Proceedings of the 9th International Congress on Deterioration and Conservation of Stone

9th INTERNATIONAL CONGRESS ON DETERIORATION AND CONSERVATION OF STONE This Page Intentionally Left Blank Proceeding...

Author: V. Fassina

74 downloads 1300 Views 61MB 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

9th INTERNATIONAL CONGRESS ON DETERIORATION AND CONSERVATION OF STONE

This Page Intentionally Left Blank

Proceedings of the 9th INTERNATIONAL CONGRESS ON DETERIORATION AND CONSERVATION OF STONE

Venice June 19-24, 2000 Edited by

Vasco Fassina

Organised by

Istituto Veneto pet t Beni Culturali

ELSEVIER Amsterdam

- Lausanne

- New

York

- Oxford

- Shannon

- Singapore

- Tokyo

ELSEVIER SCIENCE B.V. Sara Burgerhartstraat 25 P.O. Box 211, 1000 AE Amsterdam, The Netherlands 92000 Elsevier Science B.V. All rights reserved. This work is protected under copyright by Elsevier Science, and the following terms and conditions apply to its use: Photocopying Single photocopies of single chapters may be made for personal use as allowed by national copyright laws. Permission of the Publisher and payment of a fee is required for all other photocopying, including multiple or systematic copying, copying for advertising or promotional purposes, resale, and all forms of document delivery. Special rates are available for educational institutions that wish to make photocopies for non-profit educational classroom use. Permissions may be sought directly from Elsevier Science Rights & Permissions Department, PO Box 800, Oxford OX5 1DX, UK; phone: (+44) 1865 843830, fax: (+44) 1865 853333, e-mail: [email protected]. You may also contact Rights & Permissions directly through Elsevier's home p a g e (http://www.elsevier.nl), selecting first 'Customer Support', then 'General Information', then 'Permissions Query Form'. In the USA, users may clear permissions and make payments through the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, USA; phone: (978) 7508400, fax: (978) 7504744, and in the UK through the Copyright Licensing Agency Rapid Clearance Service (CLARCS), 90 Tottenham Court Road, London WlP 0LP, UK; phone: (+44) 171 631 5555; fax: (+44) 171 631 5500. Other countries may have a local reprographic rights agency for payments. Derivative Works Tables of contents may be reproduced for internal circulation, but permission of Elsevier Science is required for external resale or distribution of such material. Permission of the Publisher is required for all other derivative works, including compilations and translations. Electronic Storage or Usage Permission of the Publisher is required to store or use electronically any material contained in this work, including any chapter or part of a chapter. Except as outlined above, no part of this work may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without prior written permission of the Publisher. Address permissions requests to: Elsevier Science Rights & Permissions Department, at the mail, fax and e-mail addresses noted above. Notice No responsibility is assumed by the Publisher for any injury and/or d a m a g e to persons or property as a matter of products liability, negligence or otherwise, or from any use or operation of any methods, products, instructions or ideas contained in the material herein. Because of rapid advances in the medical sciences, in particular, independent verification of diagnoses and drug dosages should be made. First edition 2000 Library of Congress Cataloging in Publication Data A catalog record from the Library of Congress has been applied for. ISBN: 0-444-50517-2

(~ The paper used in this publication meets the requirements of ANSl/NISO Z39.48-1992 (Permanence of Paper). Printed in The Netherlands.

Organised by Istituto Veneto per i Beni Culturali

In association with ICCROM-International Centre for the Study of the Preservation and Restoration of Cultural Property SMITHSONIAN INSTITUTION-Smithsonian Center for Materials Research and Education UVO-Unesco Venice Office CNR-Consiglio Nazionale delle Ricerche. Progetto finalizzato Beni Culturali ICR- Istituto Centrale per il Restauro UniversitS. degli Studi Ca' Foscari, Corso di Laurea in Conservazione dei Beni Culturali IUAV-Istituto Universitario di Architettura di Venezia

International Scientific Committee Vsevolode Romanowsky (President), La Rochelle, France Andreas Arnold, Zurich, Switzerland Susan Bradley, London, United Kingdom A. Elena Charola, New York, USA Jos6 Delgado Rodrigues, Lisbon, Portugal Wieslaw Domaslowski, Torun, Poland Rosa Esbert, Oviedo, Spain Vasco Fassina, Venice, Italy Marisa Laurenzi Tabasso, Rome, Italy Lorenzo Lazzarini, Venice, Italy Isabelle Pallot Frossard, Paris, France Josef Riederer, Berlin, Germany Raffaella Rossi Manaresi, Bologna, Italy Ornella Salvadori, Venice, Italy Theodore Skoulikidis, Athens, Greece V6ronique Verges Belmin, Paris, France George Wheeler, New York, USA

vi

Honorary Committee Marino Folin Stefano Gasparri PierFrancesco Ghetti Angelo Guarino Vladimir Kouzminov Marc Laenen Pierre Lasserre Paolo Morachiello

Lionello Puppi Maurizio Rispoli Mario Serio Gabriele Zanetto Francesco Zofrea

Rettore Istituto Universitario di Architettura di Venezia Preside della Facolt/~ di Lettere e Filosofia dell'Universit/t degli Studi di Venezia Preside della Facolt/t di Scienze Matematiche, Fisiche e Naturali dell'Universit/t degli Studi di Venezia Presidente Comitato di Progetto Finalizzato Beni Culturali, CNR, Roma Deputy director of Unesco Venice Office General Director International Centre for the Study of the Preservation and Restoration of Cultural Property, Rome Director of Unesco Venice Office Presidente del Corso di Laurea in Storia e Conservazione per i Beni Architettonici e Ambientali dell'Istituto Universitario di Architettura di Venezia Presidente del Corso di Laurea in Conservazione dei Beni Culturali dell'Universit~ degli Studi di Venezia Rettore Universit~t degli Studi Ca' Foscari di Venezia Direttore Generale Ministero per i Beni e le Attivit/t Culturali Presidente del Corso di Laurea in Scienze Ambientali dell'Universitfi degli Studi di Venezia Presidente Eni Tecnologie, Milano

Organising Committee Vasco Fassina, Corso di Laurea in Conservazione dei Beni Culturali, Chairman Guido Biscontin, Universit/t agli Studi Ca' Foscari di Venezia Monica Favaro, Istituto Veneto per i Beni Culturali Giovanni Perego, Eni Tecnologie, Milano Lionello Puppi, Universit/t degli studi Ca' Foscari di Venezia Renzo Ravagnan, Istituto Veneto per i Beni Culturali Akatsuki Takahashi, Unesco Venice Office Alessandro Vigato, Consiglio Nazionale delle Ricerche, Progetto Finalizzato Beni Culturali

vii

Under the Patronage Ministero per i Beni e le Attivit~ Culturali Consiglio Nazionale delle Ricerche ICCROM-International Centre for the Study of the Preservation and Restoration of Cultural Property Regione Veneto Provincia di Venezia Universitfi degli Studi Ca' Foscari di Venezia IUAV-Istituto Universitario di Architettura di Venezia

Sponsors CNR, Progetto Finalizzato Beni Culturali Regione Veneto Scuola Grande San Giovanni Evangelista Eni Tecnologie, Milano Mazzali Systems S.p.A., Milano Philips Electron Optics-FEI, Milano Torggler S.p.a., Merano Dionex s.r.l., Roma

Organising Secretariat Monica Favaro, Chairperson Francesca Crivellari Gianfranco Favaro Damiana Magris Andrea Naccari Marta Pigo Raffaella Portieri Mariangela Rossi

Editing Francesca Crivellari Gianfranco Favaro


ix

Message de bienvenue de M. Vsevolode Romanovsky Monsieur le Pr6sident, Mesdames, Messieurs, En organisant, il y a 28 ans, le premier 'Congr6s International sur la D6t6rioration et la Pr6servation des Pierres en ouvre' je ne pr6voyais pas un 96me Congr6s en l'an 2000. Cette ann6e nous paraissait si lointaine et moi-m6me, je n'6sp6rais pas pouvoir y parvenir. Pourtant nous y sommes en 2000 et dans le merveilleux site de Venise que j'ai si bien connu dans les ann6es 60 et 70. A la suite de ce l er Congrbs a la Rochelle, que j'avais organis6 malgr6 de nombreuses oppositions nationales, nous avons pris l'initiative de cr6er le 'Comit6 International Permanent pour l'Organisation des Congr6s sur la D6t6rioration et la Pr6servation de la Pierre', charg6 de fixer les lieux et les dates des Congrbs successifs ainsi que d'aider les responsables nationaux dans l'organisation de ces manifestations. Le Comit6 fonctionna d'une mani6re satisfaisante et aboutit au pr6sent Congr6s. Apr6s La Rochelle, notre ami Theodore Skoulikidis d6cida, sans h6sitations, de nous inviter au suivant, fi Ath6nes en 1976. Ce fut lui qui lanca r6ellement la s6rie. Pour cela nous lui devons une grande reconnaissance. En 1979 Lino Marchesini organisa le 36me Congr6s ~ Venise. En 1982 le 46me traversa l'Atlantique car il eut lieu ~ Louisville (Kentucky) aux Etats Unis. Par suite de la conjoncture 6conomique de l'6poque, la participation europ6enne fut assez r6duite. En 1985, mon vieil ami Vinicio Furlan nous offrit un magnifique Congr6s 5, Lausanne. Ce fut le 56me de la s6rie. En 1988, le Congr6s organis6 par Wieslaw Domaslowsky fut exceptionnel car il se d6roula fi Torun (Pologne) encore sous influence sovi6tique. Malgr6 cette servitude, Wieslaw Domaslowsky parvint /~ cr6er une ambiance de libert6, de joie et de gaiet6 aussi tousles participants revinrent enchant6s de ce s6jour dans une ville trbs belle. En 1992, Josh Delgado-Rodriguez nous offrit fi Lisbonne, un Congr~s d'une efficacit~ et d'une magnificence dignes de ce beau pays et ses sympathiques habitants. En 1996, Josef Riederer organisa le Congr6s fi Berlin. Maintenant, pour la seconde fois, nous sommes ~t Venise, l'une des plus belles villes du monde, malheureusement tr6s atteinte par la 'maladie de la pierre' due fi tous les exc6s de notre civilisation. Une fois de plus, je suis Pr6sident du Comit6 Scientifique, mais probablement la derni6re fois. Je dois/t mon ami Vasco Fassina, et fi tous ses collaborateurs, mes vives et sinc6res excuses de n'avoir pas pu, compte tenu de mon fige et de mes occupations d'6crivain scientifique, les aider et les assister dans l'organisation de ce Congr~s. Je leur dois de chaleureux remerciements d'avoir bien voulu me confier, probablement la dernibre fois, la Pr6sidence du Comit6 Scientifique. Je suis persuad6 que ce Congr6s sera un grand succ6s et que l'on avancera dans la r6solution des problbmes que posent les 'maladies de la pierre'. Avant de terminer ce court expos6, off j'ai voulu rappeler l'historique des Congr6s, je souhaite fi tous les

participants un merveilleux s6jour fl Venise, off tout est fi voir et fl visiter. Je fais 6galement des voeux pour un excellent d6roulement des travaux du Congr6s.

xi

Foreword An important part of our culture is chiselled in stone, and we are in danger of losing it. The heritage we have of past and present glories of human creativity is slipping away, slowly, silently, but inexorably and at an increasing rate. Stone decay is not a new phenomenon. It starts as soon as an artefact or structure has been completed, and continues progressively for as long as the object is in contact with any kind of environment. The conservation and preservation of works in stone gradually acquired increasing numbers of practitioners in the early part of the twentieth century. However, the problems of deterioration of exposed stone proved to be intractable to the science and technology of the nineteenth century. It was not until about the second and third decades of the twentieth century that it was justified to speak of a Science of Stone Conservation beginning to take form. Traditionally, preservation restoration and maintenance, considered technically, have been closely related to crafts and prevalent contemporary building practice. This meant that earlier, problems were solved mostly within existing traditional know-how among craftsmen and architects. Successively, during the period of industrialisation, conservation was characterized by a scientific approach strongly influenced by the dominant role played by natural sciences and technology. The conservation of historic monuments, sites and structures constitutes an interprofessional discipline co-ordinating a range of aesthetic historic, scientific and technical methods. Conservation is a rapidly developing field, which, by its true nature, is a multidisciplinary activity with experts respecting one another's contributions and combining to form an effective team. Conservation is an artistic activity aided by scientific and historical knowledge. The problem of conserving architecture and the fine and decorative arts is not simple. Even in a scientific age that has developed the technology of space travel and atomic power, the solution to local environmental problems and the prevention of decay still present a major challenge. Only through understanding the mechanisms of decay and deterioration can we increase conservation skills for prolonging the life of cultural property for future generations, but we must admit that decay is the Law of Nature and we can only slow the process down. The first meetings on stone conservation took place in 60's, while only in 1972, in La Rochelle, prof. Romanovsky organised the first International Congress on Deterioration and Conservation of Stone. The first series of events started then, and were systematically repeated every 4 years. At that time the panorama was completely different, very few people were working in this field. This series of Congresses has strongly contributed to the increase in the quality of research and the number of people working in this field. A significant degree of understanding of the nature and mechanism of action of the various decay processes has been developed, and detailed understanding can be expected to be followed by successful techniques of intervention.

xii

The present is, therefore, a propitious time to survey the state of our knowledge, to review the treatment methods that have been proposed and tried out, and to consider what needs to be further explored, and what experience should not be repeated. 21 years after the third edition of the International Congress on Deterioration and Conservation of Stone, organised by prof. Lino Marchesini, Venice is hosting the ninth edition of this prestigious Congress. In these two decades the number of scientific meetings held each year around the world has increased a lot and is a definite proof of the great interest of the Scientific Community to give appropriate answers to stone conservation problems. The IX Congress is a good opportunity for presenting the most recent developments of research on stone decay and to discuss the ways of having them transformed in methods and procedures of practical use in stone conservation. The main purpose of this Congress is to point out: 9 the most appropriate methodology for the assessment of the degree of the weathering of stone, 9 the development of new methods and instruments for the diagnosis of the state of conservation, for the study of alteration mechanisms and for conservation treatments, 9 the definition of Technical European Standard Methods for the evaluation of conservation treatments of artistic and historic stone objects and monuments. The Scientific Committee is deeply interested in having valuable research results and demonstrations of their actual or potential applications to real life degraded stone monuments. The Congress is addressed to: 9 Restorers of works of art who want to improve their knowledge in conservation problems 9 Architects who seek full information on restoration problems 9 Conservators who want to exchange their knowledge and experience 9 Scientific people (geologists, chemists, physicists, biologists, mineralogists, etc.) involved in the conservation field 9 Public authorities and governmental institutions with responsibility for conservation of the Cultural Heritage 9 Art historians dealing with conservation problems It is highly gratifying to acknowledge the interest and enthusiasm that this Congress has raised worldwide. About 400 people coming from 45 countries have expressed their will to participate in the Congress and more than 270 abstracts were received, from which 176 papers have been selected. The members of the Permanent Scientific Committee were asked to review the papers and to approve their publication by dividing them into the following seven themes: 1. Weathering of natural stone: causes, mechanism and measurement of stone damage 2. External factors of decay: environmental influence on stone decay 3. Biological damage on stone

xiii

4. Laboratory methods and techniques 5. In situ evaluation of damage 6. The conservation of stone: treatment methods and products 7. Case studies of conservation of Cultural Heritage The diversity of subjects dealt with in the submitted papers and their scientific level are a demonstration of the opportunity of this event and will certainly make these proceedings a very important reference book for people involved in the field of stone conservation. The view expressed by the individual authors of these papers are not necessarily those of the editor nor of the Scientific Committee. Since no modifications were asked for, the content of the paper remains, the full responsibility of the authors. Achievement of the Congress' aims is due, in great measure, to sponsoring organisations, Institutes and the staff of the Organising Committee and the Secretary of the Istituto Veneto per i Beni Culturali which made a precious work of editing a cameraready copy of the papers for the International publisher. Sincere thanks are expressed to the members of the Permanent Scientific Committee for their assistance in the selection process of these proceedings. The editor wishes to thank all the participants attending from all over the world for their interest towards this Congress and hope that people have a fruitful discussion and exchange of opinions in order to improve their knowledge thus allowing a better preservation of our heritage of the past and present glories of human creativity. It should be desiderable that this Congress could represent a milestone in the slow change of the classical approach to the study of conservation of cultural property based on the knowledge of the causes of decay of specific objects. In fact a new focus on historic buildings, structures and materials is recently developing. Major attention was paid to management of conservation worksites, regular inspection of historic structures and maintenance strategies. Regular inspections of cultural property are the basis of sound management and can be used to develop a preventive maintenance strategy, which can greatly reduce the cost of caring for our cultural heritage. Developing a preventive maintenance strategy is the most important step in preparedness for a natural disaster. Half of the damage caused by an earthquake to historic buildings can generally be attributed to lack of maintenance. It is a great pleasure for the Scientific Venetian Community to host this Congress on the threshold of the third millennium in a scientific age that has developed a sophisticated technology but has not completely solved the problems of safeguard of our Cultural Heritage. Vasco Fassina


XV

TABLE OF CONTENTS VOLUME 1

Theme 1 - Weathering of natural stone" causes, mechanism and measurement of stone damage

Restoration of the historical brick masonry Bajare D., Svinka V. Comparative study of different methods for gap filling applications and use of adhesives on the biocalcarenite surfaces of the 'Tempio della Concordia' in Agrigento Bennardo C., Meli P., Biscontin G., Berlucchi N., Ginanni Corradini R., Mattolin F. Gr6dener sandstone, a historical building material in south Tyrol/Italy- The problem of large variability of stone properties for monument conservation Franzen C., Mirwald P. Evaluation of materials used in the replacement of sculptures in historical monuments Boutin F,. Bromblet P. White granites used in lombard architecture Bugini R., Pavese A., Borroni S., Folli L. A research into intrinsic parameters material to the durability of highly porous building stones Calia A., Mecchi A. M., Quarta G. Preliminary contribution on durability of some macroporous monumental stones used in historical towns of Campania Region, Southern Italy Langella A., Calcaterra D., Cappelletti P., Colella A., de' Gennaro M., de Gennaro R. Durability of tuffeau stone in buildings: influence of mineralogical composition and microstructural properties Dessandier D., Bromblet P., Mertz J-D. Water-rock interaction and monuments stone decay" the case of Basilica da Estrela, Portugal Figueiredo A. M. C., Marques J.M., Mauricio A.M., Aires-Barros L. Analyses of the physical parameters correlated to bending phenomena in marble slabs Garzonio C. A., Fratini F.F., Manganelli del F~ C., Giovannini P., Cavallucci F. Geoegyptology of A1-Muzawaka tombs, Dakhla oases, Egypt Helmi F. M. Thermal stress and weathering of Carrara, Pentelic and Ekeberg marble Lindborg U., Dunakin R. C., Rowcliffe D. J.

13

25

31 41

49

59

69

79

89 99 109

xvi

Weathering of runestones in a millenian perspective Lbfvendahl R., Gustavson H., Lundberg B. A. Petrophysical analysis of the sculptures decay at the Cathedral of Burgos, Spain Fort Gonzdles R., L6pez de Azcona M. C., Mingarro Martin F. Durability of sandstones in Serbian ancient monasteries and modern buildings Matovic V.B., Milovanovic D. J., Joksimovic S. M. Sandstone architectural deterioration in Petra, Jordan Paradise T. R. Preliminary studies for the consolidation of Guadalupe tuff from the Philippines Paterno M. C., Charola A. E. Secondary phosphate phases in altered trachyte from S. Miguel Island (Azores, Portugal) - A possible contribution to the stone degradation Pruddncio M. I., Nasraoui M., Trindade M. J., Sequeira Braga M. A., Figueiredo M. O. Comparison between traditional and chamber accelerated ageing tests on granitic rocks Rivas T., Prieto B., Silva B., Birginie J. M. Physical properties and durability of fresh and impregnated limestone and sandstone from central Sweden used for thin stone flooring and cladding Sahlin T., Malaga-Starzeg K., Stigh J., Schouenborg B. Stress from crystallization of salt in pores Scherer G. W. Stone materials used in the masonry of Ortigia (Siracusa, Sicily) Calia A., Mecchi A. M., Scudeler Baccelle L. Control of marble weathering by thermal expansion and rock fabrics Siegesmund S., Weiss T., Tschegg E. K. The relationship between deterioration, fabric, velocity and porosity constraint Weiss T., Siegesmund S., Rasolofosaon P. N. J. Saline pollution in trachyte monuments of the Azores Islands (Portugal) Alves C. A. S., Sequeira Braga M. A., Trancoso A. Trachyte stones in monuments of the Silo Miguel and Terceira Islands, Azores (Portugal) Sequeira Braga M. A., Figueiredo M. 0., Prud~ncio M. I, Delgado Rodrigues J., Alves C. A. S., Costa D., Silva T., Trindade M. J., Waerenborgh J. C., Nasraoui M., Gouveia M. A. Determination of structural anisotropy of Carrara marble with ultrasonic measurements Sheremeti-Kabashi F., Snethlage R.

119

125

135 145

155

165

171 181 187 195 205 215 225

235

247

xvii The stone of Piraeus at the monuments of the Acropolis of Athens Theoulakis P., Bardanis M. Petrophysical properties modifications of Strasbourg's Cathedral sandstone by black crusts Thomachot C., Jeanette D. Freeze-thaw resistance of the Yazilikaya tufts Topal T., SSzmen B. An evaluation of geology and weathering in the preservation of marl objects Ventikou M., Halls C., Lindsay W., Batchelder M., Hubbard C.

265

Theme 2 - External factors of decay: environmental influence on stone decay

293

Topoclimatic mapping, a tool for cultural heritage conservation: the case of Roman Theater of Lisbon, Portugal Aires-Barros L., Dionisio A. Characterization of surface morphology of carbonate stone and its effect on surface uptake of SO2 Bede E. A. Sea water absorption, permeability evolution and deterioration assessment of building stones subjected to marine exposure Birginie J. -M. Colour changes and reactivity to SO2 of some cladding stones at the 'Gran Theater del Liceu' (Barcelona-Spain) Grossi C. M., Esbert R. M., Alonso F. J., Valdeon L., Ordaz J., Diaz-Pache F. Early mechanisms of development of sulphated black crusts on carbonate stone Ausset P., Lefbvre R. A., Del Monte M. Past air pollution recordings on stone monuments: the heads of the r~Nigs'oc Juda statues from Notre-Dame cathedral (Paris) Ausset P., Lefbvre R. A., Del Monte M., Thidbault S. Laboratory investigations of weathering behaviour of fresh and impregnated limestone and sandstone from Central Sweden Malaga-Starzeg K., Sahlin T., Lindqvist O. The influence of building orientation on climate weathering cycles in Staffordshire, U. K. Mitchell D. J., Halsey D. P., Macnaughton K., Searle D. E. Corrosion of limestone in humid air containing sulphur and nitrogen dioxides: a model study Moroni B., Poli G. The Doria Pamphilj exhibition Gallery: the study of environmental conditions Artioli D., Giovagnoli A., Nugari M. P., Ivone A., Lonati G.

255

275 283

295

303

313

323

329

339

349

357

367

375

xviii Analytic methodologies for carbon compound identification: Leaning Tower and Baptistery of Pisa Sabbioni C., Ghedini N., Gobbi G., Riontino C., Zappia G. The effects of coal and diesel particulates on the weathering loss of two major building stones in the United Kingdom- A comparative microcatchment study Searle D. E., Mitchell D. J., Halsey D. P., Dews S. J., Smith J. P. Evaluation of the environmental influence on sulphate salt formation at monuments in Dresden (Germany) by sulphur isotope measurements Siedel H., Klemm W. Granitic building stone decay in an urban enviroment: a case of authigenic kaolinite formation by heterogeneous sulphur dioxide attack Schiavon N. Theme 3 - Biological damage on stone

Polysaccharides as a key step in stone bio-erosion Albertano P., Bruno L., Bellezza S., Paradossi G. The temples of the archaeological area of Paestum (Italy): a case study on biodeterioration Altieri A., Pietrini A. M., Ricci S., Roccardi A., Piervittori R. Biological patinas on the limestones of the Loches Romanic Tower (Touraine, France) Zagari M., Antonelli F, Urzi C. Chemiolithotrophic bacteria on stone monuments: cultural methods set-up Bartolini M., Monte M. New methods to study the detrimental effects of poikilotroph microcolonial micromycetes (PMM) on building materials Dornieden T., Gorbushina A. A. Diversity of heterotrophic bacteria isolated from three European mural paintings Heyrman J., Mergaert J., Swings J. A study of biologically decayed sandstone with respect to Ca and its distribution Jones M. S., Wakefield R. D., Forsyth G. Microbial environmental monitoring of stone cultural heritage Pitzurra L., Giraldi M., Sbaraglia G., Bistoni F., Bellezza T., Spera G. The Silo Sebastiio Church of Terceira Island (Azores, Portugal)Characterisation of the stones and their biological colonisation Romeo P., PrudYncio M. I., Trindade M. J., Nasraoui M., Gouveia M. A., Figueiredo M. 0., Silva T.

383

391

401

411 423

425

433

445 453

461

469

473 483

493

xix Rapid diagnosis of microbial growth and biocide treatments on stone materials by bioluminescent low-light imaging technique Ranalli G., Pasini P., Roda A. The action of Caloplaca Citrina on concrete surfaces: a preliminar study Rosato V. G., Traversa L., Cabello M. N. Endolithic lichens and conservation: an underestimate question Pinna D., Salvadori O. Biological colonization features on a granite monument from Braga (NW, Portugal) Leite Magalh6es S., Sequeira Braga M. A. Efficiency of biocide in 'in situ' and 'in vitro' treatment. Study case of the 'Templete de Mudejar', Guadalupe, Spain Urzi C., De Leo F., Galletta M., Salamone P., Balzarotti R. Theme 4 - Laboratory methods and techniques

Instrumental chemical analysis of the more common marbles historically used for decorative purposes or to create works of art Campanella L., Gregori E., Grossi R., Tomassetti M. Bandini G. Presence of D, L amino acids in oxalate patinas on a stone monument Casoli A., Negri S., Palla G. Unite Technologique portable et autonome de diagnostic- Analyse Investigation, de choix d'intervention avec video assistance a distance et banque de donnees Catalafini J. Large scale experimental facilities at ENEA for seismic tests on structural elements of the historical/monumental cultural heritage De Canio G. Evaluation of stone pore size distribution by means of N M R Alesiani M., Capuani S., Curzi F., Mancini L., Maraviglia B. New results in the application of innovative experimental techniques for investigation of stone decay's processes Giorgi R., Baglioni P., Alesiani M., Capuani S., Mancini L., Maraviglia B. Fractal geometry description of the permeability of a natural fissured rock Miguel A. F., Rosa R., Silva A. M. Microstructural changes in granitic rocks due to consolidation treatments: their effects on moisture transport Mosquera M. J., Rivas T., Prieto B., Silva B. The use of sound velocity determination for the non-destructive estimation of physical and microbial weathering of limestones and dolomites Papida S., Murphy W., May E.

499 507 513

521

531 541

543 553

557

565 579

587 595

601

609

XX

Analytical techniques for characterizing polychromated coatings on quartzite samples from a prehistorical cave Carmelo Prieto A., Jimdnez J., Pdrez B., Leal L. Stone drying: an approach of the effective evaporating surface area Tournier B., Jeanette D., Destrigneville C. Fractal modelling of particulate deposition in the development of black crusts on stone Watt J., Massey S., Kendall M. Characterization and physico-chemical action of condensed water on limestone surfaces Zendri E., Biscontin G., Kosmidis P., Bakolas A. Author index

TABLE

OF CONTENTS

VOLUME

619 629

637

647 657

2

Theme 5 - In situ evaluation of damage

The effects of the strong use of cements in restoration: the case of Barga Duomo (Northern Tuscany) Baccaro M. L. P., Balzi S., Del Chiaro L., Vannucci S. A simple technique for rapid field assessment of stone decay on buildings Ball J., Young M. E. 'La Fenice' T h e a t r e - Foyer and Apollinee r o o m s - Consolidation of firedamaged stucco and marmorino decorations by means of combined applications of ion-exchange resins and barium hydroxide Berlucchi N., Ginanni Corradini R., Bonomi R., Bemporad E., Tisato M. A petrographic atlas as a decision-tool in replacement and substitution of ornamental stone in historical buildings and monuments Dingelstadt C., Dreesen R., Thorez J., Lorenzi G., Bossiroy D., Antenucci D., Banier J. Study on the deterioration and conservation of the stone monument in 'Dell'Aquila' Square, Ravenna (Italy) Macchiarola M., Fiori C., Belacchi S. Deterioration of rock monuments in Petra/Jordan Heinrichs K., Fitzner B. The risk map and the blackening index: a new recording apparatus Giovagnoli A., Marabelli M., Canegallo P., Ivone A. A four-year survey of the water contents and movements within a masonry core after a restoration campaign: a case study in Notre-Dame la Grande (Poitiers, France) Godin J., Pithon M., Vergks Belmin V. Conservation of stone flooring, ancient and modern Hunt B. J., Grossi C. M.

13

23

33

43 53 63

73 83

xxi The restoration of the Ursino Castle (century XIII) in Catania Barone G., Ioppolo S., Majolino D., Migliardo P., Muscarh A., Neri N. F. Decay mapping of polishable limestone Martin B., Mason D., Bryan P. Conservation of the pigmented plaster in "Red Temple" at Monte D'Accoddi (North Sardinia) Massidda L., Meloni P., Piras M. G., Sanna U. Innovative strategies for the preservation of historic cities by ND monitoring techniques and GIS management of data regarding environmental impact on historic materials and structures Moropoulou A., Koui M., A vdelidis N. P. Deterioration features in the apse of Orvieto Cathedral (Terni, Italy): a mechanical model Moroni B., Poli G. Ultrasonic measurements on weathering alpine marble. A study on field exposed samples and on the medieval marble. Portals of Schloss Tirol/South Tyrol-Italy Recheis A., Bidner T., Mirwald P. Deterioration characteristics of columns from the Marmorpalais Potsdam (Germany), by ultrasonic-tomography Siegesmund S., Pretzschner C., Ruedrich J., Lindner H., Weiss T., Richter I., Richter D., Woyde M. Damages in monuments produced by the corrosion of metallic junctions. The Acropolis case Skoulikidis T., Vassiliou P. Durability of consolidants on a French altered limestone after eighteen years of natural ageing Vallet J.M., Simon S., Mertz J. D., Martinet G. Deterioration and conservation of monuments of Latvia Lgtsis R., Vit~ia I., Igaune S. Quantification of the long-term effects of stone-cleaning on decay of building sandstones Young M. E., Ball J., Laing R. A. The Coltea Church in Bucharest Zbirnea I.M., Bonafede L. Color and weight evolution of limestones protected by water repellents after three-year ageing period in urban conditions Boutin F., Leroux L.

91 101

111

119

129

139

145

155

165 175

179 187

197

xxii Theme 6 - The conservation of stone" treatment methods and products

An integrated approach to design fluoro substituted 'smart' polymers for protection of monumental buildings Aglietto M., Castelvetro V., Ciardelli F., Matteoli U., Botteghi C., Chiantore 0., Lazzari M., Alessandrini G., Peruzzi R., Toniolo L., Fassina V. Effect of fluorinated groups on the photooxidative stability of polymeric protectives applied on calcareous stone Chiantore 0., Poli T., Aglietto M., Castelvetro V., Peruzzi R., Colombo C., Toniolo L. Effects of combined application of biocides and protectives on marble Malagodi M., Nugari M. P., Altieri A., Lonati G. A comparative study of the efficiency of siloxanes, methacrylates and microwaxes-based treatments applied to the stone materials of the Royal Palace of Madrid, Spain Fort Gonzgtles R., Ldpez de Azcona M. C., Mingarro Martin F., Alvarez de Buergo Ballester M., Rodriguez Blanco J. Study of porosity and specific area evolutions on porous material depending on hydrophobic treatments Naizot S, Barbary P., Mark S. Performance testing of transparent protective coatings on Globigerina Limestone Cassar J., Tonna G., Torpiano A., Zammit G. Evaluations of the effectiveness of innovative perfluoropolyurethanes polymers as consolidants for porous materials Croveri P., Chiavarini M. Assessment of durability of water repellents by means of exposure tests Ferreira Pinto A. P., Delgado Rodrigues J. D6consolidation par absorption d'eau de gr~s trait6s avec le silicate d'6thyle. Mesures non destructives de E, G e t v. Fdlix C., Ferrari P., Queisser A. Injectable slurries for the in situ conservation of pavement mosaics Flatt R. J., Girardet F. J. Silica bound mortars for the repairing of outdoors granite sculptures Rolland 0., Floc'h P., Martinet G., Vergks Belmin V. Ultrasonic testing method for the characterization of Pietra d'Istria structural elements Almesberger D., Geometrante R., Rizzo A., Suran P. Ionexchange resins for historic marble desulfatation and restoration Guidetti V., Uminski M. Development of lime mortars with improved resistance to sodium chloride crystallization Henriques F. M. A., Charola A. E.

207

209

215 225

235

245

251

263 273

287 297 307

317 327

335

xxiii Slaked lime mortar: comparison between two samples supposed to be alike Jornet A., Romer A. A comparative study of mortars containing barium hydroxide (Ba(OH)2). Application on monument's conservation Lambropoulos V. N., Ghiossi S., Karatasios I. Measuring the penetration depth of consolidating products: comparison of six methods Leroux L., Vergbs Belmin V., Costa D., Delgado Rodrigues J., Tiano P., Snethlage R., Singer B., Massey S., De Witte E. Preliminary study on the set up of mortars displaying biocidal activity Ferone C., Pansini M., Mascolo M. C., Vitale A. Durability of tufaceous stones treated with protection and consolidation products Dell'Agli G., Ferone C., Mascolo G., Marino 0., Vitale A. Mineral inorganic treatments for the conservation of calcareous artefacts Lanterna G., Mairani A., Matteini M., Rizzi M., Scuto S., Vincenzi F., Zannini P. Change in properties of the stone treated with historical or modern conservation agents Maxov~ I. Criteria and methodology for restoration mortars compatible to the historic materials and structures Moropoulou A., Bakolas A., Moundoulas P. The role of consolidants in the conservation of Sydney sandstone buildings O'Connor J. Chemistry for conservation of cultural heritage: application of in situ polymerisation for the consolidation and protection Vicini S., Parodi V., Simonetta M., Moggi G., Pedemonte E. Study of the colourings of the St. Peter Fagade (Vatican) Previde Massara E., Perego G. A new white cement resistant to sea-water. Development of a white repairing mortar Puertas F., Blanco-Varela M. T., Palomo A., Vhzquez T. Hydrophobic materials - how effective are they? Puterman M. Physical properties of fine grained marble before and after conservation Rohatsch A., Nimmrichter J., Chalupar I. Compatible consolidants from particle-modified gels Escalante M. R., Valenza J., Scherer G. W. Performance evaluation of preservative coatings on stone surface of heritage buildings having hygric state Sharma R. K., Saxena V. K., Saxena K., Tewari S. K.

343

351

361 371

379

387

395

403 413

419 425

435 443 453 459

467

xxiv Dispersed hydrated lime for the preservation and conservation of stone monuments Strotmann R., Maryniak-Piaszczynski E. Ceramic additions in the restoration of stone sculpture and ceramics Tcheremkhine V. I. Effectiveness of surface treatments for sedimentary limestone in Greece Theoulakis P., Tzamalis A. Assessment of the performance of silane treatments applied to Egyptian limestone sculptures displayed in a museum environment Thickett D., Lee N. J., Bradley S. M. Scientific investigation and large scale sandstone treatments: the Washington State Legislative Building Twilley J., Leavengood D. The conservation of the sculpture work of the National Monument in Amsterdam Van Hees R. P. J., Larbi J. A. Development and assessment of a conversion treatment for calcareous stone Weiss N. R., Slavid I., Wheeler G. Evaluation of alkoxysilane coupling agents in the consolidation of limestone Wheeler G., Mdndez-Vivar J., Goins E. S., Fleming S. A., Brinker C. J. Regularities of conservation of porous material-ancient terracotta of Prichernomorye by acrylic polymer solutions Levko L. V., Yemelyanov D. N. Impregnation and strengthening of porous stone by acrylic polymer solutions Yemelyanov D. N., Volkova N. V., Pavlovskaya M. V. New proposals for the conservation-consolidation of stones and plasters: analytical characterization and trial applications of Ba aluminates Messori M., Zannini P., Mairani A., Matteini M. Integration of laser with conventional techniques in marble restoration Siano S., Pini R., Salimbeni R., Giamello M., Scala A., Fabiani F., Bianchini P. In field tests and operative applications of improved laser techniques for stone cleaning Pini R., Siano S., Salimbeni R. Results of Laser cleaning on encrusted oolithic limestone of angel sculptures from the Cologne Cathedral Siedel H., Hubrich K., Kusch H. G., Wiedemann G., Neumeister K., Sobott R.

477 485 493

503

513

523 533 541

547 553

561

569

577

583

Theme 7 - Case studies of conservation of Cultural Heritage

591

Study of stone deterioration in the Palficio do Freixo in Oporto Begonha A., Teles M.

593

XXV

Polychromy traces and stone decay in the church of S. Maria dei Miracoli in Venice Fassina V. Survey on the polychromy and stone materials of funeral monuments dedicated to Jacopo and Ubertino da Carrara in Eremitani Church in Padua Favaro M., Portieri R., Crivellari F., Naccari A., Spiazzi A., Fassina V. New findings on past treatment's effects on the Lunette of San Giovanni Evangelista in Venice Favaro M., Naccari A., Crivellari F., Magris D., Pigo M., Burtet B., Fumo G., Fassina V. La pierre du portail peint de la Cath6drale de Lausanne" nature, 6tat de conservation, et consolidation Furlan V., Fdlix C., Queisser A. The Roman aqueduct of Carthage: a minerochemical study on water conduit mortars and deposited crusts Figueiredo M. 0., Veiga J. P., Pereira Da Silva T., Alvarez A., Torrens F., Khosrof S., Ferchiou N. CastelManiace, Syracuse (Sicily): the deterioration of the marble of the monumental portal and window Alberti S. A., Antonelli F., Cancelliere S., Lazzarini L., Mannuccia F., Santalucia F. The South Portal of Sint Martinusbasiliek in Halle, Brabant: technical study and conservation de Henau P., Leirens I. Documentary and analytical analogies in the study of patinas of the 'Quattro santi coronati' by Nanni di Banco Giusti A. M., Lalli C., Lanterna G., Matteini M., Rizzi M. The case of the 'Portale of S.Maria a Mare' in Giulianova: results of physicochemical inquires Amorosi E., Di Marco F., Rosignoli R., Quaresima R., Scoccia G., Volpe R. Study of stone deterioration in the cloister of the Mosteiro de Grij6, Portugal Begonha A., Sequeira Braga M. A. Study of weathering factors and evaluation of treatments for the stones of 'Santa Maria de la Encarnaci6n' Church, Constantina (Sevilla, Spain) Villegas Sanchez R., Espinosa-Gaitan J., Alcalde Moreno M. The Aragonese Portals in eastern Sicily: relationship between form, materials, decay and the environment Salemi A., Sanfilippo G. The Columbus monument at Huelva (sw Spain): preliminary survey on stone decay Galgm E., Carretero M.I., Bernabe J.M., Fernandez-Caliani J.C., Requena A.

603

613

623

633

641

649

661

671

679 689

697

707

715

xxvi Conservation problems of the statue of Saint Michael by Raffaello da Montelupo, Castel Sant'Angelo, Rome Fiori C., Lorusso S., Casalicchio G., Pentrella R., Prestileo F. Painted sandstone as protection and as an architectural and historical concept Andersson T., Von Haslingen B. Detaching methodology for fresco paintings. The case study of a renaissance cycle Casadio F., Colombo C., Toniolo L. Weathering of painted marly limestones in the Temple ruin of Merenptah, Qurna/Luxor, (Egypt) Zehnder K., Arnold A., Kung A. Ossuccio (CO): a case study to assess the causes of degradation in some terracotta statues Valentini M. Old Khmer styled sandstone monuments in Thailand. Aspects of weathering and development of conservation concept Wendler E., Prasartset C. Ethical issues in the restoration of stone sculpture in the State Tretiakov Gallery. Moscow: evolution of methods and elaboration of new polymer materials Vassilieva O. A. The deterioration of Nubian sandstone blocks in the Ptolemaic temples in upper Egypt Abd El-Hady M. M. Conservation and safeguard of stone rural buildings: an example in a mountain area Agostini S., Calvi G. A historiography of recent past interventions at the ancient theatre of Ephesos Aktiire Z. The implication of stone cleaning for planned building maintenance Laing R. A., Ball J., Scott J., Young M. E. Campo dei Greci in Venice: the case of conservation of San Giorgio of the Greeks Ioannidou N. Masonry of Abruzzo historical buildings D'Anselmo M. The rupestral monument of Basarabi-Murfatlar. Conservation of the decoration incised in the chalk walls Niculescu G., Vlad A. M. Investigations on technology of joint mortars in brick walls Domastowski W.

721

731

739

749

759

765

775

783

793

801 813

819 829

837 843

xxvii The knowledge of the plasters typical of the buildings of Ortigia (Siracuse, Italy). Part 1 - Finishing layers Alessandrini G., Negrotti R., Bocci A. M., Amadori M. L., Ercolani G., Fabbri B., Campisi T. The knowledge of the plasters typical of the buildings of Ortigia (Siracuse, Italy). Part 2 - The mortar and mortar/finishing layer combination Alessandrini G., Negrotti R., Bocci A. M., Amadori M. L., Ercolani G., Fabbri B., Campisi T. Preliminary studies about the ancient mortars of the church of Santa Maria de Irache Monastery (Navarra, Spain) Alvarez J.I., Montoya C., Navarro I., Martin A. Study of the lime renderings decay from Plaza de la Corredera, Cordoba, Spain Gonzdles Limdn T., Alvarez De Buergo Ballester M. Problems and solutions in practical restoration of freshwater limestone-tufa Sidraba I., Krage L., Graudums I. Technical aspects of stone conservation in Jerusalem Lobovikov-Katz A. Sacrificial layers for conservation of calcareous stone in Austria-theory, practice and evaluation Nimmrichter J., Koller M., Paschinger H., Richard H. An example of a practical 'cleaning' of the architectural facades of Jubilee Rome Pecoraro I. Investigation of damage in old stone structures caused by the latest strong earthquake in Northern Greece Stavrakakis E. J., Karaveziroglou-Weber M. K., Mavrikakis S. P. Author index

853

863

873

881 889 897

903

913

921 929


Theme 1 Weathering of natural stone" causes, mechanisms and measurement of stone damage


RESTORATION OF THE HISTORICAL BRICK MASONRY Diana Bajare*, Institute of Silicate Materials, Riga Technical University, Riga, Latvia Visvaldis Svinka, Institute of Silicate Materials, Riga Technical University, Riga, Latvia

Abstract

The architectural heritage of Latvia mainly consists of manufactured materials. The Building of Riga City Council, Fortification Wall of Riga, Turaida's Palace, Ventspils Castle, etc. are some examples of historical buildings of Latvia from 12th-17th centuries. These buildings were made from brickwork. The major causes of damages in the structure of historical bricks are: migration of the water-soluble salts, and frost influence in wintertime - mechanic damage of structure by crystallisation and growing of salts or ice crystals. There is an increased corrosion of single brick or brick masonry fragments in the architectural monuments of the Middle Age observed. There is a necessity to replace these bricks with the new ones of the same size, color and physical properties. Nowadays, bricks produced in Latvia do not meet all these requirements, Therefore there is a need to develop new materials suitable for restoration works of brick masonry. The aim of the research is to analyze the old bricks, to get knowledge about the manufacturing process used, and to develop similar materials from local clays for replacement of damaged original bricks in the historical buildings. This paper is mainly aimed to analyze the historical bricks. Key words: brick, mineralogy, texture, pore size distribution, porosity, absorption, density, salts, durability of soluble salts and frost. 1. Introduction

Ventspils Castle is one of the symbols of the military fortifications of Livonian Order, which was built in 13t~ century. During centuries Castle had several rebuilding periods concerning with the growing needs of the city. The poor state of conservation of certain areas needing to be restored, together with the limited information available on the exact chronology of the erection of particular zones, have led to an extensive research project aimed to characterise the materials used in the monument. Investigation of historical bricks taken from different areas could help to estimate the age of materials in zones of uncertain date in ease there are noticeable differences between periods. It could give information about the economical situation in different periods, skills and development of civil engineering. In 1995, the reconstruction and restoration of Ventspils Castle started. The gallery of Ventspils Castle inner yards is the first and most powerful impression of the castle's actual age that the visitor gets passing through the gate vault./1/The interior of the castle in the beginning was designed as a castle- museum, without claiming restoration of any style. One should exhibit things that help to understand essential coherence of the building's development. One should preserve and not exhibit as much as possible out of the rest of the things that exist in the castle and are less essential. At present, when making some changes * Author to whom correspondence should be addressed.

4

9th International Congresson Deterioration and Conservation of Stone, Venice 19-24June 2000

in the building, the restorers have to keep the inherited feeling of the Castle Awareness. By "exhibiting" one must understand priorities of accentuation of certain elements. These priorities have definite aesthetic and cultural historical aims, and they operate with various methods like volume, surface, texture, lights and shades, patina, etc, at the same time preserving the original in as much unchanged condition as possible. Exhibition priority is part and fragments of the brick wall created under impression of the order's construction traditions, namely, large size bricks in jointed brickwork. All other parts of the wall do not fall under priority. This work is the first major event in conservation of large-scale brick wall and plaster in Latvia. Different lime mortars were used. The historical brick walls are conserved with maximum preservation of medial parts. Bricks are widely stuff with lime mortar providing a possibility to preserve partly destroyed parts of the wall in situ in opposite to replacing them with new materials. In many cases dents and grooves of unknown purpose, traces of earlier made repair works, pegs, metal anchors, fragments of painting and plaster are lett open. In places where it is allowed by composition aesthetics, wall erosion traces are preserved, thus indirectly pointing at the condition of wall during a certain period of its history. Even very hopeless brick walls can be saved and made attractive. The conservation of brick walls definitely is a creative process and must be carried out together by architect and brick wall master. 2. Experimental Methods This paper mainly tries to characterise the Middle Age bricks taken from Ventspils Castle (built 13th - 17th). To this aim, the following techniques were applied: visual inspection, Xray diffraction, mercury porosimetry, physical laboratory tests (water absorption, density, open porosity, saturation coefficient) and chemical analysis. 3. Damage of historical brick masonry Brick masonry is composed of brick that are silica based and relatively acidic while the mortar in the most of old buildings is made of lime that is more alkaline. The influence of water on the degradation process is complex. Water affects not only the possibility of the reactants solution but also the presence of salts in the materials. The major cause of damage in the structure of historical bricks is a migration of salts originating from the materials in the masonry, and also from reactants from outdoor pollution or microbiological conversions in the porous system of the brick and the mortar. The frost causes damage only in a wet structure of bricks. Due to frost the bricks usually are damaged - smaller or lager flakes split from the brick surface. A structural fault can make the brick frost susceptible even though the brick frost material was strong. Pieces can also split from the bricks or they can flake off the surface./2/ The low concentration of soluble salts in the masonry do not influence durability of bricks, if the intensive water migration is not observed. Salts can not be incorporated into the hydrogen bonds of ice crystals./3/The dissociated ions in the water can bind so strongly with the water molecules that water can not freeze - depression of freezing point. Salts solution of low concentration attacks a porous material more strongly in the freeze/thawing cycles than solution of high concentration. Highly concentrated salts solution freezes in the form of ice slurry of limited volumetric expansion and relatively low compressive strength. The historical bricks were fired at low temperatures comparatively to the ceramic materials produced at the end of 20th century. The corrosion resistance and porosity of historical bricks are usually higher, but other general properties are similar with the bricks

9th International Congress on Deterioration and Conservation of Stone, Venice 19-24 June 2000

5

produced nowadays. From the deterioration point of view ceramics are characterised by the following aspects: 1. Physical corrosion: erosion of particles in air or water; absorption of water in pores leading to expansion and cracking; freezing of water in pores leading to expansion, cracking and spilling. 2. Chemical corrosion: - acid- base and solubility reactions with surrounding moist air or water. 3. Physico-chemical corrosion: - recrystallization of salts to high- water and high- volume substances; - slow crack growth due to stress corrosion; - efflorescence from soluble salts is usually not harmful to ceramics, but could give unaesthetic appearance to the surface and may indicate future recrystalization damage. In the development and implementation of new materials and structural solutions, the reliable knowledge of frost and soluble salts resistance is always needed. The main purpose is to make bricks with good resistance to soluble salts, with similar physical properties as original ones.

4.

Visual

inspection

The oldest historical bricks (13-14 th centuries) have dimensions up to 30x15x10 cm. They compose the biggest part of the large bricks used in Ventspils Castle. Historical bricks, which are dated of 14-15th centuries, have dimensions up to 30x15x10 cm. That type of historical bricks is more yellowish in comparison with others. They are quite similar with the oldest bricks, but with different color and different additives used for producing. The newest bricks (15-17th centuries) are more reddish and have smaller dimensions: 28x15x6.5 cm.

The historical bricks were hand-made using wooden moulds (sander or water slopmoulded bricks) to perform the shaping. The moulds were smoothed with a appropriate moulding board and, depending on the nature of the body, immediately released from the mould or with the help of soft - mud bodies left for a short time to dry in the mould. The pieces of broken bricks (shamote) or pieces of soft clays and organic like grass or straw were used like additives. The organic components burned out during the firing, though creating more porous materials. Firing conditions and maximum firing temperature of historical bricks were unknown. There were not recognised any visible damages after visual inspection. The historical bricks still are in a good condition.

6

9th International Congresson Deteriorationand Conservationof Stone, Venice 19-24June 2000

5. Chemical analyses Table 1 Chemical analysis of historical bricks N O. Sample sio 2 Ai203 1. Historical brick 59.44 14.99 (13-14~ century) . 2. Historical brick 68.99 14'.58 (14-15~ century) 3. Historical brick 64.96 19.87 (15-17m een.tury)

Fe20i 3.71

CaO., M~O 4.22 2.26

Na20 2.93

K 20 2.93

1.37'

4167

3.41

1.46

1.53

4.97

2.56' 0.83

1.41

3.14

,,

According to the results of chemical analysis, all types of historical bricks were made in the different places and centuries. There were used different types of clays for the production of these bricks. The newest bricks were made of limless days, but oldest bricks were produced from more calcareous days. (Table 1) Bigger amount of iron makes newest bricks more reddish. There was found small amount of soluble salts in the historical bricks. This concentration of soluble salts is not dangerous for ceramic materials. The presence of soluble salts in historical bricks came from the pollution of air and earth. An important role plays geographic location of Ventspils Castle- very close to the Baltic see.

6. Physical properties of historical bricks Table 2: Physical properties of historical bricks 13-14th centuries 1. Waterabs0rption,% ' 17.83 2. Density, g/era3 1.70 3. Open porosity, % 26.39 4. Saturation coefficient .... 0.75

13-14na centuries 19.4~8 1.68 30.84 0.86

13-i4thcenturies ' 18.21 ' 1.81 27.81 0.86

All types of bricks, not depending on chemical composition, have quite similar physical properties. The porosity of these bricks is from 26 to 30 %. The bulk density is between 1.7 and 1.8 g/era3. The water absorption is between 18-19%. (Table 2) The saturation coefficient gives information about possible frost resistance of bricks. The results indicate that bricks are frost resistant if saturation coefficient is smaller than 0,78. The saturation coefficient of oldest bricks is 0,75 (this is one of the characteristic measurements of good frost resistance), but for other types it is higher- 0,86. All types of historical bricks have higher porosity and water absorption to compare with commercial bricks produced nowadays in Latvia.

9th International Congress on Deterioration and Conservationof Stone, Venice 19-24June 2000

7

7. X-ray diffraction None types of historical bricks have illite connection with X-ray diffraction analysis. It means that sintering temperature of historical bricks was higher than 900~

Figure 1 X ray diffraction of historical bricks: A - historical brick from 13-1~.th centuries, B - historical brick from 14-15th centuries; B-light - historical brick from 14-15th centuries, light pieces; C - historical brick from 15-17th centuries: Q-quarts, A-anothite, C-calcite, Dgelenite, H-hematite There was found calcite in the historical bricks. Normally calcite disappears at the temperature below 800 oC. (Figure 1) It means that calcite found in the bricks is secondary calcite. The secondary calcite got into the structure of bricks from lime mortars by influence of water migration

8


8. Pore size distribution There have been made many suggestions in the literature what pore size, as well as the quantity can influence durability of bricks. /4, 5/Bricks with low durability have pores between 0.04 and 1 ~tm, but with good durability - majority larger than 2 ~tm. The majority of pores 0.04-1om charaeterise deficient frost damage. Water in smaller pores requires temperatures below - 40oC to freeze. Pores >1 ~tm do not saturate with water. The pore size depends on the composition of raw materials and methods of bricks' formation: 9 grog and shale particles larger than 200 mesh favored the formation of larger pores; 9 pure clays produced pores smaller 0.1 ~tm 9 soft moulding produces larger pores to compare with the extrusion; 9 higher firing temperature and increasing soaking time reduce the quantity of pores but not the predominant pore size in a span of normal temperatures. The pore size distribution was determined by mercury porazimeter. The weight of samples used was approximately 0.5 g.

Figure 2: Pores size distribution of historical bricks: A - historical brick from 13-14m centuries; B - historical brick from 14-15th centuries; B-light - historical brick from 14-15th centuries, light pieces; C - historical brick from 15-17th centuries

9th International Congress on Deterioration and Conservation of Stone, Venice 19-24June 2000

9

The dimensions of pores are mainly in interval from l to 2 ~tm for all historical bricks. (Figure 2) The oldest bricks from 13-14th centuries have bigger pores than other bricks. It means, that these historical bricks have better resistance to frost and soluble salts. The newest bricks (15-17 th centuries) have different pore size distribution and shape. The dimensions of pores are around 1lam. It indicates that bricks have lower resistance to frost and soluble salts. Historical bricks B have light pieces of additive. The dimensions of pores of additive is concentrated in interval 0.8-0.9 pm. The pieces of additive are more porous with smaller pore dimensions. The durability of these pieces to soluble salts and frost is lower. The pieces of additive could be shamote or soft pieces of clay - with smaller proportion of water. In reality, all historical bricks are in rather good condition atter 5-7 centuries. The pores can be divided into small, middle and big ones. The possibility of ice formation and crystallisation of salts crystals is most obvious in middle size pores (0,0041pm) of bricks. Relatively bigger amount of middle size pores is observed in historical bricks B - 56%, because they have light peace of additive with big amount of middle size p o r e s - 84%. (Figure 3) Historical bricks A have only 25%, but C - 48% middle size pores.

Figure 3: Cummulative pore volume: A - historical brick from 13-14th centuries; B historical brick from 14-15 th centuries; B-light - historical brick from 14-15th centuries, light pieces; C - historical brick from 15-17th centuries.

The historical bricks A have 55% big size pores, C - 47%, but B - 7%. Due to that historical bricks A have better durability to soluble salts and frost.

10


Magge suggested an equation that allows one to calculate a Durability Factor for a brick based on its pore structure. /6/ A sufficiently large value of the Durability Factor is associated with a durable brick. The equation is: DF=(3,2/PV)+2,4xP3

(1)

DF - durability factor of brick; 3,2/PV- maximum intruded pore volume of brick, cm3/g; 2,4xP3 - percentage of the pore volume lying in pores with diameter greeted 3 lam. The first term of Maage's equation, 3,2/PV, expresses the obvious trend that a greater pore volume may lead to less durable brick. This portion of Magge's criterion is in parallel with the ASTM criterion that sets limits on water absorption for several climates. The second term, 2,4 x P3, recognises the fact that large pores, while contributing to the total volume, are unlikely to full of water. They drain easily. Thus a brick's durability rating can be increased be the proportion large pores that are present. Or, put another way, a brick should not be penalised for having a large pore volume if those pores are large and drain easily. The results indicate that bricks are frost resistant if DF>70 but not frost resistant if DF 1,5%); a yellow shade in Baveno. Moreover dark marie xenoliths are diffused only in Montorfano and rust spots coming from oxidation are diffused in Mortorfano and Alzo. The results were tested on different buildings from Milan and Turin where the use of different granite was witnessed by book references. Keywords: building material, stone, granite, quarry. 1. Introduction

Granites were employed as building stones in Northern Italy starting from 16th century. Different kinds of artifacts as ashlars, slabs, column shafts, pilasters, portals, balustrades, stairs, paving stones, etc. were made by granites. The use was facilitated by the good characteristics of the rock and by the transport facilities (i.e. the water way called Naviglio) from the quarry area to the main cities (mainly Milano and Pavia, Lombardy). The quarry area is located northwest from Milano near the lake Maggiore and Val d'Ossola (see map). The most employed is the pink granite from Baveno, but also the white granite is important as building stone; white granite comes from Baveno, Montorfano and Alzo. The three granites are almost equal and distinctive markers useful for macroscopic observation were not found in literature. The aim of the present study is to define distinctive markers among the varieties in order to facilitate the distinction; the markers must be easy to find at macroscopic observation without long and expensive analyses and they must be recognizable also on buildings with minimum sampling.

* Author to whom correspondence should be addressed.

42

9th International Congresson Deterioration and Conservationof Stone, Venice 19-24June 2000

2. Geology The so-called "Graniti dei Laghi" belong to a granitic batholith intruded in the "StronaCeneri zone" (metamorphic rocks, marie rocks and kinzigites) and composed by different plutons (fig. 1). Three different plutons are noticeable: Mottarone-Baveno, Montorfano, Alzo-Roccapietra; details on datation (about 275 m.y.) and on the mechanism of emplacement are reported in references (Boriani et al. 1992). The Mottarone-Baveno pluton (lake Orta - lake Maggiore) is composend mainly by medium grained, white, granite with quartz, K-feldspar, plagioclase and biotite as main components (GalliteUi 1937). A pink variety (pink is the K-feldspar) occours in the NorthWestern part of the pluton, the main components are the same as white variety. The Montorfano pluton (lower Val d'Ossola) is composed mainly by medium grained, white granite with plagioclase, quartz, K-feldpsar, biotite as main components. The northern part of the pluton is occupied by a green variety composed by albite, chlorite, quartz, sericite (Gallitelli 1938). The Alzo-Roccapietra (lower Val Sesia - lake Orta) pluton is a medium grained, white granite composed by quartz, K-feldspar, plagioclase, biotite (Ccallitelli 1941).

Figure 1 - Geological sketch map of"Graniti dei Laghi" area (Boriani 1974)

3. History Another aim of the study is to determine the exact date of the beginning of the quarry exploitation. The start of the exploitation of Baveno and Montorfano granites is not clearly


43

reported, but it is about the first years of 16th century and still continues; Alzo quarry was opened in 1847 (Peverelli 1922) and closed in 1956 (Gazzetta del Popolo 1956, june 21). Granites of Baveno and Montorfano are not known by Giorgio Vasari (Vasari 1568). P. Morigia (Morigia 1603) uses the local name "meiarolo" and Vineenzo Seamozzi (Seamozzi 1615) describes a granite with small black and red spots and white background, called "migliarolo", and a black and white granite without red spots. Authors in 18th and 19th centuries also witness the use of granite mainly in Lombardy (Vagliano 1710, Pini 1779, Amati 1829, Rondelet 1832). An accurate report is made by G. Casalis (Casalis 1839-40) where Alzo granite is not reported. Carlo Amoretti (Amoretti 1814) indicates the increase of the use of "migliarolo" when the cardinal Carlo Borromeo (born 1538) was archibishop of Milan (1565-1584). G. Barzan6 reports a catalogue of the Milan's monuments made by granite and this report is very important because of many buildings were aiterwards destroyed (Barzano 1853). Alzo white granite was first reported by Giovanni Jervis (Jervis 1889); Jervis also remarks the presence of dark xenoliths in Montorfano white granite. The diffusion of pink and white granites as building stones is reported by Authors at the end of 19th century (Salmojraghi 1892, Blangino 1895): Salmojraghi remarks the presence of iron sulphides in Montorfano white granite as a defect. In the XX century the features of granites are reported in handbooks and papers (Anon. 1939, Fagnani 1956, Pied 1964).

4. Sampling and methods of study Different quarries were sampled in order to detect chemical and physical characters of the rocks: Fedolo (Mottarone-Baveno pluton) 2 samples; Donna (Montorfano pluton) 1 sample; Cirla (Montorfano) 2 samples; Alzo (Alzo pluton) 2 samples. Data from the same quarry are very similar and also the samples from Donna and Cirla (Montorfano) are quite similar. Samples were taken from Milan and Turin buildings in order to compare the analytical data with quarry samples; sampling points are reported in chapter 8. Samples were studied by optical microscopy on thin section, X-ray diffraction on powder, X-ray fluores~nee, scanning electron microscopy, eolour measurements, mercury posimetry. 5. Characteristics of white granites 5.1 Colour Baveno granit is easy distinguished from Montorfano and Alzo granites: the first one is yellowish, the others are white. An exact measurement of the eolour according to the CIELab system was impossible in this ease because of the area investigated by the eolorimeter is wider than the feldspar crystals that bear the eolour in the granite. Furthermore this method is not suitable on the stone used in buildings because of the eolour changes caused by the stone decay. 5.2 Mineralogical composition White granites are medium grained plutonie rocks with high biotite content. Quartz, K-feldspar, plagioclase and biotite are the most important minerals determined by optical microscopy in the "Graniti dei Laghi". Quartz: xenomorphie grains (size 4-8 turn), undulose extinction, rare inclusions. K-feldspar: xenomorphie grains (size 4-9 turn), often perthitie, Carlsbad twinning.

44


Plagioclase (about 25% anorthite): automorphic grains (size 2-6 mm), regular zoning (more calcic in the core), polysynthetic twinning. Biotite: automorphie lamellae (size 2,5-3,5 mm), pleoehroism from light yellow to redbrown, zircone and apatite as inclusions, sometime transfomed in chlorite. Accessory minerals are: zircone, apatite, epidote, fluorite, titanite, etc. The composition shows small differences among the quarries, but it is impossible to find distinctive markers among these using the optical microscopy; because of a comparison must be based on a great number of thin sections and this is incompatible with the care of the buildings. The presence of iron sulphides mainly in the Montorfano granite and in a lesser extent in Alzo granite, together with the absence of these minerals in Baveno granite, leads to an easily recognizable difference. Rust spots are very diffused on Montorfano granite surface and they are dearly visible also in ancient buildings. This is one marker to distinguish the Montorfano granite: only the freshly cleaned surfaces don't show the spots. Dark, microgranular xenoliths are diffused in the Montorfano granite only. They have some marie composition (biotite, plagioclase) and look as black spots with size ranginng from some millimetres to some decimetres, more compact and resistant to the decay than the surrounding rock. Xenolith is a good marker to distinguish Montorfano granite from Baveno and Alzo granites. X-ray fluorescence detects a higher silica content and a lower iron content in Baveno granite than those of Montorfano and Alzo granites (tab. 1). The values are: Baveno - Silica >75%, Iron 7513) which are observable both macroseopically and on thin sections, but not measurable by means of mercury porosimetry; this discrepancy is more marked in the G.e. From the difference between P and Pt values the percentage of pores with a radius greater than 7513 Fig.1. Pore size distribution: C.c. vs. G.e has been deduced. Each


53

class of pores obtained by means of mercury porosimetry has been calculated again, related to the Pt value. In this way we can compare the two materials and each of them before and after ageing, considering also the contribution of the macro-pores. As will be seen in the Fig.2. Pore size distribution in G.c. before and after ageing next paragraph 3.6, this consideration is more valid for the G.c than for the C.c.. For this latter if we consider the water absorption on a long time, we can probably suppose the presence of pores under the minimum size measurable by mercury porosimetry. However this approximation is still acceptable in the light of the microscopic observations which reveal a prevailing presence of bigger than micro-pores. In fig. 1 we see the average porosimetric distributions relative to the two nonaged materials, while in fig.2 and 3 those of the same materials before and after ageing. They are expressed as percentages of the total volume of the samples. From the observation of fig. 1 we can see that for the G.c. more than 60% of the Pt is represented by pores with a radius greater than 75 IX and by 20-25% of pores with a radius between 0.01 and 0.5 IX, while lower percentages make up the other classes. The C.c. shows a very different distribution: the pores with a radius over 75 IX are about 30%, like those falling in the range between 10 and 40 IX, while the other classes of pores are all represented in small percentages. The diverse distribution existing between the two materials, with a clear predominance of large pores in the G.c., could be an important factor influencing Fig. 3. Pore size distribution in C.c. before and after ageing their behaviour regarding the deterioration caused by soluble salts. After ageing the porosimetric distribution in the G.c. roughly resembles that of the non-aged material both in the trend and in the percentages fractions. Before and after ageing the C.c. shows variations, though not marked ones: a reduction in the pores with a radius between 0.1 and 5 Ix and an increase in those between 5 and 40 Ix.

54


3.4 Ultrasonic wave velocity The average values of the ultrasonic wave velocities, before and aider ageing, are respectively of 3457 rn/s (st. dev. = 175) and 3319 m/s (st. dev. = 132) in the G.e., while in the C.c. they are 3649 m/s (st. dev. = 117) and 3369 m/s (st. dev. = 74). The velocity value for each sample has been obtained as average of three measures performed on the three orthogonal directions. More marked differences before and after ageing are in the C.e. than in G.e., reflecting the stronger damage induced on it by the ageing. The velocity values for the two non-aged materials are different from those already obtained in a previous study (Calia A. et al., 1999). That eould be attributed to the different dimensions of the samples.

3.5. Water absorption by capillarity The test was carried out on 10 samples measuring 5x5x2 cm for each lithotype (Doe. Normal 11/85, 1986). The absorption in very short periods was also calculated (30 seconds and 5 minutes) since these, although with a notable margin of error, can help us to calculate the first inflection point of the absorption curves. If we compare the curves shown in rigA, we can observe that both materials before ageing absorb large quantities of water. The G.e. absorbs about 20% more water than that of C.c., because of its greater porosity, for both materials the inflection point in the initial part of the absorption curve is reached between 30 seconds and 5 minutes. The asymptote value of absorption (Moo), evaluated in compliance with Doc. Normal 11/85, is reached after 6 days for both the materials and is equal to 619 mg/cm 2 for the G.c. and 475 mg,/cm2 for the C.c.. The G.c. absorbs 35% of water at 30 seconds .... increasing to 92% in the time (sec-1/2) first 5 minutes. The standard deviation is of Fig.4 Water absorption by capillarity 30% on the first measurements but it declines to 3% in the second measurements and remains roughly unvaried until the end of the triM. The average coefficient of capillary absorption (C.A.) is equal to 41 mg/cm 2 s '~ The C.c. already absorbs 82% of water at 30 seconds increasing to about 90% in the first 5 minutes. The standard deviation is of 13% on the first measurement; it declines to 3% on the second measurement and remains more or less unvaried until the end of the triM. The average coefficient of capillary absorption (C.A.) is equal to 78 mg/em2 s "~. After ageing the samples of both materials absorb a slightly smaller amount of water: Moo for G.c. is 581 mg/cm2 while that of C.e. is 430 mg/em2. This reduction seems mainly correlatable with the weight losses occurring in the samples due to weathering and therefore with the consequent loss of absorbing volume. The maximum quantities of water absorbed,


55

related to the weight of samples, are equal to 21 and 20,3% for the aged and non-aged G.c., 14,4% and 14,5% for the C.c.. To carry out a more thorough investigation into the behaviour around water, on the nonaged samples the test was held over a period of time which was notably longer than that required by the Normal document. The asymptote value was considered to have been achieved when the difference between two successive value of the quantity of water absorbed was 0,06%. This occurs in 36 days with a quantity of water absorbed equal to 510 mg/cm2 for the C.c., in 33 days with 636 mg/cm2 for the G.c..

3.6. Water absorption by total immersion The trial was carried out on 10 samples of 5x5x5 cm for each lithotype (Doe. Normal 7/81, 1981). The fig. 5 illustrates the absorption curves before and aider ageing. The maximum value of water absorbed, as advised by the above-mentioned Doe. Normal, is achieved on the 6th day by G.c. and on the 7th by C.c., with an imbibing capacity (I.C.) equal, respectively, to 20,9 and 13,2 %. The st. dev. are rather low, around 3% for all the measurements. The near-totality of water absorbed is reached in both materials in very early stages of the trial. The saturation index S.I. = time (see''~) I.Cv/Pt (I.Cv = I.C. x Yb) is equal to 75% for the Fie. 5. Water absorotion bv total immersion G.c. and 62% for the C.c.. The lower S.I. value in the C.c. is explained by its porosimetric characteristics, and, in particular, by the greater presence of small sized pores. The same samples, alter artificial ageing, show water absorption values coinciding in the C.c. as slightly superior in the G.c.. The measurements on the non-aged samples were prolonged well over the times recommended in the Doe. Normal in this ease too, for the same reasons as in the previous paragraph. Water absorption continues in the C.c. up to 234 days of immersion with a corresponding S.I. of 83% and I.C. 18%. For the G.c. the water absorption continues for up to 35 days with S.I values of 74% and I.C. of 22%. This prolonged absorption leads us to suppose the presence of very small pores which slowly saturate. Therefore the difference between total and integral open porosity is not attributable only to the class over 751a, but probably also to that under 37A.

3.7. Evaporation of water and Drying Index The trial was carried out in compliance with Doe. Normal 29/88, 1991 on 10 samples for each lithotype. In fig.6 we can see the average drying curve relative to the two materials. If we consider this curve and the values of the Drying Index calculated after 48 hours with

56


Simpson's integral and equal to 0.374 for the G.c. and 0.417 for the C.c., we can see how the G.c., while having a greater water content, dries more quickly than C.c.. In fact, in the G.c. aider 48 hours the water residue inside the samples is equal to 1% of the initial quantity, time (hours) while the same percentage is found in Fig.6 Water evaporation the C.c. a~er 4 days. The standard deviations in both materials are variable over the trial, starting from values around 2-3% in the first measurements, rising gradually to 10% in the measurements at 4 days. 3.8. Water vapour permeability This was measured by following the procedures indicated in the Normal 21/85, 1986. The values obtained are 378 g/m2 (st. dev. = 34), 322 g/m2 (st. dev. = 21) for the G.c. and 308 g/m2 (o=21), 277 g/m 2 (st. dev. =19)for the C.c., respectively before and after ageing. For both materials, therefore, after ageing a modest reduction can be seen. 4. Conclusions On artificial ageing the C.c. deteriorates more radically than the G.c.. These two diverse reactions resemble what is observable when the materials are employed. The change can be seen in considerable loss of material in the C.c. and in an increase of superficial roughness; the G.c. loses more a contained amount of weight and great variation on the surface layer are not observable. None of the other properties measured have shown clear distinctions between the materials before and after ageing. By comparing the two materials before the ageing, we observe that they have different values for almost all the measured parameters. However, only some of them make the difference in the behaviour during the ageing. The total porosity, which is higher in the G.c., though it does mean greater absorption of water (trough capillarity and immersion), is not reflected in a weaker resistance to ageing. So the mere behaviour in water can not be an index of the durability of the material. The pore size distribution and some textural characteristics are more material to the response of the deterioration. In fact, the G.c. has about 60% of its total porosity composed of pores with a radius greater than 751~, while the same for the C.c. is distributed in the classes inferior to 751~, with a concentration of the pores between 10 and 401~. Moreover the prolonged water absorption of the C.c., probably due to pores that very slowly saturate, let us suppose the presence inside it of very small pores, under the minimum size measurable by mercury porosimetry. After the ageing G.c. shows no changes at all in pore size distribution, while slight variations are observable in the C.c. (see paragr. 3.3). The cement,


57

although it is in poor quantity in both materials, probably influences their resistance: its microsparitic texture in the C.c. could make this material more v-alnerable to the stress induced by the artificial ageing. On the contrary, the presence of expandable lattice clay minerals, among the secondary components, does not seem to have any influence in virtue of the low percentage involved. The absence of marked variation in the properties measured before and after ageing in contrast to the considerable loss of material, suggest that the deterioration does not involve the entire volume of the sample, but only superficial strata. In fact, the variations observed are only those relative to the trial which react to the influence of superficial conditions, also because of the greater surface/volume ratio of samples such as the capillarity and permeability. Its intensity however, is different and, considering the previous observations, probably connected to the mierostructural characteristics of the two materials. Pore size distribution and textural characteristics of cement are not quantifiable to the deterioration measurement, because of its superficial location. On the contrary the most significant parameter for the evaluation of durability would seem to be the weight variation due to the loss material. References Cal6 G., Di Pierro M., Federico A., Mongelli G, 1985. Caratteri geologici, petrografici, mineralogiei e meceanici dei "earpari" della provincia di Lecee, Quarry and Construction. Calia A., Mecehi A., Luprano V.A.M., Rubino G., Rota P.,1999. 6th International Conference on 'r Destructive Testing and Mieroanalysis for the Diagnostics and Conservation Of Cultural and Environmental Heritage-ART 99, Roma. 147-162. Calia A., Mecehi A.M., Quarta G., Rota Rossi-Doria P., 1999. Le pietre naturali da costruzione in Puglia. I1 "earparo": impiego e conservazione. Science and Technology for the Safeguard of Cultural Heritage in the Mediterranean Basin, Proceed.1 th Int. Congr. Catania, 1995.885-891. Mecchi A., Calia A., Quarta G., 1998. Caratterizzazione di un materiale da costmzione della Puglia: il Carparo. Recuperate l'Edilizia, I, 3. Alberto Greco Editore, Milano. 33-37. Doe. Normal 4/80, 1980. Distribuzione del volume dei pori in funzione del loro diametro, C.N.R.-I.C.R., Roma. Doe. Normal 7/81, 1981. Assorbimento d'acqua per immersione totale - Capacit/t di imbibizione, C.N.R.-I.C.R., Roma. Doe. Normal 11t85, 1986. Assorbimento d'aequa per capillarit/l- Coefliciente di assorbimento capillare, C.N.R.-I.C.R., Roma. Doe. Normal 22/86, 1987. Misura della velocit/t di propagazione del suono, C.N.R.I.C.R., Roma. Doe. Normal 29/88, 1991. Misura dell'indice di aseiugarnento (dryng index), C.N.R.I.C.R., Roma. Doe. Normal 32/89, 1991. Determinazione gas volumetrica della CO2, C.N.R.-I.C.R., Roma. Doe. Normal 34/91, 1994. Analisi di materiali "argillosi" mediante XRD, C.N.R.-I.C.R., Roma. Doe. Normal 21/85, 1986. Permeabilith al vapor d'acqua, C.N.R.-I.C.R., Roma. Dunham R.J., 1962. Classification of carbonate rocks according to depositional texture. Classification of carbonate rocks. Mem. American Association Petrology and Geology, 1, W.E. Ham Ed. 108-121.


59

PRELIMINARY CONTRIBUTION ON DURABILITY OF SOME MACROPOROUS MONUMENTAL STONES USED IN HISTORICAL TOWNS OF CAMPANIA REGION, SOUTHERN ITALY

Alessio Langella* Faculty of Science, Sannio University, Benevento, Italy Domenico Calcaterra Dept. of Geotechnical Engineering, Federico II University of Napoli, Italy Piergiulio Cappelletti, Abner Colella, Maurizio de' Gennaro, Roberto de Gennaro Dept. of Earth Sciences, Federico II University of Napoli, Italy

Abstract

A preliminary study of the decay phenomena of three maeroporous volcanic rocks (Neapolitan Yellow Tuff, Campanian Ignimbrite, Piperno) from Campania, Italy is here presented, carried out by means of ageing tests. The latter were chosen taking into account the basic environmental and climatological characteristics of the Campania region. According to standard procedures the wet-dry and salt crystallization tests were carried out. In order to assess the evolution of physico-mechanical and mineralogical features the following parameters were measured at regular intervals: open porosity; water absorption by total immersion; ultrasonic velocities, uniaxial compressive strength. SEM observations allowed to follow changes of intergranular relationships. Aiming at a better comprehension of durability, IRD (Index of Rock Durability) was determined, based on rock swelling tests. The overall results so far obtained pointed out a poorer durability of Neapolitan Yellow Tuff, if compared to the other stones, especially as wet-dry tests are regarded. Key words: macroporous rocks, Piperno, Neapolitan Yellow Tuff, Campanian Ignimbrite, ageing tests, durability, Italy. 1. Introduction

In many Italian towns the use of local stones for architectural purposes has been a very wide practice throughout all the historical ages of the country. In Campania region (Southern Italy) the large availability of volcanic products, characterized by ease of workability, good physico-mechanical properties and an agreeable aspect as well, determined the utilization of these materials both for structural and for ornamental purposes. Among the most widespread volcaniclastic products widely used in the historical architecture of the main towns of Campania region, an outstanding role was played by the Neapolitan Yellow Tuff (NYT) and the Piperno (PI) in Napoli, and by the Campanian Ignimbrite (CI) in its different facies used over the whole Campanian territory (Calcaterra et al., 1999a). Literature on these volcanic products is, in some instances, almost updated. In fact, as far as NYT and CI are concerned, a wide volcanologieal, mineralogical, and petrographical bibliography is available. The behavior of these materials, when used as dimension stones ('Tacciavista"), has been widely investigated, as well. In contrast, this kind of information is lacking for PI even though some researches in progress are trying to fill this gap. Some hypotheses have also been drawn on the decay phenomena in different


60


microenvironmental contexts, whereas very little has been carried out in order to experimentally verify the actual causes of the stone weathering (Rossi Manaresi, 1976; de' Gennaro et al., 1993, 1995). The present study aims at improving the knowledge of the mineralogical and petrographical features of these materials in order to discriminate similar lithotypes and to understand the decay phenomena by means of laboratory simulations which reproduce the ageing processes of the stone. 2. Materials

The investigated materials represent the products of different eruptive episodes of Campi Flegrei volcanic area (Napoli) (Di Girolamo and Morra, 1987; Cole and Scarpati 1993). Their widespread diffusion favored an intensive exploitation (Calcaterra et al., 1999a) mainly concentrated in Napoli and Caserta provinces (fig. 1). Hereat~er, a short description of these volcanic rocks is reported.

Figure 1: Sketch map of Napoli and Caserta provinces, along with quarry locations. Legend: cross = Piperno quarry; full circle = NYT active quarry; full lozenge = NYT inactive quarry; open circle = CI active quarry; open lozenge = CI inactive quarry. Neapolitan Yellow Tuff (NYT): trachytic volcaniclastic rock made up of pumice, obsidian fragments, crystals and lithies set in an abundant zeolite-bearing ashy matrix. Age 12,000 years b.p. Widely used as dimension stone since the former Greek settlings, it has been employed with structural and architectural function, so far. It represents about the 20% of the "facciavista" walls of the ancient centre of Napoli and almost all the plastered ones. The macroporous texture along with the presence of zeolites make this rock particularly prone to the action of weathering agents. Investigation of these processes were carried out on specimens collected in a quarry, located in Quarto (Napoli).


61

Campanian Ignimbrite (CI): trachytic volcaniclastic rock made up of pumice and scoria~ in a cineritic matrix. Two distinct facies, both lithified, are present: a gray facies, with epigenetic feldspars, and a yellow facies, characterized by the presence of zeolites. Age: 37,000 years b.p. Widespread over the entire Campanian territory, both facies have been utilized for the production of dimension stones. The gray one, displaying better physicomechanical features, has been used for particular architectural parts. The Medieval town of Casertavecchia represents the highest expression of the massive utilization of this rock. The material used for the present research comes from one of the quarries (Pozzovetere) nearest to the urban settlement. Piperno (PI): traehytic volcanic rock characterized by eutaxitic texture with black flattened fiamme, with sanidine phenocrysts set in a light-gray ashy matrix. Age: ~ 37,000 years b.p. It is the most used ornamental stone in Napoli, either as facing slabs or particular architectural dements, and covers about 50% of the exposed surfaces in its ancient centre. It is sometimes also found in other provinces of the region. Physico-mechanical parameters are sensibly better than those of NYT even though they scatter in a quite wide range. Notwithstanding these good features, the stone is affected by diffuse weathering phenomena. The material used for tests was sampled in one of the rare outcrops still accessible located in Pianura, western periphery ofNapoli (Calcaterra et al., 1999b). 3. Methods

Mineralogical characterization was carried out both by SEM observations and by means of X-ray powder diffraction analysis (XRPD - Philips PW1730/3710) using a CuKF1 radiation, incident- and diffracted-beam SoBer slits, curved graphite crystal monochromator, 2[] range from 3 to 100 ~ step size 0.02 ~ 2B and 10s counting time per step. Quantitative mineralogical analyses were also performed by XRPD using an internal standard, D-A1203 (1 lxm, Buehler Mieropolish) added to each sample in amount of 20 wt %. Powder data set were analysed both by RIR (Chipera and Bish, 1995) and Rietveld methods (Bish and Post, 1993), the latter using GSAS package (Larson and von Dreele, 1995). Open porosity was calculated by means of apparent and real volumes with a Hepycnometer (Micromeritics Multivolume Pycnometer 1305) on at least 10 specimens for each material. Water absorption by total immersion: the total absorbed water after immersion in deionized water at room temperature and pressure was evaluated according to NORMAL 7/81. Ultrasonic tests were carried out according to Italian suggested standards (NORMAL, 22/86), taking into account international recommendations (I.S.R.M., 1978), as well. PUNDIT (CNS Instruments Ltd.) ultrasonic non-destructive digital tester was used with a pair of 24 kHz transducers, in direct arrangement (i.e. transmitter and receiver positioned on opposite sides). Uniaxial Compressive Strength (UCS): tests were made on the specimens previously used for the ultrasonic measures. I.S.R.M. (1979) suggestions have been followed, even though the tests were conducted on cubic samples. Before each test, rock density was determined. Ageing tests: wet-dry and salt crystallization tests were performed according to standard procedures (Rossi Manaresi, 1976; RILEM, 1980; Topal and Doyuran, 1998). The tests were chosen taking into account the basic environmental and climatic characteristics of the Campania region. The total number of cycles for each test was determined on the basis of

62


macroscopical changes of the samples, i.e. presence of cracks on the specimens, disgregation, etc. In order to assess the evolution of physieo-mechanical and mineralogical features the following parameters were measured at regular intervals: weight, open porosity, water absorption by total immersion, ultrasonic velocity and UCS. Careful SEM observations allowed to follow changes of intergranular relationships. Swelling strain: determined on cubic samples following the procedures suggested by Nascimento et al. (1968). Depending on the tested material, the swelling strain was measured on variable time spans (usually 24 or 48 hrs.), with a precision of l laxn. 4. Results

Table 1 reports the quantitative mineralogical composition of representative samples of the studied rocks. The NYT shows its typical mineral assemblage with prevailing epigenetic phases (phillipsite, chabazite and analcime), feldspar, and minor amount of mica, hydrated iron oxides and volcanic glass. As regards the other two rocks, they have a very similar mineralogical composition characterized by predominant feldspar, both pyrogenic and epigenetic, and in very subordinate amounts unaltered volcanic glass. The presence of sodalite is characteristic of the PI, and may represent an useful marker which can enable to discriminate PI s.s. from a very similar facies of Campanian Ignimbrite, as far as these rocks are used as dimension stone (fig. 2). Table 1: Quantitative mineralogical evaluation ofvoleanielastic units. Sampte

feta

bio

sod

magn

pyrox

amph

phi

cha

aria

cm

am*

NYT 24.0 tr. tr. tr. 42.5 6.2 7.0 3.4 16.5 PI 85.5 0.4 3.5 '0.3 ' tr. tr. 10.3" CI 88.9 1.0 0.3 tr. 9.8 Note: feld= K- and Na-feldspars, bio= biotite; sod= socialite; magn= magnetite; amph= amphibole; phi= phillipsite; cha= chabazite; ana= analcime; cm= clay minerals and am= glass (including hydratediron oxides); *= calculated by difference , ,

Figure 2 XRPD spectra of Pipemo and Campanian Ignimbrite samples. The variation of the physico-meehanical properties are reported in figure 3 and in table 2. As far as weight is regarded, a slight loss (0.7-2.8%) was measured in the wet-dry tests for all the materials, while the salt-crystallization tests caused a greater reduction (4.4-11.8%).


63

On the other hand, open porosity is the parameter less influenced by the ageing tests, with increases ranging between 0.3 and 4.4%. The NYT seems more sensitive to water absorption (increase by 13.7% in wet-dry and 6.6% in salt crystallization tests) if compared to CI and PI (0.7-3.0% and 2.1-4.7%, respectively). The P-wave velocities as a rule, decrease for all the materials, regardless the kind of test; however NYT shows again the worst behavior (2.3% - wet-dry test in CI; 39.8% - wet-dry test in NYT). Considering the uniaxial compressive tests, NYT turned to be affected by a significant reduction in strength, showing severe decrease of this parameter both for wet-dry (43.6%) and salt crystallization (76.6%) tests: the latter result was achieved after only six cycles. On the contrary, UCS values for Piperno did not seem to be influenced by the tests carried out, thus testifying a substantial unaltered mechanical behavior. CI holds an intermediate position as far as wetdry tests are considered, whereas is more sensitive to salt crystallization cycles (-31.3%). A different approach to the evaluation of durability considers the swelling tests performed on unaltered samples which, along with UCS and porosity, enables to calculate the IRD (Index of Rock Durability- Delgado Rodrigues and Telmo Jeremias, 1990). The tests gave IRD mean values of 0.04 for the NYT samples, 0.066 for IC and 0.439 for Piperno.

Table 2: Variation of the physico-mechanical properties after ageing tests (mean values). Neapolitan Yellow Tuff Bwd F,,d B,, F,, Porosity,(%) 53.9 55.2 54.8 57.2 1-120absorption,(wt %) 43.0 48.9 49.1 52.4 Vp,(m/s) 1789 1076 1891 1380 UCS, (MPa) 4.64 2.62 4.07 0.95

Campanian Ignimbrite B~

55.6 39.3 1812 4.18

Fwd B,,

56.6 39.6 1770 3.50

55.0 39.3 1726 4.11

F,,

56.9 40.5 1652 2.82

B., 45.1 21.6 2597 6.43

Piperno V., B.~ 45.2 44.8 22.1 25.1 2394 2274 6.60 6.32

v~ 45.1 26.2 2008 5.25

Bwa= Blank values for wet-dry tests; Fwd= Final values after wet-dry tests B~= Blank values for salt crystallization tests; F~= Final values after salt crystallization tests 5. Discussion and conclusions

The obtained results give a further contribution to the interpretation of the behavior of these three important Campanian rocks towards the decay agents. From the mineralogical point of view, a substantial difference was found out between the NYT on one side, and CI and PI on the other. The matrix of the former is mainly constituted by zeolites (phillipsite and subordinate chabazite) which also act as cement of the rock and, subordinately, by amorphous phases and volcanic glass; on the contrary, the latter (CI and PI) have a matrix almost totally constituted by feldspars and minor amount of glass. The response of these materials to ageing tests (figs. 4 and 5) is that the NYT is the most affected, followed by CI, whereas PI undergoes only slight variations of its physical and mechanical features (tab. 2). Notwithstanding any particular alteration of the crystalline phases of the matrix (fig. 4d), as far as salt crystallization test is regarded, NYT exhibits weathering evidences since after few cycles which determine a weight loss and a sensible reduction of the original volume (fig. 5). Wet-dry tests, conversely, preserve the original shape of the specimens but in the final cycles (30) a net of cracks appears on their surface. This phenomenon is also recorded on SEM observation (figs. 4b and c).

iii

Wet-d~

Salt crystallization

Wet-dry

Piperno

Campanian Ignimbrite

Neapolitan Yellow Tuff


i


Wet-d~

105

~oo ~9

~o

=-,

80

110 , 105 "~

0

P_ 95

0 O'Q

,

90

1 2 0 ~ ~

~

115

r~ 0

110 105

s

ca 90 110 lOO

m .~ ~,

~

0 ~t 0

80

7o

~o

5O 0

120 ..-.,

100

~ 8o ~ 0 ~

--,._.______

6o

0

40

20 0

0

0

10 20 Number of cycles

30 0

,

Number of cycles

e

o

,o

20

Number of cycles

3o o

;

4

Number of cycles

6 o

20

4o

Number of cycles

~o

o

~

4

6

Number of cycles

0 (1)

Figure 3: Variation of physical and mechanical properties for the considered materials after ageing tests.

< t~ i

4~


65

Figure 4. SEM micrographs of NYT. a) untreated specimen; b) and c) after 30 wet-dry cycles; d) aider 6 salt crystallization cycles.

Figure 5: macroscopical features evolving during the wet-dry (let~ column) and saltcrystallization tests (fight column). First row =NYT, second r o w - CI, third row = PI

66


However, both tests bring about an overall abatement of the measured properties. As reported in Colantuono et al. (1991) zeolite-rich tufts display an initial shrinkage induced by heating, much higher than any other materials. Consecutive wet and dry environments and the consequent contractions and swellings seem to be responsible for loss of cohesion and disgregation of the stone. CI macroscopically showing modificatiom similar to those of NYT in the salt crystallization tests, denotes only slight decrease of the checked parameters. On the contrary, PI does not reveal appreciable variations either of its macroscopieal character or of its physieo-mechanical features. A further confirmation of the different behavior is given by the swelling tests and by the IRD data. The lowest value of I R parameter pertains to NYT (0.04) which, along with CI (0.06), is characterized by values even lower than those found out in other macroporous rocks, such as the Cappadocian tufts, which are in the 0.10 + 0.13 range (Topoi and Doyuran, 1998). The same parameter turned to be quite variable for PI giving values between 0.123 and 0.439. This wide range can be explained by the presence, within the same formation, of very different facies already evidenced in previous papers (Caleaterra et al., 1999b). It likely seems that the most discriminating parameter is represented by the compressive strength, much higher in PI if compared with NYT and CI. Based on the classification proposed by Delgado Rodrigues and Jeremias (1990), all the studied materials are considered as low durability rocks since their IRD is lower than 2. The reported data enables to deepen the knowledge on the studied materials, providing distinctive features between Piperno s . s . and piperno belonging to CI, otherwise hardly macroseopieally distinguishable, once both used in the same monument. Furthermore, they can help to better understand some weathering phenomena of the studied materials. However, further investigations are still required in order to explain, for example, the marked response of NYT to swelling test. A first hypothesis accounts for the textural relationship of the different mineralogical composition of the studied rocks. The above considerations suggest, as a consequence, that only an interdisciplinary approach which consider all the fundamental aspects of the stone will allow to get the results necessary for a correct evaluation of the restoration of these materials when used as dimension stone.

Acknowledgements Work carried out with the financial support of Italian National Council of Research (C.N.R.) Progetto Finalizzato "Beni Culturali", contr, n. 97.00630.PF36 granted to Prof. Maurizio de' Gennaro and MURST - Progetto di Rieerea di Interesse Nazionale, Cofinanziamento 99. Thanks are due to Mr. Antonio CanzaneUa (Federico II University, Napoli) for his help in SEM observations. 6. References Bish D.L., Post J.E., 1993. Quantitative mineralogical analysis using the Rietveld fullpattern fitting method, Am. Mineral., 78, 932-940. Calcaterra D., Cappelletti P., Carta L., de' C~nnaro M., Langella A., Morra V., 1999a. Use of local building stones in the architecture of historical towns: some ease histories from southern Italy. Proe. 2nd Inter. Congr. on "Science and technology for the safeguard of cultural heritage in the Mediterranean basin", Pads, 5-9 July 1999. In press.

9th International Congresson Deterioration and Conservationof Stone,Venice 19-24June 2000

67

Calcaterra D., Cappelletti P., Langella A., Morra V., de Gennaro R., ColeUa A., 1999b. The building stones of the ancient centre of Naples (Italy): the Piperno from Phlegrean Fields. Contributions to the knowledge of a long-time used stone. Journal of Cultural Heritage. In press. Chipera S.J., Bish D.L., 1995. Multireflection RIR and intensity normalizations for quantitative analyses: applications to feldspar and zeolites, Powder Diffraction, 10 (1), 47-55. Colantuono A., Dal Vecchio S., Marino O., Maseolo G., 1991. On the mechanism of water movement inside zeolitized tuff stones. Atti 1~ Convegno Nazionale di Scienza e Tecnologia delle Zeoliti, De Frede, Napoli, 115-121. Cole P.D., Searpati C., 1993. A facies interpretation of the eruption and emplacement mechanisms of the upper part of the Neapolitan Yellow Tuff, Campi Flegrei, Southern Italy. Bull. Volcanol. 55, 311-326. de' Gennaro M., Colella C., Fusealdo M., 1993. Weathering typologies of monumental tuff-stone masonries in the Naples downtown area. Science and Technology for Cultural Heritage, 2, 53-62. de' Gennaro M., Colella C., Langella A., Cappelletti P., 1995. Alteration and decay of campanian ignimbrite dimension stones in some monuments sited in Caserta area. Science and Technology for Cultural Heritage, 75-86. Delgado Rodrigues J., Telmo Jeremias F., 1990. Assessment of rock durability through index properties. Proc. 6th Int. IAEG Congress, Balkema, Rotterdam, 3055-3060. Di Girolamo P., Morra V., 1987. "The Campanian Ignimbrite", Petrographical, petrochemical and volcanological characters. In: Di Girolamo P., (ed.) The voleaniclastic rocks of Campania (Southern Italy): Rend. Ace. So. Fis. Mat., Special Issue, 177-199. International Society for Rock Mechanics, 1978. Suggested methods for determining sound velocity. Int. J. Rock Mech. Min. Sci. Geomech. Abstr., 15, 53-58. International Society for Rock Mechanics, 1979. Suggested methods for determining the uniaxial compressive strength and deformability of rock materials. Int. J. Rock Mech. ~ . Sci. & Geomech. Abstr., 16, 135-140. Larson A.C., von Dreele R.B., 1995. GSAS. General Structure Analysis System. Report LAUR 86-748, Los Alamos National Laboratory, NM, USA. Nascimento U., Oliveira R., Gra~a R., 1968. Rock swelling test. Proc. Int. Symp. on Determination of the properties of rock masses in foundations and observation of their behaviour, Editorial Blume, Madrid-Barcelona, 363-365. NORMAL 7/81, 1981. Assorbimento d'acqua per immersione totale e capacit~ d'imbibizione. Ed. CNR-IC1L Rome. NORMAL 22/86, 1986. Misura della velocit/t di propagazione del suono. Ed. CNR-ICR, Rome. RILEM, 1980. Recommended tests to measure the deterioration of stone and to assess the effectiveness of treatment methods. Materiaux et Constructions, 13, 175-253. Rossi Manaresi R., 1976. Causes of decay and conservation treatments of the tuff of Castel dell'Ovo in Naples. Proc. 2na Int. Symp. on the Deterioration of Building Stones. Athens, September 27- October 1, 1976, 233-248. Topal T., Doyuran V., 1998. Analyses of deterioration of the Cappadocian tuff, Turkey. Environmental Geology, 34, 5-20.


69

DURABILITY OF TUFFEAU STONE IN BUILDINGS : INFLUENCE OF MINERALOGICAL COMPOSITION AND MICROSTRUCTURAL PROPERTIES David Dessandier* Bureau de Recherches G6ologiques et Mini6res, Od6ans, France Philippe Bromblet, Jean-Didier Mertz Laboratoire de Recherche des Monuments Historiques, Champs-sur-Marne, France

Abstract

Five types of tuffeau of poor to good durability were selected for a comparative study. After a mineralogical and petrophysical characterization in the laboratory, the samples were subjected to an accelerated ageing test by salt crystalli~tion and hydration, from which an experimental comparative durability index was determined. At the same time, a general investigation of the physicochemical mechanisms of weathering led to the determination of a theoretical durability index calculated from some of the petrophysical properties of each sample. The influence of mineralogical composition and of certain mierostructural properties on the overall durability of the tuffeau was also studied. Key words: tuffeau, durability, mineralogy, mierostructure. 1. Introduction

The principal dimension stone used in the architectural heritage of the Loire Valley, is a chalky limestone facies and more precisely, a variety of Upper Cretaceous (Middle Turonian) chalk of the Paris Basin. It is tuffeau that, due to its intrinsic properties, is liable to severe weathering, entailing the continual restoration of the monuments. A comparison of the state of preservation of tuffeau historical monuments, however, reveals extremely variable situations, attesting to the diversity of the types of stone gathered together under the general name "tuffeau'" and the broad range of corresponding durabilities. Five types of tuffeau of poor to good durability were thus selected for a comparative study. 2. Materiel and methods 2.1 Samples analysed

The term "tuffeau" includes a wide variety of microfacies, reflecting the space-time variability of the sedimentary depositional conditions: the expression "tuffeau series" is employed. This variety of stone types used in monuments is reflected by a wide variety of states of preservation and hence of durability of the material. Within the framework of this study, five stone types belonging of the "tuffeau series" were selected (Table 1) from an existing database [1] according to the "durability" criterion estimated from in situ observations of the state of preservation.


70

9th International Congresson Deterioration and Conservationof Stone, Venice 19-24June 2000 Sample CHAM

~ Sampling location

Durability observed in situ Poor

' Chateau de Chambord ' Restoration 19th contrary .... FoNT Abbey of Fontevraud Good Built in the 12th century LOUD Saint Hilaire Church in Loudun Average Built in the 15th century LUZE Abbey of Bois-Aubry in Luz6 .... Good' Built in the 12th century VILL Villentrois Quarry Not observable Table 1 Sampling location and in situ durability of the five tuffeau facies studied.

2.2 Test procedures The mineral components of each tuffeau were identified by X-ray diffraction on a Siemens D5000 diffractometer. The major minerals (calcite, opal-CT and quartz) were quantified: i) by X-ray diffraction and optical microscopy in polarized light for quartz, ii) by infrared spectrometry, using the calibration curves of calcite- quartz - opal-CT mixtures [2] on a Perkin Elmer 16PC FT-IR spectrometer for opal-CT; and iii) according to standard NF ISO 10693 for calcite. The total content of accessory minerals (clays [smectite, muscovite, glauconite] and clinoptilolite) was calculated from the difference with the major mineral contents obtained by infrared spectrometry. To evaluate the influence of clays in the behaviour of the ~mples in the presence of water, tests were also carried out with methylene blue according ' " to standard NF P 94-068. Each tuffeau was subjected to basic petrophysical characterization, including the determination of." i) the total porosity Nt, according to standard NF B 10-503; ii) the water absorption coefficient S4s, according to standard NF B 10-504; iii) the capillary coefficient C, according to standard NF B 10-502; iv) the compressive strength Re, according to standard NF B 10-509; and v) the specific surface area SSEXby the BET method, according to standard NF X 11-1021. Certain hydrodynamic characteristics, representative of the fluid transfer properties, were also determined for the five tuffeau facies: i) The kinetics of water capillary suction of each sample was measured by a procedure derived from standard NF B 10-502. This test helped to determine two parameters denoted A and B [4, 5], according to the Washburn flow law (1921) [3], reflecting the "water weight gain rate" and the "linear migration rate of the capillary fringe" as a function of the square root of time; ii) Linear deformation during imbibition by capillary suction (unidirectional water expansion) of each sample was determined using a contactless optical feeler by laser triangulation. The use of an automated 1-micron resolution prototype [6] served to quantify the maximum deformation and allowed tracking of the deformation rate as a function of the water saturation of the free pore network of the stones; iii) Each sample was evaporated by drying following the Rilem No. 11-5 experimental procedure for regulated evaporation conditions (T=20~ and H.R.=33%) maintained by a salt solution saturated with MgCI2. In these conditions, the shape of the curves reveals two parameters that are representative of drying: the evaporation flux is constant and can be used to calculate a drying rate denoted ct. This evaporation regime only affects a fraction of the water present in the porous medium: beyond a certain degree of residual water saturation, corresponding to the critical water content Se of the stone, the


71

regime changes and drying mainly takes place by diffusion according to Fick's law [7]. Evaporation tests were performed for an initial total water saturation obtained by imbibition under vacuum, and for a partial saturation according to the procedure of standard NFB 10504. 3. Results 3.1 Mineralogical composition and parameters associated with clay minerals Mineralogical composition As a first approximation, the tuffeau consists of an assembly of microcrystals of calcite and opal-CT giving the rock a micritic texture. Other phases are disseminated in this microcrystalline matrix: elastic phases, chiefly grains of quartz of various size and accessory minerals including clays (illites, smeetites, glauconites) and zeolites (clinoptilolite). In detail, most of the illites are present in grains of glauconite about 50 lam in diameter. Similarly, part of the calcite is present in bioclasts, but in amounts that can be considered negligible. Quantitatively, the mineralogical compositions differ in the five facies analysed (Table 2). As to the cementation of the rock, it varies between a ealeitic pole (FONT) and a siliceous pole of opal-CT (LUZE). The elastic quartz accounts for 2 to 6% of the total composition, and clinoptilolite was identified in three samples with contents not exceeding 3%. The difference from 100 can be attributed to other minerals: clays and micas, the latter present in small amounts. The measured quartz contents and the calculated contents of clays, micas and other accessory minerals, are clearly correlated, justifying ex post facto the classing of the clays, micas and other accessory minerals in the elastic fraction of the rock, at least as a first approximation.

Cementation ' Calcite (%) OP'aI-CT

z Z Clastic minerals '(%)' 'Quartz (%) Clays, nficas ' Clmoptilolite and other (%),* (%), ' CHAM 57'.5 25 83.5 5 12.5 ,. 'FONT ' 67' 27 94 2 . . . . . 4. . traces LOUD 41 31 72 6 15to20 lto5 LUZE 53.5 39 92.5 2 5.5 40%), the expansion of the VILL and CHAM tuffeau samples was slightly greater than that of the LUZE and FONT samples. This deformation, produced by the effect of the pressures applied by the water meniscus at the grain boundaries, is fairly well correlated with the weaker mechanical properties and the extent of the suite of accessory minerals (table) in these tuffeau facies. Conversely, the tuffeau facies displaying the lowest capillary properties (FONT and LUZE) display less contrasted and less intense expansion. These different water related behaviours emphasise structural differences in the pore networks, which were not identifiable from the analysis of the macroscopic imbibition tests alone.

Figure 2: Water expansion as a function of water content in the five tuffeau samples.

Characteristics of water transfer by evaporation The shape of the evaporation curves (Table 7, Fig. 3) conforms to the conventionally described kinetics [7, 10, 11]. Regardless of the initial saturation, evaporation is broken down into two distinct periods: rapid evaporation at constant flux r which occurs until the stone has reached a critical residual saturation Se followed by a slower drying phase. " Sample

Initial saturation Under vacuum Rate cq-cq,(. 103) Criticalwater (g/cmZ./h) comem,so, (0,6) VILL 1.68-1.97" 38 CHAM i.79-1.88 ~ ' 4i LOUD 2.12 35 FONT . . . . 2.03. . . . 66 LLIZE 1.84 .... 68 Table 7: Petrophysicaiproperties of drying dynamics. .

.

.

.

.

Initial partial saturation (N48) Rate at2 (. 103) Criticalwater

(g/omVh)

, cont., sc~ (%),

2.01

54 54 60 70

1.93 2.27 1.98 1.83

.

.

.

.

.

79

During the first phase, evaporation occurs primarily along the free water surface, and since gravitational effects are ignored, the water movements resulting from capillary effects predominate. Below the water content Sc, the capillary forces causing transfers of water present in the samples are no longer sufficient. This lack of liquid movements attests to a reduction of the driving forces, which increases as the residual water content of the samples decreases. During desaturation, the capillary pressures applied at the water meniscii are very high. These pressures, which only affect the volumes of water in the smallest pores or in the dihedral angles at the grain boundaries, progressively decrease and no longer exert a sufficient force to drain the water towards the evaporating surface. This transition phase,


75

which conditions the remaining water coment Sc in the stone, is generally associated with the fact that the liquid network becomes discontinuous [7, 12], only partially connected by water films that are so thin as to be no longer mobilizable. The break in the hydraulic continuity of the network then results in a significant decrease in the evaporation rate. This marks the onset of the new, slower drying phase, at decreasing flux rate, primarily governed by water diffusion mechanisms in a pore network, in which the saturation vapour pressures are no longer present.

Figure 3: Characteristic evaporation curves (LUZE and VILL). The drying tests suggest contrasting behaviours between the different tuffeau facies, depending on the initial saturation mode. In the case of drying atter total imbibition of the sample pores, the duration of the capillary water remobilization phase corresponds to a fairly long interval during which the flux oq can be considered constant as a first approximation. The corresponding residual content Sc~ is relatively low in comparison with the duration and rate of this phase (Table 7). The values of Sol between the different samples are sufficiently different to class the facies into two categories: tuffeau with a low residual water content (Sc65%; FONT and LUZE). These two families are thus distinguished by the nature (by mass) of the predominant evaporation regime governing their drying. A detailed analysis of the morphology of the curves (Fig. 3) reveals that the rate OCl is not exactly a constant. A break in slope occurs above a certain desaturation value of the tuffeau, corresponding to an acceleration of the drying phase before reaching the value SOl. This feature, which is only observable in tuffeau with low water content Sc (LOUD, but above all CHAM and VILL), suggests that capillary transfer occurs with two successive kinetics, with slopes OCl then cq' with Ctl

9

/

/

/

/

20,00 ~/ / 0,00 v 0,00

10,00

1

i

i

20,00

30,00

40,00

50,00

Ca2. (mg/I)

Figure 14:S04 2- and Ca2§ scatter diagram Concerning seepage waters, S.I. values increase for most of the minerals, but saturation is only reached with respect to calcite and aragonite (tab. 2). However, only a small number of these waters are over-saturated with respect to the above mentioned minerals. Table 2" Log S.I. (Saturation Indexes). Seepage waters Mineral Sample

S1 $2 $3 $4 $5 $6 $7 $8 $9 SI0 SI1 S12 S13 S14 S15 S16 S17 MAX MIN

Calcite Aragonite Dolomite Gypsum Halite 1,01 0,86 0,43 -3,77 !' -6,68 2,07 1,92 0,67 -2,80 -7,47 0,22 0,07 -0,64 -4,67 -7,47 0,79 0,64 0,43 -4,49 -7,31 0,87 0,72 0,52 -4,23 -6,97 0,47 0,32 0,02 j -4,59 . -7,08 0,72 0,57 0,11 . -4,08 . -6,85 0,46 0,32 -7,46 -0,87 -1,01 -2,46 "i -4,18 . -7,10 0,58 0,43 0,61 -4,59 -7,18 0,59 0,44 -0,24 -4,30 -7,24 0,60 0,45 -0,05 -4,50 -7,33 0,60 0,45 -0,07 -4,42 -7,25 0,01 -5,33 -7,2, r 0,57 0,42 0,53 0,38 -0,26 -4,82 -7,38 0,42 0,27 -0,34 -5,03 -7,58 0,66 0,52 -0,02 -4,65 -7,45 2,07 1,92 0,67 -2,80 -6,68 -0,87 -1,01 -2,46 -5,33 -7,58 .

.

.

.

Trona Thenardite -10,96 -7,78 -14,08 -8,81 -13,59 -8,99 -13,75 -9,22 -11,62 -8,57 -12,08 -8,85 -11,68 -8,25 -13,51 -14,00 -8,80 -12,41 -8,72 -12,45 -8,62 -12,63 -8,99 -12,67 -9,07 -12,99 -9,65 -12,68 -9,24 -13,20 -9,50 -13,22 -9,27 -10,96 -7,78 -14,08 -9,65

Another relevant aspect is that seepage waters are under-saturated with respect to trona and thenardite (locally found soluble salts). This implies that these salts could only be precipitated due to seepage water evaporation. Regarding the scatter diagrams that have been prepared from seepage samples some remarks should be made:

86

9th International Congress on Deterioration and Conservationof Stone, Venice 19-24June 2000 9 The pattern of points plotted in the scatter diagram SO4 2" concentration versus Ca 2+ concentration (fig. 12), shows a rather narrow range of variation and considerable clustering of points so that a regression line cannot be drawn through them. However, they are plotted either over or very near to the theoretical SO42/ Ca 2+ gypsum ratio. So gypsum could be pointed out as a probable sulphate salt that can be precipitated from seepage water evaporation. 9 The regression straight line (with a slope of 0,1986) fitted to the data points of the scatter diagram SO42" concentration versus Na + concentration (fig. 13), is very far from the theoretical SO42/Na § thenardite ratio. This shows that sources of these ions other than thenardite should be considered. Seepage waters show high concentration of Na § The thenardite hardly will be the most important sulphate salt precipitating from seepage waters evaporation. 9 Trona will not be the most important nor the only carbonate salt precipitating from seepage waters evaporation, according to a similar interpretation of its respective scatter diagram. In spite of HCO3" concentration in seepage waters is strongly correlated to the Na § with a correlation coefficient equal to 0,87 (fig. 14), the regression line fitted to the points plotted in the scatter diagram is relatively far from that oftrona dissolution. This is, the molar ratio of bicarbonate to sodium in seepage waters is relatively far from the theoretical value of 0,8847 suggested by CO32/Na § ratio of trona.

4. Conclusions Since seepage waters are under-saturated with respect to trona and thenardite, these salts could only be precipitated through seepage water evaporation in this monument. As environmental conditions for this soluble salt precipitation are the same all over the inside Basilica da Estrela, one has to think on a local source and/or enrichment of salt solution that could intensify the precipitation of these salts. This fact could be attributed to cleaning activity and repair (maintenance/restoration) works performed in the last few years. Besides, the evolution of seepage water composition under the environmental conditions found inside Basilica da Estrela, shows that the stone decay induced by salt deposition cannot be attributed to trona and thenardite, given the punctual spatial distribution and quantity expected to occur. However, this is not expected for calcite as is confirmed for example by the existence of secondary calcite deposition as crusts, stalactites and stalagmites observed inside the church. Although there is strong evidence of the contribution of sea water to rain water composition, there is no evidence, inside the church, of the influence of chlorides dissolved in the seepage waters as likely causes of the monument stone decay in the sampled areas.

5. Acknowledgements This study was partly financed by Praxis project 2/2.1/CTA/437/94. 6. References Aires-Barros L., 1991. Altera~:ao e Alterabilidade de Rochas. INIC, Lisboa. Arnold A., Zehnder K., 1989. Salt Weathering on Monuments. The conservation of monuments in the Mediterranean Basin, Proc. of the First International Symposium, Bari, Grafo Edizioni. 31-58.


87

Begonha A., Sequeira Braga M.A., Gomes da Silva F., 1995. A Ac~.o da /kgua da Chuva na Meteoriza~ao de Monumentos Graniticos. Mem6ria n~ 4, Universidade do PortoMuseu e Laborat6rio Mineral6gico e Geol6gico. 177-181. Begonha A., 1997. Meteorizaqao do Granito e Deteriora~ao da Pedra em Monumentos e Edificios da Cidade do Porto, PhD. Thesis, Universidade do Minho, Braga. Carvalho M.R., Almeida C., 1989. HIDSPEC, um programa de especia~ao e chlculo de equilibrios ~iguaJrocha. GeociSncias, Revista da Universidade de Aveiro, 4 (2), 1-22. Figueiredo C., 1999. Altera~ao, Alterabilidade e Patrim6nio Cultural Construido: O caso da Basilica da Estrela. PhD. Thesis, Universidade T6cnica de Lisboa, IST.


89

ANALYSES OF THE PHYSICAL PARAMETERS CORRELATED TO BENDING PHENOMENA IN MARBLE SLABS. Carlo A. Garzonio* Dip. Urbanistica e Pianificazione del Territorio, Firenze, Italia Fabio Fratini F & Carlo Manganelli del F/t CNR,Opere d'Arte, Firenze, Italia. Prisca Giovannini & Fabio Cavallucci Dip. Storia dell'Architettura e Restauro, Firenze, Italia

Abstract

The paper analyses mechanical decay processes in stone materials, which are represented mainly by Carrara marble used in architecture, historical buildings and monuments. These processes are often related to the bending of marble-slabs and architectural elements which are affected by creep processes. Mechanical and physical parameter analyses were carried out on some of the more representative slabs, and significant variations were found in the mechanical resistance, density and above all porosity values. Other slabs and bars were prepared from quarry material and they were subsequently bent with long term flexure load tests. The analyses of thin sections and of SEM images allowed us to identify the various structures of the blasts, the microfractures caused by pressure or disconnections, etc. In particular the results of systematic analysis carried out on over 80 slabs from some cemetries in Florence- are discussed and using these data preliminary models of creep phenomena are being drawn up. Key words: creep phenomena, physical and mechanical decay, Carrara marble slabs.

1. Introduction Research has been underway for a few years now on mechanical decay processes of stone materials - represented mainly by Carrara marble, in particular white marble- used in architecture, historical buildings and monuments. These time-dependent deformation phenomena are often triggered off by effects of stress release, due, in turn, to residual stress (Voight, 1966; Kieslinger, 1967). The processes are connected to the tectonic and geomorphological history, to the effects of new environmental conditions and to the new geometry of the stone elements. It is very important to identify and study these phenomena as it has already been proven that they play an important role in controlling and increasing the effects of environmental and meteoric degradation and atmospheric pollution, and of physical-chemical surface weathering. At the same time, in many cases, creep processes are the vehicle for the more rapid, and sometimes obvious, effects of bowing type deformation, caused by thermal cycles or by the incorrect fixing of stone materials. There have been some well known cases in recent architectural works which present bending phenomena in the covering material, e.g. the Amoco Tower (Chicago), the Arc de la D6fense (Paris), the Finlandia House (Helsinki). Other cases are related to ancient monuments, e.g. the many facades and external walls of Romanesque and gothic churches, in Tuscany in particular. In many cases we have a modest deformation effect, however micro-deformations, and

90


sometimes micro-cracks, are present in the intergranular and transgranular structure of the marble and they are a vehicle of degradation processes. As regards the state of art of the research in this field we should mention in particular the papers by Winkler (1972;1996), Logan et alii (1993), Sorace (1996). A more detailed analysis of the literature on creep phenomena linked to the problem examined is contained in previous papers by the research group (Garzonio et alii, 1995, Cavallucci et alii,1997). Some of these international contributions refer to creep phenomena which involve rock masses, where we can nevertheless deduce behaviour laws which are useful for defining the bending processes of slabs and of architectural elements (Cristecu,1985; Fakhimi & Fairhurst, 1994; Fossum, 1977; Hudson et alii, 1993; Ottosen, 1986 ).

2. Experimental Methodology Having said this, it is clear that it is necessary, not only to continue the studies and the tests for characterizing and modelling creep phenomena, with the identification of the constitutive laws, etc., but also to examine and apply analysis and forecasting methods to small scale processes. Deformation tests were therefore carried out on various bent, white Carrara marble slabs in order to evaluate physical and mineralogic parameters. Mechanical and physical parameter analyses were carried out on some of the more representative ones, and significant variations were found in the mechanical resistance, density, and above all, porosity values. Other slabs and bars were prepared from quarry material and they were subsequently bent with long term flexure load tests. The parameters refer to different patterns of the values in relation to the strains and the position of the samples with respect to the concave and convex surfaces. The study then focused on the microscopic analysis of thin sections and on the analysis of SEM images of various points of the slabs. These analyses allowed us to identify the various structures of the blasts, the evolution of the deformations and discontinuities, the presence of cracks, the microfractures caused by pressure or disconnections, etc. In particular the results of systematic analysis carried out on over 80 slabs from some cemetries in Florence are discussed. The data resulting from this analysis is contradictory at times if we correlate some geometric and physical parameters with the deformation pattern. The slabs refer to the period between 1833 and 1986. Five of them were examined in detail, applying all the above mentioned tests, the aim being to perfect and improve study methods in order to normalize all the various analyses and test phases. This operation becomes even more necessary in the very frequent case of interacting processes in order to understand the role played by creep phenomena linked exclusively to the weight of the slab itself, to stress release, to thermal or environmental variations and to fixing techniques.

3. Previous results In the initial phase of the research programme, tests were carried out, (and they are still being carried out) on slabs of Carrara marble from the fagade of the Collegiata di Sant'Andrea in Empoli (near Florence), and from some cemeteries in the Florence area. It is a well-known fact that the Carrara marble in the Apuan Alps in the North of Tuscany derives from limestone, as a result of a tectonic-metamorphic deformation, which occurred from around 27 to 12 x 10 6 years B.P. The latter greatly deformed the pre-existing stratigraphic structure and created new ones. White Carrara Marble and Ordinary White Carrara Marble are homogeneous white fine grained marbles. Grain size is bimodal,

91


clustered around 0.1 and 0.3-0.4 mm; (Blasi et al. 1990). Table 1 contains data regarding some physical parameters and mechanical properties. . . . .

Table 1" Mechanical parameters and physical properties of the White Carrara marbles unit

A

B

Uniaxial compressive strength MPa 133.2 MPa 128.2 Compressive strength after gelivity MPa 18.5 Bending rupture load 10-6 C 5.93 Linear thermal expansion coeff. % 0.131 Imbibition coefficient KN/m 3 26.92 Unit weight cm 57.80 Impact collapse weight MPa 71370 Young's modulus mm 5.27 Wear thickness loss A" White Marble (Bianco P); B' Ordinary White; C: Quarry E:Slab (cemetery)

C

E

D

136.6 128.1 36.7 53.7 129.8 122.0 16.5 18.8 2.0 7.12 7.4 0.129 0.090 0.16 0.19 26.97 26.85 27 27.2 58.2 65.0 71620 64800 27800 53800 5.31 5.55 (Ordinary); D:Slab (Collegiata);

In particular, the geomechanical features, the extent of the viscous deformations of the marbles and the results of the laboratories tests are described, the objective being to obtain a preliminary definition of the creep phenomenology. It is within this context that the slab from the facade of the Collegiata of Empoli was analyzed from the point of view of the variations in some physical and mineralogic-petrographic properties linked to the deformations (Garzonio et al. 1995). This analysis highlighted the different geometric distribution of the points of physical-mechanical decay. On the basis of the results of the physical analysis, we find an interesting correlation (see example in Fig.l) between the deformations the material has undergone and, above all, the porosity values (mean porosity value is 5.2 % for the concave side of the slab, and 6.1% for the convex one, with an elevated coefficient of variation CV=35%, max value 10.5%, min 3.9% ). This property determines and highlights the degradation processes. 3O

10

% o

20

4

%

~2

"~ 10 5 0

0 0

5

10

15

bending strength(MPa)

Figure 1.

Correlation

between

20

25

0

5

10

15

20

25

bending strength(MPa)

bending strength, porosity and strains

As a confirmation of these results, two white Carrara marble from two different quarries, located in zones with slabs of different tectono-metamorphic histories were analysed (Barsottelli et alii, 1998). The experimental results highlighted two different granoblastic microstructures, characterized by regular, straight grain boundaries or, on the contrary, by

92


irregular and suturated grain boundaries. These different structures are correlated to different physical properties as table 2 shows. Table 2: Physical parameters and microstructural features Marble type ~'s P P(Hg) ICv SI A 2.65 2.4 0.7 0.79 34 B 2.69 1.3 0.2 0.16 12 Per S/L Sg D 0.4152 0.173 0.118 0.0198 A(Mean) 0.66 0.146 B(Mean) 0.0159 0.7908 0.156 0.108 0.68 0.132 7s =bulk density (t/m3); P= total open porosity(%); P(Hg) = mercury porosity (%); ICv = volume imbibition coefficient (%); SI = saturation index; Sg = grain surface; Per = grain perimeter; L = long axes; S = short axes; S/L = axial ratio; D = mean grain diameter.

Figure 2. Thin sections of type A and type B marble. Cross polarized light.

4. Sampling and analysis A systematic analysis was carried out on eighty slabs of white marble characterized by clear signs of deformation, from six different cemeteries in Florence. These slabs which are directly exposed to the elements and face different directions, form the covers of ground tombs (horizontal slabs) and wall tombs (vertical slabs) respectively. As far as the marble is concerned, it was identified macroscopically on the basis of the main colour and the absence of veining of a different shade. As regards dating, the year which figures in the inscription on the slabs was taken as the year the slabs were laid. In this way the time interval was set at about 160 years (1830-1990). The deformations were measured with respect to many sections; the maximum linear strain (s) with respect to the plane on which the slab is laid (both horizontal and vertical), or the tangent at the base of free standing vertical slabs. A total of 45 horizontal slabs and 16 vertical slabs were examined. The slabs were of varying lengths, ranging from 51 to 237 cm. Taking into consideration this parameter, it was possible to identify three groups of horizontal slabs ( I, II, III,) and two groups of vertical slabs (IV, V). The mean value of these intervals is equal to about 93 cm Group IV; 98 cm Group I; 174 cm Group II; 208 and 210 cm Groups III and V respectively. Four horizontal slabs and three vertical slabs were not included in these intervals, two horizontal ones have a length of 237 cm, one 125, and the last one 140 cm, with widths varying from 78 to 90 cm. The three vertical slabs have lengths of 51, 77 and 103 cm. The other dimensions and the strain are comparable with the ones shown in Table 3. The bending B, both longitudinal and transversal, was measured with B-1/r (where r is the bending radium).


93

Table 3" Dimensions, strain and age of the slabs Group length width thickness total cm cm cm number I II III IV V

strain B1 B2 Age mm cm ~ cm l 11.6+3.6 0.00132 0.00139 1880 97.8_+8.8 75.5_+4.3 3.5+0.8 12 1961 13.5+5.7 0.00056 0.00135 1867 174.1+11.5 77.3+9.8 4.9+2.5 22 1986 14.2+3.1 0.00028 0.00102 1891 208.3+4.4 78.9+9.1 4.7+2.1 11 1956 17.5+9.1 0.00135 1833 92.5+0.7 60+12.7 2.0+0.2 4 1851 16.5+8 0.00044 1908 210+0.6 64.2+0.4 2.2+0.4 12 1924 B 1= longitudinal bending ; B2 =transversal bending (mean values)

As we can observe in Table 3, given that the width (average value 77.26 cm) and thickness (average value 4.38 cm) of the horizontal slabs are similar for all groups, the length does not influence the maximum linear strain at all. The vertical slabs show greater strain values up to about 40 mm (and B1 = 0.00073), due to the positioning, the free edges and the greater stress. Regarding this point, vertical slabs measured showed outward bulging even though inverse phenomena were observed on other slabs not considered in this study (Garzonio et al. 1995). The deformations of the horizontal slabs laid at the edges and rarely fixed, present downward bending in the centre and upward bending at the edges, with the exception of two cases which still have to be clarified. On the contrary, the vertical ones present more irregular deformations. Nevertheless situations are often found where the distribution of the deformations is irregular and they do not correspond with the configuration of the stresses due to their own weight. However as far as dimensions are concerned, it is possible to make some observations on the different extent of the deformations. Greater "normalized" deformation values are found as a function of the 1/s (length-strain) e w/s (width-strain) ratios, and these are confirmed by B values. With thickness of less than 2.5 cm the relative strain (~%) increases, almost independently of the dimensions. Creep tests have been carried out and are still underway, based on the results of the previous tests, and in particular of the bending tests, which identified a reduction of up to 34% as to the instantaneous breaking load. They consist of bending tests (Ito & Sasajima, 1987), compression tests and tension tests with 0.35 ~. In addition some triaxial tests were carried out, the main aim being to try to identify the anomaly of the main stresses. The material analysed regards white marble slabs from quarries which are well known from a geo-structural point of view. The diagrams of bending tests are given in Figures 2 and 3.

94


bending test

bending test

Figure 3. Time to rupture points

Figure 2. Bending Tests (0.35 of instantaneous ~)

Table 4 shows the limit parameters of the three creep phases obtained by the tests. These experimental data are in agreement with the measurements and data relating to the tests on the ancient deformed slabs, even though they are almost equal to breaking values at times less than 25% of the unaltered instantaneous ones. Long term tests are still underway on 5 slabs chosen to represent the ones used for the census. However the instantaneous tests have already given even lower breaking values and, above all, very low moduli of elasticity. Table 4. Creep parameters from bending test g0 tI ~;i t" (xl 04) 8.52

(days) 75

(xl 05) 10.84

(days) 635

tm

III

(days) 710

(xl0 -5) 23.16

~EII

(xl0 -5) 18.15

5. Preliminary analytic model The second phase of the research concerns an initial classification of the viscous deformation processes which are only indicative of some of the cases observed (Garzonio 1995). This leads us to place them within the general creep law (Hudson et al. 1993 ): = r l & / d t + ~E (1), which can be attributed to simple linear visco-elastic behaviour (Kelvin-Voight) where:

~(t) = A tn (0 .

,

1200

.

.

.

,

1400

.

.

.

,

1600

.

Year of documentation

.

.'

,

1800

.

.

.

,

2000

Figure 4: Damage development of the five granite monuments. Note the rapid damage progression during most of the recent periods.

2200

124


natural weathering. This also indicates that the studied material is not ideal for the intended purpose. It will, however, not alter the general picture of accelerating weathering. We have also observed that the increase in weathering of granite with time is smaller than that of the other rock types. The reason why the sedimentary rocks are more vulnerable is their anisotropy. They often tend to flake along their bedding; the stones can be split into several thin slabs. This fact makes many of the sedimentary stones to weather strongly in the long run. The damage development of individual stones with time is also of interest. The granites are the least influenced by human interference. Hence, these have been plotted in fig. 4. All reliable documentations are included in this diagram. On one of these stones (U 455) all runes are preserved undamaged. On the most weathered one (U 58), 1/3 of all runes are damaged. We can also note that during some shorter periods (decades) the damage is negligible, during others the deterioration is rapid (large slope). This is a consequence of the episodic nature of physical weathering. The overall pattern during a millennium is initially more or less continuous chemical (and biological) weathering weakening the surface. This slow continuous process is in course of time interrupted by episodic and extensive physical weathering resulting in discrete material loss (erosion).

4. Conclusions The present study shows that chronological century-long studies of stone weathering are feasible and necessary to define the total deterioration. Especially the physical processes can only be estimated after long periods of time. More or less continuous chemical dissolution and biological influence are in time interrupted by episodes of physical material loss, giving a steep decay curve. All types of rock show this pattern, although the rates differ strongly. In order to prolong the survival of a stone object, it is important to establish when the physical weathering becomes the major decay factor. When, or still better before, this process starts, the environment of the object must be improved or the object will have to be moved to more harmless surroundings.

5. Acknowledgement We thank Stig Englund, Visby for the construction of fig. 2.

6. References Brate, E. & S~derberg, S., 1900-06. The runic inscriptions of r (In Swedish). Books 1-2. Stockholm. Jansson, S.B.F., 1981. The runic inscriptions of G~strikland (In Swedish). Book 1. Stockholm. Jansson, S.B.F., 1987. Runes in Sweden. Stockholm. GOransson, J., 1750. Bautil. Stockholm (In Swedish; in the Royal Library, Stockholm). L6fvendahl, R., Andersson, T., Aberg, G. & Lundberg, B.A., 1994. Swedish building stones and damage atlas (In Swedish with English summary). Central Board of National Antiquities (now National Heritage Board), Stockholm.

125

PETROPHYSICAL ANALYSIS OF CATHEDRAL OF BURGOS, SPAIN

THE

SCULPTURES

DECAY AT

THE

Rafael Fort Gonz~dez* Institute of Economic Geology (CSIC-UCM), Madrid, Spain 1VPConcepci6n Lopez de Azcona Institute of Economic Geology (CSIC-UCM), Madrid, Spain Francisco Mingarro Martin Department of Petrology, Institute of Economic Geology (CSIC- UCM), Madrid, Spain

Abstract The statues located on the facade known as Fachada de Santa Maria in the Cathedral of Burgos are sculpted on a variety of limestone (biosparite) extracted from the Hontoria quarries. The petrophysical characteristics of the statues portraying the Infantes, added to the monument in the mid-14th century, are slightly different than those of the Kings, dating from the 19th century. The limestone used for the Infantes is finer-grained (has smaller pores) than that used for the Kings. The limestone used to sculpt the Infantes is more homogeneous, uniform and dense than that used for the Kings, which is more fragile and breakable. On account of their petrophysical characteristics and of the durability assays carried out in a climatic chamber, we can conclude that the Kings statues are more sensitive to decay processes derived from freeze-thaw cycles whereas the Infantes owe their decay mainly to wetting-drying variations. The state of preservation the Kings statues was assessed using ultrasound prospecting techniques and the following results were obtained: 14 % of the damage was found to correspond to cracking, 30 % to fissures and 30 % decay due to a loss of cohesiveness. In the Infantes, 10% of the decay corresponded to fissures and 55 % to degradation owing to dissolution. Keywords: Decay, limestone, Hontoria, durability, hydric behavior, ultrasounds, heritage

1. Historical and architectural issues On July, 20, 1221, King Fernando III and Bishop Mauricio lay the first stone of the Cathedral of Burgos. Sculptures were placed in the monument around 1350. 8 sculptures portraying unidentified members of the Royal Family (known as the Infantes) were placed on the third section of the Santa Maria facade (fig. 1). A great number of sculptures was also placed on the archivolts of the first section. On the four buttresses that divide the three entrance doors to the cathedral there are 16 additional sculptures. Between 1753 and 1768, after remodeling of the atrium carried out in 1663, many of these statues were removed in order to prevent their collapsing, and in 1788 they were declared "in decay due to the passage of time", broken and unrecognizable (Orcajo, 1856). All imagery was removed from

126

9th International Congresson Deteriorationand Conservationof Stone, Venice 19-24June 2000

Figure 1" The eight statues of the Infantes on the third section of the Santa Maria facade the first section in 1790, the fagades were refurbished and all sculptures disappeared from the archivolts and jambs. The four statues currently seen in the Cathedral were placed there in 1805 in place of the former 16 ones that, according to some scholars, were sculpted expressly for the Cathedral. The two statues located on the buttress on the let~-hand side portray King Femando III and Bishop Mauricio, whereas the other two, on the right-hand side buttress, represent King Alonso VI and Bishop Asterio. We refer to these four statues as the Kings. 2- Atmospheric characteristics There is any weather station in the Cathedral itself. In consequence, the only existing data for the last 30 years was obtained from the air base of Villafranca, 7-km northeast from the monument. The main weather conditions in the area are summarized in Table 1. Thermal winter in Burgos is long, as temperatures remain below 10~ for six months; summer is mild, with temperatures above 17~ during July and August. The minimum temperature is -22~ with strong thermal oscillations. Absolute values for these oscillations have reached 39.0 ~ in January, but there are yearly thermal oscillations of up to 60~

Tablel Climatic data in Burgos (Station of Villafranca) 9,9 ~ Annual average rainfall Annual average temperature Rainfall (maximum) 15,5 ~ Average temperature (max.) Rainfall (minimum) 4,3 ~ Average temperature (min.) Annual rainfall 38,0 ~ (Ag.) Absolute temperature .(max) Absolute temperature (min.) -22 ~ (Jan.) Maximum rainfall (24 h) Minimum rainfall (24 h) 39.0 ~ Thermal Oscillation (maxi.) 31,4~ (April) Diary relative humidity Thermal Oscillation (min.) Directions winds Wind velocity SW; W ,,

9

.

,

571,9mm J 89,2 mm (Jan. 83,7 mm (Jul) 1702,3 mm 51,6 mm(Jun) 30,2 mm (Febr) 74% 15 Km.h"1


127

Average rainfall in a year is of 47.7 mm, spread out over the year between May, with 65.0 mm, and August, with 24.8 mm. This abundant rainfall, added on to the effect of river Arlanz6n, which flows nearby, makes relative humidity in the area very high with an average of 88 % at 7 a.m. As a result of these conditions the climate in Burgos can be defined as COLD and HUMID, with some oceanic influence. According to Kreppen's categories, the climate in Burgos could be classified as humid mesothermic oceanic weather. The statues in the Santa Mafia facade, that faces west and is subject to strong westernly and southwesternly winds that coincide with the most abundant rainfall, suffer greatly from these extreme atmospheric conditions There is little information available regarding atmospheric pollution, but its effects are noticeable on the stone, that suffers sulfation, dissolution and soiling revealing the existence of sulfuric oxides, acid rain and suspension particles.

3. Method In order to classify the materials and monitor changes, ten 53-mm diameter and 200-mm long samples were collected. These samples were then subdivided into various test specimens of different morphology according to the type of assays to be done. Assays were carried out following the RILEM (1980) and the CNR-ICR NORMAL (1981-86) guidelines. The following assays were carried out: petrographical analysis under polarized light microscopy, X-ray diffraction, chemical analysis, hydric behavior (water sorption and desorption, capillarity, water steam permeability, contact angle, density, chromatic parameters, mercury porosimetry, ultrasound propagation velocity). Durability tests (carried out in a climatic chamber) included 24 freeze-thaw cycles and 20 wetting-drying cycles.

4. Petrological characteristics 4.1 Petrography The statues are cast in white limestone with some yellow hues. Fossil traces can be found in the stone as well as wide pores filled partly by microcrystalline calcite. The oldest statues, Infantes, made of medium-grain stone, are more homogeneous, while the Kings are made of coarse grained heterogranular stone with abundant fossil traces. These materials were extracted from the Hontoria quarries located on calcareous formations from the Cretaceous (Turonian), located south of Burgos. Under the microscope, the limestone shows abundant bivalves skeletons with lesser amount of echinodermata, bryozoans, foraminifers, corals and micritic peloids. The Kings are also made with fossiliferous limestones. Fossil remains in these limestones are smaller in the Infantes are richer in peloids. The skeletons are cemented with sintaxial sparite, butther whereas poikilotopic sparite is pervasive in peloidal-rich parts of the rocks. The stone of the sculptures can be classified, according to its composition, as biosparite and, according to its texture, as grainstone. This indicates the absence of micrite matrix.

128


4.2 C h e m i c a l analysis

The average results obtained from the Hontoria limestone samples are shown in table 2. These results indicate that we are dealing with a very pure and homogeneous stone, made up of calcite, with only 1.5 % of impurities. Dolomite, orthose and albite constitute these impurities. Table2: Chemical anab,sis of the SiO2 A 1 2 0 3 Fe203 0,33 0,17 0,09 +0,15 +0,17 +0,06

stones from Hontoria CaO MgO Na20 55,27 0,14 0,02 +0,50 +0,02 +0,01

K20 0,04 +0,02

SO4-0,02 +0,01

LOI 43,57 +0,37

4.3 P e t r o p h y s i c a l characteristics

The chromatic characteristics of the surface and internal areas of the statues are shown in table 3. A high degree of lightness is characteristic for both types of sculpture, although lightness is slightly more pronounced in the Infantes. As is reflected in their chromatic indexes, the Kings statues tend to show yellower hues, while the Infantes are whiter. Chromatic differences between internal and external areas of the statues indicate that a protective treatment was applied to the stone in the past. Table3: Chromatic parameters of the stone L

KINGS

Surface Internal

~

83.65 86.76

a*

b*

C*

IB

IA

2.72 1.99

14.31 10.19

14.57 10.38

4.43 23.86

23.28 16.51

,

Surface 86.65 2.57 12.11 12.38 15.81 19.37 INFANTES Internal 89.36 1.72 8.77 8.94 33.44 13.95 L*= Lightness a*=Red hue b *= Yellow hue. C *= Chroma. IB= Whiteness IndexlA= Yellow Index. The main petrophysics results obtained are shown on table 4. Stone porosity for both the % for the former and 20.2 % for the latter. Pore size distribution is shown in figure 2, which illustrates the similarities

Infantes and the Kings is high, with average values of 23.3

Figure 2: Pore size distribution


129

between both types of stone. The main differences are found for pores bigger than 5 ~tm (macroporosity), where a modal gap between 10-20 ~tm was found for the Infantes, and 3040 ~tm for the Kings. Table4: Physical properties of the sculpture Saturation (%) PorOsity accessible to water (%) Real density (Kg-m3) Bulk density (Kg-m-3) Compacity Ultrasonic velocity (m-s1) Classification (A.S.T.M.) Coefficient Of Capilarity (Kg~m2.s-1) Coefficient of water sorption (%) Water vapor permeability (g.m2.hl) ' Porosity by Intrusion Hg (%) (200-0.006 :m) Microporosity (%) Macroporosity (%) Mean size of the pores (~tm) Specific surface area (200-0.006 ~tm) m2.g

....

,

....

INFANTES KINGS 7.42 9.50 16.9 20.44 2,703 2,708 2.252 2,159 0.83 0.79 2,648 2,408 Medium-low Low 0.0533 0.1215 4.59 6.21 9.34 12.39 20.20 23.30 10.50 9.80 9.70 13.5 1.4 1.3 0.24 0.30

4.3.1 Mechanical properties Both types of statues have a similar density. The Infantes have slightly higher values as the stone is finer grained. In general, density is higher for the inside of the stone than for the outside as a result of the changes the stone has undergone. Bulk values are higher for the Kings than for the Infantes by 93 Kg-m3. This difference indicates that the Infantes have undergone more acute dissolution processes and that the Kings have a higher recrystallization rate. The bulk-real density ratio reveals that the stone with which the Kings statues are made is more compact. The ultrasonic waves propagation velocity values reveal low mechanical quality, lower for the Infantes than for the Kings. In determining the dynamic modulus, the values obtained are also quite low, which highlights the structural differences between both types of stone. According to the ASTM C-568 guidelines, and taking the dynamic modulus and the stone density into account, the mechanical quality of the materials can be described as medium to low for the Kings and low for the Infantes 4.3.2- Hydric behaviour On account of the atmospheric conditions to which the sculptures are subject, their resistance to decay processes is very directly linked to their hydric behaviour. The main results obtained are shown on table 4. It should be noted that the amount of water the limestone used for the Infantes absorbs is higher than the amount of water the Kings absorbs. Another important characteristic of the Kings statues is that they absorb more

130


water (% saturation) into their internal zones, which might indicate that a protective treatment was applied in the past reducing open porosity on the surface of the stone. Although sorption kinematics is similar for both types of material, some differences are observed owing to the textural variations (fig. 3). Hence, sorption takes place at a faster rate for the Infantes. Aiter the first 30 minutes, the Infantes have already taken in an 57 % of the water necessary to reach saturation levels, while the Kings have only taken in 54.31%. Aiter this, sorption becomes very slow and after, 7 days, the Infantes have taken in up to 70.2 % and the Kings 66,12 %.

%80 0

~

60

=

m

m

m

m

m

Im

Im

Ira|

Im

|

.f

|

|

n

,m

m

|

|

m

m

KINGS - - -INFANTES

50 ~l

40 30

. 1

0

. . . . 2 3 4

. . . . . 5 6 7 8 Time (hours1/2 )

. 9

.

10

11

12

13

Figure 3: Kinematic of the desorption of water

%

100

90

~

'

80

KINGS

" " " INFANTES

70

~

60 50

40 30 1

. . . . . . .

20

0

1

2

3

4 1/2 5 Time (hour ) Figure 4: Kinematic o f the desorption o f water

6

7

Desorption is a slow process which behaves in a similar way for both materials (fig. 4). Up till 11 hours, the Kings lose more water. At that point there is an inversion in the process and more water is evaporated from the Infantes. This is due to the pore size and morphology of the Kmgs that causes obstructions that slow down evaporation. Capillary water sorption for the Infantes is more than double that of the Kings (fig 4). The Infantes take in water quickly from the first hour, whilst the Kings do so more slowly


131

up to the first 1.77 hours. Absorption gradually slows down reaching the end of the assay 6.25 hours later. After this period of time, the Infantes have taken in 7.57 Kg-m-2, while the Kings have only taken in 5.21 Kg-m2. The Infantes have smaller pores, so they can absorb more water faster than the Kings can.

(Kg.m z)

d

j

.

KINGS -

0

i

i

50

Tim I oo (s 1• )

-INFANTES i

150

200

Figure 5: Capillarity water sorption Water vapor permeability is the process that controls evaporation in the inside of the sculptures. This petrophysical parameter, higher for the Infantes, allows for better kinematics in water desorption after 11 hours. These data are indicative of possible decay the Infantes may undergo due to wetting-drying cycles. Solid-water contact angle is close to 40 ~ for both types of stone, thus indicating it is akin to water and liable to suffer severe alterations

5. Durability of materials Durability assays, carried out in a climatic chamber, indicate that the Infantes sculptures are more sensitive to wetting-drying cycles, with 0.3 % weightloss after 20 cycles. On the other hand, these statues are more resistant to freeze-thaw processes, where petrophysical parameters hardly suffer any variations. The Kings statues suffer an average weight loss of 0.1% as a result of freezing with a significant increase in open porosity (2.66 %), and, most notably, a decrease in ultrasound wave propagation velocity (410 m.sl). This alteration indicates that a decay process has already started in these materials.

6. Statue decay In view of the compositional and petrophysical characteristics of the sculpting materials, as well as the location of the statues in an aggressive environment, it can be deduced that the stone has decayed dramatically. Already the carving of the statues results in the opening of innumerable chiseling microfissures conducive to water intake. Water absorbed increases its volume with temperature variations and acts as a wedge. This is especially so when it is transformed into

132


ice. As this happens within a closed system, volume increases by 9.5 % and at -30 ~ C it generates a pressure of up to 2,100 Kg.cm -2 The statue portraying Bishop Asterio had a severed head, and the one representing King Fernando was also split at the waist. Both statues were restored and held together using natural resin (quite possibly pine). They were then reassembled using a 3 cm-wide wooden pole. Later, the wood expanded with humidity and caused severe fissures. All the statues were attached from the shoulders to the walls with iron clamps, and secured with pressed lead. The Infantes were similarly secured onto their base. As iron oxidizes and becomes hydroxide (Goethita), it increases its volume by 291%. As a result cracking and fissuring occur. Acid water loaded with carbonic, sulfuric and nitric acid, etc. penetrates cracks and fissures and drips down the surface of the sculpture. Limestone undergoes dissolution processes when the pH level of water dripping on its surface or its porous system is below 7.8. These dissolution processes are more noticeable for small-grain minerals (micrites) than for larger-grained stones (sparites). This effects a differential dissolution that causes the sculpted ornamentation to lose definition. Soot and smoke, responsible for the soiling of statues, also carry metal oxides (iron, vanadium...) that catalyze sulfur oxides, giving way to sulfuric acid which, in turn, produces sulphations in areas protected from rainwater. This transformation of calcite into gypsum (sulfation) induces an increase in volume of 83.24 % that adulterates the superficial morphology and leads to the appearance of blisters. After these detach from the surface of the stone, the whole process is ready to start again. This leads, on the one hand, to the stone pulverizing and, on the other, to the formation of blemishes that wears off the ornamental surface of the sculptures. In the case of the lnfantes, this is mainly due to processes of dissolution, and in the case of the kings to sulphation mechanisms (CaSO4. 21-120), dissolution and blisters detaching.

7. Ultrasound propagation prospecting Ultrasound propagation prospecting allowed us to determine the advanced state of decay the stone (fig. 6). The Kings statues have an average velocity of 2642~1087 m.s1. The average velocity for the Infantes is of 2408a:384 m.s"~. Standard deviation for the Kings is high, which indicates greater heterogeneity of material as a result of fissures and cracks. The statistical Figure 6: Decay of the sculpture Bishop Asterio. Lines continuous: cracking, Lines discontinuous: fissuring. Points: loss of material.


133

analysis of measurements reflects three types of differential groups that coincide with areas with cracks, fissures and decay. 14 % of readings correspond to cracks, with values bdow 1,500 m.s-1. Ultrasound propagation velocity in zones with fissures ranges between 1,500 m.s~ and 2,400m.s'l,which correspond to 30 % of readings. Decayed zones where there has been a loss of material or of material consistency represent another 30 % with ultrasound propagation velocity values ranging between 2,400 m . s "1 and 3,000 m.s"1. Of all the Kings statues the decay process is more advance~ in those portraying Fernando HI and Bishop Asterio (fig. 6). They register the lowest ultrasound propagation velocity values (528 msl). On the other hand, the sculpture portraying King Alonso reaches higher velocity, in places above 6,000 m . s "i. The Infantes have a lower average ultrasound velocity than the Kings do. Standard deviation is of 384m.s ~, which reflects a more homogeneous state of preservation. Statistical analysis does not reveal any internal cracking, although some fissures can be observed. Fissure percentage corresponds to nearly 10 ~ of readings, but 55 ~ of the readings reveal a significant decay, with ultrasound propagation velocity values below 2,300 m.s ~. Maximum velocity recorded is of 3.700 m.s ~, below values recorded for the Kings. All Infantes statues are in a similar state of preservation.

8.Conclusions All sculptures were carved on the same type of stone (Hontoria limestone). Differences between the two types of statues, Kings and Infantes can be found in the textural contrast that their petrophysical properties reveal. These differences are owed to the different quarry sections the stones were extracted from and to the different carving periods. The Kings are more water permeable than the Infantes. They are, therefore, more fragile. According to the ASTM standards, the limestone used for sculpting the Kings is of medium to low quality, whereas the limestone used for the Infantes is of low quality. The Kings are more sensitive to freeze-thaw cycles, while the Infantes are more sensitive to wetting-drying variations. The Kings have suffered more decay as a result of cracking and sulphation mechanisms, while the Infantes owe their decay mainly to processes of dissolution. Ultrasound velocity prospection indicates that the Kings statues have severe cracks and fissures, while the lnfantes have suffered more homogeneous decay as dissolution affects the whole of the sculptures. Acknowledgements We are grateful to the National Heritage Conservation Department of Junta de Castilla y Le6n for their assistance in carrying out this study. References CNR-ICR 1981, 1985,1994. Raccomandazioni NORMAL. 7/81, 21/85, 43/93. RILEM, 1980. Recommended test to measure the deterioration of stone and to assess the effectiveness of treatment methods. Comisission 25 PEM. Mat6riaux et Constructions, 13,75: 175-253. Orcajo P., 1856. Historia de la Catedral de Burgos. Facsimil 1997 Fundacion para el apoyo de la Cultura. Amigos de la Catedral.


135

DURABILITY OF SANDSTONES IN SERBIAN ANCIENT MONASTERIES AND MODERN BUILDINGS Vesna B. Matovic* Faculty of Mining and Geology, Belgrade, YU Dragan J. Milovanovic Faculty of Mining and Geology, Belgrade, YU Slobodan M. Joksimovic The Highway Institute, Belgrade, YU

Abstract The arkose sandstones of the Lower Miocene age, from the village Bele Vode and the Red Permian sandstones from the river Grza were used extensively as building stones since medieval time. These rocks were built in the Lazarica Church, in the Monastery Manasija as well as in some new buildings in Belgrade, for example in the St. Marco Church and for the riverbank pillars of Branko's bridge. The sandstone blocks display different stages and shapes of decay. The most important damage types on the Lazarica church and the Manasija monastery are honeycomb weathering and granular disintegration, while on new buildings dominated shapes are exfoliation, spalling, blistering, peeling and granular disaggregation. In this paper causes of decay of the mentioned sandstones based on the results of their petrography and physical properties, exposed to different environment conditions are determined and presented. Keywords: sandstones, decay, cause, honeycomb weathering, exfoliation.

1. Introduction Sandstones are traditional building material in Serbia. Various objects were built of it due to its colour, easy exploatation and superficial working. Sandstones marked our medieval parochial architecture of Raska and Morava shool, but also were built in facades of some important objects in Belgrade between two World War, or later. Ancient monasteries were built with different styles and a plenty of architectural elements and sculptural decorations. In the sense of their historical and artistic value, they are orginal and eternal cultural treassures of Serbian architectural heritage. Their architecture reveal the sensitivity and creative imagination of people of that period. Sandstones used for the most of churches orginated from two localities: Bele Vode (20 km NW of Krusevac) and Red Permian sandstones from the river Grza (20 km east of Paracin). Majority of ancient monasteries from 14th and 15th century were built from the former or by bricks and sandstones together. In the aim to protect the view of ancient churches, facades of some new ortodox buildings were also built from the same sandstones. Within contemporary objects sandstone were mainly used as stone panel in facades of big and important buildings.

Faculty of Mining and Geology, Djusina 7, 11000Belgrade, Yugoslavia

136


For this investigation were chosen two ancient objects (the Church Lazarica and the Monastery Manasija) situated in rural area and two new buildings (the St. Marco church and the riverbank pillars of Branko's bridge) situated in urban area (fig. 1). Shapes of decay are the most intensive and obvious on mentioned objects, that enabled correlation between the influence of environmental pollution on velocity and types of degradation.

2. History and architecture of objects 2.1. The church Lazarica The church Lazarica was built as court church from 1374. to 1377. in Krusevac (fig.l). The architectural conception of this church derived from the Morava architecture school. Its ground plan is in the form of reduced trefoil while the dome from the central part is supported by the pilasters projecting from the walls. It was built as alternate courses of sandstone ashalrs (from Bele Vode quarries) while brick are connected with courses of mortar. Their facades are divided by cordon cornices and colonnetes into zones and richly ornamented with sculptured stonework with plastic ceramic details and chequered panels (fig. 2A). Bas-releifs decoration in a form of cable moulding and palmete, are located around single and biforium windows, on archivoltes, on rosettes, arcade friezes and on the arches above the windows of the dome and on tower above the nartex (Janicijevic,1998). 2.2. The monastery Manasija It was built as well as its surrounding buildings between 1407. and 1418. about 20 km north from Paracin (fig. 1). The whole complex was encircled by massive wall with towers. The entrance is in a form of barrel arch, and the ground plan of the church is in a form of developed trefoil with central cube and four small cupolas (fig. 2B). Its facades are covered with sandstones ashlars, and have relatively little plastic decoration: only cordon cornice by arcade friezes supported by the console and small capitals of colonnetes above apses. The massive wall of monastery fortification was built of limestone blocks held by courses of mortar. Vaults and intrados of the main entrance and windows on the other buildings are of travertine while flanks and edges of towers are of sandstone achlars. The facade of the church was re-built during 70 th years of this century, when some parts of massive wall and towers were redesigned partly. 2.3. The St. Marco church The St. Marco church in Belgrade was built from 1931. to 1939., and represent the first building of Byzantine style made of ferro-concrete in our country. Roofs are also of ferroconcrete and covered with copper, while walls are of cement-mortar.


137

The church facades were made of the sandstone from Bele Vode and of Red Permian sandstones. Plinth with outflows, out of walls are facing with granitic rocks. Ground plan of the church is in the form of trefoil with belfry on the parvis (fig. 2C). Central doma is like eight limb tambour, while small cupolas are located outside of its central part. The porch of the church as well as the other part of facade are without representative sculpture's decor that is typical for parochial architecture. Triforiums on sidewise facades, are also without plastic decoration. Four castels with small cupolas, are connected by arcades.

2.4. Riverbank pillars of the Branko' s bridge The Branko's bridge is an important utilitarian object connecting the old part of the Belgrade on the right bank of the river Sava with the new part of the city, situated on its left bank (fig. 1). Riverbank pillars of the former suspension bridge (destroyed during Second World War) anchor the terminal parts of the bridge construction.The old, decorated pillars represent an extraordinary union of modem engineering and old architecture. Some of them are still used for communication, as stairways (fig. 2D). Pillars built from 1931. to 1933. are made of ferro-concrete while stone paneling were made of Bele Vode sandstones. In the aim of estetic appearance, the most of the surface blocks (upper part of facades) were made by bushhamer and in a form of rock-faced quoin (lower part of facades). They were constructed in Romano-Byzantine style adorned by numerous architectural elements and sculpture's decorations. Friezes with small blind arcades are located above three limb arcades with New-Byzantine capitals. Upper parts of facades are ornamented by cornices, rosettes and consoles (Kadijevic, 1996).

3. Analytical techniques 30 samples of fresh sandstones from the four quaries of Bele Vode and 22 samples of Red Permian sandstones from abandoned quaries of Grza were used for petrographic study. Sampling were done according to their similarity with sandstones built in objects. Number of analyzed samples depended on the type of sandstone, its position on the objects, type and degree of decay. Nine samples were taken from Lazarica from the most damaged elements all over facades;, two samples from arch of windows, two samples from lower part of southem facade and two samples were taken from the main entrance the monastery Manasija; two samples were taken from rossetes of the riverbank pillars of Branko" bridge; two samples from lowest part of western facade were taken from the St. Marco church.

138


From the sandstone surfaces with well developed efflorescence were taken samples for identification salts using the method of X-ray diffraction.

4. Petrography and physical properties of sandstones Sandstones of Bele Vode" The Lower Miocene sandstones are exposed in Krusevac area (vilagge Kukljina and Bele Vode) and represent shallow lacustrine sediments. They cover an area of aproximately 5 km 2. They are transgressive to crystalline schists and their boundaries with neighbouring geological units are tectonic. Some blocks slide downward within the sandstone mass along dm-m faults (Petkovic et al., 1973). The lower part of sandstone is massive while the upper is layered and banked. The thickness of productive horizon is from 3.5 m to 7 m. Sandstones are gray to yellow-brown, sometimes pink or red coloured. They are massive to well bedded, often with cavities (aproximately 0.5 mm in diametar) and coarse- to finegrained (grain size vary from 0.2 to 1 ram). Sandstones are composed of partly rounded to angular slightly sorted clasts of quartz (50-60 vol. %), K-feldspar (orthoclase and microcline, to 15 % vol.), plagioclase (to 5 vol. %) muscovite (up to 10 vol. %) and fragments of rocks, mainly metamorphic (to 10 vol. %). Contact-pore filling of yellow and red varieties of sandstones is ferrous-calcitic cement, while in gray varieties (subordinated) is siliceous-clay. The cement content is about 10 vol. %. Correlation among physical and mineralogic-petrographical characteristics is the base for determination the decay causes. The obtained data for bulk density (1880-2420 kg/m 3) and density (1940-2720 kg/m3), absolute porosity (7.4-27.9 %) and water absorption (2.43-9.16 %) considered sandstones of Bele Vode as medium heavy, high porous rocks with moderate to high water absorption. Red Permian sandstones are generated in lacustrine basins and river vallies by fast erosion with simultaneously transgression. They built horizon of 50-700 m thick (well bedded to banked with rare lamination and cross-bedings) over thrusted Cretaceous limestones. Permian sandstones are of characteristic red colour with occasional occurences of colorless and gray-colored parts. Within the sandstones interbeds of conglomerates, coarse-grained arkoses, siltstone,clay, rarely dolomite and limestone occur (Veselinovic et al., 1964). Red Permian sandstones are mainly arkoses, rarely quartz sandstones. They are composed of subrounded to angular, slightly sorted clasts of quartz (50-70 vol. %), feldspar (orthoclase and plagioclase, up to 25 vol. %). Micas (biotite and muscovite) made about 2 % of rock. Fragments of rocks are determined as: volcanic rocks, cherts, quartzites and schists (5 vol. %). Cement in red sandstones occur (ferruginous and carbonaceous or siliceous matter) as thiny coating on detritial grains or as pore filling. Physical properties suggest on uniform quality of Red Permian sandstones. They are heavy (2630-2692 kg/m3), extremely to moderate porous (4.9-11.2 %) rocks with slightly to moderate water absorption (1.53-3.06 %). 5. Enviromental conditions The velocity of sandstone decay is influenced by climate and atmospheric pollution. Ancient buildings in the area of Krusevac are exposed to more sevear climate than those in Belgrade (tab. 1).


139

Table 1. Summary of climatic data from 1961 to 1990.

Climatic data Yearly average temperature average maximume (monthly) average minimum (monthly) Frost days per year (min T< 0~ Ice days per year (max T< 0~ Month with more frost days Number of sunny days

Yearly average precipitation The rainiest month The driest month

Relative humidity -yearly average Average of maximume (monthly) Average of minimume (monthly)

Belgrade

Krusevac

11.9~ July (21.7~ January (0.4~ 61.5 17.6 January (7.9) 2047.6 h 684.5 mm June (90.4 mm) October (40.3 mm) 69.0% December (79.8%) April (61.5 %)

10.8~ July (20.6~ January (-0.3~ 92.3 17.6 January (8.2) 1790.4 h 650.4 m June (86.0 mm) October (38.9) 76.8 % December (86.0%) April (70.7 %)

The climatic data of the Meteorology National Institute

Generally, all buildings are exposed to the medium continental climate with wet and hot summers, and a long cold, winters, often with fogs and snow. Average yearly precipitation, number of the frost-icy days and the value of high relative humidity during winter, undoubtely accelerate the degradation of sandstones. The Manasija monastery and the Lazarica church are located in rural area while riverbank pillars of the Branko's bridge and the St. Marco church are located in urbane area with intensive traffic, especially trucks (tab. 2). Table 2.Pollutant concentrations data for the period 1991 to 1999. Concentrations (~g/m 3) SO2 Yearly average 25.6 Max. yearly average 174 Number of days over limit 3 Sanitary limit for: 802- 150 (~tg/m3); smoke- 50 (~g/m 3)

Smoke 39.2 280.7 74

Yearly average poluttant concentrations are relatively low but maximum yearly average for individual years (1991, 1992) were appreciably over limit in Belgrade,. The main part of SO2 in the air changes pH of rain. These acidic rains attack on building facades. During last ten years, contents of SO2 and smoke show descending trend even their lower content are also pernicious for stone. 6. Types of decay Observing of the sandstone elements enabled identification of the different destruction contours: granular disintegration, contour scalling, honeycomb weathering, exfoliation, spaling, efflorescences and black crust. The most important damage types on the Lazarica church and some part of the Manasija monastery are honeycomb weathering, granular disaggregation along with contour scalling,

140


while on the St. Marco church and on the Riverbank pillars dominated shapes followed by granular disintegration are :exfoliation, spaling, flaking, and blistering. 6.1. The Lazarica church Honeycomb weathering appears on elements of the eastward side of facade within the zone of ground water capillary rise and on the elements directly exsposed to rain, blowing wind and sun, in a shape of irregular cellular depressions or as alveoles to 20cm in size (fig. 3A). Granular disaggregation occurs on the most of sandstone blocks. The scale and depth of decay are influenced by water, sun and moisture. On southern and eastern facades it occupied shallow zone (0.5-2cm in depth) of elements that are overlaping on plinth with projecting part and on external part of cordon cornices also, while contour scalling were noted on sandstone block of church portal on it's western fasade. After destroying of an outer layer (about 2-5cm thick) independently from the bedding , newopened surface consisting of sand, spalls and flakes are more sensitive for further decay. On the western and northern facades, granular disaggregation along with contour scaling produced the crumbling zones (to 10cm in depth) leading to important losses of volume,and to a rounding of comers and edges (fig. 3B). These blocks have scars and sanding surface with pieces that have a tendency to fall easy due to weak cohesion between grains. Becouse of that some part of cordon cornices already have lost its architectural shape (fig. 3C). Generally, granular disaggregation and contour scalling take place in the areas where rain and snow can be collected during winter time. 6.2. The Manasija monastery Honeycomb weathering and granular disintegration developed on the main entrance of the Manasija monastery. Honeycomb weathering, associated with granular disintegration, occurs on sandstone blocks in flanks of arc, and also in intrados of portal, on sheltered blocks exposed to moisture (from the wall above) and fast drying by current air. Differential weathering with forming alveoles (size of 5cm) is controlled by bedding of sandstones (fig. 3D,E). Granular disintegration with rounding, occurs on sandstone blocks directly exposed to rain and sun. This type of damage develope along mortar joints where water can easy move towards outer edges and towards the inner parts of block. Deep granular disintegration is evident on the sandstone block in towers of monastery wall. These elements are directly exposed to rain, sun, wind erosion and their surface are eroded up to 15 cm in depth. The church facade in monastery complex was re-built during 70 th years thus the sandstones block are relatively sound. On its northern facade, efflorescence occurs along mortar joints (fig. 3F). Unfavourable effect of efflorescences is aesthetic and for now it hasn't destroying influence on sandstons blocks. The X-ray difraction data of salt samples from surface sandstones, show the presence of gipsum, thenardite and calcite. 6.3. The St. Marco church The sandstone elements from the church facade were observed and established on deep destruction in some sandstone elements especially on its southern and on the eastward side. According to height of built elements deep destruction appears on those located in the lower part of facade.


141

Figure 3. Decay types: Lazadca church- A- honeycomb weathering; B,C- contour scaling. erosion and granular disaggregation; Monastery Manasija - D, E- honeycomb weathering; F- efflorescences; St. Marco church - G-granular disintegration; H-chipping and spalling; I- exfoliation; Riverbank pillars of Bmnko's bridge - j_ erosion and granular disintegration K- exfoliation, L- contour scaling and granular disintegration.

142


The first bed from red sandstones is over laping on plinth with projecting part (thickness about 15cm), where during winter time and rain months rain and snow are collecting causing the deep granular disintegration (up to 10cm in depth, fig. 3G). Peeling develop on shaltered surface sandstones of the first bed (in arcades), as superficial loosening of 0.5-3mm thick sheets which tend to blister and fall off. New opened surface underwent chipping and granular desintegration besides clear prominence of more durable cement-mortar ofjoints (fig. 3H). Dominant shape of decay on the southern and on the eastward side of facade is exfoliation. The primary surface of the most elements was completely destroyed and sheets are separating from the stone element perpendicular to layering. Hard surface crusts are forming in the upper parts of facades, especially in the southern side beside triforiums, and after breaking of hardened surface the sheets (up to 3 mm thick) separate and exfoliation and granular disintegration continue (fig. 3 I). Efflorescence occurs on the sandstone blocks in arcade's portch and it is composed of mirabilite and thenardite. ,

6.4. The riverbank pillars of the Branko's bridge The riverbank pillars are heterogeneous structure. The sandstone blocks, depending from type of sandstones, kind of superficial working, location, and distance from the river, microclimate factors (temperature, wind, sun, rain), atmospheric pollution, show different type of degradation: exfoliation, case hardening, granular disintegration, contour scaling, spalling and peeling. Besides efflorescence, black crust occurs too (Matovic, 1999). Deep erosion and granular disintegration occupied all sandstone blocks in lower part of the facades. Blowing wind, influence of rain, ground water etc. caused physical and chemical decomposition, why some elements lost superficial shapes, rock-faced quion, with rarely preserved primary parts (fig. 3J). Deep granular diisintegration, contour scaling, peeling caused that the most of architectural elements lost their own shape (rossete, capitals and base of colonnas). Black crust occur on sheltered sandstone blocks, untouchable for rain. Practicles of dust, smoke and other atmospheric pollution are easy deposited on blocks with rock-faced quion surface. During dry months, in arcade's porches efflorescence of thenardite and thermonatrite occur in the upper zone of ground moisture rising. It acceleratied peeling, chipping and later exfoliation of sandstone blocks.

7. Conclusion The Bele Vode sandstones were used for the most of ancient churches while in the last 80 th years it was combined with Red Permian sandstones as stone panel for facades in Belgrade. Sandstones are strongly disintegrated on all examined objects. Type and velocity of sandstone decay depend of: mineral composition, texture, physical properties, atmospheric influence (including water and temperature changes as the most important), atmospheric pollution, their position in objects (exposed to sun and wind, or not), type of mortar joints, as well as of way of superficial working. Used sandstones are medium to coarse-grained arkoses with contact and pore filling cement (carbonaceous, ferruginous or clay-siliceous). According to porosity and water absorption, i.e. the most important properties for durability, they are high porous, able to absorb and keep water.


143

Even almost all of surfaces of the original sandstone elements are completely destroyed the difference between intensity and morphology of sandstone decay in urbane or rural places is evident. On the ancient objects dominant types of decay present almost 600 years are honeycomb weathering and contour scalling followed by granular disintagration. The shape and intensity of decay imply on chemical-physical influence of water and frost as the main causes. Sandstone porosity enable penetration of rain, as well as of ground water. Majority of secondary calcite in sandstone surfaces resulted from the atmospheric influence or from lime-mortar, as well as from cement material within sandstone. Dissolved material were unequally deposited parallelly to evaporate surfaces or parts with faster evaporation, during sunny or windy period (Winkler, 1994) The sandstone structure prevailing lamination and more intensive diferential dissolution, enable accelerated honeycomb weathering. Granular disintegration and contour scalling are evident in the lower part of facades, i.e. on elements permanently exposed to moisture (ground water), as well as on elements where rain and snow could be longer kept (cordon cornices, rossetes, window frame). Chemical dissolution as well as mechanical disintegration of sandstones are accelerated and more intensive in rural than in urbane areas, due to climate, especially due to the number of frosty and icy days ( cyclic freezing-thawing) and to the amount of precipitation. On new objects sandstones have been deeply destroyed after 60 years by exfoliation, case hardening, spaling followed by granular disintegration. The main cause of these decay types is physical activity of water and frost, that enables further chemical dissolution As sandstones are high porous rocks, water absorbed to a certain depth enable freezing and thawing of the outer oarts of block. Repeated cycles, even daily formed weak surface zones. At the beginning it is hardened crust that soon loose contact with the sound parts of rock, thus its blistering and breaking, perpendicular to layering caused exfoliation and spaling. The influence of water activity and of position in objects on decay degree is especially notable on riverbank pillars on the Branko's bridge, whose base parts are next to river, and thus permanently water saturated. Due to weak current air all elements within the zone of capillary rise present in arcades sufered scalling and grain disintegration. Same or similar situation included elements of the upper fagade, that are built just under steel bridge construction. They are ptotected from the rain influence, but due to bad sluice system from the bridge floor, water and mud continualy flow over their surfaces, producing intensive erosion of the sandstone surfaces. The influence of the building type and superficial working on type and velocity od decay is visible on the same object. Perpendicular to layering, exfoliation and scalling occur on all surfaces, whereas fractures could be seen on blocks cut normal to layering. Bushhammering enable degradation of sandstone elements and produce an mechanical, degraded layer with high porosity and sensitivity on water and frost, as well as microfractures that enable intensive penetration of water in surface parts of sandstone, causing weakly cohesion and stronger influence of frost. Sandstone blocks with rock face of the riverbank pillars, sheltered from the rain influence, are covered with black crust due to deposition of smoke and dust from traffic, common in an urbane area. There are no traces of granular disintegration or other types of alteration beneath these black surfaces. Efflorescence, the common occurrence on objects in urbane area, developed during dry periods on surfaces that are sheltered from the rain, i.e. inside or over the zone of capillary rise, where following salt could be deposited: halite, thenardite/mirabilite (Na2SO4 with, or without of water). Majority of sodium originated from soil, but some parts could be from

144


streets salted during winter, whereas sulfate ($04) come from atmosphere. Over the zone of capillary rise, the main salt is gypsum and calcium for it is from mortar or sandstone. Efflorescence usually follows and accelerates exfoliation. Mentioned processes of sandstone decay can't be interrupted or stopped, but it is necessary to slow down them by properly sanation. That will enable many valuable monuments of our cultural heritage to be saved and last for further generation. References

Article reference: Kadijevic A., 1996. istorija i arhitektura zamunskog mosta Kralja Aleksandra I Karadjordjevica, Pinus, Zapisi 4, 4-10. Matovic V., 1999. Decomposition of stone in the riverbank pillars of Branko's bridge. Heritage, Insitute for the protection of cultural monuments of Belgrade, II,107-114. Petkovic K., Vukasinovic M., Smiljanic R., 1973. Krusevacki zemljotres 1. oktobra 1972. i geolosko seizmotehnicke karakteristike terena krusevackog tercijarnog basena i njegovog oboda. Geol. an. Balk. polu., knj. XXXVIII, Beograd, 371-467. Veselinovic M., Antonijevic I., Milosakovic R., Micic I., Krstic B., Ciculic M., Divljan M., Maslarevic Lj.,1970. Tumac za list Boljevac 1:100000, Savezni geoloski zavod, Beograd 1-73. Book references: Janicijevic J., 1998. Kulturna riznica Srbije, IDEA, Beograd, 321-323. Winkler E. M., 1994: Stone i arcitecture, Properties, Durability. Springer-Verlag, New York.

145

SANDSTONE ARCHITECTURAL DETERIORATION IN PETRA, JORDAN Thomas R. Paradise, Ph.D. Associate Professor of Geography & Environmental Studies University of Hawaii, Hilo, Hawaii USA 96720 email: [email protected]

Abstract Petra, Jordan was an extensively occupied city during the Nabataean and Roman eras (300 BCE to 500 CE). Incorporating Hellenistic, Assyrian, Egyptian and Roman architectural motifs and techniques, the Nabataeans created an architecture style that was both derivative and unique. However stable and 'eternal' is this spectacular setting, its architecture is deteriorating from natural and induced influences. Comprehensive networks of surface baseline measurements were made in Al-Khazneh; Petra's most celebrated structure, the Roman Theater, and the Anjur Quarry. Since this architecture was hewn, it is an ideal environmental laboratory for the study of weathering features, causes and rates they have not been moved, altered or obscured since their construction 2,000 years ago, and their lithologic composition is relatively consistent and well-documented. This study investigated intrinsic (lithology) and extrinsic (climate, sunlight, human impact) influences on sandstone deterioration. It was found that iron and silica constituents in the sandstone matrix decreased overall weatherability, while calcium constituents increased deterioration in areas that exceed 5500 megajoules/meter/year (a typical southern aspect in arid regions). In fact, when iron matrix concentrations exceeded 4% (by weight), original stonemason dressing marks were still clearly evident, indicating a nearly unweathered state in 2,000 years. Insolation (sunlight) was found to have its greatest deteriorating effect on southwestern and southeastern aspects, indicating that sunlight is most effective in stone weathering when in tandem with increased wetting/drying and/or heating/cooling cycles. Visitors to Petra have increased from 100,000 (1990) to 350,000 (1998), and anthrogenic deterioration is accelerating. Tourist groups entering the chamber of Al-Khazneh caused interior relative humidity to increase from 20~ to 50%. While, surface recession from visitor touching, leaning and rubbing has receded as much as 40mm in less than 10 years. A 4 by 3 meter wall area has lost sandstone volume of approximately one half meter of sandstone in 100 years from 0.5 to 2m above the floor indicating surface recession from human contact. Keywords: sandstone deterioration, Petra, Nabataean Architecture, weathering 1. Introduction The ruined city of Petra lies hidden in a deep valley surrounded by steep, impassable sandstone walls and winding, earthquake-defined gorges in the arid expanse of Jordan's great southern desert. However, it is the spectacular architecture rather than its beautiful setting that has drawn international attention and wonder. Although its structures and archaeological evidence indicate occupation in the Petra area since 7,000 BCE, it was its Nabataean occupants and Roman neighbors that gave Petra notoriety then and now. These * Author to whom correspondence should be addressed.

146


residents worked the valley walls into simple and elaborately carved tombs and structures, sculpted directly into the reddish-brown and yellowish sandstone cliffs, many exceeding fiity meters in height. Since then, however unique and precise their construction, natural and unnatural influences have been effective in deteriorating this architecture. Prior weathering studies have separated weathering influences into two distinctive categories: those affected by the characteristics of the stone itself or intrinsic effects (i.e. lithologic constituents, fractures), and those affected by external influences or extrinsic effects (i.e. climate, human contact). The decay of Petra's sandstone architecture can be similarly identified as those surface features related to variability in rock composition and/or caused by running water, human touch, etc. Previous research has emphasized the importance of intrinsic agents like rock composition and integrity, but recent research indicates that extrinsic influences like climate and human contact (tourism) are also important. This report addresses recent studies in Petra, Jordan that are attempting to answer these questions of sandstone weathering causes, effects and rates and possible solutions they might yield. Weathering studies for sandstone architecture in arid environments are relatively rare. Early observations on stone and architectural deterioration in the Near East and their oiten unusual features were made by Herodotus (c.450BCE), Strabo (c.CE10), Pliny (c.CE50), J.L. Stephens (c. 1830) and R.F. Burton (c. 1850), however it isn't until the 20th Century that we begin to see the conceptual development of weathering studies (Paradise 1995). Bryan (1922, 1928) and Blackwelder (1929) discussed many of the processes responsible for weathering. These were some of the first Western works that addressed sandstone weathering processes and not just the descriptions of weathering features (i.e tafoni). Later research in arid regions established important relationships between weathering and various influences including lichen overgrowth (i.e. Jackson & Keller 1970, Jones et al. 1980, Paradise 1997), case hardening (i.e. Conca & Rossman 1982), permeability (i.e Pfluger 1995), tafoni development (i.e. Mustoe 1983), salt (i.e. Smith & McGreevy 1988, Young 1987), insolation and moisture availability (i.e. Robinson & Williams 1992, Paradise 1995, 1998). These studies indicate that sandstone weathers two ways. Since sandstone is made of sandy clasts in a binding matrix, either the clast fractures or dissolves to fall out, or the matrix fractures or dissolves to release the clast. These weathering types represent the processes of disaggregation that produce loose sand as the by-product of deterioration - - the source of many of the sand dunes throughout the Near East.

2. Methodology Historical architecture and monuments represent under-utilized resources for the investigation into weathering rates and influences. Limestone and marble tombstones have been used for weathering analysis since Geikie's work (1880) in Scottish cemeteries. Marble and limestone weathering research using structures and objects has been studied by Rahn (1971), Hoke (1978), Meierding (1981), Klein (1984), Dragovich (1986), Neil (1989), and Trudgill (1989). Granite weathering and architecture was studied by Winkler (1965), Amoroso and Fassina (1983), and Smith and Magee (1990). While other rock types used in construction such as slate, arkose (Matthias 1967) and calcarenite (Spencer 1981) have also been studied and discussed. These studies used similar techniques whereby the known relative or absolute date of the


147

structure (or objects) enabled the researcher to determine the change due to weathering since its creation or construction. Basically, it is this difference between the original and contemporary surfaces that explain the causes and relationships of the influences affecting stone surfaces. The difference between the current weathered surface and the original Roman surface represents how much sandstone has receded from weathering and erosion over 2,000 years. Since many of these previous works were conducted in Europe, Australia, and the U.S. friable sandstones are not as frequently used as the more resistant and accessible stone building materials such as granite, limestone or marble. This is why sandstone is an often under-studied building material, potentially valuable for weathering studies in arid regions where its use is common because its intrinsic weatherability is decreased due to reduced precipitation which decreases overall weathering rates (Paradise 1995). 3. The Roman Theater: weathering and lithology An extensive study was conducted on the Roman Theater of Petra (Paradise 1999, 1995a, b), a huge arena that may have seated up to 10,000 persons. The sandstone Theater was carved during the first century CE according to the recommendations of the great Roman engineer, Marcus Pollio Vitruvius m the highest Roman engineering and construction standards of the period. These early requirements for theater and building construction were so standardized that the level of the original surfaces can be estimated from contemporary weathered surfaces. Five hundred locations were examined in the Theater for intrinsic influences like variations in rock composition and particle size; and extrinsic effects like sunlight angle, lichen coverage, slope, and surface temperature. Important relationships were found. First, it was discovered that sandstone deterioration is accelerated by sunlight (where temperatures exceed 50~ and the sandstone contains high amounts of calcium carbonate: CaCO3), and when the sandstone sand grains are spaced far apart (high matrix-to-clast ratio). These are often observable influences since calcium-rich and/or high-matrix sandstones are often softer and more friable than the darker-colored, iron-rich ones. Second, other relationships were found where natural effects decreased weathering. It was discovered that iron oxides (Fe203, FeO2) and silica (SiO2) in the sandstone matrix (the binding agent that holds the sand grains together in the rock) dramatically decreased weathering. These findings were so sound that original Roman stonemason tool marks were obvious atop iron-rich sandstone while large weathered cavities had developed on the rock areas that were iron-poor. In fact, statistical analysis (r~) found that the iron and silica constituency of the sandstone matrix explained for 50% of all deterioration in the Roman Theater.

Statistical Correlations (50%; +++: 50-30%; ++: 30-10%; +: 10-3%; tr: traces; -: not detected). Roan Figueiras , ~Axeitos . . . . Barbadelo Muros 6P 4.33 1.96 2.90 2.05 3.50 Type of fissures XRD Q F M K V

Inter and insular

Inter, intra and transgranular

Intra and transgranular

inter,'intra and ,transgranular

+4+ -r

+++

++ +++

+++

tr

tr

tr

tr

tr

tr

tr

,~,~| t~t Tq

tr tr

|

9

Inter, iTntraand transsrmmlar +++

Figure 1 9Orientation of transgranular fissures relative to the reference plane (bedding or horizontal plane of the quarry), as observed by fluorescence microscopy in Roan, Figueiras, Axeitos, Muros and Barbadelo granites. Salt crystallization tests The conditions of the series of salt crystallization experiments, including the salt solutions used, test sample sizes, drying chamber conditions, etc. are outlined in Table 2. In the first two experiments 14% (w/w) sodium sulphate was used and this solution was absorbed by the test prims by capillary action. The granite test prisms were orientated in three different positions during absorption of the solution, in order to determine whether the direction of absorption of the salt solution relative to a reference plane (the horizontal plane of the quarry, represented by the shaded area in the diagrams in Table 2) influenced the durability of the granite. Five samples each of R, AX and F were tested in each position in both

Table 2: Details of the salt crystallization tests. The shaded face of the granite blocks corresponds to the horizontal plane of the quarry and A,C and D the orientation of the samples during; the tests. i Experiment

1

Solution

14% Na2SO4

Granites

Test sample

Axeitos Figueiras Roan

Drying Chamber conditions

Time

Temperature of drying

Humidity of drying

7h 5h

20~ 60~

80%RH 40%RH

Analyses

Study of thin sections by fluorescence Microscopy

=r

g~ o

c) o

2

14% Na2SO4 2h partial immersion

Axeitos, Figueiras Roan

4% NaCI 2h total immersion

Figueiras

4


Figueiras


Figueiras

1' sea-salt spray

60~ 20~

40%RH 80%RH

Determination of mass and open porosity after every 10 cycles.

i'D

3


16h 16h

o~ i'D r~ r~ o

Roan

Roan

5x5x5cm

20h

20~

80%RH


o

5x5x5cm

10h 10h

60~ 20~

40%RH 80%RH


o
5.0 km/s 3.0 km/s - 5.0 km/s 2.0 km/s- 3.0 km/s 1.5 km/s- 2.0 km/s < 1.5 km/s

fresh increasingly porous sugar-like disintegration fragile crumbling rock

< 0.5 1.3 - 0.5 3.0- 1.3 5.3 - 3.0 > 5%.

This study evaluates the relationship between ultrasonic velocity, porosity and anisotropy of marbles using laboratory measurements. The aim is to gain a more


217

comprehensive method for rock diagnosis. Laboratory measurements are performed under better defined conditions than field measurements on e.g. sculptures and, thus, the interpretation of data is more confident. At first, the effect of anisotropy on ultrasonic marble diagnosis will be investigated. The elastic properties of a rock are controlled by a number of physical (temperature, pressures, pore pressure, etc.) and lithological (chemical composition, fabric, pore space) parameters (Duerrast et al., 1999). A directional dependence (anisotropy) of elastic wave velocities is a frequently observed characteristic of almost all sedimentary, magmatic and metamorphic rocks (see compilation in Siegesmund, 1996). It is caused by a lattice preferred orientation (here referred to as texture) of anisotropic minerals and can be altered, under low confining pressures, by microcracks (e.g. Siegesmund, 1996). In particular calcite shows an extreme anisotropy of petrophysical properties (e.g. compressional wave velocities, thermal dilatation coefficient c~; Fig. lb) Marbles in the open country can be more or less water saturated. Thus, we investigate the effect of the two different pore fluids air and water on elastic wave velocities. Finally, petrophysical data and theoretical models are combined to characterize the pore space which leads to the velocity decrease during weathering. The results of the ultrasonic investigations are correlated with microstructural characteristics of the respective marble type to gain more suitable and comprehensive parameters for the quality assessment of a marble.

2. Experimental methods For the ultrasonsonic measurements spherical rock samples with a diameter of 50mm and an accuracy of 0.02ram have been prepared. Spherical samples allow measurements in all spatial directions, i.e. a complete determination of anisotropy. We measured transient times of ultrasonic pulses (piezoceramic transducers, resonant frequency 1 MHz) in 90 directions using the pulse transmission technique (Birch 1960, 1961). The measurements were performed at dry and completely water-saturated samples to simulate conditions found in the field. For measurements under pressure cylindrical samples (30o30mm)have been used. All specimens were oriented according to macroscopically visible fabric elements (e.g. metamorphic foliation). The porosity of the marbles was determined by buoyancy weighting. Therefore, the spheres for the ultrasonic measurements have been used. For two strongly weathered marbles the pore space distribution has been determined by mercury porosimetry. Therefore, samples were mounted in a vessel filled with mercury which allows a simultaneous application of hydrostatic pressure. The amount of mercury incorporated in the pore space is measured as a function of pressure. The basic principle is that a higher pressure is required to press the mercury in narrow pores than in large pores. The amount of mercury at a given pressure can be inverted in a volumetric number characterizing pores with a certain dimension. Fabric analyses were carried out on standard thin sections. Quantitative fabric properties (grain size distribution, grain shape, grain shape anisotropy, etc.) were obtained using digital image analysis. High resolution fractographic images from controlled cracked specimens were investigated with a SEM to characterize the mechanical strength in the initial condition of the samples.

218


3. Sample description Marbles from different localities in Italy (Carrara, two samples; Lasa, three samples) and Poland (Kauffung, Prieborn, Grosskunzendorf) have been used for the investigations. Their fabric properties are shown in detail in Weiss et al. (1999, except Lasa marble) and are summarized in Table 1. Table 1 Summarized fabric properties of the investigated marbles (for explanations see text) Marble type sample grain size microstructure composition condition [gin] Lasa marble I flesh ~-400pm Type I/II calcite Lasa marble II flesh -400pm Type II calcite fresh Lasa marble III -400gm Type II calcite Carrara marble I flesh -~95gm Type Ill calcite Carrara marble II flesh -~130pm Type Ill calcite Carrara marble III flesh -150pm Type I calcite Carrara marble IV weathered ---140pm Type I calcite Prieborn marble I weathered -150~tm Type I calcite Kauffung marble I flesh -80~tm Type III calcite/dolomite Kauffung marble II weathered -80pm Type III calcite/dolomite Kauffung marble III weathered -80pm Type Ill calcite/dolomite Kauffung marble IV weathered ---80pm Type Ill calc ite/d o lom ite Grosskunzendorf flesh -1500gm Type IV calcite marble I Grosskunzendorf weathered -~1500pm Type IV calcite marble II Some of the specimens were from the quarry (fresh samples), others are weathered and have been replaced in the buildings. The different marbles cover a broad range of textural and microfabric types with different states of preservation (Weiss et al., 1999). All of the marbles with the exception of the Kauffung marbles are more or less pure calcitic marbles. The Kauffung marble has dolomitic veins predominantly parallel to the macroscopically visible metamorphic foliation (Weiss et al., 1999). Some of the marbles show an equilibrated microstructure with straight grain boundaries comparable to that of Carrara Statuario marble (microstructural Type I). Others show serrated grain boundaries (Type II) or a bimodal grain size distribution with small recrystallized and large relic grains with irregular grain boundaries (Type III) or a microstructure with evidence for grain boundary migration (Type IV). Of course, the classification into four fabric types requires some simplification of the naturally very heterogeneous rock fabric. Thus, the prevailing presence of one of the above mentioned fabric characteristics does not necessarily exclude the others.

4. Results A deteriorated rock shows a certain amount of micro- and macrocracks as a consequence of weathering. In particular open microcracks decrase ultrasonic velocities (see compilation in Siegesmund, 1996; Siegesmund et al., 1999). The effect open cracks can be best demonstrated using ultrasonic measurements under increasing confining pressure. With increasing pressure microcracks are closed and the velocities increase strongly (Fig. 2).


219

At high confining pressure, the elastic properties of a rock are close to those of a noncracked rock (Fig. 2). The increase in velocity as a consequence of confining pressure is very similar to that of a water saturation. The interaction between porosity, pore fluid, velocity and anisotropy is exemplary shown for weathered Kauffung IV (Fig. 3) and Prieborn marble (Fig. 2,3).

Figure 2: Relationship between pressure and elastic wave velocities of Prieborn marble. Notice the strong increase in ultrasonic velocities (-2.6km/s) with increasing pressure. Onsite, a uniaxial load e.g. on columns) can produce the same effect in certain sample directions.

Figure 3: Directional dependence of Vp for Kauffung (a-c) and Prieborn marble. The velocities of the water-saturated (a,d) and dry samples (b,e) are given as well as their differences (c,f) in an equal area projection, lower hemisphere (isolines in kin/s). Maximum and minimum Vp are given in the bottom at the right and left side and the corresponding anisotropy (Avp=(Vpmax-Vpmin)/Vpmax 100) is given at the top of the polefigures.

220


These marbles exhibit significant differences in fabric and petrophysical properties (c.f. Table 2, 3). The Kauffung marble has a small grain size, irregular grain boundaries and a low porosity. In contrast, the Priborn marble exhibits a significantly larger grain size, an equilibrated microstructure and a larger porosity. Accordingly, fracture patterns determined with the scanning electron microscope (SEM) of the two marbles are completely different. The Kauffung marble shows mostly intragranular (cleavage) cracks while the Prieborn marble is characterized by intergranular (grain boundary) cracks (Weiss et al., 1999b). The on-site damage scenario for Kauffung marble is characterized by break-outs along preexisting fractures, the prevailing part of the rock remains intact. In contrast, the Prieborn marble shows a penetrative degradation very similar to that proposed for some equilibrated types of Carrara marble. Table 3 Petrophysical properties of the investigated marbles (for explanations see text).

Lasa marble I Lasa marble II Lasa marble III Carrara marble I Carrara marble II Carrara marble III Carrara marble IV Prieborn marble Kauffung marble I Kauffung marble II Kauffung marble III Kauffung marble IV Grosskunzendorf marble I Grosskunzendorf marble II

0.43 0.37 0.38 0.20 0.14 0.44 0.94 0.77 0.23 0.32 0.26 0.27 0.31

5.39 6.42 6.53 3.45 1.54 3.03 6.58 6.34 6.97 6.72 5.05

4.61 6.16 6.41 2.67 1.43 2.11 5.43 5.04 5.99 5.84 4.37

14.62 4.13 6.26 22.5 7.39 30.12 17.78 20.44 14.11 13.05 13.59

6.50 6.76 6.83 6.14 5.55 5.65 7.17 6.97 6.95 7.00 6.83

6.09 6.62 6.61 5.94 5.35 5.31 6.31 6.19 6.08 6.19 6.48

6.35 2.12 3.26 3.46 3.53 6.11 12.04 11.22 12.57 12.57 5.18

0.46

4.59

3.63

20.81

6.42

6.10

5.09

4.1 Velocity reduction due to structural disintegration The damage type of the respective marble can be clearly identified in the Vp polefigures. The weathered marble from Kauffung shows relatively high Vp in the range of 6.0-7.0 km/s under dry sample conditions. In contrast, the velocities for Prieborn marble are significantly lower (Fig. 3b,e). Accordingly, the Kauffung marble would correspond to the damage Class 0 and the Prieborn marble to Class II from Koehler (1991). Both marbles are from the same building with almost the same time in built-in condition. However, it should be kept in mind that the effect of break-outs can not be considered in the laboratory scale and, therefore, a detailed on-site mapping of damage structures is indispensable for a comprehensive expertise. 4.2 Effect of anisotropy Both samples exhibit a pronounced anisotropy which approaches even 30% under dry conditions (Prieborn marble, Fig. 3b). Velocities measured on a column of Prieborn marble can range between 2.2km/s and 3.0km/s (Fig. 3e). Notice that almost the entire Class II (Vp from 2.0km/s to 3.0kin/s) from


221

the damage classification of K6hler (1991) is covered by only one dry and penetratively weathered marble in different directions.

Figure 4: Average compressional wave velocities for different marble samples as a function of porosity and sample condition: a) dry samples, b) water-saturated samples. The anisotropy of the respective marble is given as error bars. The velocity/porosity-curve of Koehler (1991; bold line) and theoretical predictions according to the models of O'Connel & Budiansky (1974; hatched lines) are added for comparison. For the latter model calculations, the aspect ratio of the cracks assumed in the computations are given. 4.3 Effect of fluid saturation

Marbles can be very sensitive on the type of pore filling medium (i.e. water or air). The difference of compressional wave velocities in the water-saturated (Fig. 3a) and dry (Fig. 3b) condition of weathered Kauffung IV marble is remarkably low (less than lkm/s, Fig. 3c). The corresponding porosity is about 0.27%. In contrast, the Prieborn marble shows very low velocities (-2.6km/s) at dry condition (Fig. 3e). The corresponding porosity is about 0.77% (Tab. 2). At a water-saturated condition (Fig. 3d) velocities increase up to 5.5km/s. The large differential velocities between the water-saturated and dry sample condition of about 2.9km/s (Fig. 3f) account for a strong degradation of this marble. 4.4 Vp as a function of porosity?

The velocity data compiled for all the marbles is shown in Fig. 4 and Tab. 2. At dry sample conditions (Fig. 4a), the porosity of the samples never exceeds 1Vol%. However, the compressional wave velocities cover a broad range from 6.5km/s down to less than l km/s. Thus, the pore space is small but very efficient in reducing the velocities. Note the very different ultrasonic velocities for the Carrara II and Carrara III marble at dry conditions considering that both marbles are fresh from the quarry. All of the marbles exhibit water-saturated velocities within the range of 5.5 to 6.9km/s. The Vpmaxof the Carrara IV marble (weathered) is for example close to the Vpminof the Carrara III marble (fresh) and, thus, these 0.4km/s are within the accuracy of the measurement. However, at dry sample conditions the Vp of the respective marbles differs remarkably.

222


4.5 Theoretical prediction of the Vp/porosity relation In order to find constraints for the magnitude and shape of this pore space we modeled the velocity reduction as a function of crack geometry using the well known theoretical prediction of O'Connel & Budiansky (1974). The basic principle is that a given porosity is formed by certain types of ellipsoidal cracks. The crack geometry is only defined by the aspect ratio (i.e. the ratio between small and large axis of the ellipsoid) of the cracks. Spherical cracks have an aspect ratio of about 1 flat cracks of smaller than 1. The model calculations reveal that strongly decreasing Vp as a function of porosity, as it is observed experimentally, can only be caused by extremely flat cracks with an aspect ratio of about 0.005 (see Fig. 4). 5. Diskussion and conclusions A clear interpretation of Vp in terms of rock degradation deserves a comprehensive knowledge of rock fabrics and petrophysical properties of the marble types under investigation. A number of parameters has to be considered: The basis for any interpretation of ultrasonic data must be a comprehensive knowledge of the rocks fabric. Macroscopic and microscopic fabric properties are responsible for the type and magnitude of weathering of a marble. To quantify degradation on new monuments or sculptures the initial situation has to be documented since the quality of fresh marbles from the quarry can be significantly different.

Figure: 5 Distribution of the effective pore radii of extremely weathered Prieborn (a) and Carrara (b) marble. Thermal cracking is supposed to be the first step of marble weathering. Weathering increases the pore spaces due to microcrack generation and subsequent solution/precipitation activities. Thus it is important to know for a given marble type the proneness to thermal cracking and its promoting factors which is controlled by grain size, grain shape and preexisting open or healed cracks (Siegesmund et al., this volume). The result is a sugar-like disintegration like it is observed for some types of Carrara marble. Preexisting microcrack systems can be activated by weathering leading rather to a partial loss of larger parts of monument due to break-outs than to a total disintegration. This effect is on-site hard to detect when no superficial evidence for cracks is observed but it can be predicted when macro- and microfabric data are available (Weiss et al., 1999b). On-site ultrasonic velocities measurements have to be treated with care. Anisotropy is a typical characteristic of rocks. The Vp-anisotropy can account up to 30% and, consequently, a large difference in Vp can not unequivocally be explained with locally restricted disintegration. Even for Carrara marble, which is generally assumed to be isotropic, the Vp-anisotropy can be about 22% (Carrara III dry, Tab. 3).


223

But also the anisotropy in water-saturated marbles can be considerable as it was shown for the marbles from Kauffung. The tremendous velocity reduction within a very limited porosity range of only 0-1% (see Fig. 4) can only be explained by very flat pores with an aspect ratio of about 0.005. For calcite rocks it is well known that at relatively low temperatures (i.e. below 200~ grain boundary cracks are the predominant crack types (e.g. Fredrich & Wong, 1986). The socalled "pore throat diameters distributions" of weathered Carrara and Priborn marble give a maximum of the pore throat diameters close to l gm. A facet of a calcite grain can easily reach 150gm and, thus, the quotient between the pore throat diameter (i.e. the height) and the lateral dimension of the supposed crack (i.e. the crystals grain boundary) is close to the aspect ratio of 0.005 which was estimated from modeling. However, specimens in the open country are probably never completely water saturated. When the samples are only partially saturated the corresponding velocities approach rapidly those shown for the dry samples. Thus, it is very important to get information about the degree of water saturation of a specimen under investigation.

References

Birch, F., 1960: The velocity of compressional waves in rocks up to 10 kilobars, Part I. J. Geophys. Res., 65, 1083-1102. Dandekar 1968: Variation in the elastic constants of calcit with pressure. AGU Trasactions, 49 (1), 323pp. Duerrast, H., Siegesmund, S. & Prasad, M, 1999: Schadensanalyse von Naturwerksteinen mittels Ultraschalldiagnostik: M6glichkeiten und Grenzen. Z. dr. geol. Ges., 150, 2, 359374. Frederich, J.T., Wong, T.f., 1986: Micromechanics of thermally induced cracking in three crustal rocks. J. Gephys. Res., 91 B12, 12,743-12,764. Grimm W., 1999: Beobachtungen und lJberlegungen zur Verformung von Marmorprojekten dutch Geffigeauflockerung. Z. dt. Geol. Ges., 150 No. 2, 195-236. O'Connell, R.J., Budiansky, B., 1974, Seismic velocities in dry and saturated cracked solids, J. Geophysical Res., 79(35), 5412-5426. Siegesmund S., 1996: The significance of rock fabrics for the geological interpretation of geophysical anisotropies. Geotekt. Forsch., 85, 1-123 Siegesmund S., Vollbrecht A., Ullemeyer K., Weiss T., Sobott R., 1997: Anwendung der geologischen Geft~gekunde zur Charakterisierung nat~irlicher Werksteine - Fallbeispiel: Kauffunger Marmor. Int. J. f. Restoration of Buildings and Monuments, 3,269-292 Siegesmund S., Weiss T., Vollbrecht A., Ullemeyer K., 1999: Marble as a natural building stone:rock fabrics, physical and mechanical properties. Z. dt. geol. Ges., 150, 237-257. Tschegg, E.K., Widhalm, C., Eppensteiner, W., 1999: Ursachen mangelnder Formbest~ndigkeit von Marmorplatten. Z. dt. geol. Ges., 150(2), 283-297 Weiss T., Leiss B., Oppermann H., Siegesmund S., 1999a: Microfabric of fresh and weathered marbles: Implications and consequences for the reconstruction of the Marmorpalais Potsdam. Z. dt. Geol. Ges., 150 No. 2, 313-332. Weiss, T., Siegesmund S., Leiss B., Oppermann H., 1999b: Microfabric of fresh and wathered marbles and its implication on weathering phenomena observed at the Marmorpalais in Potsdam (Germany). Proceedings Eurocare/Euromarble 1998, Bayerisches Landsamt f~ir Denkmalpflege, Forschungsbericht, 17, 4-15


225

SALINE POLLUTION IN ISLANDS (PORTUGAL).

TRACHYTE

MONUMENTS

OF

THE

AZORES

C. A. S. Alves* Centro de Ci~ncias do Ambiente/Dept. Ci~ncias da Term, U.M. 4700-320 Braga, Portugal, e-mail: [email protected]. M. A. Sequeira Braga Centro de Cidncias do Ambiente/Dept. Ci~ncias da Terra, U.M. 4700-320 Braga, Portugal. A. Trancoso INETI, Estrada do Pa~;o do Lumiar, 1649-038 Lisboa, Portugal.

Abstract

The decay aspects of trachyte stones applied in churches from Ribeira Grande (Sao Miguel Island) and Angra do Heroismo (Terceira Island), both in the Azores Islands, were studied in order to characterise the associated soluble salts and to discuss contamination sources and decay processes in these monuments. Granular disintegration, scales and flakes are the main pathologies affecting the trachyte stones on the facades. The aqueous extracts compositions of these pathologies have a clear trend towards the chloride and sodium poles, and halite and gypsum were identified. Sodium salts also dominate the efflorescences, which occur on diverse building materials. The decay patterns and the mineralogical and chemical results suggest that rainwater is the main saline pollution source. The pollution by rainwater and the high precipitation contribute to the intense deterioration of trachyte stones applied in these monuments of the Azores Islands. Key words: trachyte stones, saline pollution, Azores Islands monuments

1. Introduction

Soluble salts are recognised as an important agent in the decay processes that affect historical monuments. The sources of the soluble salts are varied (Arnold and Zenhder, 1989), including geogenic sources (the stone itself, rainwater, groundwater and sea-spray) and sources related to anthropogenic activities and products (atmospheric pollution, mortars, organic wastes, etc). In order to adopt the appropriated conservation measures, the salt systems and the pollution sources must be identified. Papers dealing with the decay of volcanic stones are rare, being most of them dedicated to volcanic tuffs, a group of rocks common in some European regions and that are also applied in the famous Eastern Island monuments. Papers dealing with trachytes are rarer (Fitzner, 1994; Laue et al., 1996). The scarcity of trachyte stone decay studies is explained by the scarcity of the trachyte rock and, therefore its rare use as building stone. "Author to whom correspondence should be addressed.

226


In the Azores Islands trachyte has been, historically, an important building resource, due to its abundance in the islands and its lighter aspect, when compared with the other rocks found in these islands. However, the trachyte stones applied in the Azores monuments show intense decay. In some buildings the stone was rendered with mortars, lime and paintings. However the decay has affected also the rendering products. As consequence, the use of trachyte in building facades has declined and other darker rocks are now been preferred, with the consequent disruption in the building traditions of the islands. The present paper attempts to characterise the saline pollution that affects selected monuments of the two main islands of the Azores archipelago: S~.o Miguel e Terceira. The discussion of salt systems, pollution sources and decay consequences is also attempted.

2. Materials and methods

The decay aspects of several selected churches, built with trachyte stones, were studied in localities of Sao Miguel and Terceira Islands (fig. 1). In Sao Miguel Island, were studied monuments in the village ofRibeira Grande: the Espirito Santo, Matriz and Nossa Senhora da Conceigao churches. In the capital of the Terceira Island, Angra do Heroismo (UNESCO world heritage), were selected the Cathedral and the Nossa Senhora da Concei~ao church. Both localities are placed near the coast (fig. lb and lc). Yearly rainfall values (mean values for the period 1961-1990 from the portuguese Instituto de Meteorologia) are very high: 1125.6 mm for Angra do Heroismo and 1027.1 mm for Ponta Delgada (climatic station in Sao Miguel Island).

Figure 1: Localisation of a) Azores Islands; b) Ribeira Grande in S. Miguel Island and c) Angra do Herofsmo in Terceira Island.

A survey of the decay features (pathologies) of the monuments was conducted, in order to define a sampling program. Granular disintegration, scales and flakes are the most important decay forms affecting the trachyte stones, but there are also the occurrence of efflorescences on several building materials. Efflorescences were sample in all the churches and, to obtain a regional view of the salt systems, also in the Casa da Cultura of Ribeira Grande. The other decay forms (granular disintegration, scales and flakes) were collected in the following buildings: Espirito Santo and Matriz churches (S. Miguel) and Angra do Heroismo Cathedral (Terceira). Samples from old quarries in both islands and from a stair stone, without decay evidences, in the interior of Espirito Santo church were also collected for comparison.


227

Effiorescences were studied by optical microscopy (immersion method, following the recommendations of Arnold, 1984), X-ray diffraction (XRD) and scanning electron microscopy with EDS analysis system (SEM/EDS). Granular disintegration, scales and flakes were also studied by SEM/EDS and chemical analysis of the aqueous extract. The cations (Na +, Ca2+, K § and Mg 2+) were determined by flame atomic absorption spectrometry and the anions (CI, NO3 and SO42) by ion chromatography.

3. Results and discussion 3.1. Field study of decay aspects Intense deterioration has been observed in the trachyte stones applied in the facades of the selected monuments, but there are some differences regarding the pathologies. In the facades of the Angra do Heroismo churches, scales and flakes (with thickness between 1 mm and 18 mm) are the dominant decay forms affecting trachyte stones, but granular disintegration is also present. These decay features have a very wide distribution from the base to the top of the facades (fig. 2a). In the Ribeira Grande churches fa~;ades, the decay is mainly linked to granular disintegration, which in the Espirito Santo church is concentrated at the higher zones of the main Fa~;ade (fig. 2b,c). In the Matriz church, besides granular disintegration, there are also thin flakes (between 1 mm and 3 mm) and a discrete occurrence of effiorescences in the main fa~;ade. Inside the monuments were not found evidences of stone physical deterioration, but there are occurrences of efflorescences. The efflorescences are found on the wall rendering mortar (leading to the disruption of the wall painting), in mortar joints between stones and in a tile panel (in the mortar joints between tiles and in places of the files were the glazing has peeled off). Discrete occurrences of efflorescences in trachyte stones (that are covered with cement) were found inside the churches of Angra do Heroismo and in the N. S. da Concei~o church of Ribeira Grande.

3.2. Characterisation of pathologies In the efflorescences of the several churches were identified the saline minerals presented in table 1. With the exception of the Matriz church in Pdbeira Grande, all the efflorescences were found indoors. Sodium minerals are clearly dominant in the efflorescences. Inside the monuments the efflorescences mineralogy is monotonous, dominated by either mirabilite (Na2S04* 10H20) or sodium carbonate (trona and natron). However, inside the Angra do Heroismo churches, in stones that are covered with cement, halite and gypsum (Cathedral) or gypsum (N. S. Concei~o church) were identified. Halite and gypsum are also the saline minerals present in the efflorescences of the main facade of the Matriz church (Ribeira Grande). Calcite crusts were found on mortar joints in the churches of Angra do Heroismo, being probably a carbonation product of the mortar and, therefore, are not considered in the discussion of the soluble salts in effiorescences.

228


Figure 2: Trachyte stone decay aspects: a) deterioration by granular disintegration, scales and flakes in Angra do Heroismo Cathedral; b) and c) granular disintegration and erosion of decorative features in the Espirito Santo church of Ribeira Grande.

In the samples of granular disintegration, scales and flakes that affect trachyte stones, SEM/EDS studies (fig. 3)~showed the presence of halite and calcium sulphate that in some samples was possible to confirm, by XRD, as gypsum. The chemical analyses of the aqueous extract show enrichment on the soluble salts (determined by the sum of all the ions analysed, in %) in the trachyte stone pathologies, when compared with samples from the quarry and from the stair stone that was inside the Espirito Santo church (Table 2).


229

Table 1- Saline minerals found in the effiorescences of the studied monuments. Saline minerals (with indication of substrate) Church Locality Mirabilite (wall mortar rendering) Espirito Santo Ribeira - Halite + gypsum (trachyte) Matriz Grande - Mirabilite (trachyte covered with cement) N. S. Concei~ao (Sao Miguel) - Natron (mortar joint; trachyte covered with cement) - Mirabilite (mortar joint; trachyte covered with cement Casa da Cultura - Mirabilite (wall mortar rendering; mortar joints) Cathedral - Natron (tiles; mortar joints between tiles) - Halite + gypsum (trachyte covered with cement) Angra do Heroismo - Mirabilite (wall mortar rendering) N. S. Conceigao (Terceira) - Trona (mortar joint) - Natron (wall mortar rendering) - Gypsum (trachyte covered with cement)

Figure 3 - SEM/EDS studies of soluble salts in trachyte stone pathologies. Examples from Angra do Heroismo Cathedral: a) halite crystals (h) in feldspar (f) fractures in a trachyte scale; b) gypsum aggregates in the separation interface of flakes.

230


Table 2 - Soluble salt amount (as sum of ions analysed, in %) in trachyte samples (number of samples in bold between parenthesis). Zions (%) Locality Type of sample 0.0214 (1) Ribeira Quarry 0.022(1) Grande Stone from inside Espirito Santo church (Sao Miguel) Granular disintegration (Espirito Santo church) 0.6634-1.6405 (5) 0.449- 1.786 (8) Granular disintegration (Matriz church) 0.189- 1.106 (7) Flakes (Matriz church) 1.5197-6.15(3) Mortars (Espirito Santo church) 0.0268 (1) Angra do Quarry 0.643- 1.552 (4) Heroismo Granular disintegration (Cathedral) (Terceira) Scales and flakes (Cathedral) 0.1405- 1.742 (14) 1.473- 4.3403 (3) Mortars (Cathedral) The miliequivalent proportions of anions and cations in the aqueous extract of stone pathologies samples show a clear trend towards the chloride and sodium poles (fig. 4). There are however, some samples in the Ribeira Grande churches facades that clearly approach the calcium and sulphate poles. Nitrate is the less important anion in all the samples (always less than 30% of the anions determined).

c~2+

a)

CI-

NO 3

Na+ Ca2+

b)

C1-

K+ + Mg2+

NO 3

Na+

K+ + Mg2+

Figure 4: Proportions of main anions and cations (in meq) in the soluble extract of samples of granular disintegration (e) and scales and flakes (o) from a) Ribeira Grande churches and b) Angra do Heroismo Cathedral.


231

In the Matriz church of Ribeira Grande and in the Cathedral of Angra do Heroismo was possible to collect stone pathologies samples at diverse heights from the floor, allowing to study the evolution of the salt content with height. It is visible (fig. 5) that there is no relation between the height of the sample from the floor and its soluble salt content. When the miliequivalent proportions of the different anions and cations (considering separately the anions and the cations) are considered, instead of the total soluble salt content, there are also not chemical trends with height (fig. 6), excepting the nitrate, that seems to show a weak trend to decrease with height. _

|.6

"

1.2-

9

0.80.4

"

0

[]

A

9

[]

O

~

9 AA 9

A A

A

A A

[]

&lll []

[]

[]

!

!

1

2

h (m)

!

i

i

3

4

5

Figure 5: Soluble salt content (as sum of all the analysed ions, in %) against heigth from the floor for samples from Matriz church in Ribeira Grande ( 9 , granular disintegration; [], scales and flakes) and from Angra do Heroismo Cathedral ( 9 granular disintegration; zx, scales and flakes).

3.3. Considerations on salt systems and pollution sources

The extensive distribution of the decay aspects in the facades and the similarity of chemical characteristics (salt contents and proportions of the different ions) indicate a pollution source that affects the whole of the monuments fagades in a similar way. The chemical and mineralogical characteristics of the stone pathologies indicate a chloride and sodium dominated pollution source, presumably rainwater with an important oceanic influence. The high rainfall in the islands contributes to the intense decay of the trachyte stones. However, some dispersion is observed towards the sulphate and calcium poles, which could indicate some contribution from other pollution sources like automobile traffic or the products that were used for the rendering of the trachyte stones. The mineralogy of the efflorescences indoors is also dominated by sodium salts, mainly sulphates and carbonates. These results seem linked to the interaction between the sodium and chloride rich solutions from the outside salt system (linked to rainwater) with the mortars applied in the walls. Similar observations were made by Gurrera et al. (1994), who referred the formation of thenardite, trona and aphthilite as a consequence of the reaction between sea-spray and cement compounds. Alves and Sequeira Braga (1999) also observed the zoned evolution of sodium salts by interaction between solutions from a deposit of common salt and the wall mortars.

232


100

80

ZX 9

60

AAu

80

9

zk

,.-,w

O

A

r.~

~

oo

~x

A A A

'~n

A

-

A

o

[]

0 30

I

I

1

2

Zk

A=

~

[]

I

I

3 h(m)

A

A

l

4

AA

/k

ZX~ o

9

i

-

60-

ON

4

A

5

[] []

AA

4020 0

[] I

2

I

I

1

2

9

9

I0

1%

AoA

{]

A

I

3 h(m)

I

A A

4

o

AAA

A

AD ZX 9 ~ mZk'~x

9

3 h(m)

9 []

+

o

I

I

1

0

m o

O

20

[]

/k

....

ZXO

A

A

zx

n

0

9A

O

Zk

i

A

A

zll

10-

mA

9 [] ~

A

A

4

9 A

20

AA

3 h(m)

A

ZX A

A

2

9

A

40

o

o

0

50

30

A

)

A

+ ~

A A

O

%, 8 0 -

I

3 h(m)

1

60

0

100

I

I

20

5

A

10"

AA

40

o

/k O 20 -

[]

Ao

A ~k []

zx

Am

A

60

40 20

o

I

I

4

5

0

0

I

I

1

2

t::~km ~k Z~Im I

3 h(m)

I

I

4

5

30 NA

+

[]

20 zx

10 -

m~ 9

~

~o m

A A AA A

0

%

9

I

I

1

2

I

A

3 h(m)

9

o

IAA

i

4

5

Figure 6: Miliequivalent proportions of the different anions and cations against height (in m) for samples from Matriz church in Ribeira Grande (m, granular disintegration; [] , scales and flakes) and from Angra do Heroismo Cathedral (A, granular disintegration; zx, scales and flakes).


233

The monotony of the saline mineralogy in the efflorescences of these Azores monuments is deafly different from what has been observed in other Portuguese monuments affected by several pollution sources (Alves and Sequeira Braga, 1999). In those monuments, the variety of saline minerals allowed the authors to identified several saline systems in the same monument, while in the Azores Islands the saline mineralogy is similar in both localities (placed in two different islands), indicating a common salt system. The soluble salts characteristics are similar in the different types of pathologies. Therefore, in order to understand the variations in the pathology type and in its morphology other factors must be investigated.

4. Conclusions The decay patterns and the chemical and mineralogical characteristics of the decay products indicate similar salt systems in these trachyte monuments of the Azores islands and a saline pollution mainly linked to geogenic sources (rainwater with an oceanic component). The pollution effect of rainwater and the high rainfall values in these islands contribute to the intense deterioration of trachyte applied in the monuments facades.

5. Acknowledgements Supported by the Funda~ao para a Ci~ncia e Tecnologia (Portugal) through PRAXIS X X I - project n ~ 2/2.1/CSH/254/95 and CCA-CT/FCT-R&D Contract-Program.

6. References 9Alves, C.A.S. and Sequeira Braga, M.A. (1999) Soluble salts in pathologies of granitic monuments of Braga (Northwest Portugal). Ninth Annual V.M. Goldschimdt Conference, pp. 5-6. 9Arnold, A. (1984) Determination of mineral salts from monuments. Studies in Conservation, 29, 129-138. 9Arnold, A. and Zehnder, K. (1989) Salt weathering on monuments. 1st Int. Syrup." The Conserv. of Monuments in the Mediterranean Basin, Bari, 31-58. 9 Fitzner, B. (1994) Volcanic turfs: the description and quantitative recording of their weathered state. Lavas and Volcanic Tufts, Int. Meeting, Eastern Island, Chile, pp. 33-51. 9Gurrera, M.A., Raventos, X.D., Bou, V.E., Perez,J.L.P., Vifias, R.R., and Horta, A.V. (1994) Degradation forms and weathering mechanisms in the Ber/t Arch (Terragona, Spain). 3rd. Int. Symp.:The Conservation of Monuments in the Mediterranean Basra, Venice, 673-679. 9Laue, S., B6hm, C.B., Jeannette, D. (1996) Saltweathering and Porosity - Examples from the Crypt of St. Maria in Kapitol, Cologne. 8th lnt. Congr. on Deterior. and Conserv. of Stone, Bedim, pp. 513-522.


235

TRACHYTE STONES IN MONUMENTS OF THE SAO MIGUEL AND TERCEIRA ISLANDS, AZORES (PORTUGAL). M.A. Sequeira Braga CCA/CT, Universidade do Minho, 4700-320 Braga, Portugal M.O. Figueiredo Inst. Investiga~.o Cientifica Tropical, AI. Afonso Henriques, 41, 4~ 1000 Lisboa, Portugal. M.I. PrudSncio Instituto Tecnol6gico e Nuclear, Aptdo. 21, 2686-953 Sacav6m, Portugal J. Delgado Rodrigues* Laboratorio National de Engenharia Civil, Av. Brasil, 101, 1799 Lisboa Codex C.A.S. Alves CCA/CT, Universidade do Minho, 4700-320 Braga, Portugal D. Costa Laborat6rio Nacional de Engenharia Civil, Av. Brasil, 101, 1799 Lisboa Codex T. Silva Inst. Investiga~o Cientifica Tropical, A1. Afonso Henriques, 41, 4~ 1000 Lisboa, Portugal. M.J. Trindade, J.C. Waerenborgh, M. Nasraoui and M.A. Gouveia Instituto Tecnol6gico e Nuclear, Aptdo. 21, 2686-953 Sacav6m, Portugal

Abstract Among the volcanic rocks from Azores Islands, trachytes have been preferentially used through centuries in many monument fagades, mainly due to their colour and softness. However, aider a few centuries the decay stone of monument stones is remarkable. Angra do Heroismo Cathedral, in Terceira Island, and Miseric6rdia Church, in Sao Miguel Island, were the monuments selected for the present study on degradation causes and mechanisms. The same methodology was applied to quarry and monument samples from both Islands by using petrographic, mineralogical, chemical and petrophysical techniques. Well-marked differences were noticed between the trachytes from Sao Miguel and Terceira Islands but within each island, outcrops or the monument building stones, show several similarities. However, some textural aspects observed in the field showed some variability. The stones have a pore-type porosity of genetic nature. The values of porosity accessible to water are very high for monument samples and higher than the values for quarry samples. The samples of Miseric6rdia Church show a high porosity, dominated by large pores (radius > 1 lam). It was concluded that these highly porous materials trachytes are very susceptible to degradation, namely by salt crystallisation mechanisms. Keywords: stone decay, trachytes, monuments, Azores Islands 1. Introduction Volcanic rocks are the typical building materials used in the Azores Islands. Among these rocks, trachytes have been used through centuries in many monument facades mainly due to their colour (the so-called white stone) and softness (easy to work). However, time


236


has shown that trachytes have one of the worse behaviours as building materials in those islands. Two monuments were selected for the present study. The Cathedral of Angra do Heroismo is located in Angra do Heroismo (Terceira Island, Azores), a city included in the World Heritage List by UNESCO. Its building began in 1570 on the site of a 15th century church and was completed in 1618. The Cathedral was badly damaged by an earthquake in 1980. The Misericrrdia Church (also named Espirito Santo Church) in town of Ribeira Grande (S~o Miguel Island) exhibits in its main facade one of the most important sculpture works in Baroque style. Terceira and S~o Miguel (fig. 1) are two of the nine volcanic islands comprise by the Azores Archipelago located near the Mid-Atlantic Ridge, between 36~ 35'N and 39 ~ 43' N (latitude) and between 25 ~ and 31 ~ W (longitude). High annual values of rainfall have been recorded: 1125.6 mm/yr in Tereeira (mean values for the period 1961-1990, according to the Portuguese Institute of Meteorology) and in Sao Miguel between 3000 mm/yr at the Lagoa do Fogo to 1000-1500 mm/yr in the dryer part of this Island (Ricardo et al., 1977).

Figure 1: Location of Azores archipelago. Several publications report on the geology of the Terceira Island (Zbyszewski et al., 1971; Self, 1973; Forjaz et al., 1990) and of the Sao Miguel (Zbyszewski et al., 1958, 1959; Moore, 1990, 1991). The oldest formations of each island are Pliocene in age and the most recent are Quaternary. Some historic eruptions have occurred in both islands. Rocks of the volcanic complexes from Terceira Island consist essentially of lava flows, pyroclasts and ignimbrites. In S~o Miguel Island these rocks comprise trachytes (lava flows and domes), basalts and pyroelastic deposits. Trachyte building stones in the Azores monuments, in particularly in S~o Miguel and Terceira Islands, show an intense decay state. Alves et al. (2000) attempted to characterise the saline pollution that affects some monuments of these islands. Regarding the Misericrrdia Church in Ribeira Grande, the stone decay was mainly linked to granular disintegration that is concentrated in the higher zones of the main facade. The decay state of the trachyte building stone in the Angra do Heroismo Cathedral is intense and widespread by several facades.


237

In this work a comparison of trachyte rocks from quarries and monuments was carried out by using petrographic, mineralogical, chemical and petrophysical techniques. The main goals of the present study are: (1) to characterise the trachyte from the quarry and from the monument; (2) to point out similarities and differences, in order to investigate the building stone provenance and (3) to contribute to the understanding of the decay processes. 2. Materials and methods Angra do Heroismo Cathedral, in Terceira Island, and Misericordia (or Espirito Santo) Church, in Ribeira Grande, from S~.o Miguel Island - themonuments selected for study (fig. 2) - were both monuments built with trachyte stones. The field survey in the Terceira Island showed that the rock in the Grota dos Calrinhos quarry, located about 2 km Northeast of Angra do Heroismo, has macroscopic similarities with the monument stones from South and East facades. In this quarry, trachyte lava flows show some textural and chromatic variability and are associated with Santa Barbara Volcanic Complex (Forjaz et al., 1990). Samples were collected from Grota dos Calrinhos quarry and from decayed building stones (scales and granular disintegration) in the monument. In S~o Miguel Island, samples were collected in nine outcrops of lava flows and domes between Porto Formoso and Ribeira Grande. In this attempt of covering the potential places of extraction of trachyte, it should be remarked that old quarries (known by the older local habitants) are today inaccessible. A sample from one of these old quarries that was stored in the so-called Casa da Cultura was studied. The sampled trachyte outcrops are related to precaldera units of Agua do Pau stratovolcano (zone 3 - Moore, 1991). The samples showing more macroscopic similarities with the monument stones were those from Porto Formoso and Lameiro trachytes. Samples of the monument were collected on a stone fragment from a fallen decorative element of the main facade (fig. 2) and on a stone previously used as an altar step. The same methodology was applied to quarry and monument samples from both islands. Mineralogical, petrographical and chemical characterisation were developed by optical microscopy, scanning electron microscopy with EDS analytical system (SEM-EDS), X-ray diffraction (XRD) of the whole rock and of the u

90 -

-*-

w "o

-

-

|

80-

E L_ o

70

._

Z

, 0

5

i

i

i

i

i

White tuff Pink tuff

i

i

i

i

10 15 20 25 30 35 40 45 50 55 Number

of Cycles

Figure 3" Variations of the effective porosity of the tufts after freeze-thaw tests

=~"=11ol '~

100

=

~

90

"o

|

80

E o

70

._ L_

Z

. . . .

i

.;

- - o - -White tuff =

0

5

Pinktuff

10 15 20 25 30 35 40 45 50 55 Number of Cycles

Figure 4: Variations of the dry unit weight of the tufts after freeze-thaw tests A

=~

110 1

o

~

L

'I"

,1~ a,,

100

I=

.......

""

90

.__.~

80

o

70

E= r

z

=

=

-

-,-

-

=

i

0

i

i

i

i

i

i

White tuff Pink tuff

i

i

i

5 10 15 20 25 30 35 40 45 50 55 Number

of Cycles

Figure 5: Variations of the water absorption under atmospheric pressure of the tufts after freeze-thaw tests


279

110 - -.-

~0o

. W h i t e tuff Pinktuff

=

u -o 0 .N

i

.9 .

90

-- 9

t~

E

'oz

80

70 0

i

i

i

I

i

I

i

i

5

10

15

20

25

30

35

40

Number

45

50

55

of Cycles

Figure 6: Variations of the uniaxial compressive strength (UCS) of the tufts after freezethaw tests 11o

~

o 100~ U

.m

U ,i

=

o u) "u q)

-o-

.

=

White tuff Pinktuff

..

90-

,-~

80

E o z

70

L.

I

i

5

10

i

|

t

15 20 25

[

i

~

i

30 35 40 45

i

50 55

Number of Cycles

Figure 7 Variations of the sonic velocity of the tufts after freeze-thaw tests Fig. 2 shows that no significant change in weight occurs at the end of 20 freeze-thaw test cycles for both white and pink tufts. Slight weight losses are observed only in white tuff at the end of 40 test cycles. Weight loss is very significant in white tuff after 40 test cycles. It is in the form of disintegration and is mainly concentrated at the upper and lower corners of the white tuff samples. At the end of 52 test cycles, weight loss in the white tuff is reached to 14.4%, although it is about 0.2% in the pink tuff. Effective porosities of both tufts are also not changed at the end of 20 test cycles (fig. 3). It is increased by 5.7% in white tuff and 2% in pink tuff at the end of 52 test cycles. Similar variations are observed in dry unit weight and water absorption values of both tufts (figs.4 and 5). Effect of freeze-thaw tests in UCS and sonic velocity of the tufts is very significant. UCS is decreased by 5.6% in the white tuff and 4.4% in the pink tuff, at the end of 10 test cycles. The loss in UCS is more pronounced at~er 40 test cycles. UCS is dropped by 29.4% in the white tuff and 9% in the

280


pink tuff at the end of 52 test cycles (fig. 6). Sonic velocity of the tufts also decreases as the number of test cycles increases. Decrease in sonic velocity in the white tuff (2.4%) is almost the same as the pink tuff (1.9%) at the end of 20 test cycles. However, sonic velocity is much more reduced in the white tuff as the number of test cycles increases. Reduction in sonic velocity is 5.6% in the white tuff and 2.2% in the pink tuff, at the end of 40 test cycles. It is 23.8% in the white tuff and 2.6% in the pink tuff at the end of 50 test cycles (fig. 7). In general, except the dry unit weight, all the other physico-mechanical properties of both tufts are influenced as the freeze-thaw test cycles increased. The change in the physico-mechanical properties of the white tuff is more than that of the pink tuff. Therefore, the pink tuff is more resistant to freeze-thaw activity than the white tuff. The field observations also reveal that the pink tuff is more durable than the white tuff. Most of the ornaments and pediment decorations carved in BC 600 in the pink tuff can be easily seen even today. Considering the fact that the pink tuff is also slightly affected from the freeze-thaw cycles, both tufts should be protected from freeze-thaw activity.

4. Conclusions

The Midas monument formed within nonwelded white and slightly welded pink Yazfl~kaya tufts, shows some signs of deterioration. Freeze-thaw tests are carried out to assess the freeze-thaw resistance of the tufts. The physico-mechanical propeties such as weight loss, effective porosity, dry unit weight, water absorption under atmospheric pressure, uniaxial compressive strength (UCS) and sonic velocity of both white and pink tufts are recorded at different test cycles, and compared with those of the fresh samples. Based on the findings of this study, except dry unit weight, all the other properties of the tufts are found to useful to assess the damage. However, deterioration of the tufts by freezethaw tests can be better followed by UCS and sonic velocity measurements. Although the pink tuff is found to be more resistant to freeze-thaw activity, both white and pink tufts are adversely affected from the freeze-thaw tests. Therefore, the Yazfl~kaya tufts should be protected from freeze-thaw activity.

5. References

Ayday, C. and G0ktan, R.M., 1990. A preliminary engineering geology study directed to the conservation of Midas monument, Proc. International Earth Sciences Colloquium on the Aegean Region (IEASCA), GOll0k-Izmir, Turkey, 102-108. Ayday, C., and G0ktan, R.M., 1993. Yazlhkaya (Midas) anm civarmda g0zlenen kaya blok devrilme ve kayma mekanizmalan, Tiirkiye Jeoloji Kurultayl B01teni, 8, 155-159. Binal, A., Kasapo(glu, K.E., and GOk~eo(glu, C., 1997. The surficial physical deterioration behaviour ofNeogene volcanosedimentary rocks of Eski~ehir-Yazfl~kaya, NW Turkey, Proc. Int. Cong. on Engineering Geology and the Environment, Athens, Balkema, Rotterdam, 3065-3069. Binal, A., Kasapofglu, K.E., and G0k~eo~lu, C., 1998. Variation of some physical and mechanical parameters of the volcanosedimentary rocks around Eski~ehir-Yazfl~kaya under freezing-thawing effect, Yerbilimleri, 20, 41-54. ISRM, 1981. Rock characterization, testing and monitoring, International Society for Rock Mechanics Suggested Methods, Pergamon Press, Oxford.


281

RILEM, 1980. Recommended tests to measure the deterioration of stone and to assess the effectiveness of treatment methods. Commission 25-PEM, Materials and Structures, 13, 175-253. Rossi-Doria, P.R., 1985. Laboratory tests on artistic stonework. The Deterioration and Conservation of Stone. Studies and documents on the cultural heritage. No. 16, 23 5-242. Topal, T. and Doyuran, V., 1998. Analyses of deterioration of the Cappadocian tuff, Turkey, Environmental Geology, 34, 1, 5-20. Balkema, Rotterdam, 2939-2944. TS699, 1987. Methods of testing for natural building stones (in Turkish). Ttirk Standartlarl EnstittisO, Ankara.


283

AN EVALUATION OF GEOLOGY AND WEATHERING IN THE PRESERVATION OF MARL OBJECTS

Metaxia Ventikou* Sculpture Conservation, Victoria and Albert Museum, London, UK. Chris Halls Royal School of Mines, Imperial College of Science, Technology and Medicine, London, UK. William Lindsay Conservation Unit, Department of Palaeontology, The Natural History Museum, London, UK. Meryl Batchelder Department of Mineralogy, The Natural History Museum, London, UK. Charlotte Hubbard Sculpture Conservation, Victoria and Albert Museum, London, UK.

Abstract

The deterioration of an Italian marl fireplace in the Victoria and Albert Museum led to the research described in this paper. The purpose was to study the geology and weathering of marl in order to explain the poor state of preservation of such objects and to develop a suitable conservation strategy. Marl is an argillaceous carbonate rock rarely encountered in works of art. The mineralogical and geological characteristics described here are particular to the sample examined, but these properties can be considered typical for all marls. The fireplace marl consists of clay minerals and calcite and the rock has a characteristic fabric. Natural weathering, combined with the effect of the conditions, under which the fireplace has been used and stored, were identified as the cause of its partial disintegration. Among the factors governing deterioration, atmospheric humidity has played a major role and is the one most likely to affect the condition of marl objects. Experiments were made in which a sample was exposed to various levels of relative humidity and the resulting dimensional changes were measured. The outcome of this research was to establish a deterioration model confirming that variations in ambient relative humidity are the main cause of weathering in a marl artefact. Keywords: marl, sedimentary, delamination, weathering, relative humidity, displacement.

1. Introduction

Marl, due to its composition and conditions of formation, is a sedimentary rock highly susceptible to weathering. Although marl is not a common material in sculpture or architecture, when used, it tends to show marked deterioration which poses significant problems for conservation. The sixteenth-century Italian fireplace, in the Victoria and Albert Museum (inv. no. 72531861), provides a case study (Fig. 1). On acquisition, in 1861, it was recorded as originating from a house in Savona, North Italy. Until the present study revealed it to be marl, it was *Sculpture Conservation,Victoria and Albert Museum, CromwellRoad, LondonSW7 2RL, UK.

284


believed to be carved from black slate (or pietra di lavagna). The fireplace consists of 11 pieces of various dimensions. All pieces show delamination, parallel to the bedding of the rock, which is most severe at the carved surface. Many of the carved details do not survive and those still intact are very friable, or even detached from the substrate. The geology of marl is very important in understanding the material and thus, crucial in establishing a conservation treatment. Marl has been fully described in the geological literature, but because it is rarely used as a sculptural medium, it has not been broadly characterised in conservation terms. In addition, the term 'marl' has been used to describe a variety of rocks and its classification depends on several parameters. It was therefore necessary to establish a suitable definition of the material before proceeding to study weathering mechanisms and potentially suitable methods of preservation.

2. The geology of marl 2.1 Definition Marl is a sedimentary rock with a composition between that of claystone and carbonate rocks (Millot 1970, Cornell and Aksoyoglou 1990). Specifically, marl is a mixture of clay and carbonate minerals, the latter between 35 and 65% of the total composition (Greensmith 1989). However, any sedimentary rock with significant clay and calcite content may be further classified to one of the marl types given by Kirsch (1968), ranging from marly clays to marly limestones. It should also be noted that marls are heterogeneous rocks, a feature common to many sedimentary formations. In geological publications that include compositional analyses of marl, the carbonate content is found in various percentages ranging from 20 to 80%. (Cornell and Aksoyoglou 1991, Antoine, Giraud, Meunier and Van Ash 1995.). Marl is further defined by its petrology. The term marl is applied to chemical and biogenic sedimentary rocks originating from unconsolidated calcareous or dolomitic mud which, if it contains clay minerals and calcite in the above percentages, when consolidated will become marl (Greensmith 1989). The texture of marls is governed by the rate of deposition of the sediments and local variations in the proportions of the clay and carbonate components. 2.2 Identification of the fireplace marl The properties of the fireplace material were identified by optical microscopy, scanning electron microscopy and quantitative x-ray diffraction. The samples used for the analyses were fragments of unknown original position already detached from the surface of the fireplace. The results from the quantitative XRD analysis are given in Table 1. In particular, clay minerals constitute 41-52% of the stone and the smallest percentage of calcite detected was 31%. Given this mineralogical composition, the rock can be classified definitely as marl. The clay minerals were found to be of sedimentary origin and have not undergone significant alteration in the diagenetic environment. Among the clay minerals present, kaolinite is well crystallised. During XRD identification, heating to temperatures above 500~ does not alter the illite structure. Although illite may coexist in mixtures with smectite in sedimentary rocks, the treatment with glycol during the same analysis did not indicate any swelling phenomena. The presence of muscovite indicates a late diagenetic process, but its coexistence with illite and chlorite does not suggest significant prograde recrystallization or alteration. Finally, chlorite in the marl samples is well-crystallised and its complete structure is confirmed by the stable XRD peak when treated with ethylene glycol. Spectrum 1 shows the detected clay minerals and illustrates their stability under different treatments. The fireplace


285

marl does not contain expandable clay minerals of the smectite group. Table 1. Minerals and their percentages identified in the fireplace marl Mineral

Weight (%)

Kaolinite Illite Muscovite Chlorite Calcite Gypsum Pyrite Quartz Total

3-5 23-27 4-5 11-15 31-33 3-4 1-2 19-21 100

Figure 1. XRD spectrum of minerals in a sample from the marl fireplace. A 550~ 2 hours, B 400~ 2 hours, C Ethylene Glycol, D Air dried. The first peak corresponds to chlorite, the second to illite and the third to kaolinite.

In thin section, marl would be expected to show a planar stratification corresponding to the gradual deposition of the sediment. Instead, foliations of clay minerals and carbonates are observed to form lenticles around carbonate domains. Clay minerals exhibit a layer structure that is easily arranged in a shape-preferred orientation. In argillaceous sedimentary rocks like shales, for instance, clay minerals show quite well developed planar arrangement as a result of deposition and compaction (Weaver 1989). Under late diagenetic conditions, clay minerals are recrystallised to form larger structures, whereas in metamorphic states, minerals are transformed into completely new assemblages and the fabrics and preferred orientation from the original rock are not readily distinguished (Weaver 1989). Because of the thin, layered structure of clay minerals, they are visible in thin sections only as crystal aggregates and foliations. Grains consisting of chlorite/muscovite aggregates and individual flakes of muscovite reaching sizes of approximately 55/L m were observed in marl from the fireplace. The presence of foliated crystals suggests a mechanism different from ordinary sedimentation and, in the case of this marl, the foliated fabric is attributed to the effect of pressure exerted by overlying sediments, causing diagenetic growth of clay minerals with a lattice-preferred orientation along the foliated planes of bedding. Since the minerals do not show significant metamorphic modification, the rock acquired its properties at a late stage of diagenesis (Fig.2). The formation of foliations is further assisted by the general behaviour of calcite and carbonate minerals during diagenesis and compression. The stress differences generated during burial and diagenesis produce elongated and lenticular structures from the aggregates of carbonate. Burial overpressure causes partial dissolution, redistribution and re-precipitation of the soluble carbonates which thus form lenticular domains parallel to the bedding planes. These are surrounded by insoluble clay minerals and other carbonaceous organic material, all of which are moulded around the carbonate lenticles. The conclusion is that the rock acquired its foliated texture at a late stage of diagenesis but it can be characterised as marl on the basis of its dominantly sedimentary features and the clay and calcite composition.

286


Figure 2. Marl fireplace piece 7253.7-1861 (from view) Figure 3. Chlorite aggregate in the foliated arrangements of clay minerals and calcite in a thin cross section from the fireplace marl, under transmitted polarised light (field width 550 ~t m)

3. Weathering mechanisms of marl Until recently removed for study, the fireplace has remained in the same store for as long as records have survived. The conditions in the store were generally stable, as appropriate to sculpture collections. However, it is evident by comparing the present state of the fireplace with a photograph taken of the piece 7253.1-1861 in 1899 that its condition has deteriorated progressively over the last one hundred years. The exposed surfaces of the marl fireplace, particularly those carved in relief, are splitting and becoming detached. The decision was therefore taken to determine the composition of the fireplace and establish the weathering mechanisms responsible for this breakdown. According to conservation terminology, the fireplace has been subject to extensive delamination which has taken place parallel to the bedding of the rock. This is most extreme on the carved surface. The present state of the marl fireplace can be attributed to a combination of mechanisms that have acted on the rock and the clay minerals in it from the time when it was brought to the surface by erosion and quarried from its place of origin. The main causes of deterioration can be classified as follows.

3.1 Weathering due to mineralogy The combination of clay minerals and calcite makes marl a very soft stone to carve. It has low strength and is susceptible to shrink-swell behaviour. In consolidated rocks, the existence of clay minerals significantly contributes to the degree of water penetration and swelling. Water entering the rock due to the presence of clay minerals is a cause of deterioration. Two main mechanisms govern the absorption of water by clay minerals. These are surface adsorption and the attraction of water molecules to the crystal lattice resulting in expansion. Lattice absorption does not occur in the fireplace marl because the clay minerals present are not expandable As far as adsorption is concerned, all clay minerals attract water to their surfaces. When water molecules approach the mineral surfaces they become adsorbed to form a molecular film which has a different density and viscosity compared to that of liquid water (Grim 1968, Kirsch 1968). Adsorption of water by clay minerals is a characteristic phenomenon and is the main response of those minerals to water (Mitchell 1993). In addition, water may dissolve the carbonates and thus alter the chemical composition and


287

porosity of the rock. In particular, water containing carbon dioxide can dissolve and carry away calcium carbonate from calcareous argillaceous rocks, including marl. Calcite becomes more soluble at high pressures of carbon dioxide and low temperatures (Deer, Howie and Zussman 1991). Some porosity can be created by this process in marls in contact with sedimentary waters during burial and diagenesis and later, when in contact with carbonic meteoric waters during exhumation as erosion of the land surface proceeds. The porous fabric of some marls, with carbonate domains and void systems, is attributed to the dissolution and precipitation of carbonate minerals in this way. Decalcification during weathering would also occur along planes of weakness created by material discontinuities and fractures where water gains access more readily (Hawkins and Pinches 1992).

3.2 Wetting-drying cycles and temperature variations Temperature plays a very important role in the weathering of rocks by affecting the state and chemical action of water, the solubility of salts, the rate of chemical reactions and by causing thermal expansion and contraction. When the effects of temperature and humidity are combined, the mechanisms governing deterioration of rocks become even more complex. The generally heterogeneous behaviour of minerals during wetting-drying cycles and temperature variations under ambient conditions can cause stone to disintegrate. However the fireplace may have been exposed to extreme variations when hot and dry conditions alternated with cool and humid conditions during seasonal domestic use. Relative humidity must have contributed to the weathering mechanism. The use of the fireplace, presumably on a daily basis during winter, would have resulted in frequent and intense RH changes in the environment of the marl. Since RH depends on temperature, when the surface of the marl cooled, water would be adsorbed on the surface of the fireplace, penetrating the outer layer of the marl as the stone cooled. When the temperature increased again as a result of fire, moisture would evaporate and the cycle would be repeated. Temperature variations also cause deterioration because of the mineralogy of the rock. It has been reported that swelling of clay minerals takes place more rapidly if water is absorbed under dry conditions (Price 1995). During drying, clay minerals shrink and their structures progressively break down. Shrinkage during drying becomes even more destructive in rock masses like marl which show textures in which voids are interspersed with closely compacted clay mineral aggregates and carbonate grains (Mitchell 1993). Consequently, both adsorbed and free water will be lost during drying and the rock must have been strongly affected by wetting-drying cycles. 3.3 Weathering due to fabric and overconsolidation All sedimentary and consolidated rocks may suffer some expansion, and even softening, when pressure is reduced during the removal of overburden and stored strain energy is released (Cripps and Taylor 1981). This stress relief readily causes expansion. Fractures open in a direction parallel to the release of force, and separation of foliations and bedding planes in the rock can occur. As a result, water absorption increases because of the increased permeability and opening of fractures (Price 1995). The state of preservation of the fireplace marl and the extensive delamination observed on the outer surface, are primarily attributable to the effects of unloading of the rock in combination with the effects of the humidity and temperature variations described above. In addition, for at least fifteen years, the fireplace was subject to the weight of a marble fireplace of similar dimensions which rested on it due to a dense storage arrangement and the misidentification of the fireplace material as slate. The storage was improved in 1995 after

288


concerns about these unsatisfactory conditions led to the removal of the outmoded system of storage. The weight of marble would have imposed a stress which, when removed, would have caused small-scale expansion added to that induced by removal of the original geological overburden. There is the added possibility that shear stresses imposed by this loading during storage could have facilitated delamination, given that the carved relief of the surface cuts across the sedimentary laminae which lie parallel to the principal plane defined by the slabs of which the fireplace is formed.

4. Analyses of the effect of relative humidity on the fireplace marl 4.1 Aim of experiments Of the main weathering mechanisms discussed, variations in humidity had the most dramatic influence on the degradation of the fireplace during its existence, and can be expected to continue to have an effect in the future. Extreme temperature variations are no longer like in the museum environment and the fracturing due to stress release which accompanied geological unloading must have been completed at the site of origin during the past. To confirm the role of relative humidity and measure its impact on the marl, a series of experiments were conducted. The aim was to determine the response of the fireplace to humidity and thus establish appropriate guidelines for conservation and determine treatment appropriate for the preservation of the fireplace and other marl artefacts.

4.2 Experimental procedure Experiments were designed in order to examine the response of the marl sample to extremes and cycles of relative humidity. The tests were performed in the Palaeontology Conservation Unit of the Natural History Museum. The experimental arrangement consisted of a micro-climate generator that provided the relative humidities and a measuring device for recording dimensional change. The micro-climate generator can accurately control the RH by conditioning and recirculating the appropriate amount of humid air in order to maintain the value set by the operator. The generator is equipped with software that simultaneously monitors the RH in the case and enables the user to program values from 10 to 90% RH. The measuring device used was a displacement transducer connected to a digital LED indicator. The temperature is not controllable but is monitored and remained at 23~ with occasional variations of _+ I~ The transducer responds to dimensional changes of _+2500x10 -4 mm, which are displayed in real time. A sample of volume 1.5cm 3 was taken from the marl fireplace piece 7235:13b-1861. The sample was placed inside the sealed case of volume 33736cm 3, with the marl foliations parallel to the level of the base. The transducer was adjusted to record the response of the laminations. The case was connected to the micro-climate generator at the RH set to 35%. A sensor inside the case sends the values of temperature and RH to be processed by the recorder in real time The dimensions of the sample were allowed to stabilise under the relatively dry conditions of 35% RH. The dimension in that state was taken as a standard zero with any subsequently induced expansion or contraction registering as positive or negative values respectively.

4.3 Results and discussion Experiment 1: From an initial environment of 35% RH, the sample was gradually exposed over a two hour period to relative humidities up to 75%. It was held at 75% RH for 75 minutes


289

when the RH was progressively decreased over a period of two hours, to 35%. Under these conditions, RH changed by 1% every three minutes. Comparable values have been recorded for some periods in the storeroom where the fireplace was kept. As illustrated in Figure 4, displacement and RH show a direct positive correlation. The marl sample expands as the RH rises, and it contracts as the RH falls. The response to the RH changes is rapid, although there is some delay in the change which takes place as RH is reduced. During contraction the dimensional changes are smaller than the expansive changes which accompany increasing RH, and the sample does not regain its original dimensions. Experiment 2: The micro-climate generator was programmed to provide a RH of 35% and when the above condition was achieved, the RH was increased as rapidly as possible to 75%. After the RH inside the case reached 75%, it was held at that level for six hours in order to examine the maximum displacement that could occur in the sample under constant high RH. During the second experiment, as shown in Figure 5, the sample responded almost immediately and the rate of displacement is fast when the RH change is rapid. The displacement tends to slow down or stop when the RH stabilises at a high level. Moreover, it is notable that the sample showed a displacement of 71x10 ~ mm, the maximum measured experimentally. The sample was left under 75% RH for six hours, though the expansion reached a maximum and stabilised after five hours. This time is therefore taken to be the time needed for the marl to respond to equilibrate under conditions of high humidity. Experiment 3: The third experiment involved the exposure of the same sample to repeated changes of relative humidity within the range of 40 to 60%. These values correspond to the ambient RH under which the fireplace is kept in storage. Given that the first and second experiments show that the sample continues to change its dimensions after RH has stabilised, in this case the sample was not allowed to reach dimensional equilibrium as the RH was varied. This experiment was designed to determine whether these short cycles of expansion and contraction had a permanent effect on the rock. The displacement curve in Figure 6 shows a very clear response to the RH changes. The pattern of displacement is repeated during the second cycle, confirming the consistent effect of short-term variation of RH on the dimensions of the sample. The high RH was maintained for less than 20 minutes and the sample showed a maximum displacement towards the end of that period. The abrupt halt in dimensional expansion suggests that further deformation would be experienced had the RH continued to increase. The sample presented the same pattern of displacement during both contraction and expansion. Experiment 4: The sample was stabilised in 50% RH and wetted with a brush to determine the possible response caused by wetting of the fireplace. The RH inside the case was set to 80% for two hours so that the wet condition was maintained. Figure 7 illustrates the behaviour of marl under these extreme conditions. During the first 5 minutes after wetting the fastest displacement took place. Thereafter the sample continued to expand rapidly for 15 minutes, followed by a slower dimensional increase. The sample reached the maximum expansion (180xl 0 -4 mm) measured in these experiments. These experiments demonstrate the susceptibility of marl to changes in relative humidity and the effect of wetting. While it is apparent that the sample deforms in response to RH changes, further study is required to establish the extent to which the dimensional changes are permanent. Moreover, given that the sample was much smaller than the object, it can be assumed that changes in RH would have less impact on the fireplace as a whole and that dimensional changes would be slower. Because adsorption and penetration of humidity depends on the area exposed, once delamination has begun, the rate of disintegration will increase. However, the experiments confirm that marl responds to ambient RH variations in

290

9th International Congress on D e t e r i o r a t i o n a n d C o n s e r v a t i o n of Stone, Venice 19-24 June 2000

the museum environment and is especially susceptible to wetting by liquid water. The response of the fireplace to ambient humidity and water demonstrates that the effective preservation of the fireplace and related marl objects requires an environment in which RH is kept relatively constant and that water should not be used during conservation treatment. Figure 4 Displacement and RH vs time (experiment 1)

Figure 6 Displacement and RH vs time (experiment 3)

Figure 5 Disolacement and RH vs time (exoeriment 2)

Figure 7 Displacement vs time (experiment 4)

5. Conclusions

Marl is an uncommon sculptural medium and therefore the study of its constitution and behaviour in a museum environment contributes to the limited references from the point of view of conservation. Marl is susceptible to deterioration because of the synergism of the factors involved. The mineralogical composition of the rock and the geological conditions under which it formed and consolidated have an irreversible effect on its preservation once taken from the ground. The condition of the marl fireplace in the Victoria and Albert Museum confirms the continuing deterioration of the rock. The characterisation of the material has proved crucial in determining a strategy for conservation. The study of weathering of the fireplace marl provided significant information that can be used both for restraining deterioration and for improving its stability. Specifically, deterioration had begun before the rock was quarried due to its geological nature and continued when the fireplace was used in its architectural setting. During storage the process continued as a result of a variety of factors in the domestic and museum environment. Marl is particularly responsive to fluctuations in relative humidity. Under ambient environmental conditions, it is subject to deformation as a result of gradual or abrupt changes in relative humidity, even within a narrow range. Thus, the primary aim in conservation


291

should be to establish and maintain a constant relative humidity in the environment where marl artefacts are stored and displayed. 6. Acknowledgements The research was carried out under the auspices of the V&A/RCA post-graduate Conservation Course, in collaboration with the Sculpture Conservation and the Sculpture Department of the Victoria and Albert Museum as part of their on-going conservation programme. We are indebted to Mr Martin Gill, Ms Elisabeth Morris and Mr Dick Giddens from the Royal School of Mines, Imperial College of Science, Technology and Medicine, to Mr Adrian Doyle from the Palaeontology Conservation Unit, The Natural History Museum, and to Mr Pedro Gaspar from the RCA/V&A Conservation Course who assisted with the technical preparation of material, analyses and scientific advice. 7. References Antoine, A., Giraud, A., Meunier, M., Van Ash, T., 1995. Geological and geotechnical properties of the "terres noires" in Southeastern France: weathering, erosion, solid transport and instability, Engineering Geology, 40, 223-234. Cornell, R. M., Aksoyoglou, E. S., 1991. Simultaneous determination of the cation exchange capacity and the exchangeable cations on marl. Clay Minerals, 26, 567-570. Cripps, J. C., Taylor, R. K., 1981. The engineering properties of mudrocks. Quarterly Journal of Engineering Geology, 14, 325-346. Deer, W. A., Howie, R. A. and Zussman, J., 1991. An Introduction to the Rock-Forming Minerals, Essex, Longman Scientific & Technical. Greensmith, J. T., 1989. Petrology of the Sedimentary Rocks (7 th edition), London, Unwin Hyman. Grim, R. E., 1968. Clay Mineralogy (2na edition), New York, McGraw-Hill. Hawkins, A. B., Pinches, G. M., 1992. Engineering description of mudrocks. Quarterly Journal of Engineering Geology, 25, 17-30. Kirsch, H., 1968. Applied Mineralogy (translated by K. A. Jones), London, Chapman and Hall Ltd. Millot, G., 1970. Geology of Clays (translated by W. R. Farrand and H. Paquet), New York, Springer-Verlag. Mitchell, J. K., 1993. Fundamentals of Soil Behavior, New York, John Wiley & Sons. Price, D. G., 1995. Weathering and weathering processes. Quarterly Journal of Engineering Geology, 28, 243-252. Weaver, C. E., 1989. Clays, muds and shales. Developments in Sedimentology 44, Amsterdam, Elsevier. 8. Equipment Micro climate generator: Microclimate Technology Inc., Preservation Equipment Ltd. Displacement transducer: Model D5/10G8, RDP Electronics Ltd. Displacement indicator: Model E309, RDP Electronics Ltd.


Theme 2 External factors of decay" environmental influence on stone


295

TOPOCLIMATIC CONSERVATION: PORTUGAL

MAPPING, A TOOL FOR C U L T U R A L HERITAGE THE CASE OF THE R O M A N THEATRE OF LISBON,

Lufs Alres-Barros Lab. Mineralogy and Petrology, Mining Dept., Av. Rovisco Pais, 1049-001, Lisbon, Portugal Am61ia Dionfsio Lab. Mineralogy and Petrology, Mining Dept., Av. Rovisco Pais, 1049-001, Lisbon, Portugal

Abstract In this paper some preliminary results regarding the microclimatic survey that is being made at Lisbon's Roman Theatre are presented 9 Some of the variables that are being monitored at different points in the archaeological site are near-surface relative humidity, dew point, air temperature and surface stone temperature. Preliminary conclusions show that it is imperative to take measures in order to maintain and preserve this site, since high variations of relative humidity, air temperature and stone surface temperature occur daily and along the year. Besides, and according to the actual conditions it rains in several parts of the Theatre, which can severely damage the artefacts. The most critical areas of the Roman Theatre are the SW comer and the central zone where are usually obtained the highest values of relative humidity and the lowest of dew point spread. These seems to be the areas that are more prone to form condensation and to allow micro biological life and weathering to occur. Key-words: microclimatic survey, topoclimatic mapping, conservation, Roman Theatre.

1. Introduction The Roman Theatre of Lisbon (Portugal), constructed or rebuilt during the 1~t century AD, was one of the most important parts of the Roman city Felicitas Olisipo Julia, and was dedicated to the emperor Nero. It is located on the hillside of St. George Castle, in the vicinity of Lisbon's Cathedral, slightly NE (fig. 1 and fig. 2). The space occupied by the Roman Theatre is located in a very populated urban area surrounded by buildings, in the confluence of two streets. The ruins of the Theatre were discovered fight after the big earthquake that occurred in Lisbon in the 18 Ih century (1755) and then again buried during the rebuild of the city. Only in our century, in the 60's decade, they were "'rediscovered" (fig. 3). The sector already excavated occupies approximately 285 m 2. According to archaeological studies this theatre is supposed to have a diameter of almost 60m. During the excavation works performed in the 60"s different types of artefacts (made of limestone, sandstone and marble) like columns, shafts, bases, statues, tiles, elements from the proscaenium and coins were discovered.. Some of these elements remain in this Author's to whom correspondence should be addressed

296


archaeological space in their original places. The remaining were stored in a warehouse and at Lisbon's Township Museum.

'Figure 1 (left): General view of the SW part of the Roman' s Theatre Ruins (Moita, 1994). Figure 2 (fight): Partial plant of Lisbon's city with indication of the Roman Theatre (Hauschild, 1990).

Figure 3: Plant of the Roman Theatre, after 1966-1967 excavations (Hauchild, 1990), with location of the points where the indoor microclimatic survey is being done. One of the most important and original parts of this theatre is the orchestra ground. This ground has a pattern shape, where different kinds of limestones with different colours were used (fig. 4). However only a small amount of this mosaic can be, observed nowadays, and most of the tiles present severe decay phenomena (fig. 5).

Figure 4 (left): Reconstitution of orchestra ground according to Hauschild, 1990. Figure 5 (fight): Two elements of orchestra's mosaic showing severe decay phenomena.


297

Since 1966, the area exposed and studied by the archaeological team is partially covered with a metal shed, that semi-protects it from the direct action of the weathering agents. The north and east elevations have no protection against meteoric agents. One of the main problems that needs to be solved by the people running this monument is not only the conservation of this space, where different types of materials coexist, but also to allow visitors without damaging the Roman monument. To evaluate possible causes of damage and to predict future deterioration, a microclimatic study is being carried out since October 1998 at the Ruins of Lisbon's Roman Theatre by our laboratory. This study addresses: 9 Evaluation of weathering decay phenomena and their intensity (through a qualitative method - visual assessment and through a quantitative method - direct ultrasound method. The results already obtained allow us to say that most of these artefacts are sound, slightly or moderately damaged. 9 Characterisation of air pollutants levels, rainwater properties, temperature and relative humidity, solar radiation, wind speed and direction of the outdoor environment (since these ruins are in a semi-sheltered regimen). Mostly of these data are obtained at Lisbon's Cathedral where our microclimatic monitoring station is located. 9 Characterisation of indoor environment through thermo-hygrometric parameters like near-surface relative humidity, dew point, air temperature and stone surface temperature that are being monitored non continuously at different points. The study of their spatial distribution enables the elaboration of topoclimatic maps. Through these maps, singularities like anomalies, intensity and shape of gradients can be visualised. It is also possible to study diurnal or seasonal variations. It is precisely this last theme that will be developed and discussed in this paper. As atmospheric variables continually fluctuate in a short period of time what is intended with this type of survey is to obtain the main trends, for example analyse if diurnal or seasonal variations occurs and check which are the most deleterious areas. The results and conclusions are preliminary. In what regards the indoor environment, it is also being continuously monitored the relative humidity and air temperature, at two different vertical levels (approximately at 2m and 5m).

2. Lisbon's environmental conditions Lisbon experiences a temperate maritime climate. During the year the temperature does not show great variations. The coldest months are January and February (mean minimum value 10~ and the warmest are July and August (mean maximum value 23~ The rainfall is scarce in Summer and very abundant in Winter, when monthly mean values reach 1 ! lmm. The pH and conductivity range values are comprised between 5.3 -7 and I-8.7 ~S, respectively. The average chemical composition of rain water is presented in tab. I. Table 1" Average mineral content of Lisbon's rain water (units are mgl -~) Na K Ca cr NO~ SO42 Mg Si 7.50 0.50 8.00 12.00 7.50 5.00 1.30 0.75

NH4 + 1.20

298


In what concerns the wind direction, it is mainly SW during winter whereas in summer is NW ("nortada" wind). Regarding other factors that can damage this archaeological site it must be considered not only the influence of marine spray (tab. 2), through the effects of wet and dry deposition on stones surface but also the air pollutants levels (fig. 6). Table 2" 85 percentile for the daily deposition rate (units are mg cm-2d-I) for dry deposition of the principal marine spray's ions (Dionfsio, 1999). NO3" ] SO4 z Ca Na K Mg CI 11.00 3.52 I 6.62 5.50 5.91 0.59 0.79

Figure 6: Hourly averaged concentration for relevant gases.

3. Methodology and analysis of the topoclimatic maps In order to get a better knowledge and to create the most suitable microclimate for conservation of this monument some thermo-hygrometric parameters (air and stone temperature, relative humidity, dew point, difference in degrees between the air temperature and the dew point, i.e., the dew point spread) are being analysed and monitored at different points of this archaeological site (fig. 3). The points (located on stone artefacts) are distributed without following a regular grid, but they cover approximately all excavated area. With exception of the sampling point n~ 6, all the points are in the same horizontal cross section. The survey it is performed not only in different days but also at different hours. To perform this type of discreet measurements it is being used a psychrometer positioned near the artefact surface, equipped with a thin insulating disk, to avoid the perturbation of the local air and a surface temperature probe. As a first approach it is assumed a simple geometry for the sampling area, i.e., without intermediate architectural structures. For the construction of the two-dimensional maps (topoclimatic maps), it was used the triangulation method with linear interpolation.


299

In fig. 7, 8, 9, l 0, 11, 12, 13 and 14 are presented some of the topoclimatic maps for the mean values of two variables - relative humidity and dew point spread - obtained in different days at different hours (the black balls corresponds to the sampling points that are precisely located at fig. 3).

Figure 7 (left): Near-surface relative humidity (%) at Lisbon's Roman Theatre. Measurement taken the 11 October 1999 at l 3.15. Figure 8 (right): Dew point spread (~ at Lisbon's Roman Theatre. Measurement taken the l I October 1999 at 13.15.

Figure 9 (left): Near-surface relative humidity (%) at Lisbon's Roman Theatre. Measurement taken the 14 October 1999 at 14. l 0. Figure 10 (right): Dew point spread (~ at Lisbon's Roman Theatre. Measurement taken the 14 October 1999 at 14.10.

Figure l l(left): Near-surface relative humidity (%) at Lisbon's Roman Theatre. Measurement taken the 7 December 1998 l 1.00. Figure 12 (right): Dew point spread (~ at Lisbon's Roman Theatre. Measurement taken the 7 December 1998 l 1.00.

300


Figure 13 (left): Near-surface relative humidity (%) at Lisbon's Roman Theatre. Measurement taken the 9 February 1999 at 10.45. Figure 14 (fight): Dew point spread (~ at Lisbon's Roman Theatre. Measurement taken the 9 February 1999 at 10.45. These graphics correspond to the main trends that seems to occurred at this monument, i.e., the central and SW area are the ones that usually have the smallest dew point spreads and highest relative humidity. The zones having the smaller dew point spread are probably the most deleterious since they are more prone to form condensation and to allow microbiological life and weathering to occur. The variations of the termo-hygrometric parameters during each survey at the different points, are not remarkable, for instance, the variation of relative humidity or dew point was always lower than 10% and 5%, respectively. However, at different days or at different hours in the same sampling point there are significant variations. The values are not kept homogeneous, which induces internal stress on the artefacts and contribute to their deterioration. Regarding the central area of the Theatre where most of the times it is attained the highest values of relative humidity and the lowest of dew point spread, it corresponds approximately to the area where an metallic element of the shed is missing and when it rains all this area is wetted. The SW area where the same phenomena occurs it is in the vicinity of one of the surrounding buildings it is also the area where the artefacts classified as moderately damaged are located.

4. Conclusions In this paper is presented the methodology that is being applied at Lisbon's Roman Theatre to describe its actual microclimate and to recommend, if necessary, the modification of some parameters, to properly conserve this ancient monument of our city and to allow the entrance of visitors. The results and conclusions here presented are preliminary and mainly correspond to the analysis of the isoline maps. Their analysis shows that it is imperative to take urgent measures to maintain and preserve this archaeological site. During this sampling period it was possible to verify that the indoor ambient conditions showed great variations in terms of air and stone surface temperature, dew point and air relative humidity. So, the first need for the conservation of this monument is to maintain it in a steady state climate.


301

According to the actual ambient conditions, the most deleterious areas of the Roman Theatre seems to be the SW comer and the central zone. It was in the central area that most of the times the highest and lowest values of relative humidity and dew point spread respectively, were obtained. The metal shed that is partially covering the excavations does not avoid great variations of air and stone temperature and relative humidity during the day or during the year. Besides, the ruins have two elevations that are directly exposed to weathering agents, which contributes to monument's damaging (through, for instance, the run-off and wetting of artefact's surface). These microclimate measurements will continue for a longer period (at least more six months) to be statistically representative and thus to give a more complete and accurate description of the Roman's Theatre environment variability.

5. Acknowledgements This study was partially financed by project PRAXIS XXI/P/ECM/12012/1998 and by the Township of Lisbon.

6. References Article reference: Dionisio A. et al., 1999. Deposiqfio de aerossol marinho na ~ e a urbana de L i s b o a - Um factor determinante no decaimento geoquimico das rochas dos monumentos. II Congresso lb6rico de Geoquimica/Xl Semana de Geoquimica, 157-160. Book reference: Camuf~b D., 1998. Microclimate for Cultural Heritage. Developments in Atmospheric Science, 23. Hauschild T., 1990. Das R~mische Theatre von Lissabon. Planaufnahme 1985-88. Verlag. Moita, I., 1994. Das origens pr6-hist6ricas ao Dom/nio Romano. O dom/nio romano, in O Livro de Lisboa, cap/tulo ll-B. Livros Horizonte, 35-68.


303

CHARACTERIZATION OF SURFACE MORPHOLOGY OF CARBONATE STONE AND ITS EFFECT ON SURFACE UPTAKE OF SO2

ElizaBeth A. Bede University of Delaware, 303 Old College, Newark, DE, 19716, USA

Abstract This laboratory study evaluates the role of receptor surface variables in the dry deposition process. It focuses on the effect of surface variables (i.e., porosity, texture, roughness, and effective surface area) on SO2 (g) uptake in ambient outdoor conditions (50 ppb; 65% RH; 4 m/s; 25~ Experiments are undertaken in a custom-built recirculating environmental deposition chamber (Spiker, 1992(1)). The primary components of this project include developing a surface characterization methodology and evaluating the deposition velocity of SO2 (g) onto limestones. Four U.S. domestic limestones (Salem; Cordova Cream; Cottonwood Top Ledge; Tennessee Pink) and one imported limestone (Monks Park, GB) are utilized in this study. Both polished and simulated weathered surfaces are examined to gauge the effect over a range of surface variables. The understanding of the role of surface morphology in the pollutant uptake process from this study will be used to develop a methodology to assess the long-term effects of surface treatments (i.e., cleaning, consolidation) of buildings and outdoor sculptures. Key words: dry deposition, sulfur dioxide, limestone, surface morphology

1. Introduction The deterioration of carbonate stone is greatly accelerated by the action of air pollutants such as SO2,NOx, H2SO4,HNO3, 03 and particulates (Amoroso 1983; Hoffmann 1986; Butlin 1991; Haneef 1992). While the deleterious effects of these atmospheric pollutants on stone was first cited over a century ago (Voelcker 1864), the apparent increase in stone degradation has been attributed to the increase in anthropogenic pollutants (Shaffer 1932; Babu Rao 1983; Livingston 1983). This is supported by observations of damage much more severe in urban than rural areas (Feddema 1987; Vella 1996). Of these pollutants the sulphurous compounds have been identified as playing a major role in the deterioration of carbonate stone (Winkler 1966; Spedding 1969; Braun 1970). Dry deposition is governed by the mass transfer in the gas phase and by adsorption and absorption at the surface. In the case of carbonate stone the latter results in an irreversible chemical reaction. Of primary concern to architectural conservators is the formation of the alteration product, gypsum, resulting from this reaction. This mechanism of stone decay occurs in protected areas such as under stone sills and cornices, or in less protruding portions of sculptures or monuments. The effect of dry deposition in these protected areas is a slow continuous process that eventually leads to catastrophic surface losses from spalling and delamination. Simplistically, the overall dry deposition process is often presented in terms of three resistance steps. In the first step, the pollutants are transported from the lowest levels of the atmosphere into the very thin quasi-laminar sublayer, sometimes referred to as the viscous sublayer, that surrounds each object. This is termed as the aerodynamic resistance, ra, and is

304


governed by wind speed and air turbulence. The second step, known as the boundary layer resistance, rb, requires just that, breaching the boundary layer turbulence and molecular diffusivity. The final step is the physical and/or chemical capture/reaction of the pollutant with the receptor surface and is termed surface uptake resistance, re. Surface chemistry, wetness, roughness, and porosity control this resistance. Each step may be rate limiting and occur at varying speeds depending on the characteristics of the system. The sum of these three resistances is inversely proportional to the deposition velocity and can be stated as: Vd - - ( r a + r b + rc) "1

(1)

ra and rb Can be estimated by micrometeorological data and chambers have been constructed in order to define and control these resistances (Spiker 1989; Preston 1972; Johnson 1990; Lewry 1992; Girardet 1996(1)(2)). rc is dependent on the surface properties of the stone and is relatively poorly understood (Wu 1992). Hence, it is the magnitude of re that distinguishes the rate of dry deposition for different materials. Methods for measuring deposition velocities on materials are reviewed by Lipfert (1989(2)). Primary dry deposition of 8 0 2 onto carbonate stone has been shown to vary from one to two magnitudes of order (Lipfert 1989(1); Furlan 1992). It is dependent on the environmental factors and the characteristics of the receptor surface. More specifically, dry deposition of SO2 onto carbonate stone is intimately related to: 9 SO2 concentration; 9 boundary layer characteristics (i.e., temperature, moisture content, turbulence); 9 temperature and humidity gradients of the receptor surface; 9 the presence of environmental and/or surficial catalysts; 9 and surface morphology (i.e., roughness, porosity, effective surface area). These parameters and their subsequent damage have been well documented in field studies. In addition, numerous laboratory studies have been undertaken to isolate the various environmental components in order to determine their contributions to primary dry deposition process. However, to date the interdependence of re and morphological properties have only been tangentially investigated and are poorly understood. This study focuses on the effect of some surface and bulk properties of carbonates stone (primarily porosity, texture, roughness and effective surface area) o n SO2 uptake in simulated outdoor conditions. The long-term aim of this research is to gain an understanding of this relationship that will enable conservators to evaluate conservation treatments more effectively. Conservation treatments, such as consolidation and cleaning, change the effective surface. An understanding of how this may or may not affect the long-term pollutant uptake of the stone will aid in choosing appropriate treatments.

2. Research Goals And Parameters The primary goal of this research is twofold:

".9 To determine the most appropriate methodology for characterizing surface morphology variables (e.g., surface texture parameters; effective surface area; porosity; pore size, shape and spatial distribution; permeability; capillarity). ".9 To utilize quantitative surface uptake resistance data to evaluate the effect of these


305

variables on primary deposition of SO2 under controlled ambient conditions. The overriding design parameter is to simulate field conditions whenever possible. The choice of test specimens, surface textures, sample preparation and chamber conditions are dictated by this criterion.

3. Sample Preparation 3.1 Stone types Five commercially available limestones that are commonly utilized as structural or ornamental elements in outdoors cultural resources were chosen. Since the focus of this study is to evaluate the effect of the physical properties of the surface on pollutant uptake it is desirous to limit the number and types of chemical interactions occurring on the surface. Therefore, the limestones considered were limited to one category -- calcitic. One imported and four domestic United States limestone types were chosen. These are: 9 Salem limestone (Bedford, Indiana) 9 Cottonwood Top Ledge Limestone (Chase County, Kansas) 9 Cordova Cream Limestone (also known as Austin stone) (Austin, Texas) 9 Tennessee Pink Marble (highly crystalline limestone; orthomarble) (Knoxville, Tennessee) 9 Monks Park Limestone (also known as Bath stone) (Wiltshire, Great Britain) Salem limestone was utilized for the US NAPAP studies and for similar international studies. In addition, in depth petrographic (McGee 1989) and porosity (Leith 1996) analysis are available. Monks Park Limestone was chosen since it is a standard test limestone in researched sponsored by the UNECE. In addition it has been widely utilized and characterized by research conducted by Great Britain's National Materials Exposure Program. Inclusion of this limestone type will allow for correlation and comparison with similar international dry deposition studies.

3.2 Texture categories In the initial phase, two texture categories are utilized: a smooth and a rough texture. The smooth and rough surface textures are achieved by polishing and acid etching respectively. Etching was chosen since it most closely simulates the preferential attack of the carbonate grains and matrix of weathered surfaces under field conditions. The goal was to achieve a similar surface roughness parameter (Ra) for each texture category. The roughness parameter is determined by 3-D non-contact laser profilometry (Proscan 1000, Micro Photonics). The limestones are cored utilizing a water-cooled thick-walled drill bit and cut to approximately 1/4" (0.64 cm) thickness. Due to possible variations within the stone, cores are taken as close together as possible. In order to remove all previously deposited and inherent sulfate the test specimens are placed in a constant temperature deionized water bath (43 C) until a constant sulfate background level is attained. The deionized water was changed daily and sulfate levels were determined biweekly by ion chromatography (Dionex DX-500). Up to six months of bathing was required for some limestones to reach background. All test sample surfaces are hand-polished lightly with successive grit sized silicon carbide paper to remove saw marks or protruding inclusions and to provide a level surface for polishing

306


or etching. Polishing residue is removed by three successive 15-minute ultrasonic baths in deionized water followed by a 5-minute bath in methanol. Test specimens are then dried in an oven at mild temperatures. To achieve the 'smooth' texture a polishing wheel with various grades of polishing cloths was employed. To etch the test specimens in the rough texture category a 23% solution of glacial acetic acid in water was utilized. In order to insure a uniform etch across each individual surface a stainless steel apparatus was designed. It allows the simultaneous etching of up to eight test specimens. Each limestone type required a different etching time to achieve a universal roughness parameter for all the test specimens in the rough texture category. At the completion of these processes each sample was rinsed thoroughly for five minutes in deionized water. The test specimens were then placed in an ultrasonic bath with deionized water for three successive baths at 15-minute intervals and followed by oven-drying overnight at a mild temperature. The surface morphologies of these test specimens were characterized utilizing the methods described below.

4. Surface Morphology Evaluation The surface morphology of a stone strongly influences the uptake of SO2. Porosity and microporosity affect the manner in which moisture and pollutants are transported through the stone, as well as, influence both the development of moisture films and the sites of reaction. Even for dry stone, the more porous the surface the greater the reaction with pollutants. In addition, the total surface area available for uptake and reaction of pollutants increases with increasing surface roughness. However, the dry deposition mechanisms at work on nonhomogeneous stone surfaces have not been systematically studied. It is the aim of this research to study the interactive relationships of surface morphology parameters (e.g., porosity, roughness, and surface area) with uptake rates in order to correlate a better insight into the SO2 deposition process. The first step towards this goal is the identification and utilization of techniques to evaluate pore structure and surface texture. 4.1 Pore structure evaluation The following methods are utilized to analyze pore structure and geometry of the limestone test specimens. Specifically the size, shape and distributions of the pores in relation to open porosity. Meng (1993) has developed a similar characterization for sandstone.

Thin Section Analysis utilizing epifluorescent microscopy (Leith 1996) 9 evaluate pore network structures (inter- vs. intragranular) 9 quantify and map spatial distribution of macropores (> 50 nm) 9 determine pore shape (i.e., round, inkbottle) and distribution 9 Scanning Electron Microscopy 9 evaluate fine pore network structures (inter- vs. intragranular) 9 quantify and map distribution of mesopores (2-50 nm) and micropores (< 2 nm) 9 determine pore shape (round; inkbottle) and distribution # Photomicrographs 9 illustrate pore geometry and pore connectivity [ 1]

9th International Congress on Deterioration and Conservationof Stone, Venice 19-24June 2000 r

Mercury Intrusion Porosimetry 9 determine pore entry size distribution 9 quantify pore surface area 9 relate pore surface area to pore volume

r

Nitrogen Adsorption Porosimetry 9 determine pore size and distribution 9 quantify surface area distribution of smallest pores

307

Pore-Surface Fractal Measurements (Mossotti 1992) [2] 9 define the effective surface capacity 9 determine pore type 9 ascertain distribution of pores and mass

4.2 Surface Texture Characterization The following were utilized to quantitatively characterize surface roughness and surface area in order to evaluate qualitatively and differentiate the surficial parameters affecting pollutant uptake rates. Nitrogen Gas Adsorption 9 quantify surface area and specific surface area (SBET) 3-D Non-contact Laser Profilometry 9 generate a quantitative topographic map 9 quantify surface parameters (i.e., roughness (Ra); peak to valley height (Rz); peak spacing (Rsm); peak count (Pc); material ratio (tp) Surface Fractal Measurement [2] 9 calculate surface roughness

5. Primary Deposition Study Primary deposition of S02 on a variety of stone surfaces was conducted under controlled ambient conditions in an environmental deposition chamber. The surface uptake resistances and their corresponding surface morphology characterizations for each sample are compared with the goal of identifying those surface variables that dominate dry deposition mechanisms.

5.1 Deposition chamber The experimental environmental deposition chamber utilized for this study was designed by Spiker et al. (1989). The chamber is essentially a recirculating flow wind tunnel where ra + rb is defined and constant within the design parameters. Temperature, relative humidity, wind speed, and pollutant (SO2, NOx, O3) concentration are computer controlled and monitored within the chamber by commercially available data acquisition cards and computer software (LabTech). These values are utilized to calculate rc and subsequently Yd. The elliptical chamber is approximately 15' (4.6 m) in length, 7' (2.1 m) wide and stands approximately 6' (1.8 m) high. It utilizes an electric steam generator to produce and maintain

308


relative humidity; an industrial recirculating cooling unit for temperature control; an industrial fan for circulation and wind speed; and a pump pack (Monitor Labs 8551) for gaseous SO2 introduction. The custom-design airfoil test specimen shelf allows for analysis of eight test specimens per run (four on top; four on bottom). The test specimens are discs of 1.5-inch diameter. An SO2 analyzer monitors SO2 levels in the chamber with an ultraviolet fluorescent detector (Monitor Labs 8660). A PITOT tube assembly (Baratron) measures the wind speed in the tunnel. Dew point cells and temperature probes gauge relative humidity and temperature. 5.2 Environmental conditions Based on previous studies and in an attempt to replicate, as closely as possible, ambient outdoor conditions, the following parameters are monitored in the chamber: Relative humidity: 65% Temperature: 25 C SO2: 50 ppb Wind speed: 4 rn/s These parameters replicate the general northeastern United States environment. These conditions remain constant throughout each 24-hour exposure run. 5.3 Overall methodology To degrease the surface and lower its surface tension the test specimens are uniformly sprayed with methanol. Test specimens are dried in an oven at mild temperature (less than 50 C) for a minimum of an hour followed by equilibration to 65% RH in a desiccator with conditioned silica gel for a minimum of 24-hours. Nine test specimens comprise a run. Eight are exposed in the chamber and one remains in the desiccator as a blank. The eight test specimens exposed consist of two limestone types and the two texture categories. For example a typical run might contain: 2 smooth textured Monks Park Limestones; 2 rough textured Monks Park Limestones; 2 smooth textured Salem Limestones; and 2 smooth textured Salem Limestones. Each of these test specimens is run in each position on the chamber shelf- a total of eight runs per test specimen group. This sample placement compensates for any anomalies of exposure along the airfoil shelf and for fluctuations in the environmental conditions over the 24-hour period. [3] At the completion of the 24-hour run the test specimens are removed expeditiously from the chamber and placed in 60 ml. screw top containers. 30.0 ml 0.6% H202 solution added to each container to oxidize and leach the sulfur deposited on the surface during exposure. The blank is removed from the desiccator and also placed in a 30.0-ml solution. The nine test specimens are placed on a Reax 3 platform mixer for 24 hours. The asymmetrical movement of this mixer causes differential pressures within the pores of the stones as the leachate levels rise and fall thus providing effective and thorough washing of the stone's substructure. The leachate is poured off and a P81 filter paper (Whatman) is placed in the leachate to provide solid phase cation exchange. This pre-filtering was necessary in order to reduce the number of calcium ions in the solution prior to analysis by ion chromatography. The leachate is placed again on the platform mixer for an additional 24 hours. An additional 30.0-ml H202solution is added to the original containers with test specimens and a second 24-hour leach is conducted followed by the same 24 hour filtering process. Simplex Optimization was utilized to develop this protocol. [4] Ion chromatography is performed on leachate to determine the ppb of the SO42. Sulfate levels of both washes are corrected in accordance with the established background levels of the individual test specimens and then added to provide the total amount of deposited


309

sulfate. The depositional velocity (Vd) and the surface resistance (re) are calculated.

6. Conclusions The correlation and comparison of this data with the surface morphology characterizations is currently underway. Preliminary investigations suggest that deposition increases with increasing roughness. Furthermore it appears that the characteristics of porosity are paramount factors in the deposition process. Preliminary analyses demonstrate that numerous factors seem to act, not independently, but rather in a complicated system of mutual interferences. It is however, the task of conservation scientists to unravel these systems and decipher the role of each dominant variable. Through this understanding effective mitigating procedures and/or appropriate treatments are developed.

7. Acknowledgements Funding for this research was generously provided by the National Center for Preservation Technology and Training, Materials Research Program and the University of Delaware, Art Conservation Research Program. This paper presents portions of an in-progress dissertation towards the Doctorate of Philosophy, University of Delaware, Newark, DE, USA.

8. Endnotes [1 ] To date there is not a method for quantitatively assessing these features. Since they are conceptually highly significant in controlling SO2 uptake and subsequent gypsum formation rates, photomicrographs are utilized to illustrate these properties and a descriptive methodology is under development. [2] MORPH-I and MORPH-II are components of a software package that analyzes SEM micrographs for the assessment of the fractal dimension of pores and a surface edge respectively. Modifications of these programs were made for the purposes of this research to enable the assessment of thin sections. (Mossotti, V., Eldeeb, R., 1998. USGS Open File Report 98-248; http://caldera.wr.usgs.gov/OF98-248.) [3] Evaluation of the data produced from these runs in underway with the assistance of a Ph.D. statistician. The design of future runs, if necessary will follow this evaluation. [4] Rebus Simplex Optimization is a multivariate, non-parametric, non-linear computer software program for response optimization. (Simplex Optimization, WindoChem Software, Inc., USA, 707-864-0845)

9. References Amoroso, G., Fassina, V., 1983. Stone Decay and Conservation: Atmospheric Pollution, Cleaning, Consolidation, and Protection. Elsevier, New York. Babu Rao, G., 1983. Effect of pollution by sulphur dioxide on marble and sandstone. Journal of Archaeological Chemistry: 1, l, 31-38. Baedecker, P. A., et al., 1990. Effects of acidic deposition on materials. NAPAP Report 19, Acidic Deposition: State of Science and Technology. NAPAP, Washington, DC.

310


Braun, R. C., Wilson, M., 1970. The removal of atmospheric sulphur by building stones. Atmospheric Environment, 4, 371-378. Butlin, R. N., 1991. Proceedings of the Royal Society of Edinburgh: 97B, 255-272. Camuffo, D., M. Del Monte, and C. Sabbioni, 1983. Origin and growth mechanisms of the sulfated crusts on urban limestone. Water, Air and Soil Pollution, 19, 351-359. Fassina, V., 1988. Environmental pollution in relation to stone decay" Air Pollution and Conservation: safeguarding our architectural heritage. Elsevier, New York, 133-174. Feddema, J. J., Meierding, T. C., 1987. Marble weathering and air pollution in Philadelphia. Atmospheric Environment, 21, 1, 143-157. Furlan, V., Girardet, F., 1983. Considerations on the rate of accumulation and distribution of sulphurous pollutants in exposed stones. Materials Science and Restoration: proceedings of the international congress, Filderstadt, Lack und Chemie, 285-290. Furlan, V., Girardet, F., 1992. Atmospheric pollution and reactivity of stones. Proceedings of the 7th Intemational Congress on Deterioration and Conservation of Stone, Lisbon, Laborat6rio Nacional de Engenharia Civil, 153. Gilardi, E. F., 1965. Absorption of Atmospheric Sulfur Dioxide by Clay Bricks and Other Building Materials. Unpublished doctoral dissertation, Rutgers State University, NJ. Giradet, F., et al., 1996(1). Experimental study of sulfation of a limestone and an calcareous sandstone in the atmospheric chamber of Laussanne. The 8th International Congress on Deterioration and Conservation of Stone, Berlin, 349-358. Giradet, F., et al., 1996(2). Reactivity of stones to atmospheric SO2: study in simulation chamber and correlation with measurements in situ. The 8th International Congress on Deterioration and Conservation of Stone, Berlin, 341-347. Hamilton, R. S., et al., 1995. Sulphur and nitrogen particulate deposition onto building surfaces. The Science of the Total Environment: 167, 57-66. Haneef, S. J., et al., 1992. Effect of dry deposition on NOx and SO2 gaseous pollutants on the degradation of calcareous building stones. Atmospheric Environment, 26A, 16, 29632974. Hoffmann, M. R., 1986. Fog and cloud water deposition. Materials Degradation Caused by Acid Rain, Washington, D.C., 64-91. Honeyborne, D. B., Harris, P. B., 1958. The structure of porous building stone and its relation to weathering behavior. Colston Papers. The Structure and Properties of Porous materials: proceedings of the 10th symposium of the Colston Research Society: 10, New York: Academic Press. Johnson, J. B., et al., 1990. Laboratory exposure systems to simulate atmospheric degradation of building stone under dry and wet deposition conditions. Atmospheric Environment, 24A, 10, 2585-2592. Leith, S. D., et al., 1996. Limestone characterization to model damage from acidic precipitation: effect of pore structure on mass transfer. Environmental Science & Technology, 30, 7, 2202-2210. Lewry, A. J., et al., 1992. A chamber study of the effects of sulphur dioxide on calcareous stone. Proceedings 7th International Congress on the Deterioration and Conservation of Stone, Lisbon, Laborat6rio National de Engenharia Civil, 641-650. Lipfert, F. W., 1989(1). Atmospheric damage to calcareous stones: comparison and reconciliation of recent experimental findings. Atmospheric Environment, 23, 2, 415-429. Lipfert, F. W., 1989(2). Dry deposition velocity as an indicator for SO2 damage to materials. Journal of Air Pollution Control Association: 39, 446-452.


311

Livingston, R. A., Baer, N. S., 1983. Mechanisms of air pollution-induced damage to stone. Proceedings of the 6th World Congress on Air Quality, Paris, SEPIC. McGee, E. S., 1989. Mineralogical studies of the Shelburne Marble and Salem Limestone. U. S. Geological Survey Bulletin, Denver, GPO. Meng, B., 1993. Characterization of pore structure for the interpretation of moisture transport. Conservation of Stone and Other Materials: Proceedings of the International RILEM/UNESCO Congress, New York, E & FN Spon, 155-162. Mossotti, V. G., Eldeeb, A. R., 1992. The fractal nature of Salem Limestone. Proceedings of the 7th International Congress on Deterioration and Conservation of Stone, Lisbon, Laborat6rio Nacional de Engenharia Civil, 621-630. Preston, P. E., 1972. An industrial atmosphere test chamber. Transactions of the Institute of Metal Finishing, 50, 125-131. Serra, M., Starace, G., 1978. Study of the reactions between gaseous sulphur dioxide and calcium carbonate. Deterioration and Protection of Stone Monuments, Paris, Reliure. Shaffer, R. J., 1932. The Weathering of Natural Building Stones, DSIR Building Research Special Report 18, London, HMSO. Spedding, D. J., 1969. Sulphur dioxide uptake by limestone. Atmospheric Environment, 3,683-684. Spiker, E. C., et al., 1989. Environmental chamber for study of the deposition flux of gaseous pollutants to material surfaces. Open-File Report 89-296. Washington, DC. Spiker, E. C., et al., 1992(1). Environmental chamber for study of the deposition flux of gaseous pollutants to material surfaces. Atmospheric Environment, 26A, 16, 2885-2892. Spiker, E. C., et al., 1992(2). Dry deposition of SO2 on limestone and marble: the role of humidity. Proceedings 7th International Congress on the Deterioration and Conservation of Stone, Lisbon, Laborat6rio Nacional de Engenharia Civil, 397-406. Spiker, E. C., et al., 1995. Laboratory study of SO2 dry deposition on limestone and marble. Water, Air and Soil Pollution, 85, 4, 2679-2685. Vella, A. J., Camilleri, A., Adami, J. P., 1996. Limestone surfaces in built-up environments as indicators of atmospheric pollution. Environmental Geochemistry and Health, 18, 165-170. Voelcker, J., 1864. On the injurious effects of smoke on certain building stones and on vegetation. Winkler, E. M., 1966. Important agents of weathering for building and monumental stone. Engineering Geology, 1, 5, 381-400. Wu, Y. L., et al., 1992. Aerosol Science and Technology, 16, 65.

10. Materials

Conditioned silica gel: Art Preservation Services, 235 East 85 th Street, Suite B2, New York, NY 10028; (212) 988-3869 Proscan 1000 Non Contact 3D Profilometer: Micro Photonics, PO Box 3129, Allentown, PA 18106; (610) 366-7105; www.microphotonics.com


313

SEA WATER ABSORPTION, PERMEABILITY EVOLUTION AND DETERIORATION ASSESSMENT OF BUILDING STONES SUBJECTED TO

MARINE EXPOSURE Jean-Marc Birginie Laboratoire de Construction Civile et Maritime, Universit6 de La Rochelle (IUT), La Rochelle, France

Abstract

The deterioration effect developing on building stones subjected to marine exposure, results mainly from the salt crystallisation, which occurs during the repeated drying and wetting phases. This phenomenon can be artificially re-created on samples in salt spray dynamic simulators. By recourse to these accelerated ageing tests, it is possible to characterise the general form of the evolution of the stone regarding the seawater accumulation, the incubation time before deterioration and the intensity of degradation. In this paper, the effectiveness of two non destructive techniques used to quantify, on one hand, the stone deterioration (visual scanning of the surface) and, on the other hand, the transfer properties evolution (measurement of permeability to air), is shown in the case of Richemont, Tuffeau, and Sireuil, three calcareous stones from West of France. The experimental results emphasise an acceleration of seawater accumulation before the developing of deterioration. It is also shown that the evolution of the stone is very dependent on the salt spray quantity received by the samples during each cycle of ageing. In order to anticipate the decay of monument stones due to the marine pollution, a general approach consists in determining the critical salt content which creates the surface disintegration, according to the time of exposure, as well as the distance from the sea. By associating the results obtained from accelerated ageing tests and from sampling on site, this general presentation of the evolution of the stone may be determined for representative lithotype in order to provide an assessment tool useful for the intervention on monuments. Keywords : building stone deterioration, permeability to air, accelerated weathering.

marine

exposure,

degradation

analysis,

I. Introduction

The deterioration of building stones subjected to atmospheric pollution, such as marine salt, depends mostly on the time of exposure, as well as on the distance of the monument from the pollution source (Auger, 1991). Several other factors concerning the stone itself, mainly the porosity and the permeability, and the climatic parameters play an important role on the deterioration evolution (Auger, 1996). The eharacterisation of lithotypes with respect to their initial microstructure and their initial transfer properties, can provide us information regarding the sensitivity of materials to atmospheric pollution (Meng, 1994; Parrott, 1994). It is also important to follow the evolution of the above mentioned characteristics during the exposure because, for example, transfer properties modifications are usually precursory of the deterioration processes. In

314


particular, near the marine zones, the sea salt accumulation at various concentrations in the stone, causes irreversible transformations on the microstructure which affect the kinetics of mass transfer. The deterioration effect by salt crystallisation that results from repeated wetting and drying phases, can be greatly amplified or reduced. When the environmental conditions are optimal, severe sand disintegrations can develop on the stone surface, favouring depth damages (Auger, 1996; Birginie, 1999). Ageing tests carried out in salt spray dynamic simulation (Auger, 1988), confirm that the disintegration intensity developing on the surface is very dependent on the micro-structural aspect of material and its fluid transfer properties such as liquid water progression (wetting phase) and vapour diffusion (drying phase) In this paper, three non destructive methods are proposed to assess the evolution of porous building material, with respect to the mass absorption and the surface deterioration, and their effectiveness are compared through the results of ageing test by salt spray concerning three French limestone : Richemont, Tuffeau and Sireuil. On the basis of this results and previous works (Auger, 1991), a global approach is then proposed with the objective to forecast the salt accumulation and the decay in the stones.

2. Accelerated ageing test and sampling The weathering in dynamic simulation consists in creating cyclic atmospheric conditions by repeated drying and salt spray phases, so that the disintegration phenomena by salt crystallisation are developed quickly on the surface of the stone. In order to accelerate the deterioration from a natural exposure, the temperature in the simulator is maintained at 40~ and the time of cycle reduced to half hour, that corresponds to an average cycle of one week under the South European climate. The cycle in simulation includes one minute of salt spray and 29 minutes of drying. The ageing procedure is explained in detail in (Auger, 1998) and has received the recommendation from the AFREM (French association of material testing ) in 1996. Core samples were extracted from blocks of two fine grain calcareous stones, Richemont (open porosity : 25%) and Tuffeau (porosity : 45%), and a coarse grain limestone, Sireuil (porosity of about 35%). The cylindrical samples of 50 mm diameter and 100 mm high, have been exposed to dynamic simulation with five different salt spray intensities corresponding to the following seawater quantities received, per cycle, by each sample : 0.3, 1, 2, 5 and 8 mg/cm 2. The evolution of material is appraised from the use of three non-destructive analysis: weight variation, surface analysis by camera-laser scanning and permeability measurement.

3. Non destructive analysis 3.1 Mass variation of samples In an ageing simulation, the weight evolution of samples can be checked during the test. The provided results give information about the solution accumulated into the samples. In fact, when the surface disintegration occurs, the weight variation corresponds to the difference between the mass gain due to the solution absorption, and the mass loss due to the loose grains on the surface of samples by effect of salt crystallisation. For this reason, it is difficult to deduce, in all case, the solution content evolution using the weight variation


315

measure. If the sequence of cycles is constant in the time, the beginning of disintegration could be approximately located on the point of mass curves where the sample weight reaches a maximum value before decreasing. Graphic of Figure 1 shows that, in the case of Richemont stone, the mass variation is very dependent on the solution quantity received per cycle, by the stone surface. A high salt spray pulverisation (8 mg/cm 2) leads to the quick saturation of the stone, preventing the salt crystallisation. The optimal pulverisation that favours a fast disintegration in depth seems to be situated between a level of 1 to 3 mg/cm 2 per cycle.

Mass evolution according to the received salt spray (Richemont)

Figure 1 : Influence of salt spray quantity received by the Richemont samples on the evolution of mass. Continuous lines : supposed mass curves without the loss due to the disintegration

3.2 Evolution of surface aspect (visual analysis) A more direct method to quantify the disintegration consists in a surface analysis by using a camera-laser scanning system (Birginie, 1999). The system provides a relief matrix (topographic data) and a laser light reflection matrix (evolution of reflection properties of the surface such as colour change). The principle of the relief analysis is based on the use of a triangulation technique between the laser light source, the CCD camera and the surface to be analysed. In previous works (Birginie, 1999), the authors showed that it is possible, by the use of the above mentioned device, to characterise objectively the degradation morphology. In other respects, the deterioration can be quantified computing global statistical parameters from matrix, such as the standard deviation. Regarding the relief data, the standard deviation corresponds to a surface roughness criteria.

316

9th International Congress on Deterioration and Conservation of Stone, Venice 19-24June 2000 a) Evolution of surface disintegration (Richemont)

Figure 2 a-b 9Detection of the stone deterioration threshold from the visual analysis of the surface roughness" a) Richemont, b) Tuffeau


317

This global statistical parameter variation is linked to the evolution of degradation developing on the surface during the simulation (Birginie, Rivas, 1999). For example, figure 2 a-b compares, by means of this statistical criteria, deterioration suffered by the Richemont and Tuffeau samples for the intensities of salt spray that have favoured the sand disintegration effect. These graphics shows that the evolution of deterioration is very dependent on the salt spray pulverisation and follows different steps of progression. By a fixed threshold of roughness (initial roughness + 0.2 ) the significance of disintegration can be considered. At that point, it is easy to locate, for each salt spray intensity, the corresponding number of weathering days need for the disintegration phenomena to develop of significant manner.

3.3 Evolution of permeability to air

Measurement of permeability to air gives a very precious result, as this transfer property of the stone is linked to the pore obstruction due to the solution accumulation. Modifications which happen on the sea water penetration process, influence directly the evolution of the permeability to air. This physical property is measured by letting pressurised air on core samples and by gauging the inlet pressure P and air flow Q. A suggested formula including P, Q, the stone dimensions and the gas viscosity vl, gives the global permeability of the sample in m s. The fact that the measurement method is easy and fast to achieve, is very important to prevent the micro-structural transformations due to the air flow into the material. Unlike the weight measurement, the permeability is much less dependent on the disintegration effect and it varies essentially according to the presence of salt solution in the pores. Graphic of figure 3-a shows the evolution of permeability in the case of Richemont stone exposed in dynamic simulation to different salt spray intensities. These curves well emphasise the incubation times followed by a decreasing of permeability corresponding to the acceleration of salt accumulation due probably to the hygroscopic effect of salt. After these two steps, the samples tend to saturate when they receive a strong salt spray and seem to reach a stability zone with respect to the permeability property. Figure 3-b tends to show the same evolution feature in the case of lithotype Tuffeau but, due to its higher open porosity, the decrease of permeability began later and this stone reaches the saturation more slowly than for Richemont. In order to deduce the solution accumulated and salt content in stone from the permeability measurement, it is necessary to establish an empirical curves and a mass transfer modelling. These physical data require several previous test achieved on each considered lithotype.

318


a) Evolution of permeability to air according to the received salt spray

Figure 3 a-b: Influence of the salt spray quantity received by samples, on measurement of permeability to air" a) Richemont, b) Tuffeau


319

4. Global approach of the marine salt accumulation effect 4.1 General presentation of results The susceptibility of the three stones subjected to different seawater spraying in cyclic conditions, can be characterised by two types of parameter, taken from the preceding analysis: - variation of permeability and roughness at 35 days of ageing test - day of the simulation corresponding to a significant variation of roughness and permeability.

Table 1: Characterisation of evolution of the three limestone studied, from the visual analysis and the permeability measurement. Visual analysis (roughness) Spraying 0.3 mg/cm2

1 mg/cm2

3 mg/cm2

5 mg/cm2

8 mg/cm 2

Sat. Richemont 0.3 / 26 1.1 / 14 1.5 / 4 Sat. Sat. Tuffeau 0.1 / 48 1.3 / 21 2.2 / 10 3.9 / 2 Sat. Sireuil 0.3 / 17 0.9 / 17 1.5 / 8 1.1 / 8 First value : deterioration at the 35 m day of simulation test (Threshold, Fig. 2) ( = roughness - initial roughness ) Second value : days of simulation at 10% increase from the initial roughness Permeability Spraying 0.3 mg/cm2

1 mg/cm2

3 mg/cm2

5 mg/cm2

Richemont 10% / 50% / 23 85% / 9 Sat./5 Tuffeau 0% / 20 % / 45 50% / 28 Sat./18 Sireuil 0% / 30 % / 31 50% / 26 70% / 17 First value 9Permeability drop (in %) at the ~35m day of the simulation test Second value : days of simulation at 30% decrease of permeability

8 mg/cm2 Sat.~3 Sat./ 9 Sat./ 7

Through these results, it appears that the salt spray quantity which optimises the sand disintegration on Richemont and Sireuil is about 2 to 3 mg/cm 2 per cycle, and 5 mg/cm 2 for Tuffeau. Beyond these spraying levels, the material saturates, that prevents the disintegration phenomenon. The less porous the stone (Richemont, 25% of porosity) the more rapid the saturation. The more porous (Tuffeau, 45% of porosity) the more depth and rapid the deterioration. When the permeability of stone reduce to 30% before the first ten days of simulation, it means that the sample is going to saturate without disintegrating.

4.2 Relationship between natural exposure and simulation On site, the distance from the coast is a determining factor to forecast the quantity of salt present into the material, when exposed to salt contamination that originates from marine spray. Due to the fact that the quantity of salt provided by the atmospheric environment is

320


cumulative in the time, the exposure duration is also a fundamental factor to describe the evolution of building stones. In a preceding paper, Auger (1991) proposed the use of a three-dimensional diagram including distance from the sea and time, in order to anticipate the possible quantity of marine salt found in monument stones, and the resulting deterioration. Considering that the experimental data available on site are very insufficient, the proposed numerical approach is more an extrapolation based on experience, some localised experimental data and the knowledge of the stone behaviour regarding the diffusion phenomena, than an exact modelling. This 3D plot could be extrapolated from the preceding results of permeability measurement, associating the number of cycle of accelerated ageing to the real time of exposure and the received salt spray quantity to the distance from sea. The analogy between simulation and natural exposure from the sea in our region can be defined by the following linear correspondence : - one day in simulation (48 cycles) is equivalent to one year in natural exposure - the relationship between the intensity of pulverisation and the distance from the sea depends on the specificity of site of exposure, in particular the wind direction and celerity. This analogy is mainly established on the base of the comparison of deterioration observed on site on samples aged in simulation. Experience shows that the sand disintegration can appear when the salt concentration reaches the value of about 500 mg per 100 g of dry stone (Auger 1991). We underline that the porosity and the transfer properties of the stone can influence a lot this critical salt content. In order to make a systematic study with respect to the specific characteristics of each lithotype and exposure conditions, it is necessary to work on samples subjected to accelerated ageing in a salt spray simulator with controlled atmospheric conditions.

5.

Summary

and

conclusions.

In the case of accelerated weathering test achieved in laboratory, the evolution of material subjected to salt spray can be effectively checked with the help of three different non destructive analysis : - the weight variation which corresponds mostly to a difference between solution accumulated and mass loss due to the surface disintegration - the visual scanning of the surface to quantify the sand disintegration degree - the measurement of permeability to air that gives an estimation, by means of an empirical relationship, of the solution quantity accumulated into the samples. Through the experimental results, three steps can be clearly distinguished in the evolution of a fine grain limestone such as Richemont during the exposure to salt spray: - an incubation time where the salt penetrates slowly without appreciably affecting the material structure, - an acceleration phase of the seawater absorption, mainly due to the increasing of salt content in the stone (hygroscopic effect, reduction of evaporation process),


321

- finely the stone reaches a more stable state, which corresponds to the saturation of the stone or the development of a progressive sand disintegration, according to the received salt spray intensity. The association of visual analysis and permeability measurements seems to constitute an objective non destructive experimental methodology to deduce the critical conditions that create an optimal disintegration. The salt content could then be deduced from the accumulated solution by implementing a mass transfer model between the stone and the environment. But, an experimental method, the electro-physic spectroscopy, represents also a promising non destructive technique to detect and quantify the salt content into the material. The relationship between the deterioration and the salt penetration kinetic in the depth of the stone has to be confirmed too by destructive chemical analysis. In order to assess the analogy between the real time ageing and the simulation, according to the climatic zone and the lithotype, an important campaign of tests conceming several stones selected with respect to their initial porosity and transfer properties, is planned now. These results may contribute to generate sufficiently information to forecast the deterioration effect by salt crystallisation on building stone.

Acknowledgements This work was founded by an EC Prject ENV-4 CT95-0100

References Auger F., 1991. Vieillissement par alt6ration atmosph6rique des mat6riaux de construction- Etude comparative in situ et en simulation. Proc. of Int. Symp. on the Deterioration of Building Materials, La Rochelle 12-14 june 1991, 115-128. Auger F., 1996. Durabilit6 des pierres calcaires utilis6es dans le patrimoine architectural, M6m. Soc. G6ol. France, 169, 415-420. Meng B., 1994. Calculation of moisture transport coefficients on the basis of relevant pore structure parameters, Materials and structures, 27, 125-134. Parrott L. J., 1994. Moisture conditioning and transport properties of concrete test specimens. Materials and structure 27, 460-468. Auger F., 1988. Simulation acc616r6e de la d6gradation des mat6riaux de construction en ambiance a6rienne saline, Proceeding of IAEG Conference on The engineering Geology of Ancient Works, Monument and Historical sites, Athens, 19-23 september, 797-804. B irginie J.-M., Auger F., 1999. Caract6risation visuelle de l'alt6ration d'6chantillons carrot6s de pierre, vieillis en simulation dynamique aux brouillards salins, Mat6riaux et Constructions, 32, 584-592. Kollek, J.J., 1989. The determination of the permeability of concrete to oxygen by the Cembureau method - a recommendation. Materials and Structures, 22, 225-230. B irginie J.-M., Rivas T., Auger F., 1999. Comparaison de l'alt6rabilit6 au brouillard salin de deux pierres calcaires de construction au moyen de mesures pond6rales, acoustiques et par traitement d'images. Submitted to Materiales de Construccion in september 1999.


323

C O L O U R CHANGES AND REACTIVITY TO SO2 OF SOME CLADDING STONES AT THE "GRAN TEATRE DEL LICEU" (BARCELONA; SPAIN) Carlota M Grossi; Rosa M Esbert*; Francisco J Alonso; Luis Valde6n; Jorge Ordaz; Francisco Diaz-Pache Area de Petrologia y Geoquimica. Dpto. de Geologia. Universidad de Oviedo. Jesfis Arias de Velasco S/N, 33005 Oviedo, Spain.

Abstract Some ornamental building stones (sedimentary and metamorphic) have been pre-selected to be used as cladding material for the restoration of the facades of the "Gran Teatre del Liceu" of Barcelona (Spain). In order to assist the final selection these were subjected to a climatic chamber SO2 exposure. A way to evaluate the effect of SO: was to measure colour changes after exposure. The methodology and the results are presented in this paper. Although no visual colour changes were observed, significant colour changes were detected by colorimetry measurements. These changes were related to stone reactivity to SO2.

Keywords: Ornamental stones, SO2 reactivity, Colour measurements, Stone conservation.

1. Introduction The main scope of the study was to select the most suitable cladding stones for the Great Theatre "Liceu" of Barcelona (Spain) from a series of pre-selected stones. In this sense, some durability tests were carded out to reproduce some of the main decay agents that can affect those stones once have been used in the building. Among the trials carried out, this paper shows the results of a SO2 polluted atmosphere exposure test. This test was selected because even very low atmospheric concentrations of SO2 can lead to decay of building stones, mainly those of carbonate nature. The test was evaluated by means of different techniques, such as binocular (low-medium power) microscopy, scanning electron microscopy (SEM) and microanalysis (EDX). Moreover, qualitative and quantitative colour measurements were undertaken both previously and after testing using a colorimeter. Colour determination is now a common practice in the field of stone conservation, mainly in studies of stone decay and treatment (Garcia Pascua et al, 1996, Ginell and Coifnan, 1998). Colour can be described by three attributes that correspond to the perceptions of hue, chroma and value (McAdam, 1985). Hue is defined by the terms blue, green, red and yellow. It creates the colour wheel. That corresponds to the normal term of colour. Chroma (saturation) refers to the degree of coloration, that is the purity of colour. It is related to the amount of white radiation that exists in the radiation of a particular colour. Colour is more or less saturated depending on the degree of dilution with white/grey. It refers to the terms "vivid" or "dull".

*Author to whom correspondence should be addressed.

324


Value (lightness or luminosity) indicates if the colour is "light" or "dark" and varies from white (maximum reflectance) to black (absence of light). Colour measurements were given in the CIE L*a*b* and CIE L*C*h systems. The system CIE L*a*b* is frequently used to measure colour differences because it represents better than others human sensibility to colour. L* is the variable lightness, which varies from 0 black to 100 white; a* and b* are the chromatic coordinates: +a* is red, -a* is green, +b* is yellow and -b* is blue. In the system CIE L*C*h, L* is again ligthness, C* is the chroma or saturation and h is the hue angle. They can be calculated by the equations (1) and (2). C* = (a*2+b*2) 1/2

(1)

h = tan-~(b*/a *)

(2)

Colour differences can be determined as follows: A L * = L * l - L * o ; A a * = a * l - a * o ; A b * = b * l - b * o ; A C * = C*I-C*0

(3)

Were: L* 1, a* l, b* 1 C* 1are the final values L* 0, a* 0, b* 0 C* 0 are the original values Total colour difference is determined as follows: AE* = (AL*2+Aa*2+Ab*2)1/2

(4)

2. Materials Six different stones were pre-selected: three light coloured sedimentary and three dark coloured metamorphic stones. All stones were honed finished. Additionally flamed specimens of two sedimentary stones were available for testing. Geological classification, mineral composition determined by X-Ray diffraction, visual colour and surface finishes are summarised in table 1.

3. Experimental methodology 3.1 Exposure to SO2 atmospheres Several nominal 50mm x 50mm x 10mm stone slabs were subjected during eight days to simulated polluted atmospheres in a climatic chamber Heraeus, which allows control of type and flow of gas, temperature and relative humidity. The gas was SO2 in a concentration of 2ppm. Temperature and relative humidity were 25~ and 90%, respectively. Samples were placed horizontally in the chamber, so that the honed or flamed surfaces were permanently exposed to SO2.

3.2 Test Evaluation The evaluation was carried out as follows : 9 Visual and low-medium power microscopy examinations. 9 Surface analysis by scanning electron microscopy and microanalysis SEM-EDX.


325

9 Quantitative measurements of colour. Colour measurements were carried out with a MINOLTA CR-200 colorimeter which possesses a xenon lamp that produces a beam of diffuse light illuminating a target area of 8 mm diameter. Colour was measured before and after testing in four specimens (50ram x 50mm surface) of each of the selected stones. The systems used were CIE L*a*b* and CIE L*C*h. The number of shots necessary to obtain a representative colour was selected taking into account the differences between two successive cumulative averages of each of the parameters L*, a* and b* (LNEC, 1994). 16 shots per specimens were selected for all the stones and 25 in the case of the gneiss, which was more heterogeneous in colour. Therefore, a total of 64 and 100 colour measurements were carried out before and after testing. The significance of colour differences was evaluated by the Mann-Whitney non parametric test. Table 1: Stone characteristics

4. Results and Discussion Table 2 summarises reaction products and colour changes during testing (Esbert et al, 1997). Values of the parameters L*, a*, b*, C* and eolour differences AL*, Aa*, Ab* and AE* are shown in this table. Differences in chroma (AC*) are graphically exhibited in fig. 1, which also indicates the significance of the differences (critical level V= 0.05). After testing, a small deposition of sulphur was detected on all samples under SEMEDX. This is almost imperceptible with the unaided eye and low-medium power microscopy (magnifications x120). No visual colour changes were observed. Sulphur distribution on the surface of the specimens was related to surface topography. Honed specimens showed a uniform distribution. Sulphur concentration was higher around

326


surface irregularities such as cavities or fissures. F l a m e d surfaces s h o w e d differential distribution o f sulphur favoured by the surface r o u g h n e s s o f this type o f finish. SO2 chemical attack w a s only detectable o n the gneiss surface. After testing, whitish salt efflorescence w a s o b s e r v e d on one o f the maffic minerals (scapolite). SO2 reacted with the calcium o f that mineral to f o r m gypsum. Table 2: Results o f S O Stone Test SEM EDX analysis Lim 1 pre . . . Honed post S deposit Lim 2 pre . . . Honed post S deposit Lim 2 pre . . . Flamed post S deposit Lim 3 pre . . . Honed post S deposit Lim 3 pre . . . Flamed post S deposit Serp 1 pre . . . Honed post S deposit Serp 2 pre . . . Honed post S deposit Gneiss pre . . . Honed post s deposit.

:st Visual Colour changes . NO . NO . NO . NO . NO . NO . NO . NO

L*

a*

74.30 2.16 73.60 L 2.37 60.86 6.53 60.46 6.71 6 7 . 0 5 5.19 65.94 5.64 65.34 3.68 65.63'3.42 70.76 4.99 70.09!| 5.34 32.74 -4.42 33.13 | -4.48 38.63 -5.63 39.74 | -5.55 47.48 -3.16 49.26 -3.27

b*

C*

AL*

15.88 16.99 17.29 18.13 13.46 14.66 21.19 21.88 15.76 16.80 0.48 0.47 0.53 0.30 2.90 2.88

16.03 17.16 18.53 19.39 14.47 15.75 21.51 22.14 16.55 17.65 4.45 4.52 5.65 5.56 4.43 4.51

.... -0.70 . . -0.40 .... -1.11 . . 0.29 . . -0.67 . . 0.39 .. 1.11 . . 1.78

Aa*

Ab*

i -0.21 l 1.11 . . . . 0.18 0.84 i "" 0.45 | 1.20 . . . . -0.26 0.69 . . . . 0 . 3 5 1| . 0 4 . . . . -0.06 1 -0.01 . . . . .1i 0.08 | -0.23 . . . . -0.11 -0.02

AE*

1.33

r

__

0.95 -.

1.70 ..

0.79 __

1.29 ..

0.39 __

1.14 -.

1.78

Gypsum from

reaction

with

scapolite

Chroma difference 1,5

Yes

I-, Ye~

1

~9

Yes

Yes

0,5 0 L1-H

L2-H

Yes

I

L2-F

L3-H

L3-F

~

$t -H

I

i

i

S2-H

I

i

'

i

G-H

-0,5

Specimen Figure 1: Chroma variations (AC*) after SO2 testing. Positive AC* means more saturate or pure colours whereas negative AC* indicates less saturate colours. The word "yes" means that the difference in values prior and after testing is significant (level cz=0.05). Chroma changes are significant in all the light limestones varying to more vivid or pure colours.


327

In the light limestones, appreciable colour changes were detected by colour measurements although no visible colour changes were observed (table 2). These stones had a substantial yellow component (positive and high b* values) and the change was mainly due to and increase in b* values (yellow direction), which led to a chroma increase (C*). Colour became more vivid or pure. That variation was similar in all the limestones. Lightness variations (AL*) were, in general, less significant. Total colour difference AE* was thought to be related to SO2 deposition and was higher in limestone 1 (which had more surface cavities) than in limestones 2 or 3 (with smoother surfaces). AE* was also higher in flamed finished (rougher) than in honed finished (smoother) surfaces. Overall, values of AE* corresponded to grey scale values from 4 to 5 (BS EN ISO 105A05, 1997), 4.5 being approximately the limit of visual appreciation. (Grey scale values vary from 5: no difference to 1 maximum difference). Neither visual colour changes nor significant chroma differences were detected in the dark metamorphic stones. The only significant change was found in lightness values (AL*) in the serpentinite 2. This difference was positive indicating, then, a change to a lighter colour. The gneiss showed higher absolute differences but these were not significant because of the poly-chromatic appearance of the stone.

5. Conclusions Colour measurements were useful to indicate degree of reaction to or deposition of SO2 in the case of mainly monochromatic building stones. In this way : No visual changes were appreciated in some light coloured limestones subjected to artificial SO2 polluted atmospheres, however significant colour differences were measured with a colorimeter. Colour differences also indicated that these limestones would with time turn to a more saturate yellowish colour due to SO2 action. Colour differences seemed to be higher in rougher surfaces, which could be related to a higher deposition of sulphur. However a quantification of sulphur is necessary to assess this supposition. From these results limestones 2 and 3 seemed to be more resistant to decay by SO2 than limestone 1. Serpentinite 1 appeared to be the most resistant among the dark metamorphic stones.

6. Acknowledgements The authors wish to acknowledge the financial support of this research to DragadosOSHSA U.T.E., CICYT (projects CC95-SEC05-01 and 1FD97-0331-C03-01) and FICYT (project PB-REC96-98). To Joan Ard~vol and Oriol Escolfi, from the Executive Direction of Restoration Works ("Gran Teatre del Liceu"). Also to Mr Richard Tews for revising the English and STATS Consultancy (St. Albans, UK) for the time and facilities allowed to complete this paper.

328


7. References

BS EN ISO, 105-A05, 1997. Test for Colour Fastness. Part A05. Instrumental assessment of change of colour for determination of grey scale rating. Esbert RM., Valde6n L., Alonso FJ., Ordaz J., Grossi CM., Diaz-Pache F., 1997. Estudio de la Durabilidad de las Piedras Preseleccionadas para el Revestimiento de las Fachadas del "Gran Teatre del Liceu". Unpublished Report, Universidad de Oviedo. Garcia Pascua N., S~.nchez de Rojas MI., Frias M., 1996. The important role of the colour measurement in restoration works. Use of consolidants and water-repellents in sandstone. 8th International Congress on Deterioration and Conservation of Stone, Berlin, 1351-1362. Ginell WS., Coffman R., 1998. Epoxy resin-consolidated stone : appearance change on aging. Studies in Conservation, 43,242-248. LNEC (Laboratorio Nacional de Engenharia Civil), 1994. Medi~oes colorim6tricas em rochas heterocronfiticas. Relatorio 89/94-GERO. I&D Geotecnia, 171. McAdam DL., 1985. Colour Measurements. Theme and Variations. Springer-Verlang.

329

EARLY MECHANISMS OF DEVELOPMENT OF SULPHATED BLACK CRUSTS ON CARBONATE STONE

Patrick Ausset*, Roger A. Lef'evre, Laboratoire Intertmiversitaire des Syst~mes Atmosph6riques, Universit6 Paris XII, 94010 Cr6teil, France. Marco Del Monte, Dipartimento di Scienze della Terra e Geologico-Ambientali, Universit~ di Bologna, 40127 Bologna, Italy.

Abstract

Experimental conditions characteristic of the urban pollution in many European cities over the last decades were reproduced in a simulation chamber in which samples of limestone were exposed for a period of 12 months, both naked or sprinkled with carbonaceous fly-ash. The development of gypsum crystals was observed overall in close proximity of fly-ash anchoring them to the limestone surface. Samples of the same limestone exposed in the field in a polluted environment for the same period of time led to similar results. The preliminary mechanisms leading to the genesis of sulphated black crusts in polluted environments were thus highlighted. Because of their roughness the embryonic black crusts increase the development of the crust by trapping new particles. This trapping is also facilitated by the wetness of the stone surface leading to the development of hydrated mineral (gypsum) in the water meniscus between fly-ash and stone surface. .(eywords: black crusts, limestone, sulphur dioxide, carbonaceous fly-ash, gypsum, simulation chamber, field exposure.

Introduction

Sulphation with the development of authigenic gypsum crystallisation below the surface of materials affects carbonate stones exposed to atmospheric pollution (Girardet and Furlan, 1983). This mechanism has been experimentally reproduced, revealing the important role played by SO2 (Spedding, 1969), enhanced by the presence of NO2 (Johansson et al., 1988), 03 (Haneef et al., 1992) and high relative humidity (Johansson et al., 1988; Spiker et al., 1992). The sulphation progresses with decreasing intensity from the surface towards the interior of the material (Ausset et al., 1996). On buildings surfaces, generally on those sheltered from direct rainfall or washout, along with sulphation products formed below the surface, one also observes the presence of black gypsum crusts above this surface containing numerous particles, particularly fly-ash, emitted into the atmosphere due to the combustion of coal and mineral oils (Fassina et al., 1979; Camuffo et al., 1982, 1983; Del Monte et al., 1981; Del Monte and Sabbioni, 1984; Ausset et al., 1994). Carbonaceous fly-ash, particulary those emitted by heavy fuel oil combustion, have a very strong physico-chemical reactivity. They contain many sulphured chemical species, in particular sulphates along with metals such as vanadium, iron and nickel. Indeed, it has been demonstrated that:

330

9th International Congress on Deterioration and Conservation of Stone, Venice 19-24June 2000 9 carbonaceous fly-ash adsorb atmospheric SO2, triggering its oxidation into sulphate (Novakov et al., 1974) 9 this reaction is intensified by the presence of NO2 and 03 (Cofer et al., 1980; 1981) 9 and by a high relative humidity (Tartarelli et al., 1978).

The reactivity of carbonaceous fly-ash in the presence of H20 or the reactivity of carbonate stones in the presence of SO2 and H20 are well known. By contrast, experiments that bring the four (fly-ash, stone, SO2 and H20) together within a single system are rare and have yielded contradictory results: some suggest that fly-ash play a slightly intensifying role in the sulphation process (Cheng et al., 1987; Ausset et al., 1996; Sabbioni et al., 1996), while others point to a completely negligible effect (Hutchinson et al., 1992). In fact the really or directly effect of carbonaceous fly-ash or other aerosol particulate matter had never been place in a prominant position. This paper tends to explain the role and the contribution of carbonaceous fly-ash in the nucleation of gypsum crystals and in the development of incipient sulphated black crust on carbonate stone. This goal is achieved by comparing simulation chamber data with those obtained in the field. Embryonic black crusts forming over 12 months through interaction between carbonaceous fly-ash and samples of Jaumont limestone in the controlled atmosphere of a simulation chamber (Lausanne Atmospheric Simulation Chamber - LASC, Ausset et al., 1996) are compared with those forming in the field on the same carbonatic stone, over the same time period in a real polluted atmosphere.

1. S i m u l a t i o n chamber observations 1.1. Materials

9 Fly-ash from heavy fuel oil combustion were collected from the electrostatic filters of an electrical power plant. The granulometric fraction lower than 100 ~tm, obtained by dry sieving, was used for the experiment. These particles are in the same size range as those found in samples collected from the atmosphere, far from their sources near the material surfaces (Ausset et al., 1992; Derbez and Lef'evre, 1996) or from within sulphated black crusts (Del Monte et al., 1981). Carbonaceous fly-ash are black and more or less round in shape with spongy (fig. la)or porous morphology (fig. lb).

Figure 1: a) spongy and b) porous carbonaceous fly-ash particles both emitted by heavy fuel oil combustion collected fromthe electrostaticfilter of a power station (Porchevillepower plant, France).


331

*Bulk chemical analysis, Raman Spectrometry and X Ray adsorption spectrum (XANES) of this fraction show that the main element is carbon (graphite), accompanied by small concentrations of sulphur, sulphates, vanadium, silicon and iron (Ausset et al., 1999). 9 The Jaumont limestone had been chosen because of its widespread use as a building material. This limestone is made up of 94% calcite and 2.5% quartz and has a water porosity of 23%. It presents macropores of 200-500 ~tm in diameter, which are observed on the surface once the stone has been cut. The fly-ash spreaded on the Jaumont limestone revealed a tendency to collect within the superficial open macropores.

1.2. Experimental procedure The Jaumont limestone samples (0.1 x 0.1 x 0.02 m) were placed inside the LASC, either naked or sprinkled with fly-ash at the concentration of 7.5 g.m-2. The operating principle of the LASC (fig. 2) is described in Ausset et al. (1996). We simply mention here the following experimental conditions: temperature 13 ~ relative -3

-3

humidity 79 %; [SO2] = 340 ~tg.m (125 ppb); [NO2] = 98 ~g.m (50 ppb). These conditions reproduced those which have existed in Milan, the temperature and relative humidity being the annual average in this city from 1931 to 1960, and the 1972 annual average for SO2 and NO2 concentrations, years when the pollution peaks were very high. Such conditions are analogous to those observed in intensely industrialised regions in the seventies and eighties. In all cases, the values chosen for this study were far lower than those utilized during most of previous simulation experiments (Braun and Wilson, 1970; Judeikis and Stewart, 1976; Johansson et al., 1988; Haneef et al., 1992; Coboum et al., 1993).

Figure 2: Schemeofone ofthe 10 cells and its nmning: gas injection system (1 to 5), main part (6 to 11) and measurement system (12 and 13) (modifiedfromAusset et al., 1996 and Girardetet al., 1996).

332


The samples were removed from the simulation chamber and studied after 3, 6, 9 and 12 months of exposure. The reaction products forming on the limestone surfaces, on fly-ash and on the stone-particle interface were studied by Analytical Scanning Electron Microscopy (ASEM, Jeol 6301F fitted with an EDS elemental analysis system Oxford Link-Isis).

1.3. Results and discussion 1.3.1. Evolution of the Jaumont limestone and of the fly-ash deposited on its surface

In a previous work (Ausset et al., 1996), it was shown how Jaumont limestone, either naked or sprinkled with fly-ash, after 12 months presented sulphation reaching a depth of 600 to 800~tm, of decreasing intensity as the depth increased. Observations by SEM also revealed the presence of small acicular gypsum crystals on the surface of the naked samples. However, even after a year of experimentation, these gypsum crystals did not cover the whole surface. In the case of the limestone sprinkled withfly-ash, the two above-mentioned phenomena, sulphation above and below the surface, are found with increasing intensity. In fact, the following observations were made: 9 small lanceolate gypsum crystals (a few ~tm), isolated or occasionally in small clusters, formed on the surface, in areas not in direct contact with the carbonaceous fly-ash (fig.3).

Figure 3: SEM micrograph of the surface of the Jaumont limestone sprinkled with carbonaceous fly-ash exposed in LASC after 3 months of exposure. Lanceolate gypsum crystals are visible in areas not in direct contact with fly-ash.

9 acicular crystals (20 x 1 lam) or tabular crystals (10 x 3 ~tm) (fig. 4a) or ~ desert rosette~ crystals (fig. 4b) developed directly on the surface of the fly-ash. EDS analysis of all the crystals revealed the presence of calcium and sulphur, with a S(Ca ratio (0.7-0.8), close to that of anhydrite (CaSO4), bassanite (CaSO4, 0.5H20) or gypsum (CaSO4, 2H20). The presence of high relative humidity in the LASC and the characteristic morphology of the crystals indicate a highly probable presence of gypsum. Among the 165 crystals individually analysed, 65% consist of gypsum (acicular, tabular or occasionally "desert rosette" in shape), the other 35% were juxtaposed micronic fine lamellar crystals composed of sulphur and iron associated with low content of sodium, vanadium and nickel or lenticular crystals composed of sulphur and sodium, with a S/Na ratio (0.7-0.8) close to that calculated for thenardite (NaESO4) or mirabilite (Na2SO4, 10H20). All of the described


333

crystals invariably contain sulphur, and a single fly-ash particle can be the site o f different crystalline t y p e s .

Figure 4: SEM micrographs of different shapes of gypsum crystals developped on the carbonaceous fly-ash deposited on the Jaumont limestone after three months of exposure in LASC: a) acicular in shape; b) desert rosette in shape.

Figure 5: SEM micrographs of gypsum crystals (Gyp.) grown on the contact between a fly-ash (FA) particle and the Jaumont limestone substrate (Cal.) a/ier 12 months in the LASC: a) general view ; b and c) blow up ofacicular gypsum, attached to the fly-ash particle (FA), anchoring it to the surrounding limestone substrate 9 regardless o f the crystals developing on the surface o f the carbonaceous fly-ash and on that o f the limestone substrate between these particles, only g y p s u m crystals were observed at the fly-ash-limestone interface. These acicular and tabular g y p s u m crystals, which appear to anchor the fly-ash to the substrate, develop on the limestone substrate

334


encircling the fly-ash basis (fig. 5a, b, c). The gypsum is clearly observed at a distance of up to 50 ~tm. Alongside the parallel experiments in the simulation chamber, some of the same fly-ash particles were placed inside a hermetically sealed opaque phial away from air, light and relative humidity. During the 12 months experimental period, they were regularly studied under ASEM at three-month intervals. No morphological transformation or crystal growth appeared and the chemical composition remained stable. It would therefore seem that carbonaceous fly-ash only become reactive in specific environmental conditions, such as those imposed inside the LASC.

2. Field observations 2.1. Materials and methods Parallelepipedic samples of Jaumont limestone (20 x 10 x 2 cm) were exposed vertically, sheltered from rainwater and washout in Milan for one year. During this period (May 1986/ May 1987), the average content of the air in SO2 was 107 lag.m"3, approximately 3 times weaker than that having existed in 1972 and than that imposed in the LASC (Furlan and Girardet, 1988). 2.2. Results and discussion The observation by ASEM reveals the following points;

9 many microparticles of industrial origin are visible on the surface of the Jaumont limestone exposed in Milan at the average density of 90 particles per square centimeter. The most abundant are carbonaceous spongy and porous fly-ash (72% in number), similar in shape and granulometry and comparable in chemical composition to those utilized during the the experiment in the LASC. Alumino-silicated spherical (18%) and ferriferous dendritic (10%) fly-ash originating probably mainly from coal combustion, are also represented. Small gypseous crystallizations appear on the surface of some fly-ash, always less numerous and less well developed than in the case of the fly-ash deposited on stone exposed in the LASC.

Figure 6: SEM micrographs of gypsum crystals grown on a fly-ash particle and at the contact between the particle and the Jaumont limestone substrate exposed in Milan during one year : a) general view ; b) blow up of small acicularcrystals of gypsum growingon the fly-ashparticle (FA) and largertabular crystals of gypsum (Gyp.) anchoringit to the surroundinglimestome substrate (Cal.).


335

While in the LASC many species of crystals develop, only gypseous crystallizations are observed on the fly-ash surface in Milan. This result can be explained by the higher solubility of many salts (i.e. sulphate of sodium, of iron...) compared to those of calcium sulphate. In the LASC the relative humidity was stable at the average relative humidity of 79%. Moreover, in the field, the soluble salts are removed by condensation water (smog, fog) and only gypsum remains on material surfaces. 9 Gypseous crystallizations are also visible on the surface of the limestone, in regions not covered by fly-ash. 9Finally, small gypsum crystals are present at the interface between fly-ash and stone surface (Fig. 6) as in the experiment.

3 - Conclusions

The above results allow us to highlight the following main phenomena: 9 Jaumont limestone exposed naked in the presence of SO2 and humidity in the LASC, as well as undergoing sulphation below the surface, also reveals a moderate degree of authigenic gypsum formation above the stone surface. 9 On Jaumont limestone sprinkled with fly-ash, one observes a combination of the two above-mentioned phenomena: predominantly gypsum crystals appear on the fly-ash, while only gypsum crystals are encountered on the limestone surface between the fly-ash particles. However, the third most original phenomenon occurs at the limestone-fly-ash interface and consists of the fixing of these particles to the material surface through the intermediate growth of authigenic gypsum crystals. This phenomenon marks the beginning of the development of an embryonic black gypsum crust, the colour of which is given by the particle content. 9 The field samples also revealed fly-ash adhering to limestone by means of a crown of gypsum crystals. The gypsum crystals observed on the surface of limestone samples between carbonaceous particles are less numerous. 9 On the basis of simulation chamber and field data obtained over 12 months, the study of the morphology, mineralogy and chemical composition of the deposits and associated neocrystallisations indicates that they represent the first stage of black crust development. Although accounting for only a modest fraction, the fly-ash appear to play a crucial role leading to the formation of black crusts, since they facilitate the precipitation of gypsum which constitutes the predominant mineral in the black crusts. 9 The deposition of the fly-ash on the stone surface is easier when this surface is wet. It is highly probable that a >of water is forming at the interface of the wet surface stone and the wet fly-ash particle. Thus, the evaporation of the water leads to the crystallization of gypsum that anchors the particle to the substrate. 9 Finally the development and thickening of the black crust from its embryo proceeds through the continued deposition of new fly-ash, the fLxation of which is strongly enhanced by the local increase in surface roughness. This phenomenon, which is a continuous deposition in the field, was impossible to reproduced in the simulation chamber.

336


Aknowledgrnent : this work was supported by the European Commision, programme >. 4. References Ausset P., Bannery F. and Lef'evre R. (1992). Les microparticules dans l'air et dans une crofite noire de Saint-Trophime d'Arles. 7th International Congress.on Deterioration and Conservation of Stone, Lisbon, vol. 1,325-334. Ausset P., Lef'evre R., Philippon J. and Venet C. (1994). Pr6sence constante de cendres volantes industrielles dans les crofites noires d'alt&ation superficielle de monuments fran~ais en calcaire compact. Comptes Rendus de l'Acaddmie des Sciences., Paris, t. 318, s6rie II, 493-499. Ausset P., Crovisier J.L., Del Monte M., Furlan V., Girardet F., Hammecker C., Jeannette D. and Lef'evre R.A. (1996). Experimental study of limestone and sandstone sulphation in polluted realistic conditions: the Lausanne Atmospheric Simulation Chamber (LASC). Atmospheric Environment, 30, 18, 3197-3207. Ausset P., Del Monte M. and Lef'evre R.A. (1999). Embryonic sulphated black crust in Atmospheric Simulation Chamber and in the field 9role of the carbonaceous fly ash. Atmospheric Environment, 33, 10, 1525-1534. Braun R.C. and Wilson M.J.G. (1970). The removal of atmospheric sulphur by building stones. Atmospheric Environment, 4, 371-378. Camuffo D., Del Monte M. and Sabbioni C. (1983). Origin and growth mechanisms of the sulfated crusts on urban limestone. Water, Air and Soil Pollution, 19, 351-359. Camuffo D., Del Monte M., Sabbioni C. and Vittori O. (1982). Wetting, deterioration and visual features of stone surfaces in an urban area. Atmospheric Environment, 16, 22532259. Cheng R.J., Hwu J.R., Kim J.T. and Leu S.-H. (1987). Deterioration of marble structures. The role of acid rain. Analytical Chemistry., 59, 2, 104-106. Coboum W.G., Gauri K.L., Tambe S., Li S. and Saltik E. (1993). Laboratory measurements of sulfur dioxide deposition velocity on marble and dolomite stone surfaces. Atmospheric Environment, 2, 193-201. Cofer W.R. III, Schryer D.R. and Rogowski R.S. (1980). The enhanced oxidation of SO2

by NO2 on carbon particulates. Atmospheric Environment, 14, 571-575. Cofer W.R. III, Schryer D.R. and Rogowski R.S. (1981). The oxidation of SO2 on carbon particles in the presence of 03, NO2, and NO 3. Atmospheric Environment, 15, 1281-1286. Del M onte M. and Sabbioni C. (1984). Gypsum crust and fly-ash particles on carbonatic outcrops. Archives for Meteorology, Geophysics and Bioclimatology, Ser-B35, 105-111. Del Monte M., Sabbioni C. and Vittori O. (1981). Airborne carbon particles and marble deterioration. Atmospheric Environment, 15, 645-652. Del Monte M., Sabbioni C., Ventura A. and Zappia G. (1984). Crystal growth from carbonaceous particles. The Science of the Total Environment, 36, 247-254. De Santis F. and Allegrini I. (1992). Heterogeneous reactions of SO2 and NO 2 on carbonaceous surfaces. Atmospheric Environment,16, 3061-3064. Derbez, M. and Lef'evre, R.A (1996) - Le contenu microparticulaire des crofites gypseuses de la Cath6drale Saint-Gatien de Tours. Comparaison avec l'air et la pluie 8 th International Congress on Deterioration and Conservation of Stone, Berlin, 359-370.


337

Fassina V., Lazzarini L., Biscontin G. and Calogero S. (1979). Influenza del materiale particellare atmosferico sui processi di degradazione della pietra a Venezia. 3rd International. Congress on Deterioration and. Conservation of Stone, Venice, 43-53. Furlan V. and Girardet F. (1988)- Vitesse d'accumulation des compos6s atmosph6riques du soufre sur diverses natures de pierre. 6th International Congress on Deterioration and Conservation of Stone, Torun, 187-196. Girardet F. and Furlan V. (1983). Mesure de la vitesse d'accumulation des compos6s soufr6s sur des 6prouvettes de pierre expos6es en atmosphere rurale et urbaine. 4th International Congress on Deterioration and Conservation of Stone, Louisville, 159-168. Girardet F., Furlan V., Ausset P., Del Monte M., Jeannette D. and Lef'evre R.A. (1996). Etude exp6rimentale de la prise en soufre d'un calcaire et d'un gr~s caleareux dans la chambre de simulation atmosph6rique de Lausanne. 8th International. Congress on Deterioration and. Conservation of Stone, Berlin, 349-358. Haneef S.J., Johnson J.B., Dickinson C., Thompson G.E. and Wood G.C. (1992). Effect of dry deposition of NO x and SO 2 gaseous pollutants on the degradation of calcareous building stones. Atmospheric Environment, 16, 2963-2974. Hutchinson A.J., Johnson J.B., Thompson G.E., Wood G.C., Sage P.W. and Cooke M.J. (1992). The role of fly-ash particulate material and oxide catalysts in stone degradation. Atmospheric Environment, 15, 2795-2803. Johansson I.J., Lindqvist O. and M angio R.E. (1988). Corrosion of calcareous stones in humid air containing SO 2 and NO r Durability of Building Material, 5,439-449. Judeikis H.S. and Stewart T.B. (1976). Laboratory measurement of SO2 deposition velocities on selected building materials and soils. Atmospheric Environment, 10, 769-776. Novakov T., Chang S.G. and Harker A.B. (1974). Sulfates as pollution particulates : catalytic formation on carbon (soot) particles. Science, 186, 259-261. Sabbioni C., Zappia G. and Gobbi G (1996). Carbonaceous particles and stone damage in a laboratory exposure system. Journal of Geophysical Research, 101,621,627. Spedding D.J (1969). The rate of sulphur-35/sulphur dioxide released in a laboratory. Atmospheric Environment, 3, 341-346. Spiker E.C., Hosker R.P, Comer V.J., White J.R., Were Jr R.W, Harmon F.L., Gandy G.D and Sherwood S.I. (1992). Environmental chamber for study of the deposition flux of gaseous pollutants to material surface. Atmospheric Environment, 16, 2885-2892. Tartarelli R., Davini P., Morelli F. and Corsi P. (1978). Interactions between SO 2 and carbonaceous particulates. Atmospheric Environment, 12, 289-293.


339

PAST AIR POLLUTION RECORDINGS ON STONE MONUMENTS: THE HEADS OF THE KINGS OF JUDA STATUES FROM NOTRE-DAME CATHEDRAL (PARIS)

Patrick Ausset*, Roger A. Lef'evre, Laboratoire Interuniversitaire des Syst6mes Atmosph6riques, Universit6 Paris XII, 94010 Cr6teil, France. Marco Del Monte, Dipartimento di Scienze della Terra e Geologico-Ambientali, Universit~ di Bologna, 40127 Bologna, Italy. St6phanie Thi6bault, Arch6ologies et Sciences de l'Antiquit6, M.A.E., CNRS, 92023 Nanterre, France.

Abstract A pollution linked to the combustion of wood is a characteristic of the atmosphere of the cities in the past. It led to the development of thin grey crusts on the surface the stone of monuments. The grey crusts discovered on the Heads of Kings of Juda Statues, present on the western facade of Notre-Dame in Paris from the 13th Century to 1792, are the material witnesses of the effects of this ancient air pollution. The particles of unburnt wood included inside these grey crusts confirm that the burning of wood was the main cause of air pollution. The high position where the statues were sited suggests that this development was not the result of a local or punctual phenomenon but the result of a general pollution of the air of Paris at that time. Keywords: Heads of the Kings of Juda, Notre-Dame Cathedral in Paris, past air pollution, grey crusts, charred wood residues, polychromy, calcite, gypsum, ASEM, DRX, anthracology.

1. Historical introduction The remains of the Heads of the Kings of Juda, exhibited as a group at the Mus6e National du Moyen Age of the H6tel de Cluny in Paris since 1980, are those of the Heads of statues that originally adorned the western fafade of Notre-Dame Cathedral in Paris (Fig.l). The statues stood there from around 1240 until 1792, when they were thrown down along with all other statuary adorning the fafade. Their remains were heaped together in the parvis of Notre-Dame until 1796, when they were cleared away by a building contractor (Fleury in Giscard d'Estaing et al., 1977). It can be supposed that the statues were not 'decapitated' until their removal in order to facilitate transport. However, it cannot be entirely excluded that the heads had broken off when the statues were knocked down. Whatever the case, the group of sculptures was dispersed among different sites in Paris and the surrounding areas. It was not until 1977, that twenty-one of the twenty-eight Heads of the Kings were accidentally brought to light during earthworks on the foundations of the French Bank of Foreign Trade in rue de la Chauss6e d'Antin (Giscard d'Estaing et al., 1977).

340


Figure 1: Western Facade of Notre-Dame Cathedral in Paris in 1699 by Antier (French National Library, Stamps department). The Gallery of Kings of Juda Statues is well visible above the three portals.

Figure 3: Headnumber 16. Note the presence of the grey crust forming an almost vertical strip from the crown down to the spacebetweenthe eyes.

2. Description of the Heads of the Kings of Juda These statues were sculpted in a free Miliolidae limestone (Middle Eocene) frequently employed in the statuary and buildings in Paris (Blanc and Lorenz, 1992). Traces of polychromy dating back to the early 13th Century are visible; indeed, the height at which the statues were erected (16 m) and the consequent lack of access to them is sufficient to discard the possibility that any maintenance, cleaning or restoration had subsequently been performed (Erlande-Brandenburg, 1982). Apart from the polychromy, the surface of the Heads locally shows a grey crust, generally away from the damaged areas. In particular, they are absent from the surfaces of major breaks.The literature makes no mention of the presence of these grey crusts. On comparison with other similar cases studied in Aries (Ausset and Lef'evre, 1994) and Bologna (Ausset et al., 1998), it is reasonable to think that the grey crusts on the Heads bear witness to the air pollution in Paris, at least at the time of the French Revolution and during the period immediately before it. The hypothesis favouring the atmospheric origin of the grey crusts is substantiated by their absence from the surfaces of major fracture sections, as mentioned above. That means that they formed while the Heads were exposed to the Paris atmosphere and not during their long period of burial from 1796 to 1977.


341

This study should be seen within an archaeological-environmental perspective and contributes to research on the character and evolution of urban air pollution during the recent past. These results will also be compared with those deriving from studies on the modem industrial black crusts.

3. Sample collection and preparation The samples were obtained by scraping the surfaces with a scalpel and took the form both of tiny scales of millimeter size and powder. Firstly, all samples were studied raw, without undergoing any particular preparation procedure. Then, the scales were set within a polyester resin to study their stratigraphic section. The powders underwent hydrolysis by HCI (6M), in order to dissolve the sulphates and carbonates and isolate the solid particles insoluble in the acid. Solid particles have been then washed with distilled water. Among the twenty-one Heads, samples of grey crusts were taken from four (fig. 2): Head 4 : centre forehead; Head 8: around the lips and back of the scalp on the leit; Head 12 : centre forehead and nose break; Head 16: from a grey crust forming an almost vertical strip from the crown down to the space between the eyes (fig. 3).

4. Analytical methods Several analytical techniques were used to highlight the main physico-chemical and mineralogical characteristics of the grey crusts: differences in colour between ancient and modem crusts were measured by colorimetric technique (Munsell Color Chart); the determination of the mineral phases present in the stone, the cement matrix and particles in the crusts and pigment layers was performed by X-Ray Diffraction (XRD, Philips PW1710). The morphological, granulometric and elemental characterisation of the isolated grains and stratigraphic sections were performed by Photon Microscopy (Leica-Wild M10 Stereomicroscope and Zeiss 2 Axiophot) and Analytical Scanning Electron Microscopy (ASEM, Jeol JSM 6301F fitted with a Link-Isis device for semi-quantitative analysis by XRay Energy Dispersive Spectrometry, EDS). 5. Results 5.1. Raw sample analysis The colorimetric analyses performed on the powders revealed a grey colour with shades running from yellow to light green. The grey crusts surface is transparent and microcrystalline, like that of alabaster. Large, solid gypsum crystals are easily identified and are highly characteristic. The crust is composed of a mixture of calcite (CaCO3) as the main constituent (65 %)and a smaller quantity of gypsum (CaSO4, 2H20) (35 %) and, less frequently, of a mixture of larger quantities of gypsum and smaller quantities of calcite, associated with traces of quartz and, in only one case, traces of weddeUite (CAC204, 2H20). Residues of the underlying stone scraped off during sampling are locally attached to the crust.

342


5.2. Analysis of the stratigraphic sections The stratigraphic sections show the following succession (fig. 4): 1. Miliolidae limestone; 2. a discontinuous white layer of a few tens of ~tm thickness, composed of white lead (cerusite : PbCO3 or hydrocerusite: 2PbCO3, Pb(OH)2), constituting a preparation layer; 3. the upper grey layer of a few hundred micrometres thickness (400 to 500 ~tm), composed of a calcite-gypsum matrix, containing numerous black particles and grains of quartz.

Figure 4: Scanning Electron Micrograph ofa stratigraphic cross section of the grey crust surrounding the Head number 12 : a) Secondary electron image : this sample is composed of three layers, from bottom to top, (1) Milliolidae limestone (2) - discontinuous intermediate layer; (3) - homogeneous upper layer, b) Distribution image of silicium (Si) : SiO2 grains (quartz) are distribuated in the limestone and in the upper layer, c) Distribution image ofsulfur (S) and lead (Pb) : the intermediate layer (2) is constituted ofPb (W.L : White Lead) and the upper layer (3) contains S in the gypsum form. d) distribution image of calcium (Ca) : calcite in the limestone (layer 1), and a mixture of gypsum and calcite in the grey crust (layer 3). The above sequence shows that the grey crust can only have formed after the decoration of the sculptures (1240). This chronology supports the hypothesis of the grey crust formation may be the result of subsequent atmospheric deposition.

5.3. Analysis of particles isolated by acidic attack Different particle types were identified after dissolving the powders in hydrochloric acid: black particles and pigment grains. They were studied and analysed by ASEM.

5.3.1. Black particles The grey colour of the crusts is due to the presence of numerous black particles, which are elongated, prismatic, longitudinally striated and, in some cases, presenting small pits


343

regularly distributed over the surface. The black particles are essentially composed of carbon associated with traces of Si, Ca, K, A1 and, occasionally, Na and/or Mg. These particles are encountered in the grey crusts removed from the unbroken surfaces of Heads 4, 8, 12 and 16 and from the breakage of the nose on Head 12. In addition, they were observed within the black patch on the lower fracture of Head 17, but not in the black patch of the vertical fracture of the same Head. The size of the particles (from a few tens to several hundreds of microns) is compatible with their transport within atmospheric aerosols. Identical particles of comparable size have been observed and measured in the grey crusts sampled on other monuments in Aries (Ausset and Lef'evre, 1994) and Bologna (Ausset et al., 1998) and have been identified as the debris of wood incomplete combustion. The identification of these wood fragments was attempted by comparing them with present-day collections of burned wood and with atlas of wood anatomy (Boureau 1956; Metcalfe and Chalk 1950; Greguss 1955, 1959; Jacquiot 1955; Jacquiot et al., 1973; Schweingruber 1978, 1990). In spite of their very small size, some characteristic anatomical structures have been recognised. It would appear that the wood debris originate from the combustion of angiosperms and gymnosperms (conifers). The three anatomic planes of wood (transversal, longitudinaltangential and longitudinal-radial) are visible on some to the samples observed using ASEM. In fact, the micrograph in figure 5 shows, in the longitudinal-tangential plane, the section of a vessel and the presence of a uniseriate ray, probably attributable to an angiosperm, as indicated also by the details of the pits on the vessel visible in figure 6. Figure 7 shows a detail of a uniseriate ray cell. A further example, presented in figure 8, shows on one side the presence of a tracheid-ray cross field in the longitudinal-radial plane and on the other a large fenestriform pit in cross-field, which cannot fail to suggest the Scotch pine (Pinus sylvestris). All of the observed samples present uniseriate rays. According to the anatomic charts, only some tree species possess such uniseriate rays" gymnosperms in general and a few angiosperms, for example, the chestnut (Castanea sativa), willow (Salix sp.), poplar (Populus sp.). Others present a combination of uni- and pluriseriate rays, like the oak (Quercus sp.) and certain Betulaceae (birche family). However, the presence of homogeneous ray cells, coupled with the absence of spiral thickenings and micro-pits in the pores tend to exclude the Betulaceae, willow and poplar. The identification of wood fuels used in Paris has been achieved by charcoal analysis from the archaeological excavations of Napol6on (Thi6bault, 1986) and Carrousel Squares in the Louvre (Solari and Thi6bault, 1991). It indicated that, among the numerous species used as domestic and craitmen fuels, the deciduous oak is the most frequently encountered. The other most commonly employed tree species were the poplar, willow, chestnut, birch (Betula sp.), hornbeam (Carpinus betulus) and hazel tree (Corylus avellana). Gymnosperms have been very rarely identified, although they are present in filigree in anthracological analysis starting from the 16th Century, the fir (Abies alba) and Scotch pine being those most frequently observed. The attribution of a definite origin to the plant residues contained in the crusts on the Heads is made more difficult by the very small size of the samples, compared to those collected from archaeological excavations. However, it is possible to attribute certain fragments to angiosperms which could belong to the oak or chestnut and others to a gymnosperm species such as the Scotch pine.

344


Finally, spherical particles are sometimes found stuck to the wood fragments, which are colourless or with slight hints of light green or brown; ranging from 1 to 5 ~tm diameter, they are composed of Si, and less often, of Si and AI. Although more rarely, these spherical particles can also be found isolated within the grey crusts. Both these features have already been observed within ancient grey crusts (Ausset and al., 1998) as well as in the microparticulate of smoke from experimental wood combustion.

Figure 5: Scanning Electron Micrograph of a vessels' section and presence of a uniseriate ray in the longitudinal-tangential plane of a fragment attributable an angiosperm.

Figure 6: Scanning Electron Micrograph of a pit of a vessel (detail of figure 5).

Figure 7: Scanning Electron Micrograph of a detail ofuniseriate ray cell.

Figure 8: Scanning Electron Micrograph of a large fenestriform pits in cross-field in the longitudinalradial section attributable to the Scotch pine (Pinus

Sylvestris).

5.3.2. Pigments As mentioned above, the Heads were painted. Indeed, the evidence suggests that the grey crust formed above of the painted surface. Pigments of different colours are observed: red pigments of cinnabar (HgS); yellow to brown pigments of oxides (hematite, Fe203) or ferrous hydroxides (goethite, FeO-OH, or limonite, FeO-OH, nH20), the constituent minerals of ochres; green pigments originate from green clays, whose chemical composition varies according to the presence and amount of minerals such as glauconites ((K,Ca,Na)_l.6(Fe 3+, Al, Mg, FeE+)4Si7.3Alo.7OEo(OH)4), celadonites (K(Mg, Fe, A1)2(Si, A1)4Olo(OH)2) or serpentines


345

(Mg3[Si2Os](OH)4), along with metallic grains of gold associated with impurities of Ag, Cu and Zn.

6. Discussion and conclusion

Many of the characteristics of the ancient grey crusts of the Heads of Kings of Juda directly contrast with those of the present-day black crusts (Del Monte et al., 1981) observed on the buildings of Paris, while they suggest comparison with other ancient grey crusts observed in different cities (Aries and Bologna). These characteristics are colour, texture, thickness, chemico-mineralogical nature of the crusts and of the prismatic or spherical particles embedded in the matrix. The grey crusts on the Heads are far lighter in colour than that of the present-day black crusts observed (very dark grey to black). Finally, the ancient crusts have a fraction of a millimetre thickness, whereas the modem crusts can reach to about one centimeter. Diffractometric analysis of the modem black crusts highlights the presence of the same minerals as those found in the ancient crusts of the Heads, but in different proportions 9 greater amounts of gypsum than calcite, followed by quartz and weddellite (oxalic acid secreted by lichens). But the main difference in the mineralogical composition of the two crusts lies in the lower gypsum/calcite ratio of the ancient crusts. In addition, the ancient grey crust is composed of gypsum crystals of three-dimensional growth, while gypsum of the modem black crusts presents lamellar crystals of two-dimensional growth. This means that the conditions of their crystallisation were not the same, particularly the humidity of the ambience and the SO2 concentrations. The studies of the genesis of modem sulphated crusts (Braun and Wilson, 1970 ; Camuffo et al., 1983) suggest that the origin of the gypsum is to be found in the atmospheric contribution of sulphur and calcium in various forms, associated with urban air pollution (Del Monte and Rossi, 1997). These results coincide with those obtained on ancient grey crusts from Bologna (calcite : 70% ; gypsum : 30%), although they differ slightly from those obtained on another grey crust, this time more ancient, observed on the Roman portal of Saint-Trophime in Arles. Here, the grey crust is mainly composed of calcite associated locally with acicular gypsum crystals of micron size (Ausset and Lef'evre, 1994 ; Ausset et al., 1998). A further major difference between the ancient grey crusts and the modem black crusts lies in the kind of particles contained in the gypsum-calcite cement. While the grey crusts only contain the residues of incomplete wood combustion, the modem crusts embed microspheres which turn out to be fly-ash released into the atmosphere by the combustion of coal or heavy fuel (Del Monte M. and Sabbioni, 1987; Ausset et aL, 1994). These microparticles, of highly characteristic granulometry, morphology and chemical composition, are excellent tracers of modem industrial pollution. Large quantities of fly-ash are always found within modem black crusts, unlike the residues of wood burning, which are very rarely observed and, when so, always only at their base. Indeed, the absence of industrial flyash confirms that the grey crust is ancient. Moreover, the presence of particles from wood combustion is proof that the Paris air, during the Revolution and in the years or centuries prior to it, was polluted by smoke originating from wood combustion. The body of observation and analyses are proof that the grey crusts locally encountered on the Heads originate from a phenomenon of atmospheric deposition on the surface of stone. It also

346


shows that the atmospheric alteration of monumental stones is not an exclusively modem phenomenon, appearing with the on-set of air pollution of industrial origin. Medieval and Renaissance cities also experienced air pollution, due, not to coal or oil combustion, but to the burning of wood and vegetal matter in general. Acknowledgements. This work received the financial support of the European Commission, contract ENV4CT 95-0092: "Archaeometric study to reconstruct the pollution and climate of the past and their effects on Cultural Heritage" and of the Ue-de-FranceRegional Council, Programme "Sesame 1995". We acknowledge particularly Mrs Huchard, Director of the National Museum of Middle Age of Paris and Mrs Lagabrielle, Conservator, for facilitiesof sampling. 7. References

Ausset P. and LeFevre R. (1994). Ddbris de bois imbrfilds dans une couche grise ddposde entre les 126me and 176me si6cles sur la pierre de Saint-Trophime d'Arles (France). 3rd

International Symposium for the Conservation of Monuments in the mediterranean Basin, Venice, 243-249. Ausset P., Lef6vre R., Philippon J. and Venet C. (1994). Pr6sence constante de cendres volantes industrielles dans les crofites noires d'alt6ration superficielle de monuments frangais en calcaire compact. Comptes Rendus de l'Acaddmie des Sciences, Paris, t. 318, s6de II, 493499. Ausset P., Del Monte M., Bannery F. and Lef'evre R.(1998). Recording of pre-industrial atmospheric environment by ancient crusts on stone monuments. Atmospheric Environment 32, 16, 2859-2863. Blanc A. and Lorenz C. (1992). Ile-de-France et Champagne; Paris et Versailles, in Terroirs et Monuments de France, BRGM ed., Orl6ans, 111-118. Boureau E. (1956). Anatomie v6g6tale. 3 vol., Presses Universitaires de France. 752p. Braun R.C. and Wilson M.J.G (1970). The removal of atmospheric sulphur by building stones. Atmospheric Environment, 11, 1157-1162. Camuffo D., Del Monte M. and Sabbioni, C. (1983). Origin and growth mechanisms of the sulfated crusts on urban limestone. Water, Air and Soil Pollution, 19, 351-359. Del Monte M., Sabbioni C. and Vittori O. (1981). Airbom carbon particles and marble deterioration. Atmospheric Environment, 15, 5,645-652. Del Monte M. and Sabbioni C. (1987). Characterization of individual fly-ash particles emitted from coal and oil-fired power plants. Atmospheric Environment, 21, 12, 2737-2738. Del Monte M. and Rossi P. (1997). Fog and gypsum crystals on building materials. Atmospheric Environment, 31, 1637-1646. Del Monte M., Ausset P., Lef'evre R. and. Thi6bault S. (1999). Evidence of pre-industrial air pollution in Paris from the Heads of Kings of Juda Statues from Notre-Dame Cathedral.

Archaeometry (submitted). Erlande-Brandenburg A. (1982). Les sculptures de Notre-Dame de Paris au Mus6e de Cluny, Ed. R6union des Mus6es Nationaux, Paris, 133pp. Giscard d'Estaing F., Fleury M. and Erlande-Brandenburg A. (1977). Les Rois retrouv6s, Cu6not, Paris, 80p. Greguss P. (1955). Identification of Living Gymnosperms on the Basis of Xylotomy. Budapest, 263 p.


347

Greguss P. (1959). Holzanatomie der Europa'fschen Laubh61zer und Straficher. Akad6miai Kiado Budapest, 330p. Jacquiot C. (1955). Atlas d'anatomie des bois des conif'eres. Centre Technique du Bois. Paris 135p, 64 pl. 2 vol. Jacquiot C., Trenard Y. and Dirol D. (1973). Atlas d'anatomie des Bois des angiospermes. Centre Technique du Bois, Paris, 175p. 72 pl. 2 vol. Metclafe C.R. and Chalk L. (1950). Anatomy of the dicotyledons. Clarendon Press, Oxford, 2 vol., 1500p. Schweingruber F.H. (1978). Mikroskopische Holzanatomie. Ztircher a.g. zug, 98p1., 226p. Schweingruber F.H. (1990). Anatomie Europa'icher H61zer. w.s.l.f.n.p, haupt. 800p. Thi6bault S. (1986). Une approche du pal6oenvironnement v6g6tal par l'6tude des charbons de bois : l'anthracologie, 30 -33 et premiers r6sultat anthracologiques : le fait 3 de la zone 10 74-78. Pal~oenvironnement etfouilles urbaines, f6vrier 1986, Grand Louvre, Cour Napol6on. Solari M.E. and Thi6bault S. (1991). Les fouilles du jardin du Carrousel, r6sultats pr61iminaires de l'analyse anthracologique in : Van Ossel : Les jardins du Carrousel ~ Paris, fouilles 1989-1991, S.R.A. lle de France, vol. III, 47-56.


349

L A B O R A T O R Y I N V E S T I G A T I O N S OF W E A T H E R I N G B E H A V I O U R OF FRESH AND I M P R E G N A T E D L I M E S T O N E AND S A N D S T O N E F R O M C E N T R A L SWEDEN

Katarina Malaga-Starzec* GOteborg University, Department of Inorganic Chemistry, SE-412 96 G/3teborg, Sweden Torgny Sahlin GOteborg University, Earth Sciences Centre; Geology, P.O. Box 460, SE-405 30 G/3teborg, Sweden Oliver Lindqvist G6teborg University, Department of Inorganic Chemistry, SE-412 96 G0teborg, Sweden

Abstract The weathering of natural stone depends on its chemical and morphological properties. Two sedimentary rocks of different mineralogy quarried in Central Sweden were used in this study, sandstone and limestone. The main purpose of the project was to investigate differences in chemical weathering behaviour between fresh and impregnated samples. Impregnation material used for the study was potassium-based water-glass diluted with water, colloidal silica and Berol 048. The experimental part consisted of corrosion experiments in laboratory. The simulation of acid rain condition was performed in two atmospheric corrosion apparatus, suitable for short-term and long-term exposure. The rocks" sensitivity to chemical corrosion was correlated to results from physical tests: threepoint load bending strength, uniaxial compressive strength, abrasion resistance, water absorption, effective porosity and air-permeability. The results from chemical experiments showed significant sulphation of both fresh and impregnated limestone surfaces while no changes on sandstone surfaces were observed. Physical tests showed very small or no changes between fresh and impregnated sandstone while impregnated limestone showed higher bending strength, water absorption and air permeability compared to fresh samples. It could be concluded that impregnation material did not improve properties of the sandstone while a small improvement was noticed for the limestone.

Key words: Acid rain, Chemical weathering, Limestone, Sandstone, Laboratory.

1. Introduction Most natural stones found near the surface of the earth are formed under high temperature and pressure conditions and are not thermodynamically stable. Exposure to the chemical weathering, particularly water and dissolved gases, decomposes the original rock into substances that are stable in the surface environment. The weathering of natural stone depends on its chemical and morphological properties. The conventional stone tile used as building material, is untreated, except for cases of surface treatment, and at least 10 mm thick. A new method makes it possible to produce natural stone tiles only 4 mm thick. The natural stone material used in the study was


350


impregnated with a potassium-based water-glass using vacuum technique including repeated cycling between vacuum and atmospheric pressure. The purpose of the impregnation process was to increase the mechanical strength and the durability of the stone material. The limestone and the sandstone are aimed to be used as exterior cladding panels of buildings, on indoor floors, and as interior decoration in bathrooms and kitchens due to its low weight compared to the traditional thicker tiles. In order to analyse the effect of an accelerated air pollution impact on the sandstone and the limestone climatic chambers for stone-air pollution-humidity simulation have been used. Results from physical studies were used for evaluation of impregnation material's efficiency for improvement of rocks properties (Sahlin et al., in press). The aim of the project was to investigate differences in chemical weathering behaviour between fresh and impregnated natural materials. Efficiency of impregnation material examined during chemical tests was analysed and correlated to results from physical tests. Image analyses of thin sections and prisms were expected to express the rate of the stone decay as a function of the different variables (corrosive gases, humidity, temperature, mineralogy and porosity).

2. Experimental 2.1 Sample preparation The red-coloured fine-grained Dala sandstone, quarried in M~ngsbodama, and the very fine-grained red J~imtland limestone, quarried in the Brunflo area, both from central Sweden, were used in this study. The mineralogical composition for sandstone had quartzitic character (quartz -~70%, feldspar- 15%, rock fragments ---10%, muscovite/sericite -~2%) and for the limestone had carbonaceous character (calcite ---80%, quartz -~10%, hematite, fossils). Samples used for analysis were both fresh and impregnated with potassium-based water-glass diluted with water, colloidal silica and Berol 048 non-ionic surfactant. Colloidal silica has been developed by Eka Chemicals in Sweden, and used for material consolidation. It consisted of very fine amorphous silica particles (500 nm) in dispersion form that easily penetrates into pores and cracks. Amorphous silica is highly reactive pozzolan. It is completely inorganic material and thus environmentally friendly. Three different sample sizes were used in the study, small prisms 30 x 30 x 4 mm and 30 x 20 x 4 mm, and thin sections. The prisms were polished in water with silicon carbide abrasive paper to 500 meshes on all sides and thereafter placed in the high purity water and washed in an ultrasonic bath 3 times for 15 rain. The washed samples were vacuum-dried in a desiccator with a drying agent for one week and equilibrated at the relative humidity for another week. Thin sections were equilibrated at the same relative humidity as the prisms. Prior to exposure, a 0.12 mm thick nylon thread was bound around the sample for suspension in the exposure chamber. One sample at the time was used in each on-line exposure while eight samples in separate chambers were used in long-term experiment.

2.2 Corrosion chamber Short-term on-line gas analysis of SO2 adsorption catalysed by NO2 was performed in a single corrosion chamber (fig.l) and for long-term exposure to corrosive gases an apparatus, of the same type as short-term set-up, with eight separate chambers was used. The atmospheric corrosion testing apparatus for short-term (-~ 20 hours) analysis consisted of: molecular sieve preceded by an air drier, needle valves, humidifier, vessels for SO2 and


351

NO2 permeation tubes, mixer, thermometer, corrosion chamber, and thermostated water tank. The main difference between short-term and long-term corrosion testing apparatus was that the latter one consisted of eight single corrosion chambers, which were open sequentially, in order to attain equal flow through all chambers, and was not connected to on-line gas analysis. The long-term exposure lasted for circa five weeks. The utilisation of corrosion chambers was ideal for the determination of deposition velocities for single and multi-pollutant gas mixtures using ppm and ppb concentrations and for investigations of chemical reaction mechanisms.

Figure 1: The experimental on-line set-up for short-term exposures. The main advantage of using corrosion chambers was the high accuracy in the control and regulation of temperature, gas flows, relative humidity (RH) and concentrations of corrosive gases such as SO2 and NO2. The presence of oxidants, such as NO2, was found to enhance the corrosion process strongly, since they catalyse oxidation of the four-valent sulphur, or calcium sulphite, to sulphate (Johansson et al. 1991, Elfving et al. 1994). The experimental conditions used in the study are presented in tab. 1. Table 1: The experimental conditions. NO2 802 (ppm) (ppm) Exposure ~ . . 1.8 0.841 0.345 1.8 Short-term 0.143 1.8 1.8 0.061 0.902 0.193 Long-term

RH (%)

Air flow rate (L/min)

T~

95

22

95

22

352


3. Results and Discussion 3.1 Deposition rates of SO2 The deposition rates of the SO2 were measured with an on-line SO2 fluorescence instrument (Environment s.a.A.F. 21M). The results for tWO sedimentary rocks, sandstone and limestone, showed significant differences in reaction with corrosive atmospheres. Impregnation material, the potassium-based water-glass contained high amounts of silica, which was expected to increase resistance to acid rain and to improve physical properties of the rock material. Sandstone, both fresh and impregnated proved to be resistant to the impact of SO2 catalysed by the presence of NO2 (fig.2-4). The Dala sandstone consisted mostly of quartz (---70%), which is one of the least reactive minerals. Variation of the input concentration of SO2 (0.841, 0.345, 0.143, 0.061 ppm) did not influence significantly the adsorption on the surface. Impregnation material improved neither chemical nor physical properties of the sandstone. Fresh limestone, which consisted mostly of the very reactive calcite, CaCO3, showed higher adsorption of SO2 compared to the impregnated limestone and both sandstones. The level of adsorption of SO2 was dependent on the concentration of SO2 (fig.4). The highest difference for the adsorption between fresh and impregnated limestone was detected for the highest concentration of SO2. Also the highest adsorption was found when concentration of SO2 was highest (fig.2). The difference in adsorption of SO2 between fresh and impregnated samples was decreasing with decreasing concentration of SO2. For the lowest concentration of SO2 used in this study, 0.061 ppm, no difference occurred in adsorption between fresh and impregnated limestone and between fresh and impregnated sandstone (fig.3).

Time of exposure (h) Figure 2: Deposition of SO2 on fresh and impregnated limestone and sandstone. Input concentration of SO2 was 0.841 ppm.


353

Time of exposure (h) Figure 3: Deposition of 802 on fresh and impregnated limestone and sandstone. Input concentration of SO2 was 0.061 ppm.

Figure 4: Trends for the deposition rate for the limestone and the sandstone after 20 hours exposure.

354


3.2 Optical changes The thin films and samples from long-term exposures were optically analysed with a polarising microscope. In order to illustrate the surfaces changes a series of images were taken before the samples" exposure in the corrosion chamber and compared with the images taken at the same locations after exposure to corrosive gases. The results for fresh and impregnated sandstones showed no optical changes after exposure while image analysis of limestone indicated severe changes on the surface areas. For the fresh limestone the corrosion products were more abundant than for impregnated (fig.5). The dark areas corresponded to the corrosion products, gypsum, CaSO4 2H20, and mostly calcium sulphite hemihydrate, CaSO4 1/2H20. The white areas on Figure 5-2c indicated points resistant to corrosion that could correspond to the natural content of quartz in the rock in combination with the impregnation material.

Figure 5: Thin section microphotographs of fresh (1) and impregnated (2) limestone before exposure in the corrosion chamber (a and b) and after exposure (c) to atmospheres containing 0.841 ppm SO2 and 1.8 ppm NO2 and 95 % R.H. (a) in polarised light with crossed polars; (b) and (c) luminated from above in plane-polarised light.

3.3 Corrosion product characterisation The Environmental Scanning Electron Microscopy connected to an X-ray detector (ESEM-EDX) and X-ray diffractometry were used for characterisation of corrosion products obtained during short- and long-term exposures. The results from ESEM-EDX showed clear differences between concentrations of adsorbed SO2 on fresh and impregnated limestone after exposure (fig.6). The sulphur content was not detectable on the fresh

9th International Congress on Deterioration and Conservation of Stone, Venice 19.24 June 2000

355

limestone before the exposure. No differences between fresh and impregnated sandstone were detected. The results from X-ray diffraction analysis indicated formation of calcium sulphite hemihydrate and gypsum on the fresh and impregnated limestone. The Secondary Ion Mass Spectroscopy (SIMS) technique was used for detection of impregnation material. The impregnation material is very difficult to detect because it is amorphous. By using SIMS technique it was possible to see a slight trend in silica concentration along the pores and grain boundaries compared to the fresh samples.

Figure 6: ESEM-EDX elements mapping of fresh limestone before exposure (1) and fresh (2) and impregnated (3) limestone after long-term exposure.

4. Conclusions The chemical degradation of the studied sedimentary rocks was a complex phenomenon depending on a number of factors. The mineralogical composition of the natural material was the major factor responsible for the natural variety in stone reactivity. The calcium carbonate-rich limestone was naturally more reactive than quartz-rich sandstone under the same environmental conditions. The adsorption of SO2, and thereafter production of

356


hemihydrate and gypsum, was dependent on the corrosive gases concentration. The effectivity of impregnation material for the sandstone could not be observed while both chemical and physical properties of the limestone became improved.

5. References Elfving, P., Panas, I., Lindqvist. O., 1994. Model study of the first steps in deterioration of calcareous stone II. Sulphate formation on calcite. Applied Surface Science 78, 83-92. Johansson L-G., Lindqvist O., Mangio R., 1991. Corrosion of Calcareous Stones in Humid Air Containing SO2 and NO2. Durability Build. Mater. 5 439-449. Sahlin T., Malaga-Starzec K., Stigh J., Schouenborg B., (in press). Physical Properties and Durability of Fresh and Impregnated Limestone and Sandstone from Central Sweden Used for Thin Stone Flooring and Cladding. Submitted for publication in the proceedings of the 9 th International Congress on Deterioration and Conservation of Stone. Venice, Italy.

6. Materials The raw sandstone material used in this study is quarried by Wasasten AB, M&ngsbodama, SE-796 99 Alvdalen, Sweden. Tel: +46 251 540 00, fax" +46 251 540 10, email" [email protected]. The raw limestone material is quarried by AB J~imtlandskalksten, Box 25, SE-834 21 Brunflo, Sweden. Tel: +46 63 21860, fax" +46 63 22595, e-mail" [email protected]. The impregnated 4 mm natural stone tile products used in this study has been developed by two companies" Eka Chemicals AB (Eka Chemicals AB, SE-445 80, Sweden. Tel: +46 31 587000, fax" +46 31 587400, e-mail" [email protected]) and Techstone AB (Techstone AB, SE-471 99, Sweden. Tel" +46 304 676560, fax: +46 304 676569, e-mail" [email protected]). For the composition of the impreganation material see tab.2. The trade name for the sandstone is Alvdal Quartzite, and for the limestone J~imtland Limestone. The impregnated 4 mm products have the traditional trade name including Techstone|

Table 2" Components of the impregnation material. Density (kg/m 3) Components Potassium-based water1200 glass (SiO2/K20) Water 1000 Colloidal silica 1200 Berol 048 1020 *0.786 g is added per litre impregnation material solution.

Concentration (Weight %) 64.6 8.5 26.9

357

THE INFLUENCE OF BUILDING ORIENTATION ON CLIMATE WEATHERING CYCLES IN STAFFORDSHIRE,UK. David J. Mitchell* School of Applied Sciences, University of Wolverhampton, Wolverhampton WV11SB, UK David P. Halsey Gyford House, Hereford, HR96EX. UK Karl Macnaughton School of Applied Sciences, University ofWolverhampton, Wolverhampton, UK David E. Searle Built Environment Research Unit/School of Applied Sciences, University of Wolverhampton, Wolverhampton, UK. Abstract

Climate or meteorological induced cycles have been associated with weathering processes of building stone for a long time. Freeze-thaw cycles can have severe effects on stone disintegration. Although less dramatic, other cycles such as heating-cooling and wettingdrying will create similar stresses in stone. Diurnal and seasonal climatic cycles from standard meteorological measurements lack detail on the effects of short-duration rates of change, variations in subsurface gradients and aspect control. During certain synoptic conditions building orientation can have a great influence on climatic extremes of different facades. The use of sensors and data loggers has, to a large extent, opened up the potential for in-depth investigations of general climatic monitoring of the exterior of buildings and cyclic changes in temperature and moisture in stone. Using temperature and humidity sensors located on the four cardinal faces of the tower of Lichfield Cathedral, Staffordshire, UK, the frequency of heating-cooling, wetting-drying and freeze-thaw cycles have been analysed. For most of the cycles, west and south faces have the highest frequency especially the west. In contrast, the north and east have the lowest values, with the north having the least. Although this is a simplified interpretation, more climatic processes are operating on the south and west faces than the east and north faces. These could be loosely termed the 'maritime' and 'continental' faces respectively. In a parallel study of 30 sandstone churches in the West Midlands, the occurrence of 18 forms of weathering with respect to aspect were recorded. Granular disintegration, spalling, multiple flaking, total case hardened stone and total autotrophic stone were all found to be the greatest on the north aspect and, to some extent, on the east aspect. In contrast, relief weathering and total blackened stone were the greatest on the south and west aspects. Keywords: weathering cycles, heating-cooling, wetting-drying and freeze-thaw cycles, Triassic sandstone.

*Author to whom correspondence should be addressed.

358


1. Introduction Temperature and moisture conditions are key parameters in mechanical weathering processes of building stone. Rather than extreme meteorological conditions, cyclic changes have greater effects on stress and strains in stone and hence disintegration. Detailed temperature and moisture measurements, using sensors and data acquisition systems, are required to examine the changes oecurring as a result of mechanical weathering processes. The aim of this paper is to assess the influence of orientation on the frequency of climatic cycles. In order to review the potential of detailed monitoring, using sensors and data acquisition systems, a case study at Lichfield Cathedral, UK will be used (Halsey 1996, Halsey et al., 1998). 2. Methodology To assess the effects of aspect on climatic parameters, temperature and humidity sensors were installed 37 m above ground level on the central tower of Lichfield Cathedral, Stafford shire, England (fig 1).

Figure 1 Central tower of Lichfield Cathedral Steel sheathed thermistor probes and 'Vaisala' relative humidity probes were installed on the four cardinal faces of the tower, which is composed of Triassic sandstones (fig. 2). The thermistors (4 mm diameter) were inserted parallel to the exposed face 5 mm from the surface and held in place by thermally conductive adhesive to ensure thermal contact with the stone. The relative humidity probes (25 mm diameter) were inserted in 25 mm deep holes, normal to the surface. Silicon sealant was used to provide airtight seal so that the R.H. of a pocket of air, trapped to a depth of 25 mm beneath the surface, was recorded. Raw data from each sensor was recorded at 15 minutes intervals on a Unidata data logger. The large data set provided a detailed record of the temperature and relative humidity of the four cardinal faces of the tower from January 1994 to November 1996. Furthermore, the time intervals enabled shorter cycles to be calculated besides diurnal changes. Using


359

mathematical and logic functions, heating-cooling, wetting-drying and freeze-thaw cycles were derived (tab. 1). To calculate these factors, it is assumed that when the sandstone RH is _>80%, moisture is present in the stone. This is based on the findings of Sereda (1974), which show at inland locations plastic and metal surfaces were wet at a RH of =80%. This assumption has also been used to measure the time of wetness by Henriksen et al., (1993). Climatic data for the same period (January 1994 to November 1996) were extracted from the 'Weather Log" for Elmdon, 24 km south of Lichfield Cathedral.

Figure 2. Temperature and relative humidity sensors in position Table 1. Factors calculated using temperature and relative humidity data, and their ....... descrip tions Factor Description " Rates of temperature Temperature changes per 15 minutes. change (monthly mean, maximum and minimum) Heating-cooling cycles (number month- 1)

Cycles consist of temperatures increasing and then decreasing. When temperatures start to rise again a new cycle begins.

Wetting-drying cycles (number month- 1)

Cycles consist of relative humidity increasing and then decreasing. When relative humidity starts to rise again a new cycle begins.

Time of wetness (minutes month -1 )

The length of time relative humidity is >_80%.

Freeze-thaw cycles (number month- 1)

Freezing occurs when relative humidity is >_80% and temperature is _0.45 lam retained are reported for 64 weeks (14/8/98-11/1/99). Chemical analysis determined from ICP and Ion chromatography is for 40 and 18 weeks respectively. 3.1 Precipitation run-off Table 2 shows the results obtained for cumulative precipitation runoff from the two stone types. When all four treatments on the limestone samples were compared using an ANOVA and subsequent post-hoc analysis, there was strong evidence (P AI4(OH)8[Si4Olo] + 2Me2CO3 + 8SIO2 In a recent review on feldspar weathering, Blum (1994) subdivides the feldspar weathering process into two main stages, dissolution and precipitation; he states that "...feldspar weathering occurs via dissolution of all components, with the subsequent precipitation of secondary minerals from solution". Dissolution rates for feldspars are very slow and these rates are determining the overall weathering reaction rate. Feldpars of all compositions have dissolution rates that increases with decreasing pH at pH50~ < 1 0 0 % ; + = < 5 0 % ; - = no growth.

-

3.2.2 Activity againstfungi In laboratory conditions, fungi were sensitive to the treatment with ALGOPHASE | In fact, reduction of the fungal growth was observed in all samples with a percentage between 88% and 99.5% (Fig. 7). It is worth noting that the new aqueous formulation pH025/d was more effective against fungi with 100% inhibition of growth in all suspensions tested.


537

3.2. 3 Activity against bacteria Both biocides ALGOPHASE | and pH025/d were ineffective against bacteria. In fact, no reduction of the bacterial growth was observed in some samples. In the other cases a percentage of reduction of the growth between 85.2% and 97.6% (Fig. 8) was observed.

Figure 7. Effect of the biocide ALGOPHASE | and of the new formulation pH025/d on the mycoflora and percentage of reduction of the growth.

Figure 8. Effect of the biocide ALGOPHASE | and of the new formulation pH025/d on the bacteria.

53 8


4. Discussion

Microbiological analysis showed that after 8 years from treatment with biocide ALGOPHASE | a re-colonisation of the surfaces of the "Templete de Mudejar" by algae, cyanobacteria, fungi and bacteria for all orientations occurred. Microbial re-colonisation was demonstrated especially in the lower parts (between the height of 0-50 cm) in which a heavy biological patina was evident. Favourable conditions for the growth of phototrophic microorganisms were due to ascending humidity caused by the position of the temple located at the centre of a small pond in the cloister garden, as already reported by Arifio et al. (2000). These Authors reported a list of phototrophic organisms isolated and put into evidence their opportunistic and ubiquitous behaviour. However, laboratory tests showed that ALGOPHASE | was able to dramatically reduce the amount of algae, cyanobacteria and fungi present in the samples suspension. These data seem to demonstrate that the biocide was no longer present (or efficient) "in situ" in the temple. Other Authors' data demonstrated that ALGOPHASE | is able to inhibit the growth of phototrophic microorganisms for a period of 5-6 years (Gomez-Bolea et al., 1999; Pietrini et al., 1999). It is known that different factors could influence the efficiency of biocide treatments; the environmental conditions of exposition of treated materials, among which sunlight radiation and rainfall washing-out, could play an important role (Nugari, 1999). In fact, in this case the "humidity" should be considered as a key factor for the durability of the biocide treatment.

5. Conclusion

We can conclude that application of ALGOPHASE | can be useful to slow down lichens, algal, cyanobacterial and fungal re-colonisation processes, if applied within intervals of time not exceeding 8 years and in environmental conditions relatively free of humidity. Because better results were obtained with the new aqueous formulation pH025/d, particularly against funsal growth, we suggest, when it is possible, to apply it. Concerning bacteria, ALGOPHASE ~ failed to inhibit their growth and thus a specific product against bacteria should be applied. 6. References

Arifio X., Canals A., Gomez-Bolea A., Saiz-Jimenez C. Re-colonisation of the Templete Mudejar, Guadalupe Monastery, Caceres, Spain, after silicone/biocide treatment. The Conservation of Monuments in the Mediterranean Basin Proc. of the 5th International Symposium, 5-8 April, 2000, Seville, Spain. Balzarotti-Kammlein R., Sansoni M., Castronovo A., 1999. An innovative watercompatible formulation of ALGOPHASE | for treatment of mortars. Microbiology and Conservation of Microbes and Art (ICMC '99), Proc. of the International Conference. Firenze. 217-220. Commissione Normal, 1990. Raccomandazioni Normal: 9/88. Microflora autotrofa ed eterotrofa: tecniche di isolamento in coltura. C.N.R. - I.C.R., Roma.


539

Commissione Normal, 1991. Raccomandazioni Normal: 30/89. Metodi di controllo del biodeterioramento. C.N.R. - I.C.R., Roma. Commissione Normal, 1994. Raccomandazioni Normal: 38/93. Valutazione sperimentale dell'efficacia dei biocidi. C.N.R. - I.C.R., Roma. Fassatiov~i O, 1986. Moulds and filamentous fungi in technical microbiology. Progress in Industrial Microbiology. Elsevier, Amsterdam. Gomez-Bolea A., Arifio X., Balzarotti R., Saiz-Jimenez C., 1999. Surface treatment of stones: conseguences on lichenic colonisation. Microbiology and Conservation of Microbes and Art (ICMC '99), Proc. of the International Conference. Firenze. 233-237. Nugari M. P., 1999. Interference of antimicrobial agents on stone. Microbiology and Conservation of Microbes and Art (ICMC '99), Proc. of the International Conference. Firenze. 211-214 Pietrini A.M., Ricci S., Bartolini M., 1999. Long-Term evaluation of biocide efficacy on algal growth. Microbiology and Conservation of Microbes and Art (ICMC '99), Proc. of the International Conference. Firenze. 238-245. Urzi, C., Realini M., 1998. Colour changes of Noto's calcareous sandstone as related to its colonisation by microorganisms. International Biodeterioration and Biodegradation, 42, 45-54. Urzi C., De Leo F., Salamone, P. Rapid survey of marble colonisation using the adhesive tape stripes. Eurocare-Euromarble EU 49, Proc. of the 10th Workshop, Stocholm, Sweden, In press.

Acknowledgements This study was supported by the financial contribution of European community through the EC projects ENV4-CT98- 0707 and Murst 60%. We also like to thank Mr. William Fenton for his kind revision of the English text.


Theme 4 Laboratory methods and techniques


543

INSTRUMENTAL CHEMICAL ANALYSIS OF THE MORE COMMON MARBLES HISTORICALLY USED FOR DECORATIVE PURPOSES OR TO CREATE WORKS OF ART L. Campanella, E. Gregori, R. Grossi, M. Tomassetti* Dipartimento di Chimica, Universitb. di Roma "La Sapienza" G. Bandini Soprintendenza Archeologica di Roma, Museo Nazionale Romano

Abstract The aim of the work is to make concise records, several prototypes of which are illustrated in the present note, of a series of experimental data obtained using instrumental chemical methods applied to the more common marbles historically used to create artistic artefacts. This chemical information, accompanied by relevant mineralogical data, should speed up or at least facilitate the identification of the type of marble used to create many marble artistic finds about which little is still known. Keywords: marbles, chemical instrumental analysis, mineralogycal data. 1. Introduction We are currently carrying out basic chemical research on the most common types of marble historically used both for decorative purposes and to make marble statues. Numerous studies and researches of different kinds have already been carried out on several of the more common types of marble [L. Lazzarini, 1980; D. Cordischi,1983; K.J. Matthews,1997], mostly of a mineralogical nature [M. Pieri, 1966]. Nevertheless we consider that a collection of data obtained using modem instrumental chemical methods can provide a valid tool to be used for different purposes, above all to identify accurately and rapidly the type of marble used to create a number of works of art as yet inadequately studied. Furthermore, a knowledge of the chemical characteristics of the various types of marble is particularly important for conservation purposes and a correct identification of the type of marble through its chemical characteristics frequently makes it possible to determine the place of origin and to confirm and supplement the historiographic data which often already exist in the case of many marble works of art. 2. Results In pratice, the study under way, focused on marbles such as Paros, Pentelic, Aphyon, Lasa, Ephesus, several Carrara varieties, etc., entails systematic analysis carried out using inductively coupled plasma (ICP), X-ray diffractometry and thermoanalysis (TG, DTG and DTA) techniques. The chemical data thus obtained were further supplemented and completed by mineralogical and stratigraphic data obtained via optical microscopy. For each type of marble investigated all the data collected have been gathered in recapitulatory records of which examples are shown in the following for two Greek marbles Pentelic and Paros, and Italian marble, Carrara-Altissimo, and a Turkish marble, white Aphrodisias.


544


2.1 Pentelie Marble

Provenance:Greece Colour: white Mineralogical-petrographic analysis under optical microscope. Structure: polygonal granuloblastic homeoblastic with prevalent straight line contact among grains. Texture: isotropic. Identifiable minerals: calcite granules containing very rare, probably dolomite, rhombohedrons. Mineral size: varies from 0.6 mm to about 0.2 mm; the most common size is about 0.4 mm. Instrumental chemical analysis.

fig. 1 fig. 2 fig. 1 TG and DTG curves of Pentelic marble. Heating rate 10~ min-l; air stream: 100 cm 3 min-1. fig.2 DTA curve of Pentelic marble. Heating rate 10~ min -1 ; air stream: 100 cm3min -l

Thermogravimetric data 9 First step mass loss % T oC 283 9.3 403 557

Second step mass loss % T oc 557 40.5 716 850

residue % at 900~ 49.6

Plasma Emission (ICP) data; values in 9pm Ca

K

Mg

A1

309754

340

314

552

Si

Fe

Sr

Mn

Zn

54

58

14

-

Xra ~ ffraction data 3.85 3.03 2.49 2.09 dhkl (12) (100) (14) (18) and intensity ( ) Calcite found in literature (*) CaCO 3 dhk 1 found 3.86 3.03 2.50 2.09 experimentally in the sample (*) Mineral Powder Diffraction File - Data Book-1976

Pb

Cr

Cu

0.06

1.93

1.91

1.87

1.62

1.60

1.52

(5)

(17)

(17)

(4)

(8)

(5)

1.93

1.91

1.87

1.62

1.60

1.52

9th I n t e r n a t i o n a l Congress on Deterioration and Conservation of Stone, Venice 19-24 June 2000

545

2.2 Paros Marble

Provenance: Greece Colour : white with yellowish lamellae Mineralogical-petrographic analysis under optical microscope. Structure: polygonal crystalloblastic; contact among granules ranges from irregular to straight line Texture: isotropic. Identifiable minerals: calcite granules including occasional, probably dolomite, rhombohedrons, rare white muscovite mica, minute opaque mineral granulations. Mineral size: varies from 1 mm to about 0.05 mm; the common size is about 0.5 mm. NB: a thin irregular vein of tiny opaque reddish brown coloured granules runs through the stone.

Instrumental chemical analysis.

fig. 3

fig. 4

fig.3 TG and DTG of Paros marble. Heating rate 10~ min -1 ; air stream: 100 cm3min -I fig.4 DTA curve of Paros marble. Heating rate 10~ min -1 ; air stream: 100 cm3min -1

Thermogravimetric data First step mass loss % 1.8

T ~C 322 448 536

Second step mass loss % T oC 536 716 43.1 850

residue % at 900 ~ 53.2

Plasma Emission (ICP) data; values in p ~m Ca

K

Mg

A1

Si

Fe

Sr

Mn

Zn

364014

1368

1307

789

-

660

122

15

5

Pb

Cr

Cu

546


Xray d~ffraction data

Calcite CaCO 3

dhkl and intensity ( ) found in literature (*) dhkl found experimentally in the sample

3.85

3.03

2.49

2.09

1.93

1.91

1.87

1.62

1.60

1.52

(12)

(100)

(14)

(18)

(5)

(17)

(17)

(4)

(8)

(5)

2.50

2.09

1.93

1.91

1.87

1 . , 6 2 1.60

1.52

3.85

3.03

2.3 White Aphrodisias Marble Provenance: Turkey Colour : white Mineralogical-petrographic analysis under optical microscope. Structure: highly heteroblastic and crystalloblastic. Texture: foliate schistose. Identifiable minerals: coarse levels composed of faintly bismarck coloured calcite granules with irregular contact among granules and little quartz; the fine grained levels are composed of elongated orientated carbonate crystals (the greatest of calcite and the smallest of dolomite) with irregular contacts among granules, little quartz, of elongated and orientated crystals and small quantities of orientated white mica. Mineral size: varies from 1 mm to about 0.025 mm. The rock is composed of alternate coarse grained (about one millimetre) and very fine grained (0.025 mm) layers.

Instrumental chemical analysis.

f

l

._

100.._.. =-.

Proceedings of the 9th International Congress on Deterioration and Conservation of Stone

Proceedings of International congress of mathematicians

Proceedings of International congress of mathematicians

Proceedings of International congress of mathematicians

Proceedings of the Ninth International Congress on Mathematical Education

Illustrated Glossary on Stone Deterioration Patterns

Proceedings of the International Congress of Mathematicians, August 21-29, 1990, Kyoto, Japan (International Congress of Mathematicians Proceedings)

Proceedings of the 9th Workshop on Algorithm Engineering and Experiments

Trends in Acarology: Proceedings of the 12th International Congress

Proceedings Of The 15th Scandinavian Congress

Proceedings of the International Conference on Colloid and Surface Science

Proceedings of The International Congress of Mathematicians 2010 (ICM 2010): Vol. I: Plenary Lectures and Ceremonies

Logic, Methodology and Philosophy of Science: Proceedings of the 1960 International Congress

Advances in Electrocardiology 2004: Proceedings of the 31th International Congress on Electrocardiology (Advances in Electrocardiography)

11th International Congress on Catalysis - 40th Anniversary

11th International Congress on Catalysis - 40th Anniversary

XIVth International Congress on Mathematical Physics

Xvith International Congress on Mathematical Physics

Chaos and Fractals (International Congress)

Logic, Methodology and Philosophy of Science VIII: Proceedings of the Eighth International Congress of Logic, Methodology and Philosophy of Science

Proceedings of The International Congress of Mathematicians 2010 (ICM 2010): Vol. IV

Proceedings of the International Congress of Mathematicians, 14 - 21 Aug 1958

THE NEW PANORAMA OF ANIMAL EVOLUTION Proceedings XVIII International Congress of Zoology

Proceedings of the International Congress of Mathematicians 15-22 August 1962

Proceedings of the Fifth International Congress of Mathematicians (Cambridge, 22-28 August 1912) - Volume II

Proceedings of the International Congress of Mathematicians, Madrid 2006 Volume 3

Proceedings of the 25th International Congress of Papyrology, Ann Arbor, July 29-August 4, 2007

Proceedings of The International Congress of Mathematicians 2010 (ICM 2010): Vol. II

Proceedings of The International Congress of Mathematicians 2010 (ICM 2010): Vol. III

Proceedings of the Fifth International Congress of Mathematicians (Cambridge, 22-28 August 1912) - Volume I

Proceedings of the Fifth International Congress of Mathematicians (Cambridge, 22-28 August 1912) - Volume I

Proceedings of the 9th International Congress on Deterioration and Conservation of Stone

Proceedings of International congress of mathematicians

Proceedings of International congress of mathematicians

Proceedings of International congress of mathematicians

Proceedings of the Ninth International Congress on Mathematical Education

Illustrated Glossary on Stone Deterioration Patterns

Proceedings of the International Congress of Mathematicians, August 21-29, 1990, Kyoto, Japan (International Congress of Mathematicians Proceedings)

Proceedings of the 9th Workshop on Algorithm Engineering and Experiments

Trends in Acarology: Proceedings of the 12th International Congress

Proceedings Of The 15th Scandinavian Congress

Proceedings of the International Conference on Colloid and Surface Science

Proceedings of The International Congress of Mathematicians 2010 (ICM 2010): Vol. I: Plenary Lectures and Ceremonies

Logic, Methodology and Philosophy of Science: Proceedings of the 1960 International Congress

Advances in Electrocardiology 2004: Proceedings of the 31th International Congress on Electrocardiology (Advances in Electrocardiography)

11th International Congress on Catalysis - 40th Anniversary

11th International Congress on Catalysis - 40th Anniversary

XIVth International Congress on Mathematical Physics

Xvith International Congress on Mathematical Physics

Chaos and Fractals (International Congress)

Logic, Methodology and Philosophy of Science VIII: Proceedings of the Eighth International Congress of Logic, Methodology and Philosophy of Science

Proceedings of The International Congress of Mathematicians 2010 (ICM 2010): Vol. IV

Proceedings of the International Congress of Mathematicians, 14 - 21 Aug 1958

THE NEW PANORAMA OF ANIMAL EVOLUTION Proceedings XVIII International Congress of Zoology

Proceedings of the International Congress of Mathematicians 15-22 August 1962

Proceedings of the Fifth International Congress of Mathematicians (Cambridge, 22-28 August 1912) - Volume II

Proceedings of the International Congress of Mathematicians, Madrid 2006 Volume 3

Proceedings of the 25th International Congress of Papyrology, Ann Arbor, July 29-August 4, 2007

Proceedings of The International Congress of Mathematicians 2010 (ICM 2010): Vol. II

Proceedings of The International Congress of Mathematicians 2010 (ICM 2010): Vol. III

Proceedings of the Fifth International Congress of Mathematicians (Cambridge, 22-28 August 1912) - Volume I

Proceedings of the Fifth International Congress of Mathematicians (Cambridge, 22-28 August 1912) - Volume I

Recommend Documents