fine powders can not be swallowed easily and the concoctions often had a bad taste. To that end, the powdered drugs were mixed with binders, for example starch (flour) and water or honey, and rolled into spherical pills. Honey was often also used to mask the taste. Because the pills thus produced were still sticky after rolling, they were sometimes coated, for example with pollen, to render them dry and easily storable. Even today, in many rural dispensing pharmacies, in developed countries for homeopathic remedies, and in lesser developed parts on Earth for many, often natural drugs, the ancient pill making is still performed by producing a formable mixture from the ingredients, proportioning the mass into the dosage size, and manually rolling these pieces into spherical pills (Fig. 6.2-1) [6.2.1]. During the 10th century AD, Arabs used a simple press to form a moist medicinal powder into shapes [6.2.2]. The mixture was “compacted” by hand between the two halves of a tongue-shaped tool, which was made from bone, ivory, wood, or stone. For many centuries, pills and manually produced compacts were well accepted by the patients as solid medicines for oral application. Nevertheless, the invention of the tabletting press during the middle of the 19th century in the UK and the USA initiated a veritable revolution in the fledgling pharmaceutical industry. Aside from the production of previously unheard of numbers of tablets, the major difference between today’s manufacturing of oral dosage forms and the technology that was successfully performed for centuries is that many powders are agglomerated dry. Tab. 6.2-1 summarizes the reasons for and/or results of size enlargement by agglomeration in pharmaceutical applications. Since larger amounts and numbers of different powders, which for reasons of uniform product composition are very fine [B.48], had to be tabletted in machines operating with ever increasing speed, flowability, determining the ease of filling the die, and compactibility, controlling the production of high-quality tablets during very short densification cycles, must be often improved. This is accomplished by pre-agglomeration into a free-flowing granular feed. It was originally carried-out by wet agglomeration, another centuries-old method, and is now increasingly done by dry compaction/ granulation. Under certain conditions, a single modern rotating table tabletting machine is now capable of producing more than 1 million tablets per hour [B.97]. The two most important aspects of agglomerated pharmaceutical products are the following.

85

86


Fig. 6.2-1 Sketches of tools for manual pill making [6.2.1]. 1, “Pill machine”: a) rolling plate, b) rope cutter, lower part, c) rope cutter, upper part, d) rolling board, e) area for ropes and finished pills. 2, Mortar (with pestle). 3, Pill rolling (disc) tool *

*

They are intermediate or final dosage forms, which must carry reliably and reproducibly the required amount of active substance. Final dosage forms are consumer products; therefore, consumer appeal is of great concern, which means for agglomerated specialties that they must have uniform, aesthetically pleasing, and reproducible shape and weight.

Tab. 6.2-1 Reasons for and/or results of size enlargement by agglomeration in pharmaceutical applications Agglomerated particulate solids contain no or low amounts of dust; therefore, they provide increased safety during the handling and processing of toxic or medicinally active materials, cause no workplace pollution or operator health risks, and, generally, result in fewer losses. Agglomerated mixtures of different particulate solids do not segregate. Agglomerated finely dispersed solids are freely flowing, therefore: Storage and handling characteristics are improved, Metering and dosing properties are better, Bulk density is increased and bulk volume is smaller. Within limits, the strength and porosity or density of agglomerates can be controlled; thus: disintegration, dispersibility, solubility, reactivity, and other properties of agglomerated intermediate or final products can be influenced. By coating or encapsulating, the agglomerate characteristics can be modified further; for example, solubility may be delayed or drug components activated by or in certain environments. Agglomerates may feature a specific shape, e.g. tablettes. Agglomerates have defined sizes and their distribution can be controlled. Monosized products or a defined amount of smaller agglomerates, the latter often enclosed in, for example, gelatine capsules, represent the dosage size. The final product has high consumer appeal and increased sales value.


The above are essential requirements because agglomerated products are supposed to facilitate measurement of or, in the case of tablets, define the dosage unit. Particularly for tablets and capsules, accurate weight is of utmost importance. Since, in most cases, the medicinally active component represents only a small fraction of the commercial drug form, uniform mixing and stabilization of the mixture (by pre-agglomeration) are necessary to avoid segregation. During manufacturing of the agglomerated specialties, a number of potential problems may have to be considered and overcome. These result from the specific properties of the highly sophisticated and often very sensitive active drug component. For example, the natural or chemically derived materials may not tolerate the presence of water or other liquids and may be sensitive to heat and/or pressure; for some substances shear forces must be avoided while still other materials are unstable or may discolor if they are brought in contact with certain chemical elements such as iron or various metal compounds. If intermediate or final pharmaceutical agglomerates are produced by tumbling, accretion, and growth, involving binder liquids, drying is required to obtain final and permanent bonding and strength (Section 6.2.1). Because many of the components of commercial drug systems are to a certain extent soluble in the liquid, not only the special case of drying of a porous body must be considered but also crust formation. Particularly if materials are temperature sensitive, which is often the case in pharmaceutical agglomeration, the phenomenon of incrustation during drying is of special importance. Agglomerates are porous bodies. The accretion and growth processes are normally controlled such that highly saturated agglomerates are produced; this means that a large portion (80–90 %) of the pore space is filled with the liquid. During drying, evaporation occurs at first only on the surface and liquid moves from the interior by capillary flow. This mechanism continues until, at about 15–25 % saturation, only liquid bridges remain at the coordination points between the particles forming the agglomerate. If the liquid flowing to the surface and evaporating there contains dissolved substances, these substances will crystallize at the pore ends and form a crust. As long as evaporation takes place, the temperature of the agglomerate remains low due to the heat of evaporation that is consumed during drying. Vacuum drying enhances this effect. However, it is possible that the crust becomes so dense that no additional liquid can reach the surface of the agglomerate from within while the core is still wet. If this happens, two alternative processes may occur. *

Because many dryers are controlled by the partial vapor pressure in the drying chamber or the dryer off-gas and will cease heating if this parameter falls below a certain value, the granules may be dry on the surface but still contain appreciable amounts of liquid inside. The same will be the case in vacuum dryers where low temperatures prevail. With time, for instance during storage of such insufficiently dried material, the residual moisture will migrate and cause lumping or, if the granules are destroyed, for example during tabletting, liquid will be liberated and cause a number of different problems.

87

88


If drying is carried-out at relatively high temperatures and, for temperature sensitive materials, the cooling effect caused by the removal of the heat of evaporation is used to keep the temperature of the granular mass at acceptable levels, this benefit is lost when formation of a dense crust occurs. As evaporation at the surface ceases the temperature in the agglomerate rises until the build-up of internal vapor pressure results in cracking of the crust and, at least temporarily, a continuation of drying, or the temperature will reach that of the dryer atmosphere. Both can be high enough to damage heat sensitive materials.

To avoid these problems care must be taken that crust formation is light. This can be accomplished by either limiting the contents of soluble materials in the formulation or producing agglomerates that contain only small percentages of liquid, thus realizing the bridge model. Incrustation can be also avoided if agglomeration and drying are taking place at the same time as, for example, in fluidized bed agglomeration when hot gas is used for fluidization. In this case drying and recrystallization of the dissolved substance(s) begin directly after liquid bridge bonding so that agglomerates never reach high liquid saturation, which is the precondition for capillary flow and crust formation. All concerns with drying and incrustation are eliminated if agglomeration is carried out dry by compaction/granulation or tabletting (Section 6.2.3). Other problems encountered in pharmaceutical applications are associated with the need to process small batches and to avoid cross-contamination when switching from one formulation to another. Therefore, equipment is typically small, designed for quick, easy, and complete cleaning, and frequently made from stainless steel or special materials with and without protective coatings (e.g., plating with chromium, nickel, or, in extreme cases, noble metals). Increasing concerns about worker protection lead to one-pot-technologies, in which several process steps, for example mixing, stabilization by agglomeration, drying, cooling, and de-dusting, are carried-out without the necessity to open the equipment to transfer the charge, and complete containment during operation is realized. “Through-the-wall designs”, whereby parts handling and processing the drug are totally separated from the technical components by an impermeable wall until a final, safe dosage form is produced, become increasingly common. This development requires new cleaning methods such as WIP (washing in place) and CIP (cleaning in place). For some applications where successful CIP is not possible and cross-contamination must be absolutely avoided, the use of specific equipment must be limited to only one formulation. Agglomeration is a technique by which particulate solids, comprising one or multiple components, are bonded together to form larger entities (granules, extrudates, tablets). According to the binding mechanisms of agglomeration, such permanent bonding can be accomplished by solid bridges or physical adhesion and chemical or molecular forces. Because of their short range, the latter require densification of the particulate mass to produce sufficient strength. Although, today, from dosage and/or marketing (appeal, acceptance) points of view, agglomeration of solid drugs is a necessity in most cases, compromises in regard to product characteristics must be often accepted. For instance, high strength (for stability during handling, packa-


ging, and storage) requires high density obtained, for example, by compaction or matrix binders which, in turn, results in reduced solubility. Therefore, special components are sometimes added, which assist in the break-up of the agglomerated product when it comes in contact with water or other liquids (e.g., effervescent or swelling components, fibers) [B.97]. On the other hand, agglomerated pharmaceutical formulations are ideally suited for the application of coatings to accomplish a multitude of tasks (Section 6.2.3). Such coatings may simply furnish a more pleasing shape and color, or enhance or mask the taste (e.g., by applying sugar coatings), and, generally, facilitate swallowing. Coatings can also provide characteristics that are required to render the drug medicinally more effective by, for example, retarding dissolution in the digestive system. These coatings are also applied by agglomeration methods (Section 6.2.3). Another “coating” technique is microencapsulation. It coats liquid droplets or solid particles with a skin and forms “microcapsules” with dimensions in the range 1– 5000 lm. The skin consists of natural or synthetic polymers and may be dense, permeable, or semi-permeable. Therefore, this technology allows the production of microcapsules containing a reactive substance, which can be liberated in a controlled fashion by destruction of the skin or by permeation. It is also possible to carry-out reactions within the capsules after permeation of reaction partners from the outside. The outer shape of the microcapsules depends on the type of material in the core and the method of depositing the wall material. Microcapsules may be smooth, spherical particles and grape-like conglomerates or irregular particulates with smooth or rough surfaces. Since all methods of size enlargement by agglomeration still need to be fine tuned by experimentation in a vendor’s testing facility (Section 9.1) or corresponding installations at an engineering company or the final user [B.97], the new, very strict requirements for process validation, which is part of the permitting and ongoing quality control in the pharmaceutical industry, put a severe burden on the development of new pharmaceutical products and the processes that include one or more agglomeration steps for manufacturing them. Because validation also includes the need to select and define equipment sizes and operating parameters of the final commercial systems already during the permitting phase (development) and scale-up is often not easy, definition of all the requirements that must be satisfied by the mechanical processes is challenging, to say the least. Also, difficulties with scale-up do normally not allow to build a larger plant if a new drug experiences unexpected success in the market. For these reasons, the trend in the pharmaceutical industry for the production of solid dosage forms is towards relatively small, standardized systems and processes, which can be already used and optimized during the development phase. Larger capacities that may become necessary during commercialization are obtained by installing identical additional systems at a central location or in regional facilities. Contract manufacturing (Section 9.2), where the burden of equipment validation is shifted to external manufacturers (Section 15.1), is also becoming more and more attractive and important. The manufacturing of primary components, both drugs and excipients, and intermediate feed materials, such as pre-agglomerated powders and powder mixtures, may still be produced in large central locations for a multitude of finishing plants.

89

90


As will be shown in the following sections 6.2.1, 6.2.2, and 6.2.3 all the main types of processes and techniques of size enlargement by agglomeration (Chapter 5) are being used in the pharmaceutical industry. Characteristic design features are based on what has been discussed above. The equipment is typically small, built from stainless steel, and can be cleaned easily inside and out to avoid cross-contamination. In addition, more and more installations are executed with a physical separation between the (dirty) drives and other mechanical parts that require technical maintenance and the containments that process the charge. Sophisticated control and recording devices are a common part of all systems, in part to satisfy the requirements of process validation. In accordance with the importance in the pharmaceutical industry of size enlargement by agglomeration and the use of agglomeration phenomena for other purposes, an increasing number of publications becomes available, particularly also in book form. Section 13.2 lists many of the more important books although it is not exhaustive and does not include works that where undoubtedly published in other languages.

Further Reading

For further reading the following books are recommended: B.3, B.6, B.7, B.8, B.16, B.19, B.21, B.22, B.24, B.26, B.33, B.34, B.36, B.40, B.41, B.48, B.49, B.51, B.56, B.58, B.64, B.66, B.67, B.68, B.70, B.71, B.72, B.73, B.78, B.82, B.89, B.93, B.94, B.95, B.97, B.99 (Chapter 13.1). Books mostly devoted to the subject matter are printed bold.

6.2.1

Tumble/Growth Agglomeration Technologies

As discussed in Chapter 2 and Section 6.2, the very first use of agglomeration in medicine was the making of pills from ground dry animal and plant matter, minerals, and other solid remedies to define the dosage and improve ingestion. Pills are spheroidal bodies that are produced from portioned moist, formable material by rolling with the flat hand or between two flat boards (Fig. 6.2-1). While wetting the medicinal mixture with water is often sufficient to render the material plastic and suitable for pill making, more sticky binders, such as honey, were frequently used. The sweet and relatively strong flavor of honey was also used to mask the often unpleasant taste of the concoctions. Although, in some countries and/or pharmacies, pill forming was and still is carried out by manual pressure during rolling, this technique can be considered the first application of wet agglomeration in pharmacy because the binding mechanism is caused by a liquid or viscous binder and, at least directly after manufacturing, a “green” agglomerate is obtained, which may be cured by drying or coated with a dry powder to reduce the product’s stickiness. Apprentice pharmacists still learn to manually make pills with the “pill machine” (Fig. 6.2-1). Since the first step of forming the mixture, which has been moistened in a mortar, is the production of a rope, which is then


divided into the portions to be rolled, today, industrial manufacturing of spherical dosage forms is performed by extrusion and spheronizing (Section 6.2.2). Granulation, to yield a free-flowing, dust-free, and non-segregating product, was also carried out by wet agglomeration for a long time. The formulation is blended, moistened, and passed through a screen by hand, and then dried to yield the granular product. Today, this is accomplished by low-pressure agglomeration with modern mechanized and typically motorized equipment (Section 6.2.2). Other wet agglomeration techniques, which were developed during recent decades for different applications, have been modified for use in the pharmaceutical industry and become important, as they can be particularly well designed to meet the special new requirements of validation, control, and documentation. They employ tumble/ growth agglomeration and are almost exclusively used for granulation [B.48, B.68]. In most cases, contemporary pharmaceutical wet agglomeration processes yield granules from pharmaceutical blends that are directly compressible and are, therefore, produced as free-flowing, dust-free, and non-segregating feeds for the modern high-speed tabletting machines (Section 6.2.2). Other applications manufacture a limited number of granular products that are either used directly, that is, for dosage by spoon or packaged in small envelopes, or as fill for gelatin capsules. Most other pharmaceutical specialties resulting from wet agglomeration feature instant characteristics (Section 6.4.1), for example mineral supplements and vitamin drink formulations. Shangraw [6.2.1.1] stated that the process of (pre-)granulation is historically embedded in the pharmaceutical industry. It produces in a single process (although many steps may be involved, see Fig. 6.1-13 and 6.1-14, and below) the two primary requisites for reproducibly making a high-quality compact/tablet (Section 6.2.2), that is, good flowability and compressibility. In more detail, the advantages of granulating a pharmaceutical press feed are summarized in Tab. 6.2-2. When all particles of a mixture are proportionally incorporated in agglomerates, thereby effectively stabilizing the blend, each granule has the ideal composition and any segregation of the granulated product has no effect on the ingredients in a tablet (1). By increasing the apparent particle size and sometimes also the sphericity of the granulated feed, the material becomes free-flowing, thus guaranteeing a conTab. 6.2-2 Advantages of (pre-) granulation in the pharmaceutical industry (adapted from Shangraw [6.2.1.1]) 1. No segregation: Permits handling and feeding without loss of mix quality. 2. Freely flowing: Improves flow of powders by increasing the particle size. 3. No dust: Reduces the level of workplace dust and cross-contamination. 4. Good bulk density: Increases and improves uniformity of feed density. 5. Improved degassing: Reduces the amount of air entrapment. And for wet granulation processes only 6. Better bonding: Improves cohesion during and after compaction. 7. Liquid addition: Allows for (small!) addition of liquid phase to powders. 8. Surface modification: Makes hydrophobic surfaces hydrophilic.

91

92


stant fill of a container (for example gelatin capsule) or the die cavity(ies) of a tabletting machine (Section 6.2.2) even at very high fill speeds (2). Since, by definition, dust particles are bonded within the agglomerate and attrition is minimized, the level of dust in the manufacturing area is low. Health risks to the workers and cross-contamination between production lines is essentially eliminated (3). Although components of the formulation may have different densities, all are incorporated in the granules, which themselves feature a uniform, average density. Within limits, the bulk density of the granulated feed can be adjusted by controlling the agglomerate size range through post-treatments (e.g., screening) or directly in the process (4). Because the agglomerates are porous and feature a certain strength, densification first takes place by rearrangement of the relatively big granules (Fig. 5-9 and Fig. 10-15) whereby the large amount of gas (air), expelled in this phase, can easily escape from the diminishing interparticle volume. Only when the pressing force approaches the high-pressure region (Fig. 5-9 and Fig. 10-15), are the granules destroyed and the original powder particles incorporated into the uniform tablet structure. Porosity and pore sizes are drastically reduced resulting in a considerable increase in diffusion resistance. However, because in this final densification phase only a relatively small amount of air remains to be displaced, the volume of trapped, compressed gas in isolated pores, if any, is small and does not negatively influence tablet quality (5). The above are valid for all granulation processes, wet or dry (below and Section 6.2.2). If wet granulation is selected, additional advantages may apply. In many cases, a thin layer of binder (added as a liquid, in suspension, or in solution) forms around substances that have poor bonding characteristics. This effect, often called functionalizing (Fig. 10-14), improves cohesion during and after tabletting (6). Although, typically, granulated products are dried prior to their use in compaction, liquids are added during mixing and/or agglomeration. This allows for the addition of drugs or other ingredients (e.g., dyes) in solution thereby achieving a more homogeneous distribution (7). By chemical or physical surface modification, initiated by special liquid phases added during wet granulation, hydrophobic substances can be made more hydrophilic (Section 10.1), which improves the dissolution rate of the tabletted products (8). One technology is wet granulation by mixer agglomeration [B.48, B.68, B.97] which, after adaptation from other industries, has evolved from simple drums with mixing elements and integrated spray-nozzles for liquid additives and/or binders to highly sophisticated, specially designed, equipped, and instrumented “one pot” granulating and processing systems. Another utilizes spray drying and/or fluid-bed agglomeration [B.48, B.49, B.68, B.93, B.97] One of the first high-shear mixers, suitable for pharmaceutical applications, was the L€odige drum with plows as mixing tools on a horizontal shaft and high-speed choppers in the drum shell. After expiration of the basic patents, in addition to the original manufacturer, this design is now offered by many vendors worldwide. Fig. 6.2-2 shows this equipment. The size, number, positioning, geometric shape, and peripheral speed of the mixing elements are selected to achieve a three-dimensional movement of the particulate solids within the mixing drum [6.2.1.2]. The resultant turbulence with total particle mobility prevents the formation of dead zones and results in gentle precision mixing and agglomeration within the shortest possible time. During


wet granulation, particularly if liquid is added, the separately driven high-speed choppers are operated to destroy larger lumps that may have formed and control agglomerate size [B.48, B.97]. For pharmaceutical applications small to medium sized, versatile, and often modular equipment is desired that must meet extreme cleanliness requirements (below). Fig. 6.2-3 shows a laboratory or small production dust-tight plow mixer/agglomerator with cantilevered shaft, interchangeable drum, easy and quick opening front door for good accessibility, chopper, and liquid addition. Larger cantilevered units are often mounted on the wall to separate the potentially contaminating technical parts, such as drives, from the clean room in which the pharmaceutical formulations are handled and processed (Fig. 6.2-4). In Fig. 6.2-4, right, the front door is opened, allowing a view into the interior of the polished stainless steel drum with shaft, plows, and chopper element. To further facilitate cleaning, the shaft with the mixing elements may be fitted with a pull-out mechanism (Fig. 6.2-5). Furthermore, particularly for larger mixer/granulators, WIP (washing-in-place) can be added: Fig. 6.2-6 is a P&I (process and instrumentation) diagram of such a machine. Horizontal, batch and continuously operating drum mixer/agglomerators with plows and many other mixing tools and chopper elements [B.48, B.68, B.97] are widely used in the pharmaceutical industry, mostly for the pre-granulation of formulations prior to tabletting. The addition of steam jackets or other heat sources and/or the application of vacuum allows single-pot processing (below), which is a fast growing technology in the pharmaceutical industry. Because the equipment must not be opened between different process steps, it avoids or minimizes contamination of both the charge and the workplace. A newer, but widely applied design is the vertical high-shear system, often called “bowl” mixer/agglomerator (Fig. 6.2-7) The shape of the container promotes formation of a vortex flow and the mixing tool has minimum clearances to the inner equipment walls for maximum product yield. This mode of operation assures rapid, inten-

Fig. 6.2-2 Diagram of the operating principle of a horizontal plow mixer/agglomerator with chopper, and two alternative means of liquid addition (courtesy L€ odige, Paderborn, Germany)

93

94

6 Industrial Applications of Size Enlargement by Agglomeration Fig. 6.2-3 Laboratory or small production dust-tight plow mixer/ agglomerator with cantilevered shaft, interchangeable drum, front opening door, chopper, and liquid addition (courtesy L€ odige, Paderborn, Germany)

Fig. 6.2-4 Wall mounting of larger cantilevered mixer/agglomerators in a pharmaceutical manufacturing facility (courtesy L€ odige, Paderborn, Germany)


Fig. 6.2-5 The activated pull-out mechanism, exposing the shaft and plow mixing elements of a mixer/agglomerator for easy cleaning (courtesy L€ odige, Paderborn, Germany)

sive movement of the entire charge even if components have diverse bulk densities and particles feature widely different shapes and sizes. Often, as shown in Fig. 6.2-7a, the impeller can be lifted, sometimes hydraulically, for improved cleaning. A chopper (or multiple ones) is located such that it extends into the zone of greatest material velocity to perform the same functions as described previously.

Fig. 6.2-6 P&ID of a horizontal, batch operating plow mixer/agglomerator, equipped for WIP (courtesy L€ odige, Paderborn, Germany)

95

96


Fig. 6.2-7 Typical designs of vertical high-shear (bowl) mixer/agglomerators: a) basic design, b) equipment for “one-pot processing”, c) process diagram (courtesy Diosna, Osnabr€ uck, Germany)

Cleaning requirements in the pharmaceutical industry may necessitate installation of the chopper(s) through a removable roof as shown in Fig. 6.2-7c. This apparatus is also designed for efficient mixing, granulating, gas stripping, and vacuum drying in a “one pot” manner. In this case, to avoid condensation, the bowl and lid are doublewalled and heated. Fig. 6.2-8 depicts the complete design of a modern one-pot mixing, granulating, and drying system in which the particular advantages of microwave drying are utilized. In this drying method, the internationally standardized microwave energy of 2450 MHz causes the water molecules in the moist agglomerates to vibrate at high speed. Heat that results in evaporation is generated by friction between the water molecules throughout the mass to be dried. Therefore, drying does no longer occur from the outside by transfer and conduction of heat; it now proceeds at a faster rate, particularly if the solids exhibit poor heat conductivity. For pharmaceutical applications, cleanliness and avoiding cross-contamination are important requirements. Fig. 6.2-9 shows that even the smallest vertical high-shear mixers can be executed such that the stainless steel vessels are exchangeable to accommodate different materials by dedicating a specific bowl to a particular formulation or to allow easy external cleaning. Larger units may feature a similar design. Although


Fig. 6.2-8 Diagram of one-pot mixing, granulating, and drying system featuring microwave drying (courtesy FUKAE Powtec Corp., Kobe City, Japan)

Fig. 6.2-9 Small vertical highshear mixer/agglomerator with exchangeable stainless steel vessels of 1, 2, 4, and 8 L volumes (courtesy Diosna, Osnabr€ uck, Germany)

97

98

6 Industrial Applications of Size Enlargement by Agglomeration Fig. 6.2-10 Typical through-thewall installation of a mid-sized vertical high-shear mixer/agglomerator (600 L bowl) with container feeding through the ceiling in a pharmaceutical processing plant (courtesy Diosna, Osnabr€ uck, Germany)

vertical high-shear mixer/agglomerators always operate in batch mode, new equipment sometimes has a rather large volume (up to 2000 L) of which, depending on the material and application, 30–80 % is useable per batch. This allows large production rates because, typically, processing times are short (Fig. 6.2-10). In addition, if the drying process is carried-out externally, as shown in Fig. 6.2-11, in which a fluidized bed dryer is applied in line, a quasi-continuous process is obtained. In such an arrangement, a closed system from loading the raw materials to the discharge of dry granular product is also achieved. It has been already mentioned that in order to maintain a clean room environment in the processing department, “through-the-wall installation” is increasingly applied. The execution of all equipment for the granulation of pharmaceutical formulations, whether employing wet tumble/growth or any of the pressure agglomeration methods, can be either free-standing or through-the-wall, depending on the specific cleanliness requirements. Fig. 6.2-12 depicts the difference between the two system designs, using a simple vertical high-shear mixer/agglomerator as example.


Fig. 6.2-11 A quasi-continuous system for mixing, granulating, and drying in a pharmaceutical setting featuring a bowl mixer/granulator of 600 L volume

and an external fluidized bed dryer (courtesy Diosna, Osnabr€ uck, Germany)

Another commonly desired equipment feature in the pharmaceutical industry is a modular design. As already shown (Fig. 6.2-9) one of the reasons for this is the easy exchange of machine parts for cleaning or the commitment of, for example, a vessel or tooling to only one material to avoid any cross-contamination. The latter is particularly necessary if the machine parts are complicated and not easily washed and/or disinfected. Modular attachments (Fig. 6.2-13) also allow modifications to include, for

99

100


Fig. 6.2-12 Diagram of through-the-wall and free-standing installations of a vertical high-shear mixer/agglomerator (courtesy Diosna, Osnabr€ uck, Germany)

example, vacuum processing with and without solvent recovery, CIP, different lids with or without filter attachments, various internal tools, instrumentation and so on. Although the above discussions were limited to two types of high-shear mixer/agglomerator, the irregular, stochastic movement that is required for high-performance blending of particulate solids in all types of blenders also produces ideal conditions for growth agglomeration by coalescence. Agglomeration (granulation) is achieved in the apparatus during or after the mixing phase by a controlled addition of binder. In the pharmaceutical industry, maintaining a uniform distribution of all components in the granular product is of particular interest since an often extremely small amount of very finely divided active substance (the drug) must be mixed uniformly and reliably with a relatively large amount of inert filler material (excipient) and segregation avoided by stabilization through agglomeration.

Fig. 6.2-13 Diagram of modular design features offered with a vertical high-shear mixer/agglomerator (courtesy H€ uttlin, Steinen, Germany)


Fig. 6.2-14 Diagrams, including indication of particle movements, and photographs of different batch-operating low-shear mixer/ agglomerators: a) double-cone (courtesy Abbe´, Little Falls, NJ, USA), b) slanted-cone (courtesy Gemco, Middlesex, NJ, USA), c) V-shape (courtesy Abbe´, Little Falls, NJ, USA) [B.97]

For some applications, low-shear, batch-operating mixer/agglomerators, such as double-cone, slanted-cone, or V-shape blenders (Fig. 6.2-14) are successfully applied in the pharmaceutical industry if the particles are soft or brittle and degradation during blending should be avoided. However, considerable problems can arise if components of the mixing and agglomerating tumbling mass have different characteristics and/or sizes. In such cases, particles that, for any reason whatsoever, feature higher adhesion tendency and/or the smaller size fraction(s) of the formulation may selectively agglomerate, thus making it impossible to achieve the ideal mixture in the granulates. The gentle tumbling in low-shear mixers does not produce sufficiently high forces to destroy these agglomerate and make the particles available for renewed, more uniform attachment to granules. The incorporation of choppers, which in mixers with stationary vessels and moving tools help in the destruction and rebuilding of agglomerates, is difficult but can be accomplished if absolutely necessary (Fig. 10-11). Aside from the gentler processing, the low-shear mixer/agglomerators produce loosely assembled granules with low density and high bulk volume. The first increases somewhat as larger agglomerates acquire more mass (Fig. 6.2-15). Furthermore, it is more difficult to achieve uniform distribution of the binder liquid. Upon impact with the powder mass, droplets wet larger areas (Fig. 5-5) and, because of the low shear, the

101

102


liquid remains trapped in this volume. Although, during tabletting, the low-density intermediate products require longer punch strokes and higher amounts of gas need to be removed, necessitating some dwell time (Section 10.2), the compacted solid dosage forms obtained from such granules often feature improved structure, higher porosity, and better dispersibility. Combinations of high-shear blender and low-shear agglomerator, the P-K blender agglomerator, for example, unite continuous operation with a more uniform binder distribution. As shown in Fig. 6.2-16, this design consists of a slowly turning eccentric drum to which a V-shaped (zigzag) shell is mounted. Inside the drum is a dispersion head that rotates at high speed, aerates the powder, and supplies finely atomized liquid via special slots [B.97]. The charge is uniformly wetted and seeds are formed, which through repeated internal recycle from the zigzag portion back into the drum, grow uniformly into loosely bonded agglomerates. During the forward flow in the final part of the V-tube, granules are rounded and grow somewhat more. Because near the discharge end of the machine essentially no shear occurs, some segregation and selective agglomeration may still be experienced. If granulated pharmaceutical formulations with uniform composition and low density are required, application of the fluidized bed [B.41, B.48, B.49, B.68, B.93, B.97] has become increasingly prevalent. The technology, in this industry originally mostly used for the gentle and uniform drying of green agglomerates (Fig. 6.2-11), has lately emerged as an important method for the production of granulated formulations. Because the agglomerated particles are relatively small and have low density, most of the instant pharmaceutical specialties are now produced by fluidized bed processes. Particularly if dry powder is produced in a spray dryer plant from solutions, suspensions, or slurries, agglomeration can be accomplished if the partially solidified but still moist particles are tumbled in an associated fluidized bed where, in most cases, final drying also takes place. Fig. 6.2-17 is the schematic flow diagram of a continuous fluidized spray dryer (FSD). As compared with the conventional spray dryer [B.48, B.49, B.71, B.93], a somewhat modified gas handling system is the most obvious new feature of the FSD. Drying gas (9) not only enters the top of the tower for cocurrent drying but also a so-called “plenum”, a specially designed chamber at the bot-

Fig. 6.2-15 Typical agglomerates that were produced in a batch low-shear mixer/agglomerator with V-shaped shell. Agglomerate sizes: left, about

0.750 mm; right, about 4 mm (courtesy PattersonKelly, East Stroudsburg, PA, USA)


Fig. 6.2-16 A P-K zigzag continuous blender/agglomerator (courtesy Patterson-Kelly, East Stroudsburg, PA, USA)

tom of the tower, from which the hot gas is introduced through a distribution plate. The amount of drying gas entering with the dispersed feed (3) is controlled so that, while the droplets descend in the tower, only partial drying is accomplished. The still partially wet, slightly sticky particles are captured in a fluidized bed (5) where they

Fig. 6.2-17 Diagram of a continuous fluidized spray dryer (FSD) with open plant gas flow (courtesy GEA/NIRO, S€ oborg, Denmark) [B.97]

103

104


collide and form larger agglomerates. Fines (7) that are removed from the off-gas (10) in a dust collector (6) are recirculated to the fluidized bed where they are attached to the growing agglomerates. While the solids tumble in the fluidized bed and grow by agglomeration they are also dried with the hot fluidizing gas. Dry, agglomerated product (8) is removed from the fluidized bed in a suitable manner [B.97]. In the pharmaceutical industry, fluidized bed technology is also commonly used for the granulation of powders or mixed powder formulations by re-wet agglomeration. For reasons of containment and cleanliness, batch operating is the preferred process execution. Fig. 6.2-18 shows the principle of the process (right) and the outline of an apparatus (left) depicting, from bottom to top, a container for collecting the product, the plenum with gas distribution plate, the tower in which the fluidized bed develops and liquid binder is sprayed onto the charge and integral bag filters for capturing

Fig. 6.2-18 left) The principle of batch fluidbed granulation; right) the outline of an apparatus (courtesy Glatt, Binzen, Germany)

Fig. 6.2-19 P&ID of a typical batch fluid-bed granulation system (courtesy Glatt, Binzen, Germany)


Fig. 6.2-20 Germany)

Two batch fluid-bed granulation units (courtesy Glatt, Binzen,

entrained fines. Fig. 6.2-19 is the process and instrumentation diagram of a typical installation showing the supply of hot air into the plenum at the bottom and the discharge of gas through filters at the top. The flow is adjusted and balanced by means of a number of automatically controlled dampers (called “control flaps” in the figure). The top-spray arrangement for binder liquid is clearly visible. Fig. 6.2-20 shows actual equipment and Fig. 6.2-21 presents a typical product. Another advantage of the fluidized bed technology is the relatively simple shape of the processing chamber which, in general terms, is a vertical tower consisting of several easily removable and exchangeable sections. For example, as seen in Fig. 6.2-22, two mid-sections may be installed on the frame of the apparatus, designed for quick exchange. While one is in operation the other can be cleaned or modified. It is also easy to perform CIP, a fully automatic, reproducible cleaning process without the need to open the apparatus (Fig. 6.2-23a). Of course, problem zones such as the filters, distribution plate, sealing joints, and bulls eye windows must be specially adapted.

105

106

6 Industrial Applications of Size Enlargement by Agglomeration Fig. 6.2-21 Acetaminophen (paracetamol) granules produced in a top-spray fluid-bed processor: a) surface, b) cross section, revealing internal voids (compare with Fig. 6.2-32). Magnification 60 (courtesy Glatt, Binzen, Germany)

Fig. 6.2-22 Batch fluid-bed granulation unit with two mid-sections installed on the frame for easy and quick exchange (courtesy Glatt, Binzen, Germany)

6.2 Pharmaceutical Applications Fig. 6.2-23 CIP of fluid-bed granulation units: a) principle of automatic cleaning, b) principle of washable pulse blow-back filters and view into the filter area of a unit, c) metal cartridge filter, d) hydraulically extendable washing nozzles (courtesy Glatt, Binzen, Germany)

The filters are a particular concern. Fig. 6.2-23b shows schematically and as a view into the filter area the washable pulse blow-back filters of one manufacturer. The shape of the metal filter cartridge (Fig. 6.2-23c), which is entirely made from stainless steel, combines optimum filtration with outstanding cleaning properties and maximum durability. The spray nozzles for washing are installed in the shell of the apparatus but, because they would disturb the material and gas flows if they were protruding into the chamber during operation of the fluidized bed, they are retracted or flush with the wall inside when not in use and extend automatically if connected with the line providing pressurized cleaning liquid (Fig. 6.2-23d).

107

108

6 Industrial Applications of Size Enlargement by Agglomeration Fig. 6.2-24 SC (“SuperClean”) fluidized bed granulator equipped for CIP and total containment (courtesy Glatt, Binzen, Germany)

The ability to perform WIP or CIP is part of the process validation in the pharmaceutical industry. While WIP means a thorough automatic pre-cleaning but, to achieve the required results, still needs opening of the unit for a manual final cleaning, CIP is a fully automated, reproducible cleaning process with a defined result. Both are applied depending on machine design and material to be processed. Equipment for the latter is specially executed (Fig. 6.2-24) and must not be opened between campaigns or when changing from one formulation to the other. Because there may be several units in a production facility that periodically undergo WIP or CIP, the equipment supplying washing liquid with the required quantity, pressure, and temperature is often mounted on a skid (Fig. 6.2-25), which is moved and connected to a particular apparatus as required. The unit may also include the metering of the correct amount of cleaning agent(s) into the liquid.


Fig. 6.2-25

WIP/CIP skid (courtesy Glatt, Binzen, Germany)

Fig. 6.2-26 Diagram of the principle of continuous fluid-bed drying and granulation (courtesy Glatt, Binzen, Germany)

109

110


For certain applications, the fluid bed can be operated continuously. Fig. 6.2-26 depicts the principle of this design. In most units a liquid raw material is dried while particle size is building up. Entrained fines, collected with external filters, are recirculated. It is also possible to continuously add other particulate solids. Both act as seed material for the granulation process. The fluid bed guarantees that all raw materials are homogeneously mixed in the agglomerates, which are discharged continuously through a centrally arranged pipe in the circular distribution plate. The final product size is determined by the velocity of air flowing upwards in the discharge pipe. Fluidized bed equipment lends itself particularly well to vacuum processing. Fig. 6.227 is the process and instrumentation diagram of a batch vacuum top-spray fluid-bed granulator with liquid (solvent, binder) recovery. The advantages of vacuum processing are: * *

* *

*

constant conditions, independent of atmospheric influences, a reduction of the vaporization temperature and shortening, up to 90 %, of the drying time (Fig. 6.2-28), the production of granules with higher porosity (Fig. 6.2-29), an expensive system to provide an inert gas atmosphere, if necessary in “open” systems, is not required, and up to 99 % of the liquid used in the process can be recovered and, in most cases, reused.

Fig. 6.2-30 shows photographs of two batch vacuum fluid-bed granulators with liquid recovery, indicating the extent of the vacuum and recovery systems. In a relatively new development, the distribution plate in a batch fluid-bed granulator is replaced by a solid rotating plate fitted into a conical bottom section [B.97]. Process air enters the apparatus through the annular space between the plate and the container walls. Its amount can be varied by changing the annular gap width

Fig. 6.2-27 P&ID of a batch vacuum top-spray fluid-bed granulator with solvent recovery (courtesy Glatt, Binzen, Germany)

6.2 Pharmaceutical Applications Fig. 6.2-28 Comparison of the drying temperatures and times achieved in contact and vacuum fluid-bed dryers (courtesy Glatt, Binzen, Germany)

by moving the plate up or down. Fig. 6.2-31 depicts the principle and the outline of an apparatus. The rotation of the plate produces a torus-like movement, which rounds and densifies agglomerates much in the same manner as in a spheronizer (Fig. 6.2-65). Fig. 6.2-32 shows microphotographs of a pharmaceutical product which, in spite of some densification, still features a considerable amount of internal void spaces. A similar rounding effect but without the same high densification can be obtained in a continuous fluid-bed drying/granulation process in which the bottom plate features rings of different perforations to modify airflow (Fig. 6.2-33). The airflow pattern induces a rotating ring of material in which the liquid droplets impinge solids and are dried. Granules are growing, rounded, and classified before they dis-

Fig. 6.2-29 SEM image of a granule produced in a vacuum fluid-bed granulator (courtesy Glatt, Binzen, Germany)

111

112


Fig. 6.2-30 Two vacuum batch fluid-bed units with closed loop and solvent recovery systems (courtesy Glatt, Binzen, Germany)

charge through the central exit pipe (Fig. 6.2-26) and are cooled in an external fluid-bed cooler. Fig. 6.2-34 depicts the process and instrumentation diagram of such a process and Fig. 6.2-35 shows photographs of an apparatus and a typical product. Worldwide, there is a growing number of manufacturers that offer specialized equipment for the pharmaceutical industry using tumble/growth agglomeration/ granulation, particularly in mixers and fluid beds. Special process and equipment execution of all types in this industry (see also Sections 6.2.2 and 6.2.3) is driven by an extreme need for cleanliness and containment and by government-imposed registration and validation [6.2.1.3]. Stainless steel and other, sometimes exotic materials of construction, smooth external and internal surfaces, special seals, valves, probes and instrumentation, modular design, and one-pot processing, during which multiple steps are performed in one container or a sealed production line without the

Fig. 6.2-31 Diagrams of the principle and outline of a fluid-bed granulator with solid rotating bottom plate (courtesy Glatt, Binzen, Germany)


Fig. 6.2-32 Acetaminophen (paracetamol) granules produced in a rotary fluid-bed processor: a) surface, b) cross section, still revealing internal voids (compare with Fig. 6.2-21), magnification 60 (courtesy Glatt, Binzen, Germany)

Fig. 6.2-33 Principle of a special rounding continuous fluid-bed drying/granulation process (courtesy Glatt, Binzen, Germany)

Fig. 6.2-34 P&ID of a rounding continuous fluid-bed drying/ granulation process (courtesy Glatt, Binzen, Germany)

113

114


Fig. 6.2-35 A rounding continuous fluid-bed drying/granulation process apparatus and a typical product at two different magnifications (courtesy Glatt, Binzen, Germany)

need to open, handle, and transfer intermediate product forms, are the most distinguishing characteristics of pharmaceutical processing and equipment. While the techniques of only a few vendors are being introduced and discussed above and in Sections 6.2.2 and 6.2.3, many additional suppliers are available, some of which are mentioned in Section 15.1.

6.2.2


As already discussed, the rolling of pills from moist and sticky medicinal blends is most probably the oldest application of agglomeration in pharmacy although, to the author’s knowledge, there is no historical document available that proves its exact origin. These older forms of solid dosage are the direct precursors of today’s primary solid pharmaceutical product, the tablet, in all its forms and presentations. While references in the Egyptian “Papyrus Ebers” already seem to confirm the use of some sort of agglomerated preparation, the manufacturing of “pills” was first reported by the Greeks and Romans well before the beginning of our calendar and in Persia and Arabia in the 9th and 10th centuries AD (Tab. 6.2-3). Such agglomerates were also often coated to improve their taste and/or appeal. Originally pills were, and sometimes still today are, rolled by hand whereby manual pressure is applied directly or with rolling boards and discs (Fig. 6.2-1). During the

6.2 Pharmaceutical Applications Tab. 6.2-3 [B.48]

Early history of tabletting for pharmaceutical specialties

Time

Source/Inventor(s)

Shape or manufacturing method

BC BC 850 – 923 980 – 1037 Tenth century 1448 1606 1837 1837

Greeks Romans Rhazes al-Razi (Persia) Avicenna (Perisa) Al-Zahrawie (Arabia) Apothecary, Florence (Italy) Jean de Renou (France) M. Labelonie (France) A. Fortin (France)

1838 1840

M. Deschamps (France) E. Mayer, F. Roman (France)

1843 1848 1874

W. Brockedon (England) US Dispensatory, 12th edition (USA) J.A. McFerran (USA)

1874

Th.J. Young (USA)

1875 1876

J.P. Remington (USA) J. Dunton (USA)

1877 1877

G. Gercke, Jr (Germany) Th.J. Young (USA)

1878

Ch. Charter (USA)

1881

A. Edler von Hofmann (Germany)

1882

J.T. Jones (USA) Ch.T. Jones

1885 1889

J. Lusby (USA) O. Smith (USA) H.K. Mulford

1891 1894 1895 1896 – 7

W. Kilian P.M. Justice (England) F. Kilian (Germany) P.J. Noyes (USA)

1897 1898 1898

P.E.M. Jamain (France) W. Kra¨mer (Germany) W.R. Dodd (USA)

1900

F. Kilian (Germany)

1901

Allen and Hanburys Ltd (England)

1903

Henning and Martin (Germany)

’Katapotia’ ’Pilulae’ Pills with Psyllium seed Silver- and gold-coated pills Pastille shapes Silver- and gold-coated pills in Europe ’Tabellae’ Sugar-coated polls (dragees) Patent 5116 for the manufacture of sugar coatings of pills Honey and Acacia powder Patents 6222 and 6449 for the manufacture of pill coatings of sugar and Acacia powder English Patens 9977 for a tabletting press Manufacturing of dragees US Patent 152666, tabletting press (improvement in pill machines) US Patent 156398, improvement in machines for making pills, lizenges, etc. Tabletting press US Patent 174790, tabletting press (improvement in pill machines) German Patent 5006, rotating press US Patent 189005, improvement in machines for marking pills US Patent 207013, improvement in coated, compressed pills, lozenges, etc. German Patent 15535, machine for pressing of flowery, powdery, or grainy materials US Patent 256573, machine for making pills, lozenges, etc. US Patent 323349, pill-making machine German Patent 54817, machine for the production of pills German Patent 63185, ’pastille’ press German Patent 81470 pastille and pill press German Patent 88514, ’pastille’ press US Patents 568488 and 582794, press for sugar-coating pills, pill machine German Patent 99282, pastille press German Patent 112286, tablette press German Patent 113018, press for the manufacturing of pills, pastilles, tabletts, etc. German Patents 120903 and 126493, rotating table press German Patent 146340, press for the manufacture of medicinal tablettes German Patent 158023, press with rotating table and punches

115

116

6 Industrial Applications of Size Enlargement by Agglomeration Tab. 6.2-3

(continued)

Time

Source/Inventor(s)

Shape or manufacturing method

1904

F. Kilian (Germany)

1906

F.F.W. Stieler (Germany)

1907

Th. O. Kend (England)

1908

F. Kilian (Germany)

1914 1917

Tietz and Co. (Germany) F.J. Stokes (USA)

German Patents 160550 and 165401, press with rotating table German Patent 190355, method for making pastilles in presses with movable upper and lower punches German Patent 202270, manufacture of pharmacuetical tablettes with a movable upper and lower punch German Patent 204824, press with rotating table and upper punches which are held coaxially in relation to the respective lower punch German Patent 287776, coating press US Patent 1248571, press for the manufacture of coated tablets

second part of the 10th century AD, Arabian “pharmacists” used a simple manually operated press to form moist powder [6.2.2]. Material was compacted between the two halves of a tongue-shaped tool made of bone, ivory, or wood. For many centuries, these forms of solid medicines for oral application had been well accepted by the patients and were produced in relatively large numbers. Nevertheless, the invention of the tabletting die press in 1843 by William Brockedon (Tab. 6.2-3) initiated a revolution in the manufacturing of pharmaceutical solid dosage forms. Although very simple and manually operated, his patent (British Patent 9977, 1843) entitled Shaping pills, lozenges, and black lead by pressure described the still-valid basic principle of punch-and-die presses, Fig. 6.2-36 [6.2.2.1]. It took some time until this method was accepted but the pace accelerated in 1874 when two additional patents for “improved pill machines” were granted in the USA. Both new machines were suitable for use with a belt drive. Thomas J. Young’s patent (US Patent 156 398, Fig. 6.2-37 [6.2.2.1]) used an eccentric movement that was disengaged when the upper punch reached its highest position. After manually filling the cavity with powder, the operator connected the eccentric drive with the fly-wheel causing the upper punch to descend, compact the powder, and then withdraw to the highest position where the drive was automatically disengaged. After manually pushing out the tablet, the cycle could begin once more. Joseph A. McFerran’s machine (US Patent 152 666, Fig. 6.2-38 [6.2.2.1]) already used a round, indexed table carrying several dies with their associated lower punches. A spindle drive, automatically turning left and right, moved the upper punch up and down and the press table was moved ahead by one step when the upper punch reached its highest point. Both filling the die and removal of the finished tablet were also already accomplished automatically. After that time, further improvements were invented around the world in rapid succession (Tab. 6.2-3). They were aimed at more accurate tablet shape and weight and increased the capacity of each unit, mostly by additional state-of-the-art automa-

6.2 Pharmaceutical Applications Fig. 6.2-36 Patent drawing of Brockedon’s hand tool for tabletting. British Patent 9977, 1843

Fig. 6.2-37 Patent drawing of Young’s eccentric drive tabletting machine. US Patent 156 398, 1874

117

118

6 Industrial Applications of Size Enlargement by Agglomeration Fig. 6.2-38 Patent drawing of McFerran’s indexed table tabletting machine. US Patent 152 666, 1874

tion. While the basic principles, defined in the early inventions, remained largely unchanged, improvements which, during the past decades, have been strongly influenced by Government imposed cGMP (current good manufacturing practice), validation, and cleanliness requirements, result in the design and offering of quickly and perpetually improved compaction equipment for the pharmaceutical industry [B.34, B.41, B.48, B.66, B.71, B.97, B.99, B.105]. Today, high-pressure agglomeration is still the most widely used technology for obtaining dry pharmaceutical dryosage forms, so-called tablets. Because individual units are small, with diameters usually in the range 5–17 mm and weights of 0.1– 1.0 g, and must be of uniform, highly reproducible shape, only machines operating with punches and dies are applicable. Although other punch-and-die presses, such as simple ejection and withdrawal machines [B.97], are used for special applications and for research and development, to meet the throughput requirements of modern pharmaceutical production, measured by the number of tablets made per unit time, special equipment, so-called rotary tabletting machines [B.6, B.48, B.66, B.97, B.99, B.105], has been introduced, which is capable of producing up to 1 million compacts per hour.

6.2 Pharmaceutical Applications Fig. 6.2-39 The layout of a rotary punch-and-die press [B.71, B.97]

The basic operating principle of rotary punch-and-die presses is similar to that of all reciprocating ejection machines. The difference lies in the fact that a series of dies is mounted into a circular steel table (the so-called “turret”) near its periphery (Fig. 6.239) and that two punches (one upper and one lower, Fig. 6.2-40) are associated with each die. The punches are moved by stationary cams while the turret with the dies and punches is rotating. An evoluted presentation of one pressing cycle is shown in Fig. 6.2-40. Feed is supplied to the table through a frame, often called the “feed shoe”, which may include a mechanical, in most cases rotating flow stimulator. The feeder is connected to a hopper above. When a particular die moves under the feed shoe, the bottom punch that is associated with that die is pulled down to its lowest position by its cam thus allowing the die to fill with powder. The punch then rises up an adjustable ramp to eject excessive powder from the die. The powder surplus is scraped off flush with the top of the turret at the highest point of the “weight adjustment ramp”. Assuming uniform fill density, this always leaves the same volume of powder to be compacted in the die. It is common practice to let the lower punch drop down slightly after the surplus material has been scraped off. This is done to prevent uncontrolled displacement or “blow-out” of powder from the die when the upper punch enters. Both punches are then moved together by their cams to achieve densification and compaction. If, optionally, the ramps moving the punches remain parallel for some distance after reaching maximum densification, a so-called ’dwell time” is introduced. During this time, the compact remains under pressure so that additional deaeration and conversion of elastic deformation into permanent plastic deformation can occur and expansion upon pressure relief is minimized (below and Section 10.2). The overall opposing movements of both punches during densification and compaction produce the effect of double pressure and, therefore result in a relatively uniform structure of the tabletted product (Fig. 6.7-23). Finally, the upper punch is lifted from the die and the lower punch travels up to eject the finished compact. As shown in Fig. 6.2-40b, another but more technically detailed evoluted presentation of a typical high speed, high-pressure rotary tabletting machine, quite often, the maximum pressure is produced by one (or two) set(s) of press rollers that oppose each other. One or both are supported by springs to provide overload protection. In such machines, the final compaction takes place very quickly and is followed by a sudden

119

120


Fig. 6.2-40 a) Evoluted (straightened) diagram of a rotary punch-and-die press. b) Paths of the punches in rotary punch-and-die (tabletting)

presses with one or two sets of press rollers in evoluted presentation [B.71, B.97]


pressure relief. This is similar to what happens in roller presses [B.13b, B.48, B.97] but, because in tabletting machines the roller diameter is very small and the table speed is high, this process takes place extremely fast. Therefore, capping (below) is a commonly observed problem in high-speed tabletting. To avoid this and still maintain high-production capacity, the preparation of special particulate feeds by pre-granulation (below) is frequently a necessity to overcome this defect. The simplest type of rotary machine is “single sided” with one feed location and a certain number (as few as four) of “stations” (dies) on the table. One rotation of the turret produces as many compacts as there are dies (and punch sets) on the machine. Therefore, the output of single sided rotary machines depends on the maximum allowable speed of and the number of stations on the table. It is normally in the range 300–800 tablets per minute. It can be doubled by installing two feed locations. In this case, the stations are filled twice on opposite sides of the rotating table and two compressions are carried out in each die per revolution of the turret. Obviously, to maintain the rate of densification and compaction the number of stations on the correspondingly larger table would have to be doubled, too. Outputs of more than 3000 tablets per minute can be obtained from well compacting material with double sided machines. Although the above production numbers also seem to indicate large volumetric capacities this is not the case because the individual compacts often weigh less than 1 g each. For example, with a tablet weight of one gram an output of 3000 tablets per minute translates to a capacity of 180 kg/h. A further increase in numbers of (typically small) compacts produced per minute in rotary punch-anddie presses can be achieved by dual or multiple tooling (two or more die sets) per station (below and, for example, Section 6.3.2). In many reciprocating and most rotary presses for the pharmaceutical and similar industries, the original and still most common “standard” shape of compacts is a more or less cylindrical tablet. As depicted in Fig. 6.2-41, this description includes flat, faceted, and crowned products. For these shapes, simple die and punch configurations are applicable. Since tabletted dry dosage forms in the pharmaceutical industry are consumer products, aesthetics and requirements that are dictated by the medical application (easy identification of a particular formulation by the user) and the marketing-driven desire to distinguish between manufacturers have more recently resulted in the development of special shapes, some of which are shown in Fig. 6.2-42. Additionally, the punches may be engraved as demonstrated, for example, in Fig. 6.2-43. Finally, as already mentioned above, the tooling for smaller tablets can be designed such that in a single pressing station two or more die cavities are associated with correspondingly shaped punches to produce several compacts at once (Section 6.3.2, Fig. 6.3-14). Of course, such punch and die designs are very delicate and require high-precision press designs and excellent maintenance. Expulsion of entrapped gas (air) from granulated or (particularly) powder feeds is very important because it reduces lamination and capping of the tablets. As repeatedly mentioned (also, for example, Section 10.2 and [B.48, B.97]), if gas is entrapped in compacts where it becomes compressed in the residual pore spaces and/or elastic deformation is still present when the compaction pressure is released, products from pressure agglomeration methods are partially or totally destroyed during ejec-

121

122


Fig. 6.2-41

“Standard” tablet shapes [B.48, B.97]

tion. During the manufacturing of pharmaceutical dry dosage forms with high-speed rotary tabletting presses, capping, the separation of a thin layer of material from the main body of the tablet on one or both faces (Fig. 6.2-44), is a particular problem. It is caused by particulate solid feeds that are not suitable for quick, high-pressure densification or, in other words, by too high compaction forces and/or excessive speed of densification (Section 10.2).

Fig. 6.2-42 Some special tablet shapes [B.97]

6.2 Pharmaceutical Applications Fig. 6.2-43 An assortment of engraved punches (courtesy Kilian, K€ oln, Germany)

Raw particulate solids for tabletting may be of three types: (1) non-compressible powders, (2) compresssible powders possessing poor flow characteristics, and (3) compressible powders featuring good flow properties. Non-compressible powders are either pre-granulated wet, which, in some cases, may add a binder component that also renders the granulate better compressible (Section 6.2.1, Tab. 6.2-2), or, if the dosage level is sufficiently low, they are mixed with a powder excipient of type (3) so that the blend becomes compressible and free-flowing. The same methods are used to improve the characteristics of type (2) whereby, if pre-granulation is selected, application of the dry compaction/granulation methods (below) may be advantageous. Type (3) powders are called directly compressible. Validation and cleanliness requirements considerably burden the designs of all equipment for the pharmaceutical industry. To avoid cross-contamination it is necessary to include on modern machines CIP or at least WIP features. It is easily understandable that such techniques are difficult, to say the least, when considering the com-

Fig. 6.2-44 Tablets with “capping” and sketch explaining the capping phenomenon [B.48]

123

124

6 Industrial Applications of Size Enlargement by Agglomeration Fig. 6.2-45 Glove box design of the processing part of a rotary punch-and-die tabletting machine demonstrating WIP (courtesy Fette, Schwarzenbek, Germany)

Fig. 6.2-46 Photographs showing: a) complete turret assembly, b) handling of the turret assembly in a smaller machine, c) turret assembly removal from a larger press, d) special handling system (courtesy Fette, Schwarzenbek, Germany)


plicated mechanical design of multi-station (up to more than 75 per turret [B.66, B.97, B.99, B.105]) rotary tabletting presses. Nevertheless, WIP is one of the latest features of such machines (Fig. 6.2-45). To meet the stringent requirements of the regulatory authorities and reduce downtimes, from newer machines (that were designed during the last decade or so) the entire turret assembly, complete with die table, upper and lower punches and upper and lower cam tracks (Fig. 6.2-46a) can be removed for cleaning, exchange, or maintenance. Smaller machines are equipped with integrated handling and mounting devices (Fig. 6.2-46b) while the assemblies of larger machines require separate handling systems (Fig. 6.2-46c and d).

Fig. 6.2-47 a) Modular system concept; b) modern automated standard tabletting system with rotary tabletting machine and tablet discharge units (courtesy Kilian, K€ oln, Germany)

125

126


The available space for and requirements of tabletting lines vary greatly. This is why many manufacturers have adopted a modular concept consisting of presses with various feed systems, tablet discharge, de-dusting, de-burring, metalcheck, and collection devices, automatic sampling and analysis, and in-process control and documentation (Fig. 6.2-47a). Fig. 6.2-47b shows a modern automated standard tabletting system. Depending on the equipment used, such lines feature good tablet unloading, bad tablet channel, and sample extractor. Electromagnetic gates may also separate tablets that are off-specification during machine start-ups and shut-down. Since punch-and-die presses of various designs are widely used in the pharmaceutical industry and represent the most important equipment for the manufacturing of solid dosage forms, many books have been published on the subject in essentially all languages around the world. It exceeds the context of this book to go into more detail. Rather, the specialized literature should be sought and reviewed, which can be obtained from the pharmaceutical associations existing in all major countries, from international sellers of academic, scientific, and technical books that can be found on the Internet, and from vendors (Section 15.1). The books cited by the author [B.6, B.34, B.41, B.48, B.66, B.97, B.99, B.105] are those that are in his personal library. Originally, when tabletting was first introduced, the powder feeds to the machines were almost always somewhat moist. While, normally, for punch-and-die presses with relatively few strokes per minute accurate and reproducible filling of the die is not a problem and the rate of densification in those machines can be adjusted to match the compressibility of the particulate feed, the very high speed of rotary presses often causes problems. As finer, dry mixtures had to be tabletted and an ever increasing pressing speed was used to optimize the machines’ production rates, it was found that some excipients which, up to that point, behaved very well during densification did not yield products with acceptable quality. At that time (the early 20th century), without knowing all the reasons for this behavior, the term “direct compression” was coined and later (after 1950 [6.2.1.1]) used to identify a process by which tablets are pressed directly from powder blends of the active ingredient and suitable excipients without any pretreatment. With time, the composition of solid drug forms became more complex and feed mixtures for direct compression now also include fillers, disintegrants, and lubricants. It is now increasingly difficult to formulate suitable blends even though directly compressible tablet vehicles (excipients), such as spray-dried lactose, microcrystalline cellulose (MCC), modified starches, and others, are commercially available, increasingly and cheaply. Although the simplicity of the direct compression process is obvious, if pharmaceutical blends are pre-treated in one way or another prior to tabletting, flow characteristics can be improved, compressibility adjusted, and a good particle size and distribution can be selected that yields an optimal feed bulk density. However, the pretreatment of dry dosage formulations becomes more and more complicated to render them suitable for use with the sophisticated high-speed tabletting machines [B.34, B.41, B.48, B.66, B.71, B.97, B.99, B.105]. In simple terms, for those presses it is a requirement that the feed flows quickly and uniformly into the die cavities where it can be converted into well-shaped and firm compacts in a very short time. Tablet quality is a result of the blend’s “compressibility”. This characteristic not only includes an excel-


lent densification behavior, obtained by low interparticle friction, possibly assisted by the addition of lubricants, and suitable deformation properties, often defined by the nature of the main excipient, but also the quick and complete escape of air that is displaced during the fast densification stroke. For that, the absence of elasticity, which may be associated with large crystallites, and a uniformly diminishing porosity are required so that no elastic spring-back occurs upon pressure release and no isolated pores are formed in which air could be trapped and end-up as compressed gas pockets. As the active components in solid dosage forms become more potent and represent a decreasing percentage of the overall composition, an extremely uniform distribution is required to guarantee the correct content of the drug in each individual tablet [B.48]. Therefore, an increasingly important reason for pre-treatment is the necessity of avoiding segregation after mixing during handling and while feeding the press. (Pre-) Granulation of the powder feed blends for tabletting machines solves all above mentioned potential problems (Tab. 6.2-2). Flow characteristics of pre-granulated formulations, owing to the larger apparent (agglomerate) size, are usually superior, even to those of naturally free-flowing powders. While there are indications that some dry granulation for tabletting was already performed by “slugging” (below) prior to 1900 in the pharmaceutical industry, the “classic” process is carried-out by wet granulation (Section 6.2.1). Alternatively, the formulation is blended, moistened, and originally was passed through a screen by the eminence of the hand; drying then yields the granular product. Technically however, this is a low-pressure agglomeration method; modern mechanized and typically motorized equipment is shown schematically in Fig. 6.2-48. Since it is possible that lumps are produced during drying, an optional mill may be used to obtain a uniform, granular product (Fig. 6.2-49). Fig. 6.2-50 and 6.2-51 show of equipment according to 6.2-48 (b) and (c) [B.48, B.68, B.97]. While, as shown in Tab. 6.2-3, wet granulation may offer certain advantages, the addition of liquid(s), which is required to initiate agglomeration, is often objectionable. With liquid addition care must be taken: *

*

*

that the pharmaceutical components are not modified, for example by chemical reaction(s), (partial) dissolution and/or recrystallization, or physical changes, the liquid must be removed after granulation by drying and, if clean and/or expensive liquids are used, should or must be recovered and recirculated, which adds to the cost of the process, and control of particle size of the granular product is difficult; if sizing is required prior to further use, the rejected portion(s) must be recirculated to avoid excessive losses due to the often very high cost of the pharmaceutical formulation and to eliminate contamination or disposal problems caused by toxic drugs.

The modernized equipment depicted in Fig. 6.2-48 is still used for the granulation of certain powder mixtures in the pharmaceutical industry. However, other agglomeration methods, which were developed during the 19th and 20th centuries for different applications, have been modified for use as (pre-)granulation techniques in the pharmaceutical industry and have become more important, as they can be particularly well

127

128


Fig. 6.2-48 Diagram of low-pressure agglomeration equipment using gravity feed and screens or thin perforated sheets: a) screen, b) trough, c) basket [B.97]

Fig. 6.2-49 Sketch of a low-pressure agglomeration system with screen granulator, dryer, and (optional) mill: 1) wet feed mixture, 2) granular product including fines [B.97]

6.2 Pharmaceutical Applications Fig. 6.2-50 Small trough-type granulator with horizontal rotor axis (courtesy Erweka, Heusenstamm, Germany)

designed to meet the special new requirements of validation, control, and documentation. In view of the above mentioned concerns that are associated with wet granulation, improved granulation processes make use of dry powder compaction. Historically, “slugging” was the first such technique. For this, as shown in Fig. 6.2-52, left, large (in diameter and thickness) tablets were made with often old punch-and-die machines, so-called slugging presses, that are no longer suitable for the manufacturing of highquality pharmaceutical dry dosage forms and, therefore, operated relatively cheaply. The compacts were broken to yield a granular feed forgranular feed for tabletting machines. At the beginning, this was mostly done to pre-densify the powder, thereby reducing the press stroke during tabletting, and speed-up the process. Later, the additional advantages described in Tab. 6.2-3 led to a quickly increasing use of the technology. To meet the industry’s capacity requirements, slugging presses

Fig. 6.2-51 The major components (feed hopper, extrusion blades, and perforated die) of a small basket extruder (Bextruder BX 150, Courtesy Hosokawa Bepex GmbH, Leingarten, Germany)

129

130

6 Industrial Applications of Size Enlargement by Agglomeration Fig. 6.2-52 Simplified flow diagram of dry granulation in the pharmaceutical industry using slugging or roller presses with optional circuit alternatives [Section 13.3, ref. 147]

must produce large compacts. While, at the beginning, the tablets were crushed by hand with mortar and pestle, eventually, specially designed motorized screen granulators with oscillating or rotating rod cages (Fig. 6.2-53) as breaking tools were developed to arrive at a semi-continuous operation. The final tablets made from the crushed “slugs” (dry granulated feed) often showed a “checkered” appearance, that is, the surface featured alternating dull and shiny areas. Since pharmaceutical specialties are health remedies and consumer products that are closely scrutinized by the patient, the customers frequently rejected such tablets, resulting in call-backs by and financial losses to the manufacturer. Although other industries (e.g., ceramics) had already realized that density variations exist in pressure compacted powder masses (Section 6.7.2) which, for example, cause the distortion of parts during sintering and, eventually, triggered the development of isostatic pressing in the 1920/30s [B.13a, B.48, B.97], the pharmaceutical industry was unaware of this fact. In 1957 the English pharmacologist Train [6.2.2.2] determined and published lines of constant density in compacts after applying various pressures from the top in a cylindrical die (Fig. 6.2-54). He found that, even in such

6.2 Pharmaceutical Applications Fig. 6.2-53 a) Granulator for pharmaceutical applications with screen and rod cage, b) designed for easy cleaning by disassembly (courtesy Alexanderwerk, Remscheid, Germany)

Fig. 6.2-54 Density distribution in cylindrical compacts [6.2.2.2]. Progressive stages of compaction from the top after applying the indicated

pressures. The curves in the compact are lines of constant density (1 – e) in % (e = porosity)

131

132


simple shapes, due to the presence of frictional forces in the mass and on the tooling surfaces, considerable density differences and gradients develop. Because the largest density variations occur if a great volume of particulate solids is densified with the application of high forces, this behavior applies particularly to the products of slugging presses. It was then realized that, during granulation (i.e., the break-up of large tablets or slugs), hard pieces were obtained from the “press-skin” and other highly densified parts of the compact while weak granules resulted from the less densified interior. During final tabletting, hard granules on the surface were compacted again and created the shiny spots in a matrix of less, but still sufficiently compacted material, explaining the “checkered” appearance. A number of other disadvantages of using slugging presses also existed. For example, they generate large amounts of dust during compaction, resulting in workplace contamination and high material losses. Since the machines are mechanically complex, they require extensive cleaning when changing from one formulation to another and cross-contamination must be avoided. The operating costs are high because of piston and die wear, and lubricants are often necessary to obtain the required compaction ratio. However the greatest disadvantage of using slugging presses is that, in addition to the above mentioned effects of friction, density and hardness also vary from compact to compact because of feeding problems associated with fine powders. This further amplifies the variations in quality of both the granulate and the final tablets. Remedying these problems is very difficult. Today, dry granulation has become a well-accepted method for the processing of powder blends prior to tabletting but slugging has been replaced in most locations by compacting roller presses (Fig. 6.2-52, right). Up to the 1960s, roller presses were almost exclusively used for “dirty” applications in the coal (Section 6.10.2), mineral (Section 6.8.2), and metallurgical (Section 6.9.2) industries. They were typically large, heavy-duty machines and it was thought that they are not particularly well suited for the compaction of very fine powders since the need to operate the rollers at low speed for successful deaeration results in small unit capacities [B.97]. The latter, however, is of no concern for the pharmaceutical industry. Individual batches for the production of a single tabletted formulation are small, often only a few hundred kilograms and sometimes even less. In addition, the value of the material to be processed and the profit margins are much higher than in most other industries. Therefore, roller speeds can be low and capacities small while still maintaining profitability. All roller compacting presses in the pharmaceutical industry use one or multiple force feeder(s) to deliver the blend to the nip between the rollers and to keep the powder firmly pressed into the nip against the flow of air expelled during compaction. The roller surfaces are smooth or slightly profiled to improve the pull of material into the nip and are set at a narrow gap, producing a thin sheet. Feeder and rollers are driven independently with variable speed drives. The feeder/roller speed ratio determines the degree of densification and the hardness of the compacted powder. Product quality can be kept constant by employing an automatic control device by which the feeder motor is slaved to the current drawn by the roll drive motor. The roller presses for pharmaceutical applications are relatively small (Fig. 6.2-55), made of stainless steel, and designed for easy cleaning (Fig. 6.2-56) to avoid cross-


Fig. 6.2-55 Four small roller presses with: a) one, b) two integrated in-line granulators for the manufacturing of tabletting feed in the pharmaceutical industry (courtesy: a1) Powtec, Remscheid, Germany; a2) Vector, Marion, IA, USA; b1) Alexanderwerk, Remscheid, Germany; b2) Riva, Buenos Aires, Argentina)

133

134


Fig. 6.2-56 Design features for easy cleaning of roller presses for pharmaceutical applications (courtesy: a) Hosokawa Bepex, Leingarten, Germany; b) Powtec, Remscheid, Germany; c) Bonals, Barcelona, Spain)

contamination when switching from one formulation to another. To eliminate contact with oil or grease, the rollers are often fixed in the frame. This design is acceptable in the pharmaceutical industry because the blends are extremely fine, clean, and free of tramp materials. Compared with the large-scale, heavy-duty applications in other industries, the compaction forces are relatively low so that torque limiting safety features are sufficient to protect the machines mechanically. The rollers are completely enclosed in a dust-tight stainless steel housing that is kept at a slight negative pressure by an aspiration system. Therefore, workplace contamination is excluded and the loss of material is only about one tenth of that encountered with slugging presses. Compacted sheets or flakes result from the continuous rolling action between the two counter-rotating rollers. Therefore, densification is very uniform and wear is low. The thickness of the compacted material is only in the range of a few millimeters either as flat or corrugated sheets/flakes (Fig. 6.2-57). Even though, microscopically, there are small density variations in the sheet or flakes, they are so small, compared with those found in the large tablets from slugging presses, that they can be disregarded. Produc-


Fig. 6.2-57 Flat and corrugated sheets/flakes and granulated product from pharmaceutical formulations

tion of checkered tablets from granular feeds manufactured by roller press-based compaction/granulation is eliminated. Referring again to Fig. 6.2-52, for many applications in the pharmaceutical industry, the compacted sheet or flakes are simply crushed and the resulting “normal” granular material is directly used as feed for tabletting machines. Although this product features a wide particle size distribution, the top size is limited because in the granulator (mill) the entire throughput is passed through a screen. In spite of the presence of fine agglomerated particles, the granulated product is free-flowing, does not segregate, and features a relatively high bulk density, which reduces the stroke length during tabletting and allows high press speeds. However, if necessary or desired, either fines can be removed by screening the granulator discharge (alternative 1 in Fig. 6.2-52) or larger particles can be produced during crushing followed by double deck screening and recirculation of the oversized material to the mill (alternative 2 in Fig. 6.2-52). The latter reduces the amount of fines produced during crushing (Section 6.1). In both cases, fines are returned to the roller press for recompaction.

Fig. 6.2-58 Diagram of the three design alternatives for positioning the rollers in roller presses for pharmaceutical applications [6.2.2.3]

135

136


Originally and in most contemporary large-scale, heavy-duty roller presses the press rollers are arranged side-by-side and the particulate solids move vertically by gravity supported force feeding or by gravity alone into and through the machines [B.13b, B.48, B.97], but the smaller presses for pharmaceutical applications may feature side-by-side, one-above-the-other, and slanted roller positions (Fig. 6.2-58). Obviously, if a dry powder formulation flows very easily, it is possible that it passes through the gap between essentially smooth rollers without being compacted and it is difficult to start and/or maintain compaction. The use of corrugated rollers may solve this problem. Fig. 6.2-59 shows a small, instrumented compaction/granulation system. The unit features a stainless steel bucket elevator to fill a feed hopper with fresh formulation blend and/or recyclate, a horizontal discharge screw from the hopper into a vertical force feeder, a roller press with side-by-side rollers, an in-line granulator, attached to the press discharge chute, and a double-deck circular screen for separating over- and undersized particles and a narrowly sized granular product. Fig. 6.2-60 shows two views of an older, larger pharmaceutical granulation system, also employing a roller press with side-by-side rollers and vertical screw feeder. With an adjustable (through variable speed drives) production capacity of several hundred kg/h it represents one of the largest facilities used in the pharmaceutical industry. In this plant, the wellblended formulation is transferred from the mixer in special transport containers and deposited into a glove-box hopper resting directly on top of the screw feeder or the roller press. The compacted sheets are crushed in a (screen) granulator, which

Fig. 6.2-59 Small, instrumented compaction/ granulation system (courtesy Fitzpatrick, Elmhurst, IL, USA)


Fig. 6.2-60 Two views of a plant for dry, high-pressure granulation of pharmaceutical products (courtesy H€ ochst AG, Frankfurt/M.-H€ ochst, Germany)

is clamped dust-tight to the discharge chute. The granulator is mounted on wheels and can be easily removed from the system for cleaning. The feed hopper, complete with screw and feeder drive, or the screw alone can be lifted hydraulically after opening some quick release fasteners and swiveled out (Fig. 6.2-56b and c) for easy cleaning of the feed hopper, the screw, the compaction chamber and the nip between the rollers. In the case shown, the granulated product discharging from the granulator is transported by a sanitary screw conveyor to a diverter gate that allows the alternative filling of two transport containers, which are then stored or directly transferred to the tabletting department.

137

138

6 Industrial Applications of Size Enlargement by Agglomeration Fig. 6.2-61 Diagram of two typical designs and arrangements of roller presses for pharmaceutical applications [Section 13.3, ref. 147]

Fig. 6.2-61 depicts two typical designs and arrangements of roller presses for pharmaceutical applications. As mentioned above, small presses in particular are sometimes equipped with rollers one-above-the-other (alternative a2). In this case, the force feeding screw is installed horizontally. It is often said that this design allows for better de-aeration as shown in Fig. 6.2-62. In this diagram, (1) is the raw blend, (2) is a special hopper equipped with a rotating flow stimulator, a separate de-aeration chamber, which is also used for feeding recycle (3) and additives (if applicable), and air removal port (4), (5) is a vacuum pad option for further direct de-aeration from the horizontal feed screw, and (6) is the recirculation of powder leakage [B.13b, B.48, B.97]. Fig. 6.2-61 also shows two step granulation (b). In those cases where the entire output from a single granulator (Fig. 6.2-55a, 6.2-59, 6.2-60, and 6.2-64) is directly used as tabletting feed, the presence of fines is often beneficial (above). However, excess amounts of fines may again result in reduced flowability, segregation, and dusting. The amount of fines can be lowered by two-step milling (Fig. 6.2-55b and 6.2-63) whereby the first mill is equipped with a screen featuring larger openings and only the second granulator, mounted in-line, determines the largest particle size of the final product.


Fig. 6.2-62 Diagram of a new arrangement for improved de-aeration in roller presses with vertical roller arrangement (courtesy Alexanderwerk, Remscheid, Germany)

Fig. 6.2-63 is the photograph of a pharmaceutical compaction/granulation system with vertical roller arrangement and horizontal feed screw, essentially as shown schematically in Fig. 6.2-62 but without the special de-aeration hopper. Granulation is accomplished by two-stage milling. The granulator assembly can be easily detached and rolled out for cleaning. The inset in Fig. 6.2-63 is a photograph of the partially disassembled granulator assembly. Some pharmaceutical formulations contain components that become somewhat plastic during compaction due to the conversion of mechanical energy into heat. While the rise in temperature is small and very short lived and does not normally damage even sensitive materials, sticking on the roller surfaces may occur as these parts heat up during continuous operation. To avoid this, roller cooling is available (Fig. 6.2-56c and 6.2-63). Scrapers may be also installed to keep the rollers clean (Fig. 6.2-64, item 12). When the material becomes temporarily somewhat plastic, continuous sheets are produced, which do not enter the granulators (mills) easily. In those cases, a simple flake breaker is installed (indicated at the discharge in Fig. 6.2-62), which breaks the strip into short pieces (flakes). There are good arguments by the vendors (Section 15.1) why each of the basic arrangements of rollers, horizontal, side-by-side or vertical, one-above-the-other, are more advantageous. Another manufacturer proposes, however, that the rollers should be slanted. The potential problem with horizontally mounted rollers is well understood

139

140


Fig. 6.2-63 Pharmaceutical compaction/granulation system with vertical roller arrangement, horizontal feed screw, and granulation by two-stage milling (courtesy Alexanderwerk, Remscheid, Germany)

(above) and with vertically mounted rollers some powder particles may stay considerably longer in the compaction area, particularly in front of the lower roller, possibly becoming physically or chemically damaged. To avoid this, as shown in Fig. 6.2-64, the roller position in a newer design is between the two extremes [6.2.2.3]. In Section 6.2.1 it was stated that, today, the industrial manufacturing of spherical dosage forms is performed by extrusion and spheronization. The process is based on low to medium pressure extrusion and rounding (spheronizing) the moist, still plastic

6.2 Pharmaceutical Applications Fig. 6.2-64 Diagram of a roller press with slanted roller arrangement, with screw feeders and integrated one-stage granulation (courtesy Gerteis, Jona, Switzerland) [6.2.2.3]

extrudates in a cylindrical vessel with vertical wall and rotating bottom plate (spheronizer) (Fig. 6.2-65). According to a flow diagram (Fig. 6.2-66), the process includes the steps: feeding (1), (2), and (3), mixing and wetting of the powder components followed by kneading the wet formulation to obtain plasticity (4), and extruding damp cylindrical agglomerates (5). The extrudates are either simply separated from fines on a de-dusting sieve (6), yielding clean extrudates or converted into spherical particles in the “marumerizer” (spheronizer) (7). In both cases post-treatment (drying and potentially cooling, not shown) is required to remove the moisture and obtain dry strength of the final product. In the pharmaceutical industry, extrudates (Fig. 6.2-67a) are used, for example, in the field of herbal remedies, including instant teas. Spheronized particles (Fig. 6.267b), the diameter of which can be quite small (down to about 0.5 mm), but more commonly is in the range 1.0–1.5 mm, may be coated if desirable or necessary to obtain specific functionality (Section 6.2.3). Since the product, with or without coating, has excellent flow and metering characteristics it is suitable for filling into soft or hard gelatine capsules (“gel-caps”, Fig. 6.2-68). After ingestion and dissolution of the capsule, the particles are released and, according to the type of coating, may make the drug(s) available at different times in various parts of the digestive system. For these reasons, spheronized products have been quickly accepted by the pharmaceutical in-

141

142


dustry and a number of competing equipment vendors are on the market (Section 15.1). While the smaller spheres with diameters of 0.5–1.2 mm are normally made from extrudates that are produced with low-pressure screen extruders (basket, radial, and dome extruders [B.48, B.97], Fig. 6.2-69), larger ones are manufactured with axial extruders (single or twin screw [B.48, B.97], Fig. 6.2-70). As shown in Fig. 6.2-69 and 6.2-70 the green agglomerates discharge from the extruders as spaghetti-like ropes. Adjustment of their length, which, for spheronizing, should be about

Fig. 6.2-65 Diagram of a marumerizer-type spheronizer (courtesy LCI, Charlotte, NC, USA)

Fig. 6.2-66 Flow diagram of a continuous granulation system featuring a low-pressure extruder with or without spheronizer (marumerizer) [B.48]


Fig. 6.2-67 a) Extrudates with small diameter (1.0 mm) obtained after de-dusting (Fig. 6.2-64, item 6); b) spheronized particles (Fig. 6.2-64, item 7)

Fig. 6.2-68 Photographs of open and closed gel-caps filled with spheronized, coated pharmaceutical granules and of tabletted dry dosage forms (for comparison)

143

144


Fig. 6.2-69 The discharge areas of basket, radial, and dome extruders (courtesy LCI, Charlotte, NC, USA)

1.2 times its diameter, depends on the size and properties of the extrudates. Thinner ones (in the +/- 1 mm range) are charged directly, as produced onto the spheronizer’s rotating plate, which is often provided with a pattern of grooves [B.48], where they break almost instantaneously into the required short segments with rather uniform length. Products with larger diameter, normally made with axial single or twin screw extruders, are brought to length by cutting means, knifes or wires rotating in front of the extrusion plate. If the length/diameter ratio is in the desired 1.2 range, the diameter of the extrudate determines the size of the spheres after spheronization. In the spheronizer, the still plastic extrudate segments are put into rotation by the rotating friction plate (bottom of the bowl) and thrown against the vertical wall by centrifugal forces. As shown in Fig. 6.2-71, a view into an operating machine, a rotating, twisting torus-shaped rope is formed in the edge of the bowl. There, the agglom-

Fig. 6.2-70 Two low-pressure axial extruders and of extrudates/spheronized particles (courtesy LCI, Charlotte, NC, USA, and WLS-Gabler, Ettlingen, Germany)

6.2 Pharmaceutical Applications Fig. 6.2-71 View into an operating spheronizer and schematic presentation of its operation (courtesy LCI, Charlotte, NC, USA)

erates collide and rub with each other and the wall. Mechanical energy is transformed into kinetic energy, which causes a gradual deformation of the extrudate segments into spherical particles (Fig. 6.2-72). During deformation some densification can occur also and excess moisture may migrate to the surface of the rounding agglomerates. Small amounts of moisture will contribute to the pick-up and incorporation of fines by larger particles so that the product is essentially free of fines. If larger amounts of moisture are set free, a slight dusting of the mass by a suitable powder dispenser reduces the likelihood of the entire charge sticking together. Other special features include cooling or heating of the bowl through a jacket and cleaning of the friction plate between charges with brushes. Although the flow diagram of Fig. 6.2-67 seems to indicate that the equipment is interconnected, for the normally small production capacities in the pharmaceutical industry, modular components are used in most cases and transfers from one machine to the next may be accomplished manually. Since the spheronizer operates batchwise, handling typically 5 kg in 1–3 min intervals, the charging of a dosing

145

146

6 Industrial Applications of Size Enlargement by Agglomeration Fig. 6.2-72 a) Extrudates, b) products from different spheronization treatments after 5 s, c) after 30 s, e) after several minutes (courtesy LCI, Charlotte, NC, USA)

unit serving the extruder and the transfer of extrudates that were collected in trays are commonly manual procedures. Larger systems include a batch proportioner in front of the spheronizer, which renders the process quasi continuous [B.48]. Nevertheless, they are still assembled from modular units, which are either dedicated to the processing of one formulation and are easily exchanged or can be wheeled into cleaning bays for decontamination if they are used for different formulations. Continuous feed preparation and post-treatment, even including coating after drying, together with appropriate methods of transfer are used in some large scale operations. Because of their particle shape and nature, cellulose and other plant based or organic drugs and pharmaceutical excipients often exhibit elastic properties. For that reason, as mentioned several times in this book (see, for example, Chapter 5 and Section 10.2), after fast compaction densified products may experience spring-back and loose structural integrity, strength, and/or “quality” as defined by a multitude of descriptions. To overcome possible problems, the rate of densification must be lowered to potentially unacceptable (technically and/or monetary) levels to allow conversion of temporary elastic alteration in shape and volume into permanent plastic deformation. Another possibility to achieve good densification of elastic materials is to use extrusion, for example in flat die pelleting machines [B.48, B.97]. In these presses the material is first predensified between rollers and a flat perforated die and then forced through the die bores, often during several press intervals (Chapter 5, Fig. 5-12b, for explanation) during which additional displaced air can escape and elastic deformation can become permanently plastic. Carboxymethylcellulose (CMC), a compound used as thickening, emulsifying, and stabilizing agent and a bulk laxative in medicine [B.97], is such a product. To become a directly compactable component of dry dosage formulations, granulation is required. Fig. 6.2-73 depicts the flow diagram of a gran-

6.2 Pharmaceutical Applications Fig. 6.2-73 Flow diagram of a flat die pelleting/granulation plant for carboxymethylcellulose (courtesy Amandus Kahl, Reinbek, Germany)

ulation plant for CMC that is based on flat die pelleting. Although the material needs some wetting to make it extrudable and more plastic, the water evaporates during pellet cooling. While elsewhere in this book (Section 6.1, Fig. 6.1-8b) it is stated that a highly densified skin is produced on the surface of extrudates due to shear and wall friction and that the center is somewhat less densified, which leads to the production of granules with different hardness (above), the advantage of overcoming the elastic properties by using pelleting for the densification of CMC overrides the potential disadvantages during tabletting (mostly also because granulated CMC is only a partial component of the dry dosage formulation).

6.2.3

Other Technologies

It was also already discussed in Chapter 2 and Section 6.2 that the first use of coating in medicine was for enrobing pills that were made from ground dry animal and plant matter, minerals, and other solid remedies. While wetting the medicinal mixture with water is often sufficient to render the materials suitable for pill making, more sticky binders, such as honey, were frequently used to mask the often unpleasant taste of the

147

148


concoctions. A disadvantage of moist and sticky pills is that they tend to adhere to each other and feature unacceptable short shelf life by quickly forming lumps. Therefore, most probably the earliest form of coating was to tumble the pills immediately after manufacturing with very fine organic materials, such as pollen, or mineral powders, such as talcum or dried natural clays, to eliminate their stickiness. The surface of the pills is deactivated by the addition of a layer of dry fine particles. At the beginning this was done in bowls, shaking the pill/powder mixture by hand, and then removing the excess fines. Later, especially when powders were compressed into tablets, sugar coatings were applied to make the product look nicer, extend its shelf life, particularly if components of the formulation are hygroscopic and/or swell when moisture is absorbed, make it easier to swallow, and also, in some cases, to mask the drug’s taste. For this, the batch coating pan was developed [B.48, B.97]. These machines are still widely used in the pharmaceutical industry today (Fig. 6.2-74) for sugar coating and glazing. The tabletted cores are often lentil shaped or crowned (Fig. 6.2-41) and the resulting product is called drage´e. Coating is accomplished by spraying (often colored) sugar solutions onto the tumbling bed of cores and simultaneously evaporating the solvent with warm air. The shapes of the pan and the tablets have been selected to achieve stochastic movement and a uniform coating by exposing all surfaces of the cores to the liquid spray. Fig. 6.275 shows two photographs of coating pans in manufacturing plants. Special executions include, among others, double walled pans for heating or cooling (Fig. 6.2-76).

Fig. 6.2-74 Outlines of different modern coating pans (courtesy Stechel, Alfeld, Germany)

Fig. 6.2-75 Coating pans in two manufacturing plants (courtesy Stechel, Alfeld, Germany)

6.2 Pharmaceutical Applications Fig. 6.2-76 Specially equipped coating pans (courtesy Stechel, Alfeld, Germany)

More recently it was discovered that tablets can be functionalized by applying special coatings. Functional coatings may be soluble or insoluble in water, soluble only in liquids with specific characteristics and/or composition, permeable, impermeable, or partly permeable, permanent plastic or elastic, elastic featuring a defined burst pressure. [B.97]. Some are based on recrystallizing dissolved substances, others on the development of polymeric surface layers, and still others on powder coatings. All, however, are produced or assisted by spraying a liquid onto the cores and evaporating the solvent. Because for these new applications, the characteristics of the final coating, particularly its thickness, uniformity, and structure, are determining the performance of the final product, much more stringent requirements on the deposition and buildup of the layer(s) are imposed. Among the first modifications of the behavior of tablets were those that delay the availability of drugs until they reach the intestinal system by providing a coating that is not dissolved by the juices in the stomach. Later, granules were coated with materials featuring different dissolution behavior and either incorporated in varying proportions in tablets or filled into gelatine capsules. The latter are now commonly based on spheronized particles (Section 6.2.2) and the coatings are distinguished by color (Fig. 6.2-68). Such pharmaceutical specialties provide a long-term (controlled release, or retard) effect by liberating the drug(s) at different times after ingestion. Today, sophisticated drug delivery systems are being developed that target specific parts of the body and avoid the indiscriminate broad-band supply of drugs to the entire system. Some of these controlled release reactions can be accomplished by suitably coating particulate active pharmaceutical components prior to formulating the composition of the dry dosage form, for example, tablets or granules. The micron sized coated particles enter the blood stream and liberate the drug when they come to, for example, inflamed body parts and experience higher temperatures. It is understandable that these new coating tasks require more defined deposition methods. Particularly the achievement of very thin, uniform, and completely closed layers (film coating) is a major problem. The entire surface of cores, which may be well-formed compacts or spheronized extrudates, but also irregularly shaped particles or agglomerates, and sometimes very small particles, must be exposed to sprays to

149

150


Fig. 6.2-77 The flow diagram of a typical film coating facility: a) programmable logic controller (PLC), b) storage tanks for spray liquid(s) and metering/pumping system, c) equipment for supplying and processing air, d) air cleaning and exhaust system [B.48, B.97]

form layers that are often only a few molecules or powder particles thick and must have no holes. Areas of coatings become ineffective, for example, if “blobs” of coating material attach to the core or a “shadow effect” causes incomplete coverage (holes) and the desired functionalism is compromised at those points. Therefore, it is also most important to apply strict process control. Modern batch drum film coating equipment [B.48, B.97] features at least four support areas as shown in Fig. 6.2-77. To obtain uniform coverage, while the cores tumble in the apparatus, the liquid sprays must spread over the entire length of the particle bed and the flow of warm or hot air must be directed such that each particle is instantaneously dried to guarantee the production of a smooth surface. Correct movement of the core particles is achieved by installing baffles or lifters or by using polygonally shaped drums. Such high-definition film coaters always operate in batch. Spray systems have become very sophisticated whereby the stainless steel spray arms with nozzles are often telescopic and can be extracted through the front door for cleaning (Fig. 6.2-78). If slurries are used, spraying is air assisted to unplug the nozzles and keep them clean. Depending on the application, the flow of air may be directed in different ways to obtain specific effects. In such drum film coaters some or all of the panels are double walled and perforated to allow air inlet and exhaust in a controlled manner [B.48, B.97]. Another design utilizes stationary, hollow, perforated paddles that are immersed in the product and create an unidirectional, constant, and homogeneous flow of air in the tumbling particle bed. Similar to what has been achieved with double walled, perforated and segmented panels in polygonal drums, in cylindrical drums using paddles, air can be directed in different ways, too (Fig. 6.2-79). Drum film coaters, used in the ultra-clean pharmaceutical industry (Fig. 6.2-80), are made in sanitary, seamless design from stainless steel and feature smooth exterior housings. Additionally, in accordance with the rules of cGMP, the equipment parts, which are contacting product must be capable of CIP. Fig. 6.2-81 shows the automatic cleaning of a drum film coater according to these requirements. The drum and the cleaning tub that is built into the housing are separated from the “technical part” by water tight seals. After wet cleaning in four steps, the machine’s own air system is used for drying.

6.2 Pharmaceutical Applications Fig. 6.2-78 Typical telescopic spray nozzle mounting assembly for film coating drums and detail of an individual spray nozzle attached to this assembly (courtesy O’Hara, Richmond Hill, Ontario, Canada)

Small particles, either powders, crystals, or agglomerates, the shape of which may be irregular, or spheronized, are typically coated in specially designed fluid-bed equipment (Section 6.2.2). As with all other coaters, the heart of fluid-bed processes is the type and location of the delivery system for the liquid coating material. For this, three methods are available: top, tangential and bottom spraying [B.48, B.97]. The nozzles are often binary, that is, liquid is supplied at low pressure to an orifice and is atomized by pressurized gas. Such pneumatic nozzles produce finer droplets, an advantage when coating smaller particles. However, it is also an important requirement of coating that the liquid, solution, or suspension droplets impact the core particles and uniformly distribute on the surface before the liquid is dried off. Since very fine droplets evaporate quickly as they travel from the nozzle to the particle bed, solids concentration and viscosity of solutions and suspensions increase. Therefore, droplets may fail to spread satisfactorily when they contact the substrate surface, resulting in an imperfect coating. This drying-up of the coating spray can be severe in top-spray coaters (similar to Fig. 6.2-18) in which the most random particle movement exists and liquid is sprayed against the flow of drying air. Nevertheless a substantial share of coating is performed in this type of equipment because larger amounts can be processed per batch and the design is simple. Fig. 6.2-82 is an artist’s impression of a topspray coater in a pharmaceutical manufacturing facility showing the fully integrated

151

152


Fig. 6.2-79 Flow diagrams of two different air flow regimes using air-blowing paddles in a cylindrical film coating drum (courtesy GS Coating Systems, Osteria Grande (Bologna), Italy): a) hot air through the paddles into the particle bed with air exhaust through the hollow shaft; b) hot air through the hollow shaft onto the particle bed with air exhaust through the paddles; 1) inlet air handling unit, 2) control panel, 3) solution tank, 4) dosing system for liquid to be sprayed, 5) sliding support arm for spray nozzles, 6) coating pan, 7) air-exhaust or blowing paddle device, 8) dust collector, 9) outlet air fan, 10) powder dosing device

6.2 Pharmaceutical Applications Fig. 6.2-80 Two drum film coaters installed in the ultraclean environment of pharmaceutical processing (courtesy Driam, Eriskirch/Bodensee, Germany)

processing systems and matching accessories with through the ceiling separation of the technical support functions and product handling below the floor of the clean room. The rotating disc fluidized bed coater (similar to Fig. 6.2-31) combines centrifugal, high-intensity mixing with the efficiency of fluid-bed drying. A major advantage of this method is its ability to layer larger amounts of coating materials onto cores consisting either of robust granules, crystals, or nonpareil nuclei. Because of the unit’s high dry-

Fig. 6.2-81 Sketch showing the CIP of a polygonal drum film coater (courtesy Driam, Eriskirch/Bodensee, Germany). Cleaning phases: 1) drum inside by cleaning spray bar, 2) drum outside and air distributor, 3) air channels inside, 4) rinsing of tub

153

154


Fig. 6.2-82 Artist’s impression of a top-spray fluid-bed coating system (courtesy Vector, Marion, IA, USA)

ing rate, relative large gains in product weight can be achieved in a short period of time. Another advantage of the rotating disc fluidized bed coater is the possibility to layer dry powders that are dispersed in the fluid bed onto nuclei, which have been wetted with a liquid. Because the spray nozzle is located below the bed surface, the above mentioned problems with early drying are not experienced.


The same is true of the Wurster coating process (Fig. 6.2-83). This is the only bottom-spray fluid bed coating method, which is not only applicable for fine particles and coarse granules but for tablets and pellets as well. The Wurster coating chamber is cylindrical and the basic model contains a concentric inner tube with approximately half the diameter of the outer chamber. At the base of the apparatus is a perforated plate, which features larger holes underneath the inner tube. The liquid spray nozzle is located in the center of the orifice plate and the tube is positioned at a certain distance above the plate to allow the movement of material from the outside annular space to the higher velocity airstream inside the tube. This design creates a very organized flow of material, similar to that of a spouted bed [B.48, B.97]. Solids move upwards in the center where coating and highly efficient drying occur. Contrary to what happens in the spouted bed, where some mass exchange occurs between the solids moving upwards in the center and those in the outside downward flow, the high-speed upward flow regime in a Wurster coater is contained in the center tube, so that no backmixing occurs. At the top of the tube, the material discharges into an expansion area and then flows down, as a near-weightless gas/solids suspension, in the annular space outside the tube. Fig. 6.2-84 shows a Wurster coater for pharmaceutical applications. Design variations include different chamber configurations for use in coating tablets, coarse granules, or fine particles (Fig. 6.2-85). For scale-up, the outer vessel diameter and the number (rather than the size) of inner tubes increase. Each tube features its own gas distributor and spray nozzle and is essentially identical with that in the test equipment that was used during development work [B.97]. During studies of particle and gas flow patterns in traditional Wurster coaters it was found that flow patterns were dominated by the particles rather than the gas. This explains why sometimes, in spite of the well defined flow, uniform coating is difficult to control. To overcome this problem, a precision Wurster coater was developed [B.97]. An essential feature of this new design is a better controlled gas flow pattern in the coating zone. This is achieved by application of a guiding system in which the gas flow is accelerated, stabilized, and given a precise amount of swirl, which eliminates slugging, often seen in traditional coaters, and stabilizes multi-tube systems. Particles are

Fig. 6.2-83 Principle of the Wurster coater (right) and outline of an apparatus (left) (courtesy Glatt, Binzen, Germany)

155

156

6 Industrial Applications of Size Enlargement by Agglomeration Fig. 6.2-84 Wurster coater for pharmaceutical applications (courtesy Glatt, Binzen, Germany)

entrained into the swirling air on an individual basis. This results in an optimized probability of impact with the droplets of atomized liquid and an even application of coating material. Fig. 6.2-86 shows scanning electron micrographs of various particles that were coated with different fluidized bed coaters. Although, as shown in Fig. 6.2-85a, short Wurster coaters may be used for the processing of tablets, it is more common to use drum coaters for this task. Since spray systems projecting into the tumbling bed are subject to relatively high mechanical forces, are damaged easily, and become clogged quickly, usually the top-spray method is employed. As mentioned before, the disadvantage of this arrangement is that many of the fine spray droplets are deflected by the strong counterflow of hot air and dried before they ever reach the particles to be coated. This results in the loss of film forming material and an irregular surface coverage. Furthermore, the coating is non-uniform because neutral or dead zones occur in the bed so that it can not be assured that all tablets travel to and are exposed at the bed’s surface for the same time to the spray.

6.2 Pharmaceutical Applications Fig. 6.2-85 Sketches of the different chamber configurations of single tube Wurster coaters as used for: a) tablets, b) coarse granules, c) fine particles [B.48]

To overcome these problems, a vertical centrifugal coater was developed [6.2.3.1]. As depicted in Fig. 6.2-87 it consists of a perforated bowl-shaped container that rotates around a vertical axis, and a static return flow cone located therein. Centrifugal force is utilized to produce the movement of the tablets to be coated. The particles initially rise up the wall of the rotating bowl and, on reaching the top rim, are led into the return flow cone. They then descend and, in the center, drop back to the bottom of the bowl where the cycle starts again. When the tablets fall down from the return cone they are sprayed with the coating material by a radial spray nozzle with a narrow spray angle of 308. The major advantages of this new machine are:

Fig. 6.2-86 SEM photographs of particles that were coated with fluidized bed coaters: a) coated with ethylcellulose in a top-spray fluid-bed processor, b) coated with ethylcellulose in a Wurster apparatus, c) vacuum top-spray coated (retard) pellet, cut open, d) Wurster coated (retard) pellet, cut open (courtesy Glatt, Binzen, Germany)

157

158


Fig. 6.2-87 Diagram of the operating principle of a vertical centrifugal coater (courtesy Diosna, Osnabr€ uck, Germany) * * * * * *

reduction of the processing time by 50 % or more, uniform coating, intense air circulation, and little loss of coating material, high-quality through reproducible thickness and uniformity of the coating, possibility to build and use (due to shorter processing time) smaller units, complete, automatic discharge, and ease of adding WIP.

Fig. 6.2-88 is a view into the open top of an operating vertical centrifugal coater. Another coating technique is encapsulation. Although this is a relatively new technology, many different processes have been developed, a large number of applications has evolved, and many new uses are conceivable and found, literally on a daily base (Chapters 5, 11, and 12). Some encapsulation can be achieved during the drying of wet (green) agglomerates, including spheronized extrudates, the pores of which are filled with a liquid (continuous phase) [B.48, B.97]. In a first drying phase, evaporation takes place only on the surface and the liquid is replenished by capillary flow of the continuous phase from the interior of the porous body. If the liquid is a solution or suspension, solids are deposited at the pore ends on the surface and cause encrustation. If a film-forming, easily soluble polymer is dissolved in the continuous phase as emulsion or dispersion, en-


Fig. 6.2-88 View into the open top of an operating vertical centrifugal coater (courtesy Diosna, Osnabr€ uck, Germany)

capsulation occurs during drying. These encapsulation processes can be carried-out with agglomerates of any size and shape and result in a large number of special effects which, depending on the type and composition of the coating or incrustation, modify final product properties. More commonly used and widely researched is microencapsulation. In this context, the prefix “micro” refers to the dimension of the encapsulated product, which is typically 1–2 mm in size or, increasingly, 10–100 lm range. If a slurry containing a solution, emulsion, or suspension of polymer is dispersed into small particles and dried in a spray dryer microencapsulated particles of the type described above are formed. Such a process yields a dry, free-flowing powder which, in most cases, satisfies the criteria defined for instant products. However, microencapsulation becomes a more and more a sophisticated packaging method in which the “packing material” (coating) features a specific, well-defined functionality. With this technology small agglomerates or tiny portions of powders, liquids, and even gases are individually wrapped into a shell (capsule) to form free-flowing particles, which are most often spheroidal [B.97]. Fig. 6.2-89 illustrates possible structures of microcapsules. In the pharmaceutical industry, drugs are encapsulated to facilitate or improve handling and, very importantly, to bring about special product characteristics (functionalizing). Microencapsulation techniques make use of sol-gel processes, coacervation, surface and in situ polymerization methods or, generally, interfacial reactions to produce soluble or insoluble and impermeable or permeable capsule walls. In addition to the coating and spray drying methods that were discussed previously, a growing number of other processes deposit particles onto cores or solid surfaces whereby the binding mechanisms of agglomeration are utilized [B.97] (Chapter 11).

159

160


Fig. 6.2-89

Diagram of possible structures of microcapsules [B.97]

As examples, Fig. 6.2-90 depicts sketches of the principle and the equipment of electrostatic, aerosol based microencapsulation. The two components to be turned into a microencapsulated product, the coating and the core particles, are given an ionic charge of the appropriate sign using a sub-corona discharge system. To achieve sufficient encapsulation, the apparatus must be designed such that a high rate of collisions between the two components occurs in a turbulent supportive gas system. In this process, the coating materials must be selected so that they will uniformly and completely cover the core particles. They include softened wax particles, which solidify


Fig. 6.2-90 a) Principle of microencapsulation by electrostatic, aerosol based coating (1, coating particles; 2, core particles); b) sketch of a possible equipment configuration [B.48, B.97]

upon cooling or polymers, which form a skin by interfacial action between a component in the core and another in the coating material or upon exposure to a suitable gas phase. The coating (capsule) can be also finished by heating (sintering). Although, technically, agglomeration in liquid suspensions is a wet agglomeration method, for the purposes of this book it qualifies under “other technologies” in the pharmaceutical industry as a new growth process for the manufacturing of welldefined, engineered particles. Crystallized active pharmaceutical ingredients which, because of their particle size and shape, do not exhibit good handling and/or compression behavior must be modified to render them suitable for direct use in the modern high-speed rotary tabletting machines. A promising new technique is spherical crystallization [6.2.3.2]. This agglomeration method greatly improves the micrometric (size, distribution, shape, and structure) and compaction properties of crystals. Crystal agglomeration in liquid suspensions is possible according to two different mechanisms, which depend on the amount of drug solution (for example ascorbic acid dissolved in a good solvent, for example, water) added to a poor solvent system (e.g., ethyl acetate). Fig. 6.2-91 shows the two alternative methods: emulsion solvent diffusion (ESD) and spherical agglomeration (SA).

161

162


Fig. 6.2-91

Mechanisms of crystal agglomeration [6.2.3.2]

When, in the first case, the two liquids, kept at different temperatures, are mixed with a small water-to-ethyl acetate volumetric ratio (1:100), using a suitable stirrer and no baffles, a W/O emulsion forms first. Without baffles, the emulsion droplets do not break down. Then, as they cool and water/ethyl acetate diffuse, solubility in the emulsion droplets decreases and crystals precipitate on the droplet surfaces. When crystallization is complete, agglomerates have formed. This mechanism is called emulsion solvent diffusion or ESD. On the other hand, if the water-to-ethyl acetate volumetric ratio is large (4:150) and baffles are built into the agitator tank to promote turbulence, crystals precipitate in the same way after an emulsion has formed. Then, a small amount of water that was liberated from the ethyl acetate phase acts as a liquid bridging agent between the crystals, causing particles to randomly agglomerate (immiscible liquid agglomeration [B.48, B.71, B.73, B.97], see also Chapter 5). This mechanism is called spherical agglomeration or SA. The main difference between the two agglomerated crystal forms is that, as depicted in Fig. 6.2-92, in ESD particles, crystals grow towards the center (a2), while with the SA method, primary crystals were randomly assembled and formed into agglomerates (b2). Fig. 6.2-93a is another SEM image of a spherically crystallized drug showing again the random arrangement of primary crystals and Fig. 6.2-93b indicates the exceptional uniformity of the product, which is responsible for its excellent flowability and handling properties, both allowing quick feeding into the dies of high-speed tabletting machines.


Fig. 6.2-92 Scanning electron micrographs of: 1) external appearance, 2) cross section of a) ESM, b) SA agglomerated crystals of ascorbic acid [6.2.3.2]

As described above, in SA the crystallization system is in reality a ternary mixture composed of a good solvent dissolving the drug, a poor solvent for precipitating the drug, and a third liquid, immiscible with the system, termed bridging agent. The latter preferentially wets and agglomerates the crystals produced. In the two solvent approach (mixing poor and good solvents), a small amount of the binding liquid is liberated from the system and becomes available for bridging. As shown in the three step sketch of Fig. 6.2-94, a three-solvent system (containing good and poor solvents and an added bridging liquid) allows even better quality control of the agglomerates [6.2.3.3]. With decreasing amount of bridging liquid the agglomerates become more uniformly

Fig. 6.2-93 Scanning electron micrographs of: a) spherically agglomerated crystals of a drug; b) sample of the same product, showing its uniformity (courtesy AstraZeneca R&D, M€ olndal, Sweden)

163

164


Fig. 6.2-94

Mechanisms of spherical crystallization in a three-solvent system [6.2.3.3]

shaped and feature a narrower particle size distribution (Fig. 6.2-95). Also, as shown in Fig. 6.2-96, the greater interfacial tension of the added aqueous bridging liquid in the ternary system, as compared with the ethanol–aqueous bridging liquid, causes stronger bonds, producing agglomerates with a thick shell and highly densified surface, made up of needle-like crystals. Ultimately, however, the interest of the pharmaceutical industry in a new, engineered material for tabletting is not only in its improved handling and die filling behavior but also in an enhanced compactibility. Direct tabletting of a granulated formulation is only feasible if all of these combined characteristics are available. In a series of experiments acebutolol hydrochloride, an anti-arrhythmic agent, was used for tabletting trials [6.2.3.4]. The material was chosen because it is easily water soluble and in its dry crystalline form exhibits strong cohesion and high static electricity properties. From an aqueous solution of the drug a W/O emulsion with isopropyl acetate

Fig. 6.2-95 Size distribution of SA crystals produced in a three-solvent system [6.2.3.3]. Ethanol 5 mL, bridging liquid (water): *) 0.5 mL, ~) 0.8 mL, &) 1.2 mL, ^) 1.5 mL


Fig. 6.2-96 Scanning electron micrographs of SA crystals prepared with two- and three-solvent systems [6.2.3.3]

was prepared in a stirred vessel containing baffles. A small amount of seed crystals (drug powder) was then added to promote crystallization and agglomeration. After solidification was complete, the dispersing solid was decanted and the agglomerated crystals were filtered and dried. This process conforms to spherical agglomeration (SA) with a two solvent system. Fig. 6.2-97 shows scanning electron micrographs of drug crystals (5–20 lm) and of spherically agglomerated product (208–600 lm). As shown in Fig. 6.2-98, the fill volume (called relative volume, crossed symbols) in the die of the spherical agglomerates is almost twice as high as that of the crystals. This is due to the fact that crystals are dense and fines fill the voids between the particles

Fig. 6.2-97 Scanning electron micrographs of single dry crystals of acebutolol hydrochloride and of spherically agglomerated crystals [6.2.3.4]

165

166

6 Industrial Applications of Size Enlargement by Agglomeration Fig. 6.2-98 Relationship between relative volume and compression pressure during the compaction of single dry (*) and spherically agglomerated () acebutolol hydrochloride crystals [6.2.3.4]

while the micro-spheres pack less densely, contain no fines, and are porous. With increasing pressure the crystals compact uniformly while the spherical agglomerates exhibit the typical densification behavior of larger particles (Section 5.2 Fig. 5.9). The latter includes rearrangement of the spheres without change in size and shape (a–b in Fig. 6.2-98), production of some fines that fill voids between the spheres (b–c), breakage and deformation of the spheres and building of a new, dense structure (c–d), and, at the end, breakage and densification of the needle-like crystals themselves (d–e). In the final phase (c–d), a somewhat denser compact structure is formed with the agglomerates as compared with the single crystals (Fig. 6.2-99).

Fig. 6.2-99 Scanning electron micrographs of cross-sections of tablets prepared with various compression pressures (1, 10 MPa; 2, 50 MPa, 3,

200 MPa) [6.2.3.4]: a) tablets from single crystals, a1–a3 1500; b) tablets from spherically agglomerated crystals, b1 750, b2 and b3 1500

6.3 Applications in the Chemical Industry

The overall evaluation must take into account the extreme handling problems and die filling difficulties and a pronounced capping tendency (Section 6.2.2, Fig. 6.2-44) of single crystals. In contrast, the very good flowability of the spherically agglomerated drug allows a higher speed of compression whereby, because freshly created surfaces resulting from agglomerate fracture enhance the plastic interparticle bonding [6.2.3.4], capping does not occur. Therefore, utilizing spherically agglomerated drugs result in directly tablettable formulations.

6.3

Applications in the Chemical Industry The chemical industry shows the most diverse applications of size enlargement by agglomeration. They are found in both inorganic and organic chemistry and are typically used to modify the characteristics of intermediate products by eliminating dusting, improving storage, flow, and metering properties, increasing bulk density, providing good dispersibility or, generally, desirable characteristics for further use, giving high surface area in stable particulate masses, and causing many more features (Tab. 6.1). There are very few chemicals or compounds that are produced in quantities similar to, for example, agrochemicals, animal feeds, ceramics, fertilizers, foods, minerals, pharmaceuticals, or solid fuels (mentioning, in alphabetical order, other major sections of this book). Each of the many different chemicals is treated by the method that is required to obtain the desired characteristics. Processes may include wet or dry granulation, extrusion, spheronization, briquetting, compacting, or sintering and apply any of the methods of size enlargement by agglomeration (Chapter 5). Tab. 6.3-1 lists, in alphabetical order, chemicals that are cited in the sales literature of different vendors of equipment for size enlargement by agglomeration. They were compiled by the author, not to paint a concise picture of the use of the technology in the chemical industry but to suggest the great variety of potential applications. Several of the chemicals are synthesized materials that are also found in nature and may also be mentioned in connection with agrochemical (including fertilizer), feed, food, pharmaceutical, and special applications. Tab. 6.3-1 Chemicals (in alphabetical order) that are cited in the sales literature of different vendors of equipment for size enlargement by agglomeration Acrylonitrile butadiene resins Acetylsalicylic acid Adipic acid Alkyl-aryl sulfonates Aluminum chloride Aluminum formiate Aluminum hydroxide Aluminum phosphate

Chromium sulfate Cobalt sulfide Citric acid CMC (Carboxy-methyl-cellulose) Codeine phosphate Copper hydroxide Copper oxide Copper oxichloride

Nickel hydroxide Nickel sulfide Nitrates Nitro-penta-erythritol Organic pigments Oxalic acid Paracetamol Pectin

167

168

6 Industrial Applications of Size Enlargement by Agglomeration Tab. 6.3-1 Chemicals (in alphabetical order) that are cited in the sales literature of different vendors of equipment for size enlargement by agglomeration (continued) Aluminum silicate Aluminum stearate Amino acid Ammonium chloride Ammonium nitrate Ammonium phosphate Ammonium sulfate Ampicillin Ampicillin trihydrate Amylase Antrachinon Aspartame Aspirin Atrazine Barium carbonate Barium chloride Barium hydroxide Barium sulfate Bismuth carbonate Borax Boric acid Cadmium carbonate Cadmium chloride Cadmium sulfide Calcium acetate Calcium bromide Calcium caseinate Calcium gluconate Calcium hypochloride Calcium lactate Calcium oxide Calcium saccharate Calcium silicate Calcium stearate Calcium sulfate Catalysts (bismuth, iron, nickel, platinum, silica, vanadium, zinc) Cellulose acetate Chrome tannin Chrome-yellow Chromium dioxide Sodium bichromate Sodium bromide Sodium carbonate Sodium chloride Sodium chromate Sodium citrate Sodium cyanide Sodium ethylxanthate

Copper sulfate Cryolite Cupric oxide DBDMH (Dibromo-dimethyl hydentoin) Dextrose monohydrate Dicalcium citrate Di-potassium and -sodium orthophosphate Dispersing agents DMT (Di-methyl-terephthalate) Ethylene-vinyl acetate Fatty alcohol sulfate Glycerides Graphite Hypochloride Ibuprofen Indigo Iron oxide Iron chelate Lead oxide Lead stearate Lead sulfate Lithium chloride Lithopone Magnesium aluminum silicate Magnesium carbonate Magnesium chloride Magnesium hydroxide Magnesium oxide Manganese carbonate Manganese oxides Maleic anhydrate Maltodextrins MCC (Micro crystalline cellulose) Melamin formaldehyde resin Monosodium glutamate Mono-sodium and -potassium orthophosphate Niacinamide Nickel carbonate Sodium phosphate Sodium silicate Sodium silico aluminate Sodium silico fluoride Sodium sorbite Sodium sulfate Sodium thiosulfate Sorbitol

Penicillin Phenoles Phenol formaldehyde resin Phtalocyanides Phosphates Polyacrylate Polyacrylonitrile Polyamide Polycarbonate Polyethylene Polyethylene terephthalate Polyformaldehyde Polypropylene Polystyrene Polyvinyl acetate Polyvinyl pyrrolidone Potassium bicarbonate Potassium carbonate Potassium chloride Potassium monopersulfate Potassium nitrate Potassium permanganate Potassium peroxymonosulfate Potassium persulfate Potassium phosphate Potassium sorbate Potassium sulfate PVA (Polyvinyl alcohol) PVC (Polyvinyl chlorides) Rubber accelerants Saccharin Saccharose Salicylic acid Selenium sulfonate Silicic acid Soda ash Sodium acetate Sodium aluminum silicate Sodium ampicillin Sodium antimonate Sodium benzoate Thorium carbonate Titanium oxide Uranium dioxide Urea Urea formaldehyde resin Xanthates Xanthene Zeolite

6.3 Applications in the Chemical Industry Tab. 6.3-1 Chemicals (in alphabetical order) that are cited in the sales literature of different vendors of equipment for size enlargement by agglomeration (continued) Sodium Sodium Sodium Sodium Sodium Sodium

gluconic acid hydrogen sulfate hypochlorite lauryl sulfate orthophosphate perborate

Streptomycine sulfate Styrene resins Sulfates Sulfonates Synthetic tannin Tetra potassium polyphosphate

Zinc Zinc Zinc Zinc Zinc

carbonate chromates oxide potassium chromate sulfate.

In addition, a few groups of chemical products were selected by the author, which are discussed in more detail in this section as examples of what can be accomplished by modifying chemicals with agglomeration techniques. They are: aspartame (artificial sweeteners) biocides (water treatment) catalysts (laundry) detergents DMT (dimethylterephthalate) pigments plastics (master batch) sodium cyanide

improved packing, dosing and dispersion eliminate dust, increase stability improved packing and increased activity improved density, packing and dissolution increased bulk density, improved handling improved handling, metering and dispersion avoiding segregation, better metering reduce toxicity, improve handling

Further Reading

For further reading the following books are recommended: B.3, B.7, B.13.e, B.16, B.21, B.22, B.26, B.29, B.33, B.40, B.48, B.49, B.50, B.56, B.58, B.60, B.64, B.67, B.70, B.82, B.89, B.90, B.92, B.93, B.94, B.97, B.98, B.102, B.103 (Chapter 13.1). Books mostly devoted to the subject matter are printed bold.

6.3.1


As indicated above, rather than discussing a multitude of applications, a few well-established uses of size enlargement by tumble/growth agglomeration in the chemical industry will be reviewed and described in some detail as examples. Aspartame (Artificial Sweeteners) Artificial sweeteners are obtained by either hydrolysis of starch (e.g., maltodextrin) or chemically as crystalline compounds that are unrelated to carbohydrates (low calorie) and several hundred times sweeter than cane sugar (e.g., saccharin and aspartame). As produced, all products are very fine, do not wet easily, feature bad flow properties, and 6.3.1.1

169

170


can not be easily handled. They are also not stable, adhere to walls, and lump during storage. Size enlargement by agglomeration is used to overcome these problems (see also Sections 6.3.2 and 6.3.3). As an example of growth agglomeration, the production of aspartame (a-L-aspartylL-phenylalanine methyl ester or a-APM) as a stabilized ingredient in low-calorie chewing gum is presented [6.3.1.1]. Many efforts have been made to stabilize APM, including encapsulating or otherwise coating the material. Most techniques use fluidized bed technology, for example, by co-drying a solution containing APM and an encapsulation agent, such as gum arabic, or by coating dry particles in a Wurster-type fluidized bed coater (Section 6.3.3). A recent patent [6.3.1.1] describes the mixing of non-coated APM with a dry binder, such as modified cellulose (e.g., hydroxypropyl methylcellulose), and a liquid, such as water. The mixer is a planetary or other such machine that applies compressive forces between the components. The resulting moist blend, characteristically dust-free, nonflowing, and crumbly, is dried and crushed to produce fine agglomerated APM particles (< about 0.4 mm). They can be added to a chewing gum formulation without a further coating. However, since these agglomerates release their sweetness instantly, it is still preferable to use them in combination with APM particles resulting from encapsulation or other treatments (Section 6.3.3) that produce a slower release during chewing. By blending APM materials with different release rates into the gum, it is possible to obtain stable products, featuring a strong initial taste and a milder longlasting sweetness. Agglomerated products, including aspartame and other artificial sweeteners, can also be made directly in high-shear mixers with knife heads or similar high-speed, high-energy tools. Such machines include drums with horizontal axes (e.g., pin or plow mixers), bowl types with vertical axes (Diosna, Fukae, or Henschel mixers) and vertical tube arrangements (e.g., Schugi) [B.97] (Section 15.1). The first two normally operate on batches and may apply one-pot processing, includings drying, but the latter works continuously, discharging moist agglomerates, with external drying, most commonly in a fluidized bed. Because of the use of high shear, which both builds and destroys agglomerates (Fig. 5-3 and 5-4), all produce the desired particle size directly and yield products with “instant properties”. As mentioned above, these agglomerated particles can be further engineered and processed, for example by adding suitable coating(s) (Section 6.3.3), to improve their storage, handling, and application characteristics.

Laundry Detergents Washing clothes is one of the oldest human activities. Originally, it was carried out by immersing the material to be washed in water and beating, treading, and/or rubbing it to dislodge the dirt. Later, in relatively recent times, it was found that water not only flushes out the soiling substances but also reduces by a factor of ten the natural van-der-Waals adhesion forces that participate in holding the dirt. In this respect it was discovered that soft water (rain water) has a better effect. Soda ash was known in ancient Egypt as a wash additive and charcoal was used in medieval times. 6.3.1.2


Soap has been known for more than 3000 years: it is the oldest of the surfactants. For a long time it was applied as a cosmetic and a remedy. About 1000 years ago it came to be used as a general purpose washing and laundering agent [B.102]. “Modern” detergents were invented in Germany by Henkel. Synthetic soda ash and sodium silicate formed the basis for “Bleichsoda”, the first commercial detergent, introduced in 1878. With the beginning of the 20th century and the introduction in Germany of the first self-acting laundry detergent (Henkel’s “Persil”, 1907), soap, now synthetically produced by the saponification of fats with soda ash, became an ingredient in multi-component systems for the washing of textiles. Soap was combined with so called builders, usually sodium carbonate, sodium silicate, and sodium perborate, and bleach. The next important development was the transition from manual laundry to machine washing, which required appropriate changes in the formulation of detergents. Soap, which is sensitive to water hardness, was gradually replaced by other synthetic surfactants; the first neutral detergent for delicate fabrics was introduced in Germany in 1932 (Fewa) and in the USA in 1933 (by Dreft). The general acceptance worldwide of the new synthetic surfactants (mostly products from the petrochemical industry) is a development of the 1940s; Procter & Gamble introduced in 1946 the synthetic detergent Tide in the USA. Because of the increasing pollution of rivers and lakes, the biodegradability of detergent products, its testing, and corresponding environmental legislation became important tasks in the 1950s and 1960s. Subsequently fillers, such as sodium carbonate, were replaced by complexing agents (sodium di- and tri-phosphates) and more recently by zeolites, particularly in those countries where phosphate legislation was enacted. Tab. 6.3-2 summarizes the main components of laundry detergents and their functions [B.60, B.102]. To produce conventional laundry detergent powder, spray drying is the established method (the tower process). Fig. 6.3-1 shows the block diagram of such a system [B.60]. All raw materials are procured either in liquid or powder form. While ensuring a certain quality, the selection is mostly based on price. The physical characteristics of the components is not important since they will be dispersed or dissolved and mixed with other liquids to form a solution or slurry. This slurry is pumped by low- or medium-pressure pumps to the spray dryer (Fig. 6.3-2). An air vessel (pressure accumulator) is used to even-out pressure peaks. The spray towers can work in con- or counter-current fashion [B.97]. Concurrent contact yields light powders (mostly made-up of hollow particles) with a bulk density of about 100–150 g/L and a moisture content of 3–10 %. Also, the hollow beads tend to break-up and form dust. Countercurrent drying (Fig. 6.3-2) produces powder with a bulk density of 300–500 g/L and a moisture content of 7–15 % (commonly 10 %). Most plants use this method because the higher bulk density is almost always desired. The tower process is limited in its ability to produce powders with bulk densities greater than 500 g/L and the inclusion of non-ionics in the formulation. The latter, which can not be spray dried due to its volatility at the prevailing drying temperatures in the tower, must be post-added in either a rotary drum or other simple machines. The final detergent could then have a maximum content of 4 % and still feature a bulk

171

172

6 Industrial Applications of Size Enlargement by Agglomeration Tab. 6.3-2 The main components of laundry detergents and their functions [B.60, B.102] Function

Components

Surfactants (anionics)

Soap Alkylbenczene-sulphonate Fatty alcohol sulfate Alpha-olefinsulphonate Alpha-sulpho-methylester Ethoxylated alkyl-phenols Ethoxylated fatty alcohols Sodium carbonate Sodium di- and tri-(poly)phosphate Poly-carboxylates (NTA) Citrates Zeolite A (ion exchangers) Sodium perborate Sodium percarbonate (Tetra-acetyl-ethylene-diamine) Sodium sulfate Water Corrosion inhibitors (sodium silicate) Foam boosters (alkylolamides) Foam inhibitors (hydr. soap/silic. oils) Stabilizers/sequestrants Optical brighteners/fluorescent whitening agents Soil antiredepositioning agents (CMC, Bentonite) Enzymes (alcalase, protease) Minors (dyestuffs, perfume)

Surfactants (nonionics) Builders

Bleaches (and activators)

Fillers and processing aids Specific additives and minors

density of < 500 g/L. Several producers have used the post-treatment system not only to add the non-ionics but to also supply and include heavy powders in an effort to produce a somewhat higher bulk density of the laundry detergent (Fig. 6.3-3) [6.3.1.2]. The ZigZag blender depicted in Fig. 6.3-3 (and in Fig. 6.3-4 and 6.3-9) is a unique combination of high-speed, high-shear mixing in the first part (drum with eccentrically arranged mixing tool) and gentle agglomeration and rounding in the second (zigzag) [B.97]. To further overcome the limitations of spray drying and extend the applicability of the post-treatment shown in Fig. 6.3-3, fluid-bed dryers were introduced for final drying (Fig. 6.3-4). Now the mixer, which can have designs other than shown, is operated as an agglomerator, producing moist granules, which are dried in the fluidized bed that follows. Such dryers can handle larger, denser particles and have high thermal efficiency, thus allowing the removal of large amounts of moisture. Therefore, the combination of a spray dryer with an agglomerator and a fluid-bed dryer results in a very flexible plant, capable of producing different powder qualities. Fines entrained during drying may be captured in a bag filter and recycled into the agglomerator.


Fig. 6.3-1 Diagram of a spray-drying system for the production of conventional laundry detergent powder [B.60]

Non-tower methods are also used for the production of complex detergent powders. Such processes are carried out in horizontal, continuous drum, or other high-shear mixers [B.97] that are equipped with various blending tools, often assisted by knife heads. As shown in Fig. 6.3-5, in the reactor/mixer dry powder raw material, which may include spray dried components, zeolites, and soda ash, is used to bind surplus moisture that is brought in with the other ingredients. Owing to the high-energy input by the blending tools, agglomerates are formed. At the same time neutralization with sodium carbonate or sodium hydroxide takes place. Since the granules may be sticky, mainly due to the non-ionics in the composition, they may be powdered (coated) with zeolite in a second mixer. If the water binding capacity of the dry components is still not sufficient, a further step, such as a fluidized bed dryer, may have to be added (Fig. 6.3-6). In the 1980s, beginning in Japan, a strong demand developed for solid laundry detergents with a density of more than 700 g/L and a content of active matter (anionic and non-ionic) of up to 50 %. Wet agglomeration (Fig. 6.3-7), performed by adding water, polymer solutions, anionic surfactant pastes, or surfactant gels at highshear rates, is a common process for densifying. It consists of the following steps [B.102].

173

174


Fig. 6.3-2 Flow diagram of a traditional spray drying process for the manufacture of laundry detergent powder [B.102]

Fig. 6.3-3 Flow diagram of a tower process with post-addition of powders and non-ionics and agglomeration [6.3.1.2]


Fig. 6.3-4 Flow diagram of a tower process with post-addition, wet agglomeration, fluid-bed drying, and dust recycling [6.3.1.2]

1. Grinding in a high-speed mixer, for example a L€ odige CB type (Fig. 6.3-8); addition of polymer solution (7–12 %), binder (water), and other ingredients via spray nozzles; pre-granulation. 2. Granulation (potentially adding more binder liquid) and conditioning in a high- or medium-shear mixer (optional but preferred). 3. Evaporation of excess water in a fluidized bed and/or cooling down. 4. Removal of coarse particles from the product by sieving (normally, fine particles remain in the product), milling the oversize, and recirculating the powder to the agglomerator. 5. Addition of minor ingredients in a low-energy mixer. Using several mixers of different type in a cascade increases the process flexibility. Energy input, granulation temperature and time, location and method of dosage of

175

176

6 Industrial Applications of Size Enlargement by Agglomeration Fig. 6.3-5 Block diagram of a non-tower agglomeration process for the production of laundry detergents [B.102]

liquids, powdering of the particles to reduce stickiness, and selection of machine types represent the know-how of the detergent’s manufacturers. A similar system, that is an extension of the tower process, is shown in Fig. 6.3-9 [6.3.1.2]. In this case, if an increase in the amount of non-ionics in the product is desired, a portion is introduced in the first blender and the remainder, after drying of the agglomerates, is added in the second mixer, together with the minors. Fig. 6.3-6 Flow diagram of a non-tower process for the manufacture of laundry detergents with two mixers and a fluidized bed dryer [B.102]


Fig. 6.3-7 Flow diagram of a typical wet granulation process for the manufacture of heavy laundry detergent [B.102]

While the spray dryer, sometimes including size enlargement by agglomeration in the tower or in a secondary fluid-bed agglomerator/dryer/cooler [B.97], is the most commonly used method for the production of basic detergent powders, the manufacture of high-density laundry detergents or of defined product forms for special applications also uses pressure agglomeration techniques (Section 6.3.2). Pigments Pigments are particulate colorants that are not soluble in a solvent or binder liquid. Chemically, they may be inorganic or organic. Since their coloring effect is caused by the reflection and the partial absorption of light, particle size and dispersion are important characteristics. Not only does a smaller particle size produce more brilliance, in some cases, changing the particle size modifies the reflection and absorption properties of the coating, resulting in a different perception of color by the human eye. The oldest known pigments are natural inorganic compounds such as chalk, carbon black, graphite, ocher, sienna, or umber. Originally, the earthy minerals were finely ground between stones or by mortar and pestle, dispersed in water and applied as paint. Carbon black, obtained as soot, a residue from the burning of carbonaceous 6.3.1.3

177

178


Fig. 6.3-8 Type CB high-speed, high-shear mixer that is often used for the densification and pre-granulation of laundry detergent components. The inset shows details of the mixing tools (courtesy L€ odige, Paderborn, Germany)

materials, was often mixed with oil and applied as a cosmetic. Today special mills (e.g., impact (hammer or pin) mills) are used and carbon black of sub-micron particle size, is produced in large quantities by special methods (e.g., the complete combustion of oil, the thermal (often catalytic/Ni) dissociation of CO, the burning of natural gas, Chapter 11). Natural organic pigments include indigo and sepia. Artificial inorganic pigments, produced by chemical or physical modification of inorganic materials, are, for example, barium sulfate, chromium yellow, cobalt blue, iron oxides, red lead, or titanium dioxide. Particularly with the beginning of cyclic organic chemistry, many artificial organic pigments were synthesized, many reproducing and/or modifying naturally occurring compounds, such as indigo, and, more recently, particulate inorganic and organic so called high-performance pigments (HPP) are being developed, which are defined as colored, black, white, pearlescent, luminescent, or fluorescent products for a specific use, with well-defined quality, and an optimized cost [B.103]. All modern pigments must feature excellence of performance, application permanence, and compatibility with health, safety, and environmental issues. The latter are imposed and controlled by local and/or national legislation.


Fig. 6.3-9 Tower process followed by wet agglomeration, fluid-bed drying, and post-treatment for the manufacture of dense laundry detergents [6.3.1.2]

179

180


Examples of some inorganic and organic pigments

Inorganic

Examples

Organic

Examples

Elements

Carbon black Al-flakes, Zn-dust Fe2O3, Fe3O4 TiO2 Cr3O3, Pb3O4 ZnFe2O4 (Co,Ni,Zn)2TiO4 Ti(Cr,Nb)O2 ZnS, CdS CeS2 Pb(Cr,S)O4 Pb(Cr,Mo,S)O4 BiVO3 Na3Al6Si6O24S3 KFe[Fe(CN)6]

Azo-compounds Anthraquinones Benzene derivatives Carbazoles

Bisazomethines

Oxides

Mixed metal oxides

Sulfides Chromates Vanadates Silicates Cyanides

Benzimidazolones Carbazol violet

Diketopyrrolopyrroles (DPP) HP Naphtols Perylenes Phtalocyanides Polycyclics

Anilines Isoindolinones Quinophthalone

The characteristic performance (perception of a particular color) of any pigment is determined by its chemical composition and particle size, the latter being typically in a range from several hundred nano- to a few tens of micrometers, and its ability to uniformly disperse, resulting in an even coloration of the material to which is has been applied. Pigments are available as suspensions or pastes in various solvents or as dry powders. Although the liquid forms have many advantages, packing, storage, and transportation of these products are expensive, as a large percentage is represented by the solvent. Therefore, especially those pigments that are produced and applied in bulk quantities, such as carbon black for use in the manufacture of tires and black, yellow and red iron oxides or manganese oxides for dyeing concrete, are dry powders. Because of their small particle size they tend be dusty, causing dirty workplaces and endangering the health and cleanliness of workers, adhere to walls and tools, bridge in silos, settle during prolonged storage, and, generally, exhibit unfavorable flow and bad metering characteristics. Automatic handling and application systems do not work reliably and the use of such powders becomes very expensive. To overcome these difficulties, pigments are now often micro-agglomerated. Such products are dust free, easily flowing, withstand handling and shipping without degradation, can be metered easily, and will adequately re-disperse back into their fine particulate form during process application. These characteristics are those of “instant” agglomerates (Tab. 6.4-3, Section 6.4.1). Suitable granules are commonly made by tumble/growth agglomeration. Methods include spray drying and the re-wetting of powders in mechanically and gas induced fluid beds (Tab. 6.4-4, A1–A3). Mechanical high-shear mixers, employing pins, plows, and other high-speed tools, often equipped with additional knife heads, are used for the manufacture of smaller batches of specifically color formulated products while bulk masses are processed in drums, on pans, or with gas fluidized beds.


As an example, the widely used method for the production of agglomerated pigments for the coloring of concrete will be described [6.3.1.3]. The granules from this process consist of one or more pigments, one or more binders to aid dispersal of the pigments in the concrete during mixing, and optional other additives. As compared with concrete surfaces that have been decorated by painting and need periodic repair, exposed surfaces made from dyed concrete mix will keep their color for many years without further maintenance. Therefore, typical products are colored concrete blocks and slabs, concrete roofing tiles, landscape bricks and stones, composite blocks and mortars and grouting materials. Agglomerates are made from dry powder mixes by rewet agglomeration in drums and pans or from slurries in spray dryers, often followed by secondary agglomeration, drying, and cooling in suitable equipment, such as fluidized beds [B.97]. To obtain quick and uniform dispersion in the concrete mixer, the pigment is blended prior to agglomeration with a material that acts as a binder, which is specifically selected to also promote dispersion during application. A large number of materials fulfills these requirements [6.3.1.3] but a common, cheap, and effective binder/dispersion agent is lignin sulfonate, available as a by-product from spent sulfite liquor in the paper industry. The granules typically contain up to 15 % binder, feature a particle size distribution with at least 90 % > 20 lm, and have a moisture content of not more than 4.2 %, preferably less [6.3.1.3]. To determine whether the agglomerated pigment disperses well in the concrete mixer and does not produce color spots in or uneven dyeing of the final product, a comparison is made with samples obtained using powdered pigment (control) of the same composition. The best agglomerated products will mix readily with the concrete, requiring the same time, moisture content, and mineral composition, and result in the same or better uniformity of color. The dispersion agent that acts as binder for achieving the required handling, storage, and shipping properties of the micro-granules, plays an important role during final application. As mentioned above, the need for ease of re-dispersion during final application requires that many agglomerated pigments are “instant” products. As briefly discussed in Section 6.4.1, under certain conditions, press agglomeration may yield intermediates with such characteristics (Tab. 6.4-4, A4). On the other hand, pigment granules are also widely used to color plastic master batches. Since very high-shear forces are present during the manufacture of these premixed products, somewhat lower requirements for easy re-dispersion exist in this case (Section 6.3.2.6).

6.3.2


The technologies of pressure agglomeration (Chapter 5) include the most versatile methods of size enlargement because, when certain preconditions are observed, very few restrictions exist that would prohibit the application of one or the other method of this group of techniques. In particular, if the right technique is used, there is

181

182


virtually no limitation to feed size both small and large, unlike the requirement for tumble/growth agglomeration that the feed particle size must be small and generally below a certain dimension (Tab. 5-1, Chapter 5). In high-pressure agglomeration, the feed size must be smaller than the feeder dimensions because the high forces will both crush or deform and then agglomerate the material. For these reasons, pressure agglomeration is also widely used for the size enlargement of chemicals. However, only a few well-established uses of size enlargement by pressure agglomeration in the chemical industry will be reviewed and described in some detail as examples.

Aspartame (Artificial Sweeteners) There are two major applications for pressure agglomeration for artificial sweeteners, such as aspartame and saccharin: granulation and tabletting. Since both product forms must easily disperse and dissolve in cold or hot liquids, they should and in most cases do exhibit “instant” properties. Dry granulation of, for example, aspartame, is carried out by compaction/granulation (Fig. 6.1-13, Section 6.1.2) mainly to reduce dusting and improve flowability and metering. Aspartame, as produced, like many other chemicals, features long, needlelike crystals that tend to clump together in storage and during handling. Compaction in a dry granulation system is normally accomplished with relatively small roller presses that are specially designed and executed for applications in the food and pharmaceutical industries (Sections 6.2 and 6.4). Because production capacities are relatively low, the entire system is integrated in a single machine (Fig. 6.3-10). Crystallized aspartame, possibly pre-mixed with additives such as fillers and disintegrants, is deposited in a feed hopper (1), kept fluid but settling (partial deaeration) with a slowly rotating discharge aid (2), and enters a horizontal screw (4). The feed assembly is mounted on a support (3) that can be moved horizontally and swiveled for easy access and cleaning. The screw, passing through a deaeration box (5), forces the powder feed into the nip between two vertically arranged rollers (6) that produce a more or less continuous compacted strip that is guided to a sheet breaker (7) for disintegration into small pieces. Final size adjustment is accomplished by a granulator (mill) (8) featuring a discharge screen. Because all parts that are in contact with the material to be processed are cantilevered, enclosed, and separated from the drives and controls, which are located in the machine housing, automatic CIP (cleaning in place) features can be added and operated between batches. A major disadvantage of the system shown in Fig. 6.3-10 is that, while totally enclosed and sanitary, a large amount of fines is produced and, at least initially, is part of the discharge. Although the powder feed has been granulated, the fine portion, resulting from leakage at the rollers and the crushing and milling processes, gives rise to dusting, reduced flowability, and potentially lumping. Improved compaction/granulation processes employ multiple crushing, milling, and classification steps (Section 6.1.2) shown in Fig. 6.3-11. This flow diagram, which is the reproduction of a patent drawing [6.3.2.1], first removes and recycles leakage and then applies three breaking/milling 6.3.2.1

6.3 Applications in the Chemical Industry Fig. 6.3-10 Small roller press compaction/ granulation system for the processing of materials such as artificial sweeteners. A product sample is shown in the inset (courtesy Turbo Kogyo, Kanagawa, Japan)

and two classification steps, producing final granules as sieve overs and recycling the fines from crushing as sieve unders. Hydrolyzed starch products, such as maltodextrins, are produced by the partial hydrolysis of cereal (e.g., corn) or root (such as potato) base starches and are commercially available in spray dried, particulate form. As manufactured it has a relatively low sweetness level and, if used alone as a sweetener, the food product can not be characterized as 100 % artificially sweetened, a characterization that is often desired from a marketing standpoint. However, maltodextrins can be used as a bulking agent or carrier for synthetic sweeteners, such as aspartame, and then, the resulting product can be characterized as 100 % artificially sweetened. However, the commercially available spray dried maltodextrin exhibits the same drawbacks as other similar powders: due to its small particle size it tends to dusting, features low bulk density, has unfavorable flow properties, and does not readily dissolve in liquids. Most of these disadvantages could be overcome by rewet agglomeration (Section 6.3.1). During such a process, maltodextrin particles stick together

183

184

6 Industrial Applications of Size Enlargement by Agglomeration Fig. 6.3-11 Flow diagram of a compaction/granulation system with multiple crushing/milling and classification steps and fines recirculation for the reduction of undersized material in the product [6.3.2.1]

and form larger clusters, which produce less dust during handling and feature improved flow and dissolution properties. But because these agglomerates contain many large voids, their bulk density is even lower than the original material; therefore, this product is not well suited for use in automatic packaging machines. The compaction/granulation process (above and Section 6.1.2), preferably applying roller presses, produces agglomerates by breaking and screening densified strips. However, if this is used for maltodextrin, the material quickly begins to stick to the rollers, which, in spite of cooling [B.48, B.97], become warm, forcing frequent shut-downs and resulting in an uneconomic operation. Adding a lubricant to the feed allows operation for a longer time but the highly densified granules exhibit very poor dissolution properties and the lubricant leaves an objectionable scum in or film on the liquid. All drawbacks can be overcome by mixing the maltodextrin with a volatile solvent (e.g., ethyl alcohol) and forming a moist blend prior to feeding the roller press (Fig. 6.3-12) [6.3.2.2]. While, normally, the feed to high-pressure agglomeration equipment should be essentially dry (Chapter 5 and [B.13b, B.48, B.97]), according to the invention, the moist blend contains sufficient moisture that liquid filled pores remain in the compacted strip and, during passage through the nip, a small amount of the liquid is squeezed to the exterior of the sheet to lubricate and cool the rollers thus preventing the maltodextrin from sticking. The sheet is then broken in a closed loop with con-


Fig. 6.3-12 Patent drawing of a compaction/granulation system using a roller press for the densification of an artificial sweetener, mixed with a volatile liquid [6.3.2.2]

ventional equipment, whereby a substantial portion of the trapped liquid is liberated, and then classified. During the final drying step, most of the remaining liquid that occupies the spaces between the compacted particles is driven off. The end result is a mass of densified particles of maltodextrin with acceptable bulk density. Although having a crystalline appearance, the surface topography features cracks, crevices, and fissures, which causes a relatively rapid solution rate in liquids, but the flow and dusting problems of the material are eliminated. Compaction/granulation is frequently used prior to tabletting to improve the flow properties and increase the bulk density, also, if applicable, to reduce dusting and avoid the segregation of powder mixtures. As mentioned several times elsewhere (for example in Section 6.2.2), this pre-agglomeration step is applied to fill the cavities of tabletting machines faster and more reliably. Artificial sweeteners are often converted into small tablets (Fig. 6.3-13) that represent a comparative dosage size. For example, a label on the dispenser shown in Fig. 6.3-13 specifies that “one tablet (1.4 kJ) is equivalent in sweetness to one level teaspoon of sugar (70 kJ)”. Although, in the example, the sweetness of aspartame is reduced by the addition of lactose, L-leucine, and croscarmellose sodium (20 % aspartame and 80 % additives), the tablets, producing the equivalent sweetness, must be

185

186


Fig. 6.3-13 Dosage forms of pressure agglomerated artificial sweetener (aspartame): left) tablets and tablet dispenser, right) granules and portion packages. (Equal is a registered trademark of Merisant Co.)

very small. For manufacture, rotary punch-and-die machines are used featuring multiple dies per station. Fig. 6.3-14 is the photograph of a tool set featuring nine dies. It requires very good flowability to quickly (which is typical of rotary presses) and reproducibly fill such small dies. Since the feed also consists of a mixture of ingredients (above), segregation must be avoided. These requirements are fulfilled when supplying a pre-agglomerated, stabilized, granular (small sized) intermediate product to the tabletting press. Such material is best made by dry compaction/granulation. Fig. 6.3-15 depicts a tabletting line that might be used for the manufacture of artificial sweetener tablets. From right to left it shows a control panel with data recorder, the rotary tabletting machine, in-line de-dusting and transportation, and bulk packaging. The final consumer items (e.g., tablet dispensers) are produced and filled by the distributor. 6.3.2.2 Biocides

Biocides are chemical substances that are destructive to many different organisms. They are used as pesticides in agriculture (Section 6.6) and as water treatment chemicals, particularly also for swimming pools.

6.3 Applications in the Chemical Industry Fig. 6.3-14 Left) upper and right) lower punches and the die insert for the multi-die station of a rotary tabletting press (courtesy Kilian, K€ oln, Germany)

Halogenated (mostly Cl and Br) biocides (such as hypochloride or DBDMH, see Tab. 6.3-1) are the most commonly used chemicals for sanitizing water. They are strong oxidizers and, as dust and products of decomposition (below), irritate respiratory tracts. Therefore, airborne, inhalable fine particles and all dust must be avoided. This is also true for non-chlorine (or bromine) oxidizers that are used for the same

Fig. 6.3-15 Tabletting line for, for example, the manufacture of artificial sweetener tablets. From right to left: Control panel with data recorder, rotary

tabletting machine, in-line de-dusting and transportation, bulk packaging (courtesy Kilian, K€ oln, Germany)

187

188


purpose. An example of such compounds is potassium peroxymonosulfate (often simply called potassium monopersulfate). Stability of oxidizers is also a concern. Halogenated products generate chlorine and bromine. Both are irritating, poisonous gases. Other materials decompose with the production of less objectionable elements or compounds. For example, the relatively stable potassium monopersulfate (Oxone), emits oxygen when it decomposes but at elevated temperatures, may also generate sulfuric acid, sulfur dioxide, or sulfur trioxide [6.3.2.3]. Small amounts of moisture (and, depending on the product, many other chemicals) reduce the stability of all oxidizers. Since these products are often manually handled during, for instance, swimming pool cleaning and maintenance, they must be converted into shapes with properties that avoid nuisance dust and enhance stability. The pure chemicals, which, as produced, may be very fine (e.g., sometimes precipitated), are often mixed with other components for specific performance. For example, Oxone can be blended with many additives, including sodium-sulfate, -carbonate (especially the dense version), -bicarbonate, -perborate, -tripolyphosphate, -metasilicate, tetrasodium pyrophosphate, citric-, malic-, and tartaric-acids and surfactants and fragrances. To maintain stability, all ingredients must be anhydrous or hold hydrated water tightly. Formulated Oxone is used as a shocking agent (auxiliary oxidant) in swimming pools and spas for the purpose of reducing the organic content of the water. It improves the efficiency of the regular sanitizing chemicals such as chlorine and bromine. It can also be used as one part of a two-part disinfecting system for spas and hot tubs with sodium bromide. In such a system, Oxone oxidizes or activates the bromide ion to bromine, which rapidly forms the active sanitizer hydrobromous acid. Upon reaction with bacteria or other water contaminants, hydrobromous acid is reduced back to bromide ion, which can be activated over and over again, thus recycling the active bromine reactant. To reduce dust and increase stability, halogenated biocides or other products, such as Oxone, are often pressed into tablets (Fig. 6.3-16) of different sizes and shapes

Fig. 6.3-16 Tabletting of swimming pool oxidizer compound with a rotary punch-and-die press (courtesy Stellar, Sauget, IL, USA)


(Fig. 6.3-17). These tablets are used in the skimmer or in, often floating, “chlorinators” where they slowly dissolve. Tablets must be formulated and pressed to retain their shape in water without falling apart. They should be so dense that they dissolve on the surface only. For shocking, a granular material that dissolves quickly is produced by compaction/granulation (Section 6.1.2). To produce tablets of sufficient strength and density, pre-granulated material may be used to feed the punch-and-die presses. This is a common practice in all applications where the quick and reliable filling of a die is required and segregation of multiple components that are incorporated in the granulated feed must be avoided (Section 6.2.2). Because of their high activity and relatively low stability, all oxidizers have a limited shelf-life. They must be stored cool and dry. Packaging must be water tight in drums with sealed lids (granular or tabletted, Fig. 6.318a), bags with water-resistant liners (granular, Fig. 6.3-18b), or individually wrapped and welded into plastic (tabletted, Fig. 6.3-18c). 6.3.2.3 Catalysts

A catalyst is a substance that enables a chemical reaction to proceed under more favorable conditions, for example with higher speed or at lower temperature, than otherwise possible. Normally, except for potentially picking-up contaminants from the usually liquid or gaseous components that participate in the reaction(s), the material inducing the catalytic reaction (catalyst) remains chemically unchanged at the end of the process. It may, however, participate in intermediate steps, forming a temporary compound which, in turn, reacts to yield the desired product(s) and regenerates the catalyst. The rate of reaction induced by a solid catalyst is generally proportional to its surface area and the concentration of the so-called active centers or catalyst sites. The latter are locations of high chemical activity on the surface. Since the specific surface of particulate solids increases with decreasing particle size, the effectiveness of solid catalysts

Fig. 6.3-17 Some different shapes of tablets for swimming pool sanitizing (courtesy Stellar, Sauget, IL, USA)

189

190


Fig. 6.3-18 Packaging of agglomerated swimming pool sanitizing products: a) water-tight in drums with sealed lids (granular or tabletted), b) bags with water-resistant liners (granular), c) individually wrapped and welded into plastic (tabletted) (courtesy Stellar, Sauget, IL, USA)

also increases with decreasing particle size. However such powders exhibit typical drawbacks, which include dustiness, low bulk density, and, generally, unfavorable handling, storage, flow, feeding, and metering characteristics. In addition, because the reactants have to be in close contact with the solid catalyst, the vessel in which the processes take place often contains a (stationary) catalyst bed through which liquids or gases or both flow and produce the new compound(s). With time, fine powders pack densely in certain areas, caused by the capture of small suspended particles in pores, thus partially obstructing the flow (Section 8.2), resulting in a non-uniform flow pattern, and leading to an ineffective usage of the catalyst. Therefore, size enlargement by agglomeration is used to convert catalyst powder into shaped, porous bodies with sufficient strength to withstand the mechanical stresses during handling and loading and to survive the prevailing process conditions (including elevated temperatures, if applicable). A packed bed reactor in which solid-catalyzed fluid-phase reactions take place must be filled easily with an optimum amount of catalyst (just enough) to produce a permanent, non-deteriorating bed with high permeability. This has led to the production of porous but strong solid carriers, which are later impregnated with the catalyst. For impregnation the active material is dissolved or dispersed in a liquid and during a post-


treatment (for example drying) it is deposited as a thin coating throughout the pore space of the carrier. The structural integrity of the bed is provided by the carrier particles while the catalyst is made available to the reactants on the very large internal surface area of the carrier. Catalysts used in industrial processes, whether made from the active substance directly or deposited on carriers, are often porous cylindrical pellets produced by medium pressure extrusion (Fig. 5-10b1–b6, Chapter 5). While almost all methods of size enlargement by agglomeration can be used to produce porous bodies from solid powders and post-treatment processes can yield strong pieces with a high accessible internal porosity [B.23, B.78, B.97] that fulfill the previously discussed requirements, pelleting offers a number of advantages. In extrusion, final densification and shaping is accomplished in the die holes by forces that are exerted by differently designed and shaped extrusion tools (rollers, screws, wipers) and the frictional resistance opposing the flow of extruding material [B.97]. Depending on the material’s consistency and the geometry of the die openings, considerable pressures develop in front of and within the die holes whereby wall friction plays a decisive role. The materials to be agglomerated by extrusion need to be formable and exhibit some lubricity. These characteristics are most often achieved by the addition of a sufficient amount of suitable liquid, particularly if the solid itself is hard and brittle. Liquid in a particulate mass produces plasticity, creates a lubrication film, and fills pores during densification. The latter are retained in the product after drying and potentially further treatment, for example by calcination, are accessible from the outside, and provide an internal network of surfaces consisting of catalyst or, if it is a carrier, are available for coating with active material. Extruders that are most often used for the production of catalysts or catalyst carriers are screw extruders and flat die pellet presses (Fig. 5-10b1 and b2, Chapter 5). Both machines can exert high pressures and are sufficiently rugged for the application. Because of design limitations, flat die pellet presses (Fig. 6.3-19) can only produce cylindrical shapes with diameter to lengths ratios of between about 1:1 and 1:3. After drying and calcination, these carrier materials are impregnated with the catalyst and serve as products that can be easily poured into containers to form a stable bed with satisfactory permeability. If the “natural” porosity, which is partially created by pores that are filled with liquid, opened during drying, and stabilized in the calcination process, is not high enough, additional porosity may be produced by adding temporary particulate solids to the feed mix. Such pore building materials are removed from the pellet structure during drying and/or calcination, leaving behind large open voids [B.78]. Nevertheless, in some packed bed applications, the permeability created by the porous, cylindrical extrudates is not high enough. For these, in addition to rods with larger aspect ratio and hollow rods, rings or supported rings (Fig. 6.3-20) and many other, often proprietary shapes (not shown) are made as carriers with specially designed screw extruders (Fig. 6.3-21). Since many of the materials, such as high performance aluminas, kaolins, or molecular sieves, require considerable pressure for extrusion and are quite abrasive, extra heavy duty components, made of special abrasion and corrosion resistant steels, are used in equipment design.

191

192


Fig. 6.3-19 Installation of a flat die pellet press for the manufacture of cylindrical catalyst or carrier pellets (courtesy Amandus Kahl, Reinbek, Germany). For a schematic, partial cut through a flat die pellet press see Fig. 6.5-4a (Section 6.5.2).

Fig. 6.3-20 Various shapes of extruded catalyst carriers (courtesy Bonnot, Uniontown, OH, USA)


Fig. 6.3-21 Special extruder for the manufacture of catalysts or catalyst carriers. A hinged die holder allows the quick exchange of extrusion plates (courtesy Bonnot, Uniontown, OH, USA)

6.3.2.4 Laundry Detergents

As already mentioned in Section 6.3.1, in the 1980s, beginning in Japan, a strong demand developed for solid laundry detergents with a density of more than 700 g/L and a contents of active matter (anionic and non-ionic) of up to 50 %. It was also indicated that, as a general rule, agglomeration was added to the manufacture of laundry detergents to improve the dispersibility of the chemicals in water. Such materials are often described as having “instant” properties. The use of medium-pressure agglomeration (extrusion) is relatively new. The complete system (Fig. 6.3-22 [B.102]) uses mixing of solid components with a lubricant, plasticizing and forming the moist mass with a double screw extruder into small diameter, spaghetti-like strands, spheronizing [B.48, B.97] the product into uniform spherical particles, drying/cooling in a fluidized bed, screening (overs, after crushing (not shown) and unders are recirculated), and post-addition of minors in a final mixer. Extruded particles exhibit some of the highest densities (about 1400 g/L) achieved in detergent manufacture. Whereas other processes deliver small, irregular granulates, the extruded and spheronized particles, called Megaperls, are spherical, uniform, and quite large (about 1.5 mm diameter) (Fig. 6.3-23). Apart from being an extraordinary basis for the manufacture of super-compact laundry detergents, they feature additional advantages, including total absence of dust, very high homogeneity, no segregation of particles due to their uniformity, and excellent flow characteristics. Extruded detergents allow anionic surfactant contents of more than 20 % along with the very high density [B.102].

193

194


Fig. 6.3-22 Flow diagram of an extrusion process for the manufacture of Megaperls [B.102]

Fig. 6.3-23 Appearance of commercial heavy laundry detergent products: a) granular, b) extruded and sheronized (Megaperls) [B.102] (courtesy Henkel, D€ usseldorf, Germany)


Granulation by high-pressure agglomeration, although a logical choice for the densification of particulate solids, was first ruled out because it was assumed that the dense particles would feature low dispersibility. In fact, in the industry, salt (NaCl) is commonly briquetted as an agent for the regeneration of ion-exchange water softeners whereby the resulting product features mechanical stability in water and a slow dissolution rate (Section 6.8.2), clearly a performance not acceptable for laundry detergents. However, as discussed in Section 6.4.1 (Tab. 6.4-4 A4), if the binding mechanisms that develop during high-pressure agglomeration are mainly caused by molecular adhesion (van-der-Waals forces), granules from compaction/granulation (Section 6.1.2) are easily dispersible because these binding forces are reduced to about 1/10th of their strength in liquids. To achieve this condition, only solid compounds should be used for formulating the laundry detergent. The density of the resulting, irregularly shaped granules is high (> 700 g/L). A concern, that the edges of agglomerates, that were produced by crushing a consolidated sheet, break off easily during handling, packing, storage, and application, causing dust, can be overcome by tumbling the granules in a polishing drum with or without the addition of small amounts of additives (e.g., perfumes). During this procedure sharp corners are removed and/or rounded. A final de-dusting step yields a mechanically stable, dust-free product (Section 6.1.2). Tablets, produced from powder mixtures with punch-and-die presses, are always an easy and precise dosage form. They are also convenient for laundry detergents since no dosing and dispensing aids are required. Other advantages include smaller packages as compared with powder or granular products due to their highly concentrated form, extra portability, and a more accurate sense of how many wash portions remain in the detergent box. They were first introduced in Europe in 1997 and in 2000 already held a share of 10 % in the heavy-duty detergent category. In the year 2000, marketing also began in the Japanese and North-American markets. Tablets may consist of one or two (differently colored) layers (Fig. 6.3-24), each providing a specific performance. They are the most compact delivery form of non-liquid

Fig. 6.3-24 Two-layer laundry detergent tablets (courtesy Henkel, D€ usseldorf, Germany)

195

196


detergents. To achieve the best efficiency, all ingredients must be in a dissolved state from the very beginning of the wash cycle. Another specific demand made on heavyduty detergent tablets is sufficient hardness to enable handling, packaging, transportation, and use. Several problems must be resolved [B.102]. *

*

*

The low surface-to-volume ration of detergent tablets adversely affects the dissolution rate. In the presence of surfactants, especially non-ionics, gel phases may form on first contact with water. Gel formation impedes fast dissolution. Hardness or strength of tablets is synonymous with high densification and low porosity, which lowers dispersibility and dissolution.

To overcome these problems, laundry detergent tablets contain special disintegrants [B.97]. These can be classified into four groups or combinations thereof [B.102]. a Effervescents such as carbonate/hydrogencarbonate/citric acid b Swelling agents such as cellulose, carboxymethyl cellulose (CMC), or cross-linked poly(N-vinylpyrrolidone) c Quickly dissolving materials, such as halogen (Na, K) acetate or citrate d Rapidly dissolving water-soluble coatings such as dicarboxylic acids. The composition of heavy-duty detergent tablets differs from other forms of supercompact detergents mostly in their contents of disintegrants. Two-phase (two-layer) tablets allow the separation of certain detergent ingredients that might otherwise adversely affect each other during storage, for example enzymes and activated bleach. All manufacturers use rotary punch-and-die machines for tablet production, whereby twolayer presses feature two feeding stations on the turret [B.48, B.97, B.99]. Fig. 6.3-25 is the evoluted (straightened) diagram of such equipment. From left to right: the die is filled with the first component while the lower punch moves down. After scraping-off the excess amount of feed, compression begins by moving the upper punch down. Pre-compaction is achieved between the small press wheels and, after a dwell time, final compression takes place between the large wheels. The upper punch is

Fig. 6.3-25 Evoluted (straightened) diagram of a rotary double-layer punch-and-die press (courtesy Fette, Schwarzenbek, Germany)


then retracted to make room for filling and volume adjustment of the second component. Compaction follows. At the end of the cycle (one revolution of the turret) the upper and the lower punches move up to expel the finished tablet. In the design depicted in Fig. 6.3-25, most of the travel is done by the upper punch. The lower one only moves during the filling and adjustment steps and for tablet extraction. In other machines both the upper and the lower punches participate during compression, performing double sided compaction [B.48, B.97]. Other laundry washing additives are also produced and applied as tablets. Fig. 6.3-26 depicts a water-softening product for European automatic washing machines. The

Fig. 6.3-26 Two-phase (two-layer) water softening product for European automatic washing machines (Calgon is a registered trade mark of Benckiser, Ludwigshafen/Rh., Germany)

197

198


machines include a heating coil that gets coated with calcareous deposits if hard water is used. The tablets have two phases but contain no active detergent components, brighteners, or perfumes, are added with the wash load, and remove the hardness when washing begins. Therefore, they must also feature quick disintegration, which is again achieved by the inclusion of disintegrants (above). By using this additive regularly, the machine drum and the heating coil are kept clean and continue to perform efficiently for a long time. As shown in Fig. 6.3-26, to increase shelf-life, laundry detergent and additive tablets are often individually wrapped. 6.3.2.5 DMT (dimethylterephthalate)

Chemicals that are produced as a melt are cooled and solidified with drum coolers. Such machines feature a polished, internally cooled, slowly rotating metal drum that dips into a bath of molten material (Fig. 6.3-27). A thin layer of material is picked-up by

Fig. 6.3-27

Rotary drum cooler (dip feed flaker) (courtesy GMF- Gouda, Waddinxveen, The Netherlands)


and sticks to the surface of the drum as it emerges from the liquid. After the coating is cooled and solidified it is scraped from the drum and forms thin flakes. DMT, a major starting material for polyester fibers and coatings, is processed by this method. It is synthesized in chemical plants as an intermediary product and must be shipped to the manufacturers of fibers, which are normally not part of the same chemical complex. In fact, DMT is made in a few locations only in large quantities and shipped for remelting and final processing to many users throughout the world. The flakes have bad flow properties, feature extremely low bulk density, and, if packed, for example into bulk bags, initially require a large volume. On the other hand, they tend to settle in the flexible containers, forming lumps that are difficult to discharge and are not suitable for easy feeding to remelt equipment at the fiber manufacturing plant. These problems are resolved by briquetting. The DMT flakes are readily densified and bonded under moderate pressure in a roller press that is equipped with a screw feeder and pillow shaped pockets. The briquettes have high density and are well formed. During a screening operation they are separated into singles and leakage and land areas that surround the briquettes and are rubbed-off on the screen, are recirculated to the press. The briquettes have a volume of about 4 cm3, are about 20 mm square and 14 mm thick, pack, store, and handle well, and can be easily metered and remelted. Other flaky chemicals obtained from drum coolers or dryers that have similar properties as those of the fresh DMT can be similarly converted into dense free flowing intermediate products by pelleting, briquetting, or compaction/granulation. Drum dryers feature heated drums that either dip into a bath of suspended particles or are coated by means of axial dispensers with the suspension. As in the case of drum coolers, dry solids are removed from the drum with scrapers and are typically bulky and dusty before suitable agglomeration has taken place. 6.3.2.6 Pigments

Agglomerated pigments must feature instant properties. As in many other industrial applications requiring similar characteristics, such as laundry detergents (above), and food (Section 6.4.2) or animal feed products (Section 6.5.2), granulation by high-pressure agglomeration was first ruled out because it was assumed that the dense particles would feature low dispersibility. However, as discussed in Section 6.4.1 (Tab. 6.4-4 A4), if the binding mechanisms that develop during high-pressure agglomeration are mainly caused by molecular adhesion (van-der-Waals forces), granules from compaction/granulation (Section 6.1.2) are easily dispersible because these binding forces are reduced to about 1/10th of their strength in liquids. The widely used inorganic pigments for the coloring of concrete have already been used as examples in Section 6.3.1 to demonstrate the reasons, importance, and results of agglomeration. A major disadvantage of tumble/growth agglomeration is the need for a binder that must be mixed with the powder, forms bonds between the pigment particles throughout the mass to be processed, and ultimately provides the strength of the granular product. The “vehicle” for introducing the binder material(s) and facilitating granule formation is a liquid, so that initially “green” (moist) agglomerates are

199

200


formed. A post-treatment, at least involving drying for the removal of the liquid and formation of permanent bonding, yields dust-free, dry, free flowing, and easily handleable granular product. Because the binder is distributed throughout the granules, the shear forces in a concrete mixer are not high enough to fully and reliably disperse the pigment. Color spots and streaks are the result of such ineffective distribution. To overcome this problem, the binders have been selected from a group of materials that promote the particle dispersion in the presence of moisture (Section 6.3.1). Nevertheless, tumble/ growth agglomeration requires the additional cost of the binder and the post-treatment, which reduce the economics and the profit potential. A more effective way to produce pigment granules, which, if applied for concrete, have a coloring effect that is at least equal to that of the unagglomerated powder, uses compaction/granulation (Section 6.1.2) [6.3.2.4]. Although, as mentioned above, the natural adhesion forces (van-der-Waals) between dry, compacted, inorganic powder (pigment) particles are reduced to 1/10th of their strength in a liquid environment, the addition of a dry additive that promotes dispersion in a concrete mixer improves the ease and uniformity of coloration. The major advantage of the new method is the dry processing, which does not require a post-treatment to dry out liquid and achieve permanent granule strength. The process uses a flow diagram that is fundamentally in agreement with Fig. 6.1-14 (Section 6.1.2). Pigment and a small amount of dispersing agent are thoroughly mixed and then compacted with a roller press. The strip emerging from the rollers is milled and classified to yield product granules sized 0.2–0.6 mm. Oversized material is returned to the mill and re-granulated while all fines (undersized particles) are recirculated and mixed into the compactor feed. Because the highest possible yield of on-size granules, the utilization of the entire pigment feed through internal recirculation, and the absence of dust in the finished product are major objectives, an optimized flow diagram uses multiple (at least two) milling (crushers) and classifying (screens) steps (Fig. 6.3-28).

Fig. 6.3-28 Flow diagram of an optimized granulation process, based on roller press compaction, for (for example) an inorganic pigment


Instead of a roller press a pellet mill may be used for densification. The rest of the system is essentially the same. However, for successful extrusion (pelleting) some lubricant (liquid) may have to be added. Although the temperature rise caused by friction in the die holes may often be sufficient to naturally dry-out the small amount of liquid, the resulting granules do still not exhibit the same uniform quality as those originating from sheets made with roller presses. As exemplified by the application of pressure compaction/granulation for pigments that are applied for the coloring of cementitious materials, this technology may be also used for other inorganic and organic materials. In many cases the reduction in strength of the molecular adhesion forces when they are in contact with liquids may suffice to render the granules sufficiently dispersible without the need for an addition of dispersing agents. 6.3.2.6 Plastics (Master Batch)

A master batch is an intermediate material that has been formulated for the manufacture of thermoplastic parts by pressing, extrusion, injection molding, and other forming and/or finishing processes. It contains all ingredients that are necessary for achieving the structure, color, strength, and other properties of the final plastic piece. Specifically it may hold pigments, fibers, fillers, chemical reactants, and many others. Most master batches are formulated from solid powders by mixing the different ingredients in a conventional (normally batch) mixer. Since considerable differences in size and shape of the individual components exist, for example between pigments and fibers and thermoplastic powders, segregation of the dry blend is a major concern. Also, the end users prefer to only re-melt the fully formulated feed and process it into parts for industrial and consumer applications. To solve the segregation problem and guarantee a uniform distribution of all the components in the master batch, the thermoplastic components of the blends are partially or completely softened or melted in double shafted screw or other high-pressure extruders [B.97]. Heat is created by the conversion of mechanical into thermal energy in the high-shear mixing and homogenizing section of the extruder or provided by external heating. The plasticized mass is passed through holes in a die plate and the extruded strands are cut into short, normally cylindrical pieces with a diameter to length ratio of between 1:1 and 1:2. If the compound is warm or hot and very plastic, the mouth piece (carrying the die plate) of the machine is under water wherein the strands are cooled prior to cutting (“under water granulation”). 6.3.2.7 Sodium Cyanide

As produced, sodium cyanide (NaCN) is a very poisonous powdery salt. The material is manufactured and applied in relatively large quantities for the treatment of steel, electroplating, and fumigating. Dust, particularly airborne particles, must be avoided to protect workers and allow safe transportation and use. While for the formulation of fumigating agents, agglomerates must be milled again to yield an aerosol powder, applications in the metals industries can accept larger pieces.

201

202


The most common agglomeration technology for the conversion of sodium cyanide into a safe product is briquetting with roller presses. Almond- or pillow-shaped compacts are made. The systems are completely enclosed, equipped with highly effective dust collection systems to safeguard the operators, and executed in stainless steel to keep corrosion in check. The discharge from the roller press is screened within the enclosed system, fines are recirculated internally to the briquetter feed bin, and clean product is immediately packed into sealed containers. This is a typical example of the use of pressure agglomeration for the transformation of a dusty, hazardous chemical into an easily handleable compacted product.

6.3.3

Other Technologies

Other technologies of size enlargement by agglomeration (Chapter 5, groups C and D) are not as frequently used in the chemical industry as the other two major groups (tumble/growth and pressure agglomeration). The applications are more specific and not as well known to the general or even the knowledgeable public. Nevertheless, some of the materials defined as examples in Section 6.3 and discussed in Sections 6.3.1 and 6.3.2 are being modified by other agglomeration technologies as demonstrated below. Aspartame (Artificial Sweeteners) Aspartame is very sensitive to high temperatures and an alkaline pH. Therefore, without modification, it is not suitable for baking applications such as pre-mixes for cakes, sweet (Danish) rolls, and cookies. Suitably coated, it would be capable of resisting these adverse effects. However, aspartame, like other products with similar characteristics, for example niacinamide, a nutritional food supplement (Section 6.4), or ibuprofen, an anti-inflammatory drug (Section 6.2), has a long, needle like crystalline structure, exhibits very bad flow properties, and can not be encapsulated easily. Encapsulation coating is typically carried out in fluid beds (for example the Wurster coater [B.48, B.97]) where the elongated crystals tend to intertwine and form unacceptable soft, fibrous aggregate balls. A patent [6.3.3.1] proposes and industrial applications later use compaction/granulation (12, 14, 16 in Fig. 6.3-29) prior to encapsulation. During the pressure agglomeration process, the needle-like crystals break easily and are incorporated into the structure of the compacted sheet that is crushed and screened to yield granules in the size range 40–840 lm, which readily form a fluid bed in a top spray coater (18 in Fig. 6.3-29). Fines removed with a sieve (16) are recirculated to the compactor feed. Preferred encapsulation materials for the process are molten hydrogenated lipids and waxes. In the case of aspartame, it could be hydrogenated soybean oil stearine. An advantage of this fat coating is believed to be that it effectively protects the artificial sweetener from the moist heat developed during baking [6.3.3.1]. 6.3.3.1


Fig. 6.3-29 Modified patent drawing of a fluidized bed encapsulation process incorporating compaction/granulation as a preparatory step (according to [6.3.3.1])

6.3.3.2 Laundry Detergents

In Section 6.3.2 it was mentioned that several problems must be resolved if laundry detergents are tabletted and three specific areas were listed [B.102]. Two have to do with the fact that the dissolution rate of compacted material is low and that, to arrive at acceptable performance, disintegrants have to be added that assist in the break-up during application. Another approach is the manufacture of a friable, unstable detergent tablet using low forces during compaction. Such a product already breaks-up just by handling it so that it can not be packed or stored. To stabilize these tablets they have to be coated with materials that form a hard, easily melting or dissolving layer or shell. Suitable materials are C12–14 dicarboxylic acids, which are combined with a disintegrant [B.102]. Suitable equipment for such coating are enrobers, more typically used in the food industry (Section 6.4.2). Fig. 6.3-30 explains the operation of such equipment. Pieces to be enrobed are transported on a wire mesh belt and, after entering the machine from the left, are kept in place by a holding down device. Coating is accomplished with a bottoming roller and by pouring mass onto the pieces from above.

203

204


Fig. 6.3-30 The operating principle of an enrobing machine (courtesy Hosokawa Kreuter, Hamburg, Germany)

6.3.3.3 Pigments

Since the coloring effect of pigments is caused by reflection and the partial absorption of light, particle size is one of the most important characteristics. Not only does a smaller particle size produce more brilliance, in some cases changing the particle size modifies the reflection and absorption properties of the coating, resulting in a different perception of color by the human eye. To be independent of unintentional changes, more and more pigments are precipitated or synthesized as sub-micron (nano) particles, which initially exist in suspension. To be able to filter, wash, and dry the solid product without loosing the pigment properties, an agglomeration process is necessary. Later, the binder and residual water are evaporated and the dry granulated pigment can be easily redispersed for use in the coloring of surfaces. In the example presented here, an organic pigment (di-methyl-quinacridone) is produced in suspension during the drowning-out (dilution) of highly concentrated sulfuric acid in which the synthesized pigment is dissolved [6.3.3.2]. Because the precipitated particles are in suspension, an immiscible liquid binder (di-N-butylamine) is used for agglomeration. During the “classic” (immiscible binder agglomeration) method it is first necessary to disperse the binder phase into small droplets. This occurs in a stirred vessel with baffles (Fig. 6.3-31). The pigment particles coalesce with the droplets and form agglomerates [B.97], which can then be subjected to the previously mentioned processing. As in all growth agglomeration processes, binder dispersion and formation of nuclei are time consuming (Fig. 6.3-32). A new process reduces this time by adding a binder emulsion, which was prepared in a separate step [6.3.3.2]. Fig. 6.3-33 is the comparison of agglomerate growth with and without binding liquid pre-dispersion by a rotor-stator mixing device. Disadvantages are a complicated installation scheme, high-power consumption, and a modification of the physicochemical properties of the system due to the temperature rise caused by the conversion of energy.

6.3 Applications in the Chemical Industry Fig. 6.3-31 Stirred vessel with baffles for the “classic” immiscible binder agglomeration [6.3.3.2]

In the second new process, binder nano-droplets are produced chemically in-situ by using the acid–basic properties of the binding agent [6.3.3.2]. The di-N-butylamine is soluble in sulfuric acid solution resulting in an ionized compound, the sulfate of di-Nbutylamine. If the neutralization of the suspension is carried out after injection of the

Fig. 6.3-32 Comparison of growth curves obtained during different operating conditions of the “classic” immiscible binder agglomeration showing long time intervals for the formation of nuclei [6.3.3.2]

205

206


Fig. 6.3-33 Comparison of agglomerate growth with and without binding liquid pre-dispersion by a rotor-stator mixing device [6.3.3.2]

amine, very fine droplets of the binding liquid are directly and homogeneously generated within the suspension (reaction between di-N-butylammonium and hydroxide ions). This results in an optimal contacting of binder droplets with solid particles and a fast and uniform increase in agglomerate size with narrow distribution. In addition, as shown in Fig. 6.3-34, very fine particles disappear with time, which makes further filtration and washing steps easier. An already classic packaging method for pigments is the microencapsulation of toner particles. Such products are dust-free, free flowing, and feature a long shelflife. They can be used directly in copying machines where, after electrostatically assisted deposition on paper, pigment is liberated between pressure rollers by rupturing the capsules. Another application is the embedding of uniformly distributed

Fig. 6.3-34 Fast and uniform increase in agglomerate size with narrow distribution obtained during optimal contacting of in-situ formed binder droplets with solid particles [6.3.3.2]

6.4 Applications in the Food Industry

microcapsules between the fibers during paper making, yielding self-copying paper. In this case the capsules must be small enough to fit into the thickness of the paper structure and strong enough to avoid smudging during normal paper handling. They must be positioned in close proximity so that, after destruction of the membrane by the force of the writing tool, a clear and continuous line is produced.

6.4

Applications in the Food Industry Besides meats and greens, humans first consumed kernels of cereal grasses by eating them raw, green or naturally dried. Later, grains were liberated through manual threshing and separating the chaff by “wind sifting”, early applications of unit operations of mechanical process technology. Maybe before learning to grind dry grains, green kernels were flattened to yield a flaked product. As human development continued it can be assumed that agglomeration was first used during the making of bread. “Whole grain” bread or flat cakes, made with flaked kernels, were most probably the first manufactured foods. As already mentioned in Chapter 2, for the making of “advanced breads” dry grains were ground between rough (mill) stones. The resulting flour (particulate solids including an inherent binder, starch) and liquid additives (additional binder for plasticity and “green” bonding) were mixed and kneaded; this mass was then formed (agglomerated) into loaves and finally “cured”, the removal of much of the moisture that was added during the mixing, kneading, and agglomeration steps, to obtain structure and permanent bonding, first by drying in the sun and later by baking in an oven. The “technology of bread making” combines all unit operations of mechanical process technology (Fig. 2-2, Chapter 2) and the components of a complex agglomeration process, including separation of grain kernels from the chaff, preparation of solid feed particles by milling (adjustment of particle size and activation of the inherent binder, starch), mixing and kneading of particulate solids with additional binder(s), agglomerating the mass into a “green” form, and a “post-treatment” (curing, drying or baking, heating) to provide strength and texture. Very early in history it was also found that the porosity of the final product could be modified (increased) by making use of gases that are produced during fermentation (initiated by sourdough or yeast) and result in bubbles in the moist, rising mass. These voids (pores) are stabilized by “strengthening” during post-treatment (baking). Although the availability of bread is documented in some of the earliest human writings, pictorial descriptions of the process only date back to the Egyptians (Fig. 6.4-1) and the Romans (Fig. 6.4-2) [6.4.1]. Fig. 6.4-1 describes the mixing, forming and baking of bread and includes special, most probably devotional shapes (top center). It is reproduced from a painting in a tomb from the time of Pharaoh Rameses III (about 1175 BC) [6.4.1, 6.4.2].

207

208


Fig. 6.4-1

The bakery of Rameses III, about 1175 BC [6.4.1, 6.4.2]

Fig. 6.4-2 is a lithograph of a stone relief from 50–25 BC on the tomb of a well-to-do baker named Marcus Vergilius Eurysaces, which now stands at Porta Maggiore in Rome, Italy. It shows in the lower two strips the (mule powered) grinding of flour (center) and mixing of dough (lower right), the forming and baking of the loaves, and the weighing and transportation of the bread in the upper strip.

Fig. 6.4-2 Lithograph of stone relief from 50–25 BC on the tomb of a well-to-do baker named Marcus Vergilius Eurysaces in Rome [6.4.1]


Bread is still a major basic food for most people on earth and its manufacture has not changed in principle. Fig. 6.4-3 shows a modern bakery [6.4.1]. Other than obtaining pre-processed components (flour, yeast, salt, and potentially other ingredients(3)) and using mechanized kitchen tools and modern ovens (1) or other “curing devices”, even in advanced societies individuals still mix, knead, and form bread in much the same way as people did millions of years ago. Bakeries where large amounts of bread are produced for sale, have evolved into medium to large industries. In particular, big bakeries serving customers in national regions or selling to international markets, have become bread-making factories: thousands of loaves of bread are made on continuous production lines, which use extruders (17) for the agglomeration step and in which extensive instrumentation is employed for automated handling, processing, storage, and packing. During the past 100 years, many more methods of food processing have developed into sizable industries, many of which use size enlargement by agglomeration for the beneficial modification of intermediate components (Section 6.4.1) and final products (Sections 6.4.2 and 6.4.3). A major influence on the development of large food manufacturing facilities has been the availability of reliable long-distance transport by railroads, steam ships, and, more recently, trucks and refrigeration. The latter played a major role in the evolution of food processing. After Linde invented the refrigerator using ammonia compression/condensation in 1874, within a few years large cooling plants were built around the world, which allowed the long-term storage of food for transportation to centralized processing facilities where the raw components are modified and combined to yield new food products. As mentioned above, many employ agglomeration at some point of manufacturing.

Fig. 6.4-3

Diagram of a modern bakery [6.4.1]

209

210


The latest development in food technology is the manufacture of engineered products. Additives are often used as functional components [B.71] to obtain materials with specific, predetermined, and controlled properties. They are formulated from particulate ingredients and then agglomerated to yield consumer products that feature desirable characteristics. For the new food groups, descriptive names have been coined during the past years [B.97]. For example, “convenience foods” can be easily and quickly used, such as “instant” soups, sauces, and drinks, or products that were recombined from fine, ground food stuffs, contain already the correct amount of spices and other aromas, and, after quick and easy preparation, feature a texture and taste that pleases the palate. “Functional foods”, also called “designer foods” on the other hand, have been treated to eliminate unhealthy ingredients, such as fat. They are then recombined with additives that replace the removed components without sacrificing the “mouthfeel” that is expected from the untreated product. Functional foods may also contain dietary additives that make a product particularly acceptable for a special group of often chronically sick people, such as diabetics. The market for food additives is growing over-proportionately, largely due to the increasing production of more nutritious and better balanced convenience and designer foods. Calorie reduction agents represent the largest segment. “Fun foods” comprise the wide range of modern sweets and snacks in which mostly sugar and fat-based binders are applied to obtain agglomerates, for example bar-shaped products (Section 6.4.2), from a multitude of ingredients for the consumer market. As new food products are developed and their production is industrialized, problems arise and have to be solved because of the need for mechanized and automated processing. For example, for centuries the baker realized that different flours require different amounts of additives and, by experience, modified the recipe to obtain the desired consistency of the dough. It was also known that, for cakes and cookies, different results were obtained after baking if the sugar had not been fully dissolved or the fat was not finely divided. The beating of a mixture of water or milk with sugar and fat, often in a hot water bath, yielding an emulsified, foamy mixture as an ingredient of dough, was an important part of baking and was modified as needed by changing the composition and the preparation methods or times. Such individual visual and tactile evaluation by a human is, of course, no longer possible with an automated production line, producing high-quality products reliably and reproducibly, ready to be packed, shipped, stored, sold, and consumed. It became necessary to analyze and define food characteristics and their modifications and process parameters, knowledge of which were previously transferred from generation to generation and applied by experience. It also became necessary to measure, apply, and control all variables to be able to guarantee the desired product quality. While, in the past, some modifications in the presentation of individually prepared foods was accepted, merely reflecting on the skill of the baker or the cook, a new understanding of quality by the consumer requires absolutely consistent appearance, composition, texture, taste, and many other attributes, which may be specially defined for a particular product. Questions of shelf life and the need for specific methods of packing also became important.


As in many other areas, but most easily described in connection with food products, industrialization converted the manufacturing of foods from a craft to a mechanized process technology, which is controlled by chemistry and physics and their interrelationships (Chapter 12). Since the 1980s, the polymer science approach to the study of the glassy state, glass transitions, and their importance for structures, properties and water relationships in food materials, products, and processes was recognized by a growing number of food scientists and technologists [6.4.3]. As a result, the following questions arose (Chapter 3 in [B.53]): “What is a glass? What is a glass transition? Why is the temperature at which a glass transition occurs (Tg) so important to the processing and storage stability of so many foods? Why is the effect of water, used as a plasticizer, on Tg of such widespread relevance to food products and processes? Why are considerations of non-equilibrium glassy solid and rubbery liquid states in foods more germane to issues of food quality and safety rather than equilibrium phases? Why are the kinetics of heat/moisture processes for foods and of deterioration in food systems during storage more often appropriately interpreted in terms of the Williams–Landel–Ferry (WLF) rather than the Arrhenius equation? What is the “food polymer science” approach with its central concepts of “glass dynamics” and “water dynamics”, and why has this research been so useful to the study of glasses and glass transition in foods?” Answers to some of these questions are as follows. A glass is an amorphous (noncrystalline) solid; it is actually a supercooled liquid of such high viscosity that it exists in a metastable solid state in which it is capable of supporting its own weight against gravitational flow. In a food agglomerate it represents the solid-bridge binding mechanism. A glass transition in amorphous systems is a temperature-, time-, and composition-dependent, material-specific change in the physical state, from a rigid glassy solid to a rubbery viscous liquid or vice versa. In the polymer science approach, the importance of the glassy state, the glass-transition temperature, Tg, together with its location relative to the temperature of storage, and the critical role of water as a plasticizer of food glasses (depressing Tg with increasing moisture content) were recognized as controlling the quality, safety, and storage stability of a wide range of food systems. An increasing awareness of the inherent non-equilibrium nature of most food products and processes has developed. This is exemplified by the development of intermediate moisture foods (IMFs) in which amorphous carbohydrates (polymeric and/or monomeric) and proteins are major functional components. Glass dynamics deals with the time- and temperature-dependent relationships between composition, structure, thermodynamic properties, and functional behavior of foods. It focuses on the glass-forming solids in a food system containing water, the resulting glass that can be produced by (often drying and) cooling to T < Tg, and the effect of the glass transition and its Tg on processing and process control. Temperatures during individual processing steps, for example during agglomeration, may be deliberately chosen to be first above and then below Tg (providing the binding mechanism). Some selected examples of food systems whose behavior is governed by dynamics far from equilibrium and of practical problems of food science and technology posed by their non-equilibrium nature, are presented in Tab. 6.4-1.

211

212

6 Industrial Applications of Size Enlargement by Agglomeration Tab. 6.4-1 Some selected examples of food systems with non-equilibrium behavior (adopted from Chapter 3 in [B.53]) * *

* * * * * *

Water-vapor sorption/desorption hysteresis in concentrated food polymer systems. Graininess and iciness in ice cream; reduced survival of frozen enzymes and living cells; reduced activity and shelf-stability of freeze dried proteins. Caking of dry powders; sugar bloom on chocolate. Recipe requirements for gelatine deserts. Cooking of cereals and grains Expansion of bread or collapse of cake during baking; effects of flour and sugar on cookie baking. Baked goods become stale Effects of sugar-water glasses and rubbers on texture and storage stability.

Cookies and crackers have been used as examples for research on how the food polymer science approach expands the quantitative and practical knowledge of commercial processes [6.4.3]. A sucrose phase diagram (Fig. 6.4-4) was developed as a tool for the understanding of cookie and cracker baking. In many different food products and processes, the glass-forming and crystallizing behaviors of sucrose constitute key functional attributes. In Fig. 6.4-4, center left, at typical temperatures prior to baking, the locations of lean (low sugar/fat ratio) and rich (high sugar/fat ratio) cracker doughs are sufficiently above the glass curve, so that it is easy to dry these products during baking without structural modifications. When the temperature is raised to the vaporization curve (for example, vertical line from the location labeled “rich”), water in the dough begins to evaporate. As baking continues, more water is removed and the concentration of dissolved sucrose increases. Depending on how much flour, sugar, and water are added to the mixture for highsugar cookies, and on how much crystalline sugar dissolves prior to the beginning of baking, the final state of this dough may be on either side of the solidus curve (A or B in Fig. 6.4-4). Then, when baking begins, either water evaporates first and additional

Fig. 6.4-4 Phase diagram for the sucrose-water system, illustrating the locations of the glass, solidus, liquidus, and vaporous curves and the points Tg’ and Te (eutectic melting temperature) corresponding to the intersections of the liquidus/glass and liquidus/solidus curves, respectively [6.4.3]


sugar dissolves later (from A) or additional sugar dissolves first and water evaporates later (from B). Once all of these doughs are baked and cooled, their state falls into a box of final product conditions (D) that spans the glass curve. For typical cracker doughs containing sucrose (“lean” or “rich” in Fig. 6.4-4), the relatively small amount of sugar is so far below its saturation limit in water that it can be completely dissolved during dough mixing, remain in solution during baking, and then, without likelihood of recrystallization, convert to the glassy solid state in the finished product. In contrast, in the dough for a cookie the sucrose content may be already higher than the saturation concentration (67.5 % at 25 8C); then (B in Fig. 6.4-4), part of the remaining crystalline sucrose will dissolve during baking (along Tsolidus in Fig. 6.4-4), depending on the time/ temperature/moisture-loss profile of the oven. Nevertheless, sufficient moisture loss during baking of any (high sucrose) cookie dough (either A or B in Fig. 6.4-4) can create a supersaturated sucrose solution (located within the metastable rubbery region, C in Fig. 6.4-4), from which some sugar may recrystallize during final baking, cooling, or product storage (Fig. 6.4-5), conditions that are not acceptable from a quality point of view. Other food systems can be characterized and investigated with similar phase diagrams. As explained earlier (Fig. 6.4-4), amorphous foods may enter the rubbery state, in which contacting or colliding particles adhere to each other, either at constant moisture or due to an increase in temperature. Very recently the following definitions were formulated (Chapter 5 in [B.108]). “The physical state and physicochemical properties of food components affect food behavior in processing and storage. Many of the food components can exist in the amorphous state, especially at low temperatures and/or low moisture contents. Amorphous materials may exist in a solid-like, “glassy” or in a viscous, “rubbery” state. The

Fig. 6.4-5 Diagram of the transitions of food materials, which can be present as crystalline solids, glasses, rubbers, liquids, or in solution [B.108]

213

214


transitions of materials, which can be present as crystalline solids, as glasses and rubbers, as molten liquids, or in solution, are shown schematically in Fig. 6.4-5. The glassy state is not an equilibrium state but is determined by kinetic considerations, such as rates of cooling or dehydration. However, once established, the glasses are stable until the temperature exceeds the glass-transition temperature Tg.” Because water plasticizes hydrophilic food components, their glass transition is strongly dependent on water content. The effect of water on the glass-transition temperature of several amorphous carbohydrates, calculated with the Gordon Taylor equation [B.79], is depicted in Fig. 6.4-6. Within the range of the materials shown, Tg decreases with lower average molecular weight and/or increased concentration of plasticizer (water). The rubbery and glassy states of amorphous food products play important roles for agglomeration and caking. As the product temperature exceeds Tg, amorphous materials enter the rubbery state and the decreasing viscosity induces flow, deformation, and bonding. The latter may be desired and is stabilized when the glassy state of the final product is obtained at or below Tg by drying and/or cooling, or it may initiate caking, undesired build-up during food processing [6.4.4]. For example, the solids in dehydrated fruit juices contain mostly fructose, glucose, and sucrose. In combination, their Tg is estimated to be below room temperature (Fig. 6.4-6), which makes them very sticky even at typical ambient temperatures; to avoid build up during dehydration in spray dryers it is recommended to cool the walls. In another example, as the amount of high molecular weight carbohydrates, such as maltodextrins, added to infant formula grows, Tg is raised, which increases the stability against caking if the product is kept dry. Therefore, while the transition into the glassy state stabilizes the bonding of agglomerated products, a strict control of moisture content and storage at low temperature is typically required to avoid undesired caking and deterioration of many food products.

Fig. 6.4-6 Effect of water on the glass-transition temperature of several carbohydrates, calculated with the Gordon-Taylor equation [B.108]


Since in many cases, particularly if the food powders are hygroscopic, adsorption of moisture from the air during processing and packing and later during storage, due to liquid migration through package walls, is already sufficient to substantially change Tg, anticaking agents may have to be added to improve stability (Chapter 4).

Further reading

For further reading the following books are recommended: B.21, B.26, B.33, B.40, B.44, B.49, B.50, B.53, B.54, B.56, B.59, B.67, B.72, B.77, B.86, B.89, B.93, B.94, B.108 (Chapter 13.1). Books mainly devoted to the subject matter are printed bold.

6.4.1


In mammals, milk is nature’s food for the offspring and, therefore, represents one of the best and richest nutriments for this animal group, including people. Reduced-fat milk from which only cream has been skimmed and, to a lesser degree, whey from which fat and coagulated milk solids have been removed (mostly for the manufacture of cheeses) are still highly nutritional foods. They are processed for a number of reasons. The evaporation of milk has been known for many years, even as early as 1200 when Marco Polo described the production of a paste-like milk concentrate in Mongolia [6.4.1.1]. Approximately 600 years later the concentration of milk and of other liquid food products, for example extracts such as coffee, was taken-up as an industrial technology, eventually ending in the production of a dry powder. During water removal, pronounced changes in physical structure and appearance take place. Since the process starts with a thin, water-like liquid and ends with a dry powder, it was found that one method of liquid removal is not optimal for all conditions In the food and dairy industry, the methods listed in Tab. 6.4-2 have been adopted for liquid removal. Tab. 6.4-2 indicates that sometimes liquids containing small amounts of suspended and/or dissolved solid matter may be concentrated first by evaporation. This was oriTab. 6.4-2 Methods that have been adopted in the food and dairy industry for liquid removal [6.4.1.1, B.97] Evaporation Spray Drying Vibrating Fluid Bed Integrated Fluid Bed Integrated Belt Drying

Changing from a water like liquid to a high viscosity concentrate Transforming the concentrate into droplets and further evaporating water to get a dry powder Introduced for additional drying and cooling to improve drying efficiency and powder quality Further improvement of drying economy and possibility of drying difficult materials Adding a moving belt dryer at the bottom of the drying chamber; applied for materials that are extremely difficult to dry by spray drying.

215

216


ginally carried-out in an open pan, later in more economical forced recirculation evaporators, and now mostly in highly efficient falling film evaporators [6.4.1.1]. In any case, spray drying has become the heart of all liquid-removal facilities for the production of a dry powder. The associated fluid bed or belt dryers improve drying efficiency and powder quality and allow the processing of difficult materials from which, for example, the final moisture content can not be removed easily or which require cooling. The reason for liquid removal and manufacturing of a powder is to reduce the volume and mass of the liquid product and to stabilize the dry concentrate. Later the dry powder is reconstituted in water or another suitable liquid. During spray drying, particularly if the solid content of the liquid feed is low, the spray droplets must be small to achieve the necessary removal of liquid. This, in turn, results in the production of tiny solid particles which, because of the ensuing drying mechanism, are hollow (Fig. 6.4-7). The reconstitution of a powder consisting of small, light particles in liquid is difficult because they exhibit bad wetting characteristics, which results in powder floating on the liquid surface. Intensive mixing is required to accomplish reconstitution, which is still ineffective and time consuming. Tab. 6.4-3 defines the mechanisms that must occur if a powder is dispersed (nonsoluble particles) or dissolved (soluble particles) in a liquid. All three or four phases of dispersion or dissolution proceed individually whereby some overlapping may occur, depending on the amount of material involved. In the food industry good reconstitution of powders in a liquid, which is a function of time, is associated with the term “instant product” and is often synonymous with agglomerated materials [B.97]. In addition to the inter-particle porosity of the powder, pores in the agglomerates assist in the desired quick liquid penetration. Agglomerates are also larger (heavier) thus improving submergence. The binding mechanisms holding agglomerates together must be such that they easily and quickly break-up in the liquid.

Fig. 6.4-7 Micrograph of an overheated spraydried particle that has burst, showing that it is hollow (courtesy Niro A.S., Soeborg, Denmark)

6.4 Applications in the Food Industry Tab. 6.4-3 Mechanisms occurring during the dispersion and, respectively, dissolution of powder in a liquid 1. Penetration of liquids into the dry powder (also called wetting) 2. Submergence of the powder in the liquid (also called sinking behavior) 3. Break-up of the powder mass into the primary particles (also called dispersibility) (4. If the solid is soluble, dissolution of the primary particles (also called solubility)).

Manufacturers and/or users have a more or less well-defined procedure to determine the maximum allowable time. Typically, complete dispersion or dissolution should be accomplished within a few seconds in warm liquid and in about 30–60 s in cold liquid. Instant agglomerates may contain certain substances, such as fibers and other disintegrants, which swell on contact with liquid and assist in break-up during the dispersion phase [B.63, B.97]. Products from powdered food materials with instant characteristics can be obtained with a variety of different, rather conventional processes (Tab. 6.4-4); most of these use agglomeration techniques (A in Tab. 6.4-4). Because granule size should be small and porosity must be high, instant food products are most commonly manufactured by rewet agglomeration in mechanically agitated beds or fluidized beds. Spray drying combined with mixer agglomeration or fluidized bed agglomeration, in which turbulent particle movement is induced by flowing gas, are also often used. The term “instant” is normally used in the food industry for drink powders, including milk, coffee, tea, soups, sauces, and the like. All instant products must be able to disperse quickly and, if applicable, dissolve in a specific liquid at any temperature, particularly also at ambient or even cold conditions, without residue and sediment. Spray drying has been found to be the most suitable process for converting milk into a dry powder while still keeping the nutritional properties. In the modern dairy industry the drying of milk dates back to around 1800 with larger systems introduced around 1850, but early methods required the addition of sugar, sulfuric acid or alkali and, therefore, the product could not be considered pure. Breakthroughs in the production of high-quality milk powder occurred at the beginning of the 20th century. Tab. 6.4-4 Principles that are most commonly used to manufacture instant products from powdered food materials [B.97] A

Agglomeration techniques

B

Techniques utilizing other processes

A1

Rewetting of powders in fluid beds A1a mechanically induced A1b gas induced Spray drying and agglomeration Combinations of A1 and A2 Press agglomeration A4a Compaction/granulation A4b Extrusion/crumbling

B1

Improvement of wetting with additives (surfactants) Improvement of wetting by extraction (e.g. of fat) Improvement of solubility (e.g., amorphous structure)

A2 A3 A4

B2 B3

217

218


Stauf in Germany in 1901, and Grey and Jensen in the USA in 1913, applied for patents on the use of spray nozzles. Rotary atomizers were developed in 1912 in Germany by Kraus and in 1933 in Denmark by Nyrop [6.4.1.1]. Because of the danger of bacterial destruction, as for milk, other foodstuffs that are in liquid solution or in the form of a slurry or a paste have limited shelf life. Similar to refrigeration or the addition of preservatives, the removal of liquid reduces this bacterial activity. Foods that are free (or relatively free) of liquid can be stored for an almost unlimited time if kept dry and cool. Such products may contain proteins, carbohydrates (including the most important one, starch), fat, and other ingredients such as vitamins, flavorings, including salts and sugars, emulsifiers, stabilizers, colors, and chemicals. If the dry foodstuffs are powders resulting from spray drying they all exhibit difficulties during reconstitution. As mentioned before, size enlargement by agglomeration improves this situation (Discussion of Tab. 6.4-3). During spray drying, agglomeration may occur spontaneously or in a controlled manner. Spontaneous agglomeration is mostly caused by coalescence of partially dried particles with dissimilar diameters. Such particles collide with one another due to their different deceleration paths and adhere to each other. As shown in Fig. 6.4-8, the resulting agglomerates are small and normally do not markedly improve reconstitution. Controlled, forced agglomeration provides conditions that are favorable for particle coalescence and/or size enlargement by growth. These methods yield instant products since most powders achieve that characteristic by mere agglomerate growth. Favorable conditions for collisions exist in multi-nozzle spray dryers where two or more atomization clouds penetrate each other (Fig. 6.4-9) resulting in the attachment of larger, often equal-sized particles and the production of larger agglomerates. This is in contrast to what has been shown in Fig. 6.4-8, in which only smaller particles are captured by larger ones.

Fig. 6.4-8 Typical particle from single stage spray drying with small “satellites” attached [6.4.1.1]

6.4 Applications in the Food Industry Fig. 6.4-9 Forced agglomeration by the interaction of two atomized clouds of droplets from opposing nozzles [6.4.1.1]

Fig. 6.4-10

Sketches of fines return methods for nozzle atomizers [6.4.1.1]

219

220


Fig. 6.4-11

Sketches of fines return methods for rotary atomizers [6.4.1.1]

More common, however, is the spraying of concentrate onto particles in a fluidized bed of a down stream dryer. Of course, this is only possible if two or multi-stage dryers are used [6.4.1.1, B.97]. Growth agglomeration (Chapter 5) occurs after re-wetting the surfaces of the partially dried particles. A further technology for the production of instant (agglomerated) materials in a spray dryer uses the recirculation of dry, already pre-agglomerated particles into the spraying zone of the dryer. Recirculating fines are particles captured in cyclones of the dust collection system and product screen undersize. Figs. 6.4-10 and 6.4-11


demonstrate how the recycling dry particulates are introduced into the dryer near the atomization devices (nozzles or rotary atomizers) where they meet and collide with droplets (called primary particles in Figs. 6.4-10 and 6.4-11) that wet their surfaces. Subsequently, collisions form agglomerates that consist of several particles stuck together. Depending on the parameters selected, the final agglomerates discharging from the system can have a size of 100–200 lm. Fig. 6.4-12 is a schematic flowchart of a complete plant for the production of agglomerated dry solids from a liquid, such as milk [6.4.1.1]. On the left, raw liquid receiving and storage is followed by concentration (partial evaporation) in a falling film, vapor recompression system. On the right the concentrate and recycling cyclone fines are fed to a compact spray dryer which is followed by a vibrating fluidized bed dryer/agglomerator/instantizer and a vibrating fluidized bed cooler. Fig. 6.4-13 is the photograph of a fluidized bed machine in a food processing plant. Fines entrained in the effluent gases from the spray dryer and the vibrating fluidized beds are directed to the cyclones from where they are recirculated into the spray dryer. Re-wet agglomeration of dry, fine food material (e.g., from spray drying) to obtain instant products is carried out in special tower structures or in mixer agglomerators. The first instant non-fat dry milk product was marketed in 1954 based on pioneering research by Peebles [6.4.1.1]. Since steam condensing on solids wets surfaces directly and uniformly and, at the same time, transfers heat very effectively, which helps to partially dissolve solids that are then acting as binders (discussion of Fig. 4.5, Chapter 4), steam is used directly [B.97] or added for improved results (Fig. 6.4-14).

Fig. 6.4-12 Flowchart of a complete plant for the production of agglomerated dry solids from a liquid [6.4.1.1]

221

222


Fig. 6.4-13 Photograph of a fluidized bed machine (dryer/agglomerator/instantizer) in a food processing plant (courtesy Niro A.S., Soeborg, Denmark)

Another example of re-wet agglomeration was developed by Nestle´ for the instantizing of milk powders, chocolate flavored beverages, and soups, as shown in Fig. 6.4-15 [6.4.1.1]. It uses a wetting/agglomeration tower (5) in which an about 10 % solution (1) of the material to be agglomerated (4) is sprayed onto the dry feed with a high-velocity flat fan type nozzle (3). The now sticky particles adhere to each other and are dried in a vibrating (7) fluid bed (6) with warm air (8) to a residual moisture content of < 3 % (9) and bonded by solidification of the dissolved material prior to packing. So far, in all examples the method for agglomeration was based on the use of lowdensity gas fluidized beds. As discussed in Chapter 5 and in much more detail in earlier publications [B.48, B.93, B.97], adhesion by coalescence of irregularly moving particles in low-density fluidized beds yields relatively small, structurally loose, and low-strength agglomerates, which (if reconstitution in liquids is desired) exhibit instant characteristics. In most cases these processes operate continuously, but batch operations for very sensitive materials (vitamins) are also possible. Fig. 6.4.1-16 shows as examples the photomicrographs of two typical products (dry milk and coffee extract). Mechanical agitation can also be used for this purpose. A low-density mechanically activated cloud of particles is obtained in the “Schugi Flexomix”. This machine features a vertical open-ended cylindrical mixing chamber in which a shaft, equipped with


Fig. 6.4-14 Diagram of the Peebles instantizer [6.4.1.1]

Fig. 6.4-15

Schematic flowchart of the Nestle´ re-wet instantization plant [6.4.1.1]

223

224


Fig. 6.4-16 Photomicrographs of two typical products: a) coffee extract; b) milk (courtesy Niro A.S., Soeborg, Denmark)

adjustable blades, rotates at high speed (1000–3000 rpm). The number and position of the knife blades and their angle of attack are selected to suit the particular process needs. The shaft is suspended from heavy duty bearings in the drive system above the vertical cylindrical mixing chamber, resulting in a continuous, completely unobstructed discharge of the moist, agglomerated product [B.48, B.97]. The mixing chamber of the “Schugi” consists of a flexible sleeve that is continually deformed from the outside by rollers that are moving up and down, thus preventing build-up on the interior (Fig. 6.4-17). The rollers are mounted in a cage that is pneu-


Fig. 6.4-17 a) Diagram of a cross section through the operating parts of a “Schugi Flexomix”. b) Photograph of the opened-up roller cage of a “Schugi Flexomix”, showing the vertical shaft with the mixing blades after removal of the flexible sleeve that defines the mixing chamber. c) Artist’s conception of the “Schugi” (courtesy Hosokawa Schugi, Lelystad, The Netherlands)

matically operated. Roller cage and mixing chamber are easily accessible for cleaning and servicing (Fig. 6.4-17b). Wet or dry powders are dropped into the upper end of the mixing chamber so that a low concentration of solids in the mixing chamber develops. Hold-up or retention time, which is only about 1 s or less, can be to a certain extent influenced by the angle of the knife blades: they may increase or decrease the free fall speed component of the rotating charge. Agglomeration of the solid particles occurs either by disagglomeration of a wet feed and reagglomeration using the liquid that is available in the feed or by wetting dry powder with binder liquid. For liquid addition, a wide range of spray or atomizing nozzles is available; selection and installation of these wetting arrangements depends on the liquid and the desired product characteristics. Agglomerates from the “Schugi” normally feature a small but somewhat adjustable particle size in the range 0.2–2 mm and narrow distribution. With increasing rotor speed, the width of the particle size distribution tends to become smaller [B.48, B.97]. Because agglomerates have formed by accretion after impacting in the low-density particle cloud, they typically feature instant characteristics. The four photographs

225

226


shown in Fig. 6.4-18 show the quick and complete (instant) dispersion of an agglomerated product in water. Although, theoretically and sometimes practically, instant products can be also manufactured in high-shear mixer agglomerators (Section 6.3.1), in the food industry spray drying and agglomeration (A2 in Tab. 6.4-4) are used as the original, still very common

Fig. 6.4-18 Sequence of photographs showing the quick and complete dispersion of an agglomerated material in water (courtesy Hosokawa Schugi, Lelystad, The Netherlands)


method and, more recently, re-wetting of powders in gas or mechanically agitated particle bed and clouds (A1 in Tab. 6.4-4) and several combinations of the two (A3 in Tab. 6.4-4) are applied as the dominant techniques. Tab. 6.4-5 lists food materials (in alphabetical order) that are processed by the methods described above and Fig. 6.4-19 is a summary in which, in various categories, specific food materials are mentioned and one each (indicated by an asterisk *) is presented microscopically and macroscopically. As usual, the list is not complete and exhaustive but is presented to demonstrate the large variety and wide acceptance of growth agglomeration in this industry. Without going into detail, some low- and high-shear mixer and pan agglomerators are being also used in the food industry (for example, for grated cheese, dry yeast, cake mixes, fat/flour mixtures, food additives, including so-called nutraceuticals). As in the case of pharmaceutical applications (Section 6.2.1), the equipment is executed in food grade stainless steel or suitably coated on those surfaces that come into contact with the product and are designed for easy, thorough cleaning. Such methods are used to increase the bulk density of fine powders, reduce their dustiness, and improve the flowability to ease the packaging and the metering during further processing or final use. Obtaining products with instant characteristics is not a primary objective in such cases. In most applications small granules with particle sizes around 1 mm (0.5– 3.0 mm), narrow distribution, and the above-mentioned desirable physical characteristics are manufactured. For the fast growing market of cereals and snacks and the dietary and health foods, agglomeration plays an increasing role (Section 6.4.2). Agglomerates consisting of

Tab. 6.4-5 List of some dry and dried food materials as well as mixed food formulations that are converted by growth agglomeration into free flowing, granulated, often instant, and dust free products for easy packaging, metering, and reconstitution Artificial sweeteners Baby formulas Cake mixes Cereal dust Cocoa mixes (w&w/o sugar) Coffee extracts, powders, and substitutes Cream Dairy product blends Decaffeinated coffee extract Dextrose Drink mixes Flavorings Flours and flour mixes Food colors Fruit juices (w&w/o fillers/additives) Gravy mixes Herb extracts Malt extract

Milk powdered whole skim low fat non-fat mixed fat filled permeate protein Molasses Pudding mixes Sauce mixes Soup mixes Soy milk and protein Spices, powders and extracts Starches and derivatives Teas, powders and extracts Yeasts

227

228


Fig. 6.4-19 Various categories of and specific food materials that are dried and agglomerated to yield powders; * indicates which material is shown in the pictures (courtesy Niro A.S., Soeborg, Denmark)


Fig. 6.4-20 a) Growth agglomerated cereal mixtures; b) “Rice Krispies” (courtesy Kellogg Co., Battle Creek, MI, USA)

only a few large particles are produced to avoid segregation (Fig. 6.4-20a) or to yield a product that can be served with liquids, especially milk, without loosing its crispiness (Fig. 6.4-20b). Growth occurs in blenders with mixing tools after adding (often warm) binders, such as honey, syrup, caramel, fats, sugar. The resulting mass is dried and cooled, then screened to remove dust, which is used elsewhere, and chunks that are broken and rescreened. It is also possible to extrude a thick sheet with roller extruders (Section 6.4.2), cure it, and break and screen the solidified mass as described before.

6.4.2


Many modern food products are processed and/or finished by methods of pressure agglomeration, particularly extrusion. But other techniques, such as punch-and-die pressing and compaction/granulation, are also used for particular applications.

229

230


Bread making has already been mentioned and described as one of the oldest applications of agglomeration by people. Fig. 6.4-3, items 10, 11, 12, and 17, show roller presses and extruders for the shaping of dough into flat sheets or ropes from which raw bakery goods (bread loaves and rolls) are made by cutting the extrudates. Machines featuring two counter-rotating rollers are the most common equipment for the shaping of modern food products by extrusion. Fig. 6.4-21 shows a toothed roller press. The feed to be processed, solid, viscous, or, generally, plastic masses to which coarse components can be admixed, is drawn in by the teeth on the rollers, transported to the pressure zone below, and extruded through interchangeable dies as strands (ropes) or slabs (sheets). The rollers are timed such that the oscillating die, which extends over the entire working width of up to 1000 mm, wipes the surface of the rollers clean. With this press, considerable pressures can be exerted. The strands or slabs emerge with constant speed that can be varied infinitely. Fig. 6.4-22 shows the Fig. 6.4-21 Diagram of a toothed roller press for the processing of solid, viscous, or, generally, plastic masses to which coarse components can be admixed (courtesy Hosokawa Bepex, Leingarten, Germany)

Fig. 6.4-22 Photograph of the top side of a toothed roller form-press, featuring a clear feed hopper that allows a view into the nip of the toothed rollers (courtesy Hosokawa Bepex, Leingarten, Germany)


top side of such a form-press, featuring a clear feed hopper that allows a view into the nip of the toothed rollers. For some materials and applications, the oscillating motion of the extrusion die is a disadvantage, as it may make the deposition of the extrudates, that require a certain strength to avoid breakage, onto the following equipment difficult. To overcome this problem, the rotary bar roller press is now often preferred. Fig. 6.4-23 shows the principle of this design. Specially shaped bars are inserted in semi-circular grooves in the rollers and held in position with a lip exposed by spring action. In this situation, similar to the action of toothed rollers, the feed material, which is again solid and viscous (plastic) and may also contain coarse components, is pulled into the nip, transported and pressurized, and extruded through interchangeable, differently shaped die plates. Fixed scrapers clean the body of the rollers and bars by momentarily turning the latter ones. After cleaning, the bars return immediately to their basic position. Very high pressures can be produced with this press. The compression ratio and the extrusion speed are adjustable and can be varied infinitely. The two rollers must not necessarily interact with each other. This is depicted in Fig. 6.4-24. Each rotary bar roller operates in a partially closed barrel and is physically separated from the other one. On the feed side a grooved roll assists in feeding, creating a nip from which the material is transported along the barrel wall into a pressure chamber below; as usually, the rotary bar roller is cleaned by a fixed scraper. With this design principle, more than two press rollers and associated feeders can be combined

Fig. 6.4-23 Design principle of a rotary bar roller press (courtesy Hosokawa Bepex, Leingarten, Germany)

Fig. 6.4-24 Diagram of the design of a double rotary bar roller press (courtesy Hosokawa Bepex, Leingarten, Germany)

231

232


(below). Extrusion is accomplished in sheets or ropes. Particularly in the latter case, filled or multi-layer products can be made. Fig. 6.4-25 shows two possible nozzle types for the manufacturing of double layered and hollow strands. Fig. 6.4-26 is a collection of diagrams of cross sections through ropes from single, double, and triple rotary bar roller presses. All rollers can be up to 1300 mm wide and accommodate a multitude of nozzles for strand forming (Fig. 6.4-27). Wide sheets can be multi-layered by extrusion through slots in multiple rotary bar roller presses or by laying several slabs on top of each other (Fig. 6.4-28, Fig. 6.4-31a). Fig. 6.4-29 shows single, double, and triple rotary bar roller presses with “standard” (cantilevered) drive.

Fig. 6.4-25 Two possible nozzle types for the manufacturing of: a) double layered; b) hollow strands in a double rotary bar roller press (courtesy Hosokawa Bepex, Leingarten, Germany)

Fig. 6.4-26 Diagrams of the designs of single, double, and triple rotary bar roller presses and collection of cross sections through ropes that may be manufactured (courtesy Hosokawa Bepex, Leingarten, Germany)


Fig. 6.4-27 Nozzles on a double rotary bar roller press for the forming of: a) filled ropes, b) flat-single, c) double-layered strands (courtesy Hosokawa Bepex, Leingarten, Germany)

Fig. 6.4-28 Two situations in which a second sheet is deposited on top of a previously made slab (courtesy Hosokawa Bepex, Leingarten, Germany)

233

234


Fig. 6.4-29 Photographs of three cantilevered rotary bar roller presses: single (DP 200-800), double (DDP 200-1000), and triple (DP/3 250-300) formers (courtesy Hosokawa Bepex, Leingarten, Germany)


The model numbers represent roller diameter and width in millimeters. All machines are executed in stainless steel and feed hopper, housing, side walls, rollers and die plates or nozzles are double walled and can be tempered (heated or cooled) independently of each other. Individual snack bars can be made either by cutting ropes (Fig. 6.4-30a) of by slitting slabs and cutting the strips (Fig. 6.4-30b). During slitting and cutting, the resulting strips and bars are automatically separated to avoid sticking together. Fig. 6.4-31a is a close-up photograph of a triple-layered food bar and Fig. 6.4-31b and c show other snack bars, including some that are coated with chocolate or very coarse food particles. Since the equipment following the extrusion is associated with belts (Fig. 6.4-30b), the roller presses are often executed in a bridge type fashion (Fig. 6.4-32), called mill shaft design in high-pressure roller presses [B.48, B.97]. An actual combination of formers, one bridge type and the other cantilevered, and a cutter are depicted in Fig. 6.4-33. Other equipment is available that can be added to complete lines, including: *

*

*

Smooth roll formers for the extrusion of pressure-sensitive or aerated masses that must be processed in viscous, pasty, kneadable, or sticky consistency; the forming system works without pressure according to the dragging principle (Fig. 6.4-34). Shape formers that are installed after, for example, rotary bar roller presses; they are two-roll formers (lower part of Fig. 6.4-35a) in which the roll surface carries molds that correspond to the product shape (Fig. 6.4-35b). This procedure is applied for cookies and articles that are difficult to produce with other systems. The shaped products are taken off with a special belt and transferred to a subsequent transport belt. Coolers (Fig. 6.4-30b), coaters (Fig. 6.4-30b), tunnel ovens for baking, enrobers and decorators (Fig. 6.4-36), and packing units.

Fig. 6.4-37 is a summary of some of the many different foods that can be shaped by or based on low-, medium-, and high-pressure agglomeration (extrusion). Fig. 6.4-30 Production of individual bars: a) by cutting ropes, b) by slitting slabs and cutting the strips (courtesy Hosokawa Bepex, Leingarten, Germany). In (b) are shown: (5) mixer/conditioner, (6) smooth roller former, (7) cooling drum, (8) rotary bar roller press, (9) cooling tunnel, (10) strand slitter, (11) fanning (separation) belt, (12) cutter, (13) tempering unit, (14) coater, (15) cooling tunnel

235

236


Fig. 6.4-31 Close-up of a triple layered food bar and other snack bars, some coated with very coarse food particles (courtesy Hosokawa Bepex, Leingarten, Germany)

Originally, the equipment described above was developed for sweets, such as marzipan (egg-, roll- or bar-shaped), but with the advent of convenience, designer, fun, and functional foods, which are often offered in the form of bars or other shapes, its application has widened and is now a major manufacturing tool of small and large, multinational manufacturers of foods. Tab. 6.4-6 is a listing of some of the materials that are processed with roller extrusion presses and associated equipment in the food industry. It shows that, today, not only confectionery products (sweets) but many others that were not extruded, are industrially handled and converted by these methods, mostly into convenience foods (ready to prepare or eat). Of course, many of the foods listed in Tab. 6.4-6 are not made-up or do not contain solid particles in large quantities and, therefore, even in the widest sense, can not be considered agglomerates. They are mentioned to show the versatility of the technology and tendencies in modern food processing and preparation. Other low-, medium-, and high-pressure extruders are also sometimes applied for the agglomeration of food products. Many of the materials mentioned in Tab. 6.4-6 are processed in such equipment; however the relationship between machine vendor and food processor is often a close one and governed by secrecy agreements.

6.4 Applications in the Food Industry Tab. 6.4-6 Listing of some materials that are processed with roller extrusion presses and associated equipment in the food industry (in alphabetical order) Confectionary Aerated nougat Aerated sugar masses Candy creme Chewing masses Chocolate masses Coconut masses Cream masses Fat fondant Fruit caramel Fudge Hard croquant Honey nougat Liquorice masses Marzipan

Others Noisette masses Nougat Nougat Monte´limar Nut pastes Peanut brittle Peppermint toffees Persipan* Praline´ mixture Soft caramel Soft croquant Sugar pastes Nougat Truffle masses

Biscuit masses Cereal mixtures Cheeses Chewing gum Corn starch masses Crisp rice (w.binder) Diet food masses Doughs (all kinds) Fats Fish masses Fruit pastes Honey cakes Jelly masses

Liver pastes Meat masses Muesli Oat mixtures Peanut butter Popcorn (w. binder) Potato masses Protein masses Puffed rice (w. binder) Ricemix masses Semolina masses Vegetable masses Yeast

..... masses contain viscous components to produce plasticity Binders may be fat, chocolate caramel toffee, sugar, starches, etc. * Persipan is a marzipan substitute made with apricot kernels

Fig. 6.4-32 Roller extrusion press executed in bridge type design (courtesy Hosokawa Bepex, Leingarten, Germany)

237

238

6 Industrial Applications of Size Enlargement by Agglomeration Fig. 6.4-33 Combination of roller press formers, one bridge type and the other cantilevered, and a cutter (courtesy Hosokawa Bepex, Leingarten, Germany)

Fig. 6.4-34 Sketch of the principle of smooth roll formers for the extrusion of pressure sensitive or aerated masses (courtesy Hosokawa Bepex, Leingarten, Germany)

Fig. 6.4-35 Shape forming: a) two-roll formers installed after, for example, a rotary bar roller press; b) the roll surface carries molds that correspond to the product shape (courtesy Hosokawa Bepex, Leingarten, Germany)

6.4 Applications in the Food Industry Fig. 6.4-36 Enrobed and decorated food products (courtesy Hosokawa Bepex, Leingarten, Germany)

Fig. 6.4-37 Summary photograph of the some of the many different foods that can be shaped by or based on low-, medium-, and high-pressure agglomeration (extrusion) (courtesy Hosokawa Bepex, Leingarten, Germany)

239

240


Several manufacturers offer pressure-cooker extruders [B.97] for the food industry. These machines apply medium pressure because the mostly grain and/or vegetable based starchy organic feeds are conditioned by pressure and heat into easily deformable and extrudable masses. Fig. 6.4-38 shows the functional components. At the inlet on top of the equipment, the slightly pre-mixed feed components enter first a mixing and predensification screw. Pressure builds-up by the changes in shaft diameter, sometimes variable pitch of the screw flights, and by a collar at the end of the screw shaft or a “pressure piece”, both forming an annular space with reduced area through which the material must pass. Such a screw conveyor, mixer, and processor is often called expander. The processed material drops into the cooker in which pressurized steam is injected and paddles or screws move the material to accomplish optimal contact. The plug at the end of the mixing and predensification screw acts as a dynamic seal so that the cooker is kept under pressure. Depending on the capacity and cooking time required for a specific application, the cooker vessel dimensions and/or number (multiple ones are mounted on top of each other) and the type of agitators may be changed. By varying the speed of the agitator(s) and the operating pressure in the cooker, almost infinite time/temperature combinations can be obtained. Pressurized steam cooking decreases work and power use, cuts production costs, and increases production capacity as much as 50 % compared with other extruders that accomplish heating by the conversion of mechanical into thermal energy through friction. The pressure (steam) cooker accomplishes much of the work that is conventionally done by the extruder; therefore, the life span of the extruder screw, inserts, barrel, die plate, and bearings is significantly increased. The cooker is easy to maintain and operate and, because the agitators are of simple design, these may be rebuilt numerous times to regain critical clearances. The processed mass is transferred into the screw extruder where hydrostatic pressure is developed, which causes axial extrusion through the openings of the die plate. The orifices produce extrudates, which often feature different cross sections, for processed cereals, for example, tubes. The ropes, featuring various cross sectional shapes, or, for example, tubes are cut with an adjustable and, for cleaning purposes, replace-

Fig. 6.4-38 Diagram of a typical pressure-cooker extruder with main dimensions in feet and inches: 1’ = 0.3048 m, 1” = 25.4 mm (courtesy Sprout-Matador, Muncy, PA, USA)


Fig. 6.4-39 Some examples of food products obtained with pressure-cooker extruders (courtesy Sprout-Matador, Muncy, PA, USA)

able device yielding short pieces or rings (Fig. 6.4-39). The lumpy shape of the processed (often called “expanded”) mass can often be directly used in a post-treatment facility (for example, “puffing” snack pieces during drying). Pellet mills (Fig. 5-10b2–b6, Chapter 5) are used in the food industry for the size enlargement of coffee and tea meal, herb powders, yeast, products with instant characteristics (for example mixtures of cocoa and powdered sugar, effervescent drink powders), and others. The cylindrical extrudates are bagged and used mostly by institutional customers (large kitchens). The most common application of punch-and-die presses in the food industry is for the production of bouillon cubes (Fig. 6.4-40) with dimensions typically in the range 13–15 mm. To obtain efficiency, rotary presses, so-called “cubers”, are used. The multi-component mass that results from mixing salt, sweeteners, binders (such as hydrolized corn gluten, modified corn starch, yeast), emulsified cooked meats, flavors, preservatives, various amounts of fat, and water (moisture content: 45–55 %) in a blender, is dried to < 2–3 % moisture with a belt dryer. The dry porous cake is broken and screened into three fractions. “Overs” are recirculated to the beginning of the process, the fraction between 2–5 mm is sold as granular product with “instant” properties (due to the particle’s high porosity, Fig. 6.4-40, bottom right), and fines are cubed. Since the small feed particles to the cuber contain substantially fewer pores and are deformed under the high pressing force to yield dense cubes, dissolution requires hot liquid, stirring, and a long time (several minutes). With modern cooking requirements such dissolution behavior is no longer acceptable. Therefore, bouillon cubes may now

Fig. 6.4-40 Cubed and granulated beef bouillon, both with “instant” characteristics (courtesy Borden Foods/Wyler’s, Columbus, OH/Chicago, IL, USA)

241

242


contain organic fibers, which act as conduits for liquid, swell, and assist in the product disintegration, or starches and/or their derivatives as disintegrants [B.97]. Other, mostly rectangular, compacted cooking ingredients, such as concentrated gravy, sauces, and soups are formulated and produced in similar ways. Fines separated from crushed salt, sugar, and other crystalline food products tend to stick together and cake. For that reason, for a long time until recently, they were redissolved and reprocessed in crystallizers. When high-pressure roller press manufacturers began to look for alternative applications for their equipment and started building clean, pharmaceutical and food grade machines (totally enclosed, stainless steel execution, easy cleaning), compaction/granulation (Fig. 6.1-14, Section 6.1) became a feasible alternative for size enlargement. Because the granules, the size of which can be adjusted as desired, are made from fine particles, they are not as hard as crystals of the same size and are, therefore, particularly suitable as sprinkling sugar for cookies or pretzel salt. Very large granules are made as fishery salt for use on fishing vessels. The same relative softness of agglomerated granules from fines as compared with crystals of the same material and size, even if made by high-pressure compaction/ granulation, make them ideal feed materials for secondary compression in punchand-die presses. The material is dust-free, easily flowing, and, therefore, can be metered well by and into all equipment, including high-speed rotary machines. During compression, the granules disintegrate and result in a uniform structure of the final compact. To avoid segregation, additives, such as flavors, colorants, or vitamins, can be incorporated in the first granulation step. As already indicated in Section 6.4.1 (Tab. 6.4-4, A4), contrary to common belief, granules obtained from press agglomeration may have instant characteristics. One of the most important binding mechanism of high-pressure agglomeration is caused by van-der-Waals forces (Chapter 3). This short-range molecular attraction does not develop solid bridges between the agglomerate forming particles and is lower in liquid environments by a factor of about 10, causing such granules to disperse easily and quickly. The addition of organic fibers, which, in foods, may at the same time constitute dietary ballast components, and/or of disintegrants [B.97] further improves the product’s instant properties. To further demonstrate the varied uses of size enlargement in the food industry, a relatively recent development shall be described that has led to the recovery of dust, which was previously considered a waste and discarded. After grinding roasted coffee beans, the product must be dedusted to meet customer expectations. Fines at the bottom of bags with ground coffee are not accepted. On the other hand, the large specific surface area of this dust results in almost instant brewing results, if it could be offered in an acceptable form. The solution of the problem is to produce a thin (< 1 mm) sheet by pressure agglomeration between smooth rollers in a modified flaking mill and crush it to yield flakes. With these, filter pockets are filled and bags are produced, called “coffee singles” by the manufacturer, which are similar to what was already known and available for (loose) teas for a long time. To engineer the product further, freeze concentrated coffee and flavors can be added and marketed as “gourmet coffee singles” as shown in Fig. 6.4-41.


Fig. 6.4-41 Examples of “gourmet coffee singles”, single portion filter bags filled with flaked and flavored roast coffee fines and freeze-concentrated coffee (Folgers Coffee/Procter & Gamble, Cincinnati, OH, USA)

243

244


6.4.3

Other Technologies

Agglomeration by heat or, more generally, thermal agglomeration, is sometimes used for special applications in food or food related industries. The oldest thermal agglomeration method was used in large communal “ice houses” that were built for the long-term storage of perishable foods such a meats and vegetables. Since large amounts of energy had to be used for freezing water, it was economical not to let small chips and fines, which were produced during a number of processing steps (crushing), melt. One of the reasons for crushing ice was to produce smaller pieces for the packing of foods in preparation for shipping in insulated containers. It was found that if supercooled (< 0 8C) ice pieces are compressed, the conversion of mechanical energy into heat causes roughness peaks and a thin surface layer to melt momentarily and to produce a liquid phase resulting in liquid bridges between the ice particles. However, because the amount of liquid produced is very small, typically less than 1 %, and the bulk of the mass remains at a temperature that is substantially below freezing, immediately after pressure release (when the energy supply ceases) the liquid solidifies, bonding the shaped body into an ice piece with high density. Since relatively small, briquetted, almond or pillow-shaped ice can be easily poured and metered, it became a superior cold-packing material. Roller presses were preferred for this operation because the thermal contraction of machine parts does not cause operational problems. Dry ice, compressed and shaped solid CO2, is an even better material for cold packing because it evaporates rather than melts and its cooling capacity is almost three-times higher than that of water ice. Dry ice compacts are typically made with hydraulic presses and are still used today. Today, many vegetable and fruit coulis are available in frozen blocks. After the food blocks (dimensions: 500 mm 500 mm 100 mm, temperature: –20 8C) have been milled to < 10 mm, compacts are made with a roller press (Fig. 6.4-43), as depicted in the inset sketch. Briquettes, still at –20 8C, are good looking and stable. They can be identified with patterns, machined into the form pockets on the rollers by electrochemical milling [B.48], and offer a presentation apart from the usual packaging and a product that is easy to meter, quick to defrost, and useable at home, by caterers, in restaurants, and in field kitchens (Fig. 6.4-42). The roller presses are very simple (Fig. 6.4-43) and have been supplied for capacities of 2.5–15 t/h. Because the feed is very clean and does, by definition (food), not contain any foreign material, contrary to most other designs, these roller presses do not have a floating roller so that the area that is in contact with the food can be totally sealed against oil or grease and synchronization is accomplished with timing gears. During production, the machine head is enclosed with food grade sheeting and the gears are completely covered. Feeding is by gravity. Agglomeration of foods by heat is mostly limited to the production of bonding between often large food particles (such as nuts). Sugar mixed with such materials is partially melted and caramelized, producing the desired bonds and forming clusters. At the same time the process enhances the taste. This is a very old method


Fig. 6.4-42 Reproduction of a leaflet showing frozen food pulp briquettes and the rendition of a reconstituted meal (courtesy Sahut-Conreur, Raismes, France)

of making candy, originating in the Arab world, widely practiced by street vendors and at fairs. Nuts and sugar are stirred by hand in a metal bowl over charcoal fire or burners until the caramelization has proceeded to the point where clusters are formed. After cooling, so that sticking between the product pieces is minimized, the candy is packed and sold.

245

246


Fig. 6.4-43 Photograph of a special roller press for the briquetting of food coulis (capacity 2.5 tonne/h) and sketch of the briquetting process (inset) (courtesy Sahut-Conreur, Raismes, France)

Coating of food products is performed most frequently with enrobers whereby food bars of the fun food category receive a layer of, for example, a chocolate mass (Section 6.4.2, Fig. 6.4-30b, items 13 and 14). This process is, however, not an agglomeration technique. Liquid coating mass is applied to the surfaces of the food article and solidified in a cooling tunnel. True coating by agglomeration, for example with powdered cocoa or sugar, may be carried-out in a coating drum or a Wurster type coater [B.97]. This technique is used for the finishing of small compacts (tablets), called cores, which are tumbled or fluidized while the powdered solids are added, binder liquid is atomized to cause adhesion to the surfaces of the cores, and warm air flows through the mass to evaporate the liquid and strengthen the powder layer. Rubbing against and collisions with each other densify and smooth the coating (Section 6.2.3).

6.5 Applications for Animal Feeds

6.5

Applications for Animal Feeds

When humans began to keep certain animals as pets and later raised a growing number of different animal species for work or transportation and as a source of food, clothing, and many other uses, the necessity of feeding independently of natural availability and the seasons arose. In addition to the production of hay, many animals shared human food scraps with their keepers. However, certain species, for examples the trained elephants in Southeast Asia, where so much removed from their natural food sources and foraging habits that the mahouts produced large balls of plant material and minerals, sort of giant hand-made agglomerates, which are fed directly into the elephant’s mouth, a method, which can still be observed today. Until the middle of the 19th century natural products, mostly plants and plant seeds, prevailed in animal feeding. The raising and keeping of animals was decentralized and concentrated near large human settlements. Only a few products, such as cheese, dried or cured meats, and processed eggs, were suitable for longer time storage and large distance transportation. The same reasons (i.e., improved long distance transportation and refrigeration) as described for food production (Section 6.4) triggered a change in livestock farming and feeding. The animals were now raised in large numbers at geographically favorable locations and their products transported to, stored near, and distributed to a large number of consumers. In addition, at approximately the same time, more knowledge about the digestion of feedstuffs by animals and the need for minerals and other ingredients for animal health and development became known; later the addition of vitamins, and, more recently, the desire to include medications and the controversial inclusion of hormones, became common practice. All that led to the formulation of today’s concentrated or compound feeds. Also at approximately the same time, large food processing and packing facilities were developed, which produced huge amounts of byproducts from plants, vegetables, and animals. In addition to milled feed grains, the major source of starch, these materials became excellent sources of nutritional, highly digestible components for livestock feeds. However, because of their physical consistency these very desirable feed components are not accepted by many animals in their finely particulate, suspended, or dissolved state. Suspended and dissolved byproducts had to be dried and, together with other fine solid components, size enlargement became necessary or desirable for many applications. Concentrated feed is formulated with the specific needs of each animal group and the requirements of the particular industry in mind, whereby growth (fat and/or muscle development), endurance, subsistence, and special performances (e.g., high rate of egg production) are encouraged and supported. Since, in comparison to the basic nutritional feed components carbohydrates, proteins, fats, amides, fibers, and water, the relative gravimetric and volumetric amounts of additives, such as minerals and vitamins, are very small, uniform distribution in the bulk feed became a challenge (Section 6.2.1) and segregation a problem. It was also found that certain animals did

247

248


not take-up and eat feed additives if they where supplied as separate entities, even if they were processed for easy consumption. Smell and/or taste causes the animal to avoid such compounds. All these reasons led to the evolution of a large, highly efficient industry in which socalled feed mills formulate and process special nutriments for different groups of animals. While, at the beginning, this development was centered around the traditional domestic livestock, mammals and fowl, today’s products include feeds for new food production technologies, for example fish farming, where fish or shrimp are raised and fed for optimum results. Another important new market is pet food. Tab. 6.5-1 is a summary of the most important reasons for size enlargement in the animal feed industry.

6.5.1


The first applications of agglomeration in animal feeding were in veterinary medicine for the treatment of sick animals. Remedies were mixed with a feed base from which pills and later tablets were made (Section 6.2) that were dispensed to the keeper (later farmers) and fed to the animals. This technology is almost as old as the manufacturing of agglomerated medicines for humans. For mammals, milk is nature’s food for the offspring and, therefore, represents one of the best and richest nutriments for this animal group, including humans. Reduced fat milk from which only cream has been skimmed and, to a lesser degree, whey from Tab. 6.5-1 Summary of the most important reasons for size enlargement by agglomeration in the animal feed industry *

* * * * * * * * * * *

* *

* * *

Conversion of mixtures of feed compounds and additives with different particle sizes and shapes, specific weights, hygroscopicities, and physical states into a stable mixed formulation. No losses and no dust annoyance (compliance with emission control laws). Improved flow, storage, transportation, metering, and feeding behavior. Avoidance of the segregation, separation, or selection (by the animals) of feed components. Better acceptance of the feed by animals. Possibility to prepare species specific forms of feed (e.g. for chickens, fish, etc.). Prevention of lumping, build-up, and bag-set. Reduction of losses by oxidation and other reactions during storage. Improved pickup of liquids added during feeding. In some cases, improved dispersion and solution in liquids. Possibility to include a wide variety of waste materials, including liquids, with nutritional value. Particularly after conditioning in the so-called expander and pelleting, improved availability of nutrients for digestion. Increased starch-value of the feedstuff. Lower moisture requirement for (pressure) agglomeration (pelleting) after preconditioning; therefore, reduced energy requirement for drying. Specifically for pet food, pleasing appearance of the product for the buying consumer. Increased economy of the manufacturing process. Higher profit margin.


which fat and coagulated milk solids have been removed (mostly for the manufacturing of cheese) are still highly nutritional feeds. Nevertheless, they are processed for a number of reasons (Section 6.4.1). Animal feeds contain mostly carbohydrates, proteins, fat, starch, and organic fibers, all of which can spoil with time. Particularly all types of milk and whey degenerate due to the activity of bacteria destroying the nutritive value unless it is kept at low temperatures or preservatives are added. Removal of water from the product will also reduce the bacterial activity and, thus, ensure an almost infinite shelf life if the product is dry enough and kept in a sheltered, cool place. A common animal feed that is processed by growth agglomeration is milk replacer. This product is a replacement for whole milk in which the butter fat, removed for mostly human consumption, is replaced by a cheaper animal or vegetable fat. It is typically used to feed calves, chickens, and pigs. Calf milk replacer is primarily used for breeding, as, at birth, calves have underdeveloped stomachs, lack the ability to digest fibrous feed, and must receive a liquid diet. Weaning newborn calves off cows milk frees it up for other diary based operations, such as processing milk for human consumption or conversion to more valuable products, for example butter and cheeses. Furthermore, all replacers are for fattening of animals, including calves. It is possible to make modifications of components in the mix and supply unique nutrients that are normally not present in whole milk. Therefore, however, the finished product has no fixed composition, as shown in Tab. 6.5-2 [6.4.1.1]. Milk replacers may be in liquid or dry powder form. The latter offers various benefits when compared with liquids; they cost less to transport, which helps improve the overall efficiency of diary production, and, as mentioned above, microbial growth during storage is much reduced. The farmers desire powdered milk replacer that are stable until reconstitution (rehydration, dispersion and dissolution in water), undergo little or no fat separation prior to, during, or after hydration, are easily and quickly processed in cold or hot water, feature an optimum fat particle size distribution for easy digestion by the animal, and exhibit no or very little lumping during storage prior to its application [6.5.1.1]. To fulfill the above requirements, manufacturers of milk replacers seek to obtain the following: high wettability, high dispersability, low sedimentation tendencies (little residue or sediment after reconstitution), and stability of the fat-in-water emulsion. Referring back to Section 6.4.1, Tab. 6.4-3, the first three are characteristics of instant products. In addition, the stable fat-in-water emulsion is maintained with the help of emulsifiers. Tab. 6.5-2

Typical ingredients and range of compositions of milk replacers [6.4.1.1] %

Skim milk solids Fat Dextrose, lactose, possibly whey solids Emulsifying agents (lecithin, mono-glycerides, sucro-glycerides) Flour Minerals, vitamins, antibiotics (Ca, Na, Mg, Cu, vitamins A, D, E)

65–80 15–20 0–10 0–2 0–7 0–0.5

249

250


The two major components (Tab. 6.5-2) must be brought together and agglomerated. Originally, the loading of skimmed milk powder with fat was accomplished by spraying fat with a nozzle onto powder passing on a belt conveyor. However, this product had fat exposed on the surface, did not have good shelf life, and could not be reconstituted easily. Later fat was sprayed into an airstream that carried skim milk powder. In this process, milk particles adhere to and cover the fat droplets. Although this milk replacer behaves better during storage and reconstitution, the fat particles are much too big, causing poor storage properties and indigestion for the animals. Other processes used contacting a pre-made fat compound and skim milk powder in a tower with steam [B.97], secondary agglomeration of the warm and moist mixture with water in a tumble agglomeration unit, cooling, and sizing. This method suffers from inconsistency and inefficiency and does not accommodate flexible incorporation of additives. Until recently, the best product was produced by spray drying an emulsion of skim milk concentrate and fat, followed by agglomerating and/or cooling in a vibrating fluidized bed. In many ways, a spray/dryer/agglomerator/cooler system for, for example, the manufacturing of calf milk replacer (Fig. 6.5-1) resembles a plant for dry, instant milk (Section 6.4.1). Mixing (homogenizing) of liquid fat (about 65 8C) and concentrated (40–45 % solids, 65 8C) skimmed milk, producing the emulsion, is either accomplished discontinuously in mixing tanks or continuously in a “combinator” [6.4.1.1]. In both cases, the purpose of homogenizing is to reduce the size of the fat droplets (above) with an aim of < 3 lm and a minimum of 90 % less than 1 lm.

Fig. 6.5-1 Spray dryer/agglomerator/cooler system for the manufacturing of calf milk replacer (courtesy Niro AS, Soeborg, Denmark)


As documented in the patent literature [6.5.1.1], there is a constant effort among manufacturers of milk replacer to produce a material that better meets the aforementioned requirements. In a new process [6.5.1.1], a powdered nutritional composition, with components corresponding to the range suggested in Tab. 6.5-2, is introduced into a mixer. This machine is a modified Schugi Flexomix (Section 6.4.1, Fig. 6.4-17, and [B.48, B.97]). One or more agglomerating and emulsifying agents are also added to the mixer. The feed powder blend is agglomerated whereby its particle size is increased and forms a moist intermediate product, which is dried and cooled. Care is taken to maintain the shape and size distribution of the agglomerates during post-treatment. The milk replacer is then classified to obtain the final product for packaging. The mixer has been modified to deliberately produce agglomerates with a relatively large particle size of 200 lm and larger. Contrary to competitive milk replacers, which may initially feature even larger but friable agglomerates, the particle size in the new product is substantially or fully retained until the time of reconstitution and helps (due to large size and no fines) to enhance the hydration characteristics. This is caused by a rather continuous coating that is formed on and between the particulate components and on the exterior of the agglomerates. Fig. 6.5-2 shows SEM micrographs at two

Fig. 6.5-2 SEM photographs of milk replacer produced by: 1) conventional method, 2) new process [6.5.1.1] at two different magnifications (a and b) (courtesy Land O’Lakes, Inc., Arden Hills, MN, USA)

251

252


different magnifications (a and b), of milk replacer powder produced by a conventional method and the new process [6.5.1.1]. The non-uniform and discontinuous nature of the bonding and coating of an agglomerate made by the old method is visible at high magnification. At lower magnification, showing a larger area, the loose attachment of particles and the many voids, gaps, and crevices can be distinguished even better. Components are rubbed-off easily during packing and handling, causing disintegration of the agglomerates and the formation of dust. In contrast, product from the new process is completely enrobed by a continuous coating whereby the coating material is selected to allow quick and complete dispersion of wetted particles in lukewarm (about 45 8C) tap water. A number of animal feed compositions, often containing fat, proteins, vitamins, antibiotics, and special nutrients, normally not found in natural products, are granulated for grain-eating animals, particularly chicken and other farm or pet birds. Since, during agglomeration, the moist mass often becomes sticky, special blenders with various self-cleaning mixing tools [B.97] are applied. Birds of different size pick-up only narrowly sized grains that are typically related to the bird’s proportions: small ones require fine and large ones coarse agglomerates. Accordingly, even though the feed composition may be identical, different fractions may be produced and sold.

6.5.2


At the beginning of the 20th century, the fast growing world population and the concentration of humans in industrialized centers of developed and developing countries required a new, highly efficient production and distribution of food. Animals in the food chain for milk, meat, and, later, also eggs, were no longer raised and processed on relatively small farms but in ever larger breeding and rearing facilities. Depending on the type, hundreds (cattle and swine) to several tens of thousands (chicken) of animals are kept in one location, needing to be efficiently fed. Also, in countries with extensive grass lands (e.g., USA and Argentina), cattle are raised in large herds and transferred to “feed lots” for fattening just prior to slaughter. In such facilities, specially formulated, highly concentrated feeds are provided to the animals for fast growth within only a few weeks. More recently, to counteract the overfishing of streams, lakes, and oceans and to satisfy growing demand, an increasing number of fish and shellfish species are raised in fish farms and fed with dry feed. As mentioned in Section 6.5, all these modern food production plants are no longer relying on naturally available animal feed sources. Research into the needs of the various animal groups for healthy, fast, and controlled (e.g., low fat or large eggs) development has led to the definition of preferred feed components and the development of complex rations, which do not only provide food but also guarantee animal health and the desired growth rate. The multi-component concentrated feed compositions consist of a few main ingredients and an increasing number of minor elements, which often also include antibiotics. As in pharmaceutical (Section 6.2) and similar products, it is important that


each morsel of food contains the same amount of all ingredients. To accomplish this, the basic components are ground to yield feed meal into which the lesser constituents are blended. Since only a few animals take up powdery feed and those who do tend to leave bad tasting ingredients behind, the blend must be stabilized by agglomeration to make it acceptable and the taste masked by suitable means. While wet granulation (Section 6.5.1) is possible in some cases, pressure agglomeration by pelleting (Chapter 5, Fig. 5.10 b1–b6) soon became the production method of choice [B.48, B.97]. The reasons for this preference are the easy preconditioning of the components, whereby the starchy ingredients are activated and the plasticity of organic materials is enhanced, the possibility to process materials that, because of their origin, still feature a certain elasticity, the reliable shaping of the blend and the production of feed pellets with adjustable physical, particularly strength and structure related properties. A recent study [6.5.2.1] estimates that the annual “world compound feed production is presently at about 600 million tons and is expected to reach 630 million tons by 2006”. Much of this is agglomerated, typically pelleted, which means that almost every feed mill on earth is equipped with one or more pelleting systems. It also means that by volume the agglomeration (pelleting) of animal feed is the largest application of the unit operation “size enlargement by agglomeration” followed by the pelletization of iron ores (Section 6.8.1) which, with an available capacity of just over 300 million tons in 1999/2001 [6.8.1.8], is a distant second. Pelleting of animal feed started at the beginning of the 20th century with the use of screw extruders, which had just been invented for the shaping of clays (Section 6.7.2). Fig. 6.5-3 is the reproduction of a brochure from the 1920s describing screw extruders (Chapter 5, Fig. 5.10b1) and showing a flowchart for the pelleting of animal feed meal blends. Shortly thereafter, responding to the growing demand, flat die (Chapter 5,

Fig. 6.5-3 Reproduction of a brochure from the 1920s describing screw extruders and a flowchart for the pelleting of animal feed meal blends (courtesy Amandus Kahl, Reinbek, Germany)

253

254


Fig. 5.10b2) and cylindrical die (Chapter 5, Fig. 5.10b5) pellet mills and their designs were proposed by many inventors in short succession [B.48]. Although other designs were and, to a limited extend, still are offered for specific applications, the flat die design with muller-like press rollers (Fig. 6.5-4a) and ring dies with interior press rollers (Fig. 6.5-4b) are the dominant machines for the processing of animal feeds. Typically, both machines are equipped with integral feeder/conditioners (lower part of Fig. 6.5-4b). The largest parts of a feed mill (Fig. 6.5-5) are raw material receiving, storage, preparation, proportioning, and blending (left side of Figure 6.5-5) and the packing and shipping departments (right side of Figure 6.5-5). In the first the components are received, stored, and processed as required. Grains are milled to produce a meal and grasses in the form of hay or alfalfa and other dry plant materials are chopped to become suitable for blending with the major and/or minor ingredients. All components are then conventionally formulated by metering and mixing to form the feed to the pellet mill(s). The flowchart in Fig. 6.5-5 is a versatile plant for the production of 40 t/h of pellets. It is designed for the storage of a total of 400 ts of up to 24 dry and several liquid components. Compounding and mixing is carried out in 10 batches per hour of 4000 kg each. Two independently operating conditioning and pelleting lines, each equipped with one flat die pellet mill (Fig. 6.5-4a), allow the simultaneous manufacturing of two different animal feed formulations. Correspondingly, the 20 product silos are arranged in two groups, each with two loadout points for bulk carriers. Alternatively, bagging systems can be added if desired.

Fig. 6.5-4 Diagrams of the principles and designs of the two dominant types of pellet mills for the processing of animal feeds: a) flat die pellet press (courtesy Amandus Kahl, Reinbek, Germany); b) ring die pellet press, also showing integral feeder/conditioners in the lower part (courtesy CPM, Waterloo, IA, USA)


Fig. 6.5-5 Flow sheet of a typical feed mill (Courtesy Amandus Kahl, Reinbek, Germany): 1) receiving, 2) silos, 3) proportioning and weighing, 4) premixing, 5) grinding and mixing, 6) conditioning and pelleting, 7) liquids storage and metering, 8) coating (Rotospray), 9) product storage and loading, 10) miscellaneous support functions, 11) electrical and control equipment

Fig. 6.5-6 is the flow diagram of a pelleting system using a ring die pellet mill (Fig. 6.5-4b and 6.5-7) and a vertical cooler (Fig. 6.5-8) [B.3, 1971]. The pellet mill (Fig. 6.5-7) is fitted with a variable speed screw conveyor, which is necessary to provide an even feed to the conditioner. This machine is a continuous, flow-through mixer that is equipped with fixed or adjustable pins or paddles (lower part of Fig. 6.5-4b). The purpose of the conditioner, which is fitted with steam and liquid injection manifolds, is to prepare the materials properly for optimum pelleting results. Conditioning is almost always accomplished by the addition of controlled amounts of constant-quality process steam, which supplies moisture for lubrication and softening of plant material, liberates natural oils, and causes the partial gelatinization of starches. Viscous liquids, such as molasses, may also be added as nutrients and binders just prior to extrusion. In some cases the lubricating, plasticity, and binding characteristics of the dry blend are adequate for successful pelleting. In such few cases, the conditioner (Figure 6.5-7) may be omitted. The conditioned, warm, moist, plastic, and sticky feed is transferred to the extrusion section of the pellet mill. For good operation and product quality it is imperative to distribute the feed uniformly over (flat) or in (ring) the die; this is one of the major difficulties, which is addressed and solved by various distributor designs [B.48, B.97]. Since densification and shaping of the feed is due to the frictional resistance during

255

256


Fig. 6.5-6 Flow diagram of a pelleting system using a ring die pellet mill and a vertical cooler [B.3, 1971]

6.5 Applications for Animal Feeds Fig. 6.5-7 Components of a ring die pellet mill [B.3, 1971]: 1) variable speed screw conveyor, 2) conditioner, 3) ring die and roller (extrusion) section, 4) speed reducer, 5) main motor, 6) machine base

extrusion and a certain amount of moisture is required for lubrication, the pellets as produced are warm (70–95 8C) and feature a moisture content of up to 18 %. Both are too high. Therefore, in a pellet cooler (horizontal [B.48, B.97]) or vertical (Fig. 6.5-6 and 6.5-8) design) ambient air is pushed through a bed of pellets affecting at the same time cooling and drying. In some cases, particularly with the more complex and sensitive mixtures of modern animal feeds (below) a hot air dryer is installed first.

Fig. 6.5-8 Schematic of a vertical pellet cooler [B.3, 1971]: 1) feed hopper with level sensing device, 2) cooling columns, 3) plenum or air chamber, 4) discharge gate drive motor, 5) discharge star gates, 6, 7) air fan with drive motor

257

258


Fig. 6.5-9 Pellet mill floor of a conventional animal feed mill employing ring die machines with conditioners (courtesy CPM, Waterloo, IA, USA)

Many feed pellets are sold in bulk or packed and offered to the animals as cylindrical extrudates. However, it is also possible to install so called crumblers, often mills with fluted of corrugated rollers, to produce granules, mostly for fowls. Depending on the specification, such feed may be screened to remove fines. The latter are recirculated to the mixing area of the feed mill for reuse. Although the need for animal feeds is constantly growing, the conventional pelleting as described above and shown in Fig. 6.5-5, 6.5-9, and 6.5-10 is overbuilt and new capacities are not as frequently built as in the past. The feeds processed in these mills are the basic staples, which include [B.3, 1971] the following. *

*

*

*

High-grain complete feeds (50–80 % grain, 12–25 % protein). These are high-starch formulas requiring conditioning to reach high moisture content and temperature. Heat-sensitive feed with sugar, dried milk, or dried whey (5–25 %). Since sugar and milk products begin to caramelize around 60 8C, the heat of friction must be kept low by employing thin dies, low speed, and adding fat and/or water as lubricant/ coolant. High natural protein (25–45 %) supplements and concentrates. Commonly these feeds also contain 5–30 % molasses. Similar to high-grain complete feeds they require high heat but less moisture addition. Low-protein (12–16 %) complete feeds. They contain little grain and high amounts of bulky fibers. Since these feeds can accept only a small amount of moisture, addition of steam must be limited resulting in low moisture and temperature of the conditioned feed. Consequently, the manufacturing of high quality pellets is difficult.

6.5 Applications for Animal Feeds Fig. 6.5-10a Flat die pellet mill in one of the production lines shown in Fig. 6.5-5 (courtesy Amandus Kahl, Reinbek, Germany)

Fig. 6.5-10b Multiple flat die pellet mills in a conventional animal feed mill (courtesy Amandus Kahl, Reinbek, Germany)

259

260


High-urea (6–30 %), high-molasses (5–20 %) feeds. This type, particularly with higher concentrations, is very difficult to extrude, requiring little or no steam, thin dies, low speed, dusting of the pellets to remove stickiness, and drying prior to cooling.

The above indicates that, while a feed mill may be equipped with one or more pellet mills that are selected to handle the production capacity with the most common feed type, conditioning, die design, extrusion speed, and, potentially, post-treatment must be variable. This is accomplished by controls, variable speed drives, the availability of various dies, and an adaptable flowchart with diverters (circles in Fig. 6.5-6) for different process path selections. Installation of redundant equipment helps to chose the most effective system and maintaining production during machine modifications and maintenance. As already mentioned above, research into the needs of various animal groups for healthy, fast, and controlled development has led to the definition of preferred feed components and the development of complex rations, which do not only provide food but also guarantee animal health and growth. This is particularly true for feeds for specific animal farming technologies, such as the already traditional feed lot formulations for cattle or the more recently formulated nutrients for fish and shrimp farming, and for pet foods. Regarding pet food developments, pelleted dry dog food is an example [6.5.2.2]. Fig. 6.5-11 shows that the research into the nutritional needs of dogs results not

Fig. 6.5-11

Analysis of nutrients in extruded dry dog food [6.5.2.2]

6.5 Applications for Animal Feeds Tab. 6.5-3 Typical generic recipe of a high quality dry dog food suitable for pelleting [6.5.2.2] 48 % 12 % 5% 8% 8% 6% 4% 2% 2% 3% 2%

wheat or flour fish meal meat meal soy flour corn (maize) wheat bran (beer) yeast dry extracted sugar beet slices potato flour lime and minerals minor ingredients (e.g. drugs, colors, flavors, etc.)

only in a best composition but also reveals that the percentages of the main ingredients and the amount of energy that is made available should be modified according to dog age and size. Dogs are also carnivores and in contrast to, for example, swine have a digestive system that is about 2/3 shorter. Although, theoretically, dogs may be nourished with food components that originate from materials other than meat and bones, the formulation must be energy rich and easily digestible. Tab. 6.5-3 is a typical generic recipe of a high-quality dry dog food suitable for pelleting. To make this mixture easily digestible, all particles, particularly the plant-based ones must be ground to < 0.6 mm. The large surface area of finer particles facilitates the pre-gelatinization of the starchy components (Fig. 6.5-12) which, in turn, increases digestibility. Pre-gelatinization itself is accomplished in the conditioner (called expander [B.97]) by heating (with steam) and moistening. Originally, the warm and moist dog food formulation was pelleted in one of the traditional pellet mills as described above for the conventional types of animal feeds. More recently it was found that the expander itself can be modified to produce pellets by installing a hydraulically adjustable die plate with knife head (detail in Fig. 6.5-13), thereby effectively becoming an extruder [B.97]. As in the case of dog food, research into the nutritional needs of other pets and of farm-raised animals, also taking into account animal sizes and ages, led to the requirement to produce a large number of different formulations for optimal care (pets) and

Fig. 6.5-12 Starch gelatinization of pet food as a function of particle size [6.5.2.2] characterized by the diameters of the openings in the discharge screen of a hammermill (Section 6.1). “Special” refers to a screen producing particles < 0.6 mm

261

262


Fig. 6.5-13 Flow diagram of a modern animal feed extrusion system (courtesy Amandus Kahl, Reinbek, Germany)

growth (farming) in relatively small quantities. To accomplish this in an efficient, profitable way, the bulk processing conventional feed mill is no longer suitable. New systems must be small, versatile, and quickly adaptable to new formulations. Fig. 6.5-13 shows the flowchart of such a manufacturing plant. Production begins with the exact proportioning of solids that have been ground to the optimal particle size (above) and are made available from silos with feeding screws, and of liquids, steam, and water into a blender. After mixing, the material is transferred to a holding tank (conditioner) with bottom agitator/discharge feeder. There, during a residence time of about 10 min, the liquids penetrate into the solids. A screw conveyor feeds the tubular expander–extruder [B.97] that is equipped with double walls for heating or cooling and a hydraulically adjustable die. With the latter, blockages are avoided or can be removed quickly. Some designs are fitted with multiple dies for quick change [B.97]. However, in any case, even without this provision, die change of expander–extruders is fast and can be carried out without tools. Following the extrusion, several alternative flow paths are possible. In most cases, because the pellets from expander–extruders are relatively wet and have low green strength, the discharge is carefully positioned onto a belt dryer, which dries-off excess moisture at minimum mechanical stress while the final binding mechanism is developed (e.g., by recrystallization of dissolved substances, increased viscosity of molasses,


or chemical reaction between, for example, molasses and lime). Screening of the dryer discharge to remove fines is optional. If used, the fines together with dust from dust collection are recirculated to the blender. Coating of screened pellets is used when specific feed characteristics are desired (below). If the product is still too warm, cooling (with a belt dryer) may be employed. Feed for chicken, other fowl, and fish or shrimp must be granulated (crushed) by a crumbler followed by fractionating into different sizes (alternative on the lower right of Fig. 6.5-13). According to the manufacturer (Amandus Kahl, Reinbek, Germany, see Section 15.1), the expander–extruder operation allows starch modification (pre-gelatinization) of 80– 90 % for good digestibility, fat contents of 20–30 % for high energy feeds, and considerable water binding (200–300 %) prior to extrusion resulting in high porosity after drying. The latter is of importance if, to increase the water uptake of animals, the pellets must be soaked prior to feeding but retain sufficient stability. Depending on die design, the cross section of the pellets may be round, oval, cloverleaf, bone shaped and so on (Fig. 6.5-14). Final feeds may be blends of different products (Fig. 6.5-14, last). New applications include the manufacturing of feed for use in water farms. For young, small fish, granules with sizes in the range 0.1–2 mm are required (system alternative at the lower right of Fig. 6.5-13 and Fig. 6.5-15a). Coating allows the production of floating or slowly sinking pellets for tilapia, carp and catfish, of slowly sinking pellets with high fat content (up to 30 %) for trout, salmon, and perch, and of sinking, water stable pellets for shrimp and other crustaceans. Such pellets have diameters in the range 2–12 mm (Fig. 6.5-15b–e).

Fig. 6.5-14 Different extruded dry dog and cat foods (courtesy Amandus Kahl, Reinbek, Germany)

Fig. 6.5-15 Extruded dry fish and shrimp feed: a) crumbled feed for young fish, b) pellets 2–12 mm, c) floating pellets, d) slowly sinking pellets, e) water

stable (2–8 h) pellets (courtesy Amandus Kahl, Reinbek, Germany)

263

264


Dry plant material that is used as animal feed, including cereal, hay, and alfalfa, is very voluminous, even after baling. Therefore, it is not easily transported and stored. Baled material needs to be torn apart prior to feeding and milled before it can be incorporated into the formulation of complete animal feed rations, which are then pelleted as described above. Also bales take-up water if they are not protected from rain. The early development of stationary and mobile baling presses (Fig. 6.5-16) [6.5.2.3] and the successful application of high pressure for the production of relatively small, highly densified compacts (Fig. 6.5-17 and 6.5-18), which can be collected and handled in bulk, have suggested the use of punch-and-die or roller presses for the production of cylindrical and almond or pillow shaped briquettes, mainly consisting of shredded hay and alfalfa enriched with small amounts of minerals and other animal feed components. While after conditioning (heating and moistening with steam) briquetting is feasible and economical, field trials showed that such material can not be used as Fig. 6.5-16 Sketches demonstrating the development of baling presses [6.5.2.3]: a) old American box frame press, b) first stationary high-pressure press (USA, about 1870), c) first German mobile straw press (1896), d) swing-piston press (Raussendorf, Singwitz/Saxonia, 1938)


Fig. 6.5-17 Mobile hay collection equipment with high pressure ram press [6.5.2.3]. The cylindrical compacts are transferred into an open car by pushing them through the transport pipe on top of the machine

Fig. 6.5-18 Comparison of volume of compacts (left) produced with the equipment shown in Fig. 6.5-17 and a conventional bale (right) from the same material [6.5.2.3]

feed. The highly densified pieces swell to a multiple of their volume in the stomachs of, for example, cows and endanger their wellbeing. Therefore, this development was given up. High-pressure briquetting, mostly with ram or punch-and-die presses, is used, however, for the manufacturing of so called range cubes or licking stones. These large briquettes are used on farms for free-range cattle and provide salt, minerals, and minor ingredients (drugs, hormones) to the animals that lick the pieces. It is important that such briquettes have high density and often contain a binder that makes them rain resistant.

265

266


6.5.3

Other Technologies

Applications of coating and nanotechnologies are new developments in the design, formulation, and production of animal feed. Coating, as discussed in Section 6.5.2 and shown as a possible process step in Fig. 6.5-13, is being used to obtain special effects, such as resistance to moisture pickup during storage, and for the production of water stable pellets, for example for shrimp feed. Coating, often as micro-encapsulation (Chapters 5 and 11), may be also used as a preparatory step for feed components to influence the availability of nutrients or drugs in the animal’s digestive system. Such processing takes place prior to creating the formulation in the feed mill and is often carried out by the supplier of the ingredient off-site. Nanotechnologies (Chapter 11) are of increasing interest for the manufacturing of functional feed components, mostly in veterinary medicine (Section 6.2.3). Functionalized ingredients that influence animal growth and health are being developed for incorporation in complete feed formulations.

6.6

Fertilizers and Agrochemicals For centuries, through ancient and medieval times, man has been interested in improving crop yields. Various mineral and organic substances have been used to improve productivity by exploiting effects discovered by accident or empirically by trial and error. Such materials include: manure, ground bones, wood and other ashes, saltpeter, and gypsum. However, the results were not predictable and a treatment that benefited one field could have no or even an adverse effect on another. The foundation of modern fertilizer technology was laid by Justus von Liebig in 1840. He postulated that the mineral elements nitrogen, phosphorus, and potassium (N, P, and K) in the soil are responsible for plant nutrition and stressed the necessity of replacing those elements to maintain soil fertility. The availability of nutrient elements for plant life depends to a large extent on solubility. In most cases a high solubility is desired, requiring a large specific surface, which is synonymous with small particle size. Additional micronutrients are necessary, which must be added as fine powders because of the small amounts of these trace elements in a fertilizer formulation. Such powder systems exhibit a number of problems, as shown in Tab. 6.6-1. To overcome these difficulties, it is not surprising that size enlargement by agglomeration of powdered plant nutrients has been investigated almost since the beginning of their use.

6.6 Fertilizers and Agrochemicals Tab. 6.6-1 Potential problems associated with the processing and handling of powdered fertilizer formulations * * * * *

Uniform mixing is difficult and time consuming. Dusting is excessive during handling. Segregation of components occurs due to differences in particle size and/or density. Danger of caking exists during storage and transportation. Difficulties prevail during application (dusting, which may result in health hazards, run-off with water, scattering by wind, etc.).

After definition of the main plant nutrients, naturally occurring materials containing those elements were systematically used for fertilization. Ground bones constituted the first phosphate fertilizer, and wood ashes, sugar beet wastes, and saltpeter were early sources of potassium. For many years, the need to supply nitrogen was considered to be of secondary importance, as natural supplies in rain water and from the air with a system of crop rotation were deemed adequate [6.6.1]. By about 1840, treatment of phosphate rock with sulfuric acid was found to yield an effective phosphate fertilizer, called superphosphate. The first successful commercial production started in England in 1842, and by 1870 there were 80 factories operating in the UK [6.6.1]. Mining of potassium chloride salt deposits began in Germany in 1860 and dominated the world market for 75 years. The first products were low-grade unrefined ores such as manure salts (20–25 % K2O) and Kainite (19 % K2O). The development of refining methods gradually increased the grade and, today, potassium chloride (60– 62 % K2O) is the main product. Toward the end of the 20th century it became evident that the food needs of a growing world population could be met only by an increased supply of fixed nitrogen in fertilizers and three processes were developed. In 1903 the arc process was commercially introduced in Norway which, after several process steps, produces calcium nitrate. At about the same time the calcium cyanamide process was perfected and in 1913 direct synthesis of ammonia from nitrogen and hydrogen was first carried out successfully on a commercial scale in Germany [6.6.1]. More recently, urea production has grown rapidly and urea is now the leading form of nitrogen in fertilizers. First reports of fertilizer granulation appeared at the beginning of the 20th century [B.48]. Original research by the US Department of Agriculture and the Tennessee Valley Authority (TVA) started in 1922 and aroused worldwide interest. In 1927 a branch of the German IG Farbenindustrie introduced for the first time a “grained” fertilizer material (Nitrophoska) [6.6.2]. Until 1933 the process involved granulating a slurry of di-ammonium phosphate and ammonium nitrate in pug mills with the addition of potash salts. After 1933 the formulation was changed several times, but the production of Nitrophoska continued (later at BASF). One of the first commercial fertilizer granulation processes in North America was carried-out by COMINCO (Consolidated Mining and Smelting Co.) in Trail, BC, Canada [6.6.3]. This also employed a double-shaft pug mill, called a “blunger”. Sulfuric acid, a by-product from the smelting operation, was used to make ammonium sulfate and phosphoric acid with phosphate rock; ammonia was also produced from electrolytic hydrogen.

267

268


Three granular fertilizers were made: triple superphosphate, monoammonium phosphate (11-47-0), and ammonium phosphate-sulfate (16-20-0). This technology, with various modifications, was later used in numerous fertilizer plants around the world and was one of the more important granulation processes for many years. The first application of a drum for fertilizer granulation was in the early 1930s in Baltimore, MD, USA, in the Oberphos process [6.6.4], a batch method for producing superphosphate from phosphate rock and sulfuric acid. In the closed drum a chemical reaction and granulation took place simultaneously. In the 1930s the first granular fertilizers were introduced in the UK by ICI (Imperial Chemical Industries). Six grades of granular “concentrated complete fertilizers” were produced, based on ammonium phosphate with additions of ammonium sulfate and potash salts [6.6.5]. ICI probably used a slurry process, similar to the one introduced by COMINCO, except that much of the feed was solid ammonium sulfate and potash with only ammonium phosphate in slurry form. Relatively large agglomerates were formed and the final granules were produced by crushing and screening the dried product. Prilling is the spraying and solidification of molten urea, which produces spherical granules but is not really an agglomeration technology [B.48, B.97]. There are also chemical reaction methods in which granulation is a side effect (e.g., the TVA “continuous ammoniator”, in which both a chemical finishing process and granulation occur [6.6.6]). Apart from these, further large scale commercial adoption of fertilizer granulation by any agglomeration method did not take place until the 1950s. The production of mixed (NPK) granulated fertilizers from dry components began then, using mixers, drums, pans, and suspended-solids agglomerators (spouted and fluidized beds). By 1934 a patent had been issued to the machine manufacturer Eirich in Germany (DRP 647 651) for a granulating mixer intended for fertilizers. The patent describes the mixing and rolling of particulate solids on the flat bottom of a pan mixer by means of eccentrically arranged mixing tools. The mixing elements are designed as either blades or bars, which rotate and extend into the material to be agglomerated. The patent stated that, due to the intense movement of the material, practically all solid powders can be agglomerated into relatively small, uniform granules with or without the addition of binder(s). Utilization of the inclined-pan granulator in the fertilizer industry was first reported in Germany in 1953 for the granulation of ordinary superphosphate and the first documented application of suspended-solids agglomerators was in the early 1960s using a spouted bed [B.48]. Although the tabletting (punch-and-die) press was invented in 1843, its use for small-volume fertilizers did not occur until much later. Today this technology is still not very important in the fertilizer industry; it is mostly used for special garden and plant-nursery applications. In 1950, another pressure-agglomeration technology emerged as a method for the granulation of potassium chloride (mineral potash) after concentration by flotation or selective crystallization. This method uses roller presses for the production of a highly densified sheet, which is then crushed and screened (compaction/granulation) to yield granular and coarse potash fertilizer grades. The technique quickly found general acceptance around the world [Section 13.3, ref. 108]. Because roller presses can be

6.6 Fertilizers and Agrochemicals

easily adapted to a wide range of capacities and feed materials, ten years later compaction/granulation was also introduced and is now applied for the production of granulated mixed NPK fertilizers. Much more recently, during the past 20 years or so, agglomeration methods began to make inroads into other agrochemical technologies. Coating is now used to modify the availability of the nutrient by controlling the release time or to improve seeds. Agglomerates are also employed as carriers and diluents for toxic chemicals, such as herbicides, insecticides, and pesticides. These agglomerates must feature specific properties, particularly that they are easily degradable. Easily degradable carrier materials must offer a large accessible (inner) surface area (high porosity), so that they can be easily loaded (impregnated) with liquid components, and have sufficient strength to withstand processing, storage, and handling. The binding mechanism must survive the impregnation process with the active substance. For example, if molecular or electric forces were used for dry agglomeration, it is possible that, after impregnation, the active substance replaces (e.g., by recrystallization during a drying step) or enhances (e.g., by viscous liquid bonding and/or chemical reaction) the original binding mechanism. On the other hand, the granules must break down easily and quickly under the influence of moisture either from the soil or the atmosphere (rain or dew). Therefore, the product must wet easily and even small amounts of moisture should reduce the strength. Surfactants improve wetting and components that swell in the presence of moisture may assist in breakdown. Sometimes, particularly in the case of fertilizers and micronutrients, interactions with bacteria also participate in degradation. With the large-scale industrial production of straight fertilizers each containing only one primary nutrient, and the merger of land into large operations using mechanical equipment for farming including fertilization, over-fertilization and pollution by runoff became a major concern. Also, farmers striving for higher yields and greater productivity requested more and more complex fertilizers, which now also contain secondary elements and micronutrients (Tab. 6.6-2). Formulations are determined by Tab. 6.6-2 Primary-, secondary-, and micro-nutrients for plant fertilization, each in alphabetical order [6.6.2]

Primary nutrients

Secondary elements

Micronutrients

Element

Symbol

Nitrogen Phosphorus Potassium Calcium Magnesium Sulfur Boron Chlorine Copper Iron Manganese Molybdenum Zinc

N P K Ca Mg S B Cl Cu Fe Mn Mo Zn

269

270


agronomists on the basis of soil analyses and knowledge of the requirements of specific plants. Since secondary and micronutrients are required in only tiny amounts, they must be added as fine powders and uniform distribution becomes a challenge. At the same time, increased caking problems result from the production of fertilizer components with higher analyses while the common use of mechanical application equipment calls for dependably free-flowing, dust-free materials. This is why the granulation of fertilizers, either as components for bulk blending or as mixed fertilizers for direct use became popular during the second part of the 20th century. Today, even small farms, particularly in subtropical and tropical climate zones, obtain specially formulated multi-component fertilizers, which are compounded and granulated in special fertilizer agglomeration plants using compaction/granulation for size enlargement (Section 6.6.2).

Further Reading

For further reading the following books are recommended: B.3, B.7, B.16, B.21, B.22, B.26, B.40, B.48, B.56, B.58, B.64, B.67, B.81, B.82, B.89, B.93, B.94, B.97, B.98 (Chapter 13.1).

6.6.1


The first fertilizers that were industrially produced were the result of reacting a solid (such as phosphate rock), with a liquid (such as sulfuric acid), in a pug mill. The reaction yielded the fertilizer material, in this case superphosphate. Other early fertilizers used similar production methods. Granules formed naturally during mixing and reaction. Sometimes the agglomerates were too large, so they were subsequently crushed and screened into the desired size distribution. An improvement was the accretion process in which a recirculating load of granules received a thin coating of slurry in the mixer prior to drying. The dry product was screened to recover product-sized particles and the rest was recirculated. Beginning in the 1950s, for reasons already discussed, more complex mixed fertilizers were developed and produced. This trend required either compatible granules (in size and/or mass) of single fertilizers and additives for bulk blending or the production of multi-component granules from dry, mixed fertilizer formulations. To accomplish this, size enlargement by agglomeration of the particulate solids was introduced. Since some earlier processes had yielded agglomerates from solids after wetting them with liquids, the first techniques that quickly became the standard for fertilizer granulation followed these precedents. Such plants (Fig. 6.6-1) include an appropriate number of silos for the raw components (1) from which the correct amounts are metered (2). Recycling fines are added at a fixed rate. These feed materials can be alternatively mixed (3) for added uniformity or fed directly into the tumble agglomerator


Fig. 6.6-1

Flow diagram of a typical wet granulation plant for dry fertilizers

(e.g., rotating drum, (4), [B.48, B.97]) where a liquid binder (normally water or a fertilizer solution) is added and the growth of agglomerates takes place in the moving particle bed. Alternative tumble agglomerators can be inclined pans or many types of mixers (including pug mills, the Eirich mixer, ribbon blenders, and others [B.97]). The discharge from the tumble/growth agglomerator consists of a wide distribution of green (moist) agglomerates, which must be dried (5) and cooled (7) to achieve permanent final product strength. Since the formation of oversized lumps in the agglomerator and the dryer can not be avoided, the cooled material is screened (9) to yield the finished product. Oversized and undersized particles pass through a mill (10) and are returned, together with the dust from the cyclones (11) that clean the effluent air, via a surge bin and a metering device to the process. It should be mentioned that the recirculating, pre-agglomerated particles that are removed on the product screens and adjusted in size by the mill play an important role in tumble/growth agglomeration. Since nucleation, the formation of seed agglomerates from powder, is the most difficult and time-consuming step in growth agglomeration [B.97], recycling introduces nuclei, which help agglomerate formation and growth. Fig. 6.6-2 shows the discharge end of a granulation drum and the chute feeding the green agglomerates into the rotary dryer. The dryer is a co-current design with the burner (6 in Fig. 6.6-1) visible under the support structure. Fig. 6.6-3 shows the discharge of well-sized green agglomerates from two pan agglomerators. Fig. 6.6-4 is the overall view of a system specified in Fig. 6.6-1. One of the major problems associated with the wet granulation of fertilizers is the very reason why the wetted particulate mass agglomerates and forms granules. The moistened solid particles not only adhere to each other but also to equipment walls and parts. This build-up can not be avoided and must be controlled. In mixers employing agitators, such as pug mills, ribbon blenders, or Eirich-type pan mixers, the tools them-

271

272

6 Industrial Applications of Size Enlargement by Agglomeration Fig. 6.6-2 The discharge end of the granulation drum and the feed end of the rotary dryer of a wet fertilizer granulation plant (courtesy Krupp Polysius, Beckum, Germany)

Fig. 6.6-3 Discharge of granules from two pan agglomerators in a wet fertilizer granulation plant (courtesy Krupp Polysius, Beckum, Germany)


Fig. 6.6-4 Overall view of a system for the wet granulation of dry fertilizers according to Fig. 6.6-1 (courtesy Krupp Polysius, Beckum, Germany)

selves limit the thickness of the build-up to the distance between the equipment walls and the agitators. In addition, scraper blades may be installed to clean those areas that are not reached by the mixing tools. The drum and pan agglomerators, which are the most commonly used for wet granulation of particulate fertilizer components, do not employ mixing tools. Therefore, special scrapers must be installed to limit build-up. In inclined pans, the bottom scrapers (Fig. 6.6-5, left) are also used to enhance the separation of granule sizes in the downward moving part of the pan (Fig. 6.6-5, right) and, for that reason, their position is adjustable. Fig. 6.6-6 shows a pan of 4.6 m diameter with unadjusted scrapers. In agglomeration drums, stationary (Fig. 6.6-7a and b) or movable scrapers (Fig. 6.6-7c and d) are used. Fig. 6.6-7 depicts only four examples of many different designs.

Fig. 6.6-5 left) Diagram of the wall and bottom scrapers in a pan granulator; right) sketch of the particle motion that is assisted by the position of the bottom scrapers [B.97]

273

274

6 Industrial Applications of Size Enlargement by Agglomeration Fig. 6.6-6 Photograph of a 4.6 m diameter pan showing the still-unadjusted vane-type plow scrapers (E) that are individually mounted on a support structure (D), which moves with the tilt of the pan (courtesy Feeco, Green Bay, WI, USA)

Fig. 6.6-8 is a sketch of the operating principle of a drum granulator depicting the spray bar for liquid addition, the movement of the particulate charge, and the scraper. In spite of these control measures, build-up takes place. If the coating has a controlled thickness and grooves from scraper tips, it can act as wear and/or corrosion protection. It may also improve the tumbling action by modifying the friction between the coating and the charge. However, with time the coatings become very hard and increase the drive

Fig. 6.6-7 Four typical designs of internal scrapers of drum granulators [B.48]: a) single-stage, sectional adjustable straight edge; b) two-stage scraper consisting of tungsten carbide cutters and a secondary adjustable straight edge; c) hydraulically powered reciprocating; d) rotary spiral

6.6 Fertilizers and Agrochemicals Fig. 6.6-8 Sketch of the operating principle of a drum granulator depicting the spray bar for liquid addition, the movement of the particulate charge, and the scraper (B.48)

power required which, in turn, is converted into frictional heat within the apparatus. Therefore, it is necessary from time to time to stop the equipment and remove the buildup, a time consuming, dirty, manual operation. Often such cleaning must also take place before production can be switched to another formulation. Because many of the components are relatively fine high-analysis fertilizers and moisture is part of the process, many other parts of a fertilizer granulation system are also prone to build-up which, owing to chemical reactions or after natural evaporation of contained liquids, also harden. This affects dust-collection ducts, collectors, and chutes and transport devices, dryers, and screens. Generally speaking, wet agglomeration plants in the fertilizer industry must be cleaned often and good housekeeping is an important part of reliable operation. Fig. 6.6-9 is another flow diagram of a plant for the granulation of dry fertilizer materials [6.6.6]. Compared with the system presented in Fig. 6.6-1, there are the fol-

Fig. 6.6-9 Typical flow diagram of a plant for steam granulation of dry fertilizer materials [6.6.6]

275

276


lowing differences. Particularly if mixed fertilizers are to be formulated and granulated, it is important that segregation of the components is avoided as much as possible prior to and during agglomeration. Since the raw materials tend to cake or lump and, as supplied, may feature different particle size distributions they are screened prior to the intermediate storage and oversized material is crushed in a disintegrator (mill). Another, even better way to homogenize the feed to an agglomerator (granulator) is to mill the mixed formulation as shown in Fig. 6.6-25. Since the materials all have similar size distribution and are metered continuously and simultaneously onto a collection belt, mixing in the chutes, the bucket elevator, the feed hopper, and the drum granulator is considered sufficient to obtain satisfactory homogeneity. Also, recycling of fines from the product screen, the oversize crusher, and the dust collectors is added prior to the bucket elevator without going through a surge bin and metering. This is a questionable practice. It is not unusual that surging occurs in the granulator and dryer, which results in large fluctuations of the amount of recycling. For uniform plant operation it would be better to level this out by means of a recycle surge hopper. In the drum granulator, water is added as a binder and steam is also discharged under the bed of material at the feed end of the drum. Any of the tumble/growth agglomerators mentioned earlier can be installed at this position. If steam is injected, the process is often called “steam granulation”. Size enlargement is controlled by the amounts of steam and water added. For each mixture a “percentage liquid phase” exists at which granulation efficiency (agglomerate growth rate) is optimal. The liquid phase consists of the amounts of moisture and dissolved salt. Since the solubility of fertilizer salts increases with temperature (Fig. 4-5, Chapter 4), the higher the temperature the less spray water is required. For any given fertilizer formulation an optimum moisture content exists at each temperature; this may be described by an experimentally determined curve such as shown in Fig. 6.6-10. Since the condensation

Fig. 6.6-10 Influence of the relationship between temperature and water content on the agglomeration behavior of fertilizers during steam granulation [6.6.6]


of steam in a particle bed is one of the most efficient means of transferring heat to particulate solids, the overall moisture content is reduced and dryer operation becomes cheaper. At the same time a stronger and often denser agglomerate is formed in the dryer when the larger amount of dissolved salt recrystallizes to form the permanent bonding. In the granulator, conditions of mixtures that are below the range shown in Fig. 6.6-10 result in insufficient agglomeration and those above the range produce excessive growth, build-up, and even the formation of a slurry, requiring shut down of the system and extensive cleaning of the granulator and equipment downstream. As depicted in Fig. 6.6-9, oversized material is removed after the dryer on a doubledeck screen, crushed, and recirculated as fines. As considerable amounts of oversized lumps can form in the granulator and the dryer, returning the discharge of the crusher that would now operate at lower levels of stressing (Section 6.1, Fig. 6.1.9), to the screen may increase the product yield. The screen is installed after the dryer to unburden the cooler because the recirculating load in wet granulation systems is often high (up to 500 %). The flow diagram also indicates the use of a coating drum after the cooler. Although granulated fertilizers are much less prone to setting during storage (Chapter 4), the inherent hygroscopic nature of many plant nutrient materials may cause problems. Coating with anti-caking materials (Chapter 4) may prevent difficulties, and coatings with functional layers can add new characteristics (i.e., slow release, herbicide, fungicide, insecticide: Section 6.6.3). Wet granulation is mostly used for size enlargement of multi-nutrient fertilizers formulated from dry components, but processes to form a granular fertilizer product by chemical reaction accompanied by agglomeration are still carried-out. In the USA, for example, the introduction by TVA of a continuous ammoniator-granulator (US patent 2 729 554) had a significant effect [6.6.6] on the development of fertilizer granulation. The method was originally developed for a more efficient ammoniation of superphosphate; however, it was found that agglomeration occurred during this process and could be controlled by the addition of water or steam or by adjusting the formulation to provide the required temperature of 80–100 8C. When the heat of reaction of the ammoniation of superphosphate was insufficient, sulfuric or phosphoric acid was added along with more ammonia through sparger tubes embedded in the tumbling mass (Fig. 6.6-11). Granulation of two NPK grades (6-12-12 and 10-2020) was demonstrated in a pilot plant in 1953 and by 1962, 164 installations used the process in the USA, about two-thirds of all US granulation plants [6.6.6]. Later the process was adapted to receive a pre-reacted slurry (Fig. 6.6-11b) for use with formulations in which the heat of reaction is too great for release in the ammoniator– granulator. Fig. 6.6-12 is a typical flow diagram of a TVA-type ammoniation-granulation plant for the production of granulated NPK fertilizers [6.6.6]. The liquid phase that initiates granulation in a tumble/growth agglomerator can also be a melt. For example, using the TVA pipe-reactor [6.6.6] phosphoric acid reacts with ammonia to produce a melt of ammonium polyphosphate (APP), which is sprayed onto the moving bed of solids and causes agglomeration. Fig. 6.6-13 is the flow diagram of a plant producing granulated NP fertilizers by spraying a highly concentrated

277

278


Fig. 6.6-11 a) Sketch of the TVA pilot ammoniator-granulator; b) cut-away view of a large scale ammoniator-granulator designed to accept pre-reacted slurry [6.6.6]

urea solution, an APP melt, and scrubber solution onto recirculating material from the process in a pug mill, and Fig. 6.6-14 is a system for the production of granulated NPK fertilizers from solid ammonium nitrate or urea and potash with an APP melt in a special drum. Fig. 6.6-15 is a cut-away drawing of the granulating drum in Fig. 6.6-14. Other processes using melts are applied to round irregular fertilizer granules by coating or to add additional nutrients (Section 6.6.3).


Fig. 6.6-12 Flow diagram of a TVA type ammoniation-granulation plant for the production of granulated NPK fertilizers [6.6.6]

Fig. 6.6-13 Flow diagram of a TVA pipe reactor-pugmill process producing granular NP fertilizer (urea, ammonium polyphosphate) [6.6.6]

279

280


Fig. 6.6-14 Flow diagram of a granulation plant using the TVA pipe reactor for the production of NPK fertilizers [6.6.6]

Fig. 6.6-15

Cut-away drawing of the drum used in Fig. 6.6-14 [6.6.6]


6.6.2


Since 1950 granulation by compaction has been an accepted method for the manufacture of high-quality granular potash [Section 13.3, ref. 96]. Today, this technology is used by the majority of potash producers worldwide. About 10 years later, in the early 1960s, pressure agglomeration emerged as an alternative to the conventional tumble agglomeration methods in mixers, drums, pans, and suspended solids granulators for mixed NPK fertilizer granulation [Section 13.3, ref. 108]. Because, in contrast to conventional wet granulation methods, high-pressure compaction is processing dry feed materials from an almost unlimited number of sources, without special requirements on particle size or distribution, this technique is gaining increasing importance. Compaction/granulation with the most commonly used equipment, roller presses, which can be easily adapted to a wide range of capacities and feed materials, is even better suited for multi-component fertilizers then for the single potash product. Roller presses that are used for the compaction of fertilizers (Fig. 6.6-16) feature selfaligning roller bearings, optimally sized steel bearing blocks, and a hydraulic pressurizing system with proprietary functions and hydraulic accumulators. The latter allow adjustment of the pressure-response characteristic and provide for overload protection. While in some cases simple gravity feeders with flow control baffles are provided, most applications require one or more screw feeder(s) with variable (e.g., hydraulic) speed drives to force the material to be compacted into the nip between the rollers [B.48, B.97]. To maximize availability and minimize potential problems that could be caused by insufficient routine maintenance, the machines are equipped with water cooling and automatic grease lubrication. Double output-shaft gear reducers provide for completely enclosed, dust-tight drives connected with the rollers by gear-tooth couplings and spacers, which allow transmittal of full torque even at relatively high misalignment. In the case of machines with high torque requirement, the oil of the gear reducer is circulated, filtered, and cooled and the gear-tooth couplings may be optimally equipped with continuous greasing to guarantee long life and availability. Particularly if fine nutrient powders or carriers for agrochemicals are processed, deaeration, which is the removal of air from the densifying mass, requires special design and operational considerations. In machines with large production capacities this includes the “split roller” design, which is characterized by two separate compaction rolls on each shaft with a gap in the middle that provides additional venting of air at the center cheek plates. Many fertilizer materials are hard minerals that cause wear. Other nutrients or solid agrochemicals are salts or compounds which, especially in the presence of moisture, may cause corrosion. For these reasons it is inevitable that the “pressing tools”, rings or segments, which are fastened to the roller core by suitable means, must be exchanged at regular intervals for remachining or replacement [B.48]. The “hinged frame”, which is available from some manufacturers (Fig. 6.6-17) facilitates this work and minimizes downtime due to maintenance.

281

282


Fig. 6.6-16 Two views of a modern roller press for the compaction of fertilizers (courtesy K€ oppern, Hattingen/Ruhr, Germany)


Fig. 6.6-17 Schematic and photograph of a “hinged frame” for roller presses (courtesy K€ oppern, Hattingen/Ruhr, Germany)

Fig. 6.6-18 depicts the most versatile flow diagram of a fertilizer granulation plant using a roller press for compaction. Pre-mixed powder formulation (1) is fed into a day bin (2). Recycled fines (17) from the process (below) and dust from the pollution control system (20) are transported to bin (18). The latter should be sized so that in an emergency or unscheduled changeover the entire hold-up of the plant can be accepted. Prior to running a new formulation the contents of the recycling bin (18) must be dumped via diverter gate (19). During normal operation fresh pre-mixed (if applicable) feed (2) and recyclate (18) are proportioned by rotary gates and weigh belts (3). The ratio of fresh feed to recyclate should be kept constant; it is only adjusted if the level controls in bins (2) and (18) require modification of compactor feed composition. Typically a changed relationship “fresh feed to recyclate” necessitates readjustment of the compactor (8) and, sometimes, the oversize crusher (16) parameters. Fresh feed and recyclate are homogenized in a low-intensity mixer (e.g.,. pug mill, mix muller) (4) and transported by buck-

283

284


Fig. 6.6-18 Flow diagram of a fertilizer granulation plant using a roller press for compaction [B.48, B.97]

et elevator (5), metal detector (6), and drag chain conveyor (7) to the roller press (8). Independent of the type of feeder used to transport material into the roller nip, it is imperative to avoid “starved” feeding conditions at all times. Therefore, a small overflow, measured by a solids flow meter (11), is maintained. The signal from the flowmeter (11) may be used to adjust the system feed rate by controlling the rotary gates and weigh belts (3). The compacted sheets exiting the roller press (8) are pre-crushed in the flake breaker (9) and screened (10) to remove fines. Coarse material is transported to screen (13) by bucket elevator (12). On the screen, product is separated and transferred to storage silo (14) while oversized material is crushed in granulator (16) and again separated into three fractions on the double-deck screen. All undersized fines, including dust from the pollution control system, are recirculated to recycle bin (18). Potential modifications to optimize this basic flow diagram are as follows (compare Figs. 6.6-18 and 6.6-19).


Fig. 6.6-19 Flow diagram of an optimized fertilizer compaction/ granulation system incorporating changes [B.48, B.97]

a Elimination of flake breaker (9) Sometimes the sheet produced in compactor (8) breaks up easily and, therefore, flake breaker (9) is not required. b Elimination of screen (10) Since fines separated at this point amount to only 10 % (mostly “leakage” from the compactor cheek plates) screen (10) may be eliminated. However, crusher screen (13) and/or primary granulator (22), if applicable, may be less efficient. c Flake curing (21) Some materials yield a relatively soft sheet immediately after compaction (e.g., due to liquid phases resulting from energy input during compaction, for examples see below) but quickly “cure” (i.e., solidify) when cooling and reach higher strength. In such cases it is desirable to install a “time delay” of between a few seconds and several minutes (curing belt) between compactor and flake breaker.

285

286


d Addition of primary granulator (22) The yield of granular product can be optimized by lowering the reduction ratio of crushing (Section 6.1, Fig. 6.1.9). This can be accomplished by the installation of a primary granulator (22) or, more generally, multi-stage crushing. e Improved product characteristics Particularly if primary granulator (22) is installed, the loading of double deck screen (13) with finer material is so high that its separation efficiency deteriorates. On the other hand, product users often impose limitations on the amounts of oversize and fines; also, the presence of product-grade material in the oversize and/or undersize streams reduces yield; therefore, screen decks with larger and smaller openings may be selected for screen (13). Then, final separation is achieved on secondary screen (23). f Product particle rounding (24) Studies have shown [6.6.2.1] that the irregular (angular) shape of granular fertilizer obtained by compaction, crushing, and screening (Fig. 6.6-20) does not have a negative influence on the efficiency of and uniformity of distribution by modern mechanical rotary spreaders. Nevertheless it may be preferable to remove sharp edges and corners in an “abrasion drum” (24) to avoid excessive production of dust during handling and transportation. Fines produced during tumbling are separated from the product on screen (25) and recirculated. g Product conditioning (26) In some cases it is desirable to condition or treat the product with anticaking reagents, insecticides, or fungicides. Such treatment can be accomplished in a conditioning drum (26) (Section 6.6.3). Modern roller presses for the fertilizer and agrochemical industries feature special designs (Figs. 6.6-16 and 6.6-17). Operating parameters are determined during tests with a representative sample of the particular formulation. As for all applications, they include: specific pressing force, response characteristic of the floating roller (i.e., in-

Fig. 6.6-20 Compacted sheet and granular fertilizer obtained by crushing and screening


fluence of the hydraulic accumulator pressure), roller diameter, sheet thickness, and roller speed, which is limited by deaeration and potential elastic properties of feed components. As a result of testing and experience the surface configuration of the rollers (e.g., smooth, corrugated, waffled, welded), the type (gravity or force) and number of feeder(s), and the drives, size (kW) and method (single or variable speed, electric or hydraulic), are chosen. As always, the most important parameter for roller press selection is the specific pressing force necessary to obtain a highly densified sheet that (after crushing and screening) produces strong enough granules with acceptable yield in the required particle size range. The specific pressing force in kN/cm is defined as total force exerted by the pressurizing system of the machine divided by the active width of the rollers. It is different for each fertilizer or formulation and varies in the range about 30–120 kN/cm if materials are processed in presses featuring rollers with 1000 mm diameter and operating at 12–14 rpm producing sheets 12 mm thick (Tab. 6.6-3). Simple mathematical relationships make it possible to convert these figures to the conditions that are valid for different roller diameters and speeds or sheet thicknesses [B.97]. In mixed fertilizers the presence of substantial amounts of material requiring low specific force (e.g., urea) reduce the necessary force and act as a binder while the admixture of hard components (e.g., raw phosphate or Thomas slag) result in the need for higher forces. Selection of peripheral system components, such as mixers, crushers, screens, material handling, and storage, depends on the material(s) to be processed and, to a cer-

Tab. 6.6-3 Specific pressing force, water content, and feed particle size that were determined for the compaction of some fertilizer materials in roller presses [B.48] Fertilizer

Specific pressing forcea (kN/cm)b

Water content (%)

Feed particle size (mm)

Ammonium sulfate Potash 60% K2O Feed temperature > 1208C Feed temperature 208C Potash 40% K2O Feed temperature 908C Potassium sulfate Feed temperature > 708C Potassium nitrate Calcium nitrate Calcium cyanamide Urea Mixed fertilizer containing – No raw phosphate or Thomas slag – Raw phosphate or Thomas slag – Urea

100 – 120

0.5 – 1.0

< 1.0

45 – 50 70

dry dry

< 1.0, with max of 3% < 0.06

60

dry

< 1.0

1.0 0.5 – 1.0 dry dry dry

< 0.5 < 1.0 < 1.0 < 0.4 2 – 3 to < 1.0

< 1.0 < 1.0 < 1.0

< 1.0 < 1.0 < 1.0

70 100 60 60 30 – 40 30 – 80 > 80 30 – 40

a) Indicated pressing force is for machine having 1.0 m diameter rollers. b) 1 kN/cm = approximately 0.1 t/cm.

287

288


tain extent, on whether the plant is dedicated to the production of only one fertilizer or to a variety of formulations. Materials that reach final strength only after some curing time (such as formulations containing urea, partially acidulated phosphates, or superphosphates) and fertilizers obtaining strength from recrystallization, such as ammonium and potassium sulfates, must be handled, crushed, and screened gently. Other applications, such as the granulation of potash, require high energy input during crushing and screening so that a strong, abrasion-resistant product, suitable for bulk shipment, is obtained [Section 13.3, ref. 96]. At the beginning, compaction was applied to produce a strong, abrasion resistant granular fertilizer from KCl fines because dry material could be processed directly. Later, this process was used in the mixed fertilizer industry [Section 13.3, refs 108 and 110] primarily to save energy. Today the specific location and local costs of energy determine whether this feature is still a deciding factor. Other advantages often play a more important role, the most valuable being its versatility as demonstrated by the following list.

1. With the exception of a few materials, such as urea or triple superphosphate, for which maximum amounts exist that can be used in a formulation, literally all particulate solids can be processed by compaction. This includes also, for example, dry digested sludge from municipal waste treatment plants (Section 8.2). 2. To minimize cost, raw materials can be purchased on the world markets without specific requirements on particle size. Off-specification fines can be used and often are even preferred. 3. Compaction/granulation plants can be designed for economic operation at any feed rate. Production capacities of 0.1–50 t/h per line are feasible. 4. Larger plants are preferably equipped with two or more lines fed by one large compounding (batching or formulation) system. Otherwise the lines are kept separate to improve availability because only one line is down during maintenance and emergency shut-downs. 5. If a plant equipped with multiple lines features separate day bins for fresh feed, recirculating fines, and granulated product, each line can be operated on different formulations. 6. Production of small batches is feasible. Depending on the amount of cleaning necessary during change-over (determined by how much cross-contamination can be tolerated), several different formulations (batches) per 8 h shift can be produced. 7. Fertilizer granulation plants utilizing compaction can be combined with either custom designed batching systems or with standardized formulation or bulk-blending units. The latter allows easy expansion of bulk-blending to mixed fertilizer granulation. 8. Any fertilizer compaction/granulation system can be utilized as a regional production facility for the manufacturing of bulk-blend grade material from off-specification feeds. This capability also includes special formulations that are required by the local market such as indigenous fillers, with or without major nutrients, or carriers


for micronutrients. Such product can then be used together with imported bulkblend grade materials in bulk-blending. 9. It should also be mentioned that plants with roller presses can be easily adapted to the manufacturing of urea supergranules for deep placement in wetland rice production. In this case, the roller surface must be modified, the flake breaker (9 in Fig. 6.6-18) bypassed and the granulator (16 in Fig. 6.6-18) blocked-off. The freedom to select raw materials from a large number of different sources on the free market allows one to take advantage of special offers and, thus, optimize the facility’s cost structure. It is also possible to use otherwise marginally or not suitable raw materials offering special agronomic characteristics, including the incorporation of micro-nutrients. Furthermore, it is feasible to work closely with individual farmers and formulate fertilizers for their particular crops, soils, and climatic conditions. Actual operating data confirm that production runs of as short as 1 h duration can be economical and that during a typical day, three or more different formulations can be manufactured for specific customers. This exceptional versatility is most interesting for tropical and sub-tropical agricultural zones where many different crops are planted on relatively small plots. Fig. 6.6-21 shows block diagrams of different installations using compaction for the granulation of finely divided particulate solids (including materials such as fertilizers

Fig. 6.6-21 Block diagrams of different installations using compaction for the granulation of finely divided particulate solids (including materials such as fertilizers or agrochemicals) [Section 13.3, ref. 101]. F, fresh feed (possibly premixed); P, product(s); A, mixer; B, compactor; C, flake-breaker; D, screen(s); E, crusher(s); G, wet granulator; H, dryer; I, cooler

289

290


or agrochemicals). Fig. 6.6-21a depicts the simplest, basic flow diagram of a compaction/granulation plant. Feed (F), potentially premixed, coming from an outside source, is mixed in a blender (A) with recycle (R) from the product screen (D). The blend is compacted in the roller press (B) and sheets are produced, which are crushed in the flake breaker (C). The double deck screen (D) separates the discharge from the flake breaker into product (P1) and oversized and undersized material. Oversized particles are stressed again in a granulator (E) and returned to the screen (D) for sizing. Fines (recycle R) are returned to the blender and compaction. Fig. 6.6-22 is the photograph of such an installation, which is placed alongside the raw material storage facility (left). On the right are a day bin for premixed feed material and a recycle surge hopper. Below is the roller press, equipped with two feed screws (Fig. 6.6-23) and the structure behind the compactor contains the crushers, screens, and material handling equipment. Fig. 6.6-24 is the flow diagram of this particular system. When compared with the conventional wet granulation of fertilizers, the advantages of compaction/granulation became more widely known and accepted, dry systems according to Fig. 6.6-21a were installed at several sites that already operated a wet granulation plant. In fact, the system presented in Fig. 6.6-22 and 6.6-24 is in such a facility. The block diagram depicting both plants and how they are interconnected is shown in Fig. 6.6-21b. The rather extensive formulation system for the fresh feed and the mixer are common to both the wet and dry granulation plants. After this point, a diverter valve feeds either one of the two plants. A decision as to which of the two systems should be used depends on the feed characteristics and economical consid-

Fig. 6.6-22 Mixed fertilizer compaction/granulation plant (courtesy S€ udchemie AG, M€ unchen/Kehlheim, Germany)


Fig. 6.6-23 Roller press similar to the machine used shown in Fig. 6.6-22 (courtesy K€ oppern, Hattingen/Ruhr, Germany)

Fig. 6.6-24

Flow diagram of the plant shown in Fig. 6.6-22

291

292


erations. As indicated by the dotted connection, some or all of the fines (recycle) form the wet granulation system can be treated in the dry compaction/granulation plant, thus avoiding repeated re-wetting, re-drying, and re-cooling of a substantial amount of the fertilizer. A further energetically advantageous application of compaction/granulation, making use of the fact that systems with small capacities are economical, installs a compactor at the tail end of a conventional wet granulation plant (Fig. 6.6-21c). All of the off-grade (under- and oversized) dried and cooled material is compacted. The oversize crusher of the wet granulation plant may have to be upgraded in this case to also handle the compacted sheet but the screening capacity is normally sufficient. The product (P1 + P2) is a mixture of nearly spherical particles from wet granulation and irregularly formed granules from crushing a compacted sheet. Compacting off-grade dry material instead of recirculating it to wet granulation increases the capacity of the original system and, at the same time, considerably reduces energy consumption per ton of granular product. However, in a wet granulator, recyclate ( undersized but already pre-agglomerated material) often plays an important role for the kinetics and the ease of agglomeration [B.48, B.97]. Therefore, it is often preferred not to treat 100 % of the recyclate by compaction/granulation but to return 10– 20 % to the wet agglomeration system. Referring to (7) above, the flow diagram presented in Fig. 6.6-25 depicts what is also shown in Fig. 6.6-21d: the combination of a standard bulk blending system with a versatile compaction/granulation system (with two lines). On one hand, bulk blending (Fig. 6.6-25, items 1–7) can be used to accomplish the formulation and mixing of the fresh feed for the dry, multi-component fertilizer granulation plant (beginning with item 8, bucket elevator, in Fig. 6.6-25) employing one or more (two in the case on hand) roller presses for compaction. On the other hand, according to (8) above, the compaction/granulation system can be used as production facility for the manufacturing of bulk-blend grade material from off-specification feeds. Off-specification fertilizer components can be fines or, for example, micronutrients, which need to be attached to, often inert, carriers for uniform incorporation into a large mass of major nutrient materials. To avoid segregation of the bulk blended mixed fertilizer product, all components must meet special particle size and mass requirements. After adjusting these properties by compaction/granulation the products can be successfully used in the shown or any other bulk blending plant. Fig. 6.6-26 is a photograph of the actual plant according to the flow diagram of Fig. 6.6-25 showing bulk blending on the left and compaction/granulation on the right. Fig. 6.6-27 is a view of one of the compactors with compacted material on the discharge/cooling belt conveyor. The technologies, equipment, and plants described above are mostly used for single or multi-component fertilizers and for fillers (such as limestone) and mixtures of inert carriers loaded with small amounts of micronutrients. After screening-out a granular particle distribution (typical size about 1–6 mm), this product is either directly suitable for bulk shipment or bagging and use or, as discussed before, it is a component for bulk blending. As mentioned earlier, it is also possible to produce pillow- or almondshaped briquettes (about 20 cm3) as urea supergranules for deep placement in wetland


Fig. 6.6-25 Flow diagram of a plant combining fertilizer bulk blending (2–7) and compaction/ granulation (8–27). 1, front-end loader input of components; 2, 8 ,18, bucket elevators; 3, distributor; 4, silos for components; 5, compounding scale; 6, blender; 7, bulk blend receiving hopper (courtesy Sackett, Baltimore, USA). 9, mill for feed homogenization; 10, feed bin; 11, 27, screw conveyors; 12, belt mixer; 13, metal separators; 14, 25,

drag-chain conveyors; 15, roller presses compactors; 16, solids flow meters; 17, curing belt conveyors; 19, double-deck screens; 20, mills (granulators); 21, product belt conveyor; 22, belt scale; 23, 24, dust collection system equipment and building; 26, recycle surge bin (courtesy Sackett, Baltimore, USA, and K€ oppern, Hattingen/Ruhr, Germany)

rice production. Machines for this purpose are often very small, homemade roller presses, sometimes even manually driven with simple hand cranks, to make the product by the farmer himself on site. A further application of briquetting roller presses in the plant nutrient industry is the manufacturing of fertilizer spikes (Fig. 6.6-28) for the direct, long-term feeding of shrubs and trees. To produce these highly densified, strong, and hard shapes, elongated, pointed matching pockets are machined axially across the face of the rollers and the spikes are produced in normal briquetting fashion (Section 6.3.2) from the mixed fertilizer feed. During application, the pointed end of the spikes is punched (hammered) into the soil, for example near the trunk of a tree, where it slowly dissolves, thus releasing the nutrients in a controlled manner. Other high-pressure agglomeration techniques for the production of special fertilizer spikesertilizer products use punch-and-die presses with reciprocating punches (eccentric or rotary designs; Chapter 5). They are limited in capacity because of the acceleration and deceleration forces at the upper and lower dead centers. Even though rotary tablet presses with multiple dies and punches are capable of producing large numbers of typically flat, cylindrical (tabletted) compacts [B.97], the mass processed

293

294


Fig. 6.6-26 Plant described in Fig. 6.6-25. Bulk blending is on the left and compaction/granulation on the right (courtesy Ferquigua, Guatemala)

per unit time is low. Applications in the fertilizer industry are in the area of specially formulated products for house or nursery plants, where each tablet provides the ration for one plant, one container, or, after dissolution in water, for a certain amount of liquid plant food. Although, in the agrochemical industry, easily degradable products (Section 6.6.1) are more commonly made by growth and tumble agglomeration (Section 6.6.1), some are also made by pressure agglomeration. Typical examples are carrier materials for liquid fertilizers, insecticides, fungicides, and many other chemicals. In the liquid phase the active substance is highly concentrated and, in most cases, toxic. By adsorbing these liquids on the mostly inner surface of granules that are manufactured from


Fig. 6.6-27 One of the two roller presses in the plant according to Figures 6.6-25 and 6.6-26 also showing compacted material on the discharge/curing belt conveyor (courtesy Ferquigua, Guatemala)

fine particles by compaction/granulation, the toxin is diluted and the products are rendered safe for handling and application. For example, spreading of granular agrochemicals by conventional equipment is possible. Newer applications of the technology also include special micronutrients.

Fig. 6.6-28 Photograph of fertilizer spikes

295

296


In contrast to common belief, compaction/granulation is well suited for the manufacturing of easily degradable carrier particles since one of the main binding mechanism, molecular forces (van-der-Waals), diminishes by a factor of 10 in the presence of water. Therefore, if rain or dew wet the granules they lose strength, readily degrade, and liberate the active substance. A possible drawback is that high-pressure compaction results in a material with little residual porosity, which limits the amount of active components that can be incorporated. To overcome this problem, some carriers, with or without the active substance included, are extruded at medium pressure with pellet mills (Fig. 5-10b1–b6, Chapter 5). The product consists of cylindrical extrudates and can be used either directly or after impregnation with a chemical. Snail and slug control pellets are a typical application of this technique (Fig. 6.6-29).

Fig. 6.6-29

Pelleted agrochemical product (snail and slug control pellets)


6.6.3

Other Technologies

Coating (Chapter 5) is the most commonly used other technology in the fertilizer and agrochemicals fields. Agglomerated fertilizers (granulated or shaped by applying pressure) are coated in drums, to achieve rounding with an additional fertilizer component (melt coating), to supply a beneficial component (such as non-fertilizer agrochemicals), to reduce sticking tendencies (coating with anti-caking materials), or to provide functional properties (e.g., slow release). Non-fertilizer agrochemicals (e.g., herbicides, insecticides, pesticides) are typically impregnated on porous (often agglomerated) carriers and the product may then be also coated to obtain a certain functionality (e.g., release control). In the latter case the method of choice is now sometimes microencapsulation. In Section 6.6.1 the use of a melt as the liquid phase that initiates granulation in a tumble/growth agglomerator was described. Flow diagrams and a special drum design to accomplish this were shown (Fig. 6.6-13–15). A drum that accomplishes coating or rounding of irregularly shaped fertilizer granules, mostly from compaction/granulation (Section 6.6.2), by applying a melt is now introduced. Fig. 6.6-30 shows cross sections of broken melt-coated fertilizer granules [B.97]. In this case, the cores of conventionally granulated (by tumble/growth agglomeration, Section 6.6.1) TSP (triple super phosphate) were melt-coated with sulfur to provide an additional nutrient for sulfur-deficient soils and/or obtain a slow-release fertilizer. Fig. 6.6-31 is the flow diagram that is used for this process. The centerpiece of the system is the so-called fluid drum granulator (FDG). The fluid drum granulator (Fig. 6.6-32) is one vendor’s design to most efficiently achieve a uniform coating. Other coaters can be also used for the task. The fluid drum coater is a cylindrical horizontal drum, rotating around its axis, and fitted with special

Fig. 6.6-30 Cross sections through broken melt-coated fertilizer granules: cores, triple superphosphate (TSP) granules; coating, sulfur (courtesy Kaltenbach-Thuring, Beauvais, France)

297

298


Fig. 6.6-31 Flow diagram that is used for the manufacturing of products as shown in Fig. 6.6-1. The centerpiece of the system is the fluid drum

granulator (FDG) (courtesy Kaltenbach- Thuring, Beauvais, France)

anti-clogging lifters. A fluidized bed develops in a tray that is mounted inside the drum and supplied with atmospheric (cool) air. Seed solids fed to the drum can be either granules, produced by a separate agglomeration system, prills from a melt solidification process, or recyclate. The lifters transport the seed material to the upper part of the

Fig. 6.6-32 Sketch depicting the principle of the fluid granulation drum (FGD) for granulation or the coating of granular materials (seeds) (courtesy Kaltenbach-Thuring, Beauvais, France)


drum, from where it falls onto the surface of the fluidized bed and is cooled by the fluidizing air. Because the tray is sloped, the granules move across the bed and, at the edge, drop down into the lower part of the drum (falling curtain). During its fall, the material is sprayed with a melt or a solution, coating the particles, which are then lifted up and the cycle repeats itself. The drum can operate batch but more typically works continuously with finished product discharging over a weir. The process can also be used outright for the granulation of fertilizer components whereby a melt and/or solution, sprayed onto the falling curtain, provide the liquid binder component for growth agglomeration. Although, as shown in Section 6.6.2, the irregular shape of fertilizer granules obtained by, for example, compaction/granulation (Fig. 6.6-20) is not detrimental to the uniform distribution with mechanical spreaders in the field, there are various reasons for the desire to produce a rounded granule by coating, including a more pleasing appearance of those products destined for sales to small volume consumers (home and garden market) or the fattening of small (undersized) particles. Fig. 6.6-33a shows again some granules that were produced by compaction/granulation and Fig. 6.6-33b is the rounded product after coating in a fluid drum granulator. The FDG can be also used for the agglomeration of special agrochemical materials. Examples are the production of granulated degradable sulfur and of ammonium nitrate fertilizers. To obtain degradable sulfur granules a swelling agent is added, which assists in the break-up when moisture (from irrigation, rain, or dew) becomes present. Fig. 6.6-34 is the flow diagram of a manufacturing process that was designed for this purpose. Approximately 10 % of a bentonite clay (swelling agent) is added to the molten sulfur. The slurry together with make-up water is sprayed into the FDG (for startup, stored recycle from earlier production runs is introduced first) which, for this purpose is slightly modified as shown in Fig. 6.6-35. The discharge is screened and the undersized particles are returned to the granulator for fattening. In contrast to explosive low-density ammonium nitrate (LDAN, Section 6.11.3), fertilizer grade AN must have higher density (high-density ammonium nitrate = HDAN).

Fig. 6.6-33 a) Irregularly shaped fertilizer granules from compaction/ granulation (phosphate); b) rounded fertilizer granules obtained by coating by the FGD process (courtesy Kaltenbach-Thuring, Beauvais, France)

299

300


Fig. 6.6-34 Flow diagram of a system using the fluid drum granulator (FDG) for the production of granulated degradable sulfur for agricultural use (courtesy Kaltenbach-Thuring, Beauvais, France)

Fig. 6.6-35 Sketch of a modified fluid granulation drum (FGD) for the production of granulated degradable sulfur for agricultural application (courtesy Kaltenbach-Thuring, Beauvais, France)


Calcium AN (CAN) is an ammonium nitrate incorporating a filler (e.g., Dolomite) and, therefore, with reduced nitrogen content (26–34.5 %) and increased safety. The fluid drum granulator process is one of the alternatives for the manufacturing of HDAN and CAN. Fig. 6.6-36 is the flow diagram for the production of these granulated fertilizer grade ammonium nitrate materials from a 98.5 % AN solution; again, for startup, the FGD must be filled with recycle material from earlier production. The granulator discharge is screened on a double deck screen. The oversized particles are redissolved and undersize is returned for further size enlargement in the FGD. Sometimes it may be necessary to crush a part of the product to balance the closed loop recirculation system. Correctly sized particles are cooled in a fluidized bed cooler with dry air (from air conditioner). To avoid sticking of the hygroscopic product the granules are coated in a rotary drum with anti-caking agents (Chapter 4). Another interesting application of coating in this field is the enrobing of seeds with fertilizers and agrochemicals such as fungicides, herbicides, or insecticides. The technology is used particularly for light seeds, such as grass, and is often applied for reseeding golf courses. The coating with fertilizer makes the individual seeds larger and heavier thus allowing easier spreading and the addition of agrochemicals keeps birds, fungi, and insects in check. For other seeds control of weeds and grasses with suitable agrochemicals is a more important task. Microencapsulation (Chapter 5, [B.33, B.67, B.97]) is another one of the methods to obtain products with controlled release properties. It allows to isolate an active component from external media by forming a polymeric network or differently structured

Fig. 6.6-36 Flow diagram of a plant for the production of fertilizer grade granulated ammonium nitrate (courtesy Kaltenbach-Thuring, Beauvais, France)

301

302


wall in which a solid or a liquid substance are enclosed. Objectives for modern agrochemical products are to limit pollution by the uncontrolled release of active substance into the environment, to reduce the quantity of active substance required for a given application, and to improve the material’s efficiency and duration of action in the field. As an example of the use of microencapsulation in agrochemistry, a relatively recent development will be presented [6.6.3.1], although it deals with the “packaging” of a liquid and, therefore, is only connected with agglomeration through encapsulation. It is reviewed to show the principle. The goal of that work is to enclose an insecticide and to mix the microcapsules with disinfectant liquid and yield a two-effect product with long term release properties. In a first step, a liquid-liquid dispersion is formed in a static mixer. The insecticide and a monomer are in the immiscible phase and the continuous phase is an aqueous solution of polyvinyl alcohol. A second monomer is added to the liquid-liquid dispersion to achieve interfacial polymerization, a process in which monomers polymerize at the interface of two immiscible substances. In this case, the resulting microcapsules consist of a liquid core and a permeable wall. The encapsulated insecticide is stored in a concentrated disinfectant. While non-encapsulated insecticide and disinfectants are chemically incompatible and the active substance (insecticide) is destroyed after a relatively short time, the microencapsulated material remains essentially unchanged during long periods of storage. For final agricultural application, the concentrated product must be diluted with water to make a 2 % solution, which is applied with compressed air by hand held or power spray equipment. In the field, as shown in Fig. 6.6-37, the disinfectant solution surrounding the microcapsules evaporates, providing the disinfectant action. Then, the insecticide diffuses through the capsule wall and becomes available on the outside surfaces in small, highly effective amounts of the active substance. As the insecticide dissipates it is replenished by more chemical from the inside. Thus, the product is still working after most other insecticides have lost their effect.

Fig. 6.6-37 Description of the action of a microencapsulated two-effect agrochemical product [6.6.3.1]: a) microcapsule surrounded by a

disinfectant film, b) disinfectant evaporation, c) diffusion of the insecticide through the wall and dissipation of the active substance

6.7 Building Materials and Ceramics

6.7

Building Materials and Ceramics As already mentioned in Chapter 2, the manufacture of artificial building blocks (bricks) for the construction of shelters (protective walls and houses) and of ceramic materials as containers or for decorative purposes were among the earliest applications of agglomeration. Particulate solids (sand) and naturally occurring binders (clays) were mixed and formed into rectangular shapes, dried, and later fired to obtain a readily useable building material. Finer clays (later called china clay) were easily formed into hollow bodies (vases, containers) or artistic shapes that were also heat treated to yield permanent strength and were often decorated by mineral glazes, which fuse during firing into colorful and waterproof coatings. With time the manufacture of kitchen and tableware (earthenware, porcelain), tiles, and numerous industrial articles was added and synthetic raw materials (e.g., cement) and additives (e.g., silica fume, inorganic pigments) were introduced. Since heat treatment of the agglomerated solid particle mixture is a necessary and integral part of any application, size enlargement in this field is always a two-stage procedure, although, sometimes, both operations are carried-out in the same equipment. Both procedures, the production of the green pre-shape (agglomerate) and the heat treatment (sintering), determine the properties of the final product, which include porosity, cold and hot strength, and permanent changes on re-heating [B.13c]. To optimize the characteristics of the final product, particulate solid (feed, powder) treatment has become an important step in the manufacture of high-quality building materials and ceramics (Fig. 6.7-1). As shown in Tab. 6.7-1, agglomeration increasingly plays a major role in the preparation of raw materials and additives. It is used to improve the storage and handling properties of building and ceramic raw materials and additives and to enhance the flow properties and packing efficiency of fine powders prior to forming. As shown in Fig. 6.7-1, particularly for the manufacture of industrial ceramic parts, hot forming (called hot pressing or pressure sintering) or, more precisely, pressure assisted sintering (PAS), the simultaneous application of pressure and heat to a powder mass that is enclosed in a die, gains importance. In general, this technique allows the use of lower temperatures and pressures

Fig. 6.7-1 Block diagram of fabrication routes for building materials and ceramics

303

304

6 Industrial Applications of Size Enlargement by Agglomeration Tab. 6.7-1 Alphabetical listing of raw materials and additives for building materials and ceramics that have been and/or are being agglomerated and the products using or representing agglomerated parts Raw materials and additives Aggregate, alumina (aluminum oxide), cement raw mix, clay, dolomite, fluorspar, kaolin, light weight aggregate, lime, limestone, magnesia (magnesium oxide), pigments, silica (silicon dioxide), silica fume Products Bricks, building blocks, cement, decorative parts, earthenware, industrial ceramics, insulators, light weight aggregate, porcelain, pottery, manufactured concrete parts, refractory materials and parts, tiles

and shorter processing times than required for the two-stage process while, at the same time, producing finer grain sizes, lower porosity, higher purity, and precise dimensional control, resulting in near-net-shape parts. Tab. 6.7-1, without claiming completeness, lists some of the most important materials that are related to building materials and ceramics and have been agglomerated for a multitude of purposes. Further Reading

For further reading the following books are recommended: B.3, B.4, B.6, B.7, B.13a+c, B.15, B.16, B.21, B.22, B.26, B.40, B.48, B.51, B.55, B.56, B.61, B.64, B.70, B.72, B.76, B.80, B.81, B.82, B.89, B.93, B.94, B.97, B.98, B.103 (Chapter 13.1). Books mainly devoted to the subject matter are printed bold.

6.7.1


Most of the agglomeration techniques used for the manufacture of building materials andceramicsapplypressureforthedensificationandshapingofminerals(Section6.7.2). However, for better handling, metering, and feeding it is often desirable to produce a granular, free-flowing intermediate material in a preparatory step, which may use agglomeration in drums, pans, mixers, and, originating from suspensions, in spray dryers. For the manufacture of cement, two major technologies have been developed, using either dry or wet processing (Section 6.7.3). Before the preheating of the raw materials in high-efficiency cyclones became a standard method to achieve the most economic production of cement, pre-agglomeration of cement raw meal in balling drums was an accepted method of improving feeding to, environmental control of, and clinker production by the traveling grate used for sintering (Section 6.7.3). However, today this technology has all but disappeared. Finely divided clay, often after drying and grinding to destroy natural agglomerates, is a common binder in tumble/growth agglomeration. One of the more common and best known is bentonite (Section 15.1), a montmorillonite (Al[Si2O5]OH). The most


Fig. 6.7-2 Planetary intensive mixer with a flat bowl (courtesy Eirich, Hardheim, Germany)

valuable clay, used for fine ceramics, such as tableware (porcelain), is kaolinite (Al2[Si2O5](OH)4), which is cleaner but has similar binding characteristics. Therefore, it is easy to agglomerate such materials in drums or on pans with water as binder. However, production of small granules is not possible with this equipment. To achieve the manufacture of small granules, high shear forces must be created in the tumbling mass, preferably in a batch mixer. Typical equipment for this application includes the Eirich planetary mixer as a versatile rugged machine, which was originally designed for the mineral industries (especially building materials and ceramics) [B.48, B.97]. Later the same operating principle that had been carried out in flat bowls (Fig. 6.7-2 and Fig. 9.5, left, Section 9.1) was modified to accomplish better granula-

Fig. 6.7-3 Planetary intensive mixer/granulator with an inclined bowl (courtesy Eirich, Hardheim, Germany)

305

306


tion by using an inclined bowl design (Fig. 6.7-3 and Fig. 9.6, Section 9.1). Although most of the moisture must be removed prior to further use, the spheroidal shape of the dense granules results in excellent flowability, which is often of great interest in the fine ceramics industry and warrants the additional cost of wet agglomeration over dry compaction/granulation (Section 6.7.2). Today’s intermediate ceramic products, particularly for industrial applications, such as silicate molding materials for floor tiles or compounds for high-performance ceramics, must satisfy stringent quality standards. Therefore, conventional processes have been superseded by new technologies. For example, similar to one-pot-processing in the pharmaceutical industry (Section 6.2.1), mixing, homogenization, agglomeration, and drying, all in a vacuum, can be performed in a specially equipped and instrumented planetary mixer (Fig. 6.7-4). Dry prepared raw components, in the case of molding

Fig. 6.7-4 Flow diagram of an Evactherm preparation plant for ceramic molding materials (courtesy Eirich, Hardheim, Germany)


compound for ceramic floor tiles consisting of clay, feldspar, and grog, pre-ground to < 63 lm, are combined in a batching scale and transferred to the Evactherm mixer. Special (binder) additives are not required. After blending, using steam and water, the homogeneous mix is wetted to 10–11 % and agglomerated (granulated). Finally, the finished product is gently dried by the combined action of steam and vacuum in the same unit to about 6 %, the moisture desired for the next (forming) process. To meet the granule size specification (0.1–1.2 mm, Fig. 6.7-5), oversized particles are removed by screening and recirculated to the feed end of the process where, after milling, it is used as a component of the batch. The process has minimum impact on the environment as exemplified by Fig. 6.7-6 showing the extreme cleanliness in two Evactherm preparation plants. Through the use of vacuum technology, product quality and yield can be raised considerably. In response to the needs of the industry, similar processes have been developed by other vendors using different mixer designs (Section 15.1). Spray drying [B.48, B.97] is used in the manufacture of granulated press powders for tiles and electronic parts. It also plays an important role in the industrial production and development of high-performance (advanced) ceramics because of the ability of this process to meet size distribution requirements and give granules with a smooth surface and spherical shape, all resulting in excellent flowability, which is important for the subsequent forming step. Another reason for the widespread use of spray dryers in the ceramic industry is the fact that they can handle abrasive feedstocks without problems.

Fig. 6.7-5 Scanning electron micrograph of silicate ceramic granules produced in an Evactherm preparation plant (courtesy Eirich, Hardheim, Germany)

307

308


Fig. 6.7-6 Two preparation plants with Evactherm mixers demonstrating the extreme cleanliness of this process technology (courtesy Eirich, Hardheim, Germany)


In spray drying of ceramic materials, suspensions of particulate solids or pumpable slurries are divided into small droplets at the top of a tower by special wear resistant nozzle(s) [B.97] and dried with hot air. Since drying begins at the surface of the droplets, it is possible that voids are formed within the granules, which reduce the density of the product. To obtain the required density of the green (before sintering, Section 6.7.2) ceramic part, a longer stroke of the press is necessary which, sometimes, may be objectionable. Although the high cost of the energy to evaporate the suspending liquid burdens the process increasingly, the low expenditure for personnel, only a few persons are required for operation and maintenance, together with other advantages, related to granule shape and size, make the technology still attractive for the ceramic industry. In the building materials industry, particularly for the do-it-yourself markets, it becomes an increasing trend to pre-mix all components. Such blends may contain sand, lime, cement, pigments, plasticizers and others. Owing to different particle sizes and densities of the ingredients and the often limited mixing capabilities of the user, they tend to segregate in the bag or container and during application, causing quality and color variations in the finished product. Separation during packing, storage, and handling of the homogenized blend is avoided by encouraging some agglomeration during mixing, combining the components uniformly into loose, non-segregating, easily dispersible granules. Manufacturers of high-quality building materials require machines and plants that must be rugged and tough to withstand the harsh conditions caused by the abrasive raw materials. Particularly the mixer/agglomerators must be of relatively simple but efficient design to allow wear protection by hard metal liners and the application of overlay coatings on the blending tools. Fig. 6.7-7 shows the building with storage silos of a plant for the production of dry mortars and Fig. 6.7-8 is a view into a facility with high-intensity mixers for the processing of sand-lime brick aggregates. Silica fume, a by-product of the manufacture of ferrosilicon and silicon metal that is captured in the plants’ dust collection systems (Section 8.1), consists of amorphous silica particles that are spherical, have very small dimensions (0.05–0.5 lm [B.97]), and feature a large specific surface area. These characteristics make it an excellent admixture to, for example, high-strength concrete and high-performance grouts and mortars. The amorphous structure and high surface area render the particles very reactive, causing pozzolanic effects, and the small particle size results in highly impermeable structures of building materials. In pre-stressed concrete, the latter protects the reinforcement bars from attack by water. As collected, silica fume is very dusty, difficult to handle, self-agglomerating (causing bridging, build-up, and lumping), and can not be transported and handled economically. Because of its very low bulk density (160–240 kg/m3), bulk tanker trucks only hold 8–10 ts and require long pump-off times and if bags are used they are large, light, and bulky. A simple dry fluid bed agglomeration process (Fig. 6.7-9 [B.97]) converts silica fume by dry agglomeration into a product, which is less dusty, flows well, and can be handled pneumatically. Product density may be such that the tanker truck now holds about 25 ts and can be adjusted to fit different handling and end use applications. At the same time, agglomerate bonding is so weak that the product disperses easily, for example in cement mixers (compare Section 6.3.1, pigments).

309

310

6 Industrial Applications of Size Enlargement by Agglomeration Fig. 6.7-7 Plant for the production of dry mortars with storage silos

A final application of tumble/growth agglomeration for building materials that shall be mentioned is the manufacture of expanded clay, often also called ESCS (expanded shale, clay, and slate). The product is a lightweight ceramic aggregate. After mining the clay, the raw materials are cleaned (removal of coarse and foreign components), homogenized by kneading, and divided into small pieces (Section 6.7.2). At the cold end of a long rotary kiln the clay pieces are first agglomerated and rounded. Later, at temperatures of about 1200 8C, the clay expands forming a porous interior while the outer surface fuses into a strong ceramic skin (Fig. 6.7-10). For different applications, various sizes in the range 0–25 mm are produced by screening. The manufacture and raw material selection processes are strictly controlled to ensure a uniform, high-quality product that is structurally strong, stable, durable and inert, yet also lightweight and heat and sound insulating. Therefore, ESCS gives building designers greater flexibility in creating solutions to meet the challenges of wall and


Fig. 6.7-8 View into a facility with high-intensity mixers for the processing of sand-lime brick aggregates (courtesy Eirich, Hardheim, Germany)

Fig. 6.7-9 Diagram of a fluidized bed agglomerator for dry silica fume (courtesy Norchem Concrete Products, Fort Pierce, FL, USA)

311

312


Fig. 6.7-10 Expanded clays, showing the porous interior and the fused ceramic skin (courtesy Fibo ExClay, Lamstedt, Germany)

construction specifications, dead load, terrain, seismic conditions, and budgets. According to the Expanded Shale, Clay, and Slate Institute (Section 15.1), for nearly a century (originally developed in 1908) expanded clay products have been used successfully around the world in more than 50 different types of applications. The most notable among these are lightweight concrete masonry, high-rise buildings, concrete bridge decks, high-performance concrete (HPC), pre-cast and pre-stressed concrete elements, asphalt road surfaces (chip seal, asphalt surface treatment), soil conditioners, and lightweight geotechnical fills. Today ESCS aggregate, as manufactured by the rotary kiln process, is available throughout the world. ESCS ceramic lightweight aggregate improves concrete in several ways. Tab. 6.7-2 summarizes the advantages of this product in the building industry. Tab. 6.7-2 Advantages of ESCS in the concrete building industry (adapted from information provided by The Expanded Shale, Clay, and Slate Institute, Salt Lake City, UT, USA, and Fibo ExClay, Lamstedt, Germany) * *

*

*

*

Ability to design concrete with specified density, especially low density concrete. Concrete with ESCS lightweight aggregate features * better thermal properties * better fire resistance * reduced shrinkage * reduced chloride ion permeability * improved freezing and thawing durability * improved contact zone between aggregate and cement matrix * less micro-cracking as a result of better elastic compatibility * better blast resistance * better shock and sound absorption * no attack by rodents and vermin * no decay and decomposition. ESCS carrying absorbed water provides internal curing (often referred to as self-curing or water entrainment) that improves the hydration of the cement. High-performance lightweight aggregate concrete also has less cracking, improves skid resistance, and is readily placed by the concrete pumping method. Concrete with ESCS lightweight aggregate can be recycled (typically after crushing).


6.7.2

Pressure Agglomeration Techniques

In most cases, the shaping of building materials and ceramic parts (collectively called “ceramics”) is accompanied with a densification of raw mineral ingredients with or without additives (such as binders, lubricants, plasticizers, pore formers) and mixtures thereof. Also, many final parts, particularly high-performance ceramics, must in the end have net shape, meaning that they exhibit ultimate dimensions and tolerances. In this respect, the manufacturing process is similar to that encountered in powder metallurgy (P/M, Chapter 7). While for P/M parts, a near-net-shape that requires some final machining before the intended use is often satisfactory, additional machining of ceramics is normally not feasible or possible because of their low cost (e.g., building materials) and, generally, their hardness, abrasiveness, and brittleness. Therefore, during the manufacture of ceramic pre-shapes for subsequent sintering, it is most important to take into consideration possible shrinkage and/or distortion during firing of the green parts. The extent of densification during the initial forming of ceramic powders into green parts defines the amount and orientation of shrinkage during post-treatment (sintering). Within certain limits, the density of a final part sintered from a low-density green ceramic pre-form will be the same as that obtained from a green part with high density, but the higher shrinkage during firing makes it more difficult to anticipate and control the shrinkage to produce a well-dimensioned and shaped final part. Therefore, for the manufacture of true-to-size ceramic components the density of the green part should be as high as possible. Component size, particularly the relation between vertical and horizontal dimensions of a green part, also plays an important role in how the outline of a body changes due to shrinkage during sintering. For example (Fig. 6.7-11, [B.21, pp. C39–C52]), if the same mass of a ceramic raw material (steatite) is compressed by a punch in a die with square cross section using varying compaction pressures, different green densities, characterized by the height Hu (upper part of Fig. 6.7-11) of the resulting body, are obtained. With increasing compaction pressure, the density of the part increases (decreasing volume). Of course, the length of the edges, Lu, is always the same and is equal to the sides of the die. During firing, these bodies shrink and, as mentioned above, all parts attain approximately the same density (in the case reported here, 2.75 g/cm3). However, as shown by Hg, the heights, and Lg, the edge lengths of the sintered parts, the shape of each body is different. Because with increasing pressure a more pronounced orientation of non-isometric particles in the raw materials occurs, the deformation of the fired parts as a function of compaction pressure is further intensified (lower part of Fig. 6.7-11). Sh (shrinkage of the height) and Sq (shrinkage of the edges) do not change at the same rate. Many raw materials for ceramic parts contain clays. During densification, the lamellar shape of clay minerals results in a more-or-less defined texture because the foliated silicates tend to orient themselves vertically to the direction of force. Since lamellar particles shrink non-isotropically, such textures may result in high stresses in the sintered part, causing cracking during cooling. The orientation effect also plays a role in

313

314

6 Industrial Applications of Size Enlargement by Agglomeration Fig. 6.7-11 Above) dimensions of fired (sintered) steatite parts obtained from green bodies with identical mass that were densified with different pressures, Lu = unsintered length, Lg = sintered length, Hu = unsintered height, Hg = sintered height; below) vertical and horizontal shrinkage that occurred during sintering, Sh = height shrinkage, Sq = length shrinkage [B.21]

Fig. 6.7-12 Vertical and horizontal shrinkage occurring during the firing (sintering) of parts from raw materials with different densities, which were compressed to the same green density and the relation of both [B.21]


defining the final dimensions since it may influence the horizontal and vertical shrinkage in different ways. For example, tests using steatite raw materials with different densities revealed that, if green parts are produced with the same density by the application of different compression forces (high- for low-density particles and low- for high-density particles), the shapes of the sintered bodies are not the same [B.21, pp. C39–C52]. According to Fig. 6.7-12, with increasing density of the raw materials the vertical shrinkage decreases while the horizontal contraction increases somewhat. As a result, the relationship between vertical and horizontal shrinkage declines steeply. Therefore, if higher compaction pressures are used during the shaping of green bodies, the resulting sintered parts are shorter and wider. This was explained by the more pronounced flow of material during increased densification when higher forces are applied, which results in a distinctly intensified orientation of the lamellar particles perpendicularly to the direction of the force. It has been shown earlier (Section 6.2.2, Fig. 6.2-14) that during the densification and compaction of particulate solids, density variations occur as a result of the effects of internal and external friction. Together with the associated changes in primary particle orientation, which are further modified by variations in the direction of the acting forces, rather unpredictable shrinkage behavior during sintering is obtained if these effects are not considered during manufacture of the green bodies. This is particularly important if parts with different cross sections are to be made and show the importance of the agglomeration step for the entire manufacturing process and the shape and the dimensions of the sintered component. More generally, a thorough understanding of the microstructural dependence of the properties of green and fired ceramics is necessary [B.80]. It is the critical link between forming, processing, and properties and/or performance of the finished part. Forming (always an agglomeration method) is the manufacture of specific shapes by punchand-die pressing, isostatic pressing, injection molding, extrusion, and slip casting and includes as crucial parameters also the effects of binders, mixing, and consolidation for producing the actual size and shape of uniform green bodies. Processing refers to the post-treatment by the application of heat and includes the influences of particle character, green density, and sintering parameters on the microstructure of the finished components. Forming and processing are often referred to in one term, fabrication, because the microstructural features of the final part are a direct consequence of the starting material’s character and the parameters of forming and processing. Fabrication of ceramics typically begins with the consolidation of powders. After that, two alternative paths are commonly followed [B.80]. One is to obtain a desired porosity to achieve key application characteristics. The requirements may include a particular specific surface area, which increases with porosity but must be balanced against other physical properties that decrease with increasing porosity, particularly strength (Section 6.7.3, Tab. 6.7-5). The other path is used when high or maximum levels of physical properties, such as strength, optical transmission, thermal conductivity, are sought, which necessitate low to zero porosity. With increasing firing conditions (temperature and time), many characteristics of the final component, such as thermal and electrical conductivity and elastic properties, increase at different rates for

315

316


a given material to varying plateaus. Similarly, other properties, especially most mechanical attributes, initially increase at varying rates but pass through variable maxima and then decrease (Fig. 6.7-13). All these changes are related to the starting particle size and shape, their orientation and packing (density) in the green body and other parameters, which result from the different agglomeration methods. As mentioned at the beginning of this section, in most cases, the shaping of building materials and ceramic parts is accompanied by a densification of raw mineral ingredients with or without additives (such as binders, lubricants, plasticizers, pore formers) and mixtures thereof to yield green bodies, which must be post-treated (Section 6.7.3) to achieve the ceramic part’s final properties. Although a number of different agglomeration methods are used to a varying degree for the manufacture of green ceramic bodies, three pressure-based technologies are the most common ones: * * *

punch-and-die pressing extrusion, and isostatic pressing.

The oldest ceramic product is bricks, a building material made from naturally occurring moist or wetted clay, originally shaped by hand in wooden forms, sometimes reinforced with fibers (straw), and air (sun) dried to obtain a building block. In later times, the green body was fired to yield a strong and waterproof product. While today in some undeveloped areas, this basic (manual shaping and sun drying) manufacturing principle is still used and applied for the construction of adobe buildings, large quantities of bricks with many different qualities and for a large number of applications are produced by punch-and-die presses and extruders. At the beginning of industrialization, various designs of mechanical punch-and-die presses with single or multiple cavities, including “stone presses” with indexed die tables, were used, particularly for the manufacture of high-quality bricks that required dimensional precision for their application (e.g., refractory bricks). Among the earliest machines of this type was the Couffinhal press (Fig. 6.10-14, Section 6.10.2). Later, hydraulic presses were also used.

Fig. 6.7-13 Diagram of maxima of properties, such as strength, that are dependent on both porosity decreases and grain size increasing as a function of firing temperature or time [B.80]. The dashed line suggests that maxima may change with different material and fabrication parameters


Fig. 6.7-14 Typical robust double-screw extruder with integrally mounted pug sealer for the processing of stiff materials (courtesy J.C. Steele, Statesville, NC, USA)

For mass production, however, extrusion has become the technology of choice [B.97]. After first applying ram extrusion (Exter press), although this machine was originally invented for the briquetting of solid fuels (Fig. 6.10-3, Section 6.10.2), the continuous single- or double-shaft screw extruder was introduced and, today, is the prevalent equipment for the manufacture of clay-based building materials, including bricks [B.97]. Fig. 6.7-14 shows a typical, albeit smaller, robust double-screw extruder with integrally mounted pug sealer for the processing of stiff materials, such as clays. Pug sealers are conditioners that combine an open pug mill for intensive high-shear mixing with an enclosed (sealing) screw for degassing and a shredder for easy feeding into the vacuum chamber of the extruder with two counter-rotating screws. The tough, plastic and/or sticky clay is transported into a wide throat area, which ends in a flanged front barrel that accepts a wide variety of dies and die holders. A variety of dies is available for the extrusion of solid and hollow bricks and specially shaped building blocks, roofing tiles, and other hollowware for construction purposes. Fig. 6.7-15 shows some of the finished extruded products. On the extruder mouth, the dies can be mounted fixed or in single- and double-hinged holders or a hydraulic chan-

Fig. 6.7-15 Some typical extruded clay-based building products (courtesy J.C. Steele, Statesville, NC, USA)

317

318


ger, the latter to allow quick change-over between different shapes [B.97]. The continuous strands emerging from the extruder are cut to size with different cutter designs. Individual large extruders can produce more than 40 000 standard bricks per hour representing a capacity of about 80 t/h. More complicated shapes, required in much smaller numbers, are made on versatile, die-changer-equipped smaller machines as shown, for example, in Fig. 6.7-14. Lower duty, simpler machines are used during the manufacture of expanded clay (Section 6.7.1). As depicted in Fig. 6.7-16, the product still shows that it originated from cut extruded ropes. A similar expanded clay material is made using pellet mills, particularly the heavy-duty flat die machines (Chapter 5, Fig. 5.10, b.2), for the manufacture of green clay extrudates. Extruders are also used in the production of sintered industrial parts. Of special importance is their application for the manufacture of catalyst carriers (Section 6.3.2, Figs. 6.3-20 and 6.3-21). The use of wear-resistant alloys for all parts that come in contact with the material to be extruded make the extruders suitable for the processing of highly abrasive catalysts, other chemicals, and minerals, such as molecular sieves, high-purity aluminas, and kaolin carriers. Interchangeable die plates allow the extrusion of an almost unlimited variety of sizes and shapes on the same basic machine that is then equipped with a variable-speed screw-drive to adjust retention time, pressure, and production rate. For more structured industrial, sanitary, and household ceramic parts mechanical and hydraulically operated punch-and die presses are used. Fig. 6.7-17 depicts a modern mechanical automatic press showing the complexity of such equipment. Such machines are available with pressing forces of 60–4500 kN, filling height of up to 180 mm, pressing travel (densification) of up to 90 mm, and stroke rates of 4–60 per minute (decreasing with increasing press size). Hydraulic presses (Fig. 6.7-18) are stronger (up to 26 000 kN) but all feature lower stroke rates (maximum 30 per minute and typically less). With both types of presses the platens can be subdivided into smaller sections (Fig. 6.7-19) to accommodate multiple die holders for the manufacture of smaller parts and to increase the production rate for a more economical operation. Fig. 6.7-20 shows several industrial ceramic parts demonstrating the detail that can be achieved with this technology.

Fig. 6.7-16 Expanded clay produced from green extrudates (courtesy J.C. Steele, Statesville, NC, USA)


Fig. 6.7-17 Modern automatic mechanical press for the production of green ceramic bodies (courtesy Dorst, Kochel am See, Germany)

During densification, particularly of powders, density variations are obtained as a result of the effects of internal and external frictional forces. Fig. 6.7-21 depicts the simplified density distribution obtained after the unidirectional compaction of particulate solids in a cylindrical mold (Section 6.2.2, Fig. 6.2-54). In all applications that require post-treatment by heat to obtain final properties (Chapter 7), density variations or gradients in a green body must be of concern as they will cause uneven shrinkage and may result in distorted or warped final parts. During the unidirectional compaction of a part with variable cross sections, the density is higher just above an internal corner because of the elevated shear stress at this point in the mold (Fig. 6.7-22). Under the corner, the density is lower since the high-density region above prevents the powder from flowing downward. In addition to all the previously mentioned shape related problems due to this density distribution, a crack is likely to open up in the corner, which may already be present undetected in the green body [B.61].

319

320


Fig. 6.7-18 Modern automatic hydraulic press for the production of green ceramic bodies (courtesy Laeis Bucher, Trier, Germany)

Fig. 6.7-19 Example (ref. Fig. 6.7-18) of a mold use and performance table (courtesy Laeis Bucher, Trier, Germany)


Fig. 6.7-20 Industrial ceramic parts demonstrating the detail that can be achieved with this technology (courtesy Komage, Kell an See, Germany, and Dorst, Kochel am See, Germany)

321

322

6 Industrial Applications of Size Enlargement by Agglomeration Fig. 6.7-21 The density distribution obtained after the unidirectional compaction of particulate solids in a cylindrical mold

Fig. 6.7-22 Areas of major density variation obtained during the unidirectional compaction of a part with variable cross sections [B.61]

The various potential defects resulting from the changing density in a compressed powder part are of great concern. Net-shape articles are often desired because of the abrasiveness and/or hardness of many of the components preventing machining after sintering. This means, that, after taking into consideration the unavoidable dimensional changes during and after the sintering process, the finished item should posses dimensions that need no or minimal adjustment for its intended use. One way of reducing the development of uneven densities in green parts during compaction is to lower friction by the addition or application of lubricants [B.48, B.97] (Section 6.2.2). A secondary effect of lubrication is that wear of the tooling is also reduced. “Double pressing”, that is, the densification of particulate solids by the movement of both the top and bottom punches, is used to overcome at least part of the uneven density distribution caused by unidirectional pressing. If both punches move towards each other with the same speed, exerting the same force, a mirror image of the density distributions across a “neutral axis”, which, in this case, is in the middle of the body, will be created (Fig. 6.7-23). If, in addition, the die walls move with or to some extent in regard to the travel of the punches (withdrawal die), the location of the neutral plane

6.7 Building Materials and Ceramics Fig. 6.7-23 Sketch of the density distribution and the location of the “neutral axis” in a simple (cylindrical) green part obtained by symmetrical “double pressing”

(axis) can be further influenced (B.28, B.48, B.97]. It is necessary to control the position of the neutral planes (low-density zones that are perpendicular to the direction of pressing) particularly in complex parts. This is achieved by the relative motions of the tools (Fig. 6.7-24). It is also important to understand that within the developing structure of a compact, especially under pressure, particles will not move from one level or position to another. As a consequence, if parts are pressed that have more than one shape feature, separate forces and movements must be applied simultaneously, and neutral planes will exist in each level (Fig. 6.7-25).

Fig. 6.7-24 Drawings describing ways of influencing the location of the “neutral plane” in unidirectional (upper punch) pressing by controlled die withdrawal [B.28, B.48, B.97]

323

324


Fig. 6.7-25 Different “neutral planes” in single- and multi-level parts [B.28, B.48, B.97]

Today, the best method of overcoming the problem of uneven densities in a green body is isostatic pressing. In this compaction process, the pressure is applied by a fluid that is pressurized and acts uniformly from all sides on dry particulate solids that are enclosed in a flexible container (“mold”) immersed in the fluid [B.13a]. This results in the most uniform consolidation possible. Of course, there is still a density gradient across the part, the center is always less compacted than the surface, but the gradient is uniform and, typically, does not cause distortion during sintering. Isostatic pressing was patented around 1910, and about 30 years later the technology found a general and large scale application for the direct isostatic pressing of spark plug insulator pre-forms [B.48]. Technically it is quite immaterial whether or not sintered ceramic spark plug insulators are totally straight. However, as a consumer product which, at that time, needed replacement often, the part, in addition to functioning well, had to look good. After introducing isostatic pressing into the manufacturing process, the number of rejects became very low. Later, the technology was used for the production of many other ceramic parts, particularly of the high-performance variety, and for many applications in powder metallurgy (Chapter 7). Isostatic pressing is carried-out cold or hot using dry powders as feed material. The most common application is still cold isostatic pressing (CIP), which is performed at ambient temperatures. In hot isostatic pressing (HIP) the forming and densification process is achieved uniformly by heated high-pressure gas in an autoclave. Especially in powder metallurgy (Chapter 7), the material to be processed is itself often also brought to elevated temperature prior to loading. Unlike CIP, in which powders are always containerized, HIP may be applied for containerized powders and also for preformed non-containerized components manufactured by any method. Hydrostatic pressing is a term often used as a synonym for isostatic pressing. Isostatic pressing is the generic term covering liquids and gases as the pressure transmitting medium whereas hydrostatic pressing is best reserved for liquids. However, the two are used interchangeably to cover both aspects. If flexible containers are used, their arrangement may be such that they contract or dilate under the application of pressure. Whether the tooling is an integral part of the


press or loaded and removed during each compaction cycle determines if it is a “drybag” or “wet-bag” process [B.48, B.97]. The difference between the two methods is illustrated in Fig. 6.7-26. In the dry-bag process, the flexible container is fixed in the pressure vessel and the powder is loaded directly. The tool forms a membrane between the fluid and the powder. Optionally, the flexible container may be placed inside a primary diaphragm so that the powder never comes in contact with the fluid, even if the flexible mold is damaged or breaks. Therefore, dry-bag pressing also has the advantage that the fluid is not contaminated by the powder. However, because the container must stand up to many pressing cycles and since changing is time consuming, it has to be made of a very durable material. The wet-bag process, in which the container that has been filled externally with the powder, is entirely submerged in the fluid inside a pressure vessel, uses the simplest type of equipment. This process is commonly used for the production of single large components or a large number of small parts. While dry-bag tooling can be fitted with means to remove the gas displaced during densification, the material in containers for wet-bag pressing must be consolidated and evacuated prior to closing and loading into the autoclave to avoid compressed air pockets which, upon pressure release, may damage the structure of the compacted parts. The basic principles of isostatic powder pressing are summarized in Tab. 6.7-3. Pressure equipment consists of powder storage and dispensing facilities, at least one pressure vessel with means for loading and unloading the tooling or parts, pressure generator(s), and related items that enable effective and safe operation of the process. Dry-bag pressing is used for the production of small components at a

Fig. 6.7-26 Diagram of the differences between dry- and wet-bag pressing [B.13a, B.48, B.97]

325

326


Summary of the basic principles of isostatic powder pressing [B.97]

1. The wet bag pressing of large and/or complex shapes in which the flexible container is filled and prepared outside the pressure vessel and then immersed in the fluid and compacted. 2. The dry bag pressing of smaller, regular shapes in which the tooling forms an integral part of the pressure vessel. 3. The use of internal or external rigid formers to produce accurate surfaces. 4. Pressurization by systems using pumps or by direct compression with pistons in a die. 5. Handling systems for the filling, preparation, loading and unloading of powders and parts as well as for the tooling and pressing equipment.

high rate. Fig. 6.7-27 demonstrates the compaction, ejection, and filling of a dry-bag press during the manufacture of spark plug insulators. It is relatively easy to automate the operation of this process. The permanent location of the tool and small fluid volume surrounding it contribute to a fast operation. Production rates in the neighborhood of 100 parts per minute are common. The actual production speed depends on powder properties, size of the part, maximum pressure and potential dwell time requirements, the number of tool cavities and pressure and handling needs. Some automated installations feature a round, sequenced pressure-chamber system (Fig. 6.7-28) while others hold multiple dies in a

Fig. 6.7-27 Operational sequence of a dry-bag isostatic press for the manufacture of spark-plug blanks [B.13a, B.48, B.97]


common tool holder (Fig. 6.7-29). The cold isostatic press shown in Fig. 6.7-29 is designed for the production of balls (5–90 mm diameter), tubes (4–150 mm diameter, up to 250 mm long), rods (3–114 mm diameter, up to 250 mm long), filter inserts, and valve bodies. The three main, of several possible, demolding (extraction) methods are demonstrated in Fig. 6.7-30, and Fig. 6.7-31 shows some typical finished parts. While hot isostatic pressing was originally developed and used to remove defects and/or produce parts with minimum porosity and, consequently, ultimate density from powders and preforms, the study and understanding of the mechanisms of pressure agglomeration also led to a modification of the process, which is actually used to produce parts with a controlled high porosity [B.78, B.97]. During this new HIP process for making porous products, open, non-containerized powder compacts or loosely sintered bodies are HIPed at high temperatures in a high-pressure gas atmosphere. In this situation, overall densification of the part is avoided by the high-pressure gas that fills the open pores. Thus, high open porosity of sintered components can be obtained by the combination of open HIPing and simultaneous sintering from materials that normally experience considerable densification if conventional HIP or CIP followed by sintering are applied. The porous products are, for example, suitable as filters. Today, the efforts to avoid distortion of ceramic products during firing by the use of isostatic pressing for the production of preforms has even led to the design and application of this technology for tableware. Fig. 6.7-32 shows the open tooling section in a “free fall” horizontal isostatic press, the two parts of the die system, and the depic-

Fig. 6.7-28 Operational sequence of an automatic isostatic press with round, timed (sequenced) tooling table [B.13a, B.48, B.97]

327

328


Fig. 6.7-29 Photograph and diagram of a dry-bag CIP system (courtesy Dorst, Kochel am See, Germany). 1, protective enclosure; 2, control cabinet; 3, hydraulic unit; 4, high-pressure intensifier pump; 5, loading device; 6, hopper; 7, dosing device;

8, closing lever with preload cylinder at top; 9, upper tool; 10, device for parts removal from above (Fig. 6.7-30); 11, lower tool; 12, conveyor belt; 13, tool component for ejection from below (Fig. 6.7-30); 4, closing lever with preload cylinder at bottom


Fig. 6.7-30 The three main demolding (extraction) methods (courtesy Dorst, Kochel am See, Germany)

Fig. 6.7-31 Some typical parts manufactured by cold isostatic powder pressing (courtesy Dorst, Kochel am See, Germany)

329

330


tions of a raw ceramic bowl and the corresponding tooling design. In the latter, the cross-hatched parts represent polyurethane as a (protective) coating (top) and a membrane (bottom) through which the isostatic pressure for consolidation is applied. Fig. 6.7-33 presents some shapes and actual flat ware products that can be made Fig. 6.7-32 Tooling section of a horizontal isostatic press for the manufacture of tableware preforms: a) open section in a “free fall” horizontal isostatic press, b) the two parts of the die system, c) parts of a raw ceramic bowl and the corresponding tooling design (courtesy Sama, Weissenstadt, Germany)


Fig. 6.7-33 Some shapes and actual flat ware products that can be made with a “free fall” horizontal isostatic press (courtesy Sama, Weissenstadt, Germany)

331

332


Fig. 6.7-34 A “free fall” horizontal isostatic press for the manufacture of tableware preforms (courtesy Sama, Weissenstadt, Germany)

by this method and Fig. 6.7-34 shows a machine that transforms dry ceramic spray granulate into evenly compacted dry-pressed articles that do not deform or lose shape and size during firing. The method can be applied for the manufacture of articles that are round, smooth or festooned, square or multi-cornered, with reliefs or not round, small or large (from saucers of 10 cm diameter to oval plates 40 cm long) and flat or deep (including 9–10 cm deep salad bowls) from porcelain, stoneware, earthenware, vitreous china, and bone china.


6.7.3

Other Technologies

Agglomeration by the application of heat (sintering) is a technology that is used during the manufacture of one of the most important building materials, cement, and of others and for the post-treatment (finishing) of essentially all ceramics. Chapter 2 described how the production of materials for the building of shelters was among humankind’s first activities. The search for materials to bond stones together also has a long history, which extends back to Neolithic times [B.69]. Carbon dating of materials found at a site in southern Galilee indicates that a lime-based concrete, used for the construction of polished floors, was made around 7000 BC. The extensive excavated floor area reveals that considerable quantities of lime were necessary to produce such a large amount of concrete. From this it was deduced that the technology of burning (calcining) limestone to form calcium oxide (CaO, lime), slaking the lime with water to form Ca(OH)2, and then mixing the slaked lime with limestone aggregate to obtain concrete was well known to the Neolithic builders. Lime-based concrete hardens slowly by reaction with carbon dioxide forming calcium carbonate. Later, in the construction of the Pyramid of Cheops (2613–2494 BC), another inorganic cementing material was made from gypsum (CaSO4 2H2O). The calcium sulfate was first dehydrated, then mixed with sand, and, when water was added, rehydration occurred, forming interlocking gypsum needles, which give the mortar its strength. Other historians have somewhat different opinions on the composition of early cements but all confirm the long history of using inorganic cementing materials by humans. Hydraulically setting cements were first developed by the Greeks and Romans [B.69]. It was found that, after the addition of volcanic ash (pozzolana) to the slaked lime and sand, a mortar was obtained that possessed superior strength. This was discovered by the Greeks (700–600 BC) and later passed on to the Romans (150 BC) who called it Caementum. It was used, for example, for the construction of the Colosseum in Rome. Hydraulically setting concrete, a mixture of aggregate, sand, and Caementum, has been used since Roman times. Throughout the centuries, a number of modifications and improvements were made but without changing the basic composition and the use of natural pozzolanic materials. In 1818 in France, Vicat prepared an “artificial Roman cement” by calcining a mixture of limestone and clay, which was distinctly different from naturally occurring clay/lime admixtures [B.69]. It was the forerunner of Portland cement, although Portland cement was only patented by Aspdin in 1824. The modern high-temperature process was discovered in 1844 by Johnson who heated the ingredients to a temperature at which they partially melted. It was recognized that the amounts of clay and lime had to be carefully proportioned but it took until 1887 (Le Chatelier) before upper and lower limits were defined. The production of Portland cement normally involves the firing of a mixture of finely ground calcareous and argillaceous materials in a kiln at about 1500 8C and the formation of a clinker, which consists of a number of compounds that set (harden) when the clinker is ground to a fine powder (cement) and then mixed with water. Today it is known that the quality of cement is defined by the fineness and the exactly

333

334


Fig. 6.7-35

Diagram of a rotary cement kiln [B.69]

controlled composition of the mixture of raw materials, the temperature and duration of the calcination, the method of cooling the clinker, and the specific surface area (fineness) of the final product. In ancient Roman and later times vertical (now called shaft) kilns were used. Originally, they were manually charged and controlled. The irregular operation often resulted in clinker with unpredictable and inferior properties. More recently, the raw materials were finely ground, uniformly mixed, agglomerated in drums or discs, and fired employing tight process control (Section 6.8.1, Fig. 6.8-17a). The disadvantages of this kiln design are the large overburden pressure under which pellets may break, disturbing the gas flow and resulting in uneven clinker quality, and a relatively low production capacity. The development of the rotary kiln, now predominantly used for the production of Portland cement, started around the late 1870s but the technology was not patented until 1885 (Ransome [B.69]). A rotary kiln (Fig. 6.7-35) is a long refractory-lined steel cylinder that is inclined at about 3–68 to the horizontal. At the lower end is a burner (coal, oil, or gas fired) and the material to be calcined enters on the other end. While passing down the kiln, chemical and physical reactions take place in the oxidizing atmosphere. Hot clinker emerges at the burner end and must be suitably cooled before being milled to yield cement. The rotary kiln process may be carried out with wet (slurry), semi-dry, or dry feed. Two agglomeration processes occur successively in the rotary kiln using wet feed. The schematic diagram of a coal (dust) fired wet process rotary kiln, also showing ancillary equipment and components, is depicted in Fig. 6.7-36 [6.7.3.1]. The raw materials are mixed and ground before water is added to form a slurry (37–39 % moisture), which is fed into the kiln (right side of Fig. 6.7-36). In the kiln, water is driven off and, as the moisture evaporates, the flow properties of the slurry change, tumble/ growth agglomeration occurs, and granules are formed. In Fig. 6.7-37 typical temperature profiles of gas and solids along the length of the kiln are shown [B.69]. As the granules dry in the early part of the preheating zone (material temperature < 100 8C) the colloidal clay provides a permanent bond, which is enhanced by partial melting and sintering in the calcining and clinkering zones. Air (gas) is preheated during clinker cooling and used in the burners and the rotary kiln to boost the temperature. The wet process, starting with a large amount of water in the feed (slurry), requires costly pre-drying to initiate agglomeration within the rotary kiln. Therefore, as soon as balling (wet tumble/growth agglomeration) in drums and pans became a reality


Fig. 6.7-36 Diagram of a coal (dust) fired wet process rotary kiln, also showing ancillary equipment and components [6.7.3.1]

around the middle of the 20th century and was applied for the size enlargement of fertilizers (Section 6.6.1) and iron ores (Section 6.8.1, Fig. 6.8-17b and c), in an effort to develop a more economical method for the manufacture of Portland cement, cement raw meal mixtures were also converted into spherical pellets, now containing only 10– 12 % moisture. With this greatly improved feed (in addition to the lower moisture content pellets also feature more uniform composition and structure than the nodules formed in the rotary kiln), grate and grate-kiln machines were introduced to carry out the highly efficient “semi-dry” cement making processes.

Fig. 6.7-37 Typical temperature profiles of gas and solids along the length of a rotary cement kiln [B.69]

335

336


Fig. 6.7-38 Traveling grate for the drying and partial calcination of pellets (agglomerates, nodules) prior to entry into the rotary kiln [B.69]

Because the feed mixture to the cement clinker manufacturing furnace is already agglomerated, a more efficient method of heat exchange can be used. In the grate-kiln type furnaces (Section 6.8.1, Fig. 6.8-11c and 6.8-16) pellet drying and preheating are carried out on a traveling grate in downdraft fashion with the hot gases coming from the rotary kiln (Section 6.8.1 and Fig. 6.7-38). Fig. 6.7-39 shows the use of only a traveling grate for the production of cement clinker. Calcination and clinkering are accomplished by the burning of fuel (coal) that has been incorporated into the feed pellets during tumble/growth agglomeration in a drum (Section 6.8.1, Fig. 6.8-13) and/or added to the agglomerated material bed on the grate. Because the temperature must reach the clinkering temperature, the structure of the traveling grate must be protected. This is accomplished by depositing suitably sized clinker from the screen (11 in Fig. 6.7-39) into a hearth layer prior to building the bed of raw meal/fuel pellets (Section 6.8.1, Fig. 6.8-15). All kilns using grates, partially or totally, suffer from high installation and maintenance costs. Therefore, although the grate-kiln and straight grate machines have been

Fig. 6.7-39 Simplified flow diagram of a grate kiln cement clinker manufacturing facility (courtesy, Lurgi, Frankfurt/M., Germany). 1, 2, raw material receiving and pre-crushing; 3, 4, 5, raw material drying and milling; 6, 7, 8, 14, raw meal (6) and fuel

(coal, 8) proportioning and mixing; 9, 10, drum agglomeration and pellets sizing (roller screen); 11, clinker screen; 12, sinter grate; 13, 15, dust collection


and are used in a great number of cement plants and are particularly suitable if systems must be shut down or started-up quickly and without problems (e.g., rotary kilns must be turned with emergency power and drives as long as they are hot to avoid major damage due to bending), the rotary kiln process is still considered simpler and features larger production capability. A major problem associated with the wet rotary kiln is the very bad heat transfer between the hot gases from the kiln and the slurry that essentially coats the furnace walls until granulation occurs and the charge begins to tumble. To obtain satisfactory economy, a number of methods, all designed to improve the transfer of heat at the feed end of the rotary kiln, have been invented and used. The oldest and among the most effective is the installation of slack metal chains in the furnace tube (Fig. 6.7-40). They absorb heat from the hot gas and, as the kiln rotates and the hot chains mix with the feed, transmit it to the slurry. However, there are problems with slurry adhering to the chains and other heat exchanger designs have their own difficulties. When it was found that dry cement raw meal mixtures produce clinker from lumps formed in the furnace at elevated temperatures due to partial melting and sintering if the materials can be contained and the particulate solids discharging with hot gas can be controlled by collection and recirculation, the basic principle of the dry cement manufacturing process was born. The method is only feasible with the rotary kiln because fine, dry raw feed can not be retained on grates. The most important requirement is to remove entrained solids form the hot off-gas with highly efficient dust collection equipment (cyclones) and recirculate the particulates into the feed end of the kiln. Since drying of slurry (wet process) or partially pre-dried feed (semidry process) is no longer necessary, the length of the dry process rotary kiln can be reduced (Fig. 6.7-41a, b, and c) while, at the same time, the throughput capacity increases (Tab. 6.7-4). It was also found that the cyclones can be used to efficiently dry and preheat the raw feed (Fig. 6.7-42) further decreasing the length of the kiln (Fig. 6.7-41d). The latest development is to also transfer calcination of the raw materials into the external preheating section by installing a combustion chamber that boosts the temperature of hot

Fig. 6.7-40 Sketch of a chain curtain for the heat transfer from the kiln gases to the slurry in a wet process rotary cement kiln [B.69]

337

338


Fig. 6.7-41 Sketches of the relative sizes of different rotary cement kilns [B.69]

air coming from the clinker cooler, which is then injected into the system. This reduces the length of the rotary kiln further (Fig. 6.7-41e). Nevertheless, a cement manufacturing plant, including all the auxiliary equipment and processes is still a very large facility (Fig. 6.7-43). Generally, rotary kilns and sinter strands are suitable machines for the heat treatment and/or calcination of a great variety of building materials. Lime, although often produced in shaft furnaces, is calcined on sinter strands or with rotary kilns. Another example is the firing of ESCS (Section 6.7.1) for the production of lightweight aggregate. Still other applications are in the manufacture of artificial aggregate, both dense and light weight, from mineral wastes, crushed debris, and ashes, particularly fly ash

6.7 Building Materials and Ceramics Tab. 6.7-4 Comparison of diameter, length, and clinker production (throughput) of wet and dry process cement kilns [B.69]

Kiln diameter Kiln length Clinker

[m] [m] [tonnes/day]

Wet process

Dry process

5.0 165 1.050

4.0 70 2.300

(Section 8.2). The advantage of using a sinter strand for the agglomeration by heat is that the machine can be easily used for the manufacture of additives to building materials with different compositions and qualities (Fig. 6.7-44 [6.3.7.1]). For most shaped building materials, notably all types of bricks and preshaped structural parts, and for all ceramic articles, post-treatment, which almost always includes the application of heat, is an important final manufacturing step. During such treatment, the final properties and physical characteristics are produced. Tab. 6.7-5 lists some important characteristics of finished ceramic or building materials. Particularly in regard to modern high-performance ceramics, this list is not exhaustive. Also, some properties are more important for specific applications, for example for building materials, dimensions and strength, for technical ceramics, dimensions and density/por-

Fig. 6.7-42 Diagram of a dry-process rotary cement kiln employing cyclones for the collection of entrained solids and the drying and pre-heating of the raw materials [6.7.3.2]

339

340


Fig. 6.7-43 Artist’s impression of a complete dry-process rotary kiln cement manufacturing plant [6.7.3.2]. 1, primary crushing; 2, raw material storage; 3, raw meal grinding; 4, raw meal silos; 5, heat exchangers; 6, rotary kiln; 7, dust collection; 8, clinker cooler; 9, clinker storage; 10, coal

grinding; 11, coal storage; 12, main air blower; 13, cement grinding; 14, gypsum storage; 15, gypsum grinding; 16, cement silos; 17, discharge facility; 18, bagging; 19, office and laboratory; 20, power distribution and plant controls

Fig. 6.7-44 Sinter strand operating on the production of light-weight building material from fly ash [6.7.3.1]

6.7 Building Materials and Ceramics Tab. 6.7-5 *

Some important characteristics of finished ceramic or building materials

Dimensions Shape * Fit (free of distortion) Density / Porosity * Shock absorption * Sound and/or heat conductivity * Permeability Strength * Crushing / Tensile / Shear * Abrasion / Hardness / Wear * Impact resistance * Elasticity / Toughness * Freeze – Thaw * Thermal stress and/or shock resistance Quality * Surface * Color * Translucency * Electrical conductivity and other non mechanical properties *

*

*

*

osity, and for household ceramics (tableware) dimensions and certain quality indicators (e.g., surface finish, color, and translucency). Shaped building materials and ceramic parts are most often post-treated (hardened, calcined, sintered) in batch or continuously operated furnaces [B.97]. For certain building blocks and ceramics, such equipment may also include the use of special (inert or reactive) atmospheres. Fig. 6.7-45 shows the manufacturing process of (hollow) building blocks made from lime and sand. Milled burnt lime and sand are mixed and hydrated in a batch drum (10) with steam (for about 1 h). This reaction can be also carried out continuously in a silo with appropriate residence time. Blocks are formed from the moist mass in a “stone press” (punch-and-die with indexed rotating table, (14 in Fig. 6.7-45), stacked on carts, and hardened in closed chambers (17) with steam for 8–14 h at a pressure of 8–15 bar. During post-treatment, the final, permanent bonding is obtained and the building blocks are ready for use. More sophisticated and higher temperature batch furnaces include muffle, bell, elevator, or pit designs [B.97]. Tab. 6.7-6 lists the processes occurring during the sinTab. 6.7-6 * * * * * * *

Processes occurring during the sintering of most ceramic parts [B.13c]

Removal of water Removal of binder and organic media Dehydroxylation Oxidation Decomposition Phase transformation Cooling

341

342


Fig. 6.7-45 Diagram of the manufacturing process of (hollow) building blocks made from lime and sand. 1, lime ball mill; 2, elevator; 3, wind sifter; 4, 9, 11, silos; 5, metering bucket; 6, screw conveyor; 7, sand feed; 8, screen; 10, batch mixing and

hydration drum; 12, proportioning; 13, elevator; 14, stone press; 15, block cart; 16, transfer platform; 17, post-treatment (hardening) chambers; 18, steam production

tering of most ceramic parts. Ceramic ware is heated to a temperature between 700 and 2000 8C. Because raw ceramic parts are often “green” (moist), removal of water is the first post-treatment step. Following or simultaneously and immediately preceding firing, binders and plasticizers, which have provided the properties needed for forming, are also removed. The amount of residual moisture and/or additives that can be tolerated in the part when firing begins depends on its shape and structure and on the heating rate of the furnace. If the part is made by dry pressing, removal of additives can be incorporated into the heat-up stage of the sintering furnace if the time for this process is not too long. However, since additives are often cellulose, wax, or starch type products, they can be more conveniently decomposed by oxidation in air at low temperature prior to loading the parts. In the furnace, clay minerals usually dehydroxylize at 500–600 8C whereby steam is produced. The loss of strength that occurs at this stage may result in cracking. The next step, oxidation, can be accomplished by holding the temperature at a certain level, which varies with the manufacturing method that was used for and the type of the body but is often about 900 8C where a decomposition of carbonates and sulfates may produce bloating in vitrified parts. Silica, which exists in many different crystalline forms, is an important constituent of most ceramics. The conversion from one form into another is accompanied by sometimes large volume changes. Because this occurs during heating and cooling, the rate of temperature change must be considered and may have to be adjusted to avoid deformation and/or cracking.


Most ceramic products are fired in air: under oxidizing conditions. The ideal kiln for the firing of ceramics is capable of heating and cooling the parts uniformly at the maximum rate of temperature change for each of the stages mentioned in Tab. 6.7-6. In the ceramics industry, muffle furnaces are typically used for batch sintering [B.97]. For high-quality wares, temperature control is very important to avoid the above mentioned problems in different processing stages. This can be accomplished easiest and most accurately in batch furnaces. Some materials must be, at least during certain stages, fired in reducing or other special atmosphere that is also best provided in batch kilns. However, many bulk ceramic products need to be of such low cost that continuous furnaces must be used, which operate more economically. In conclusion, batch sintering furnaces are used for: * *

*

low-output production, special duties (because there are no moving parts, batch furnaces can be designed for higher temperatures; furthermore, since it is possible to seal the interior more effectively, purer atmospheres can be realized and maintained), and experimental work.

For the first two of these, manual pusher furnaces (Fig. 6.7-46) can be applied in which parts on a tray are moved through a furnace, one tray at a time [B.97]. If it features gas tight interlocks and/or doors for charging and discharging, it can be also used for processes in which the atmosphere must be well controlled. The bell furnace is widely used for parts requiring long sintering cycles [B.97]. Typical equipment consists of one or more supporting bases with removable sealed retorts, used to cover the loads and retain the protective atmosphere around them throughout the entire heating and cooling cycles, a portable heating bell, a stand-by base, a hoist, and an optional cooling bell. The elevator furnace [B.97] is useful for sintering heavy and/or bulky loads. It has an elevated heating chamber with open bottom in a fixed position, a mechanism for raising and lowering load supporting cars into and out of the furnace, a stand-by car to

Fig. 6.7-46

Sketch of a manual pusher furnace [B.28, B.97]

343

344


plug the kiln opening during idling periods, and optional cooling chambers. It is also applicable if protective atmospheres of exceptionally high purity are required. Flexible hoses carry atmosphere gas and cooling water to and from the cars. Since batch furnaces can be small and controlled easily, they are also commonly used for development work. The results from small-scale testing can be directly transferred to larger or even continuously operating kilns. As mentioned above, sintering of ceramic wares normally occurs in oxidizing atmosphere and without a special gas environment. Therefore, continuous sintering furnaces are often directly flame heated. Fig. 6.7-47 is the side elevation of a tunnel furnace for the firing of ceramic parts, indicating the direction of movement of material (car or mesh belt) and gas and the process zones. The diagram below depicts the temperature profile over the length of the furnace and shows that temperature control is normally quite unpretentious. Most tunnel kilns for ceramics are of the car type. Cars can carry more weight, are more rugged, and are more reliable than belts and other continuous methods of movement. Operation of modern furnaces is computer controlled and continuous movement is accomplished with automatic loading and unloading systems (Fig. 6.7-48). The pusher furnace that was shown schematically as manually operated equipment in Fig. 6.7-46 can be mechanized and then becomes a continuous kiln. The most commonly used continuous furnace for the sintering of small, light parts is the mesh-belt sintering furnace. A variation of the straight belt arrangement is the hump-back kiln, which is used where high purity of the atmosphere in the sintering zone is required [B.97]. Another sintering furnace used in the ceramic industry is the roller hearth furnace in which loaded trays are conveyed through the kiln by riding on individually driven rolls [B.97].

Fig. 6.7-47 Side elevation (diagram) of a directly flame-heated tunnel kiln for the firing of ceramic parts and typical temperature profile [B.13c]


Fig. 6.7-48 Modern tunnel kilns for the firing of sanitary parts, tableware, and fine china, all with fully automated car systems, and of a completely

automated handling area for loading and unloading (courtesy Eisenmann, B€ oblingen, Germany)

345

346


As mentioned above, the use of sintering for building and ceramic products (bricks, structural parts, pots, vases) is quite old. Through the centuries, even though empirically improved, it was exclusively carried out in batch (periodic) kilns. Even today, large pieces and all other types of ceramics are still frequently fired in modern periodic hearth, shuttle, elevator, or bell kilns (Fig. 6.7-49). Continuous heat treatment of pre-agglomerated parts is less than 150 years old. Fig. 6.7-50 shows a continuous kiln for drying and firing load-bearing structural bricks. The most recent development is hot isostatic pressing (Section 6.7.2) in which forming of the part(s) from dry particulate raw materials and heat treatment are combined in one piece of equipment [B.13a, B.97]. This technology is especially used for the manufacture of high-performance ceramics and of composites containing ceramic and other (e.g., metal) components.

Fig. 6.7-49 A large periodic shuttle kiln for the firing of chimney flues (courtesy Eisenmann, B€ oblingen, Germany)

Fig. 6.7-50 A continuous kiln for the drying and firing of load-bearing structural bricks (courtesy Eisenmann, B€ oblingen, Germany)

6.8 Applications in the Mining Industry (Minerals and Ores)

6.8

Applications in the Mining Industry (Minerals and Ores)

With exception of the extraction of mud and sand from the ground for building materials and clays for the same purpose as well as for ceramics (Section 6.7), in human history mining began with the search for and recovery of pure metals and minerals. Among the first minerals that were mined was coal (Section 6.10), the first metals that were found and used were gold and copper, the first decorative materials that were used together with the metals for making jewelry were precious and semi-precious stones, and the first nutrition-related material was salt (Sections 6.3.2 and 6.4.2). Metallic (telluric) iron was found much less frequently. It was either naturally reduced iron of volcanic origin or meteoric iron; however, man recognized that, as compared with bronze tools and weapons, it provided superior quality when forged into shape. After learning that iron can be produced from ores by relatively simple heating with charcoal in shallow cavities that were dug into the earth, early industrialization centers developed quickly. About 1400 BC, particularly in the Near East and in China, monopolies developed from which iron was distributed and later its manufacture was spread to other cultural areas. Until very recently, while the production of iron developed away from the primitive forms and casting of iron and the blast furnace were introduced after approximately the 14th century, the raw material was still high-quality lump ore. The sintering (Section 6.8.3) of high-quality ore fines into lumps was invented several hundred years ago and first applied mostly to enlarge the size of non-iron ores prior to smelting. Later this technology was also used for high-grade iron ore fines. Beginning during the first part of the 20th century, as a response to a higher demand for iron and steel, the capacity of blast furnaces was increased by improving the burden composition with sized and fluxed sinter. After World War II a great backlog of industrial demand, including the production of steel, triggered the interest in abundantly available low-grade ores, particularly the Taconites of North America. Although the concentration of ores, based on density differences between heavy and light mineral components, had been known for some time, the separation size of these lower grade ores, that is the particle size to which the raw material must be ground to effectively accomplish separation into concentrate and tailings, is so small (typically < 325 mesh or < 44 lm) that the enriched product can not be used directly and even sintering is not applicable (Section 6.8.3). To overcome this problem, size enlargement by agglomeration must be used. During the earlier development of an optimal burden for the blast furnace, consisting of lump ore, crushed sinter, coke, and fluxing additives, it was found that optimal particle size ranges should be 5–50 or, better, 5–30 mm [B.18]. Particularly for the former size range, which is typical when high-grade lump ores are used and, at that time, was the more common requirement, briquetting was the available and accepted agglomeration method (Section 6.8.2). Although the metallurgical behavior of briquettes in smelting furnaces is very good, iron ore briquetting was too expensive (mostly because of high

347

348


wear) and the production capacity of briquetting units too small when compared with the anticipated large quantities of concentrates to be agglomerated. Therefore, a new method had to be developed. Since the particle size of the concentrate is small enough for growth agglomeration, discs, drums, and, originally also cones were used for the production of green (wet) spherical balls, which were hardened by heat in special sintering furnaces (Section 6.8.1). It was found that the large tumble/growth (called “balling”) equipment was best suited to producing spherical agglomerates (called “balls”) with a size range of 9– 15 mm and, according to specification, does not include any particles below 5 or above 25 mm. Even after sintering, this product contains pores that are accessible from the pellet surface (open porosity), so it is easier to reduce than lump ore. Furthermore, additives can be mixed with the fine ore feed to yield a self-fluxing product. These characteristics, together with the development of an excellent burden structure, resulting in increased capacity and better control of blast furnaces, made iron ore pellets the preferred feed material for the blast furnace and, later, direct reduction plants (Section 6.8.1). Therefore it became an accepted practice to grind even high-grade ores to the required fineness and produce pellets, with or without the addition of flux. Also during the first part of the 20th century, other finely divided ores and minerals, often recovered during the mining operation, were first enriched and then upgraded by agglomeration (granulation or briquetting, Section 6.8.2) for many different uses. A major application is the treatment of by-product sodium chloride from the processing of potassium chloride (Potash) for fertilizer and chemical applications (Sections 6.3.2, 6.6.2, and 6.8.2). Fines from rock salt mining and sea salt manufacture are also agglomerated for many applications (Section 6.8.2). Tab. 6.8-1 lists ores and minerals that have been and/or are being agglomerated for a multitude of purposes. Further Reading

For further reading the following books are recommended: B.3, B.4, B.7, B.8, B.10, B.11, B.15, B.16, B.18, B.21, B.22, B.26, B.31, B.35, B.40, B.48, B.55, B.56, B.89, B.94, B.98 (Chapter 13.1). Books mainly devoted to the subject matter are printed bold. Tab. 6.8-1 Alphabetical listing of minerals and ores that have been and/or are being agglomerated Alumina Cement Raw Mix Clays Dolomite Graphite Lead Ore Magnesia Lateritic Nickel Ore Potassium Sulfate Salts Zinc Ore

Ammonium Sulfate Chalk Cobalt Ore Fluorspar Gypsum Lime Manganese Ore Phosphate Rock Rare Earth Oxides Soda Ash (heavy and light) Zirconium Ore

Bauxite Chromium Ore Copper Ore Glass Batch Iron Ores Limestone Niobium Ore Potassium Chloride Rock Salt Sodium Chloride Zeolite


6.8.1

Tumble/Growth Technologies Pelletization of Iron Ores By far the largest application of tumble/growth agglomeration in the mining industry is the pelletization of iron ores. Iron ore pelletization was developed around the middle of the 20th century in North America (particularly by the US Bureau of Mines) to make very fine (< 325 mesh or < 44 lm) Taconite concentrates suitable for use in the blast furnace. In Sweden the process was developed to agglomerate magnetite fines for the same purpose [B.8, paper 36]. Strongly metamorphic, originally sedimentary iron ores, similar to the Taconites, are found elsewhere on earth: for example, Itabirites in Brazil and similar ones in other major mining districts. Therefore, after early successes of the newly developed technology in the 1950s in the USA, many large plants were built worldwide during the following 40 years. In the USA, a production capacity of 15 million t/y was reached in 1960. Worldwide, as shown in Fig. 6.8-1 [B.48], a capacity of 50 million t/y was available in about 1963, and 200 million t/y in about 1977. After a maximum of over 300 million t/y, according to the statistics of UNCTAD [6.8.1.1], 232 million t/y of iron ore pellets were produced in 2001. This is 11.7 % (30.7 million t/y) down from 2000. The pellets are manufactured in 20 countries where the available total production capability is still about 300 million t/y. The lower actual capacity reflects the reduced volume of world steel making and the growing importance of alternative iron units (Sections 6.9 and 8.2). Excluding product that is consumed in the producer’s own semi- or fully integrated steel plants (particularly in Japan) or by domestic users, iron ore pellets are exported to foreign steel making 6.8.1.1

Fig. 6.8-1 Graph showing the development of worldwide iron ore pelletizing capacity during the first 40 years (1950–1990) and projections at that time for further growth [B.48]

349

350


facilities by 10 countries (Brazil, Sweden, Canada, USA, Chile, Peru, Venezuela, Australia, India, and Bahrain). Much of the above mentioned decline has happened in the USA, although Minntac, which began operation in 1967, has become the single largest processor in the world with an annual production capability of 16.4 million tons of pellets after several expansions and recent modernization efforts, lasting for 6 years [6.8.1.2]. In October 2002 the cumulative pellet output from Minntac had reached 2 billion tons. Because of the remarkable uniformity of size, shape (Fig. 6.8-2), and (adjustable) composition of iron ore pellets, major improvements in blast furnace operation were achieved. They include better handling characteristics of the ore feed and increased production capacity per blast furnace, which, to a large extent, is due to a more uniform burden structure, the higher reducibility of the furnace charge, and, more recently, the “engineering” of the pellets’ composition, mostly to decrease the contents of gangue and obtain self-fluxing characteristics. This has led to modified purchasing specifications of blast furnace operators who increasingly require the supply of pellets over direct shipping lump ore, in spite of their somewhat higher price. As a result, pelletizing plants have been built in several areas, processing high-grade ores that would not normally necessitate concentration, and converting them, after fine grinding and formulating a preferred metallurgical composition by mixing, into superior quality pellets. When direct reduction (DR) was developed as an alternative to the blast furnace route to steel, most of the commercially successful technologies were and stillarebasedontheuseofpelletsasorefeed.Sincedirectreducediron(DRI,Section6.9.2) is a charge material for electric arc furnaces, DR grade pellets are now produced, which are specially formulated to have a high total iron content (low gangue). An increase in blast furnace capacity and a reduction in coke requirements resulted from the first agglomerated iron ore charges used by Gr€ ondal in Finland in 1899. For that process, briquetting was used, the only large-scale agglomeration technique available at the time (Section 6.8.2). As the benefits of a sized, agglomerated feed became recognized, many other agglomeration techniques were patented in the early 20th century, of which only the one that Andersson originated in Sweden was of importance. He had the idea of agglomerating fine iron ore in a drum, even the use of a binder to strengthen the green balls, and subsequently drying and firing the product. However, Andersson’s patent was never used commercially and his work was soon forgotten [B.8, paper 36]. Independently, in 1930 a pilot plant was constructed in Rheinhausen, Germany, to use the patent by Brackelsberg in which so-

Fig. 6.8-2

High-quality iron ore pellets (courtesy CVRD, Vitoria, Espirito Santo, Brazil)


dium silicate was employed as a binder and the green balls were hardened at low temperature. The plant was unable to produce an acceptable product, was closed, and also forgotten. The methods that finally were to develop into the universally accepted iron ore pelletizing technology at the US Bureau of Mines, the University of Minnesota and in Sweden were first described by Firth (USA) and Tigerschiold/Ilmoni (Sweden) in the 1940s. Much additional information about the early research can be gleaned from a paper by Goldstick [B.8, paper 36]. The agglomeration of iron ore uses a two-stage process. Such processes are characterized by the production of discrete agglomerates by size enlargement from the particulate feed with or without binders in a first stage and the development of final strength in a second stage. For the first (size enlargement) stage, many of the known tumble/growth and pressure agglomeration techniques can be used and for the second (hardening) stage, new methods had to be developed. Although at least three major low temperature or “cold” pellet hardening processes were invented (Fig. 6.8-3) [B.3, B.18], the application of heat (sintering) became the dominant technology. Depending on the most efficient method of heating for the specific purpose, different types of furnaces are applied. In modern iron ore pelletizing, nearly spherical pellets are produced by tumble/ growth agglomeration in the first stage. Since high throughput capacities must be handled, large drums and discs (pans) are used (Fig. 6.8-4). Although the agglomeration pan, due to a natural segregation effect, has the advantage of directly producing narrowly sized agglomerates [B.48, B.97], for most installations drums are applied. In spite of the fact that the agglomeration circuit (Fig. 6.8-5) must be able to handle several hundred percent (typically 300–500 %) of the production capacity and, therefore,

Fig. 6.8-3

Pellet hardening methods without firing [B.18]

351

352


Fig. 6.8-4 Tumble/growth agglomeration equipment for the balling of iron ores. (Left) Diagram of: a) the disc (pan), b) the cone, c) the drum (including recirculation circuit). (Right) Actual equipment (c: drum only) (courtesy Feeco, Green Bay, WI, USA (a and c), file photo, Kennedy Van Saun, New York, NY, USA (b))


Fig. 6.8-5

Typical balling drum circuit for the agglomeration of fine iron ores [B.48]

needs to be much oversized, the considerably simpler operation and control of drum agglomerators is responsible for this preference. In some early installations balling cones were used (Fig. 6.8-4b). It was said that these “deep pans” produced the same segregation pattern as inclined shallow discs but, because of the larger “hold-up” in the machine, additional strengthening of the green pellets occurs due to overburden pressure and a longer retention time. However, in reality some of the defined pattern of charge movement, which is typical of the shallow discs with cylindrical rim (Fig. 6.8-4a), is sacrificed and operation of balling cones is not easily controlled. Therefore, this design is no longer used and is, for all practical purposes, forgotten. Balling drums (Fig. 6.8-4c) offer the advantages that very large throughput capacities can be handled in a single unit and operation is simple. They consist of a cylindrical tube, normally made of steel with or without a variety of liners, and are installed with a slight slope towards the discharge end (typically 108 from the horizontal). A retaining ring is often fitted to the feed end of the drum to avoid spill back. Depending on the slope and diameter of the drum and the physical characteristics and moisture content of the charge, the rotational speed of the equipment must be adjusted such that the particle bed begins to separate from the wall at approximately the 3–4 o’clock position in a counter clockwise rotating drum (Fig. 6.8-6). The mass then cascades down the inclined bed and liquid binder, if applicable, is sprayed onto the surface with nozzles from a manifold extending into the first part of the drum.

353

354

6 Industrial Applications of Size Enlargement by Agglomeration Fig. 6.8-6 Diagram of the optimum charge movement in a balling drum [B.18]

Within the kidney shaped tumbling mass of solids, production of seeds and growth of agglomerates takes place. Although some natural segregation by size occurs in the bed and even if a retaining ring would be provided at the outlet of the drum, the product consists of a wide distribution of agglomerate sizes because the entire charge moves forward in a plug flow fashion. Normally, in iron ore pelletizing no retaining ring is provided downstream and the discharge end features a scroll (spiral extension, Fig. 6.8-7) that serves to distribute the pellets over the width of a screen. As shown in the sketch of an agglomeration circuit (Fig. 6.8-5), fine ore (concentrate), dry binder (see below), and (fluxing) additives are mixed (fluffer) with recycled undersized material and deposited into the feed end of the drum. Although the charge material is normally moist and may contain enough binder (water and dry binder) to accomplish balling, a manifold with spray nozzles is installed in the first about 1/3 of the drum to initiate and control agglomerate growth if the tumbling mass is too dry. A sufficient number of nuclei, some of which are also supplied with the recyclate, must

Fig. 6.8-7 The spiral discharge of a large balling drum for iron ore (file photo, McKee [Section 13.3, Ref. 16])


be produced in the drum at all times to replace the pellets that are removed from the circuit, and the growth rate per pass must be such that the required production rate of green balls is consistently maintained. To avoid surging, the rate of pellet production must be stabilized and the balance between material in the drum and recycling rate must be kept constant. This may necessitate installation of a surge bin in the closed recycle loop from which a metered amount of undersized returns is fed into the drum. Originally, many drums were coated on the inside with cement or expanded metal to encourage build-up of material as an autogenous wear liner. Different designs of scrapers [B.48], called cutter bar in Fig. 6.8-5, are then installed to control its thickness. More modern installations use a special rubber coating, which prevents buildup and still protects the drum from excessive wear. The green (moist) balls produced in iron ore pelletizing are rather weak and require gentle handling before they reach final strength during sintering in the second stage. Because of this, and since vibrating screens tend to clog, roller screens (Fig. 6.8-8) were

Fig. 6.8-8 a) Diagram of the principle of a roller screen in iron ore pelletizing; b) the design, and c) operation of such screens [B.97]

355

356


developed [B.97]. In most cases, the rotation of the often individually driven rollers (Fig. 6.8-8b) is against the gravitational flow of material on the downward sloping machine deck. This induces a rolling movement of the pellets (Fig. 6.8-8a, upper left) causing additional cleaning and rounding. The final, typically very high strength of the iron ore pellets is obtained in the second process stage by the development of sinter bridges between the ore particles at high temperatures, the so-called sintering temperature [B.48, B.97]. During the post-treatment, the green iron ore pellets must be first dried and preheated before hardening by sintering occurs. The problem of this sequence of events is that the original binding mechanism, caused by capillary forces and the surface tension of the liquid, has disappeared after drying and sintering, which requires about two-thirds of the ore’s softening temperature, has not yet begun. Therefore, there is a time interval during which the dry pellets have almost no strength. To partially overcome this problem and increase the chance of survival, binders that retain some bonding characteristics in the dry state, are added during tumble/growth agglomeration. The traditional additive for this purpose is bentonite, a natural montmorillonite clay (Section 6.7.1). Originally, when the binder was mixed with the feed to the tumble/growth agglomerator rather inefficiently with a simple fluffer (a rotating mixing tool crudely turning-over the material on a belt conveyor, see Fig. 6.8-5), 2 % (and sometimes more) of bentonite were added to obtain an acceptable effect. However, from a metallurgical point of view, this additive is a detrimental constituent as it increases the acid gangue portion in the product. Fig. 6.8-9 depicts this effect of bentonite. Particularly in highly concentrated ores (lowest line), which have been upgraded at high cost, the increase in acid gangue components may be 100 % thus canceling some of the ore improvements. Today, bentonite is often blended into the ore concentrate with a highly efficient mixer. Fig. 6.8-10 shows the influence of a good binder mix on green and dry pellet compression strengths [B.18]. It shows that the green strength is almost unaffected. Nevertheless, the material has a beneficial effect in the wet stage as the colloidal nature of the clay imparts a certain plasticity to the pellets, which makes them less friable and better formable (for example during rounding on a roller screen, see above and Fig. 6.8-8a). During the final stage of drying, the suspended ultrafine clay moves with the retreating liquid to the coordination points between the ore particles and develops solid bridges (Chapter 3, Tab. 3.1, item I-6b). Depending on the amount of bentonite added, a considerable increase in dry strength is obtained. Since a compressive dry strength per pellet of 20 N is normally considered sufficient and because the contamination level of the product should be kept as low as possible, in most plants using high-efficiency mixing, the bentonite rate is kept at < 0.7 % and often below 0.5 %. Even with these low amounts of binder addition a certain percentage of the ore concentration effort is reversed. Therefore, numerous alternative materials were proposed and tried to produce a higher quality pellet. The most effective additives in this respect are organic materials (for iron ore concentrates, for example Peridur [B.97]), which do retain strength in the dry and preheating stages but burn-out (or otherwise disappear) at high (sintering) temperatures and do not leave binder-related contaminants behind. In the end, their application is defined by an economical evaluation [B.97].

6.8 Applications in the Mining Industry (Minerals and Ores) Fig. 6.8-9 Influence of bentonite addition on acid gangue components in iron-ore pellets [B.18]

Fig. 6.8-10 The influence of the amount of bentonite in a well-mixed iron ore concentrate feed on green and dry pellet compression strengths [B.18]

357

358

6 Industrial Applications of Size Enlargement by Agglomeration Fig. 6.8-11 Diagrams of the three major furnace types used for the hardening of iron ore pellets: a) shaft, b) traveling grate, c) grate-kiln; D, Drying; F, firing (sintering); C, cooling

Based on development work in the 19th century, in the early 20th century, modern thermal treatment units had been developed for the drying, heating, and cooling of minerals, particularly for the manufacture of cement (Section 6.7.3) and the sintering of metal ores (Section 6.8.3). These were converted for application as the second stage of the newly developed iron ore pelletization process. Three furnace types evolved as the dominant thermal treatment methods (Fig. 6.8-11): the shaft furnace, the traveling grate, and the grate-kiln [Section 13.3, Ref. 16]. Shaft furnaces had been used for several centuries for the smelting of ores and the burning of lime. Because of their high thermal efficiency and ease of operation they were the first, both in Sweden and the USA, to be adapted for iron ore pelletizing. In North America the first pilot plants were installed in Aurora and Babbitt, which started-up between 1948 and 1952. The first commercial plant was built for Erie Mining Co., Hoyt Lake, Minnesota, beginning in 1954 and began operation in 1955. Green pellets are fed into a narrow, slender shaft, which was modified to a rectangular (from the originally round) shape and is heated by burning oil or gas in chambers that are arranged on its sides. Since the sintering zone is located only about 50 cm below the top of the pellet bed and hot gas and solids move countercurrently, the green agglomerates are quickly dried and preheated. Therefore, they must be resistant to thermal shock. Also, because ore sintering and softening temperatures are close


together, the charge tends to form lumps, which disrupt the uniform flow of solids and gas in the furnace and must be broken at the bottom of the hot zone (Fig. 6.8-12). In the first shaft furnaces for the hardening of iron ore pellets, a solid fuel was added to the charge to supply additional thermal energy. Later, shaft furnaces were almost exclusively used for the treatment of pellets that were produced from Magnetite, as the exothermic oxidation reduces the heat requirements from the outside and simplifies furnace control. Additional improvements were obtained after the long shaft furnace with internal cooling (Fig. 6.8-12a) was converted to a medium shaft height with external coolers and recovery of sensible heat (Fig. 6.8-12b). Nevertheless, the shaft furnace has lost its importance. Tab. 6.8-3 summarizes the advantages and disadvantages. While at the beginning and still to date for some facilities it is an advantage that smaller units, particularly if they process Magnetite pellets, are economically feasible, in today’s environment, where iron ore pelletizing plants produce several million tons per year in one unit (below), this is considered a disadvantage. Because pellets are prone to spalling by thermal shock and easily abrade in the constantly downward moving charge they must be of the highest quality, which normally means a high amount of binder addition (particularly bentonite). The biggest disadvantage of shaft furnaces is, however, that because every process component takes place in close proximity in one unit, there is little chance of influencing the process stages, which translates into low flexibility. Even small disturbances in charge consistency (e.g., due to lumping), which are unavoidable, results in disrupted gas and charge flow patterns and inconsistent product quality. At the time when iron ore pelletizing was first developed, simple traveling grates had already been used for ore sintering. They are the oldest machines for the production of agglomerates from fine grained ores (Section 6.8.3). As shown in Fig. 6.8-13, the early traveling grate machines combined all process steps and little control was possible. The feed, consisting of a blend of fine ore and solid fuel (coal), is first placed as uniformly as possible onto an endless perforated belt, made-up of connected and hinged cast iron or steel sections, and then passes through an ignition furnace, which is a short hood with burners inside. The flames impinge on the surface of the bed and ignite solid fuel that is close to the hot interface. From that point on, air is pulled through the completely open bed, in a downdraft fashion; combustion of the solid Tab. 6.8-3 Advantages and disadvantages of shaft furnaces for the induration of iron ore pellets Advantages

Disadvantages

Simple design Small number of moving parts Refractory lining of entire furnace Intensive heat exchange Feasible for low production capacity Best for Magnetite pellets

Little chance of influencing process stages Low flexibility Charge is in constant movement Pellets must be of highest quality Limited production capacity High fuel consumption Commonly disrupted flow pattern (lumping) Inconsistent product quality

359

360


Fig. 6.8-12 Details of two shaft furnace designs: a) long shaft furnace with internal cooling, b) medium shaft furnace with external cooling and two alternative methods of sensible heat recovery [B.18]


Fig. 6.8-13

Diagram of an early traveling grate sintering machine [B.97]

fuel is sustained and enhanced by providing oxygen from the air and the burning hot zone moves downward through the bed. The amount of air that is pulled through the entire length of the bed and the speed of the grate are adjusted such that, at the end of the machine, the fuel has disappeared and the entire charge has sintered together. The porous, solidified, still hot mass is broken in a sinter breaker, the resulting pieces are screened into the desired particle size distribution, and the product is cooled externally. Solid particles that are entrained in the combustion air, settle in bins, which are part of the main suction duct and fine dust is removed in a dust collector. These solids are transported to the burden preparation plant and ultimately fed back to the sintering machine. Fig. 6.8-13 already includes two improvements. First, oversized pieces from the sinter crusher are recrushed in closed loop with a screen (not shown) and all screen fines are recirculated to provide a hearth layer, which protects the grate from excessive temperatures. Second, to achieve uniform bed depth across the grate, feed is placed onto the insulating hearth layer with a swinging conveyor and leveled with a roll feeder. This traveling grate machine was adapted for use in iron ore pelletizing plants in US test facilities at Carrollville and Babbitt between 1952 and 1954 [Section 13.3, Ref. 16]. The first commercial plants were started-up at Reserve Mining Co., Silver Bay, Minnesota, USA, in October 1955, at International Nickel Co. of Canada Ltd., Copper Cliffs, Ontario, Canada, in February 1956, and at Cleveland-Cliffs Iron Co., Ishpeming, Michigan, USA, in October 1956 [B.18]. The major modifications, which were later further optimized, comprised the separation of the process stages of drying, pre-heating, firing, and cooling, including recuperation of sensible heat. This resulted in much better control and improved the economics of the process. As shown in Fig. 6.8-14, the traveling grate is now completely enclosed and the housing is subdivided to produce the different process conditions. The belt consists of pallet cars with grate bottoms and side walls, connected with each other by hinges to form a continuous unit (Fig. 6.8-15a). To protect the pallet walls, recirculating fired pellets are deposited in the hearth and side layers in which the green pellet bed rests [B.18].

361

362


Fig. 6.8-14 Principles of two traveling grate hardening systems for iron ore pellets [B.18]: a) McKee design; b) Lurgi–Dravo design

The hoods and windboxes depicted in Fig. 6.8-14 are easily and effectively isolated from each other. While the original traveling grate was applying only a downdraft flow of process gas through the particle bed, the new, improved design employs both upand downdraft sections to optimize the efficiency of thermal energy transfer and use. Also, the originally necessary addition of solid fuel was abandoned in favor of heating with burners. Different ores require distinct heating patterns. Their behavior is determined in laboratory pot grate furnaces [B.97] and pilot plant installations. While the process sections of a traveling grate installation can be designed for a particular ore, an advantage of this machine is that by merely changing the speed of the pallet belt an existing hardening facility can be adjusted to meet the requirements of changing ore compositions. As shown in Tab. 6.8-4 the total duration of the process may change between about 33 min for magnetitic ores and 41 min if the raw material contains limonite and hematite.

6.8 Applications in the Mining Industry (Minerals and Ores) Tab. 6.8-4 Ranges of thermal treatment zones for different ore compositions (Figure 6.8-14b) [B.18] Thermal treatment

Time [min] (range)

Updraft drying Downdraft drying Preheating (including dehydration, calcination and oxidation) Firing (sintering) After firing Cooling Total duration arithmetric

4–7 2–5 7–9

% of overall process (range) 12 – 17 6 – 12 17 – 22

7 – 11 17 – 28 3–5 8 – 12 12 – 14 30 – 35 35 – 51 (actual 33 – 41)

Fig. 6.8-15 a) Side and hearth layer as well as green pellet feeding of a Lurgi-Dravo traveling grate machine. b) Sketch and thermal insulating effect of the hearth and side layers [B.18]

363

364


The traveling grate, which may be executed straight or circular [Section 13.3, Ref. 16, B.48, B.97], is the most commonly used hardening machine for iron ore pellets throughout the world. In most cases the straight grate design is employed. Advantages are its easy adaptation to operation with practically all ore types, the fact that the pellets remain stationary throughout the process, thus allowing the processing of agglomerates with lower strength (requiring less binder), the production of pellets with uniform quality (because of the processing in a relatively thin layer), the possibility of obtaining high energy efficiency (including effective heat recuperation), the potential use of many different fuels for firing, and, last but not least, its applicability for large capacities, which further increases the economy of the system (below). Still better process control can be achieved if the three components (drying/preheating, firing, and cooling) can not only be influenced separately and individually but are physically detached equipment units. This is the case in the so called grate-kiln that was developed and optimized by Allis-Chalmers Mfg. Co., Milwaukee, Wisconsin, USA, for the production of cement clinker (Section 6.7.3). Work to test the feasibility of the grate-kiln system for the hardening of iron ore pellets began at the Allis-Chalmers pilot plant at Carrollville near Milwaukee, Wisconsin, USA, in the 1950s. The first commercial plant started-up in the mid 1960s at the Humboldt Mining Co., in Michigan, USA [Section 13.3, Ref. 16]. Fig. 6.8-16 is a diagram of a grate-kiln system for the hardening of iron ore pellets [B.18]. The traveling grate, consisting of an endless chain of grate plates, is used for drying, preheating, and oxidation of magnetite. The green balls are charged directly onto the grate plates without hearth layer and remain in a relative position of rest. For drying and pretreatment, hot waste gases from the rotary kiln and sometimes also from the cooler pass through them in different directions and zones. Use of the tra-

Fig. 6.8-16 Sketch of the grate-kiln hardening systems for iron ore pellets [B.18]


veling grate with its various functions is imperative for the successful operation of the system. For thermal, operational, and quality reasons, the rotary kiln alone is not sufficient to do the task. During transfer from the preheating grate to the rotary kiln, special care must be taken to avoid breakage of the still weak pellets by installing special (often spiral) chutes. In the rotary kiln, which is equipped with refractory lining and fired by oil, gas, or, to a certain extent, pulverized coal, thermal energy to achieve uniform heating and sintering of the pellets is mainly provided by radiation. To reach the required hardening temperature of 1320–1340 8C, the wall temperature must be higher, which may lead to reactions between oxide dust and the hot lining, causing patchy or ringshaped accretions. This is one of the main problems that can occur in the rotary kiln and must be avoided. The pellets roll spirally in a very thin layer towards the discharge end where they leave the kiln as hardened hot product. Subsequently, they pass into the third unit, a circular or annular cooler. Having three different process-specific units allows excellent adaptation of the system to a specific ore or mixture of ores. Residence times and heat supply in each part can be adjusted to achieve optimum results in regard to power consumption, thermal efficiency, and pellet quality. Particularly in regard to the latter, it is often said that the roundness and surface quality are higher after rolling the pellets in the kiln at high temperature. This translates into higher abrasion resistance and less production of nuisance dust during shipping and handling. As in the case of the traveling grate, large capacity grate-kiln systems are most economical. Fig. 6.8-17 shows isometric drawings of the three major iron ore pelletizing processes. The three small shaft furnaces in Fig. 6.8-17a are each fed from a balling drum circuit, while the straight traveling grate in Fig. 6.8-17b, which always handles a much larger throughput, receives its feed from three balling drum circuits. To give an impression of the size of such systems, Fig. 6.8-18 is a partial view into the drum agglomeration section of an iron ore pelletizing plant showing five drums and associated process equipment. In Fig. 6.8-17c, depicting a plant using the grate-kiln hardening process, it is indicated that either balling pans or closed loop balling drums may be used for the production of green pellets. Of course, complete iron ore pelletization plants also include the feed preparation. As discussed in Section 6.8 and at the beginning of this Section, iron ore pelletization was originally developed for the size enlargement of ore concentrates which, after beneficiation (upgrading by removal of gangue), were too fine for direct use in the blast furnace or other smelting equipment. After crushing and fine grinding to liberate the high-grade ore particles in the raw, as mined material, separating the iron oxide from the gangue by mechanical (jigging, spiral) or chemical (flotation) sink/float processes, and dewatering, the resulting ore concentrate is fine enough for growth agglomeration by one of the balling processes (Fig. 6.8-4). The simplified flow diagram of a complete plant, which includes crushing to the required fineness for growth agglomeration of concentrate produced at the mine site and feed preparation, is shown in Fig. 6.8-19. Again to give an impression of the size of such plants, Fig. 6.8-20 shows views into the fine grinding bays of multi-million ton per year iron ore pelletizing plants using semi-autogenous mills (a) or rod and ball mills (b) for size reduction.

365

366

6 Industrial Applications of Size Enlargement by Agglomeration Fig. 6.8-17 Isometric drawings of the three major iron ore pelletizing processes [B.48, B.97]. a) Balling drum circuits with shaft furnaces. b) Balling drum circuits with straight travelling grate machine. c) Balling drum circuit(s) or balling pan(s) with grate-kiln hardening system


Fig. 6.8-18 Partial view into the drum agglomeration section of on iron ore pelletizing plant showing five drums and associated process equipment

Depending on the operating philosophy, both iron ore concentration and pelletizing can be combined in a single facility or in two separate plants. For example, Fig. 6.8-21a is the aerial photograph of National Steel Pellet Company’s Keewatin, Minnesota, USA, installation on the Mesabi (Taconite) iron range [6.8.1.3]. In this plant with a rated production capacity of 5.5 million t/y, iron ore is concentrated and 1 % limestone pellets are produced for exclusive consumption by the US National Steel Corporation. Similar plants in different parts of the world produce pellets from iron ore, mined and upgraded on site, for export to the free market. Other mines concentrate the iron ore on site and ship so called pellet fines to pelletizing plants that are owned by the same group or by others, Fig. 6.8-21b is the aerial photograph of such an installation [6.8.1.3]. The picture shows the pelletizing facility of Quebec Cartier Iron Co. at Port Cartier, Quebec, Canada. The ore is mined and beneficiated at the Mt. Wright mining area in Fermont, Quebec, Canada. Similar installations are located at import harbors, particularly in Japan, where pellet fines from, for example Australia, are received, potentially reground (below), and pelletized for use in nearby steel mills. The improvements in blast furnace operation from the use of pelletized and sometimes additionally engineered (i.e., self-fluxing) iron ore burdens has been so great that naturally high-grade iron ores are being ground to balling fineness and converted into

367

368


Fig. 6.8-19 Simplified flow diagram of a complete iron ore pelletizing plant (courtesy CVRD, Vitoria, Espirito Santo, Brazil)

high-quality pellets. Fig. 6.8-22 is the simplified flow diagram of a natural ore regrind and pelletization process [B.10]. Of course, in reality the feed preparation section consists of several ball mills with the associated equipment and the single straight traveling grate machine is fed from a number of balling drum circuits. The difference in the processing of natural ores is that the clay minerals have not been removed by either concentration or de-sliming processes and, therefore, play a part in the operation. Generally, the fineness of feed materials is not only important for successful growth agglomeration but must also be compatible with subsequent operations. For example, pellets made from very fine material, for example due to the presence of clay, can cause problems in the drying stage of the induration section. Also, the filtering of slurry produced in wet tumble mills is very difficult if clay is still included. Therefore, as shown in Fig. 6.8-22, closed circuit dry ball milling and the removal of clay in a thickener is the preferred method of feed preparation for a natural ore pelletizing plant although other flow diagrams are also used.

6.8 Applications in the Mining Industry (Minerals and Ores) Fig. 6.8-20 Views into the fine grinding bays of multi-million ton per year iron ore pelletizing plants using: a) semi-autogenous mills, b) rod and ball mills for size reduction

Fig. 6.8-21 a) aerial view of US National Steel Pellet Company’s Keewatin, Minnesota, concentration and pelletizing plant on the Mesabi (Taconite) iron range. b) aerial photograph of the pelletizing (only) facility of Quebec Cartier iron Co. at Port Cartier, Quebec, Canada. [6.8.1.3]

369

370


Fig. 6.8-22 Simplified flow diagram of a regrind and pelletization process for natural ore [B.10]

Tumble/Growth Technologies for Non-Ferrous Minerals and Ores As in the case of iron ore concentrates that are too fine for direct use in reduction and smelting processes, size enlargement by tumble/growth agglomeration is used for non-ferrous minerals and ores that feature the same characteristics and limitations, that is they are fine because they have been upgraded or they are produced as finely divided particulate solids. Among the first group are copper, nickel, platinum, zinc, and other ore concentrates [B.35] and many precipitated products and by- or waste-materials from processing plants belong to the second group. While the former need modification for their proper use, the latter are treated for recirculation (Section 8.2). Preparation of pelletized charges for hydrometallurgical, melting, reduction, or roasting processes follows the same two-stage agglomeration that was described for fine iron ores and uses similar equipment (Fig. 6.8-23) or employs a binder for the development of sufficient strength for the process that follows [B.35]. In some cases, layered pellets are produced, for example as feed to the lead-making shaft fur6.8.1.2


filter cake

dust collection oil air

drying drum

to thickener Surge bin

to dump balling pan

wring belts

green balls

traveling grate machine

roller screen

to dust collection

product screen

fines pile overs pile to storage bins Fig. 6.8-23 Simplified flow diagram of a two-stage manganese ore agglomeration plant. Above) concentrate drying and balling stage; below) pellet hardening stage [B.35]

nace (Fig. 6.8-24). For such applications, the collared or stepped pan agglomerator [B.48, B.97] is often preferred as it is better suited for the balling of multi-component mixes, enables the controlled take-up of individual materials for the layered structure, and produces pellets in a narrow size range. If binders are used, they may, at the same time, be part of the reaction system that is used for the extraction of non-ferrous components from, for example, roasted pyrite residues (Fig. 6.8-25) [B.35]. Another reason for size enlargement by tumble/growth agglomeration is that, for better reactivity or solubility, a large specific surface area of the solids (i.e., small particle size) is desirable, which is retained in the agglomerated product. After size en-

371

372

6 Industrial Applications of Size Enlargement by Agglomeration Fig. 6.8-24 Diagram of a layered pellet [B.35]: 1, core (recycled undersize); 2, flue dust; 3, limestone; 4, lead containing concentrate + SiO2 + iron ore

largement, the bigger porous entities can be easily handled and fed into reactors without experiencing excessive losses by dusting or oxidation. Examples for such applications are finely ground cement raw meal, and glass batch. Many of the fundamental studies of the pan agglomerator were carried out in connection with the pelletizing of cement raw meal [Section 13.3, Ref. 2], which was widely investigated in the 1950s and 60s. Some experimental plants were built, but later this approach was abandoned in favor of preheating fine raw materials in fluidized bed heat exchangers.

Fig. 6.8-25 Processing plant for the firing of roasted pyrite residues using CaCl2 as a binder and reactant for the extraction of non-ferrous components [B.35]. 1, balling pan; 2, belt dryer (max. 250 8C); 3, shaft furnace


Fig. 6.8-26 Flow diagram of a glass batch agglomeration system and photograph of the control panel (courtesy Philips Lighting BV, Winschoten, The Netherlands)

Glass batch agglomeration was also extensively evaluated during the second half of the 20th century [6.8.1.4–7]. For the making of good quality glass, a high-standard of batch preparation (feed to the melting furnace) is essential. Not only is the selection of the type and size of raw material (mineral or synthetic, wet or dry, fine or coarse) important, but after compounding and mixing it is necessary to ensure that either caking or segregation and/or cross-contamination do not occur. The advantages of agglomerated glass batch (pellets) include [6.8.1.7]: pellets are free-flowing, therefore, they are easily transported by many systems and are simply and reliably charged into the furnace, they do not cake, their components do not segregate, and very little dust is produced, they can be stored easily, can be packed for prolonged storage, and several recipes can be kept in stock, suitably separated and ready for use. The grains of the raw materials for growth agglomeration, particularly the sand, must be finer than those of regular batch. This is due to the fact that the agglomerate forming particles must have a certain minimum fineness to create strong enough bonds to accomplish growth and green strength (Chapter 5 and [B.48, B.97]). Green pellets may contain up to 15 % moisture. This allows the use of various liquid components, such as sodium hydroxide, arsenic acid, and soluble colorants. The result is a nearly ideal distribution of refining agents and colorants. After drying at 400 8C the

373

374


Fig. 6.8-27 Green pellets discharging from a pan agglomerator Fig. 6.8-26 and of dry pelletized glass batch (courtesy Philips Lighting BV, Winschoten, The Netherlands)

pellets are strong, their water content is below 2 %, and the bulk density of the agglomerated batch has been increased substantially [6.8.1.7]. Fine-grained components can not be used in regular glass batch because they feature a higher gas release rate and bonding of the melting batch takes place, forming a floating insulating layer on the melt. Pellets, on the other hand that by necessity incorporate finer components, are more reactive but the feed layer remains porous during meltdown and preserves the ability to freely release gas. Melting without cullet is possible. As a result of all this and in addition to the advantages described above, there is ample evidence that the use of agglomerated glass batch results in an increased melt-


ing rate and a decreased power consumption. The use of pellets allows a reduction in the temperature of the hot spot (up to 30 8C [6.8.1.7]) without sacrificing glass quality and furnace output. This and the lower amounts of dust and carryover increase the life of the furnace, the recuperators, and the regenerators and reduce pollution of the environment. In spite of these advantages, little industrial use of glass batch agglomeration takes place. However, when the Special Glass Factory of Philips Lighting BV at Winschoten, The Netherlands, was built (start-up in 1980), it was decided to operate it with 100 % pellets as batch. Fig. 6.8-26 depicts the flow diagram of the agglomeration system and a photograph of the control panel. Fig. 6.8-27, shows photographs of green pellets discharging from the pan agglomerator and of the dry product. The diameter of the pellets is 12 mm on average, each featuring the required chemical composition and physical properties. The merits of using agglomerated batch was found so great in practice that the company began to offer this high-quality material to other glass manufacturers [6.9.1.7]. As of this writing the plant is still in operation. Over eighty different raw components are used to produce agglomerated batch compositions with a high demand on homogeneity and quality for various special glass types.

6.8.2


Towards the end of the 19th century several briquetting machines for the size enlargement of clay for the production of building materials and of coal fines as shaped solid fuel had been invented and many plants for these purposes had been built and were operating successfully (Sections 6.10 and 6.10.2). Therefore, it is not surprising that, when the first commercially acceptable process for the agglomeration of fine iron ores was developed in 1899 by Gr€ondal, it was based on briquetting [B.8, paper 36]. The Gr€ondal process originated in Finland, using equipment similar to that applied for the shaping and hardening of clay bricks (Section 8.7.2). Briquettes were made by pressing fine iron ore, mixed with water as a binder, into rectangular shapes about the size of building bricks. These green agglomerates were then loaded onto cars and passed through a gas fired tunnel with a temperature of about 1400 8C in the combustion zone. It was said that the strength of the fired product resulted from the heat evolved by rapid oxidation in the combustion zone. The porous briquettes produced in this manner were hard, featured extremely low sulfur, and had a ferric iron content of more than 90 %, regardless of the originating ore type. In spite of their unfavorable size and shape, mostly because of the fineness of the original ore and the porosity of the briquettes, they made an excellent blast furnace feed and it was established that Gr€ondal briquettes increased blast furnace capacity and reduced coke requirements. The success of the first Gr€ondal plant at the Pitkaranta Iron Works in Finland led to the installation of a similar one in Sweden in 1902. By 1913, there was a total of 38 plants throughout the industrially developed world. Of these, 16 were in Sweden, 12 in the UK, and 6 in the USA.

375

376


Many other agglomeration techniques for ores and minerals were patented early in the 20th century, but, excluding sintering (Section 6.8.3), the Gr€ ondal process was the only success until pelletization was developed and found worldwide application with an unprecedented pace (Section 6.8.1). As for the size enlargement of iron ores, briquetting with roller presses, which was a widely accepted technology in the coal industry since the middle of the 19th century (Section 6.10.2), was also evaluated for the treatment of many ores and minerals. However, because the machines were not yet designed to exert high forces and withstand elevated temperatures (below), binders were required to produce sufficient product strength. These binders included asphalt, bitumen, coal tar pitch, cement, lime, silicates, sulfite liquor, and many others [B.48, B.97]. Most binders were not desirable constituents because they contained impurities, were costly, and, during final use, lost strength prematurely. In addition most briquettes were relatively weak and the economics of the process suffered from excessive wear. During the mid 1950s when coal briquetting started to decline (Section 6.10.2) and some equipment manufacturers searched for new applications of their machines, stronger presses were built, improved roller designs developed, and novel materials of construction selected [B.13b, B.48], which individually or together made applications feasible, which heretofore had been uneconomical. In particular, the segmented roller design, which had been proposed much earlier but failed due to the lack of suitable wear- and high-temperature resistant steel grades [B.13b, B.48], together with extensive cooling of machine components, allowed the processing of hot materials. Since metals, ores, and minerals become quite malleable at temperatures in the range 500–1000 8C hot briquetting was evaluated for several applications. The first hot briquetting facilities were used in pilot plants of the fledgling technologies of DR iron ores and for the de-oiling, densification, and shaping of cast iron chips (Section 6.9.2). Since it quickly became obvious that iron ore pelletization for the size enlargement of fine iron ores by balling and heat hardening (Section 6.8.1) is only economical in plants processing more than 1 million t/y and hot briquetting with roller presses promised efficient and profitable operation at much lower capacities, several pilot and demonstration systems were built and operated in the 1960s. Fig. 6.8-28 is a schematic elevation of the largest iron ore hot briquetting plant that was installed at the Steel Company of Wales in Margam, UK, with participation by the engineering companies Dravo and Davy-Ashmore [Section 13.3, Ref. 32]. Cost comparisons between sintering, pelletizing, and hot briquetting, which were based on the operation of the plant during the years 1964/65, showed that the investment for briquetting is 50 % less than for the two others and that overall production costs are 27 and, respectively, 23 % lower. A large part of the financial advantages is due to the fact that no binders are required, which eliminates its cost and the need for binder-related installations and equipment and directly influences the product value by not adding a contaminant. In spite of the use of high-quality steel grades for the segments and other machine parts, wear is considerable and, while, when compared with other methods, the total costs associated with this item are not excessive, the process interruptions, which can be only avoided if redundant stand-by equipment is installed, are not tolerable. In


Fig. 6.8-28 Large scale pilot plant (10 tonne/h) at the Steel Company of Wales, Margam, UK, for the hot briquetting of iron ore [Section 13.3, ref. 32]

addition, it was found in another pilot plant, which was built and operated at Arbed, Luxembourg, in the late 1960s for the hot briquetting of Minette ore, that the considerable amount of limestone contained in this material is converted to calcium and magnesium oxide during heating of the briquetter feed and renders the product unstable. The oxides hydrate with ambient moisture; this unavoidable reaction is associated with an increase in volume, which causes even the strongest briquettes to weaken or completely fall apart. It was found that, in a stockpile, the product deteriorates in 2–3 days so much that it can no longer be reclaimed and fed into the blast furnace without causing excessive dusting and losses. An attempt to use logistics (first in – first out) in producing, depositing, and reclaiming the product was unrealistic in the blast furnace environment, and the project was abandoned. Another relatively early application of size enlargement by pressure agglomeration uses roller presses for the compaction/granulation of phosphate rock prior to calcination in a system for the production of elemental phosphorous. The previously described grate-kiln process (Section 6.8.1) was modified by Allis-Chalmers for the calcination, but run-of-mine rock produced an arc furnace feed of varying quality due to widely different sizes and structure of the raw material. In addition, fines had to be removed and discarded because they became entrained in the off-gases from the heat treatment (calcination) facility. Allis-Chalmers modified a state-of-the-art roller mill into a “compactor mill” (Chapter 4, Fig. 4.3) to crush and densify phosphate rock into a compacted sheet that is broken and screened into a granular product with narrow size distribution. This feed to the calcining system resulted in optimal drying, heating (firing), and cooling and produced an excellent charge material for the electric arc furnaces (Fig. 6.8-29). The advantages of using roller presses for such processes over tumble/growth agglomeration are that there are essentially no requirements on the fineness of the material to be agglomerated and that normally no binders are necessary. The compaction/

377

378


Fig. 6.8-29 Flow diagram of a plant for the production of elemental phosphorus [Section 13.3, ref. 32]

granulation process also allows to adjust the product size and distribution to the optimal feed properties. Nevertheless, some other plants use roller briquetting, which always produces the same, typically larger and almond or pillow sized shape, requiring longer calcination time but resulting in an even better electric arc furnace feed. The higher cost of the pocketed rollers and more noticeable wear are offset by the elimination of the granulator (breaker) and a much smaller amount of fines to be recirculated. In addition to size enlargement, densification is often a major requirement when pressure agglomeration is used. One such application is the briquetting of magnesium oxide (MgO) prior to high-temperature sintering (dead-burning) in rotary kilns for use in refractories (Fig. 6.8-30). Magnesite, the raw material, is either mined and, after milling, concentrated or precipitated from sea water. The first calcination step, the conversion into MgO, is often carried-out in rotary hearth furnaces and the resulting product would be fine enough for successful tumble/growth agglomeration. However, the most important requirement for the sintered end-product (dead-burned magnesia) is a tolerance for extremely high temperatures in the refractory material. This is only achieved if the magnesium oxide reaches theoretical density during deadburning which, in turn, requires a high density of the feed to the sintering furnace. Therefore, although the MgO powder is very fine and aerated, a condition, which is not very desirable for briquetting [B.97], densification in roller presses must be applied.

6.8 Applications in the Mining Industry (Minerals and Ores) Fig. 6.8-30 Flow diagram of a sea-water magnesium oxide (magnesia) plant: 1, precipitating thickener; 2, pump; 3, washing thickener; 4, vacuum filter; 5, rotary hearth furnace; 6, screw conveyor/cooler; 7, 15, bucket elevator; 8, feed bin; 9, roller briquetting press with screw feeder; 10, screw conveyor; 11, screen; 12, 13, chip (undersized fines) recycling; 14, chip surge bin; 16, magnesia (sintering, dead-burning) kiln; 17, magnesia cooler

Because large amounts of air are squeezed from the powder during briquetting, a vertical screw feeder must be applied to force the material into the nip between the rollers [B.48, B.97]. Even then, particularly if the magnesium oxide originated from precipitated material and, therefore, was extremely fine, the natural densification ratio of a medium sized roller press (typical machines featured a roller diameter of about 520 mm) was not high enough to produce a high yield of well-densified briquettes. In such cases, separate equipment is used for predensification (for example, Fig. 6.8-31a) or already partially densified material (“chips” in Fig. 6.8-30) is recirculated in a closedloop system and added to the fresh feed (Fig. 6.8-31b). The systems shown in Fig. 6.8-31 have been often used to render smaller roller presses suitable for the briquetting of very fine aerated materials, such as magnesium oxide from seawater precipitation. If presses with larger roller diameters (e.g.,> 1000 mm) are used and the material is fed hot, directly from the rotary hearth calciner, a high yield of briquettes with excellent density is obtained and only a very small

Fig. 6.8-31 Two flow diagrams of precompaction arrangements for roller presses [B.48]

379

380


amount of undersized material must be recirculated. Modern plants have been equipped with large hot briquetting roller presses. As a final example of the use of pressure agglomeration for minerals, various applications of roller presses in the salt (sodium chloride) industry shall be discussed. The technology is used for the size enlargement of rock salt fines and of crystallized byproduct salt from the concentration of potassium chloride and for potassium chloride itself and for other salts. All salts deform easily under pressure and, while becoming corrosive in the presence of water, do not cause much mechanical wear unless hard solid impurities or contaminants are present, which happens rather infrequently. Most of the size enlargement of salt by pressure agglomeration is accomplished by compaction/granulation (Section 6.6.2). The application of the granulated product determines its size and distribution, which are easily adjusted by changing the openings of the screen decks and, possibly, modifying the crusher (granulator) type or operating characteristics. Sodium chloride as a household and industrial seasoning (Section 6.4.2) is the finest (about < 2 mm) while fishery salt used on fishing vessels is coarse (> 5 mm). Granular salt for industrial applications (for example potassium chloride) is also designated and produced as coarse. Sodium chloride is also briquetted into a well-densified pillow or almond shaped product that is sold directly to the consumer (Fig. 6.8-32). These briquettes are used to maintain a concentrated salt brine for the automatic regeneration of ionic water softeners. For the proper functioning of the regeneration cycle, the briquettes should not fall apart when immersed in water; rather they must dissolve, much as if they were large salt crystals. That means, they must feature high density and bonds that are waterproof. Such a binding mechanism is obtained when during briquetting the salt particles are plastically deformed, come into close contact, and a natural recrystallization (healing of the crystal structure) takes place. The quality of such briquettes is determined with a so-called mush test, which determines whether or not submerged briquettes fall apart and form a sludge. The latter is unacceptable.

Fig. 6.8-32 Salt briquettes that are used for the regeneration of ion-exchange water softeners


In Section 6.8.1.2 the advantages of glass batch agglomeration were discussed. Taking into consideration that, in addition to the common properties of agglomerated materials, that is no segregation of components, good flow and storage characteristics, and low content and production of dust, some of the specific reasons for improvements that were experienced with pelletized glass batch are higher apparent density (better penetration and sinking behavior) and better solubility (because of smaller particles), it was an obvious conclusion that briquetted (compacted) glass batch should feature even more pronounced advantages. Therefore, developments employing roller briquetting machines were piloted in several companies, particularly in the USA (Corning Glass, FMC, Ball, US Gypsum, and others). From these tests, which are carried out with essentially dry batches, thus avoiding expensive drying, it was initially concluded that the pressure consolidation of glass batch leads to significantly shorter melting and refining rates and better glass homogeneity (as compared with raw and growth agglomerated feed) and, therefore, warrants further investigation. However, it was later found that, because of the hardness and abrasiveness of most of the glass batch components (particularly silica sand), considerable wear takes place during briquetting. Although this by itself is not necessarily detrimental, as new, hard roll materials and novel roller designs (segments) have become available, even small contaminations with iron and other (alloying) metals result in a discoloration of the glass. Since batch agglomeration is particularly interesting for high-quality glass, pressure consolidation has been ruled out and testing was discontinued.

6.8.3

Other Technologies

Sintering, size enlargement by heat, has been known for a long time [B.97]. Originally, it was mostly applied on a small scale by medieval alchemists for the fusing of different metal-bearing materials and in the laboratory by assayers for the preparation of samples. It was used on a larger scale for the lumping of ore fines when larger and more numerous shaft and blast furnaces were introduced, particularly to satisfy the great need for iron at the beginning of the industrial era. In the beginning, industrial sintering was carried-out batch-wise in pan-like structures on grates that were heated and supplied with air from below (updraft) with additional heat produced by the burning of solid fuel that had been mixed with the raw ore. Later, the bed was ignited from above (see below) and air, containing the oxygen that is necessary to sustain burning of the solid fuel, was pulled in a downdraft fashion through the material layer by a fan. Even today, modern pan sintering facilities are still used where small amounts of metal ores must be agglomerated (Fig. 6.8-33). As usual, heat is provided in such systems by carbonaceous solid fuels that have been uniformly distributed in the ore charge. After ignition, for example by depositing red-hot charcoal or coke onto the bed surface, air is pulled through the machine by a suction fan. When the burning fuel, which produced the intense heat necessary for sintering, is depleted, the material is cooled. Since the entire charge has become one large cake it must be removed by

381

382


Fig. 6.8-33

Flow diagram of a contemporary pan sintering plant [B.97]

suitable means, broken, and screened into the desired sinter size. Oversize particles are re-crushed in the closed breaker loop and fines are recirculated to be mixed with fresh ore and fuel. Such plants may produce 30–50 ts/day of sized sinter [B.97]. In most cases, however, economic and processing reasons require the treatment of larger amounts of several hundred tons per day or, in the case of iron ore, of over 1000 ts/day to satisfy the needs of modern base metal smelting facilities. Such quantities also necessitate the use of a continuous system. For this, introduced by several companies at almost the same time at the beginning of the 20th century, the traveling grate, a slowly moving endless belt, also called a strand, made up of hinged steel plates with grate bars (Section 6.8.1), was used. The early machines (Fig. 6.8-13) used open belts with only a short ignition furnace located at the beginning after the bed is deposited and a de-dusting hood at the end where the sintered cake is broken off. More modern installations with essentially the same design include a number of traveling gratemprovements and cost-saving features (Fig. 6.8-34) [B.48]. They are related to better feed preparation and more controlled deposition of the hearth layer and of a deeper bed through surge bins and charging devices, energy savings by the use of hot air from a separate sinter cooler for bed preheating and in the ignition furnace, additional heat recovery in boilers, improved cooling efficiency by employing a pre-breaker in front of an optimized cooler, desulfurization and denitrification of the exhaust gas, and electric power savings from the use of more efficient fans. Although, in general, the process has not changed during further development, experience that was gained during the modification of the system for the hardening of pellets (Section 6.8.1) was applied to sintering. In particular this includes the complete enclosure of the traveling grate to accommodate specific combustion and off-gas handling procedures. While, during the induration of pellets, changes between downand updraft airflow patterns produces advantages, in sintering the downdraft method is used throughout.


Fig. 6.8-34 Flow diagram of a modern sintering plant with improvements and cost saving features highlighted [B.48]

Fig. 6.8-35 shows a traveling grate sintering plant featuring enclosures [B.40, pp.8393]. Waste gas minimization by recycling is used together with energy recovery. Only a relatively small part of the gas in the system is exhausted and, possibly after desulfurization, released to the environment. In three of the gas recycling loops the temperatures are high enough for reasonable heat utilization for electric power generation.

Fig. 6.8-35 Diagram of a traveling grate sintering plant featuring enclosures and waste gas minimization by recycling [B.40, pp.83–93]

383

384


Advantages are also obtained from two-stage cooling. Sinter is first cooled from 500–600 8C to an average temperature of 250 8C on the strand, broken to < 200 mm during discharge, and separated at 50 mm on a scalping screen. The oversize is crushed to < 50 mm by a double roll crusher arranged in closed loop with the scalping screen. The material is then again separated at 5 mm; fines are recycled “hot”, thus providing some heat to the feed mix (supplemental energy recovery), and the coarse fraction is transferred to a conventional secondary cooler. As the sinter has obtained its product size before final cooling, this process stage becomes more efficient, also resulting in a more uniform and lower discharge temperature that poses less danger if belt conveyors are used for transport. The structure of and the solid fuel distribution in the material bed on the traveling grate critically determine how well sintering proceeds. Since ore, fuel, and sometimes flux (such as limestone and/or lime) need to be blended prior to the mixture’s deposition on the strand and, if moisture is present or after the addition of moisture, the motion of solid particles in a mixer will initiate agglomeration by growth, it has become a rather common practice to granulate the feed mix, thereby improving the bed structure on the traveling grate, and causing fuel and ore to get into closer contact (Fig. 6.8-36). The use of modern, sophisticated control features is the latest development in ore sintering. Fig. 6.8-36 shows the instrumentation installed at a sinter plant in China [B.56, pp. 450–454]. The multi-variable process control is adaptive in nature, that is it keeps track of variations and automatically adjusts the operation to the most optimal conditions. To overcome the time delays that are inherently experienced in a process of long duration, a prediction algorithm has been included. However, since random, unpredictable disturbances are often experienced, a proportioning expert system is necessary to yield rational and uniform results.

Fig. 6.8-36 Sensor and control systems at the no.3 sintering plant at Anshan, China [B.56, pp. 450–454]

6.9 Applications in the Metallurgical Industry

Batch and straight or circular, and occasionally differently shaped continuous sintering plants are being used not only for iron ores but also for non-ferrous ores and lately are applied also for the treatment of metal bearing fines, such as flue dusts, mill scale, and ground slags, for recirculation (Section 8.2).

6.9

Applications in the Metallurgical Industry Applications of size enlargement by agglomeration in the metallurgical industry are in three major fields: * * *

raw materials and additives for metal production; modification of metal products; and processing of metal or metal-bearing wastes for recycling.

The agglomeration methods used for the raw materials for metal production have been covered in Section 6.8 (mining industry: minerals and ores). Applications for additives and other materials required in different steps of metal production are discussed in the following section. A major modification of metal products is carried-out in powder metallurgy (Chapter 7) while others are reviewed in Section 6.9.2. The processing of metal wastes for recycling is also included in Section 6.9.2, while the recovery of metal-bearing wastes, mostly dusts from metal making, is covered in Section 8.2. Tab. 6.9-1, without supposing completeness, lists some of the most important materials that have been and/or are being agglomerated for a multitude of purposes in the metallurgical industry. Tab. 6.9-1 Alphabetical listing of raw materials, additives, metal products, and metal bearing wastes that have been and/or are being agglomerated to obtain various benefits Raw materials and additives or materials for metal production Alloying elements, electrode mass, minerals, ores, oxidizers, refractories, fluxes, Modification of metal products Direct reduced iron (DRI, sponge iron), metal powders, titanium sponge Processing of metal or metal-bearing wastes for recycling Aluminum chips and turnings, blast furnace dust, brass turnings and swarf, cast iron turnings, converter dust, copper turnings and swarf, copper wires, electric arc furnace dust, ground metal bearing slags, metals sludges, metal swarf, mill scale, zinc and lead enriched dust

Further Reading

For further reading the following books are recommended: B.3, B.4, B.7, B.8, B.10, B.11, B.13,c,d, B.14, B.15, B.16, B.18, B.20, B.21, B.22, B.26, B.31, B.35, B.40, B.48, B.55, B.56, B.64, B.82, B.87, B.89, B.94, B.97, B.98 (Chapter 13.1). Books mainly devoted to the subject matter are printed bold.

385

386


6.9.1


One of the requirements for the successful application of tumble/growth technologies for size enlargement of particulate solids by agglomeration is a sufficient fineness of the particles (Chapter 5). Others are the need for binder liquids and post-treatment. Since size reduction by comminution of metallic materials is difficult and, if possible, expensive and the production of metal powders by atomization of melts, followed by solidification, normally produces ultraclean particles for special applications (Chapter 7), only those metals that are directly available in a finely divided form could be candidates for the use of growth agglomeration. Also, suitable binders that do not introduce objectionable impurities or become lost during post-treatment while giving sufficient strength, are difficult to find at acceptable cost. A few applications exist that combine metal or metal-bearing powders by dry mixing to yield formulated alloying components that are stabilized by wet agglomeration and post-treatment to become a dust free and easily handleable granular product, but most processes in the metallurgical industry use methods of pressure agglomeration (Section 6.9.2). Typically, this technology is also used for the recovery of metal-bearing dusts from metallurgical manufacturing and processing plants (Section 8.2). Certain special metallic parts, such as hard metal inserts for cutting tools and permanent magnets, are pressed into shape and hardened by powder metallurgy (Chapter 7). For good structure and performance the primary particles, metal carbides (tungsten or titanium carbide) and ferrites, must be small but the final parts should be dense. To achieve the required structural density, the powder (produced by chemical reaction and precipitation) has to be filled reliably, quickly, and tightly into the molds with a minimum of air space between the particles. As for all pressing applications (Section 6.2), this necessitates free-flowing granules with excellent metering characteristics and high bulk density. Such intermediate, agglomerated products may be produced by fluidized-bed spray granulation from particle suspensions [B.93]. Similar applications of tumble/growth agglomeration are conceivable if powders or suspensions of metallic solids (obtained by precipitation or other ultrafine particle formation methods, for example in nanotechnology (Chapter 11)) must be converted into a dry, freely flowing granular material for further processing.

6.9.2


During past centuries, the blast furnace was developed to become the dominant producer of raw iron in the metallurgical industry. To be competitive, gigantic installations were built with unit capacities of several million t per year, operating in integrated steel mills that produce semi-finished products from ore and coal in a single industrial complex. Modifications of the ore and coal by agglomeration methods (Sections 6.8 and 6.10) helped to improve blast-furnace operation and further established the dominance of the process. Until recently, the blast furnace route to steel was the accepted technology everywhere in the world.


In the 1960s the idea of “mini-mills” was introduced. These are regional steel manufacturing plants based on electric arc furnace (EAF) technology, melting scrap and pig iron and making a growing number of steel grades for local markets. Because, at any given time, these facilities produce and process relatively small amounts of liquid metal they are very flexible and can supply local markets economically with custommade steels, Therefore, they quickly found widespread acceptance. Pig iron, being a product of the blast furnace, is a clean but expensive material. Its availability is controlledby thelarge primarymills.Thecompositionandpriceof scrap, on the other hand, vary widely. Only the most expensive and scarce scrap (home or machining scrap and #1 bundles), which contains restricted amounts of contamination and highly alloyed or coated steel, is suitable for melting in EAF without extensive and costly adjustment of the composition. Therefore, the price of this commodity varies widely depending on its production, availability, and trader-controlled market forces. As a consequence, several companies became interested in solid-state reduction of high-grade or upgraded (pelletized, Section 6.8.2) iron ores by the direct reduction (DR) processes. They produce virgin iron with little gangue (depending on the total iron content of the ore) by utilizing gaseous or solid reductants [B.14, B.20, B.87]. After a burst of development in many parts of the world, yielding many ideas and patents, as well as pilot and demonstration plants, only a few were industrially and economically feasible and prevailed. Today DR facilities are based on gas and coal as reductants and use shaft furnaces, fluidized beds, rotary kilns, and circular grates. Processes using solid reductant (coal) are used where natural gas or other sources of gaseous reductant are not available or too expensive. The product often contains some ash and residual solid fuel. The capacity per unit is relatively small, 10 000–500 000 t/y, and this method is increasingly being used for the processing of metal-bearing waste materials for recycling (Section 8.2). Processes utilizing gaseous reductants (reformed natural gas, hydrogen, suitable process off-gases, or products from oil or coal gasification) yield cleaner products and are carried-out in shaft furnaces or fluidized bed reactors. The highest quality is obtained if the reductant is a cleaned (desulfurized) reformed gas or hydrogen. Such installations may have a capacity of up to 1.5 million t/y per unit in the case of shaft furnaces or up to 2.5 million t/y by employing multiple modules (Section 9.3). While most of the earlier large-scale DR processes were built in conjunction with EAF shops as integrated mini-mills, in which the directly reduced iron (DRI) was used on site (Fig. 6.9-1), more and more so called merchant DR plants have been installed. The latter produce DRI as premium iron units for export and sale to EAF steel-making facilities around the world. Particular advantages of merchant DRI are its cleanliness, reliable chemistry, and predictability of feeding characteristics (size and shape). To be competitive with high-quality scrap during times when this commodity is offered cheaply, merchant DRI facilities are installed where high-grade iron ore or pellets are easily available and/or the gaseous reductant (natural gas) is in ample supply. Today, such places are north-eastern Venezuela, north-western Australia, western Iran and India, and parts of Russia and China, where iron ore and natural gas are available and the added value of the merchant DRI is of great interest to the local economy.

387

388


Fig. 6.9-1 a) Schematic flow diagram of a mini-steel-mill using DR and EAF (courtesy Hylsa, San Nicolas de los Garza, NL, Mexico); b) Hadeed, Al Jubail, Saudi Arabia; foreground left: ore storage, center: two DR plants, upper right: steel mill (courtesy Midrex, Charlotte, NC, USA)

Other sites are, for example, the south central USA, Malaysia, or Trinidad where the ore feed on its way from the source to the consumer can be easily landed and value can be added by utilizing local sources of natural gas. Except for some thermal and mechanical cracking and degradation, the external macroscopic physical form of DRI is the same as that of the feed material (Fig. 6.9-2). Chemically, oxygen that was associated with the iron (iron oxides) has been removed in the solid state. This leaves a very porous (sponge) iron structure. Fig. 6.9-3a shows an SEM image of the internal surface of a DRI pellet and Fig. 6.9-3b is a polished and etched cross section through the same pellet showing


Fig. 6.9-2 a) Mixture of iron ore lumps and pellets, b) top: direct reduced iron pellets, bottom: direct reduced iron lumps. Observe the slight cracking

the amount of intraparticle porosity. As a result, the apparent density of DRI is low (about 2 g/cm3) and the specific surface is extremely high (around 1 m2/g). To a certain extent, both density and specific surface area depend on the raw material, the type of solid state DR technology, and the operating conditions during direct reduction (Section 13.3, ref. 79). Nevertheless, because of its nature and structure, DRI always behaves differently in many ways if compared with solid (pig) iron. An important characteristic, which is common to all products, is caused by the large specific surface area, and is most critical for the shipment of merchant DRI, is the material’s tendency to reoxidize at ambient temperatures (Section 13.3, ref. 69). Most of the possible chemical reactions taking place during reoxidation are exothermic in nature, that is they produce heat. Since both the thermal conductivity (DRI is an insulator)

389

390


Fig. 6.9-3 a) SEM image of the internal surface of a DRI pellet; b) micrograph of a polished and etched cross section through the same pellet

and the temperature at which spontaneous reoxidation begins are low, it is possible that DRI in bulk masses (storage piles), in silos, or in ship holds will heat due to such reactions and reach the self-ignition temperature from which oxidation proceeds, selfsustained and accelerating, as long as oxygen is available. If water is present, reoxidation is more easily initiated and hydrogen is produced during the reaction, which can lead to the formation of explosive oxyhydrogen gas. Seawater accelerates this process. To avoid catastrophic situations, such as documented in Fig. 6.9-4, the International Maritime Organization (IMO) has, as part of its code of safe practice for solid bulk cargoes, declared DRI as “material hazardous in bulk” (MHB), which requires certain precautions to allow shipment and obtain insurance. Untreated DRI can be transported if the ship is equipped with safety measures as shown in Fig. 6.9-5 [6.9.2.1]. Before loading the cargo, the floor of each hold is wired with thermocouples at specific points for temperature monitoring during the voyage (Fig. 6.9-5a). Steel pipes are also installed in each hold (Fig. 6.9-5b) to allow a purging of the cargo with nitrogen after loading. If the hatch covers are tightly sealed, safe shipment of DRI is possible for short distances (coastal, river and lake) if strict rules are observed [6.9.2.1]. The above is not acceptable for long haul and intercontinental ocean shipments where (salt) water can enter the holds during storms or other unusual conditions (accidents). It also does not solve the reaction and self-ignition problem during transshipment and storage on the receiving side. While the latter can be avoided by keeping the material cool and dry from the time DRI is unloaded until it is used in the steel mill, other precautionary measures must be taken for ocean shipping. It is also desirable for merchant DRI to be inert so that it can be safely handled without losses in conventional scrap yards. Therefore, all companies that are active in the field of DR have put considerable effort into the development of suitable passivation methods. At the same time, test methods have been developed to determine the reactivity of DRI products, simulate the conditions during storage, understand the factors controlling


Fig. 6.9-4 US national press reacting to a DRI shipping accident (Section 13.3, ref. 69)

reactivity, and ascertain techniques to reduce or inhibit reoxidation (Section 13.3, ref. 69). All DR facilities processing pellets and/or lump ores can be designed to produce cold or hot DRI. Cooling is used before discharge if the material requires only limited handling and storage before its use. This is the case if a meltshop is associated or nearby (Fig. 6.9-1b). The option of cold discharge does not exist for plants using fluidized bed reactors as the fine-grained product features an even higher exposed specific surface and, therefore, lower self-ignition temperature. To overcome this problem, based on the success and experience with the hot briquetting of cast-iron borings (below), early pilot plants (ESSO FIOR, Purofer, US Steel) were equipped with roller presses to shape and densify the hot, fine DRI into pillow-shaped briquettes (Fig. 6.9-6). It was found that, if the briquettes attained a density of at least 5.0 g/cm3, the compacted product had become essentially inert. This is caused by a marked reduction of porosity (Fig. 6.9-7) and a highly densified outer “skin”, which makes the remaining internal surface area largely

391

392


Fig. 6.9-5 Cutaway sketches of a ship’s hold equipped for the transportation of DRI [6.9.2.1]: a) locations of thermocouples and oxygen and hydrogen monitoring, b) installation of inerting pipes

inaccessible. As a result, IMO has amended the code to exclude “DRI, hot molded, pressed at temperatures equal to or higher than 650 8C and attaining an apparent density of at least 5.0 g/cm3” from the MHB classification. The reasons why this rather recent development was described in so much detail are: the author’s participation in its evolution from the first pilot plant stages to today and it is an example of how the analysis of technical problems and the evaluation of their roots results in the definition of remedies and industrial solutions. All true (producing exclusively for export) merchant DR plants use roller presses to produce hot briquetted iron (HBI) although other methods of densification and shaping are also feasible (Section 13.3, Patents 5–9). Profitable merchant plants use ores that are more likely to produce fines or avoid the additional cost of procuring highquality lump ore or pellets altogether by feeding natural ore fines or concentrates directly into fluid-bed reactors. As mentioned before, such facilities require hot den-


Fig. 6.9-6

Briquettes made from hot fine DRI (courtesy FIOR, Puerto Ordaz, Venezuela)

sification to render the product handleable and suitable as a source of iron for steelmaking. The particular advantage of briquetting is not only the morphological change from an iron sponge to a high-density compact structure (Fig. 6.9-7) but also that the densified skin incorporates most of the fines that are discharged from the process. Such fines may result from thermal decrepitation and abrasion (lump ore and/or pellet feed) or represent the entire output (fine ore feed). Fig. 6.9-8 depicts three typical hot briquetting systems for DRI. Hot feed in the form of reduced pellets and/or lumps and/or fines is forced by a vertical screw feeder into the nip between two counter-rotating rollers with matched pockets [B.48, B.97]. The briquettes are pillow shaped, 90–140 mm long, 50–60 mm wide, about 30 mm thick, and weigh 450–800 g each (Fig. 6.9-9). Although the rollers are set closely together, a distance of 2–3 mm is maintained between their surfaces to make sure that there is

Fig. 6.9-7 Micrographs comparing the structures of direct reduced iron (DRI, left) and hot briquetted iron (HBI, right) (Section 13.3, ref. 146)

393

394


Fig. 6.9-8 Schematic flow diagrams of three typical hot briquetting systems for DRI: a) left: hot pellet and/or lump feed from shaft furnaces; right: hot fines feed from fluidized bed reactors; both including hot fines recycling; b) hot pellet and/or lump feed without hot fines recycling: 1, roller press; 2, separator; 3, hot screen; 4, briquette cooler; 5, hot bucket elevator, fines recycling (Section 13.3, ref. 146)


Fig. 6.9-9

Photograph showing differently sized HBI

never any metal-to-metal contact. This is particularly important because a high specific pressing force of 150 kN/cm is required to attain the necessary density. During operation, this gap increases to 3–6 mm. Since material in the area between the briquette pockets experiences the highest forces and the resulting material bridge is 3–6 mm thick, a continuous string of briquettes is produced rather than individual ones. A separator, producing mostly single briquettes by breaking the continuous string, is an important part of all hot DRI briquetting plants. Briquettes made from reduced pellets and/or lump ore, typically produced in a shaft furnace, feature stronger connections than those made from fine DRI, typically produced in fluid bed reactors. Therefore, the separator consists of a high energy rotating wheel in the first case (Fig. 6.9-8a, left) and a tumble drum with lifters in the second (Fig. 6.9-8a, right). In the preferred embodiment, the briquettes are cooled by immersing them in water (quench cooling) on a conveyor (belt or vibrating) but other methods of cooling (e.g., evaporative) are also available. Fig. 6.9-8a also shows a hot fines separation (double deck screen) and recirculation (bucket elevator) system. Fines and chips are produced by leakage in the roller press [B.13b, B.48, B.97] and during breakage in the separator. More recently, plants are installed without fines recycling (Fig. 6.9-8b). In such cases, the product is screened after cooling and again prior to loading. The metallized fines and chips can be recirculated to the DR plant as “Remet” or, generally, the material is sold for special applications. In many publications, photographs of HBI show nicely formed briquettes. While briquettes from fine DRI obtained in fluidized bed reactors are more frequently well shaped (Fig. 6.9-6), product made from pellets and/or lump ore and after high energy separation includes not only single briquettes but also some multiples and broken ones (Fig. 6.9-10). Metallurgically, such mixtures behave the same as single, well-formed briquettes. Nevertheless, there is often a concern because broken

395

396

6 Industrial Applications of Size Enlargement by Agglomeration Fig. 6.9-10 Commercial HBI from a mixture of reduced pellets and lump ore (Section 13.3, ref. 146)

pieces expose the less densified interior. One of the commonly observed “defects” is caused by the rolling action of a roller press (Fig. 6.9-11), which results in that pockets are never completely closed (Fig. 6.9-11, left) and that edges are missing (Fig. 6.9-11, right) or open-up (so-called clam-shelling, oyster-mouthing, duck-billing). Nevertheless, the briquettes are inert and can be shipped and handled without problems by truck (Fig. 6.9-12, top left) or ship (Fig. 6.9-12, top right) and stored outside (Fig. 6.9-12, bottom).


Fig. 6.9-11 Left: five successive momentary conditions of briquetting between two counter currently rotating rollers with matching pockets showing that molds are never completely closed. Right: sketch of a defective pillow-shaped briquette from a roller press [B.48, B.97]

Fig. 6.9-12 Top left, HBI being shipped by truck; top right, HBI being unloaded by magnet from a ship hold; bottom, outside storage of HBI in Venezuela before loading it on ships (courtesy Midrex, Charlotte, NC, USA, and OPCO, Puerto Ordaz, Venezuela)

397

398


To obtain the required density, according to IMO the temperature of the feed must be > 650 8C (better > 700 8C). Therefore, all parts of the system that come in contact with the hot material, especially the roller presses and separators must be made of temperature-resistant steel, vigorously cooled, and, at least partially inert (Section 13.3, refs. 102, 125). Fig. 6.9-13 shows views into the hot briquetting bays of different merchant DR plants, depicting the specialized roller presses, and Fig. 6.9-14 shows photographs the facilities. Until hot densification of DRI was introduced for merchant plants the dominant shaft furnace technologies (Midrex and HYL) were designed to discharge DRI cold, even if the facilities were next to and the material was mostly used in an associated EAF steel mill, as shown in Fig. 6.9-1. The disadvantage of this arrangement is that to a large extent (only some heat recovery is carried-out in the DR process itself) the heat from direct reduction is wasted and the EAF charge must be preheated, requiring either special equipment or electric power. The recent development of methods for hot, inertized transport has made it possible to use hot discharging DR furnaces, heat recovery, and consequent substantial savings. Installation of roller presses for the hot densification or external cooling of excess DRI or during upsets in the steel mill (Fig. 6.9-15) results in superior layouts of DRI/EAF steel mills. With many of the blast furnaces at integrated steel mills outdated or due for costly major overhaul during the first decades of the 21st century, steel companies are trying to switch to alternative iron and steel making technologies, also using new iron units as feed [6.9.2.2]. Because of the technically limited size of EAF (maximum capacity about 150 t of liquid metal and several hours tap-to-tap time), resulting in relatively small steel mills with production rates of a few 100 000 t/y using multiple EAFs, the DRI/ EAF route is not an alternative for the large integrated steel companies for the production of mass steels for construction of every kind (automobiles, ships, industrial and domestic structures). One new development that is directed towards the replacement of the multi-million t/y blast furnaces for the production of liquid steel is carried out at POSCO, Pohang, South-Korea. In this new process, fine ore in sinter-feed quality, typically 23 % cheaper than lump ore and even more when compared with high-quality iron ore pellets, is reduced in fluidized bed reactors (process named FINEX) with off-gas from the downstream melter gasifier. The resulting DRI, which also contains a suitable amount of lime for desulfurization, is hot compacted, broken into smaller pieces (hot compacted iron: HCI), transported hot to the smelting facility, mixed with briquetted common bituminous coals (typically 24 % cheaper than metallurgical coal) and charged to the melter gasifier (modified Corex vessel) in which liquid steel and reduction gas for the FINEX DR process are produced. It is anticipated that with this technology million t/y installations, producing liquid steel at considerably lower cost than blast furnaces, can be built and operated [6.9.2.2]. Size enlargement by agglomeration with roller presses is used for the briquetting of coal (Section 6.10.3) and the hot densification of DRI. Still other processes, all using agglomeration at some point, are being developed at different locations [6.9.2.2]. As mentioned in Section 6.9, methods for the recovery and use of metal-bearing (waste) materials are covered in Section 8.2.


Fig. 6.9-13 Views into the hot briquetting bays of different merchant DR plants showing some of the roller presses (courtesy Orinoco Iron, Puerto Ordaz, Venezuela)

399

400


Fig. 6.9-14 Panoramas of two merchant DR plants (courtesy Orinoco Iron, Puerto Ordaz, Venezuela (FINMET), and Midrex, Charlotte, NC, USA (COMSIGUA HBI, Matanzas, Venezuela))


Fig. 6.9-15 Schematic flow diagrams of steel mini-mills using DR with hot discharge: a) hot (pneumatic) transport to the EAF, hot briquetting of excess hot material (courtesy HYLSA, San Ni-

colas de los Garza, NL, Mexico); b) typical hot link with transfer located under the DR furnace and external cooling of excess hot material (courtesy Midrex, Charlotte, NC, USA)

401

402


Fig. 6.9-16 Flow diagram of a hot briquetting plant for cast-iron borings [B.3, Vol. 10 (1965), 16–22]

At the beginning of this chapter it was stated that the use of hot briquetting for the densification of DRI was influenced by the earlier successful testing and use of roller presses for the briquetting of oily cast-iron borings to yield a product resembling pig iron for remelting. This development was initiated by General Motors in the USA, where in two central foundry divisions (Saginaw, MI, and Tonawanda, NY) large numbers of cast-iron motors were produced and machined leaving mountains of borings which, because they were heavily contaminated with boring oil, could not be easily reused. Disposal of the material was out of the question because it was very valuable, well-defined home scrap. The process that was developed [B.35, Vol. 10 (1965), 16–22] uses burning (removal) of the boring oil in a special multiple-hearth furnace to evaporate the liquid and heat the dry metal particles to about 675 8C. At this temperature the cast-iron borings have become malleable enough to produce a high-density briquette while requiring relatively low force and resulting in acceptable (minimized) roll wear. After the roller press, the material is cooled before further handling (Fig. 6.9-16) The briquettes feature a number of advantages as listed in Tab. 6.9-2 and are an ideal melt charge for the foundries’ cupolas. Based on this and other success stories and the need to conserve raw materials and avoid waste (Chapter 8), other, particularly non-ferrous, materials are collected and converted into high-density shaped products (briquettes) by pressure agglomeration techniques. The aim of the briquetting process is to obtain similar characteristics to those listed in Tab. 6.9-2. The aluminum industry has one of the most advanced and complete recycling programs in the metals market (Section 13.3, refs. 120, 132, 137). Among the reasons for this are that while aluminum has become one of the major metals used for construc-

6.9 Applications in the Metallurgical Industry Tab. 6.9-2 Advantages of briquettes made from hot cast iron borings as a melt charge for foundries [B.3, Vol. 10 (1965), 16-22] 1. 2. 3. 4. 5. 6. 7.

The briquettes will take several severe handling abuses without breakage. The briquettes can be stored without deterioration for great lengths of time. The briquettes can be easily transported and weighed (metered). The briquettes are a clean foundry charge material, free of moisture, oil, and other contaminants. The briquettes do not disintegrate in a hot gas stream, no fines blowing away. The briquettes’ high density results in a high heat transfer rate regardless of the type of melting. The briquettes’ curved outer surface (pillow shape) insures sufficient voids in the furnace charge and does not prevent the free flow of hot gases during melting. 8. The briquettes’ chemical composition closely resembles that of the iron being made; little adjustment of the chemistry is required and the slag volume is low.

tion and packing, its production from Bauxite requires large amounts of electrical energy so that, today, most of the aluminum producers are located in areas where cheap (hydroelectric) power is available. Transportation from these locations to the finishing mills adds substantially more to the cost of the virgin material. A large segment of aluminum recycling is based on the collection and processing of used beverage cans (UBC), other containers, and automotive scrap: “old scrap” resulting from obsolescence (Section 8.2). In the context of this chapter, the following will exclusively deal with “home scrap” and “prompt industrial scrap”. Home scrap is recycled within the mill, processing (alloying) and shaping (rolling, extruding) metal into intermediate products, while prompt industrial scrap is new scrap from fabricators who do not choose or are not equipped to remelt. The latter is collected by and marketed through secondary scrap dealers and should be segregated according to alloy, shape, and quality. Aluminum home scrap in the form of swarf is produced in considerable amounts during the trimming of ingots and blocks by milling for sheet and foil manufacturing, in large machining centers, for example in automobile wheel and motor production, and in the aircraft industry. If the turnings and borings form loose, tangled masses, they must first be crushed. This processing step reduces the length of the swarf components, which then become relatively free-flowing and pack more densely. If swarf is also wet and/or oily, an additional cleaning step by washing and drying or thermal treatment is recommended or required. Even after such processing, melting of the now dry, crushed swarf still causes problems, mostly because of difficulties in feeding the low-density bulk material (about 0.25 ts/m3) into remelting furnaces and excessive oxidation due to its large surface area. Fig. 6.9-17 is the block diagram of an improved secondary aluminum smelting operation (Section 13.3, ref. 132). Compared with earlier installations, the “compacting” process step has been added. As described above, to render the material suitable for feeding into the compaction equipment, the swarf must be clean and reasonably free flowing. The flowability and, therefore, its metering characteristics depend to a great extent on particle shape, which is influenced by material hardness. Swarf from softer alloys tends to be thicker and spirally wound; such material is bulky, interlocks, and

403

404

6 Industrial Applications of Size Enlargement by Agglomeration Fig. 6.9-17 Block diagram of a secondary aluminum smelting operation that includes compacting the processed swarf (Section 13.3, ref. 132)

entangles and may have to be crushed. Harder alloys produce thinner, straight chips, which can be used directly. For the compaction of clean processed metal swarf, two methods are available. The non-continuous confined volume punch-and-die process (Fig. 6.9-18 a) and the continuous compaction in the nip of two counter-rotating rollers (Fig. 6.9-18 b). The advantage of the non-continuous punch-and-die process is that highly entangled, very loose swarf can be compacted to yield a relatively high apparent density. Preconditions are that the material can be fed to the die and that the stroke of the punch (densification ratio) is long and/or multiple feeding and compression cycles are possible to form a single compact. In the case of aluminum, the product shape is normally cylindrical with diameters of 80–190 mm, heights of 30–120 mm, and weights of 0.4–7.3 kg. The technology also has a number of disadvantages of which the most important is the relatively slow movement of the hydraulic rams, which must be further reduced to overcome elastic behavior of the swarf, and a dwell time is necessary [B.97]. To improve the unfavorable production/investment cost ratio, it is better to produce large compacts. Owing to interparticle and wall friction during densification, such briquettes feature lower and less uniform density, particularly if the simpler singlesided machines (one punch pressing against a fixed anvil) are used. Double-sided

Fig. 6.9-18 Diagrams of the two methods used to compact aluminum swarf: a) punch-and-die, b) roller press. 1, anvil (removable for ejection of compact; 2, die; 3, punch (reciprocating); 4, feed; 5, compact; 6, rollers; 7, force (screw) feeder; 8, compacted sheet (product).


machines, in which compaction is accomplished between two hydraulically operated punches (Fig. 6.9-19), produce higher density and good transfer stability of the briquettes. With these and other characteristics in mind ([B.97], Section 13.3, ref. 132) punchand-die presses are used when relatively small amounts of aluminum swarf and/or alloys, producing turnings and borings with high elasticity, must be compacted. The flow diagram of a complete processing system is shown in Fig. 6.9-20. The vibro-screen feeders in front of the centrifuges and the press eliminate and by-pass large pieces that might damage the machines. Fig. 6.9-21 shows processed aluminum swarf and briquettes produced with a punch-and-die press.

Fig. 6.9-19 a) Diagram of a hydraulic punch-and-die press for metal swarf briquetting featuring two-sided pressing; b) sketches explaining the operating principle of the press depicted in (a) (courtesy Metso Lindemann, D€ usseldorf, Germany)

405

406


Fig. 6.9-20 Flow diagram of a complete metal swarf processing and compacting system using a two-sided punch-and-die press for briquetting (courtesy Metso Lindemann, D€ usseldorf, Germany)

Fig. 6.9-21 Photograph of processed aluminum swarf and of briquettes produced with a punch-and-die press

The continuous compaction between two rollers offers the advantages of large capacity and high apparent density. However, since an endless thin sheet is formed from elongated and often tortuous pieces, there is no hope of making individual compacts that can be easily charged into the remelting furnaces, similar to the cylindrical briquettes from the punch-and-die process. Use of the continuous roller press was developed for compacting relatively large amounts of home scrap that has been segregated by alloy to realize the highest economical advantage. Fig. 6.9-22 is the flow diagram of such a plant. Not shown in the flow diagram is that the swarf may have to be crushed to yield chips with dimensions of about 0.8 mm 12 mm 20–60 (max. 100) mm. Washing and subsequent drying are required if the oil content (from milling) is > 2 g oil/kg chips. The clean dry aluminum is delivered to the plant pneumatically (1), collected in a cyclone (2), and distributed by a shuttle conveyor (3) into silos (I to VI). From the silos the material is metered by vibrating feeders (4) onto a belt (5) and an elevator (6) into a feed bin (7). An amount, larger than required for the roller press, is removed from the bin by vibratory conveyor (8) and passes a magnetic separator. The integral force (screw) feeder (9) of the roller press (10) supplies the material into the nip between the rollers for continuous densification. The excess material that was made available to reliably avoid a starved feed condition, overflows (11) and is recirculated to bin (7). The compacted strip discharging from the roller press is cut into slabs of approx. uniform length in a double roll trimming shear (12). Fig. 6.9-23 shows three different processed aluminum home scrap samples (top) and sheared compacted strips. On a vibro-screen


Fig. 6.9-22 Flow diagram of a processed metal (aluminum) swarf briquetting system using a roller press for compaction (explanations see text)

feeder (13) potentially uncompacted chips are screened out and recycled to elevator (6) by a vibrating feeder (14) and conveyor (15). The compacted slabs are alternatively deposited by a diverter gate into one of two carts (16) which may be positioned in a sound-proof chamber that is connected to dust collection. Conveyor (4) can be reversed to clean out individual bins or the entire chip storage facility into transport containers (17). During start-up, optimization trials, or towards the end of a run, unsatisfactory compacts may be produced. In this case, the vibroscreen feeder (13) can be turned to remove the off-spec material from the system via vibro-feeder (14) and conveyor (15), which is reversed to transport the material into a dumpster.

Fig. 6.9-23 Three different processed aluminum home scrap samples (top) and sheared compacted strips produced by the continuous roller press process (Fig. 6.9-22)

407

408


Fig. 6.9-24 shows the flow diagram and a photograph of another plant for the briquetting of 8 t/h processed aluminum home scrap (swarf) in two roller press compaction lines of 4 t/h each. Although the capability is larger, after the plant was put into operation total annual production was about 23 000 ts. The slabs produced in such facilities have high density (> 2.3 g/cm3) and are particularly well-suited as feed for reverberatory furnaces. Typically an aluminum loss of 15–20 % is experienced if uncompacted particulate scrap is melted. This is reduced to 2–5 % if compacted slabs are charged, resulting in a sizable commercial advantage. Other high-quality metal refuse that is suitable for remelting is not produced or collected in such large amounts as cast-iron borings or aluminum swarf, discussed above. Such materials are densified and shaped in newly designed small hydraulic punch-and-die presses. Fig. 6.9-25 shows a collection of briquettes from different

Fig. 6.9-24 a) Flow diagram; b) Plant for the briquetting of 8 tonne/h processed aluminum home scrap (swarf) in two roller press compaction lines of 4 tonne/h each (Rhe´nalu, Neuf Brisach,

France): 1, pneumatic transport system; 2, storage silos; 3, reversible horizontal belt conveyor; 4, dumpster; 5, elevators; 6, day bins; 7, roller presses; 8, rotating shears; 9, product collection carts


Fig. 6.9-25 Collection of briquettes made from different metals with small, hydraulically operated punch-and-die presses (courtesy Ruf, Zaisertshofen, Germany)

materials and Fig. 6.9-26 depicts the principle and a photograph of such a press. A predetermined volume of the loose, clean metal dust, chips, turnings, borings, wire clips, foil, wool, etc. are fed from an agitated hopper (to avoid bridging) by a screw conveyor into the press (Fig. 6.9-26a). A tamper (vertical, hydraulically operated piston) predensifies the charge, which is then compacted by the horizontal hydraulic ram. The press features two oscillating pressing chambers and, in addition to the press ram, two rods, one of which expels the finished briquette while the other one is densified (Fig. 6.9-26a, top). Such machines have production capacities of 50– 3000 kg/h (depending on machine size and material to be briquetted) and may exert pressures of up to 5000 kg/cm2. There are many other fine materials in the metallurgical industry that are converted with methods of pressure agglomeration. A major group comprises additives that are used in the liquid metal bath to achieve a number of processes, most commonly either alloying or chemical reactions, such as desoxydation, denitrification, desulfurization (Section 6.9.1). For good dispersibility and/or solubility these components should feature small particle size, but that makes them unsuitable for feeding into metallurgical equipment. In the thermally turbulent gas atmosphere above the melt and the normally strong flow induced by the dust collection fans, small particles are entrained and end-up in the off-gas filters of the systems. Introducing the fine solids into the bath with lances is often not feasible and/or economical. Therefore, the materials should be agglomerated into larger pieces or granules, which are dense and dust free but still retain the large surface area of the originating particles. These requirement can be obtained with briquetted or compacted/granulated products. Since the manufacturing of pressure agglomerated products for the metallurgical industry requires high forces, in some cases, additional benefits can be experienced. The following describes what may occur as an example. During high pressure agglom-

409

410


Fig. 6.9-26 a) Principle; b) photograph of a small, hydraulically operated punch-and-die press (courtesy Ruf, Zaisertshofen, Germany)

6.9 Applications in the Metallurgical Industry Fig. 6.9-27 SEM image of the internal structure of a mineral briquette showing particle disintegration and cracking (Section 13.3, ref. 23)

eration plastic particles are deformed and brittle solids break (Chapter 5, Fig. 5-9). In the latter case, due to the fact that the structure of the compact becomes rather dense, in addition to producing fine particles, cracks are formed within the solids (Fig. 6.9-27). This increases the specific surface area of the material. Among the common steel making additives that are briquetted are limestone and lime. Fig. 6.9-28 shows roller presses in operation for this application. The almond shaped briquettes have better storage characteristics, particularly if quick lime has been processed, and exhibit better solubility in the melt. Before its application in the steel mill, the very porous burnt lime absorbs water and hydrates, forming the more stable calcium hydrate. The higher density of the briquetted material slows this process. Even if briquetted lime is stored outside, only a surface layer will hydrate causing those briquettes to swell, disintegrate, and form a protective coat on the pile, effectively sealing the material below. In the steel mill, the heavier briquettes easily penetrate the slag layer and are quickly submerged in the metal bath. Lime produced from natural lump still features mineralogical bridges, which require time for dissolution. Briquettes are made from small particles, which, in addition, feature microcracks

Fig. 6.9-28 left) Three lime briquetting machines; right) close-up of briquettes on the discharge conveyor

411

412


after they have been subjected to the high forces during briquetting (Fig. 6.9-27). Furthermore, they are held together by a binding mechanism (molecular forces). When the liquid melt penetrates into the pores of briquettes, the binding forces are readily destroyed whereupon the small, cracked particles are set free and dissolve quickly. The mechanisms discussed above are universally present advantages of products obtained from high-pressure agglomeration. They may be applied for all materials destined for similar uses and requiring these characteristics. The most important properties of briquettes or granules produced with high pressure agglomeration methods are: * * * * * * * *

higher density and weight, larger surface area, increased reactivity due to the presence of microcracks, binding mechanisms that are more readily destroyed, reduced dusting, improved flowability, feeding, and metering, easy storage and reclamation, and a whole host of other, more specific beneficial modifications that depend on the particular material and processing details.

6.9.3

Other Technologies

For a variety of applications, spherical particles are required. Many of these are associated with the field of powder metallurgy (Chapter 7). While it is relatively easy to produce spherical particles from low-melting materials by conventional techniques, in which melt droplets are produced and solidified in spray (prilling) towers [B.97], refractory solids in general and, specifically, high-melting metals can not be converted by this simple technique. However, if the solid is available in powder form (Section 13.3, refs. 82, 85, 89, 90, 94), various methods are available to produce spherical particles by agglomeration. One such “spherical agglomeration process” uses an immiscible binder liquid to form spheroidal products from particles that are suspended in a second liquid. These highly specialized materials are required in small amounts and, therefore are carried out in small, high-energy, batch-shaking devices as shown in Figure 6.9-29. In this apparatus, tungsten carbide spheres are manufactured, which after sintering yield ball-pen tips [B.73]. The closely sized particles of 1 mm diameter are prepared by agitating tungsten carbide and cobalt powders in a closed Teflon container with hemispherical ends on a high-speed reciprocating shaker. Halogenated solvents are used as the suspending liquid and water is the binder. The addition of about 6 % cobalt to the tungsten powder is required to lower the sintering temperature to more acceptable levels. In the batch process, compaction and rounding occurs during many collisions between the agglomerates and with the container walls. The advantage of products

6.9 Applications in the Metallurgical Industry Fig. 6.9-29 Teflon cylinder with hemispherical ends, mounted in a reciprocating shaker, and used to form small spheres by the spherical agglomeration process [B.73, B.97]

from this process is that finishing operations, such as lapping and grinding after a preliminary sintering step, are greatly reduced compared with those necessary for spherical compacts from press molding. When comparatively high (but still relatively small) production rates are required, continuous processes are better suited. Figure 6.9-30 depicts a drum agglomerator featuring an internal screen classifier for the formation of uniform spheres by immiscible liquid agglomeration [B.73]. In these tumbling agglomerators, the presence of a liquid slurry is useful to reduce dusting, especially if toxic powders are processed. The liquid environment also avoids avalanching because particles and immiscible binder liquid are uniformly distributed throughout the agglomerating mass thus allowing conglomerates to grow into larger entities in a much more controlled manner. The liquid charge also helps in the development of a desirable tumbling and cascading motion in the equipment because it is more voluminous and better interparticle lubrication occurs than would be the case if no suspending liquid were present. Further-

413

414


Fig. 6.9-30 Drum agglomerator with internal spiral screen classifier for the formation of uniform spheres by immiscible liquid agglomeration [B.73, B.97]

more, the solids are carried with the liquid, which makes internal classification possible. As shown in Figure 6.9-30 a spiral screen that rotates at a slower speed than the drum passes through the charge and picks up agglomerates. Undersized particles fall through the screen openings and return to the agglomerating liquid mass. Larger material moves along the spiral until it reaches a tube at the axis of the drum that directs the finished agglomerates to a discharge point. In immiscible liquid agglomeration, particles with a small amount of adsorbed binding liquid on their surfaces collide and coalesce to form larger entities by growth agglomeration. In the sol-gel process, another agglomeration technique that occurs in the liquid phase, fine particles are initially suspended in a binder liquid. The suspension is then formed into spherical droplets and the excess binder is removed to solidify the droplets into a particulate product (Section 6.10.3). Agglomeration by heat or sintering is also very common in the metallurgical industry. However it is mostly applied for the size enlargement of feed materials (minerals, Section 6.8.3), for the preparation of secondary raw materials from metal bearing byand waste-products (Section 8.2) and for the development of final properties of powder metallurgical parts (Chapter 7). These chapters should be consulted for further information.

6.10 Applications for Solid Fuels

Finally, the modification of metals, for example to prepare improved catalysts, and the production of novel metal or metallic compounds by product engineering, all using nanostructural assembly, is a fast growing modern technological field that often includes desired and undesired size enlargement by agglomeration. Some references to these applications can be found in Chapter 11.

6.10

Applications for Solid Fuels Humans have used solid fuels from the time fire was harnessed to provide warmth and light. In the beginning, solid fuel consisted exclusively of dry plant material, mostly wood, but nomadic tribes living with their grazing mammals in steppes and on the brink of deserts, also learned to burn dry animal excrement. Other vegetation-based solid fuel was dry peat in certain geographical areas. Later, wax, resin, tar, tallow, and fat were found to sustain fire. These materials can be liquefied by melting and torches, originally dry (porous) pieces of wood that were impregnated with such liquids, were the first manufactured solid fuels and, because the aforementioned materials are also moldable, candles incorporating a wick were formed as “advanced” sources of light. Relatively early charcoal, the residue of anaerobic burning (distillation) of wood in an earth-covered pile, was produced to enhance the heating value of the wood and natural outcroppings of mineral coal were already mined in prehistoric times for use as solid fuel. When human culture entered the metal ages (Bronze and Iron) more high-quality solid fuels were required for smelting. This increased the charring and mining activities but, because the metallurgical and other applications of these or, more generally, of all solid fuels necessitates relatively large pieces for optimal burning and heat production, fines, which are the unavoidable by-product of solid fuel preparation, had to be removed and were discarded. The development of advanced blast furnaces in the 16th century, the invention of the steam engine by Watt in 1765, the introduction of modern iron- and steel-making technologies in the 19th century, and the beginning of large scale production of electricity for lighting and power in 1882 (New York, Edison) further increased the demand for solid fuels (particularly coal) and their use grew exponentially during the 19th and 20th centuries. Coal mining and processing became a major industry and, although the share of coal in world energy production is declining, has reached annual production rates of more than 2 milliard (US: billion) tons per year. Together with this growing capacity and the ever-increasing requirements on coal quality, which includes strict product size specifications, the amounts of rejected coal fines have also increased disproportionately. In the industrialized, densely populated countries of Europe, where coal mining had been going on for centuries and mining conditions were becoming more difficult, resulting in high solid fuel cost, technologies were developed to collect, clean, and convert coal fines into useful products for a

415

416


number of industrial and domestic applications (Section 6.10.2). This increased the yield of the valuable raw material and reduced the need for scarce disposal sites. In other parts of the world, where good coal was still available abundantly and could be mined relatively cheaply and where plenty of disposal areas existed, the practice of removing the fines and dumping them into lagoons or landfills prevails to this day. For example, in the mid-1990s in the USA the US Department of Energy determined that “more than 5 US billion tons of coal fines may be locked in disposal sites across the country and that, each year, approximately 40–100 million tons of waste coal are added to these impoundments” [6.10.1]. It was stated that, in a very real sense, coal fines are an “unclaimed fuel”. Landfill areas can be reclaimed if old deposits are recovered and the need for new ones can be avoided if these and newly produced fine “waste coal” are turned into a viable fuel for power producers. While many proposals for the utilization of coal fines in the USA are based on turning fines into coal slurries for direct burning, size enlargement by agglomeration is also being considered. The oldest and most common application of size enlargement by agglomeration for solid fuels, which now also include biomass [B.25, B.48] (Section 8.2), utilizes different forms of pressure agglomeration. More modern uses of the technology try to convert abundantly available cheap sub-bituminous coal from, for example, the Powder River Basin in Wyoming, USA, into an improved product with increased heating value [6.10.2] (Section 6.10.2). Solid fuels other than carbon-based materials can also be agglomerated to obtain specific properties. For example, in recent developments spherical agglomerates are produced from enriched uranium powder as a fuel for specific nuclear reactors (Section 6.10.3). Tab. 6.10-1 lists some of those solid fuels that have been or are being processed most commonly with agglomeration technologies to improve their properties for various applications. Tab. 6.10-1 List of some materials that can be used as or converted to solid fuels and have been or are being processed most commonly with agglomeration technologies to improve their properties (see also Tab. 6.10-3) Municipal refuse Vegetable refuse Wood

Charcoal

Peat Coal

Mineral oil Uranium

Separated (paper, packing materials, organic wastes, plastic) Dried digested sludge (see Table 6.10-3) Bark Chips Saw dust Wood Nutshells Bones Bituminous (soft, brown coal) Subbituminous Anthracite (hard, black coal) From disposal sites Shale Sludge Enriched powder


Further Reading

For further reading the following books are recommended: B.1, B.2, B.3, B.7, B.13b, B.16, B.21, B.22, B.25, B.26, B.35, B.37, B.40, B.46, B.48, B.55, B.56, B.58, B.64, B.82, B.89, B.93, B.94, B.97, B.98 (Chapter 13.1). Books mainly devoted to the subject matter are printed bold. 6.10.1


The briquetting of fine coal was used extensively after the second part of the 19th century (Section 6.10.2). When, about 100 years later, tumble/growth agglomeration methods became more commonly known and were applied for many different materials, numerous studies were initiated and efforts began to use this seemingly simple technology for the size enlargement of fine coal [6.10.1.1]. The big success in the production of spherical iron ore pellets from fine concentrates (Sections 6.8.1 and 6.9.1) and the great improvements in the blast furnace process resulting from their use, triggered the interest in many quarters to also apply agglomeration pans or discs, drums, and cones (Chapter 5) for other applications. As reported during the 3rd International Symposium on Agglomeration [B.21, H36–H51], in 1977, based on pilot plant studies at the University of California, Berkeley, CA, and financed by the Electrical Power Research Institute (EPRI), Palo Alto, CA, a conceptual design of a coal pelletizing circuit was established by Kaiser Engineers in the USA (Fig. 6.10-1). Selection of the proposed equipment and the system layout was much influenced by experience from the large investments in the quickly growing application of iron ore pelletizing in North America and around the world (Section 6.8). To be economical, a plant using the circuit shown in Fig. 6.10-1 had to treat 100 t/h of filter cake consisting of < 0.6 mm coal fines with 10 % ash and 25 % moisture. The filter cake is an intermediate product from either a wet upgrading process or the recovery of high-quality coal fines from slurry disposal ponds. Corn starch was chosen as binder but the use of other additives, including liquids, was anticipated. It was assumed that 35 % moisture is necessary for the formation of good-quality green pellets, thus allowing the controlled addition of water in the agglomeration device. Six pelletizing drums, each 3 m diameter by 10 m long, or six discs of 8 m diameter were considered necessary for the proposed plant. Although the size distribution of agglomerates discharging from a pan is relatively narrow, installation of a roller screen (Section 6.8.1 and [B.97]) with closed loop recirculation of undersized material was provided in any case to improve operating flexibility. Assuming two shifts per day for 5 days per week, an annual capacity of 375 000 ts was projected. The dryer in which moisture is removed and the final, permanent binding mechanism develops, represents a major cost item, both in regard to investment and operation. Although it represents 25 % of the investment and 35 % of the operating costs, in the economical analysis it was argued that a dryer is always necessary to reduce the moisture content if a coal filter cake is to be processed. However, nevertheless, in the

417

418


Fig. 6.10-1

Flow diagram of a proposed coal pelletizing circuit [B.21]

final evaluation the conversion cost was considered too high and no large scale commercial system was built then or yet. Compared with the prevailing coal briquetting (Section 6.10.2), which also necessitates an essentially dry feed (a dryer if wet coal is to be agglomerated), the wet granulation system also requires a binder (equal to about 25 % of the operating cost) and a guaranteed fineness of the feed < 0.6 mm (or, more accurately, a surface-equivalent diameter of < 0.2 mm, see Chapter 5). The requirement that, to become suitable for tumble/growth agglomeration, coal fines (as all other materials, too) must have a certain fineness, seems to make this technology most applicable for the recovery efforts that are centered around the slurry ponds that are found in all coal mining areas [6.10.1.2]. Vast quantities of undersized coal are impounded in these disposal sites. For example, several estimates indicate that in the eastern states of North America alone, more than 2 US billion t of recoverable coal fines are already contained and, each year, coal preparation plants in this area continue to send 30–50 million t of fines to slurry ponds. About 10 years after the proposed coal pelletizing circuit (shown in Fig. 6.10-1 and described above) was developed, additional efforts in the field were summarized [6.10.1.1]. It was concluded then that several technically feasible methods are available but commercial application will depend on the particular location, economic considerations (which may change with time), and the costs of the binder and pre- and post-treatments that will be necessary to render the coal suitable for processing and/or the market. Crushing coarser fractions to meet the size requirements for wet agglomeration will always make the material prohibitively expensive.


With the development of coal–water slurry fuels and their successful use in power plants, direct combustion of fines in circulating fluidized beds and the injection of dry coal fines into a number of thermal and metallurgical processes, the interest in tumble/growth agglomeration for the size enlargement of coal fines from any source has diminished to almost zero.

6.10.2


Among the oldest known and mined coal deposits are the ones around Fushun in Northern China. Even today, in and around this location, an ancient coal agglomeration technique can be observed. After adding a binder, coal dust and fines, distributed by the mines to their workers for personal use, are manually formed into brick-shaped agglomerates in wooden frames. Binders are wet clay, flour (starch) pastes, and, more recently, oils and tars. Air-curing results in sufficient strength for handling and use as domestic fuel. In Europe, as recently as the beginning of the 20th century, similar “production facilities” were still in operation (Fig. 6.10-2) in some locations (Section 13.3, ref. 122). However, these agglomerates were of poor quality. They had low heating value and strength. In response to the large quantity of unmarketable coal fines produced as a result of the much increased need for solid fuel during the 19th century and the high cost of mining coal in Europe, efforts were undertaken to develop new products from coal fines with superior properties. Such coal products, featuring many or all the characteristics in Tab. 6.1 (Chapter 6), are made by pressure agglomeration and are commonly called “briquettes”. The first coal briquetting trials were carried out in the 1840s. In Belgium, France, and the UK bituminous coal fines were briquetted with “sticky” binders while in the USA and Germany peat, lignite, and other carbonaceous fines were dried and shaped

Fig. 6.10-2 Manual agglomeration of coal fines in Germany in about 1900 (Gewerkschaft Susanna, [B.1])

419

420


without a binder to obtain a product with higher heating value. During that time, independent of each other, two totally different briquetting technologies were developed: binderless briquetting and briquetting with binders. The “double roll type briquetting machine” or roller press, the most commonly used equipment for the briquetting of coal with binders, was invented around the middle of the 19th century. The objective of its invention was the economic conversion of coal fines into a coarsely particulate product, suitable for burning in stoves, in furnaces, and on the firing grates of industrial ovens. In the beginning, the machine was also called ‘Belgian’ press because the first successfully operating one was constructed by the Belgian Louiseau and installed not in Europe but in a coal briquetting plant at Port Richmond, USA, in the late 1870s [B.1]. Mashek, whose company merged later with the Traylor Engineering Company in New York, built an ‘American’ roller press and Komarek and Greaves, Co. (today Hosokawa BEPEX, see Section 15.1) was formed in Chicago, among others, at the end of the 19th century. However, predictably, the largest application was in Europe where companies such as Fouquemberg, Humboldt, K€oppern, Sahut-Conreur, Sch€ uchtermann & Kremer, Wedag, Zimmermann & Hanrez, and others supplied great numbers of roller presses [B.13b, B.48]. As shown earlier (Section 6.2.2), around the middle of the 19th century the basic principles of punch-and-die presses were also developed and widely patented. Particularly the indexed table arrangement of McFerran (1874, Section 6.2.2, Fig. 6.2-38) seems to have been influenced by a similar but larger press for the briquetting of coal, the so-called Couffinhal press (invented 1859). Contrary to the success of the rotating table press in the pharmaceutical industry, for coal, this machine was in industrial use for only a short period of time (below) during the same period when the first roller presses were installed. Both the Couffinhal and the roller presses produced briquettes from mixtures of fine hard coal (e.g., Anthracite) and a binder (mostly coal tar pitch). This blend was preconditioned by steam to soften the binder and moisten the mix; therefore, relatively little force was required to form the briquettes and the process entailed more shaping than densification. However, in the search for additional, easily available, and cheap solid fuels, other carbonaceous materials also became of interest for size enlargement by briquetting, notably bituminous or “brown” coal (lignite) and peat. Both were known for a long time as natural resources; they were “mined” in open pits, dried to remove the often very high content of water (particularly in peat), and then burned. For use in modern domestic and industrial furnaces it was desirable to densify and shape these materials and improve the handling, feeding, and burning behavior. Carbonization of the organic matter making up peat and lignite has not proceeded far and very little natural densification has occurred, because their deposits are on or near the surface of the Earth, are water logged, and covered with no or little overburden. Therefore, they still contain large amounts of loose fibers and fossil plant matter, which render the product elastic. Economically, such materials can not be briquetted with either punch-and-die or roller presses, even if a binder is used. After the relatively quick densification in such machines, because of high residual elastic deformation, the products expand during pressure release and weaken or


fall apart completely. To overcome this problem, in 1857 Exter invented what later became known as the ram extrusion press (Prussian Patent no. 6015). In this machine, due to repeated densification in the extrusion channel (below), elastic dry peat or lignite is converted in steps into strong, well formed briquettes without using a binder. The first plant performing binderless briquetting of lignite with an “Exter Press” was installed in 1858 at Ammendorf in the eastern part of Germany (Section 13.3, ref. 157) [B.13b, B.48]. As will be shown later, within a short period of time, beginning during the second half of the 19th century, many coal briquetting plants were built, originally using all three types of machines. The technology reached its peak during the first third of the 20th century with a short but high peak after World War II before, for a number of reasons, coal briquetting technology went into a steep decline. Although several new uses, mostly involving Biomass (below), have been and are being proposed, at the beginning of the 21st century the briquetting of solid fuels is at a historical low. Binderless Briquetting of Peat and Lignin (Soft Coal) Fines As mentioned before, the briquetting machine used for this application is the ram extrusion press, also called the Exter press, or reciprocating ram press. The principle of this equipment is shown in Fig. 6.10-3. A typical feature is the horizontal extrusion channel, which first converges somewhat to allow build-up of sufficient pressure. During continuous operation, the reciprocating punch feeds against briquettes that were formed during previous strokes and compresses the material and all the briquettes in the channel until the wall friction and a potential back pressure at the mouth of the press are overcome shortly before the end of each stroke thus incrementally moving forward the entire column of briquetted material [B.48]. Fig. 6.10-4 depicts the sequence of events during a briquetting cycle [B.97]. The reciprocating motion is produced by, for example, an eccentric drive, symbolized by the circular representation on the left. The diagram on the right indicates the progress of force that is exerted on the material to be briquetted. The figure is self-explanatory. Only a few important operating stages will be discussed. At position 3 the force produced by the ram has reached a level that is sufficient to overcome the friction of all briquettes in the pressing channel and, if applicable, the back pressure caused by the column of briquettes in the cooling channel [B.48]. The 6.10.2.1

Fig. 6.10-3 Diagram of the ram extrusion, Exter, or reciprocating ram press [B.48]

421

422


Fig. 6.10-4 Sequence of events during a briquetting cycle in a ram extrusion press [B.97]

entire line of briquettes moves forward, with the force remaining approximately constant, and a new briquette emerges from the “mouth” of the press (position 4). During the back stroke, the energy of the drive is stored in a flywheel and made available again to overcome the deceleration/acceleration at the return points and to help during compaction. At the beginning of the back stroke (when the eccentric drive has passed position 4) and if elastic materials, such as peat, lignite, or biomass are processed, at first the ram face does not separate from the newly created briquette because of considerable elastic recovery. At a typical rotational speed of the eccentric drive of 90 rpm, the duration of the compression phase is only about 0.04 s. This time is too short to achieve total conversion of elastic into plastic volume change; therefore, the elastic recovery during the back stroke is high and directed toward the feed position because the briquettes near the discharge end of the press are firmly wedged in the extrusion channel. Without the characteristic of ram presses that, during each compression stroke, all briquettes in the pressing (extrusion) channel are again loaded and compacted, whereby more and more permanent plastic deformation is achieved, successful briquetting of material with high elasticity would not be possible. Fig. 6.10-5 is a diagram of the increasing density and decreasing elastic recovery of a particular briquette during repeated pressing and forward movement in the extrusion channel. This performance is a significant difference from the densification process in, for example, roller presses [B.48]. It is important to note, however, that even after the first stroke, the surface produced by the ram face is so highly densified that, during the next stroke and for phases 2, 3 and 4 in Fig. 6.10-4, it acts as the bottom of a confined volume densification chamber until friction is overcome and the product column moves forward. During the entire production process the surfaces of adjacent briquettes do not develop significant bonding; therefore, upon discharge from the “press mouth” or the cooling channel, if applicable, the product will readily separate into single briquettes.


Fig. 6.10-5 Diagram of the decrease of elastic recovery and increase of density during consecutive press cycles in a ram extrusion press [B.97]

To accomplish this, the design of a ram press must provide a relatively long extrusion channel. However, there are physical limits to this parameter because friction and drive power and overall stressing of the equipment increase with channel length. In a technically feasible channel, reaching the conditions of Fig. 6.10-5 may not be possible. Then briquettes can retain a certain elastic deformation which, if suddenly released, could be large enough to damage or even destroy product integrity. Therefore, in most applications, a gradual release is provided by slowly increasing the cross section of the channel prior to product discharge. Fig. 6.10-6 is a cross section through a modern ram extrusion press. The upper channel wall is adjustable such that different release angles can be obtained. In addition, a flexible support system at this point serves as a safety device to avoid overloading due to tramp material in the feed or “overcompaction”. As compared with a closed mold (punch-and-die), in which a predetermined pressure is reached with no difficulty, in extrusion presses the situation is complicated (Fig. 6.10-5). The peak pressures developing at each stroke depend not only on the force exerted by the ram but also on the resistance to the forward movement of the briquettes in the extrusion channel. The latter is influenced by many things: the shape and length of the channel, the changes in cross section in relation to length, the smoothness of the channel walls, the nature of the material to be processed, including parameters such as temperature, structure, plasticity, and, if applicable, the type and length of a curing (cooling) channel.

423

424

6 Industrial Applications of Size Enlargement by Agglomeration Fig. 6.10-6 Cross section through a modern ram extrusion press (courtesy Zemag, Zeitz, Germany)

The rate of pressure increase is also important; it depends on the stroke frequency and length and the rather complicated relationship between movement of the ram and magnitude of the resisting frictional force between extrudate and die and the force caused by the column of already compressed product being pushed forward. These forces change with both the state of compaction and the rate of movement. Therefore, to control the briquetting pressure, channel cross section, configuration and length and back pressure, resulting from pushing-ahead long lines of finished briquettes in the cooling channels, are varied [6.10.2.1]. Because of its well-known, large, and highly exploited lignite (called “brown coal”) deposits, Germany and countries with similar mines surrounding it have been the leaders in binderless briquetting of this fuel with ram extrusion presses. Fig. 6.10-7 is the flow diagram of one of the latest lignite (“brown coal”) drying and briquetting plants in Germany, which was commissioned in 1956 [6.10.2.2] but has long since been shut down and is dismantled. Run-of-mine coal, containing about 60 % moisture, was precrushed in the mine (open pit) to < 150 mm. In the plant it was first dried in four indirectly heated rotary tubular dryers (6), crushed and screened to < 3 mm in a chain conveyor (10), cooled in another chain conveyor (13), and then briquetted without binder in seven three-channel ram extrusion presses (16, 17). Each press feeds briquettes into three cooling channels (a total of twenty one), which also provide back-pressure for coal densification in the press. The capacity, initially 1300 t/day (as depicted in Fig. 6.10-7) of “union-type” briquettes, was later doubled to 2600 t/day (about 0.8 million t/y). Fig. 6.10-8 shows a diagram of an indirectly heated rotary tubular dryer (drum fitted with heating tubes inside, (6) in Fig. 6.10-7). Fig. 6.10-9 and 6.10-10 are two different views of the extrusion presses and Fig. 6.10-11 shows briquettes being loaded into a rail car after cooling. Fig. 6.10-12 depicts different briquette shapes that can be obtained with ram extrusion presses. The briquette in the upper left features the most common, so called “union” shape.


Fig. 6.10-7 Flow diagram of one of the last lignite (brown coal) briquetting plants built in Germany [6.10.2.2]

Fig. 6.10-8 Diagram of a rotating tube dryer for the reduction of moisture content of lignite or “brown coal” (courtesy Zemag, Zeitz, Germany)

425

426


Fig. 6.10-9 Partial view of the briquetting bay showing the drive side of the ram extrusion presses [6.10.2.2]

Fig. 6.10-10 Partial view of the briquetting building showing the briquetter heads and the beginning of the cooling channels [6.10.2.2]

6.10 Applications for Solid Fuels Fig. 6.10-11 “Union-type” briquettes being loaded into a rail car [6.10.2.2]

Fig. 6.10-12 Different briquette shapes that can be obtained with ram extrusion presses

427

428


6.10.2.2 Heydays and Downfall of Binderless Briquetting of Bituminous Coal Fines

Fig. 6.10-13 (Section 13.3, ref. 126 and [6.10.2.3]) indicates that in Germany, historically the largest manufacturer of binderless lignite briquettes using Exter ram extrusion presses, production began to take-off in 1890 and grew until it reached about 60 million t per year shortly before the end of World War II. Particularly in East Germany (GDR) it quickly regained a dominant role in the post-war energy picture of this country. Between 1955 and 1960 production of “brown coal” briquettes in Germany (FRG + GDR) peaked at almost 80 million t per year. Although representing the larger and more populous part, the Federal Republic (FRG) accounted for less than a quarter of that total, indicating the growing environmental concerns in the west, which curbed the use of this “dirty” energy source. As shown in Fig. 6.10-13, the decline after the post-World War II peak until the early 1970s was caused by the phasing-out of “brown coal” briquette usage in the west, which was driven by clean-air legislation. In the mid-eighties some reduction of lignite briquette production and use also began to take hold in the GDR as a result of agreements between the two Germanys and after reunification many plants were closed immediately causing a precipitous drop in production and a very bleak forecast for the future of this technology. (It should be noted that “brown coal” is still a major energy source in Germany where it is used in large, modern power plants with strict environmental control for the production of electricity (Section 8.2). Similar developments are planned for other parts of the world where “brown coal” is abundant.) 6.10.2.3 Briquetting of Hard Coal Fines with Binders (Punch-and-Die Presses)

With the binderless briquetting technique, only relatively soft coals can be briquetted. The saturation or bed moisture of the coal is a good measure of whether a particular coal is soft enough. The transition is at about 40–45 % free moisture content. Coals with higher bed moisture are normally briquettable without binder, while at lower saturation this is only possible in exceptional cases and/or if special techniques are applied. Lower moisture is due to higher coal density, which results in significantly harder particles. Such coals must be briquetted with a binder. After adding a thermoplastic

Fig. 6.10-13 Development of “brown coal” (bituminous coal) briquetting in Germany [6.10.2.3]


binder material, such as coal tar pitch, for a short period beginning at approximately the middle of the 19th century, large, mostly brick-shaped briquettes were produced in punch-and-die presses. Among different designs the so called “Couffinhal” press (Fig. 6.10-14) enjoyed the widest application. For example, at the briquetting plant

Fig. 6.10-14 a) Model and photograph of a Couffinhal press. b) Vintage general arrangement drawing of a Couffinhal press (courtesy K€ oppern, Hattingen/ Ruhr, Germany)

429

430

6 Industrial Applications of Size Enlargement by Agglomeration Fig. 6.10-15 Some typical coal briquettes (bricks) produced with punch-and-die presses

“Wiesche” in the German Ruhr District, where originally a Belgian “Mazeline” press had been installed producing twenty-four 9.5 kg bricks per minute (about 13 t/h), the first of several Couffinhal presses was introduced in 1887 for the production of 1, 3 and 7 kg briquettes that were used in the firing holes of locomotives [6.10.2.4]. Although the Couffinhal press was semi-automatic, employing an indexed table with several molds and separate feed, pressing and discharge stations, the operation was only economical if large briquettes, each weighing several kilograms (up to as much as 10 kg) were made, requiring handling, stacking, and feeding by hand, mostly for use in locomotives and older (smaller) stationary steam engines. Fig. 6.10-15 shows some typical coal briquettes (bricks) produced with punch-and-die presses (Section 13.3, Ref. 126). 6.10.2.4 Briquetting of Hard Coal Fines with Binders (Roller Presses)

The quickly developing need for large amounts of briquetted carbonaceous solid fuels, allowing bulk transportation, storage and handling, necessitated the production of much smaller, egg or pillow shaped briquettes (Fig. 6.10-16). An economical method for this task uses “roller presses” (Section 13.3, refs 111 and 134) and [B.13b, B.48, B.97]. Fig. 6.10-17 represents schematically the operating principle of roller briquetting presses. These machines achieve compaction of particulate feed by squeezing the material between two synchronously counter-rotating drums. Matching pockets repre-

Fig. 6.10-16 High-quality pillowand egg-shaped coal briquettes

6.10 Applications for Solid Fuels Fig. 6.10-17 The operating principle of roller briquetting presses

senting briquette halves are cut into the working surface of the drums (rollers) and form the briquettes. During any briquetting or compacting process (Chapter 5, Fig. 5.9), densification begins by a rearrangement of feed particles resulting in the filling of larger voids in the bulk mass. The pressure rise during this phase is relatively small. This is followed by a steep increase in pressure during which further densification results in plastic deformation and/or brittle destruction of the particulate feed particles [B.13b, B.48, B.97]. Since a considerable reduction in volume and pore space is associated with the densification process, gas that previously occupied the pores between and within the particles to be compacted must be totally removed during densification to avoid compressed air pockets. This process of dissipating air is time consuming, particularly if the permeability of the mass is low (i.e., if fine particles cause small pore diameters), and, therefore, limits the speed of compaction. The single densification step in roller presses is completed in a very short period of time, typically a fraction of a second when the material passes the narrowest part of the nip between the rollers. If air is not totally expelled, pockets of compressed air, together with remaining elastic deformation will result in expansion when the pressure is suddenly released [B.13b, B.48, B.97]. In roller presses where this is the case and can not be avoided, damage to or destruction of the structure of the densified material may occur. In contrast, the ram extrusion press, discussed above, provides “dwell times” under pressure: the column of previously made briquettes is held in the press channel and then advances during the next press cycle. The briquettes (or compacts) are repeatedly redensified; thus, in addition to the conversion of temporary elastic densification into permanent plastic deformation, both effects (repressurization and redensification) assist also in the complete removal of air. Because the maximal pressure attainable with the extrusion principle is limited and sometimes not sufficient for the successful briquetting of harder, but still elastic lignites and to overcome the above mentioned problems of roller presses with such materials, the ring roller press was developed in the 1930s (Fig. 6.10-18). Significantly higher briquetting pressures could be obtained with this equipment and, due to the large diameters of the ring and the internal press roller and the long and slender nip, the speed of densification is considerably lower, thus allowing better deaeration and achieving less residual elastic energy (reduced spring-back). Nevertheless, technical and economic reasons limited its application. Only a few machines were built and since the 1950s this design principle is no longer in use.

431

432


Fig. 6.10-18

The “ring roller press” [6.7.3.1]

The shaping of briquettes in roller presses is performed during a rolling motion of the two synchronously rotating press cylinders. As depicted in Fig. 6.9-11 (Section 6.9.2) the two halves of the pockets are never completely closed. Particularly if high pressure is necessary to obtain good densification and strength, a number of shape inadequacies, characterized by such descriptive terms as “clam-shelling”, “duck-billing”, “oyster-mouthing” (Chapter 14), can be observed on the final briquettes (Sections 6.3.2 and 6.9.2). Since, during their peak acceptance for domestic (heating and cooking) applications, coal briquettes had become consumer products, which must not only perform well but also have a pleasing appearance, such defects would not have been accepted. This was another reason for applying a binder. The conditioned (below) coal and binder mix became so plastic and sticky that, after some densification and under moderate pressure, well-shaped briquettes were formed, which hardened during cooling. In fact, coal briquetting with binders in roller presses was more a shaping than a compaction process, which was so easy that it became feasible to in-


clude distinguishing marks in the pocket design to determine the source of the briquetted product for marketing reasons (Fig. 6.10-16). Fig. 6.10-19 is the flow diagram of an anthracite (hard) coal mixing, drying, conditioning, briquetting, and cooling plant demonstrating the extent and complexity of the more modern (about 1950) hard-coal briquetting systems [6.10.2.2]. To guarantee specific quality requirements, which included strength (handling properties), heating value, and burning characteristics, coals from different sources were often supplied, crushed, mixed, and dried in the feed preparation plant (top and upper right part of Fig. 6.10-19) preceding briquetting. In the specific plant shown here, the briquetting section encompasses four lines, each with three roller presses and all the other associated equipment. Three trains are sufficient to produce the 200 t/h of briquettes for which the plant was designed., Thus, one line is always available as a stand-by and for scheduled maintenance, the latter mostly for the replacement of wear parts. The starting materials are different coals with particle sizes in the range 0–80 mm. After mixing, the coal is dried to < 5 % moisture content, crushed to < 3 mm particle size, and briquetted using, in this case, pitch, clay, and sulfite liquor as binders. Fig. 6.10-20 is a major elevation of the plant showing from left to right the raw coal bins with coarse crushers above and metering below, the drying section, featuring indirectly heated drum dryers, the dry coal storage, fine crushing, and binder mixing section, the

Fig. 6.10-19 Flow diagram of a modern anthracite (hard) coal mixing, drying, conditioning, briquetting, and cooling plant [6.10.2.2]

433

434


Fig. 6.10-20 Major elevation of the plant depicted in Fig. 6.10-19 [6.10.2.2]

coal conditioning and briquetting lines, and, finally, the product drying and cooling conveyors on which excessive moisture evaporates and is removed to the atmosphere while the binder system hardens. One of the most important process steps in a conventional coal briquetting system is the conditioning of the coal/binder mixture prior to feeding it into the briquetting machine. This is accomplished in a vertical pug mill (Fig. 6.10-21) where the blend of coal and binder(s) is converted into a plastic mass by heating with saturated or superheated steam. In addition, the steam acts as a “wetting agent”, which assists in the uniform coating of all coal particles with binder.

Fig. 6.10-21 Sketch (partially cut, interior view right of the center line) of a vertical pug mill for steam conditioning mixtures of fine coal before briquetting [Section 13.3, ref. 111]


The residence time in or the throughput of a vertical pug mill are determined by the moisture content of the material entering the apparatus and the temperature to which the particular coal blend must be heated to obtain optimum plasticity. If these parameters and the desired production rate of the plant are known, the pug mill can be designed to match the application. Through a system of nozzles, distributed around and along the entire shell, steam at 250–300 8C and a pressure of 50–250 kPa is injected into the mass. Since the transfer of thermal energy to solids during the condensation of steam is highly efficient, heating of the coal mixture occurs quickly and uniformly. Although the necessary quantity of steam per metric ton of material varies, it is normally in the range 30–60 t/h. Because different amounts of steam are required in various zones of the pug mill, each nozzle is provided with a regulating valve. Excess steam collects in a dome and is vented to the atmosphere. Stirring arms (mixing elements) on a central, vertical shaft agitate the material and avoid build-up on the walls. The design, number, and distribution of the mixing elements depends on the particular material to be processed. Each pug mill is normally driven individually. At the bottom, a rotating discharge device removes the plasticized coal blend through remotely adjustable gates. Signals from level indicators are used to maintain the correct level by adjusting the feed accordingly. To control wear, exchangeable sleeves are mounted to the inner walls of the shell. External insulation and jackets heated with steam or hot water may be used to minimize heat losses. The principle, the design, and special considerations of roller briquetting presses (Fig. 6.10-17) have been covered in much detail in three earlier books by the author [B.13b, B.48, B.97] and in many of his papers (Section 13.3). Machines for the briquetting of hard coal fines with binder(s) are relatively simple because the process entails more a moderate densification and forming of the plasticized mass into defined shapes than a high-pressure compaction of solid particles. Although featuring modern design, recent equipment for this application is still very similar to the presses that were built shortly after the invention of this technology and throughout its common use. Fig. 6.10-22 depicts the general arrangement drawing of a press built in 1913 by one of the German manufacturers. The machine is equipped with a rotating distribution device to accomplish the uniform feeding of two sets of molded rings, which are synchronized by wide, rugged spur gears that are mounted in the center of the shafts, thus separating each briquetting roller into two halves of identical dimensions (1000 mm diameter 150 mm wide). Such a press, operating at a rotational speed of the rollers of 6 rpm, was capable of producing 6 t/h of relatively small (15–50 g, Fig. 6.10-16) hard coal briquettes that were suitable for home heating purposes. The required specific force, that is the applied pressing force divided by the active roller width, described in kN/cm, is low for hard coal fines briquetting with binder(s), typically 10–20 kN/cm. Therefore, to lower the mass of the rollers, originally and even in more recent designs (Fig. 6.10-23), they were made of hollow cylinders equipped with a ring that carries the pockets and is exchangeable as a wear part. Typically, the vertical pug mill conditioner (Fig. 6.10-21) was mounted close to the roller press and the plasticized coal/binder blend was metered into the rotating feed

435

436


Fig. 6.10-22 General arrangement drawing of a vintage (1913) roller press for the briquetting of hard coal with binder(s) (courtesy K€ oppern, Hattingen/Ruhr, Germany)

distributor by a horizontal screw from which excess steam was vented. Well-formed briquettes discharged from the rollers onto a perforated chute where fines were separated (Fig. 6.10-24a). Fig. 6.10-24b is a photograph of the briquetting floor of a relatively recent plant (installed in 1952, now dismantled) that had, with multiple presses, an annual capacity of 1.5 million ts. The equipment in the upper center is the vertical pug mill feeding the roller press in the left foreground. As compared with the machine depicted in Fig. 6.10-24a, the major difference in the design of this later installation is the use of electrical motors (instead of transmission belts), gear boxes, and the extensive use of vent lines to collect and remove vapors. Feed distribution is still achieved with a rotating distribution device to two pairs of briquetting rings, coupled and synchronized with open gears located between the rings, the press frame is open, and adjustment of the tongues (gates) for feed control [B.13b, B.48] is manual. 6.10.2.5 Heydays and Downfall of Hard Coal Briquetting with Binders

Shortly after introduction of the roller press technology in the late 19th and early 20th centuries, hard coal briquetting with binder(s) applying roller presses found extensive


Fig. 6.10-23 a) Elevation of an early roll type briquetting press designed by Sch€ uchtermann and Kremer [B.13b] and b) detail of the rollers of a relatively modern machine (about 1980, courtesy K€ oppern, Hattingen/ Ruhr, Germany), both featuring hollow roller cores

437

438


Fig. 6.10-24 Vertical pug mill conditioner and roller press arrangements: a) general arrangement drawing dating from 1918; b) Photograph of the

briquetting floor at the Rosenblumendelle mine in Germany (courtesy K€ oppern, Hattingen/Ruhr, Germany)

6.10 Applications for Solid Fuels Tab. 6.10-2 Development of hard coal briquetting with binder(s) using roller presses in some of the coal mining areas of Germany (Ruhr, Aachen, Lower Saxonia) Year

Number of plants

Number of presses

Production [Mt/y]

Average/press

1890 1900 1910 1921 1924 1956 1976

18 32 53 48 55 25 4

19 100 243 205 201 90 19

0.361 1.612 3.847 4.547 4.984 7.196 1.347

0.019 16.12 15.83 22.18 24.80 79.96 70.89

use in the traditional European industrial zones, located in the UK and the central part of Europe, particularly Belgium, Northern France, and Germany, and in other industrialized countries, such as USA and Japan. Before the worldwide depression in 1929, the European production of bituminous coal briquettes with binders reached a peak of tens of millions of t per year and then declined sharply. In the 1930s and during World War II a recovery occurred in Europe without reaching the pre-depression peak. The end of World War II caused another downturn followed by a recovery period. In 1962/ 63 the Cuban missile crisis resulted in a last peak of bituminous coal briquetting in the western world. At that time, Japan and western Europe together produced about 25 million ts per year while the capacity of the USA, whose production had peaked in 1947/48, was already far below 1 million ts/year (Section 13.3, ref. 111, 122, 126, 134, 142, 157). During this history of roller press briquetting, in addition to becoming more modern in design, the machines also became larger. This is shown in Tab. 6.10-2 which lists for several years the number of plants, presses, and their capacity in one of Europe’s largest coal mining districts. Today, only a few plants remain that produce “smokeless” briquettes from carefully selected coals and additives for special applications such as burning in home fireplaces. These briquettes are consumer products requiring a “pleasing” appearance and are sold shrink wrapped in small quantities in specialty stores. Most of the large industrial plants were dismantled and scrapped. 6.10.2.6 Present Status and New Developments in Coal Briquetting

In the “classic” coal mining areas of the western industrialized countries briquetting of coal has decreased to marginal levels. The reasons for this are three-fold. *

*

Briquettes were originally produced as cheap fuel for domestic use and the rapidly growing power and transportation industries. Between about 1890 and 1920 most of the briquettes were consumed by the railroads and stoker fired power stations. With increasing demand for briquetted fuel, briquetting plants became larger, more sophisticated, and more expensive to operate. These developments eventually resulted in the replacement of briquettes by newly emerging cheaper fuels such as gas and oil.

439

440


Raw materials for “cheap” fuel briquettes are bituminous coal fines with coal tar pitch, bitumen, or asphalt as binders or lignite (brown coal). These briquettes produce considerable smoke and air pollution. “Clean air legislation”, now enacted in most industrial countries, bans such fuels.

In spite of or, more likely, because of the decline in coal briquetting, new developments were achieved during the past decades. They were either directed towards a more economical production of briquettes, rendering this fuel more competitive, or the manufacturing of “compliance fuel”, briquettes or compacts that can be burnt without violating anti-pollution laws. More recent developments are using so called “opportunity fuels”, waste materials with calorific value that need to be disposed of without landfilling. To achieve more economical production, the remaining two types of equipment for the briquetting of solid fuels, the ram extrusion press and the roller press, have been redesigned extensively. In ram extrusion, the press drives were modernized for more power input and the shape of the extrusion channels was optimized to allow the application of higher pressures. Since the capacity of these machines, even if producing large briquettes, is limited to a maximum of somewhere around 20 t/h, less with smaller product sizes, higher capacities are achieved by using several channels per press (Fig. 6.10-25) and ultimately multiple lines (Figs. 6.10-7, 6.10-9 and 6.10-10). Roller presses are now available for throughputs of as high as 100 t/h per machine (Fig. 6.10-26, and Fig. 6.10-33 below) [B.97]. However, in those areas where many of the future solid-fuel related applications are envisaged, that is low sulfur sub-bituminous coals, biomass, and other opportunity fuels, the roller press has limited applicability. Unless extensive and costly preparation or conditioning steps are used, which include the addition of new, effective binders, the sometimes very elastic organic materials cannot be processed successfully in roller presses. The single, quick densification, which is followed by immediate, complete pressure release does not allow the manufacturing of permanently bonded briquettes or compacts from such solids in one pass [B.13b]. Another potential problem of roller presses is related to briquette size which, particularly if relatively small briquettes are desired (20–30 cm3), influences selection of equipment size and roller width that can still be fed uniformly and, therefore may limit machine capacity. These considerations must be taken into account when developing new solid fuel related applications for roller presses. The application of ram extrusion presses overcomes the problems of roller presses. Machines with multiple (up to four; for example Fig. 6.10-25, bottom) extrusion channels can produce high capacities per unit and, if this is desired or required, comparatively small amounts of binders will produce, for example, water repellant compacts. Since it is preferable to produce a crushed product (below), initial briquette shape is of no concern and can be optimized for capacity. 6.10.2.7 New Applications of Ram Extrusion Presses: Elastic Materials, Coal Logs

Although the production of the “traditional” brown coal briquettes has come to very low levels, the technology, based on the ram extrusion principle, is available, mature, and has been further developed until recently. While many classic manufacturers of


Fig. 6.10-25 Photographs of a triple (top) and a quadruple (bottom) channel extrusion press (courtesy Krupp F€ ordertechnik GmbH, Essen, Germany)

441

442


Fig. 6.10-26 a) Elevations and plan view of a modern large capacity roller press for the production of about 100 t/h of coal-based compliance fuel with a binder [Section 13.3, ref. 134] (courtesy

K€ oppern, Hattingen/Ruhr, Germany); b) photograph of one of the latest large roller presses that was designed for the briquetting of coal (courtesy Sahut-Conreur, Raismes, France)


ram presses in western Europe folded or gave-up this section of their business, some found new niche markets. The particular characteristic of this press type, the possibility to successfully and permanently densify and shape materials that are typically organic in nature and feature high elasticity (Fig. 6.10-5), can also be used for the briquetting of materials other than peat and lignite that have similar properties [6.10.2.5, 6.10.2.6]. Tab. 6.10-3 lists some of the materials for which ram extrusion presses are already being used. New non-coal applications also include the pressing of very fine powders and of materials that either inherently contain binders or to which binders are added. For example, typical binders in or for wood-based products are lignins, either as a natural ingredient or a waste product (lignosulfonates), and for Bagasse or spent sugar beet slices the binder is unrefined sugar or the by-product molasses. For some of the new applications, less heavy duty machines have been developed [B.25, B.46, B.97]. Many of these use screws for the development of force and, therefore, accomplish continuous extrusion. The rope emerging from the die is cut into pieces, often by a rotating knife. In a much smaller scale, the conditions in the die holes during medium pressure agglomeration or pelleting (Chapter 5, Fig. 5-10b1–b6) realize the same densification process as developed in the (long) channel of ram or screw extrusion presses. Therefore, medium pressure agglomerators can be also used for the production of “fuel pellets” from, for example, shredded paper, saw dust, plastic foil, and similar waste materials [B.97]. Fig. 6.10-27 shows two examples of lose raw materials and the resulting pelleted product. Another potential application of Exter ram presses is in the field of transportation. It has been proposed to produce cylindrical, water repellent “coal logs”, which are transported over long distances in so called “freight pipelines” [6.10.2.8]. The coal log pipeline (CLP) technology is being developed at the Capsule Pipeline Research Center (CPRC) of the University of Missouri–Columbia. One of the key requirements for the economic application of this technology is the production in a very short time of large numbers of good quality cylindrical logs with a diameter of more than 125 mm (5”) and an aspect ratio (length over diameter) of about 1.8. Although most of the laboratory and semi-industrial pilot work is still being carried out with punch and die presses, ram extrusion has also been tested with good preliminary results [6.10.2.9, 6.10.2.10]. Ultimately, the application of this technology for the com-

Tab. 6.10-3 Examples of some new materials, including biomass and opportunity fuels, for which Ram Presses have been successfully applied [6.10.2.7] Bagasse Bark Biomass Cannery wastes Grape wkins Hay

Mineral powders (various, fine) Municipal waste (processed) Paper Polymers Rubber (from shredded tires) Saw dust

Scrap leather Seed husks Straw Sugar beet slices Vegetable wastes Wood chips

443

444


Fig. 6.10-27 Two waste materials and the resulting products after medium pressure agglomeration in a flat die pelleting machine (courtesy Amandus Kahl, Reinbek/Hamburg, Germany)

mercial production of logs from coal and other materials, will require the design, manufacturing, and testing of new ram extrusion presses. 6.10.2.8 New Applications of Roller Presses: Charcoal, Smokeless Fuel, Formed Coke,

Coal-Based Solid Compliance Fuels A relatively new but already mature and presently overbuilt (excess capacity) application of solid fuel agglomeration is charcoal briquetting. While the production of charcoal from wood belongs to the oldest commercially and industrially used technologies of mankind, application of the material as a solid fuel was restricted to chunky pieces which, during their use in the furnace, developed the necessary openness (voids) of the particle bed to achieve an effective flow of air and, as a result, optimal burning. Fines were removed prior to charcoal use as a fuel and used elsewhere (ground finely as a black coloring agent for many applications, including paints and cosmetics, and later in gun powder and as additive in animal feeds). Charcoal briquetting evolved when backyard barbecuing became popular. The additional need for good quality chunks of charcoal which, in the meantime, had become a valuable material for many industrial applications (e.g., filtering, catalysts, chemical synthesis) could not be satisfied. Also, because the material became a consumer product, it was expected by the buyer that pieces looked perfect and contained virtually no fines in the bag (Fig. 6.10-28). Briquettes from charcoal fines resulted in a product with better ignition, particularly if a liquid lighter is employed, and superior heating behavior. Because charcoal briquettes are used in food preparation, government authorities soon regulated their composition. Therefore, the process, which requires a food grade binder (pregelatinized starch) to yield a well-formed, good quality consumer product, became standardized as shown in Fig. 6.10-29.


Fig. 6.10-28

Typical barbecue charcoal briquettes

Most plants use hardwood charcoal (F), which is crushed to < 3 mm (1), stored in a small surge bin (2), and then mixed with the binder. Often premixing occurs in a simple paddle mixer (3) and conditioning in a double shaft vertical pug mill ((4) called vertical fluxer), similar to the one shown in Fig. 6.10-21 but without steam injection. A typical mix formulation consists of 76 % charcoal, 4 % pregelatinized starch, and 20 %

Fig. 6.10-29 “Standard” barbecue charcoal briquetting system. Explanation see text

445

446


water. If soft wood charcoal or a mixture of different charcoals is used, the amount of starch may be somewhat greater and the water content can be as high as 45 %. The well-conditioned mixture is briquetted in a low-pressure double roll press (6) with a paddle feeder ((5) to stimulate uniform flow into the nip between the rollers). The green (moist) briquettes are uniformly deposited by a special conveyor (7) onto the mesh belt of a tunnel dryer (8) and hardened at about 120 8C for 2.5–3 h. The finished briquettes are screened (9) and fines are returned (10) to the paddle mixer for reprocessing. After natural cooling the product (P) is packed or shipped in bulk. To overcome the main problem of conventional coal briquettes, excessive pollution, beginning in about the 1960s, research started in Europe, particularly the UK and in Japan to develop “smokeless” coal briquettes. As a result of fundamental studies into and evaluations of the briquetting of coal at the British National Coal Board’s Coal Research Establishment (NCB/CRE) a book on roll pressing was published in 1976 [B.13b]. Smokeless fuel was expected to become again a consumer product with all its new requirements, especially excellent physical quality and pleasing appearance. The latter motivated one producer in the UK to dip the briquettes into a special gold bronze die during curing to eliminate the “dirty black coal look”. Although a number of processes were developed, which were based on the partial removal of volatile matter and/or either cold cure binders or hot briquetting, the meanwhile abundant availability of cheap, clean energy did not let this technology grow as was originally anticipated. The few plants built, mostly in the UK and Japan, were small and used low-capacity roller presses for briquetting. Most of these enterprises became quickly unprofitable and closed. In the 1950s and 1960s, a number of companies and organizations worked on socalled formed coke processes. The interest on such technologies developed in the wake of an unprecedented growth in iron and steel production and forecasts predicting a continuation of this trend for quite some time. At the same time, fears came up that coking coal reserves might soon be insufficient to satisfy the demand and a growing awareness of the need for a better protection of the environment became a topic of much concern. Therefore, the goals of all formed coke developments were as summarized in Tab. 6.10-4 [Section 13.3, ref. 97]. While some of the processes combined known technologies, for example hard coal briquetting with pitch binder and conventional coke making in indirectly heated slot type ovens, others used a number of novel partial steps which, for the complex overall

Tab. 6.10-4 * *

* *

Goals of formed coke developments

A coke product that would perform as well as or better than conventional coke A process that permits use of a wide range of either coking or non-coking coals and produces a consistently uniform coke A process that meets present and future environmental laws A process that produces coke at essentially the same cost as that from conventional coke making while meeting the above requirements.


system, presented the inherent difficulty of arriving at a continuous operation. Further problems in process development resulted from the large number of goals, particularly the intended increase in the range of applicable coals. New technologies included the process steps volatilizing or semi-coking by low-temperature carbonizing, adapted mixing and agglomeration methods and final carbonizing and/or post-treatment. While during development most of the researchers adopted roller briquetting for agglomeration, balling in drums or discs (Section 6.10.1) and extrusion were also tried. For technical and economic reasons, in those processes reaching commercial acceptability, briquetting is ultimately favored. Fig. 6.10-30 is a generally valid schematic into which all formed coke technologies can be fitted. A fundamental distinction was made into those using hot or cold agglomeration. Parallel to the efforts to develop a process in which formed coke is produced as a new commodity, the partial briquetting of coke oven charges was introduced (Fig. 6.10-31). Only part of the coke oven feed is densified and shaped by briquetting, subsequently mixed with loose coal (Fig. 6.10-32), and charged into conventional slot type coke ovens. This increases the bulk density of the coal feed and, thereby, improves coke quality while, at the same time, allowing the utilization of lower grade or non-coking coal with a high percentage of volatile matter and unfavorable swelling characteristics.

Fig. 6.10-30 Block diagram depicting the different processes for the manufacturing of formed coke [Section 13.3, ref. 97]

447

448

6 Industrial Applications of Size Enlargement by Agglomeration Fig. 6.10-31 Block diagram of the partial briquetting techniques for coke making [Section 13.3, ref. 97]

As in the case of formed coke, the economical and environmentally safe production of briquettes requires large agglomeration equipment, with roller presses as the machines of choice. Fig. 6.10-33 is a collection of photographs showing details of machines, which are used in several SUMICOAL partial coal briquetting plants in Japan and South Africa. Each press is capable of producing between 90 and 120 t/h of coal briquettes. Similar to the machines built at the beginning of the century (Figs. 6.10-22, 6.10-23a, and 6.10-24a) each roller is equipped with two pocketed rings (1400 mm diameter, 900 mm wide). Each set of rings is fed by a gravity feed chute with overflow and tongue control [B.48]. Although a binder system had been added, the mixture to be briquetted is still free flowing; therefore, this simple type of feeder and feed control could be selected. In the early 1970s, when most processes had developed into large pilot plant or early production stages, the initial enthusiasm of the project sponsors disappeared. This was due to first indications of the approach of what since has been termed the 1970s steel crisis and the beginning of a fundamental technological change in the steel industry involving a move away from the blast furnace, the major consumer of (metallurgical) coke. Some plants experienced premature closure, others never reached full production capacity, and plans for big installations were canceled. Nevertheless, interest in the development of alternative coking processes continues, particularly in regard to economy (i.e., use of lower quality coals) and better environmental control. Fig. 6.10-34 shows the flow diagram of the 6 t/h pilot plant of one of the latest attempts [6.10.2.11]. The process is designed for a feed mix including fines that contains 50 % coking coal and 50 % lower-cost non-coking coal (as compared with 25 % in conventional coke making). The fines are hot briquetted before feeding the coke


Fig. 6.10-32 Mixture of briquettes and loose coal on its way to the coke ovens in one of the plants for partial briquetting (courtesy Iscor, New Castle Works, South Africa)

449

450


Fig. 6.10-33 a) Large briquetting machine for the briquetting of coal (courtesy K€ oppern, Hattingen/Ruhr, Germany); b) partially assembled roller press showing the two sets of rings (courtesy K€ oppern, Hattingen/Ruhr, Germany); c) roller press installed in a plant for the partial briquetting of coke oven charges (courtesy Iscor, New Castle Works, South Africa)


Fig. 6.10-33c

oven together with preheated coarse coal. Called SCOPE 21 for “Super Coke Oven for Productivity and Environmental enhancement toward the 21st century”, the project is a joint development of the Center for Coal Utilization, Japan (CCUJ, Tokyo) and 12 Japanese steel and chemical companies. Coking is accomplished at lower temperatures (750 8C vs. 1000 8C) and in a shorter time (10 h against 30 h). The entire system is enclosed and quenching is done by nitrogen. Both reduce dust generation and the latter provides better heat transfer for energy recovery.

451

452


Fig. 6.10-34 Flow diagram of the 6 t/h pilot plant of the proposed SCOPE 21 coal coking process [6.10.2.11]

The low-sulfur sub-bituminous coals from, for example, the Wyoming Powder River Basin (PRB), which are natural compliance fuels, typically contain high amounts of moisture (30–50 %, sometimes more) and, therefore, must be dried prior to briquetting, also to improve their heating value [6.10.2.12]. During drying these coals disintegrate into a finely divided particulate mass and then exhibit a pronounced tendency of auto-ignition. Therefore, the dry coal must be processed with a suitable technology that either directly or by the application of binders (which must not introduce undesirable, polluting components and/or reduce the heating value and must be low in cost) forms highly densified, inert briquettes or compacted pieces. The production of coal-based compliance fuel must yield a bulk commodity at high production rate and with low conversion cost to be competitive with alternative measures to achieve compliance (for example desulfurization processes). While in other briquetting applications well-shaped, uniform briquettes are required or preferred, compliance fuel must meet the size criteria of coal for power plants, which is, for example, 0–50 mm with only a small amount of “dust”, a characteristic that is not defined but represents a subjective requirement for “safe” loading and handling. Coal for power plants is transported in bulk in large trucks, railroad hopper cars, unit trains, barges, and seagoing vessels. Because the volume of these transportation means is predetermined by their existing designs, the specific mass of a new, briquetted product in t per m3 must be close to that of raw coal to guarantee delivery of the same amount. Monosized pieces, such as briquettes of any size or shape, do not pack closely enough. Therefore, it is desirable to produce broken briquettes or compacts by passing them through special “calibration” equipment. As shown in Fig. 6.10-35 (Section 13.3, ref. 134) the screened product after calibration is similar to regular screened coal and, due to some break-down of larger pieces, will also contain a certain amount of finer particles, which fill the voids and increases the bulk density but do not produce airborne “nuisance” dust.


Fig. 6.10-35 Screened coal-based compliance fuel after the sizing (calibration) of briquettes (Section 13.3, ref. 134)

So far, numerous processes were proposed and evaluated. Many laboratory tests and pilot plants and several larger scale demonstration plants were conceived and operated. All were based on roller presses and the application of binders. The developments, each of which by itself represents an important technological advance, did not arrive at an economical large-scale concept, suitable for the conversion of hundreds of millions of tons of PRB coal annually. To overcome the deaeration and elastic spring-back problems, the roller presses had to be run at low speeds requiring multiple briquetting lines to produce the necessary amount of high-quality compacts. Alternatively, extra high percentages of binders could be used to enhance the binding characteristics and produce compacts, which will survive the combined spring-back caused by compressed air and stored elastic energy. However, binder cost, its on site storage and handling and the need for complicated systems for mixing them into the roller press feed render the technology uneconomical. In spite of the fact that such briquettes retain their integrity, microcracks in the structure contribute to the breakdown during bulk handling and shipping or excessive fines. As a result, to date, no system is successfully operating in the multi-million t per year range that is required for electrical energy production. Instead, in the USA, for example, in 1999 about 400 million tons of PRB coal were shipped in run-of-mine condition, mostly by 10 000 t unit trains (100 cars at 100 t each), to power plants throughout the country. Because, as an average, run-of-mine PRB coal contains 35 % water, each unit train transports 3500 t of water, a total of about 14 million tons annually, and, as a consequence, coal with low heating value. This should be a strong incentive to continue the search for an economical briquetting system. Another new application of coal briquetting that is already being used, is the densification and shaping of solid reductant for the direct reduction of iron ores. Such briquettes are mixed with iron ore and fed into the reactor where controlled combustion provides heat and reducing gas for the production of direct reduced iron (DRI).

453

454


The application of pressure agglomeration by extrusion has been also proposed for the beneficial recovery of impounded coal fines from slurry ponds (Section 6.10.1 [6.10.1.2]). One process produces coal-fiber pellets from a 50:50 mixture of fine coal and waste paper fiber. Since the latter contains little bound sulfur and nitrogen, burning the coal-fiber product can yield lower emissions of sulfur dioxide, nitrogen oxide, carbon dioxide, and possibly particulates (including trace metals) than the coal alone. Fig. 6.10-36 is a simplified block diagram of the manufacturing process. The “pelletizer” may be a (flat die) pellet mill (Fig. 6.10-27) or a screw extruder (Fig. 6.10-37). Another application is the BioBinder process. In this method, small quantities of digested municipal sewage or similar sludges are mixed with the coal and allow the formation of pellets. Since the plasticizer/binder is a waste product that is costly to dispose of, its price is negative (assuming the disposal fee is applied as a credit) resulting in low pelleted fuel cost. Fig. 6.10-37 is a flow diagram of the proposed process. To arrive at a weather-resistant pellet an additive is added to the sludge and coal mixture. Finally, another extrusion process for the manufacturing of a fuel from waste materials will be mentioned. After mixing saw dust and other finely divided wood waste with waxy materials the blend is extruded and cut into artificial coal logs for burning in

Fig. 6.10-36 Simplified block diagram of the coal-fiber pellet manufacturing process [6.10.1.2]


Fig. 6.10-37

Flow diagram of the proposed BioBinder extrusion process [6.10.1.2]

domestic fire places. The product ignites easily, burns cleanly, has high output of heat, and, after the addition of traces of suitable chemicals, emits colorful flames for added enjoyment.

6.10.3

Other Technologies

Oil-agglomeration is increasingly important for the treatment of suspension of fine coal in water [6.10.3.1]. Although, referring to the beginning of Chapter 5, the technology is part of group A, sub-groups 6 and 7, the methods applied for the cleaning of fine suspended coal are sufficiently different to warrant their coverage in this chapter. As with froth flotation, the separation effect of oil-agglomeration relies on differences in the surface properties of coal and impurities. Coal fines are preferentially wetted and agglomerated by oil, which is mixed with the suspension and is immiscible with water. Ash forming impurities remain in suspension and are rejected when the agglomerates are recovered, for example by screening. The major factors affecting oil-agglomeration of coal are amount and type of oil, degree and type of agitation, density of the suspension, particle size distribution, and wetting properties of the coal. The most significant is the oil concentration. As progressively larger amounts of bridging liquid (oil) are added, a variety of agglomerates form [6.10.3.1]. In the capillary region (Chapter 5, [B.48, B.97]), requiring 10 %

455

456


or more of the bridging liquid, oil filled coal pellets (spheroidal agglomerates) are formed. In spite of the ability to recover an excellent coal product from sludges that are normally discarded in conventional coal cleaning operations, the high cost of the oil used was an impediment to the commercial adoption of the technology. Therefore, during recent years, decreasing oil levels were used for the production of micro-agglomerates. This approach requires a highly efficient mixer to disperse the oil and centrifugal drying to reach the desired product moisture. Oil-agglomeration can be tailored to form the type of product that is necessary for a given application. Micro-agglomerates may be used in coal dust burners of power plants and larger agglomerates are easily handled and transported. The properties of the hydrocarbons (oils) selected as collectors are important as they afford the wetting of the coal and, hence, the selectivity and yield of the process. Since different oils can be used and their price represents the major operating cost it is necessary to choose the most economical oil for the required ash rejection and coal recovery levels. It is recommended [6.10.3.1] that lighter (< 0.9 g/cm3 density), more refined oils of higher paraffin content be used when optimum rejection of ash is an important consideration. The denser, more viscous, and, generally, more aromatic oils are preferred if recovery and dewatering of coal fines are a major and ash rejection a secondary objective [6.10.3.1]. Because highly variable natural materials are involved, laboratory testing of the wetting behavior is advised for each specific application. Oxidized coals or those of lower rank are relatively more hydrophilic than bituminous coals and, therefore, respond less well to selective wetting by hydrocarbons. Improvements may be obtained by the use of chemical additives that influence surface wetting [6.10.3.1]. Agglomeration at low oil levels produces a coal product with fewer impurities. This is due to the fact that only particles with lower ash content are wetted sufficiently to produce aggregates. On the other hand, the absorption of oil on coal fines during agglomeration displaces water. The moisture content of the recovered product is primarily made up of surface water which, in turn, depends inversely on agglomerate size and oil level. It may be reduced to less than 10 % by agglomeration that is only followed by simple mechanical dewatering, provided the diameter of the agglomerates is larger than a certain minimum [6.10.3.1]. Smaller aggregates with less oil content require centrifuging to obtain the same low moisture level. Fig. 6.10-38 is the basic flow diagram of a commercial oil-agglomeration process for the recovery of fine coal that has been installed in a coal cleaning plant in Pennsylvania, USA [6.10.3.1, B.73]. It uses a low oil level for the production of micro-agglomerates with few impurities. This approach requires both an efficient, high shear mixer to finely disperse the oil and centrifugal drying to reach the product moisture requirements. A skimmer tank was also added to recover very small agglomerates that were lost in the screening operation. After the screen bowl centrifuges, the agglomerate size of the dewatered product is enlarged by rolling the mass in a balling disc pelletizer. The plant is designed to recover 20–30 t/h of clean coal agglomerates from the underflow of a wet coal cleaning plant thickener containing particles with 50 % < 40 lm; this sludge was previously discarded as waste.


Fig. 6.10-38 Flow diagram of a commercial oil agglomeration process for the recovery of fine coal [B.73]

A technique related to immiscible liquid agglomeration in suspensions is using the sol-gel process for the formation of agglomerated spherical oxide fuel particles with a diameter of up to 1 mm for nuclear reactors [B.73]. Fine particles are initially contained in an excess of a bridging phase (sol), which is then dispersed by spraying into an immiscible liquid, forming a droplet suspension. The immiscible liquid is selected such that it extracts bridging phase and causes gelation. The gel particles are removed from the suspension, dried, calcined, and sintered to yield the final product. The immiscible fluid is recovered for reuse.

457

458


Fig. 6.10-39 is the flow diagram of a sol-gel process for the formation of gel microspheres from a water based sol in a fluidized extraction column and the cleaning of the water extracting fluidizing liquid [B.97]. The post-treatment steps (drying, calcining, and sintering) are not shown. In the process the aqueous sol of colloidal particles is dispersed into drops at the top of a tapered vessel. The drops are fluidized by the upward flow of a water-extracting fluid, such as 2-ethyl-1-hexanol. A surfactant is added to the immiscible liquid to prevent coalescence of the droplets, sticking to the walls, or lumping together. As water is removed and the sol is converted to a gel, the particles become denser and settle toward the product collection port. Vessel design and flow rates are controlled such that densified gel particles drop continuously into the product receiver while fresh droplets are added at the top. Extracting liquid is separated from the system and for cleaning. At least part of it is sent to distillation for the removal of excess water. It is possible to further disperse the droplets in stirred and baffled vessels. As compared with the fluidization method depicted in Fig. 6.10-39, smaller gel particles are obtained in such systems.

Fig. 6.10-39 Flow diagram of a sol-gel process for the formation of gel microspheres from a water based sol in a fluidized extraction column and the cleaning of the water extracting fluidizing liquid [B.97]

6.11 Special Applications

6.11

Special Applications Agglomeration is the action or process of gathering particulate solids into a conglomerate, and an agglomerate (another word for particles adhering to each other) is an assemblage of particles, which is either loosely or rigidly joined together with or without featuring a specific shape (Chapter 14). These occurrences or procedures are caused by a natural phenomenon, which may be enhanced by a number of means, such as use of wet and/or dry binders, pressure, and/or temperature. Therefore, literally millions of applications, some happening without intent but still producing beneficial results, are known, carried-out, or occur in all fields that handle and/or process particulate solids. Many applications can be associated with specific industries and, therefore, have been covered in some detail in this book. It should be recognized, however, that all are based on the same fundamentals and could be used for other materials in different industries, often after only minimal application-related modifications (e.g., materials of construction, equipment size, duty). In the following three sections a few examples of special industrial applications of the technology will be described to round-off this book’s compilation. It will be pointed out that agglomeration methods and techniques are often used repeatedly during the preliminary, intermediate, and final manufacturing processes. Although not specifically mentioned in other chapters, this repeated application of agglomeration is a general trend when, in one way or another, fine particulate solids are involved. A further large new special application of controlled size enlargement of small solid items uses the fifth binding mechanisms of agglomeration (Chapter 3, Tab. 3.1 V: interlocking bonds), involving fibers and various additives to achieve specific product characteristics. This technology yields “non-woven” engineered fabrics for a multitude of industrial and consumer applications. 6.11.1


As discussed in several sections of this book (Sections 6.1, 6.2, 6.3 and Chapter 11), many physical characteristics of individual particles improve with decreasing size (Tab. 11.1, Chapter 11). This tendency is often maximized when the solid entities Tab. 6.11-1 Some areas that are cleaned with gas-phase filters (adapted from Purafil, Doraville, GA, USA) Airports Archives Casinos Clean room manufacturing Hospitals Hotels

Laboratories Museums Office buildings Prisons Restaurants

Retail Stores Schools Zoos

459

460


reach nanoscale dimensions where sometimes even chemical properties are changed, producing or allowing new and beneficial reactions. Ultrafine particles (UFP), defined as being in a size range from a few micrometers down to nanometers, feature natural adhesion tendencies, which strongly increase with decreasing linear dimensions or increasing specific surface area. On the one hand, this may be a disadvantage, because nanosized particles always exist as agglomerates (Fig. 11.1, Chapter 11) and, if individual nanoscale particles are required, special

Fig. 6.11-1 Photograph of a pan agglomeration system for gas cleaning media (courtesy Purafil, Doraville, GA, USA)


environments must be provided to produce and keep them separate or costly disagglomeration steps must be added. In bulk, UFP form accretions and adhere to surfaces, and have generally bad handling properties due to low flowability, inaccurate metering characteristics, and lumping. On the other hand, such particles easily form agglomerates which, if sufficient accessible internal void space is present, retain to a large extent the desirable properties of the agglomerate-forming UFP. Such granular products are best made by tumble/growth agglomeration since only small forces are exerted and minimal structural densification occurs. One of the main qualities of these granules is their large specific surface area that is available through open porosity for surface related reactions (as catalysts and catalyst carriers, Section 6.3) while the product is dust-free, free-flowing, and, generally, can be handled effortlessly. Also, such mainly spheroidal agglomerates can be packed easily and reproducibly into static beds in columns or containers of various sizes and shapes. The conditioning and control of indoor environments is an ever increasing problem. Lately this does not only include the traditional control of temperature and moisture but also the elimination of contaminants and odors from the atmosphere. Particulates, including organic matter, such as mold, germs, and viruses, can be captured and retained by sometimes electrically assisted ultrafiltration. Cartridges are either discarded or cleaned/reactivated. For the elimination of chemicals and odors, absorption, adsorption, and chemical reactions with air purification media are required. While diluting the indoor environment with outside air is a simple method to reduce the levels of contamination, gas-phase filtration has been shown to be more efficient and cost effective. Indoor sources of contamination include people, cleaning compounds, disinfectants, emissions from new carpeting and furniture, office equipment, such as copiers and printers, and so on. Nearby exhausts from, for example, industrial manufacturing, furnaces, and incinerators and gasoline or diesel engines can also participate when the fumes are brought inside through building HVAC systems. As microelectronics and microcircuitry continue to develop and businesses are more and more depending on computers for data processing, the control of operations, and the management of manufacturing processes and facilities, indoor environments must be kept very clean to avoid interference of gaseous contaminants with computer controls and the corrosion of electronic components. Tab. 6.11-1 lists some of the applications of these decontamination technologies. While fiber-based filters or other porous media can be impregnated with customized decontamination compounds, these air cleaning units are only suitable for relatively low contamination levels and small air flows. For larger cleaning requirements, powdered carrier and/or active materials (e.g., sodium bicarbonate, activated alumina, activated carbon) are impregnated with chemically active substances (for example potassium permanganate, potassium hydroxide, sodium thiosulfate) and agglomerated by pelleting or pan agglomeration. Fig. 6.11-1 shows a pan agglomeration system, which makes use of the classification effect of this equipment. The quite uniform spherical wet agglomerates are collected in trays, transferred into and passed through an oven with a trolley system, and dried. The cured agglomerates are screened to remove fines and are then ready for use (Fig. 6.11-2).

461

462

6 Industrial Applications of Size Enlargement by Agglomeration Fig. 6.11-2 Vials with different pan agglomerated gas cleaning media (courtesy Purafil, Doraville, GA, USA)

Fig. 6.11-3 depicts a deep bed filtering and decontamination system and a so-called deep bed scrubber, which contains a blower section, one or more deep beds of media in series, and particulate pre- and high-efficiency final filters. Such units sit outside and provide pressurized recirculating air, maximum contaminant removal, and long service life. Fig. 6.11-4 shows the principle of an extended surface system and the artist’s conception of a partially open unit. The large surface of the V-shaped gas filter media module banks increases the contact area in spite of relatively small bed thickness (< 75 mm) and, therefore, features a low pressure drop. The powered or nonpowered unit is designed to filter low to moderate levels of gaseous contaminants in less polluted areas of the process industry. Other units are for cleaning the air in instrumentation rooms or for computer racks, for compressor intake filtration, and many more. One manufacturer lists over 1000 specific gases and 100 general gas categories as a guide for selecting the most appropriate cleaning media. Fig. 6.11-5 is an example of media solutions for contaminant gases, indicating some of the more important pollutants that can be removed.

Fig. 6.11-3 a) Diagram of a deep bed filtering and decontamination system; b) photograph of a so-called deep bed scrubber (courtesy Purafil, Doraville, GA, USA)


Fig. 6.11-4 a) Principle of an extended surface gas cleaning system, b) artist’s impression of a partially open unit (courtesy Purafil, Doraville, GA, USA)

To guarantee the efficient and reproducible manufacturing of the final product, a number of other particulate solids must be granulated, too. Because of their fineness, these materials are difficult to handle but do feature properties that are important for their ultimate application and, therefore, must be retained (Section 6.1). The beneficial characteristics of particulate matter formulations, consisting of several, often widely disparate ingredients, may be even greater after mixing the components in one or more preparatory steps. For these blends, not only the handling problems but also the danger of segregation are a great concern. For example, brake linings contain finely divided, hard, abrasion resistant particles (Tab. 11-1, Chapter 11) in a complex matrix that provides the bonding in the lining itself and with the metallic supports. The formulation, containing all components, may also include such materials as metallic, mineral and, synthetic fibers. During production of the break drums, discs, or shoes, the blend is pressed onto the structural supports with specially designed presses to produce uniform, homogeneous wear pads.

Fig. 6.11-5 Examples of media solutions for some of the more important contaminant gases (courtesy Purafil, Doraville, GA, USA)

463

464


First, however, the mixture must quickly and evenly flow into the die, whereby no segregation can be tolerated. This is only possible if the mass has been pre-agglomerated, which is often accomplished in specially executed batch high shear mixer/agglomerators. Another example is found in the manufacturing of dry-alkaline batteries. The active ingredients of the cathode (positive electrode) in primary (non-rechargeable) batteries are MnO2, either as chemical (CMD) or finely divided ground electrolytic manganese dioxide (EMD), graphite, and a KOH electrolyte solution and a small amount of binder. The ultimate task is to pack a maximum of uniformly distributed active substance into the battery container. Again, to avoid segregation, pre-agglomeration is required. Although, to achieve the desired performance, particularly long life and good cell-discharge efficiency, high density is necessary in the final product and, for that reason, granulation by pressure is in many cases preferred (Section 6.11.2), agglomeration with high shear mixer/agglomerators is also feasible.

6.11.2

Pressure Agglomeration Techniques

The particular advantage of pressure agglomeration over tumble/growth agglomeration is that external forces are applied and act upon particulate solids, causing particles to approach each other more closely, even break or deform, and, depending on the pressure level, result in sometimes considerable densification of the powder mass. The latter may be used to reduce the reactivity of materials, for example of directly reduced, “sponge” metals, such as iron or titanium (Section 6.9.2). Other special applications of pressure agglomeration take advantage of the fact that the solid components of a compacted mixture are in intimate contact. This does, for example, allow the fast progress of chemical reactions from particle to particle or the immediate availability of elements (e.g., oxygen) for the sustenance of interactions between chemical entities (e.g., by oxidation or reduction). In general terms, special pressure agglomeration techniques can yield products with controlled reactivity [B.97]. In the case of explosives or, for example, chemicals that activate airbags, large amounts of gas (products of combustion) are produced during a very fast reaction which, in addition, expand because of high system temperatures thereby creating the destructive forces or the pressure in an airbag. After activating a primary detonation, the chemical reaction of these materials is a rapid decomposition by oxidation whereby the necessary oxygen may be part of the solid system itself or made available from oxygen-rich ingredients, such as chlorates. The effect of an explosive depends on its energy density (the relative stored energy concentration), the amount of energy released during the reaction, and the speed of the reaction. Particularly if oxygen is made available from separate oxygen carriers in a mixture, such as airbag chemicals, it is necessary to provide good contact between the components. Product characteristics can be adjusted by modifying the degree of densification resulting from pressure agglomeration, for example compaction/ granulation. With higher density the stored energy concentration increases but, because the available surface area is


lower, the unassisted reactivity decreases. The reaction of such explosives or chemicals must be initiated with primers. A primer is a highly reactive material that is easily ignited by friction, percussion, or electricity that will detonate the somewhat less-reactive explosive. Another product with precisely predetermined reactivity is the chemical oxygen generator [6.11.2.1]. These “oxygen candles”, which are, for example, installed in aircraft to provide each passenger with oxygen if an emergency arises, must produce oxygen while the plane descends to lower altitudes. The production of oxygen occurs during the thermal dissociation of chlorates and perchlorates. The chemical core in which the reaction proceeds is an agglomerate containing chlorate, fuel, and catalyst that control the decomposition of the “candle” (core) and is ignited by a primer charge that is activated when the oxygen mask is pulled down. Fig. 6.11-6a depicts a cross-sectional view of a typical chemical oxygen generator and Fig. 6.11-6b is the photograph of actual equipment for aircraft that is housed behind the panels above each seat. The heart of such an oxygen generator is the chemical core. By varying the physical shape and the amount of chemicals that make up the core, a wide range of oxygen outputs may be achieved. Generally, the larger the core diameter, the more oxygen is produced, and the greater its length, the longer the duration of the oxygen production. If in a high-flying aircraft, an aneroid, operated by the effect of ambient air pressure on a diaphragm, senses that decompression has occurred, a signal is supplied that

Fig. 6.11-6 a) Cross-sectional view of a typical chemical oxygen generator, b) actual equipment for aircraft (courtesy Puritan Bennet, Lenexa, KS, USA)

465

466


releases the overhead compartment door and the masks are presented within the reach of the passengers. The flow starting mechanism (primer), which is activated by releasing the spring loaded pin while pulling down the mask, initiates the chemical decomposition. As the chemical reaction zone travels along the core from one end to the other (right to left in Fig. 6.11-6a), the oxygen produced flows through the insulation until it reaches the core locator, a metal plate separating the candle from the filter media. The gas passes through holes around the perimeter of the plate into the filter where particulate matter and traces of contaminating gases, resulting from core decomposition (Section 6.11.1), are removed. Clean oxygen is supplied through a manifold and tubing to the mask. The chemical oxygen generators for airline application are designed to produce oxygen at a rate varying with time that coincides with the physiological needs of humans during descent. The requirements are stated in the US FAA (Federal Aviation Administration) regulation FAR 25.1443. Fig. 6.11-7a shows a typical altitude profile (cabin altitude against time after decompression). It is a requirement that immediately after decompression a large amount of oxygen is made available as, typically, the emergency begins at great altitude. Because it is assumed that, after a short period, during which the crew reacts to the situation, the aircraft quickly descends to a denser atmosphere, the production of oxygen can diminish during the “candle’s” 12–25 min burning time. In Fig. 6.11-7b a trace of the actual output of an aviation chemical oxygen generator is indicated on the flow requirement profile. Once the aircraft reaches an altitude of 10 000 feet (about 3000 m), oxygen is normally no longer required and production ceases. The variable production of oxygen is determined by the dimensions, shape, and chemical composition of the core. To make sure that the reaction proceeds uniformly, the core is made by pressure agglomeration. For that purpose, different mixtures of chemicals, which may be pre-agglomerated to avoid segregation and improve metering, are filled in layers into the die of a mechanically or hydraulically operated punch-anddie-press. After compaction, the chemical components of the blend are in very close contact so that the influences of the different layers are only visible as little defined steps in the curve representing the actual production of oxygen in Fig. 6.11-7b. Since a major

Fig. 6.11-7 a) Typical altitude profile (cabin altitude against time after decompression), b) trace of the actual output of an aviation chemical oxygen

generator as compared with the flow requirement profile (courtesy Puritan Bennet, Lenexa, KS, USA)


concern when working with oxygen is fire susceptibility, high-pressure oxygen systems are dangerously sensitive to nearby fires or heat, particularly in aircraft. Chemical oxygen generators will not explode, are a low fire contributors, and even gunfire fragmentation tests resulted in no signs of combustion within the chlorate core. Referring to some of the discussions in Section 6.11.1, manganese dioxide based cathode blends for the manufacturing of dry, alkaline batteries may be preferably agglomerated by compaction/granulation to produce higher density of the dust free, easily handleable, and non-segregating particles. This can be of particular interest if CMD (chemical magnesium dioxide) is used which, contrary to crushed solid EMD (electrolytic magnesium dioxide), is made of porous spherical particles. The interaction of certain types of synthetic graphite, which perform better in battery blends, is also improved by pressure granulation. Ultimately, it is necessary to pack the cathode tightly into the battery container. This is especially difficult for the small (diameter) cells. Therefore, the cathodes of many modern high performance batteries are made from pre-compacted rings. It is necessary to make these rings with high precision, particularly in respect to thickness, since several rings are filled into the container and must fit its defined (standardized) height with only very little adjustment possible during final pressing. Even these rings are made from pre-granulated cathode mass to improve metering and avoid segregation of the components. The application of rings, which feature small density variation, if any, avoid the otherwise marked density gradient that would result from the considerable volume reduction required to compact the pre-granulated mass into the final cathode in one step. In the latter case, the friction on the container wall and on the mandrel that is necessary for the production of a central hole for the anode, does cause the density gradient that is unacceptable in high quality batteries. In Section 6.11.1, the manufacturing of brake linings from pre-agglomerated masses with special presses has already been mentioned. This additional processing is a further application of pressure agglomeration, yielding highly densified brake pads in which all components are well bonded. A post-treatment may be necessary to obtain final properties (Section 6.11.3). Another special application of agglomeration is in the field of polyimides. In this group of polymers thermoplastic, thermosetting, and non-melt processable characteristics exist. The resins of the latter type, in conjunction with fillers (e.g., graphite, MoS2, PTFE) or discrete fibers (chopped or spun glass, carbon), comprise typical molding compounds [6.11.2.2]. Fabrication of these formulations into engineered products occurs in many ways and, for non-melt processable compounds, includes methods of powder metallurgy for the manufacturing of near-net or net shaped parts (Chapter 7.0). Other techniques are conventional injection, transfer, and compression molding and standard plastic extrusion. Stock shapes, produced by these processes, are available as sheets, rods, tubes, plaques, rings, discs, and bars (Fig. 6.11-8) and can be machined (much the same as brass) on standard metalworking equipment [6.11.2.2]. Many products, particularly the often very complex near-net or net shape parts (Fig. 6.11-9), which also include simpler gears, self-lubricating or high temperature bearings, bushings, retainer hubs, roller guides, thrust washers and discs, wire guides with molded-in holes, wear strips, and many more, are made on punch-and-die

467

468


Fig. 6.11-8 Stock shapes (sheets, rods, tubes, plaques, rings discs, and bars) and simple parts (bushings) of Vespel polyimide non-melt processable compound (courtesy, DuPont, Newark, DE, USA)

presses with special tooling. Taking into consideration the superior quality of the engineered items, beginning in some cases with batch sizes of 500 pieces per design, development and manufacturing of new parts become feasible end economical. The process begins with the formulation of the molding compound. Particulate resin is uniformly mixed with specific, application related additives and then agglom-

Fig. 6.11-9 a) Complex near-net or net shape parts for wafer handling and processing, b) IC (integrated circuits) handling and testing and other semiconductor manufacturing made of Vespel polyimide non-melt processable compound (courtesy, DuPont, Newark, DE, USA)


erated by compaction/granulation, mostly, as always, to avoid segregation and improve handling, particularly the metering and packing into the dies of the densification and shaping equipment. Especially if complex parts of the types shown in Fig. 6.11-9 are produced, not only the bulk density and flowability, but also the particle size distribution, the macro- and microsurface roughness and the strength and structure of the granules must be tightly controlled and maintained. Regarding particle size distribution, a sufficiently large amount of “fines” that fill the voids between larger particles, produce a denser packing in the die, reduce the necessary stroke length of the punch, and increase the structural uniformity of the compact should be present. The granule shape and macro- and microsurface roughness will influence interparticle friction and, therefore, the fill density and ease of densification during the manufacturing process. Removing edges and roughness by abrasion in a tumbling drum (Section 6.6.2), with or without de-dusting, may improve the packing characteristics. Lastly, because the structure of the final part must feature uniform and high density, it is necessary that the pre-agglomerated molding compound is totally dispersible (i.e., destructible) under the prevailing pressure in the die during compaction. Therefore, the strength of the granules must meet this requirement. The objective of pre-agglomeration will be to stabilize the uniform mixture of all components and produce not the highest possible abrasion resistance but the best strength for obtaining the optimal structure of the final parts (i.e., a compromise between “high enough for good handling” and “low enough for easy disintegration under pressure”). Compaction/granulation is the most versatile pre-agglomeration technique for this task. By modifying the pressing force the density and strength can be controlled, granule size and distribution is adjustable, using two- or multi-step crushing and screening, and, within limits, shape and surface roughness can be changed. Parts produced by compaction and shaping with the various, previously mentioned techniques are cured by heating in specially designed furnaces. While “stock shapes” (Fig. 6.11-8) can and will be machined to final tolerances, net or near-net shape parts (Fig. 6.11-9) must have so small density variations after pressing that they do not distort during heat treatment. Well granulated molding compounds, taking into consideration the above discussions, and the use of optimally designed dies will yield products that meet this requirement.

6.11.3

Other Technologies

Most of the “other agglomeration technologies” in the field of special applications are related to the use of nanoparticles and to coating, without but, increasingly, with nanoparticles and techniques. Agglomeration by heat, mostly as a post-treatment, is also commonly present. Another large new special application of controlled size enlargement of small solid items uses the fifth binding mechanisms of agglomeration (Chapter 3, Tab. 3.1 V: interlocking bonds), involving fibers and various additives to achieve specific product characteristics.

469

470


It has been repeatedly mentioned in the previous chapters that one objective of agglomeration, particularly during early preparatory manufacturing steps, is the stabilization of a uniform mixture of different dry solids to avoid segregation during storage, handling, and further processing. In addition to the conventional granulation methods, in which binders are added or inherent binding mechanisms are activated to achieve bonding between the particles, a few years ago the term “ordered mixture” was created to describe blending of particulate solids in the presence of interparticular attraction with and without forces causing organization [6.11.3.1]. Nanoscale particles easily agglomerate naturally and, therefore, make mixing of, for example, two nanosized solids extremely difficult (Chapter 11). If the particles of one component, suspended in a suitable fluid, are electrostatically charged with the same polarity, for example tribo-electrically by using high shear forces [6.11.3.1], they become optimally dispersed due to repulsion between the entities. A second component of equal size and similarly prepared but with opposite polarity, can then be mixed uniformly with the first material whereby the two types of particles attract each other. This is best achieved in a liquid environment because van-der-Waals forces are lower in liquid than in air and the electrostatic charges prevail. Because the electrostatic charges dissipate after mixing, natural re-agglomeration occurs, which stabilizes the blend. As shown in Fig. 6.11-10, re-agglomeration by aggregation also occurs with time after only one type of nanosized particulate solids (in this case SiO2, Aerosil OX50, manufacturer Degussa, Chapter 11 and Section 15.1) was first electrostatically dispersed. Fig. 6.11-11 shows another result of electrostatically assisted mixing. The coating of relatively large spherical lactose with nanosized SiO2 particles (Aerosil R972, manufacturer Degussa) occurs if the lactose is charged positively and the Aerosil negatively [6.11.3.1]. The result is an interactive, homogeneous particle blend. Small particles adhering to the surface of larger solid entities, which, prior to this modification, featured unfavorable flow properties, do improve flowability because they impose a distance between the large particles and avoid sticking.

Fig. 6.11-10 Re-agglomeration (aggregation) of electrostatically dispersed SiO2, Aerosil OX50 (manufacturer Degussa): left) original sample;

center) at the end of dispersion/charging; right) 10 min after dispersion/mixing [6.11.3.1]


Fig. 6.11-11 SEM photograph of relatively large spherical lactose with immobilized (adhering) nanosized SiO2 particles [6.11.3.1]

A purely mechanical modification of surface structure and characteristics of particles is accomplished by mechanofusion and hybridization [B.48, B.97] (Chapters 5.0 and 11.0). With these technologies nanosized particles are embedded into or coated onto core substrates. As discussed above, flowability of the solid cores may be improved but, depending on the properties of the adhering substances, many other characteristics, such as wettability, electrical conductivity, magnetic properties, color perception, dispersibility, solubility, etc. can be changed, too. Generally, as shown in Fig. 6.11-12 the technologies are used to alter particle shape and structure, allowing particle engineering. Recently, another method was developed that can be used for particle modification by allowing dispersion and processing of powders of 0.1–50 lm by fluidization [6.11.3.2]. As depicted in Fig. 6.11-13, left, conventional fluid bed systems are unable to disperse powders < 50 lm because the attraction forces between the particles are greater than the fluidizing forces exerted by the gas (drag and buoyancy). A new rotating fluid bed processor, termed Omnitex by the manufacturer (Fig. 6.11-13, right), consists of a plenum chamber, a cylindrical porous drum as gas distributor, and a filter in the center of the drum (Fig. 6.11-14). The porous drum is made of sintered stainless steel with a pore size between 1 and 20 lm, depending on the application, and rotates within the chamber to create centrifugal forces of up to 50g. Fine powder is placed into the drum. During operation, due to centrifugal forces, the particles move to the inside wall where they form a layer and (hot) fluidizing gas flows radially inward through the porous drum. By proper selection of the rotational speed of the drum and the gas flow rate, sufficiently high drag forces are created to overcome the centrifugal and cohesive forces, thus dispersing the particles and creating a turbulent movement. A spray nozzle may be installed in the drum for the addition of coating or binding agents as required. Product is collected on the internal filter. From time to time during operation, particles and fines are reintroduced into the process by a compressed air blow-back action and, at the end of the cycle, product is removed from the filter using

471

472


Fig. 6.11-12

Different possible effects of mechanofusion [B.48, B.97]

the same mechanism. In addition to the conventional powder modifications to improve flowability, produce direct tabletting formulations (Sections 6.2.1 and 6.2.2), coat particles or agglomerates to produce drug delivery systems or mask taste (Section 6.2.3), other surface alterations, the manufacturing of synthetic composite materials, and many more fine powder processing techniques are feasible with this method.


Fig. 6.11-13 Comparison between the fluidization mechanisms in a conventional fluid bed and the rotating Omnitex FB processor (courtesy Nara, Tokyo, Japan)

In Section 6.11.1, the manufacturing of brake linings from pre-agglomerated masses with special presses has already been mentioned. After application of the pads by pressing, a thermal post-treatment is required in most cases to obtain final bonding and properties. A similar thermal post-treatment is also necessary to obtain the uniform, high product quality of polyimide stock and engineered parts (Section 6.11.2). So far in this book, the binding mechanism V: interlocking bonds (Tab. 3.1, Chapter 3) has been mentioned in connection with the sometimes heat-induced plastic deformation during press agglomeration (Section 6.6.2) and the application of fibers for the reinforcement of agglomerates (Section 8.2) and for assisting disintegration of

Fig. 6.11-14 Diagram of the principle of the Omnitex FB processor (courtesy Nara, Tokyo, Japan)

473

474


compacts in liquids (Section 6.2.2) ([B.97]). With the exponential increase in the production of both natural and artificial industrial fibers during the 20th century, in the second half of the century, inspired by military developments for and during World War II, a number of fibrous materials were no longer woven but consolidated and bonded in different ways. This technology yields “non-woven” engineered fabrics for a multitude of industrial and consumer applications [B.107]. Although one type of fiber-based agglomerate, paper with all its derivatives, which is, technically, a non-woven material, dates back to the beginning of the 2nd century in China and the methods of paper making were further developed and improved during all the following centuries, real advances were only made during the past 100 years. The origins of other non-wovens resulted from the recycling of fibrous wastes and low quality fibers left from industrial processes such as wood processing for the manufacturing of cellulose, weaving, or leather processing and also from raw material restrictions during and after World War II. Therefore, non-wovens go beyond the limits of paper products or textiles; the fibrous web may be also made of plastics and foils and their humble beginning is responsible for two misconceptions that are still lingering today: they are often assumed to be (cheap) substitutes and/or associated with disposable items, which are cheap and of low quality. Non-wovens neither depend on the interlacing of yarn for internal cohesion nor do they have an organized geometrical structure. Non-wovens are the result of relationships between individual fibers and feature new properties. Therefore, they are by no means cheap or substitutes but high quality materials in their own right offering special characteristics and performances. Mostly due to their irregular, random but relatively permeable structure, non-wovens are applied in filters, hygiene, health care, cleaning, household, and automotive articles, agriculture, civil engineering, food wrap and packaging, and clothing to just name a few areas of end-use. Tab. 6.11-2 is a more detailed compilation. For the manufacturing of non-wovens, there are three main routes to web forming. *

* *

The dry-lay processes with carding or airlaying of natural, man-made, and inorganic fibers. The wet-lay methods using short fibers and cellulose. The polymer-based web bonding originating with particles from which fibers or film are formed.

Fig. 6.11-15 is a diagram of these routes indicating the applicable raw materials and their preparations, the web forming and bonding, the post-treatments (processing of non-wovens) and the type of products [B.107]. Since their early development about 50 years ago, many different non-woven products have been developed for use in a wide variety of applications (Tab. 6.11-2). In spite of this relatively short history, so much information is already available that it goes beyond the scope of this book to describe in detail the various manufacturing methods and the special properties of the products. Reference is made to a recent book that is a coherent presentation of the present state of the art. It is available in both German and English [B.107].

6.11 Special Applications Tab. 6.11-2 [B.107])

Some current application of non-wovens (adapted from

General area

Specific applications

Filtration

Dry air and gas, HVAC and process air technologies, particle filters, gas sorption, liquid particle filters, surface filters for > 5 mg/m3, deep filters for low (< 1 lg/m3) and average (10–500 lg/m3) concentration. Baby diapers, feminine hygiene products, adult incontinence items, dry and wet pads, nursing pads, nasal strips, etc. Surgical drapes, gowns, and packs, face masks, dressings and swabs, bag liners, etc. Bed and table linens, furnishings, fabric softeners, filters, food wraps, tea bags, scouring media, wipes, dusters, etc. Trunk liners, molded hood liners, heat shields, airbags, tapes, trim, decorative fabrics, oil and air filters, etc. Rooting pads and pots, soil stabilization, etc. Asphalt overlay, drainage pads and pieces, sedimentation and erosion control, etc. Roofing and tile underlay, thermal and noise insulation, floor pads, house wrap, etc. Cable insulation, abrasive pads, reinforced plastics, battery separators, satellite dishes, artificial leather, coating, etc. Leisure and travel, school and office, interlinings, insulation and protective clothing, industrial work wear, chemical defense suits, shoe components, etc.

Personal care and hygiene Health care Home, household, cleaning Automotive Agriculture Geotextiles Construction Industrial Clothing

Nevertheless, a few statements are in order to show how non-wovens are truly a part of agglomeration. The main characteristic of all non-woven products is that fibers, with a wide range of properties and originating from a multitude of sources, are at first loosely and, most importantly, irregularly deposited to form a 3D web. For many applications layers, made-up of different fibers with specific characteristics and forming distinct structures, are laid down. For example, Fig. 6.11-16 is a SEM micrograph of the cross section through a laminated non-woven product that is composed of synthetic leather and microfiber non-wovens. This household cleaning material combines high cleaning performance on glossy surfaces with smear free drying due to the presence of an absorbent core [B.107]. Many other high volume consumer products make use of one or more core layers with absorbent properties. Among the most quickly and highly developed applications are baby diapers, which use super absorbent fibers. Super absorbent polymer (SAP) molecules can trap and hold hundreds to thousands of times their own weight in fluid, ultimately forming a gel. The super absorbent core layer in a diaper is between a nonwoven cover stock, a one or two layer non-woven fluff/pulp sheet that takes up, distributes, and draws liquid into the core, and a microporous back sheet. In addition, elastomeric materials and waterproof elements are incorporated. The super absorbent core not only stores liquid but actively pulls moisture out of the damp or even wet fluff/ pulp, thus leaving the contact areas soft and dry.

475

476


Fig. 6.11-15

Diagram of the manufacturing routes to non-wovens [B.107]

While laying down (dry or wet) one or more fiber layer(s), a weak bonding occurs by entanglement between the fibers (Tab. 3.1, Chapter 3), which can be enhanced by crimping or looping the fiber strands. Particulate components and/or other solid or liquid additives that provide special properties (e.g., odor absorbents or perfumes, anti-bacterial substances) or later act as binders during a post-treatment are also incorporated during web forming. To obtain final and permanent bonding, structure, and strength, finishing processes are required prior to shaping, cutting, packing, and distributing the product. Post-treatments include shrinking, compacting and creping, glazing and calandering, embossing and goffering, molding and stamping, and a

Fig. 6.11-16 SEM micrograph of the cross section through a laminated non-woven household product that is composed of synthetic leather and microfiber non-wovens. Layer sequence (from top to bottom): macroporous synthetic leather surface, non-woven absorbent core, microfiber non-woven [B.107]


Fig. 6.11-17 Summary of the different bonding processes as described in an ISO standard (ISO/DIS 11224 [B.107])

multitude of chemical finishing methods and web bonding by needling, stitch bonding, knitting, and drying, heat setting, ultrasound and/or chemical methods to activate binders. Bonding is very important for the performance and life of non-wovens [B.107]. Fig. 6.11-17 is a summary of the different bonding processes as described in an ISO standard. While needling, such as stitch bonding, knitting, and other, similar techniques, uses special tools to reorient some of the fibers or introduces continuous filaments or threads in a sewing fashion, other bonding processes are based on friction, cohesion, and adhesion forces between fiber surfaces. A cohesive bond occurs between two identical web fibers and, therefore, no binders are present while in adhesively bonded non-wovens, cross-linked or coagulated binder fluid, solidified droplets, originating from binder fibers or powders, attach the matrix fibers to one another. Fig. 6.11-18 shows schematically some possible bonding sites. Non-wovens that contain binder fibers“,4› have both cohesive and adhesive bonds. The latter form after melting or softening and are predominantly bonds at fiber intersection points (Fig. 6.11-18c). In Fig. 6.11-19 two microphotographs of non-wovens depict fiber bonding.

Fig. 6.11-18 Diagram of some possible non- woven bonding sites: a) large area, enveloping fiber intersection points, b) small area and punctiform bonding, c) bonding of fiber intersection points [B.107]

477

478


Fig. 6.11-19

Two micrographs of non-wovens depicting fiber bonding [B.107]

Further Reading

Once again, for further reading the book “Nonwoven Fabrics – Raw Materials, Manufacture, Applications, Characteristics, Testing Processes” [B.107] is recommended.

479

7

Powder Metallurgy Powder metallurgy (PM) is a method for manufacturing ferrous and non-ferrous industrial parts. Although relatively young, the technology is highly developed and very reliable. In accordance with its great importance for modern high-tech metal products and components, many publications are available on the subject (see Further Reading) and everywhere in the developed world scientific groups, industrial and trade organizations, and other institutions exist that promote the knowledge and applications of powder metallurgy. Many are easily accessible, for example: www.mpif.org; www.ifam.fraunhofer.de; www.PowderMetalWeb.com; www.ipmd.net. Tab. 7.1 lists the advantages of the PM process and products. The technology is cost effective because it produces parts, simple or complex, at or very close to final dimensions. This is often called “near-net-shape” production (Section 6.7.2). As a result, only minor, if any, machining is required making it a “chipless” metal working process. Typically more than 97 % of the starting material (metal powder) is in the finished part. Because of this, powder metallurgy is a process that saves energy and raw materials. Most PM parts are small and even larger ones weigh typically less than 2 kg, although pieces of up to 15 kg can be fabricated with conventional PM equipment.

Tab. 7.1 *

Advantages of the P/M process and of parts made by it (adapted from MPIF, Princeton, NJ, USA)

Eliminates or minimizes machining (near-net-shape) * Eliminates or minimizes scrap losses * Maintains close dimensional tolerances * Produces good surface finishes

*

Permits wide variety of alloy systems * Facilitates manufacture of complex or unique shapes which would be impractical or impossible with other metal working processes * Provides materials which may be heat treated for increased strength or increased wear resistance * Metallurgically immiscible materials can be combined to produce uniform structures and “exotic alloys” that are not attainable by pyrometallurgy

*

Provides controlled porosity for, for example, self-lubrication or filtration

*

Offers long term performance reliability

*

Suited to moderate and high volume production requirements

*

Cost effective

Agglomeration in Industry. Wolfgang Pietsch Copyright ª 2005 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN 3-527-30582-3

480

7 Powder Metallurgy

Fig. 7.1 Some complex parts produced by powder metallurgy (courtesy of Komage-Gellner, Kell am See, Germany, and Dorst, Kochel am See, Germany)

7 Powder Metallurgy

Production rates range from a few hundred or less to several thousand parts per hour. While many early powder metallurgical items, such as cutting tool tips, bushes, or bearings, had very simple shapes (cubes, rectangular blocks, discs, rings, cylinders), parts with complex contours and multiple levels are produced economically today (Fig. 7.1) [B.28, B.48, B.97]. Many gears, cams, and intricately shaped parts that would require expensive machining when produced from cast, forged, or wrought stock can be made from metal powders. Counterbores, flanges, hubs, and holes as well as keyways, keys, D-shaped bores, and other fastening devices can be an integral feature of the part and two or more parts may be combined into a single unit if product design permits. Compared with other metallurgical manufacturing processes, PM offers the additional advantage of precise control. Powder metallurgists are able to exercise their influence over the entire production process from the pure metal to the equally pure powder, the (agglomerated) pre-form, and the finished part. By mixing metal powders, which may be of different elemental or alloy origin, size, and/or shape, various material compositions and finished products can be created. Powder metallurgical manufacturing eliminates impurities and inclusions, avoids uneven internal stresses, and excludes poor finishes, unworkable tolerances, and many other factors that might affect the rate of production and the quality of the finished part. PM assures uniformity and optimum performance characteristics of products and offers long-term performance reliability in critical applications. As shown in Fig. 7.2, after the formulation and blending step, shaping (forming) of the part and the development of strength are accomplished with methods of agglomeration by pressure and heat followed by a choice of post-treatment steps, if required. While powder production is accomplished by the atomization of pure molten metals or alloys [B.13d and Section 13.3, Refs. 82, 85, 87, 89, 90, 91, 93, 94], the manufacturing process for parts occurs in the solid state. Therefore, metallurgically immiscible materials can be combined to produce uniform structures and “exotic alloys” that are not attainable by melt methods. Dissimilar metals, non-metallics, and other components of widely different characteristics can be mixed, compacted, sintered, and further processed, if required, into components that exhibit unique properties. For example: ceramics can be blended with metals to make cermets (Section 6.7.2). Carbon and copper form electrical brushes with high electrical conductivity and wear resistance. Tungsten and silver are combined in the manufacture of electrical contacts and switch gear. Copper, tin, iron, lead, and graphite are compacted into heavy-duty friction material. The combination of materials through the use of PM techniques is essentially unlimited. Application for a new task only requires research and experimentation. Since powder metallurgy is a technology in its own right, this book is not going into more detail. As far as agglomeration tools for the densification and shaping of the powders and of powder blends are concerned, Fig. 7.2 classifies the methods under hot and cold compaction. The equipment available and used for these tasks has been described in detail in an earlier book by the author [B.97]. In many respects the shaping and densification of metal powders is similar to that of high-performance ceramics and many of the presses can be used for both applications (Section 6.7.2). The same is true for sintering. Here too, some related material is included in Section 6.7.3

481

482

7 Powder Metallurgy

Fig. 7.2 Process diagram of powder metallurgy (courtesy of MPIF, Princeton, NJ, USA)

7 Powder Metallurgy

and the author’s earlier book [B.97]. Additional information can be gleaned from the numerous websites associated with the search word “powder metallurgy” and its foreign language forms as well as from the books listed at the end of this chapter. As can be seen from the process diagram in Fig. 7.2, hot compaction combines the steps of densifying, shaping, and sintering in one piece of equipment. For many applications it is the modern way of accomplishing the task. In most cases, after hot compaction only additional post-treatment (optional operations) is necessary if required. Isostatic pressing, either hot or cold [B.13a, B.97], results in the most uniform structure, which minimizes distortion during heat treatment (sintering) that is used either for the development of strength or the modification of density (reduction of porosity) during re-sintering (Section 6.7.2). Punch-and-die pressing is used for the mass production of simple small parts but also for complex symmetrical designs (Fig. 7.1) [B.28]. Rolling with roller presses is a relatively new process for the manufacture of sheet with special properties from powders that are often mechanically alloyed and could not be produced by traditional methods.

Further Reading

For further reading the following books are recommended: B.4, B.13a, c, and d, B.28, B.47, B.57, B.65, B.76, B.78, B.79, B.84, B.85, B.86, B.88, B.95, B.96, B.100, B.101 (Chapter 13.1). Books mainly devoted to the subject matter are printed bold.

483

485

8

Applications in Environmental Control Minimization of solid wastes, pollution control, and conservation of natural resources have become important and popular topics in all countries of the world. We live in an age of increasing government intervention, resulting in environmental laws, regulations, and enforceable penalties for offenders. Legal federal, state, and private citizen actions are taken increasingly often. They often result in considerable fines and sometimes even jail for those found responsible. Every year billions are spent to capture, contain, and remedy solid wastes after their generation, and costs for “hazardous waste” treatment and disposal have multiplied during the past few years because of the limited availability of treatment and disposal sites and facilities.

Tab. 8.1 Origins of particulate wastes and sources of pollution (presented in alphabetical order) Primary wastes * Agricultural wastes * Excrements * Mining wastes Processing wastes Chemical processing wastes * Metallurgical processing wastes * Mineral processing wastes * Vegetable and other food processing wastes * Waste by-products * Wood processing waste *

Production wastes Energy production wastes * Food production wastes * Waste by-products * Wastes from the forming, shaping, and finishing of solid products *

Secondary wastes Wastes created during the disposal of wastes * Wastes created during the handling of wastes * Wastes created during the processing of wastes *

Agglomeration in Industry. Wolfgang Pietsch Copyright ª 2005 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN 3-527-30582-3

486

8 Applications in Environmental Control

Most solid wastes are particulates with an increasing portion in the fine particle size range. As shown in Tab. 8.1, they are produced during primary activities, the processing of raw and advanced materials, and the handling, processing, and disposal of wastes from previous sources. Agricultural wastes include, for example, straw. Formerly straw was burnt but this practice has been outlawed in many regions. Excrements used to be applied as a natural fertilizer. Today, concerns about public health forbid the direct application of raw feces, manure, and sewage on farmland or gardens in most developed countries. People have mined the earth for thousands of years. Originally only high-grade ores and minerals were removed for further processing and use. Many of the oldest mine sites can be identified by the associated dumps of low-grade or undesirable material. Even today, tailings, below-grade slurries, and slimes that remain after high-efficiency upgrading and may now also be contaminated with chemicals, are still frequently disposed of in ponds or other deposits from which ground water is threatened by seepage. During all processing steps, wastes develop featuring very diverse characteristics. In chemical processing, the solids are often very fine (dispersions), toxic, or hazardous. Metallurgical processing produces large amounts of fine (often metal-bearing) dusts. In mineral processing, size reduction is an important step that yields fines with particle sizes that are too small for direct further use and particulate, often contaminated, tailings are discarded from concentration plants. Vegetable, food, and wood-proces-

Tab. 8.2

Origin, forms, and characteristics of particulate solid wastes

Origin * Single-component or composite materials * Primary or secondary (Tab. 8.1) Forms (organized by size) Large and/or bulky parts * Lumps and pieces * Granular materials (0.1–10 mm) * Fibrous materials (turnings, borings, organic fibers, etc.) * Dusts (0.1–100 lm) * Ultrafine (nano) particles (0.1–100 nm) * Cakes, sludges, slurries, and slimes * Dispersions: suspensions (solids in liquid), smokes (solids in gas) * Ultrafine particulate systems (colloids and aerosols) *

Characteristics (arranged alphabetically) Hazardous (by composition, definition, nature, or size) * Inert * Obsolete * Organic * Radioactive * Reactive * Toxic * Useless * Valuable *

487

sing facilities require the disposal of large amounts of organic matter, which is today summarily called “biomass”. By-products, such as molasses from sugar making, lignins from paper mills, oils and tars from the coking of hydrocarbons, or slags from metallurgical processing may be considered wastes because either their amounts are too large, contaminations prohibit their direct use, or, more recently, legislation requires mandatory processing, often to reduce secondary pollution and/or health hazards. In many of these by-products, the particulate solids are dissolved or suspended in a liquid phase. Similarly, various particulate wastes develop during the production of industrial or consumer products. In particular, the shaping, forming, and finishing of articles from solid material yields turnings, borings, chips, grindings, and dust. By-products, for example gypsum or ammonium sulfate obtained during flue-gas desulfurization as required by “clean air acts”, are increasingly obtained in response to anti-pollution legislation. The production of food, such as canning and fish or meat packing, again results in large amounts of biomass, which must be dealt with in an environmentally safe manner. The broad category “energy production” comprises wastes from all fuelpowered motors and generators of any kind and size. Secondary pollution is a cause for great concern during the handling, processing, and disposal of solid refuse. Since primary waste materials are often fine, they tend to

Tab. 8.3 Possible effects and applications of size enlargement by agglomeration in environmental control (A) Collection of contaminants * Improvement of the collection efficiency after agglomeration by – Particle contacts in turbulent flow regimes – Sonic agitation – Coalescence in the presence of moisture – Flocculation by stirring with or without polymer addition. (B) Handling of particulate solid wastes * Elimination of dust by size enlargement * Increased bulk density or decreased bulk volume * Improved flow, including derived characteristics of bulk materials (metering) * Optimized or modified product shape. * Reduced reactivity (C) Processing of particulate solid wastes * Same as (B) * Additionally – Modification of properties (strength, porosity, dispersibility, solubility, reactivity) – Adjustment or engineering of composition. (D) Disposal of particulate solid wastes * Same as (B) * Additionally: – Production of permanently strong and/or dense pieces – Encapsulation of toxic or radioactive materials – Achievement of leach-prove bonds

488


dust if they become dry, may be washed away with liquids, or seep into the ground where they may contaminate soil and aquifers. Tab. 8.2 lists typical origins, forms, and characteristics of particulate solid wastes. They must be considered during handling, processing, and disposal. Solid wastes may consist of materials containing one single component or of composites comprising two or more substances. They may be primary discards, such as domestic refuse and solids contained in sewage, or secondary wastes such as ashes from refuse incinerators and digested filter cakes from municipal waste-water treatment plants. Secondary pollution can also arise from the emanation of dust or ultrafine particles from primary wastes. The dimensions of solid wastes span the range from meters (for example automobile scrap and obsolete structures and equipment) to nanometers (in particulate residues from combustion and products of sublimation). Originally, the removal of particulate pollutants from liquids and gases was accomplished by settling or with relatively simple filters and collectors. Ultrafine particles are so tiny that they can not be observed with the eye or by light microscopy [B.70]. Also, because of their small mass they remain suspended in gases (especially in air) and dispersed in liquids. Furthermore, they follow the streamlines of flowing fluids and so can not be captured by conventional means. Electron microscopy made such particles visible and relatively recent research revealed that very fine particles enter the respiratory tract during breathing or are resorbed from drinks, causing previously unrecognized health risks (for example silicosis and asbestosis or intestinal infections). For these reasons, the characterization of hazardous solid materials has changed and, with it, the requirements for pollution control. In addition to its nature, described by certain properties (e.g., radioactive) and compositions (e.g., containing toxic components), legislation often defines the term “hazardous material” by “below a certain particle size”. With this definition, totally inert materials, consisting of or containing ultrafine particles, become hazardous by law and require processing. Since in most cases the dimensions of particulate solid wastes, particularly those of contaminants, are small and, therefore, cause problems during collection and handling and with secondary pollution, size enlargement by agglomeration offers technologies that can assist in solid waste handling, processing, and disposal in many different ways. Tab. 8.3 summarizes the possibilities. Increasingly, wastes are no longer just disposed of but processed for direct recycling or as secondary raw materials. With these technologies environmental protection laws are fulfilled, raw material sources are conserved, problem-causing waste deposits or extensive land fills are avoided, and, over all, energy is saved. This results in considerable reductions in costs and increases in profit potentials in all industries.

8.1 Collection, Stabilization, and Deposition of Particulate Solids

Further Reading

For further reading the following books are recommended: B.3, B.7, B.8, B.15, B.16, B.18, B.19, B.20, B.21, B.22, B.24, B.25, B.26, B.35, B.36, B.37, B.40, B.43, B.46, B.48, B.49, B.51, B.55, B.56, B.69, B.70, B.83, B.87, B.89, B.93, B.94, B.97 (Chapter 13.1). Books mainly devoted to the subject matter are printed bold.

8.1

Collection, Stabilization, and Deposition of Particulate Solids 8.1.1

Size Enlargement by Agglomeration during the Collection of Particulate Solids

Most effluents from processing plants of any kind are contaminated in one way or another. Contaminants may be gaseous, liquid, or solid. Gaseous pollutants are often removed by chemical reaction, which may result in a solid by-product that needs to be processed, for example the removal of sulfur from the flue gas of power plants, or they are burned, which may produce particulates (soot) that become an increasing environmental concern. Liquid contaminants may be precipitated and then also form finely dispersed, suspended particles. While for quite some time particles with sizes down to a few hundred micrometers have been successfully removed by conventional dust collection (mostly using gravitational and centrifugal or other field forces) and dry and wet filtration, air borne (Fig. 8.1) and suspended solid pollutants continued to be a great problem. Because of the small mass of fine and ultrafine particles they do not settle, even if high-field forces are produced, and they follow the flow lines in filter media so that impacts, which are necessary for collection (Fig. 8.2), do not take place. For a long time, natural agglomeration has helped to improve the collection efficiency of, for example, filters and cyclones. Knees, bends, restrictions (e.g., valve seats), and even build-up in the pipes and channels of the cleaning device caused turbulence, particle collisions, and the formation of aggregates that, as a whole, have the combined mass of all the particles adhering to each other and a correspondingly larger cross section. Both effects allow the removal of particles that, individually, would be too small (Fig. 8.3). Originally, the introduction of such turbulence was unintentional and led to sometimes surprising results: low efficiency if short straight connection lines are used between the source of particulate pollution and the collection device or unexpectedly high efficiency if, often for design purposes, tortuous, twisting lines had to be installed. More recently, process designers have applied this knowledge in the routing of effluent lines to assist dust collection. Ultrafine particles that are suspended in a fluid exhibit Brownian motion, a random movement resulting from the impact with molecules of the fluid surrounding the particles. In spite of the randomness of the motion, it is very unlikely that particleto-particle impacts will occur because the amplitudes of the movement and the par-

489

490


Fig. 8.1 Example of an industrial plant (pyroprocessing of minerals) heavily polluted with airborne particulate solids

ticle sizes are very small. Therefore, other methods must be used to cause particle collisions that will result in coalescence and agglomerate growth. In addition to artificially induced turbulence, such techniques use uni- or bipolar charging in so-called electrostatic precipitators, magnetism, and ultrasound [B.48]. Although, from a process point of view, it is very advantageous that with these methods particle accretion occurs in the free-flowing fluid, a major drawback is that the collision probability changes with the square of the particle concentration. Therefore, as the number of particles becomes smaller, during the cleaning process itself by the incorporation of dust particles into the growing agglomerates, the collision probability decreases and the desired low concentrations of ultrafine particles in the off-gas may not be reached.

8.1 Collection, Stabilization, and Deposition of Particulate Solids Fig. 8.2 a) Enlarged view of the interior of a filter mat with particles sticking to the fiber surfaces. b) Detail of a dust laden fiber from (a) showing how particles extend into the gas stream [B.71]

Fig. 8.3 Naturally formed agglomerate of small (8 lm) glass spheres adhering to a filter fiber photographed during a laboratory experiment [B.48, B.97]

491

492


Fig. 8.4 String-like agglomerates of “brown smoke” particles produced by natural magnetic coagulation [B.48]

For a long time, images similar to the one shown in Fig. 8.1 could have been seen at any steel-making complex in the world. The air was heavily polluted with “brown smoke”, consisting mostly of submicron c-Fe2O3 particles that are individually too small for collection by the early, rather basic gas-cleaning systems. After the invention of high-efficiency cyclones and multi-clones which, for physical reasons (low particle mass), can not collect submicron particles from contaminated gas, and their use in the steel and other metals or mineral-processing industries, it was found that, nevertheless, some of the brown smoke and similar airborne dust was removed anyway. In the meantime, it had become possible to observe ultrafine particles with the newly developed transmission electron microscope (TEM). It was determined that, because they are ferromagnetic dipoles, c-Fe2O3 particles naturally attach to each other, resulting in string-like agglomerates (Fig. 8.4). The larger of these aggregates were removed but an objectionable reddish plume was still discharged. The ultimate clean-up became possible after the naturally produced agglomerates (Fig. 8.4) were charged in an electrical field or by spray electrodes causing them to grow into larger dendritic structures (Fig. 8.5) that can be captured and collected [B.48]. It must be a goal during the collection of particles in filters that they collide with the medium and adhere upon impact. As shown by the results of model calculations,

Fig. 8.5 Dendritic growth of “brown smoke” agglomerates (Fig. 8.4) in an electrostatic field [B.48]


various particle sizes not only move differently but modified system conditions also influence the behavior of the particulate solids, for example if particles and/or fibers carry a natural or acquired electrical surface charge, which causes electrostatic forces [B.97]. Another new strategy for the removal of ultrafine contaminants from a gas uses a particulate collector medium that is fluidized in a suitable vessel by the gas to be cleaned. Aerosol particles adhere to the large surface area of the fluidized medium and form a coating that is densified when the collector particles collide with each other. As the coating becomes thicker, attrition results in the formation of secondary particles. These are agglomerates and substantially larger than the original aerosol so that they can be easily separated in conventional dust collectors [B.97]. Like many original methods of mechanical process technology, an old technique for the successful removal of airborne dust was first observed in nature. The capacity of rain to “clear the air” has been used since ancient times to remove suspended dust particles by passing the contaminated gas through a water spray (wet scrubbers). The liquid droplets capture the solid (and some of the gaseous) pollutants and collect them in a sump. While this reduces air pollution it transfers part of the separation problem to a secondary cleaning process: the removal of fine particulate solids from a liquid. Although these pollution control devices are cheap and efficient, it becomes more and more difficult to satisfy the increasingly stringent legislation of environmental control that is introduced almost everywhere. Traditional sprinkler sprays that require pumps and high-pressure water systems with substantial water usage, tend to exacerbate the problem by producing contaminated water runoff, potential spillage, and more high-cost clean-up. This is of particular concern if the locations of the sources of pollution and, with it, the collection points are not stationary. For example, dockside mobile loaders (DML, Fig. 8.6) are used during the unloading and transfer of dusty materials [8.1.1]. To avoid the aforementioned disadvantages, a new development causes dust suppression by wet agglomeration and accomplishes collection of the enlarged particles with fabric filters in the same unit. For the wet agglomeration step, an almost dry fog is produced by an air or (inert) gas-driven oscillator. This device atomizes liquids by passing them through high-frequency sound waves. The air or gas expands in a convergent section (nozzle) into a resonator cap where it is reflected back to complement and amplify the primary shock wave. The result is an intense field of sonic energy focused between the nozzle body and the resonator cap. Any liquid that can be pumped into the shock wave is sheared and forms fine droplets. Gas bypassing the resonator carries the atomized droplets away and forms a soft plume. The droplets have low mass, a low velocity, and ensure uniform distribution of liquid with minimum overspray and waste. It was also found that large, relatively fast-moving droplets (from sprinklers) displace so much air that the flow around it prevents dust particles from contacting the droplets (Fig. 8.7, left) while the fine dust particles easily impact small droplets, triggering agglomeration (Fig. 8.7, right; see also Chapter 5, discussion of the importance of droplet size for wet agglomeration). Sonic energy can be also used to initiate and/or accelerate dry agglomeration of ultrafine particles that are suspended, for example, in flue gases [B.43, B.48]. An acoustic field imposes sound pressure and energy. For a typical pressure of 160 dB the acoustic velocity is about 5 m/s and a typical frequency of 2000 Hz causes a fully en-

493

494


Fig. 8.6 Loading of a truck by heavy duty clam shell grab via dockside mobile loader (DML) to avoid pollution of the environment [8.1.1]

trained particle to move back and forth 2000 times a second over a distance of about 600 lm. Particle entrainment is explained by a factor that is a function of acoustic frequency, particle diameter and density, and the dynamic viscosity of the gas [B.48]. If the factor is 1, full entrainment is achieved, while with a value of 0, no entrainment occurs. The latter means that such particles are not affected by the acoustic field and remain still. For each acoustic frequency a particle size exists below which particles are so much entrained (factor > 0.5) that they are sufficiently moved. For example, for a sound frequency of 2000 Hz, this “cut size” is about 4.5 lm. The flight paths of larger particles remain essentially unchanged, while smaller particles move with large displacements, colliding with the large particles, and adhering because of high van-der-Waals forces.

Fig. 8.7 Demonstration of the importance of droplet size for particle agglomeration


In the hot gas cleaning system of a coal-burning power plant, acoustic agglomeration could be installed between the first cyclones, capturing coarser particles, and highefficiency (multi) cyclones, removing the agglomerated fines. The power requirement to operate the acoustic agglomerator would be about 0.02–0.5 % of the power plant output [B.48]. This means that for a 250 MW power plant several hundred kilowatts of acoustic power are needed, which is very high (compared with about 36 kW, the acoustic power output of a four-engine jet aircraft on take-off). To harness such large acoustic powers, considerable additional research and development are required. Aggregation of fine particulate solids also takes place in liquids. In environmental control, the removal of particulate solids from liquid process effluents is of great importance. As for gas/solid separations, when the size of the solids diminishes and reaches the micron or submicron (nano) range, the mass of individual particles becomes so small that they remain in suspension and cannot be removed by settling. Because of the fineness, membranes would be required to retain particles on a diaphragm, which is uneconomical for the cleaning of large volumes of contaminated liquids from industrial plants or waste-water treatment facilities. However, remembering the mechanisms of growth agglomeration (Chapter 5), if particles can be made to impact with each other, it is possible that they will adhere to one another in liquids, too. Therefore, when water that is contaminated with suspended fine solids is stirred, flocs may form naturally. If this happens, the size and shape of these aggregates depend on the circumferential speed of the stirrer and the processing time. Fig. 8.8 shows that flocs are larger if the shear forces are low and the

Fig. 8.8 Natural flocculation of solid contaminants in river water [B.48]. Parameters are the circumferential speed of the stirrer and the processing time

495

496


processing time is short. But further investigation revealed that higher speed of the stirrer and/or longer duration of mixing ultimately result in denser and more stable agglomerates. This is because of the previously discussed mechanism of growth agglomeration (Chapter 5) whereby loosely attached particles are removed under the influence of ambient forces (in this case, shear) and later have the chance to become re-attached in energetically more favorable positions thus yielding denser and stronger products. It depends on the process that will be used for solids removal, which of the two agglomerate structures is required, the loose flocs resulting from relatively gentle movement or stronger agglomerates from a more vigorous stirring for a longer time. In the large diameter circular thickener/clarifier (Fig. 8.9) that is commonly used in municipal water-treatment plants and in many industrial applications, water, flowing slowly from the feedwell in the center to the overflow around the periphery of the circular tank, is gently moved by a slowly rotating arm. Loose flocs are formed, which settle by gravity to the slightly conical bottom. Differently shaped scrapers (rakes) are used to move the sludge to the discharge cone at the lowest point from where it is transported to conventional liquid filters. The more vigorous stirring that produces stronger agglomerates is used when the resulting agglomerates are moved with some of the water to an off-site filtering system and, therefore, must survive transport. In many cases, even if collisions between particles do take place, the naturally available binding mechanisms, mostly molecular forces, which are considerably lower in a liquid environment than in a gas atmosphere, do not create bonds with sufficient strength to withstand the various separating effects and satisfactory flocculation does not occur. For quite some time it has been known that polymers, added to liquid-based particulate systems, have a dramatic influence on particle interaction. Molecules may attach themselves to solid surfaces and, depending on the characteristics of the exposed radicals, can cause particle attraction [B.29] or dispersion [B.63]. The second, dispersion, is applied to avoid agglomeration (Chapter 4) or enhance disintegration of aggregates. There are two ways in which polymers can promote aggregation: either by making particles more susceptible to salts or by flocculating the system without the aid of electrolytes. These processes are known as sensitization and adsorption flocculation, respectively. The second is more common. To create aggregates or flocs, the polymer adsorbs on different particles simultaneously, which is best accomplished by using substances with high molecular weight and a strong affinity to the particles to be agglomerated. Fig. 8.10 explains the principle. In nearly all applications of polymeric flocculants, the polymer addition and the subsequent flocculation process are carried out under conditions in which the suspension is agitated in some way, for example by stirring. In this way, the polymer molecules are distributed uniformly throughout the system and adsorb onto the particles, which are then encouraged to collide and form aggregates. As described earlier in other contexts, bridging may be followed by break-up if the bond is not strong enough and, later, re-attachment during another impact. Fig. 8.11 is a sketch of a flocculate. Care must be taken not to oversaturate the suspension with polymer. If too much polymer is adsorbed, the par-


(a) (b)

Fig. 8.9 a) Diagram; b) photograph of a circular thickener/clarifier (according to EIMCO, Div. Baker Hughes, South Walpole, MA, USA)

497

498

8 Applications in Environmental Control Fig. 8.10 Principle of polymer adsorption and flocculation [B.29]: a) adsorption of polymer molecule on the particle; b) rearrangement of adsorbed chain; c) collisions between destabilized particles and bridging to form aggregates (flocs); d) break-up of flocs

Fig. 8.11 Structure of a flocculate (floc) bonded by a polymer [B.48]

ticles may become restabilized (deactivated) because of surface saturation or by steric stabilization [B.29]. Fig. 8.12 demonstrates diagrammatically bridging, which results in the desired flocculation, and restabilization. Commercial flocculants are used extensively, for instance in water purification. By influencing the affinity of the polymer, it is also possible to obtain selective agglomeration. Less well known is the fact that, more often than not, solids and immiscible droplets dispersed in aqueous solution are electrically charged because of preferential adsorption of certain ion species, charged organics, and/or dissociation of surface groups [B.48]. Depending on such variables as the nature of the material, its pretreatment, pH, and composition of the solution, these charges can be either positive or negative. Since the surface charges on particles are compensated by an equal but opposite

Fig. 8.12 a) Diagram of polymer bridging between particles; b) restabilized particles [B.29]

8.1 Collection, Stabilization, and Deposition of Particulate Solids Fig. 8.13 Diagram representation of two particles with electrical double layers in a liquid [B.48]

countercharge surrounding them (Fig. 8.13) an electrical double layer develops (Chapter 5). Even though, as a whole, the system is electrically neutral, repulsion between the particles occurs. Upon addition of an indifferent (non-adsorbing) electrolyte (a salt), the double layers become less active and, as a consequence, the particles can now approach each other more closely before repulsion sets in. If enough salt is added, the particles may eventually come so close that van-der-Waals attraction binds them together. This is, in principle, the explanation of the sensitivity of colloids and suspensions to salts and why flocculation may be caused by the addition of salts. For certain technical applications, electrocoagulators may be used to charge the solids in contaminated effluents [B.48]. Metal hydroxides are produced by a system of soluble electrodes (anodes) which, in suitable electrolytes, cause coagulation of suspended solid particles into larger flocs.

8.1.2

Size Enlargement by Agglomeration for the Stabilization and Disposal of Particulate Solid Wastes

Current legislation in most developed countries classifies as hazardous most fine and all ultrafine particulate solids separated and collected from fluids with environmental control devices of any kind. In many cases, the mere fact that solids are micron or submicron sized, constitutes a reason for this classification. Owing to their fineness, dusts and slurries or moist residues from pollution abatement that are or become dry particulate matter easily become airborne causing a renewed threat to the environment and to human and animal or plant life (secondary pollution). Also, the large surface area of fine particulate solids results in high solubility so that toxic or otherwise undesirable substances may leach out and spill into surface water or end-up in aquifers if such fines are stored outside or deposited in unprotected landfills. Therefore, a common task is to use agglomeration for the size enlargement of fine materials that result from pollution control measures. Emphasis in this case is on the production of large and heavy aggregates to avoid scattering by wind or water and of permanent bonds that are waterproof, survive freeze–thaw cycles, and, preferably, immobilize leachable compounds. Because in this section recycling and the manufacturing of secondary raw materials are not a topic (for that, see Section 8.2), the only purpose of size enlargement is to render fine particulate solids suitable for safe handling, storage, and disposal. The cost

499

500


for carrying out the necessary processes is an extra burden and, consequently must be kept low. In Section 8.2, in connection with the size enlargement of low-grade ores or recovered fines for heap leaching, stockpile agglomeration is discussed as a cheap method. For the final disposal of terminal wastes, the same principle of agglomeration can be applied. Agglomeration is achieved at multiple transfer points from belts to belts (Fig. 8.14), on shaking or vibrating conveyor decks (Fig. 8.15), or on steeply inclined belt conveyors where material tumbles downward against the upward motion of the belt (Fig. 8.16) [B.48, B.97]. To obtain permanent, strong, and waterproof bonding, cement is often added as a low-cost, easily applicable binder. Agglomeration is initiated and bonding is activated by water sprays. Post-treatment for the production of final strength and long-term aggregate properties occurs in most cases naturally in curing piles prior to loading the material for transfer to the final disposal site, which does not need to be environmentally secured. Other low-cost processing of terminal waste may simply use mixing of the material with a suitable matrix binder (cement is again an option but hot mixing with bitumen Fig. 8.14 Sketch of belt conveyor agglomeration [B.48, B.97]

Fig. 8.15 a) Diagram of shaking trough; b) sketch of vibrating deck agglomeration [B.48, B.97]


Fig. 8.16

Reversed belt agglomeration [B.48, B.97]

is a common alternative) and forming the blend into crude, brick-like bodies that harden into inert disposable pieces. Depending on the hazard level, more sophisticated (and expensive) methods may have to be chosen that can finally include some sort of additional sealing by surface coating or encapsulation. In this context, encapsulation could mean, for example, the pressing of the prepared mass into drums or barrels (Fig. 8.17) that are closed after hardening of the contents has been completed.

Fig. 8.17 Heavy-duty drum compactor (courtesy S&G Enterprises, Germantown, WI, USA)

501

502


In a growing number of cases, particularly if radioactive wastes are involved, glass is used as the matrix binder. Since, in this case, achieving safety requirements is more important than cost, rather elaborate processes are developed and used. They include the impregnation of evacuated porous agglomerates and, sometimes, also coating with liquid glass or the high-temperature post-treatment of agglomerates that contain ground waste glass. Although recycling, the production of secondary raw materials, waste minimization, and resource conservation (Section 8.2) are the declared goals of all industrial societies, useless, terminal waste is also produced in growing amounts, which needs to be processed to avoid the costly disposal on secured sites or in managed landfills. Size enlargement by agglomeration is a new and quickly growing field of activities for this application. Knowledge of the fundamentals of this unit operation [B.48] and of the methods and equipment used to accomplish a multitude of tasks [B.97] and their interdisciplinary evaluation allows the development of suitable processes for safeguarding the environment, nature, enjoyment of life, and the future of the Earth.

8.2

Recycling/Secondary Raw Materials

Since industrial wastes are commonly particulate solids, of which a large portion becomes increasingly finer, size enlargement is often required to render these materials suitable for beneficial use. Agglomeration, the sticking together of particles with the help of various binding mechanisms, results in products with controllable composition, strength, size, shape, density or porosity, and other desirable properties. The latter include lowered or increased reactivity, improved solubility or dispersibility, increased abrasion resistance, reduced dustiness, better flow and metering, or, generally, greatly enhanced handling characteristics.

8.2.1

Historical Review of Waste Production

People have mined the earth for thousands of years. Initially these activities were limited to the gathering of metal, mostly gold nuggets and gem stones from the surface, but soon digging and tunneling into the earth began and materials other than the desired ones were also excavated. At the same time crushing was introduced to liberate valuable ingredients from unwanted solids. Fines, produced during crushing, and tailings (residue) were discarded in waste piles or dumped into holes. The location of many of the earliest mines is revealed by the waste heaps found nearby. Between 4000 and 3000 BC the Bronze Age began: humans learned to extract copper and tin from ores by heat and produced the alloy bronze. Accordingly, mining activities increased and wastes were still simply discarded into heaps. Somewhat before 1000 BC, the Iron Age started in western Asia and Egypt as man learned to smelt

8.2 Recycling/Secondary Raw Materials

iron from its ores. Large amounts of smoke and fumes were emitted from the increasing number of small furnaces, but no one felt a need to capture or control it and slag was dumped. Relatively early during the history of mankind carbonaceous materials, such as peat, lignite, and other coals, were collected and later mined to be burnt for various purposes. Like other mineral mine waste, coal fines were discarded, often in ponds after washing the coal lumps. There are indications that even in ancient times poor people retrieved the fines, mixed them with fat or oil, and manually formed them into bricks that hardened during open-air drying, a method that can still be observed today in coal mining districts of less-developed countries and in China. Such manually recovered agglomerated coal fines were used for home cooking and heating. Nevertheless, and in spite of these isolated recycling efforts, until the early part of the 20th century and sometimes even today, wastes from mining, including fine coal, ores, and many other minerals, from the processing of organic materials, such as wood, plants of all kinds and types, and fruits, and from the production of industrial goods were mostly discarded. During the past 100 years this situation has drastically changed. Industrial processing and production, yielding wastes of all kinds, skyrocketed. The desire and later the requirement to effectively clean gaseous and liquid plant effluents created large amounts of dusts and other fine or ultrafine particulate solids. Today, many of these materials are classified as hazardous because they may again pollute the environment and can be inhaled or ingested and resorbed, causing health risks. At the same time, industrial production and output are no longer concentrated in a few developed areas but distributed worldwide. Therefore, solid waste collection, treatment, processing, recycling, and disposal have become very important (often mandatory) industrial activities whereby handling of the finest materials poses the biggest challenge and costliest efforts. In addition to the wastes produced during the collection, extraction, upgrading, and processing of natural raw materials, the quickly and still increasing production capacities in all industries result in large quantities of rejects and by-products, which inherently represent great value. Furthermore, the useful life of many industrial products is relatively limited and tends to become ever shorter. The obsolete items: equipment, such as machinery, cars, electronic components; packing materials, including paper, cardboard, foil, plastics, bottles, cans; wastes, for example newsprint, glass, wood, organic residue; and many more, contain a multitude of potentially valuable ingredients that must be separated and/or cleaned for further use. Tab. 8.4 summarizes the most important advantages of agglomerated solid mineral and metallurgical wastes. Bullet points 1 to 3 are self-explanatory since the particle size of most solid wastes is too small for direct processing or is ultrafine after they were removed from gaseous or liquid plant effluents as solid pollutants (Section 8.1). Because of their size-related large specific surface area, fine and ultrafine solid wastes, especially metallic fines or those containing metals or cellulosic dusts, are very reactive. Even at ambient conditions they combine easily with oxygen in an exothermic reaction that may cause accelerated, catastrophic heating or “dust explosions”. Size enlargement by agglomeration, accompanied by densification, sometimes

503

504

8 Applications in Environmental Control Tab. 8.4

Advantages of agglomerated solid wastes

*

Agglomerated particulate solid wastes contain no or low amounts of dust; therefore, they provide increased safety during handling, even of hazardous materials, no secondary pollution, and, generally, fewer losses.

*

Agglomerated solids are freely flowing, featuring: – Improved storage and handling characteristics, – Better metering and dosing properties, – Increased bulk density and lower bulk volume.

*

Agglomerates have larger sizes and their size distribution can be controlled.

*

Within certain limits, depending on the process used, product porosity or density can be influenced; leachability and reactivity as well as other properties can be controlled.

*

Secondary raw materials can be crafted by: – Enlarging the size of wastes containing valuable ingredients, – Combining different solid waste components to obtain improved feed materials with controlled composition, – Including additives to bring about desirable processing characteristics.

hot, renders such dusts sufficiently inactive for safe handling and use. Terminal solid wastes, destined for final storage, sometimes contain toxic or other hazardous components which, if untreated, may become airborne or leach into groundwater, requiring lined and/or covered disposal sites. Size enlargement by agglomeration using permanent binders, often in the form of a matrix, filling all interparticle voids, eliminates the need for special conditions for the deposition of such materials. 8.2.2

Agglomeration Technologies for the Size Enlargement of Wastes

A common classification of methods for the size enlargement of particulate solids distinguishes between two types of process (Chapter 5), * *

natural and growth or tumble agglomeration (no external forces) and pressure agglomeration (low, medium, or high external forces)

and two techniques, binderless agglomeration and * agglomeration with the addition of binders. *

For fine metal ores and, more recently, for iron-bearing waste materials, agglomeration by heat (sintering), a less frequently used size-enlargement technology, is used. All agglomeration methods can be used to manufacture products for recycling and secondary raw materials. Growth/tumble agglomeration happens in a similar fashion to natural agglomeration. Because the particles to be agglomerated are larger (surface equivalent diameter 10 < xo < 200 lm [B.97]), the particle-to-particle adhesion must be enhanced by binders and the collision probability must be increased by providing a high particle con-


centration. In most cases, growth/tumble agglomeration first yields “green” agglomerates, which are only held together temporarily by liquid binders. Final strength and potentially other product characteristics are obtained during post-treatment. Because of the growth mechanism, the agglomerated bodies are more or less spherical. If a narrowly sized product is desired, screening is used. Oversize (after milling) and undersize are recirculated to the agglomeration equipment. Pressure agglomeration processes use external forces to enhance bonding by densification and to shape the product. As far as applicability is concerned, high-pressure agglomeration is the most versatile technology for the size enlargement of particulate wastes. If certain characteristics of the feed materials and conditions occurring during densification are considered and controlled during equipment selection, design, and operation, any kind and size (from nano- to millimeters) of particulate material can be successfully processed. Since high-pressure agglomeration is essentially a dry process, a limitation exists with regard to the highest tolerable moisture content of the feed. Filter cakes, for example, a common initial form of waste, can not be agglomerated by high pressure without first reducing the moisture to very low levels by additional drying.

8.2.3

Applications in the Mineral, Metallurgical, and Energy Related Industries

In the mineral and metallurgical industries, resource conservation by the recovery of valuable components from waste and the recycling of materials containing recoverable ingredients has already reached a high level of acceptance. New recovery technologies make the secondary processing of old mine waste deposits, particularly of ores and coal, economical and deposition of new wastes is minimized by the application of methods that convert a host of different, formerly obsolete materials into “secondary raw materials”, which are recycled, replacing primary raw feed sources. The latter also include iron and non-iron scrap, turnings, borings, grindings, dusts, etc. Mixing different raw materials, additives, and binders, agglomerating the blend, and subjecting the product to different post-treatment methods to achieve special properties is known as material engineering. Industrial wastes can be included in such raw materials and additives can produce, for example, a fluxed feed for metallurgical operations, secondary raw materials with predetermined alloying ingredients, or smokeless fuels (Sections 6.8, 6.9, and 6.10). Tab. 8.5 lists the areas in which size enlargement by agglomeration can benefit solid waste management in mineral and metal processing. Waste minimization is a preventive measure by which particulate solids are agglomerated to avoid pollution in the first place. Examples are in the bulk handling and transportation of naturally dusty minerals. In pollution control agglomeration has many applications. In most cases, collection of fine solids is improved or the removal of ultrafine particles is made possible (Section 8.1). Examples are the agglomeration of smoke particles or the flocculation of solid contaminants in effluent water. By agglomeration, by-products can be converted into materials that are suitable for beneficial use. Examples are fertilizers containing FGD

505

506

8 Applications in Environmental Control Tab. 8.5 Areas in which size enlargement by agglomeration benefits solid waste management *

Waste minimization

*

Pollution control

*

Beneficial use of waste by-products

*

Recycling/reprocessing

*

Production of secondary raw materials

*

Terminal wastes disposal

*

Prevention of secondary pollution

(flue gas desulfurization) ammonium sulfate or pulverized Thomas slag, lump gypsum from FGD gypsum for use in cement production, and building materials from ashes. Recycling identifies the re-utilization of off-size solids after agglomeration in the same process in which this material was originally rejected. Examples are metal-bearing dusts from smelters, metal turnings, borings, grindings, etc., and coal, ore, and mineral fines from crushing and/or beneficiation. With modern processing technologies, many fine or low-grade raw materials which for decades and sometimes centuries were discarded at the mine or former processing sites can be used after recovery and agglomeration for reprocessing. Examples are fines deposits of rejected bauxite, coal, and ores (copper, gold, iron). With the exception of waste minimization all aforementioned applications are part of “resource conservation” by increased emphasis on secondary raw materials, process input obtained from wastes, rejects, and obsolete products, in most instances involving agglomeration. In some cases, the nature of terminal solid wastes, those which can no longer be processed for beneficial use, require agglomeration for disposal (Section 8.1). One reason for this can be to avoid leaching of toxic or radioactive components into the groundwater by the application of matrix binders such as cement or glass for stabilization. Another is to avoid secondary pollution because, after separation from plant effluents, many terminal wastes contain very fine particles in dusts, slurries, sludges, or slimes. Size enlargement by agglomeration is increasingly required to avoid costly disposal in scarce, small, specially designed hazardous waste deposition sites. Some examples will be presented in more detail. They are selected such that, with the exception of natural agglomeration (used, for example, for the densification of silica fume, Section 6.7.3), at least one type of the available agglomeration processes is demonstrated. Concerns regarding acid rain led to legislation in all industrialized countries aimed at limiting the sulfur content of flue gases, typically from power plants (Section 6.10.2, compliance fuel) and metallurgical processes. One of the most widely used desulfurization processes uses limestone as absorbent. The resulting calcium hydrogen sulfite is converted by oxidation into (synthetic) gypsum, called FGD gypsum (calcium sulfate dihydrate). Fig. 8.18 shows an absorber and associated processing facilities in a coalfired power plant.


Fig. 8.18 Sulfur absorption in the flue system of a coal-fired power plant (EVS, Heilbronn, Germany): left) absorber; right) limestone processing and gypsum recovery (briquetting) systems

The product from such a desulfurization system is a slurry and, after mechanical dewatering, a filter cake, containing finely divided solids. Disposal of the wet FGD gypsum is expensive and can result in secondary pollution if it becomes dry and is not stabilized. It has been found that, after agglomeration, the material can be used in cement kilns (replacing natural lump gypsum), economically transported to and used in plants producing gypsum board for the building industry, or safely discarded. Principally three methods are used to convert the finely divided moist gypsum into the required lumpy form, all requiring no additional binder. 1. Tumble agglomeration of the moist gypsum in discs, drums, or mixers (Chapter 5, Fig. 5.2) and hardening the green, spherical pellets by thermal curing. 2. Densification of gypsum with reduced moisture content (8–10 %) in pellet mills (Chapter 5, Fig. 5.10b1–b6) and hardening the green cylindrical pellets by thermal curing. 3. Drying the gypsum to < 1 % moisture content and pressing and shaping it into finished, pillow-shaped briquettes with roller presses (Chapter 5, Fig. 5.11, lower right).

507

508


Fine particulate solids, dry or moist and from any source can be converted into larger products by growth/tumble agglomeration. The example shown in Fig. 8.19 depicts a simple system using an inclined disc for agglomeration, a rotary kiln for post-treatment, and screening and crushing for finishing. Moist particulate solid wastes, such as predried FGD gypsum filter cake, can be extruded (Fig. 8.20) in, for example, flat die pellet presses (4 in Fig. 8.20) yielding cylindrical agglomerates, which are dried and screened prior to load-out. Since in high-pressure agglomeration with roller presses densification is so high that the residual porosity can only hold very small amounts of moisture (Chapter 5),

Fig. 8.19 Size enlargement of particulate solid waste by growth/ tumble agglomeration (courtesy Eirich, Hardheim, Germany)

Fig. 8.20 Pelleting of wet FGD gypsum by medium-pressure agglomeration (flat die pelleting, courtesy Amandus Kahl, Reinbek, Germany)


if this technology is used, the FGD gypsum must be thermally dried to < 1 % moisture content prior to briquetting. The latter is very simple as shown in Fig. 8.21. To avoid starved feeding of the roller press and guarantee the production of high-quality briquettes, a small stream of excess material overflows at the end of the horizontal conveyor, is measured by a solids flow meter [B.97], and controls the discharge from the day bin, that is, the amount of press feed. Fig. 8.22 depicts photographs of a roller press in a FGD gypsum briquetting plant and of product samples. Unless only fine dusts must be treated, the more versatile processes for size enlargement in recycling and particularly the secondary raw materials industry use pressure agglomeration technologies. Especially with the high-pressure machines (ram extrusion, hydraulic punch-and-die, and roller presses, Chapter 5, Fig. 5.11) a wide variety of feed materials and particle sizes can be processed. The sometimes very large energy input results in extensive densification, yielding high molecular forces for bonding, disintegration of brittle particles followed by recombination bonding, momentary melting of roughness peaks at the contact points, resulting in solid bridges, and plastic deformation of malleable components. The latter can be enhanced if hot feed materials are used (Sections 6.6.2 and 6.9.2). As early as 1885, briquetting of iron-bearing fine residues from cupriferrous pyrites after roasting and the extraction of copper was carried out commercially for recycling in the UK [Section 13.3, ref.103]. The processes used punch-and-die presses and a binder, producing relatively large, brick-like pieces that were subsequently fired to obtain final strength. Around the same time, great success was reported in the bri-

Fig. 8.21 Flow diagram of a FGD gypsum roller press briquetting plant. Thermal drying of the feed is not shown

509

510


Fig. 8.22 Roller press for the briquetting of (synthetic, FGD) gypsum in a flue gas desulfurization plant: inset, synthetic gypsum briquettes (courtesy K€ oppern, Hattingen/Ruhr, Germany)

quetting of blast furnace dust since this by-product contains lime, alumina, and soluble silica and, therefore, like cement, already possesses hydraulic binding properties. However, although they had been considered as quite important at the time, these early technologies were soon abandoned because, due to increasing costs of labor, energy, and, where applicable, binder, it became more economical to simply dump waste materials into industrial landfills. The desire for environmental protection and the growing need to conserve raw materials, such as, fuels, ores, or strategic alloying components, and natural resources, such as land and water, have revitalized these efforts, leading to new processes utilizing modern equipment and technologies. Therefore, today’s briquetting systems have little in common with early brick-making procedures. Fig. 8.23 is the flow diagram of a briquetting plant for metallized, metal-bearing, or other recyclable fines. The possibility of adding solid or liquid binders, either alone or in combination, render this simple system very versatile. As shown, the feed material(s) is (are) dry and cold and the binders are self-curing, that is, they gain final strength at ambient temperatures in a ventilated (to remove heat, moisture, and gaseous reaction products) briquette storage bin. Incorporation of certain materials, such as oxide fines or certain mill scales, and/or restrictions on binder amount or composition may require the installation of a curing oven. Typically the briquetted product is pillow shaped, strong, free of dust, and can be easily stored, handled, and metered during recycling (Fig. 8.24).


Fig. 8.23 Flow diagram of a generic briquetting plant for metallized, metal-bearing, or other recyclable fines

For recycling and the manufacture of secondary raw materials, successful size enlargement by agglomeration often hinges on finding a binder that is chemically (no contamination) and economically (low cost) acceptable and physically effective (adequate strength, high abrasion resistance, good flow and metering characteristics). The addition of fibers or other elongated particles is an innovative method for reinforcing agglomerates [B.97].

Fig. 8.24 Close-up view of a pile of typical briquettes produced by roller presses for recycling

511

512


Fig. 8.25 Flow diagram of a plant for the production of paper fluff from old newsprint, as a binder for use in a roller-press briquetting plant [8.2.1]

A simple, low-cost fibrous binder component that is available everywhere can be obtained by grinding waste paper. Fig. 8.25 is the flow diagram of a plant for the production of paper fluff from, for example, old newsprint as a binder for use in a roller press briquetting plant. This cellulosic material mixes quickly and uniformly with dry powders in any kind of blender. Even small percentages (2 %) result in a marked increase in (crushing) strength if compared with two other common binders (Fig. 8.26) [8.2.1].

Fig. 8.26 Effect of binder type on the crushing strength of briquettes made from metallurgical dust [8.2.1]. Amounts of binder added: waste paper, 2 %; molasses, 6 %; lime, 6 %; starch, 6 %


Another application of “fibers” as a reinforcing binder component is the addition of grinding swarf during the briquetting of metal-bearing filter dust [B.48, B.97]. Up to 50 % of high-grade steel grinding swarf < 30 mm and a small amount of the conventional binder lignosulfonate powder (sulfite waste from paper making) were added to residue (filter dust and sludge) from specialty (alloyed) steel production. Since the product is destined for recirculation into the metal-making process, it is important to produce high strength of which at least a certain part is retained at high temperature, until the surface of the liquid metal bath is penetrated and melting occurs. Thereby, secondary contamination because of premature release of dust is avoided. This is achieved by the presence of reinforcing fibers in the structure. Fig. 8.27 indicates that the briquette strength increases with addition of swarf while the necessary amount of chemical binder, constituting contamination and non-temperature resistant bonding, decreases. During simple atmospheric curing a considerable increase in strength occurs because of hardening of the chemical binder component, which improves the product’s handling and storage properties. Fig. 8.28 shows a broken cylindrical compact that was manufactured during process development with a laboratory punch-and-die press and actual briquettes obtained in an industrial plant. Concerns that, when produced with a roller press, the addition of “fibers” would prohibit separation into mostly single, handleable pieces (see also Section 6.9.2) were unfounded. Most dusts from pyroprocessing plants in the minerals and metals industries contain calcium and/or magnesium oxides. They hydrate easily, especially also during storage at high or even ambient humidity, only more slowly in the latter case. Since this reaction is coupled with a large increase in volume, when natural hydration occurs, with time even small amounts (> 0.5 %) of alkaline earth oxides tend to lower the strength of any agglomerate, including high-quality briquettes, and, in many cases, eventually result in complete destruction. Therefore, if such components are present, an appropriate amount of water is mixed with the dust in a preparatory step

Fig. 8.27 Cold crushing strength of briquettes from metal-bearing dust and sludge containing different amounts of dry lignosulfonate binder (called “sulfite waste powder”) with and without reinforcement by the addition of swarf [B.48, B.97]

513

514


Fig. 8.28 a) Broken cylindrical compact that was manufactured during process development with a laboratory punch-and-die press; b) actual briquettes obtained in an industrial plant [B.48, B.97]

and the oxides are allowed to hydrate prior to agglomeration. If afterwards additional moisture is required alone or as part of a binder system, a second blender must be installed in the agglomeration plant. Because the design of agglomeration plants, especially of those using briquetting, can be very compact, they are often suitable for installation at or near the dust-collection system(s) (Fig. 8.29). They are fully automatic, needing practically no operator attention, and can be monitored from a nearby control room. In the case of dust recirculation from specialty steel production, the briquettes that were reinforced with alloyed steel grinding swarf were an excellent feed addition for electric arc fur-

Fig. 8.29 Compact briquetting system for metallurgical filter dust. Capacity 11–20 t/h, depending on feed characteristics (courtesy: Thyssen Stahl AG, Krefeld, Germany)


naces. Some 98 % of the chromium and 99 % of the nickel contained in the residue can be recovered by this practice [Section 13.3, ref. 103]. Fig. 8.30 shows how, by employing size enlargement by agglomeration, almost all solid wastes within, for example, the steel industry can be recycled (furnace and converter dusts) or processed into secondary raw materials (Waelz oxide, sinter), resulting in very little terminal residue going to disposal or to outside use (road building material). Hot and cold briquetting with roller presses are the predominant methods for size enlargement in this case because it can handle materials of varied composition, sizes, and properties. The rotary kilns shown in front of the hot briquetting machines are used for heating the feed to the required temperature. Alternative methods of heating can be chosen, too. Other agglomeration processes shown in Fig. 8.30 apply hot nodulizing (a tumble/growth method) in the Waelz kiln and sintering (agglomeration by heat). Sometimes, when wastes are recycled, technological, physical, and/or chemical, limitations must be considered. For example, it is not possible to continue re-using steel mill dust (as shown in the right part of Fig. 8.30) indefinitely. Because of decrepitation and attrition, this would result in the production of an ever increasing amount of new dust and briquettes. Therefore, a maximum amount of briquettes in the feed per charge must be determined and the rest must be sold to other users. Also, an

Fig. 8.30 Size enlargement by agglomeration in iron and steel making for the minimization of terminal residue through recycling and the production of secondary raw materials [Section 13.3, refs 129, 166]

515

516


enrichment of contaminants, particularly zinc, takes place in the dust with each cycle. To determine its level, dust must be sampled and analyzed regularly. Once a certain concentration of zinc is reached, the entire dust that is in the system at this time is passed to the Waelz plant. The Waelz oxide is briquetted hot and treated in an Imperial Smelting (IS) shaft furnace to recover zinc and lead. As an example [8.2.2], in one plant, 3–4 t of briquettes are charged per heat and enrichment from about 5 % to more than 17 % of zinc occurs in 3–4 weeks. At that time, 350–400 t of briquettes are removed from the cycle for reprocessing in the Waelz kiln. Certain processes shown in Fig. 8.30, such as the manufacturing of Waelz oxide and the recovery of zinc and lead, can not be supported by and operated economically in a single steel mill. Therefore, independently owned plants are often processing the waste from several producers and are, preferably, located in the geographical center of an industrialized area where suppliers of the metal-bearing dusts, users of the recovered metals, and customers for the by-products (road building materials) are close by. Also, with a trend away from the giant integrated steel mills towards smaller regional producers and the application of alternative iron units for steel making (direct reduced iron, Section 6.9), the treatment and recovery of metallurgical wastes is no longer economically feasible in these new facilities, although, in plants that are also using rolling mills and wire drawing operations for the manufacturing of intermediate products such as slabs, sheets, rods, and coils, they may additionally include mill scale and metal grindings, sludges, and slimes. This situation is generally found in all industries generating recoverable and recyclable waste. In response, maximum amountolling processors (Section 9.2) and specialized service companies are being created to take wastes and convert them into beneficially useable products or treat them for safe disposal. These companies fulfil the desires, requirements, and/or charter of the industry and of governments to conserve resources and to reduce the need for increasingly expensive disposal and scarce, often specially designed and protected deposition sites. While many are serving a small number of customers and are strictly of regional importance (see below), some operators are becoming large, even multi-national, globally active corporations. An example of a large, multi-national service provider is Heckett MultiServ (Section 15.1). In their own words, this company, “delivers specialist services to more than a quarter of the world’s steel production, operating at over 160 sites in 33 countries and employing more than 9000 people. (Currently) every year (they) process 40 million tons of slag and debris, recover 9 million tons of steel, handle 11 million tons of scrap, and transport 10 million tons of liquid slag and 6 million tons of liquid steel. (Heckett MultiServ) design, build, and operate facilities for materials handling and preparation, waste processing, and environmental services ... (They work) within customer’s own premises and plants, providing tailor-made services using specialist technologies and equipment operated by their own highly-trained people”. The above indicates that the services offered by this group include much more than processes that are related to size enlargement by agglomeration. In regard to the latter the corporation states that they have “extensive experience with a variety of steel industry by-products and that all processes offer the recovery of iron units and valuable metal-bearing components as well as reduced disposal costs”. Using binder systems,


both pellets (Fig. 8.31) and briquettes (Fig. 8.32) are made from the most appropriate by-product blends, meeting the demands of the iron and steel making process and the local environment. Tab. 8.6 lists materials that are successfully pelletized or briquetted by Heckett MultiServ around the world. Fig. 8.33 depicts stacks of pellets (left) and briquettes (right) that are ready for recycling. Pelletizing (Fig. 8.31) is applied to convert waste and by-products from a steel mill that also operates a sinter plant into agglomerates with improved handling characteristics. The pellet composition is site specific. It depends on the number of components to be incorporated and their physical and chemical characteristics, including moisture content. By preparing the materials in this way, not only is recycling achieved but the productivity of the sinter plant is also enhanced. The inputs and ratios depend on the mill’s specifications for so-called control elements, such as zinc, alkali, and oil. To date, Heckett MultiServ operates two pelletizing plants, both in Europe, with annual capacities of 100 000 and 200 000 t/year.

Fig. 8.31 Disc pelletizer for the recycling of waste dusts, slurries, and plant fines (courtesy Heckett MultiServ, Butler, PA, USA)

Fig. 8.32 Briquetting plant for and typical briquette from steel mill by-products (courtesy Heckett MultiServ, Butler, PA, USA)

517

518

8 Applications in Environmental Control Tab. 8.6 Materials that are successfully pelletised or briquetted by Heckett MultiServ (Section 15.1) around the world *

Sinter plant dusts

*

Blast furnace dusts

*

BOS grits and sludges

*

Magnetic fines from desulfurization and steel slags

*

Mill scale and oily mill scale sludges – EAF dusts – AOD/CLU/VOD (converter) dusts – Grindings and swarf

Fig. 8.33 Stacks of pellets (left) and briquettes (right) that are ready for recycling (courtesy Heckett MultiServ, Butler, PA, USA)

Briquetting (Fig. 8.32) is used for the recycling of particulate wastes and by-products in steel mills if a sinter plant does not exist or the inputs are unsuitable for the sinter plant/blast furnace route. In addition to the accurate proportioning and mixing of the materials with a suitable binder, control of the blend’s moisture content is critical. At the present time Heckett MultiServ is operating seven briquetting plants with tonnages ranging from 7500 to 150 000 t/year in Canada and Europe and is in the process of installing another one in Australia.

8.2.4

Applications in Regional and Municipal Material Recycling Plants

As a further example of major recycling efforts that use agglomeration methods to convert sorted wastes into recyclable secondary raw materials, regional material recycling facilities (MRF) [8.2.3] and municipal waste processing plants shall be discussed. Beginning in the 1980s, MRF were developed in several countries (USA – MRF; Germany – DSD, Duales System Deutschland) to separate, through process design, tech-

8.2 Recycling/Secondary Raw Materials Tab. 8.7 Commonly considered advantages of material recycling facilities [8.2.3] *

Since MRFs permit co-mingling, public participation is made easier as less sorting is required by the waste generator. This leads to higher recycled volumes and unburdens the conventional municipal solid waste collection and processing.

*

Collection costs can be reduced since less expensive equipment can be employed and time consuming sorting at the curb is eliminated or reduced.

*

Higher volumes of recyclable materials offer greater market influence and acceptance since industry quality standards can be better met by producing superior forms of secondary raw materials.

nology, and human inspection, co-mingled recyclable materials and convert them into marketable commodities. These facilities are in addition to the conventional municipal solid waste processing plants. While for the latter wastes are picked up indiscriminately from households or collection points, the feed for MRF, although still more or less co-mingled, has been pre-sorted by the originators and is collected from curb sides or special collection containers. Tab. 8.7 lists the three principal characteristics of MRF that are commonly considered as advantages and Tab. 8.8 provides a breakdown of MRF separated recyclables in a specific plant, typical processing costs, and expected market prices of the resulting secondary raw material [8.2.3].

Tab. 8.8 Breakdown of MRF separated recyclables in a specific plant as well as typical processing costs and expected market prices (US Midwest, end of 1992) of the resulting secondary raw material (adopted and modified from [8.2.3]) Material

% of total

Paper Newsprint Cardboard Mixed

50

Cans Steel Aluminum (UBCs)

10.50

Glass Clear (flint) Brown (amber) Green Mixed

29.00

Plastic HDPE PET Polystyrene Foil

7.75

Waste Residue

2.75

Processing cost

Market price

US$/t

US$/t

33.59 42.99 36.76

0–20 10–30 5–25

(6.4) (4.1)

67.53 143.51

49–78 500–900

(11.8) (3.9) (4.7) (8.6)

72.76 111.52 87.38 50.02

50 25–40 5–15 0–30

(5.1) (2.5) (0.08) (0.07)

187.95 183.84 – –

40–160 40–200 – –

–

–

519

520


Tab. 8.8 shows that the most valuable material in MRF recycling streams are used (aluminum) beverage cans (UBC) which, with the exception of paper (see below), represent the only material that, at the assumed market conditions, can be sold for a profit. More recently polymer recycling has become feasible too (see below). As discussed in Section 6.9.2, the aluminum industry has one of the most advanced and complete recycling programs in the metals market [Section 13.3, refs 120, 132, 137]. The major reason for this is that the production of aluminum from Bauxite requires large amounts of electrical energy so that, today, most of the producers are located in areas where cheap (hydroelectric) power is available. Transportation from these locations to the finishing mills adds substantially more to the cost of the virgin material. Aluminum is becoming one of the major metals used for packing, replacing steel and, in many cases, glass. As a result, a large segment of aluminum recycling is based on the collection and processing of used beverage cans (UBC) and other aluminum (food) containers [Section 13.3, ref. 132]. Competitively priced high-quality can stock can be made from re-melting aluminum containers after they have been decoated (delaquered) by solvent or, more commonly, thermal processes. The objective of decoating is to remove paint, lacquer, plastic, paper, and other contaminants (food residues) with minimal modification of the metal. The nature and amount of the coatings vary widely. They include organic chemical compounds, which are predominantly volatile (VOC, volatile organic compound), inorganic components, which are added for coloring and mass, plastic or paper laminates, and oil and water from machining and forming operations. After proper decoating, it is also possible to recycle, in addition to UBC, such materials as clean foils, printed foils, painted packaging, paper- and plastic-laminated foil, food containers, litho plate, and contaminated extrusions. The raw aluminum scrap separated, for example in an MRF, from the other materials is rather voluminous and, if decoating is not available on site, which is the most common situation, are baled, an agglomeration process typically using hydraulic presses (Section 6.9.2, Fig. 6.9.19) for transportation to the thermal treatment facility. There, the key to efficiently producing a quality decoated aluminum scrap product is feed preparation, which includes shredding during which bales are broken-up and larger parts are reduced in size to less than about 50 mm. This material is then subjected to screening to remove fines < 3 mm, which are discarded, and to magnetic separation. Decoated aluminum scrap leaves the thermal treatment unit in particulate form with a temperature of about 500 8C. Since, in most cases, the remelting furnaces are also at a different location, the product is again baled (hot) and sold as premium “old” aluminum scrap to secondary aluminum smelters. A disadvantage of using baling at this point is that, prior to feeding the scrap into remelting furnaces, these bales must be torn apart again because: * * *

the bales are too large for direct charging, operators want to make sure that the scrap quality is consistent throughout, and bales may have picked up excessive amounts of moisture, which must be removed.


A disadvantage of using the resulting loose scrap, although found of acceptable quality, is that considerable oxidation losses (dross) occur during melting. As an alternative to baling, particulate decoated aluminum scrap can be converted into highly densified (>2.0–2.3 g/cm3) slabs using roller presses (Section 6.9.2) [Section 13.3, ref. 140]. Fig. 8.34 shows in the upper part different processed MRF aluminum wastes. They are from left to right: light foil, dense foil, and UBC. Densifying these materials in roller presses as described in Section 6.9.2 yields products as depicted in the lower part of Fig. 8.34. Although UBC do not reach a density of more than 2.0 g/cm3, even if they are hammer-milled or mixed with other, heavier aluminum wastes before compaction, a considerable improvement of recovery is obtained with a product as shown in the right lower part of Fig. 8.34. The slabs are particularly advantageous as feed for reverberatory furnaces. They can be easily charged and the metal losses experienced during the melting of loose, particulate aluminum scrap, which typically amount to 15–20 %, are reduced to 2–4 %. In general, for the compaction of clean processed metal swarf, two methods are available. The non-continuous confined volume punch-and-die process and the continuous compaction in the nip of two counter-rotating rollers (Section 6.9.2). With the punch-and-die process clean, delaquered, and crushed UBC can be compacted to yield a relatively high apparent density. The product shape is normally cylindrical with diameters between 80 and 190 mm, heights between 30 and 120 mm, and weights between 0.4 and 7.3 kg (Section 6.9.2). Although the punch-and-die technology is by far the most widely applied choice for the compaction and shaping of UBC-based aluminum scrap it also has a number of disadvantages of which the most important is the relatively slow movement of the (hydraulic) rams. Therefore, it is used when relatively small amounts of processed aluminum cans and containers must be compacted. Influenced by developments in connection with the briquetting of aluminum home scrap (Section 6.9.2), the continuous compaction between two rollers is now a feasible

Fig. 8.34 Three different MRF aluminum wastes as collected and processed (top) and after roller press compaction [Section 13.3, ref 132]. From left

to right: loose foil, densified (granulated) foil, delaquered and hammer-milled used beverage cans (UBC)

521

522


alternative. It offers the advantages of large capacity and high apparent density. However, since an endless sheet is formed with these machines, there is no hope of producing directly individual compacts, such as cylindrical briquettes from the punch-anddie process that can be easily charged into the remelting furnaces. Similar to the solution presented in Section 6.9.2 a suitable cutting device (shear) must be employed to divide the strip into slabs, pieces with acceptable dimensions. Other examples of regional processors and recyclers are modern modifications of the conventional municipal waste disposal plants. Until recently there were two alternatives for the removal, handling, and elimination of municipal wastes. Originally, after pick-up from the curb or from specific collection points, the entire mass was disposed of in landfills, either filling natural or excavated holes or building mountains, which, after completion, were covered with dirt and topsoil and left to rot. Aside from the scarcity of suitable land near population centers to accept the exponentially growing masses of waste, problems arose with time from the leaching of harmful chemicals into aquifers and the production of increasing amounts of methane within the waste. These required the rehabilitation of many of the older, sometimes already abandoned landfills and triggered environmental laws, severely restricting the use of existing and opening of new landfills and/or requiring special design considerations (such as leakproof liners, collection and removal or use of methane gas, etc.) for new disposal sites thus making this method expensive and unattractive. Efforts to entice consumers (households) to segregate wastes into recyclable components, such as paper, metals, and plastics (see above), and rejects, reduced the amount of materials to be disposed of but created mostly organic discards, which become smelly and produce gases and other environmental contaminants if not properly handled and stored. The logical solution to this problem, the municipal waste incinerator, often faces public opposition. Therefore, communities and waste processors, supported by new laws that require the conversion of solids with a certain heating value into (industrial) fuels, are forced to develop new methods for further segregation of wastes and the processing of the resulting components. Since not all regions have adopted the separate collection of recyclables and not all waste generators accept the desire to segregate, the first processing step in a modern municipal solid waste treatment plant is to segregate the waste by scalping (screening, size), manual and visual selection (paper and paper products, wood, plastic containers), shredding of large pieces, re-screening (size), magnetic separation (metals), jigging (density, plastic), etc. Metals, divided into magnetic and non-magnetic (aluminum) varieties, are sold as iron-bearing (magnetic) and non-iron, heavy scrap or transferred to aluminum scrap processing (see above). Increasingly, natural and man-made organic materials that can be burnt are converted into a secondary solid fuel product. The latter is exemplified with the description of a recently installed (1999) large plant in Schwarze Pumpe, Germany, for the production of pellets from municipal (household) solid waste, the light fraction from a shredder plant, DSD (see above) plastics, and waste wood (chips, saw dust, crushed, or reduced to fibers). Typically, the mixture to be pelleted consists of 50 % household waste, 10 % light shredder fraction, 25 % DSD plastics (with up to 75 % foil), and 15 % wood. Depending on the composition of these ingredients (Tab. 8.9), the feed can be coarse and light (20–60 mm long,

8.2 Recycling/Secondary Raw Materials Tab. 8.9 Approximate compositions of two common feed mixtures to the pelleting machines of the secondary solid fuel manufacturing plant at Schwarze Pumpe, Germany (Courtesy Amandus Kahl, Reinbek, Germany) Description/Component

Light mixture

Fine mixture

Size range Thickness Moisture content Bulk density

[mm] [mm] [mass %] [t/m3]

20–60

Agglomeration Processes: Phenomena, Technologies, Equipment

Agglomeration Processes: Phenomena, Technologies, Equipment

Astrophysics Processes: The Physics of Astronomical Phenomena

Rotary Kilns Transport Phenomena and Transport Processes

Astrophysics Processes: The Physics of Astronomical Phenomena

Surface phenomena in fusion welding processes

Markov-Modulated Processes & Semiregenerative Phenomena

Equipment

Equipment

Equipment

Equipment

The Agglomeration Theory of Sleep

Resource-Adaptive Cognitive Processes (Cognitive Technologies)

Flat-rolled steel processes: advanced technologies

Surface Engineering of Metals - Principles Equipment and Technologies

Microwave and Millimeter Wave Technologies Modern UWB antennas and equipment

Microwave and Millimeter Wave Technologies Modern UWB antennas and equipment

Advanced Transport Phenomena: Fluid Mechanics and Convective Transport Processes

New Equipment

Refrigeration Equipment

Storm Equipment

Storm Equipment

Refrigeration Equipment

Combustion Phenomena

Diffusion phenomena

Transport Phenomena

Electrokinetic Phenomena

TRANSPORT PHENOMENA

Wave phenomena

Transport Phenomena

Electrokinetic Phenomena

Agglomeration Processes: Phenomena, Technologies, Equipment