Cutting tool technology: industrial handbook

Cutting Tool Technology Previous books for Springer Verlag by the author: Advanced Machining: The Handbook of Cutting ...

Author: Graham T. Smith

623 downloads 4228 Views 21MB 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

Cutting Tool Technology

Previous books for Springer Verlag by the author: Advanced Machining: The Handbook of Cutting Technology (1989) CNC Machining Technology series: Book 1: Design, Development and CIM strategies Book 2: Cutting, Fluids and Workholding Technologies Book 3: Part Programming Techniques (1993) CNC Machining Technology: Library Edition (1993) Industrial Metrology: Surfaces and Roundness (2002)

Graham T. Smith

Cutting Tool Technology Industrial Handbook

123

Graham T. Smith, MPhil (Brunel), PhD (Birmingham), CEng, FIMechE, FIEE Formerly Professor of Industrial Engineering Southampton Solent University Southampton U. K.

ISBN 978-1-84800-204-3

e-ISBN 978-1-84800-205-0

DOI 10.1007/978-1-84800-205-0 British Library Cataloguing in Publication Data Smith, Graham T., 1947– Cutting tool technology: industrial handbook 1. Metal-cutting 2. Metal-cutting tools I. Title 671.3'5 ISBN-13: 9781848002043 Library of Congress Control Number: 2008930567 © Springer-Verlag London Limited 2008 Apart from any fair dealing for the purposes of research or private study, or criticism or review, as permitted under the Copyright, Designs and Patents Act 1988, this publication may only be reproduced, stored or transmitted, in any form or by any means, with the prior permission in writing of the publishers, or in the case of reprographic reproduction in accordance with the terms of licences issued by the Copyright Licensing Agency. Enquiries concerning reproduction outside those terms should be sent to the publishers. The use of registered names, trademarks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant laws and regulations and therefore free for general use. The publisher makes no representation, express or implied, with regard to the accuracy of the information contained in this book and cannot accept any legal responsibility or liability for any errors or omissions that may be made. Cover illustration: eStudio Calamar S.L., Girona, Spain Printed on acid-free paper 9 8 7 6 5 4 3 2 1 springer.com

Preface

Just over twenty years ago I began writing a book, the forerunner to this present volume for Springer Verlag, entitled: Advanced Machining – The Handbook of Cutting Technology. This original book covered many of the topics discussed here, but in a more general and less informative manner. Since this previous volume was published, many of the tooling-related topics are now more popular, or have recently been developed. Typical of these latter topics, are both High-speed and Hard-part machining that have now come to the fore. While Micro-machining and Artificial Intelligence (AI) coupled to neural network tool condition monitoring have become important, the latter from a research perspective. These machining and tooling topics, plus many others have been included herein, but only in a relatively concise manner. It would have been quite possible to write a book of this length just concerned with say, drilling techniques and associated tooling technologies alone. With the concerns raised on the health hazards to operational personnel exposed to cutting fluid mists in the atmosphere, the permissible exposure levels (PEL’s) have been significantly reduced recently. Further, with the advent of Near-dry and Dry-machining strategies, they have played a important role of late, particularly as their disposal and attendant costs have become of real consequence. Tool management issues previously discussed in the ‘Advanced Machining’ book have hardly changed, because when I wrote this chapter over two decades ago, most of today’s tooling issues by then had been addressed. However, the toolpresetting machines and associated software now, are far more advanced and sophisticated than was the case then, but the well-organised and run tool preparation ‘rules’ are still applicable today. One area of cutting tool development that has seen significant design novelty, is in the application of

Multi-functional tooling. Here, the chip control development is facilitated by both chip-narrowing and -vectoring, being achieved by computer-generated insert design, to position raised protrusions–‘embossed dimples’, on the top face. Further, some cutting insert toolholders are designed for controlled elastic compliance – giving the necessary clearance as the tool is vectored along and around the part’s profile, allowing a range of plunge-grooving and forming operations to be simultaneously undertaken by just this one tool. Coating technology advances have enabled significant progress to be made in both Hard-part machining and for that of either abrasive and work-hardened components. Some coating techniques today approach the hardness of natural diamond, particularly the aptlynamed ‘diamond-like coatings’ (DLC). Recently, one major cutting tool company has commercially-introduced an ‘atomically-modified coating’, such is the level of tool coating sophistication of late. Potential problems created by utilising faster cutting data often without benefit and use of flood coolant in cutting technology applications, has had an influence on the resulting machined surface integrity of the component. This sub-surface damage is often disguised, or not even recognised as a problem, until the part catastrophically fails in-service – as a result of the instability produced by the so-called ‘white-layering effect’. While another somewhat unusual factor that has become of some concern, is in either handling, or measuring miniscule components produced by Micromachining techniques. Often a whole month’s mass production of such diminutive machined parts could easily be fitted into a small shoebox! All of these previously mentioned tooling-related challenges and many others have to a certain extent, now become a reality. While other technical and machining factors are emerging that must be techni-

VI

Preface

cally-addressed, so that cutting tool activities continue to expand. It is a well acknowledged fact that if one was to list virtually all of our modern-day: domestic; medical; industrial; automotive; aerospace, etc; components and assemblies, they would to some extent rely on machining operations at a certain stage in their subsequent manufacturing process. These wide-ranging manufactured components clearly show that there is a substantive machining requirement, which will continue to grow and thus be of prime importance for the foreseeable future. This present book: ‘Cutting Tool Technology – Industrial Handbook’, has been written in a somewhat pragmatic manner and certain topics such as ‘Machining Mechanics’ have only been basically addressed, as they are well developed elsewhere, as indicated by the referenced material at the end of each chapter. Any book that attempts to cover practical subject matter such as that of cutting technology, must of necessity, heavily rely on information obtained from either one’s own machining and research experiences, or from indus-

trial specialist journals. I make no apology for liberally quoting many of these industrial and research sources within the text. However, I have attempted – wherever possible – to acknowledged their contributions when applicable, in either the references, or in the associated diagrammatical and pictorial figures herein. Further, it is hoped that the ‘machining practitioner’ can obtain additional information and some solutions and explanations from the relevant appendices, where amongst other topics, are listed a range of ‘trouble-shooting guides’. Finally, it is hoped that this latest book: ‘Cutting Tool Technology – Industrial Handbook’ will offer the ‘machining practioner’ the same degree of support as the previous book (i.e. Advanced Machining – The Handbook of Cutting Technology) achieved, from the significant feed-back obtained from practitioners and readers who have contacted me over the past decades. Graham T. Smith Fortuna, Murcia, Spain

Acknowledgements

First and foremost, I would like to express my sincere thanks to my wife Brenda for her support and for the time I have taken, whilst writing this book: Cutting Tool Technology – Industrial Handbook. I could not have achieved such an in-depth treatment and reasonably comprehensive account of the subject matter without her unstinting co-operation and help. A book that relies heavily on current industrial practices could not have been produced without the unconditional support from specific tooling manufacturers and the machine tool industries. I would like to particularly single-out one major cutting tool company, to genuinely thank everyone at Sandvik Coromant who have provided me with both relevant and significant: information; photographic; and diagrammatic support – the book would have been less relevant without their indefatigable co-operative help and discussion. Likewise, other tooling companies have been of much help and assistance in the preparation of this book, such as: Seco Tools; Kennametal Hertel and Kennametal Inc; Iscar Tools; Ingersoll; Guhring; Sumitomo Electric Hardmetal Ltd; Mitsubishi Carbide; Horn (USA); Shefcut Tool and Engineering Ltd; Rotary Technologies Corp; Diashowa Tooling; Centreline Machine Tool Co Ltd; DeBeers – element 6; Walter Cutters; Widia Valenite; TRW – Greenfield Tap and Die; Triple-T Cutting Tool, Inc; Hydra Lock Corp; Tooling Innovations; and Microbore Tooling Systems. Several machine tool companies have been invaluable in providing information, notably: Cincinnati Machines; Yamazaki Mazak;

Dorries Scharmann; DMG (UK) Ltd; Giddings and Lewis; Starrag Machine Tool Co; and E. Zoller GmbH and Co KG. While other tooling-based and associated companies have also provided considerable information, including: Renishaw plc; Kistler Instrumente AG; Taylor Hobson plc; Mahr/Feinpruf; Cimcool; Kuwait Petroleum International Lubricants; Edgar Vaughan; Pratt Burnerd International; Lion Precision; Westwind Air Bearings Ltd; Third Wave AdvantEdge; Susta Tool Handling; Tooling University. I have listed the main companies above, rather than attempting to name individuals within each company, otherwise the list would be simply vast. However, I would like to express my gratitude to each one of them, personally. I would also like to acknowledge the breadth and depth of information obtained from industrially-based journals, such as: Cutting Tool Engineering; American Machinist; Metalworking Production; Machinery and Production Engineering. The publishers of this book Springer, have been most patient with me as I have attempted to meet extended deadlines for the manuscript, for which I am indebted to and can only offer my sincerest thanks. Lastly, if any unfortunate mistakes have inadvertently slipped into the text, or misinterpretations in the draughting of any line diagrams have occurred, it is solely the author’s fault and does not represent any of the companies, or their products, nor that of the individuals mentioned. Graham T. Smith

Contents

1 Cutting Tool Materials . . . . . . . . . . . . . . . . . 1 1.1 Cutting Technology – an Introduction .. . . 2 1.1.1 Rationalisation .. . . . . . . . . . . . . . . . . 2 1.1.2 Consolidation .. . . . . . . . . . . . . . . . . . 4 1.1.3 Optimisation .. . . . . . . . . . . . . . . . . . . 4 1.2 The Evolution of Cutting Tool Materials 7 1.2.1 Plain Carbon Steels .. . . . . . . . . . . . . 7 1.2.2 High-Speed Steels . . . . . . . . . . . . . . . 7 1.2.3 Cemented Carbide . . . . . . . . . . . . . . 8 1.2.4 Classification of Cemented Carbide Tool Grades .. . . . . . . . . . . . 12 1.2.5 Tool Coatings: Chemical Vapour Deposition (CVD) .. . . . . . 14 1.2.6 Diamond-Like CVD Coatings . . . 14 1.2.7 Tool Coatings: Physical Vapour Deposition (PVD) . . . . . . . 17 1.2.8 Ceramics and Cermets .. . . . . . . . . . 19 1.2.9 Cermets – Coated . . . . . . . . . . . . . . . 23 1.2.10 Cubic Boron Nitride (CBN) and Poly-crystalline Diamond (PCD) .. . . . . . . . . . . . . . . . . . . . . . . . . 25 1.2.11 Natural Diamond . . . . . . . . . . . . . . . 29 2 Turning and Chip-breaking Technology 2.1 Cutting Tool Technology . . . . . . . . . . . . . . . . 2.1.1 Turning – Basic Operations . . . . . 2.1.2 Turning – Rake and Clearance Angles on Single-point Tools .. . . . 2.1.3 Cutting Insert Edge Preparations 2.1.4 Tool Forces – Orthogonal and Oblique .. . . . . . . . . . . . . . . . . . . . 2.1.5 Plan Approach Angles . . . . . . . . . . . 2.1.6 Cutting Toolholder/Insert Selection .. . . . . . . . . . . . . . . . . . . . . . .

33 34 34 34 36 39 41 43

2.2 History of Machine Tool Development and Some Pioneers in Metal Cutting . . . 2.2.1 Concise Historical Perspective of the Development of Machine Tools . . . . . . . . . . . . . . . . . . . . . . . . . 2.2.2 Pioneering Work in Metal Cutting – a Brief Resumé . . . . . . 2.3 Chip-Development .. . . . . . . . . . . . . . . . . . . 2.4 Tool Nose Radius . . . . . . . . . . . . . . . . . . . . . 2.5 Chip-Breaking Technology . . . . . . . . . . . . 2.5.1 Introduction to Chip-Breaking 2.5.2 The Principles of Chip-Breaking 2.5.3 Chip-Breakers and Chip-Formers .. . . . . . . . . . . . 2.5.4 Helical Chip Formation .. . . . . . . 2.5.5 Chip Morphology . . . . . . . . . . . . . 2.5.6 Chip-Breaker Wear .. . . . . . . . . . . 2.6 Multi-Functional Tooling .. . . . . . . . . . . . .

50 50 51 54 62 66 66 68 69 71 75 79 79

3 Drilling and Associated Technologies 87 3.1 Drilling Technology . . . . . . . . . . . . . . . . . . 88 3.1.1 Introduction to the Twist Drill’s Development . . . . . . . . . . . 88 3.1.2 Twist Drill Fundamentals .. . . . . 88 3.1.3 The Dynamics of Twist Drilling Holes .. . . . . . . . 96 3.1.4 Indexable Drills . . . . . . . . . . . . . . . 103 3.1.5 Counter-Boring/Trepanning . . . 107 3.1.6 Special-Purpose, or Customised Drilling and Multi-Spindle Drilling .. . . . . . . . . . . . . . . . . . . . . . 110 3.1.7 Deep-Hole Drilling/ Gun-Drilling . . . . . . . . . . . . . . . . . 113 3.1.8 Double-Tube Ejector/ Single-Tube System Drills .. . . . . 115

Contents

3.1.9

Deep-Hole Drilling – Cutting Forces and Power .. . . . . 3.2 Boring Tool Technology – Introduction 3.2.1 Single-Point Boring Tooling .. . . 3.2.2 Boring Bar Selection of: Toolholders, Inserts and Cutting Parameters .. . . . . . . 3.2.3 Multiple-Boring Tools . . . . . . . . . 3.2.4 Boring Bar Damping .. . . . . . . . . 3.2.5 ‘Active-suppression’ of Vibrations .. . . . . . . . . . . . . . . . . 3.2.6 Hard-part Machining, Using Boring Bars .. . . . . . . . . . . . 3.3 Reaming Technology – Introduction . . . 3.3.1 Reaming – Correction of Hole’s Roundness Profiles . . . 3.3.2 Radially-Adjustable Machine Reamers . . . . . . . . . . . . . 3.3.3 Reaming – Problems and Their Remedies . . . . . . . . . . . 3.4 Other Hole-Modification Processes . . . . 4

117 117 118 122 124 126 127 128 133 135

6

139

6.1 6.2

142 142

4.4 4.5

172 177 177

5 5.1 5.2 5.3

Threading Technologies . . . . . . . . . . . . . . Threads .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . Hand and Machine Taps .. . . . . . . . . . . . . . Fluteless Taps .. . . . . . . . . . . . . . . . . . . . . . . .

4.2 4.3

6.3 6.4

Milling Cutters and Associated Technologies . . . . . . . . . Milling – an Introduction . . . . . . . . . . . . . 4.1.1 Basic Milling Operations .. . . . . . 4.1.2 Milling Cutter Geometry – Insert Axial and Radial Rake Angles 4.1.3 Milling Cutter – Approach Angles .. . . . . . . . . . . . . . . . . . . . . . . 4.1.4 Face-Milling Engagement – Angles and Insert Density .. . . . . 4.1.5 Peripheral Milling Cutter Approach Angles – Their Affect on Chip Thickness 4.1.6 Spindle Camber/Tilt – when Face-Milling . . . . . . . . . . . . Pocketing, Closed-Angle Faces, Thin-Walled and Thin-Based Milling Strategies . . . . . . . . . . . . . . . . . . . . . Rotary and Frustum-Based Milling Cutters – Design and Operation . . . . . . . Customised Milling Cutter Tooling .. . . . Mill/Turn Operations . . . . . . . . . . . . . . . . .

4.1

5.4 Threading Dies . . . . . . . . . . . . . . . . . . . . . . . 5.5 Thread Turning – Introduction . . . . . . . . 5.5.1 Radial Infeed Techniques . . . . . . 5.5.2 Thread Helix Angles, for Single-/Multi-Start Threads 5.5.3 Threading Insert Inclination . . . 5.5.4 Thread Profile Generation . . . . . 5.5.5 Threading Turning – Cutting Data and Other Important Factors . . . . . . . . . . . . . 5.6 Thread Milling .. . . . . . . . . . . . . . . . . . . . . . . 5.7 Thread Rolling – Introduction . . . . . . . . . 5.7.1 Thread Rolling Techniques . . . .

149 150 151

6.5

155 158 160 163 166 169

181 182 182 189

7

189 191 193 195 195 198

Modular Tooling and Tool Management .. . . . . . . . . . . . . . . Modular Quick-Change Tooling . . . . . . . Tooling Requirements for Turning Centres .. . . . . . . . . . . . . . . . . . Machining and Turning Centre Modular Quick-Change Tooling . . . . . . . . . . . . . . . . Balanced Modular Tooling – for High Rotational Speeds . . . . . . . . . . . . Tool Management .. . . . . . . . . . . . . . . . . . . . 6.5.1 The Tool Management Infrastructure .. . . . . . . . . . . . . . . . 6.5.2 Creating a Tool Management and Document Database .. . . . . . 6.5.3 Overall Benefits of a Tool Management System .. . . . . . . . . . 6.5.4 Tool Presetting Equipment and Techniques for Measuring Tools . . . . . . . . . . . . . . 6.5.5 Tool Store and its Presetting Facility – a Typical System . . . . . 6.5.6 Computerised-Tool Management – a Practical Case for ‘Stand-alone’ Machine Tools

Machinability and Surface Integrity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.1 Machinability .. . . . . . . . . . . . . . . . . . . . . . . . 7.1.1 Design of Machinability Tests and Experimental Testing Programmes .. . . . . . . . . . 7.2 Machined Roundness .. . . . . . . . . . . . . . . . 7.2.1 Turned Roundness – Harmonics and Geometrics .. . . 7.3 Chatter in Machining Operations . . . . . .

200 203 206 209

211 212 216 221 230 233 238 240 244 245 261 264

269 270 270 285 291 294

Contents

8.8.1

7.3.1

7.4 7.5

7.6 7.7

7.8 7.9 7.10

Chatter and Chip Formation – Significant Factors Influencing its Generation .. . . . . . . . . . . . . . . . 7.3.2 Chatter – Important Factors Affecting its Generation . . . . . . . 7.3.3 Stability Lobe Diagrams . . . . . . . Milled Roundness – Interpolated Diameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . Machined Surface Texture . . . . . . . . . . . . 7.5.1 Parameters for Machined Surface Evaluation .. . . . . . . . . . . . 7.5.2 Machined Surface Topography 7.5.3 Manufacturing Process Envelopes .. . . . . . . . . . . . . . . . . . . . 7.5.4 Ternary Manufacturing Envelopes (TME’s) . . . . . . . . . . . . Machining Temperatures . . . . . . . . . . . . . . 7.6.1 Finite Element Method (FEM) . . . . . . . . . . . . . . . . . . . . . . . Tool Wear and Life .. . . . . . . . . . . . . . . . . . . 7.7.1 Tool Wear .. . . . . . . . . . . . . . . . . . . . 7.7.2 Tool Life .. . . . . . . . . . . . . . . . . . . . . 7.7.3 Return on the Investment (ROI) Cutting Force Dynamometry . . . . . . . . . . Machining Modelling and Simulation Surface Integrity of Machined Components – Introduction . . . . . . . . . . . 7.10.1 Residual Stresses in Machined Surfaces .. . . . . . . . .

XI

297 297 300 301 305 308 317 324 326 326

328 330 331 337 342 343 350

360 360

8 Cutting Fluids .. . . . . . . . . . . . . . . . . . . . . . . 8.1 Historical Development of Cutting Fluids .. . . . . . . . . . . . . . . . . . . . . 8.2 Primary Functions of a Cutting Fluid .. . 8.3 High-Pressure Jet-Assisted Coolant Delivery .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.4 Types of Cutting Fluid .. . . . . . . . . . . . . . . . 8.4.1 Mineral Oil, Synthetic, or Semi-Synthetic Lubricant? .. . 8.4.2 Aqueous-Based Cutting Fluids 8.4.3 Water Quality .. . . . . . . . . . . . . . . . 8.5 Cutting Fluid Classification – According to Composition .. . . . . . . . . . . . . . . . . . . . . . 8.6 Computer-Aided Product Development 8.6.1 Cutting Fluid – Quality Control 8.7 Selecting the Correct Cutting Fluid .. . . . 8.7.1 Factors Affecting Choice .. . . . . . 8.7.2 Selection Procedure . . . . . . . . . . . 8.8 Care, Handling, Control and Usage – of Cutting Fluids .. . . . . . . . . . . . . . . . . . . . .

381 382 383 383 387 392 395 397 398 398 404 407 407 408 409

8.9 8.10 8.11

8.12

Product Mixing – Preparation of a Aqueous-Based Cutting Fluids . . . . . . . . . . . . . . . . . . . . . . . . 8.8.2 Monitoring, Maintenance and Testing of Cutting Fluid – in Use . . . . . . . . . . . . . . . . . . . . . . . . Multi-Functional Fluids . . . . . . . . . . . . . . . Disposal of Cutting Fluids .. . . . . . . . . . . . Health and Safety Factors – Concerning Cutting Fluid Operation and Usage . . . . 8.11.1 Cutting Fluid-Based Health Issues .. . . . . . . . . . . . . . . . . Fluid Machining Strategies: Dry; Near-Dry; or Wet . . . . . . . . . . . . . . . . . . . . . 8.12.1 Wet- and Dry-Machining – the Issues and Concerns . . . . . . . 8.12.2 Near-Dry Machining . . . . . . . . . .

410 411 417 417 418 420 425 425 426

9 Machining and Monitoring Strategies 9.1 High Speed Machining (HSM) .. . . . . . . . 9.1.1 HSM Machine Tool Design Considerations .. . . . . . . . . . . . . . . 9.2 HSM Dynamics – Acceleration and Deceleration .. . . . . . . . . . . . . . . . . . . . . 9.2.1 HSM Dynamics – Servo-Lag .. . 9.2.2 Effect of Servo-lag and Gain on Corner Milling . . . 9.2.3 Effect of Servo-Lag and Gain Whilst Generating Circular Paths . . . . . . . . . . . . . . . . . . . . . . . . . 9.2.4 CNC Processing Speed . . . . . . . . 9.3 HSM – with Non-Orthogonal Machine Tools and Robots . . . . . . . . . . . . . . . . . . . . . 9.4 HSM – Toolholders/Chucks . . . . . . . . . . . 9.4.1 Toolshank Design and Gripping Pressures . . . . . . . . 9.4.2 Toolholder Design and Spindle Taper .. . . . . . . . . . . . 9.5 Dynamic Balance of Toolholding Assemblies . . . . . . . . . . . . . . . . . . . . . . . . . . . 9.5.1 HSM – Problem of Tool Balance 9.5.2 HSM – Dynamic Balancing Machine Application . . . . . . . . . . 9.6 HSM – Research Applications .. . . . . . . . . 9.6.1 Ultra-High Speed: Face-Milling Design and Development . . . . . . 9.6.2 Ultra-High Speed: Turning Operations . . . . . . . . . . . 9.6.3 Ultra-High Speed: Trepanning Operations .. . . . . . . . . . . . . . . . . . .

431 432 434 445 446 448 448 449 451 458 458 465 467 469 472 474 474 480 484

XII

Contents

9.6.4

9.7 9.8

9.9 9.10

Artefact Stereometry: for Dynamic Machine Tool Comparative Assessments .. . . . . HSM: Rotating Dynamometry . . . . . . . . . Complex Machining: of Sculptured Surfaces .. . . . . . . . . . . . . . . . 9.8.1 Utilising the Correct Tool for Profiling: Roughing and Finishing . . . . . . . . . . . . . . . . . 9.8.2 Die-Cavity Machining – Retained Stock . . . . . . . . . . . . . . . . 9.8.3 Sculptured Surface Machining – with NURBS . . . . . . . . . . . . . . . . . . 9.8.4 Sculptured Surface Machining – Cutter Simulation . . . . . . . . . . . . . Hard-Part Machining .. . . . . . . . . . . . . . . . 9.9.1 Hard-Part Turning . . . . . . . . . . . . 9.9.2 Hard-Part Milling . . . . . . . . . . . . . Ultra-Precision Machining . . . . . . . . . . . . 9.10.1 Micro-Tooling . . . . . . . . . . . . . . . . 9.10.2 Micro-Machine Tools .. . . . . . . . .

486 493 496 496 498 502

505 507 508 511 516 518 525

9.10.3 Nano-Machining and Machine Tools . . . . . . . . . . . . 9.11 Machine Tool Monitoring Techniques 9.11.1 Cutting Tool Condition Monitoring .. . . . . . . . . . . . . . . . . . 9.11.2 Adaptive Control and Machine Tool Optimisation .. . . . . . . . . . . . 9.11.3 Artificial Intelligence: AI and Neural Network Integration .. . . . . . . . . . . . . . . . . . . 9.11.4 Tool Monitoring Techniques – a ‘Case-Study’ . . . . . . . . . . . . . . . .

526 531 531 535 538 538

Appendix .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 549 About the Author .. . . . . . . . . . . . . . . . . . . . . . . . . 587 Subject Index .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 589

1

Cutting Tool Materials ‘What is the use of a book’ , thought Alice, ‘without pictures or conversations?’

LEWIS CARROLL

(1832–1898) [Alice in Wonderland, Chap. 1]

Chapter 1

1.1 Cutting Technology – an Introduction Previously, many of the unenlightened manufacturing companies, having purchased an expensive and sophisticated new machine tool, considered cutting tool technology as very much an afterthought and supplied little financial support, or technical expertise to purchase these tools. Today, tooling-related technologies are treated extremely seriously, as it is here that optimum production output, consistency of machined product and value-added activities are realised. Often companies feel that to increase productivity – to offset the high capital investment in the plant and to amortise such costs (i.e. pay-back), is the most advantageous way forward. This strategy can create ‘bottlenecks’ and disrupt the harmonious flow of production at later stages within the manufacturing environment. Another approach might be to maximise the number of components per hour, or alternatively, drive down costs at the expense of shorter tool life, which would increase the non-productive idle time for the production set-up. Here, the prime tooling factor should not be for just a marginal increase in productivity and efficiency, nor the perfection of any particular operation. If ‘bottlenecks’ in component production occur, they can readily be established by piles of machined parts sitting on the shop floor awaiting further valuedadded activities to be undertaken. These ‘line-balance’ production problems need to be addressed by achieving improved productivity across the whole operation, perhaps by the introduction of a Taguchi-type component flow analysis system within the manufacturing facility. The well-known phrase that: ‘No machine is an island’ (i.e. for part production) and that manufacturing should be thought of as ‘One big harmonious machine’ and not a lot of independent problems, will create a means by which increases in productivity can be achieved. The cutting tool problems, such as: too wide a range of tooling inventory, inappropriate tools/out-dated tooling, or not enough tools for the overall operational

Tooling refers not only to non-consumable items such as: cutting tools and inserts, tool holders, tool presetters, screws, washers and spacers, screwdrivers/Allen keys, tool handling equipment, but also consumable items, such as hand wipes, grease/oils employed in tool kitting and cutting fluids, etc.

requirements for a specific manufacturing environment, can be initially addressed by employing the following tooling-related philosophy – having recently undertaken a survey of the current status of tooling within the whole company: • Rationalisation • Consolidation • Optimisation NB These three essential tool-related factors in establishing the optimum tooling requirements for the current production needs, will be briefly reviewed.

1.1.1 Rationalisation In order to be able to rationalise the tools within the current production facility, it is essential to conduct a thorough appraisal of all the tools and associated equipment with the company. This tooling exercise will be both time-consuming and costly, because it necessitates a considerable manpower resource and needs a means of identifying all the tools and inserts currently utilised, in some logical and tabulated manner. Such surveys are often best conducted by utilising a primitive but efficient tool-card indexing system in the first instance. Details, such as: tool type and its tooling manufacturer, quantity of tools in use and the current levels of stock in the tool store, their current location(s), feeds and speeds utilised, together with any other relevant tool-related details are indexed on such cards. Once these tooling facts have been established, then they can be loaded into either a computerized tool management system database, or recorded onto an uncomplicated tooling database for later interrogation. Having established the current status of the tooling within the manufacturing facility, this allows for a tooling rationalisation campaign to be developed. Tool rationalisation (Fig. 1) consists of looking at the results of the previous tooling survey and significantly reducing the number of tooling suppliers for particular types of tools and inserts. This initial rationalisation policy has the twin benefits of minimising tooling suppliers with their distinct varieties of tools, while enabling bulk purchase of such tools from the remaining suppliers, at preferential financial rates of purchase. Moreover, by using less tooling companies whilst purchasing bulk stock, this has the bonus of making you one of their prime customers with their undivided at-

Cutting Tool Materials

. Figure 1. Rationalisation of cutting inserts, can have a dramatic effect on reducing the tooling and workholding inventory. [Courtesy of Sandvik Coromant]

Chapter 1

tention, should the need for later ‘tool problem-solving’ of manufacturing clichés in production occur.

1.1.2 Consolidation For any tooling that remains after the rationalisation exercise, these should be consolidated, by reducing the number of insert grades, by at least half – which often proves to have little effect on production capability. By grouping inserts by their respective sizes, shapes and say, nose radius for example, this will eliminate many of the less-utilised inserts, enabling the potential for bulk purchase from the tooling supplier, with an attendant reduction in tool costs. From this consolidation activity, it may now be possible to purchase higher-performance grade cutting inserts, that meet a wider application range, enabling the consolidation to be even more effective. Furthermore, such improved inserts, will probably have a longer tool life and can be utilised at higher speeds, which probably negates their extra cost, over the previously used inserts. If fewer grades of insert are stocked, the tooling/application engineers will be acquainted with them much more thoroughly and this will result in a added effectiveness and a consistent application, for the production of machined components – more will be said on this latter point in the next section on Optimisation.

1.1.3 Optimisation By consolidating the tooling, it allows productivity to be boosted by optimisation of the cutting insert grades. For example, in turning operations, the depth of cut can probably maximised and, as a result, the number of passes along, or across the part can be minimised. It can be argued that increasing the depth of cut leads to a reduction in the subsequent tool life (in terms of minutes of cutting per edge). However, there are fewer cuts per part, so each machined workpiece requires less overall cutting and as a result, many more parts per edge can be produced. More important, are that the cycle times for roughing operations be reduced: a reduction in the number of roughing passes from three to one, results in a 66% reduction in the cycle time. This increase in productivity may justify any potential decrease in tool life, on the basis that it could reduce, or eliminate a potential ‘bottleneck’ in latter production processes of the part’s manufacture. To extract the maximum productivity from today’s

high-performance grades, they must be worked hard and pushed to their fullest capabilities. When tool life is reduced by increasing the depth of cut, there are several ways that a such loss can be minimised. For example, it is known that the size of the insert’s nose radius has a pronounced effect on tool life, so by doubling the depth of cut this can, in the main, allow for a larger nose radius – assuming that the component feature allows access. If an increase in nose radius cannot be utilised, then increasing the insert’s size will help to offset any higher wear rates, by providing a better heat dissipation for the action of cutting. The accepted turning practice when roughingout, is that no more than half the insert’s cutting edge length should be utilised, because as the depth of cut approaches this value, a larger insert is recommended. Where large depths of cut are used in combination with high feedrates, a roughing grade insert geometry promotes longer tool life, than a general-purpose insert. Often, by using a single-sided insert rather than a double-sided one for roughing cuts, this has the twin benefits of increased productivity and longer tool life (in terms of machined parts per edge). Normally, single-sided inserts are recommended whenever the depth of cut and feedrate are so high that the surface speed must be reduced below the grade’s normal range, in order to maintain an adequate tool life. Such inserts should be considered if erratic insert breakage occurs. Later to be discussed in the chapter on Machinability and Surface Integrity, is the fact that the highest temperature region on the tool’s rake face is not at the cutting edge; but in the vicinity on the chip/tool interface where chip curling occurs – this is some distance back and where the crater is formed. The position for this highest isothermal region can vary, depending upon the feedrate. For example, if the feedrate is increased, the highest temperature zone on the insert’s face will move away from the cutting edge; conversely, if the feedrate is now reduced, this region moves toward the cutting edge. This phemomena means that if the feedrate is too low for the chosen insert geometry and edge preparation, heat will be concentrated too near the cutting edge and insert wear will be accelerated. Thus, by increasing the feedrate, it has the affect of moving the maximum heat zone away from the insert’s edge and is so doing, extends tool life – in terms of minutes of actual cut-time per edge. As a result, each machined part will be produced in less time and at higher feeds, so the tool life in terms of parts per edge will also increase.


As a result of the inappropriate use of cutting data, such as incorrect feedrates employed for the chosen insert geometry, this can produce a number of undesirable symptoms. These symptomatic problems include: extremely shortened tool life, edge chipping and insert breakage are likely if feedrates are too high, whereas when feeds are too low, chip control becomes a problem. Once the insert grades have been consolidated with their associated geometries, it is relatively easy to determine the feedrates for a selected grade of workpiece materials. Tooling suppliers can recommend a potential insert grade for particular component part material, with an initial selection of insert grade, such surface speeds being indicated in the Appendix. These inserts can be optimised by ‘juggling’ the grades and geometries marginally around the specified values, this may allow feedrates to be increased and should provide a significant pay-off in terms of improved productivity, at little, or no additional capital expenditure. If the cutting speed is increased rather than the feed, a point is reached where any increase in surface speed will result in a decrease in productivity. In other words, cutting too fast will mean spending more time changing tools than making parts! Equally, by cutting too slowly, the tool will last much longer, but this is at the expense of the number of machined parts produced per shift. If these statements are correct, what is the ‘right’ surface speed? This question will now be discussed more fully. If we return to the theme previously mentioned, namely: ‘No machine is an island’ and treat the production shop as: ‘One big machine’ , it can be stated that every shop should determine its own particular manufacturing objectives – when considering both cutting speeds and tool life. Typical objectives for tool life might be the completion of a certain number of parts before indexing the insert, or adopting a ‘sister tool’ , or alternatively, insert indexing after one/part of a shift. If very expensive components are being machined, the main goal is to avoid catastrophic insert

‘Sister tooling’ is the term that refers to a duplicate tool (i.e. having the same tool offsets) held in the turret/magazine and can be automatically indexed to this tool, to minimise down-time when changing tools. Such a ‘sister tool’ , can be pre-programmed into the CNC controller of the machine tool, to either change after a certain number of parts has been produced, or if the tool life has been calculated, then when the feed function on the CNC has decremented down to this preset value, then the ‘sister tool’ is selected.

failure, which on a finishing cut, would probably result in scrapping the part. When exceedingly large parts are to be machined, the objective may simply be to complete just one part per insert, or in an even more extreme situation, just one pass over the part. When small parts are being produced, then the tool life can be controlled in order to minimise dimensional size variation with in-cut time. This strategy of tool life control, reduces the need for frequent adjustment of tool offset compensations in the CNC controller. However, one idea shared by all of these strategic production approaches, is that by optimising the surface speed, the manufacturing objectives will be realised. As a consequence of this approach to production, there is no correct surface speed for any specific combination of material and insert grade, the optimum surface speed depends upon the company’s manufacturing requirements at this time. When long production runs occur, these are ideal because it allows cutting data experimentation to discover the optimum speed for a particular production cycle. Sometimes it is not possible to find the speed to exactly meet the production demands and, a change of insert grades, to one of the higher-technology materials may be in order. If short production runs are necessary, this can often rule out any experimentation with insert grades, but by consultation with a ‘cutting tool expert’ , or reference to the published cutting literature the answer may be found to the problem of insert optimisation. However a cautionary note, care must be taken when utilising published recommendations, as they should only be employed as guidelines, to help initiate the job into production. Comparison with a known starting point within the recommended range for specific production conditions, namely for: large depths of cut, high feedrates, very long continuous cuts, significant interrupted cuts, workpiece surface scale and the absence of coolant, would all suggest that reductions in surface speed should be initially considered. Conversely, production conditions that result in: short lengths of cut, shallow depths of cut, low feedrates, smooth uninterrupted cuts, clean pre-turned, or bright-drawn wrought workpiece materials and flood coolant, having a very rigid setup, suggests that the recommended ranges for the insert could be exceeded, while still maintaining an acceptable tool life. It should be remembered that the main requirement is for an overall increase in production output and not perfection. After the analysis, when the tooling inventory has been consolidated, there will be fewer and

Chapter 1

more versatile insert grades and geometries that need to be considered. This smaller insert inventory allows a detailed appreciation of how to optimise the speeds and feeds in combination with depths of cut more effectively, for the desired production objectives. By optimisation here of the machining parameters, this allows full utilisation of the capital equipment, with the result that large improvements in overall manufacturing efficiency can be expected. It is evident from this discussion concerning optimisation, that the parameters of: tool life, feedrate and cutting speed form a complex relationship, which is illustrated in Fig. 2a. Consequently, if you change one parameter, it will affect the others, so a compromise has to be reached to obtain the optimum performance from a cutting tool. Preferably, the ideal cutting tool should have superior performance if five distinct areas (see Fig. 2b): • Hot hardness – is necessary in order to maintain sharp and consistent cutting edge at the elevated temperatures that are present when machining. NB If the hot hardness of the tooling is not sufficient for the temperature generated at the tool’s tip, then it will degrade quickly and be useless.

• Resistance to thermal shock – this is necessary in

order to overcome the effects of the continuous cycle of heating and cooling that is typical in a milling operation, or when an intermittent cutting operation occurs on a lathe (e.g. an eccentric turning operation).

NB If this thermal shock resistance is too low, then rapid wear rates can be expected, typified in the past, by ‘comb cracks’ on High-speed steel (HSS) milling cutters.

• Lack of affinity – this condition should be present between the tool and the workpiece, since any degree of affinity will lead to the formation of a builtup edge (BUE) – see the chapter on Machinability and Surface Integrity.

. Figure 2. The main factors affecting cutting tool life, under ‘steady-state’ cutting conditions

NB This BUE will modify the tool geometry, leading to poorer chip-breaking ability, with higher forces generated, leading to degraded workpiece surface finish. Ideally, the cutting edge should be inert to any reaction with the workpiece.


• Resistance to oxidation – a cutting edge should mately 5 m·min–1. These early cutting tools frequently have the desirable condition of having a high resistance to oxidation.

NB This oxidation resistance of the cutting tool is necessary, in order to reduce the debilitating wear that oxidation can produce when machining at elevated temperatures.

• Toughness – allows the cutting edge of the insert to absorb the cutting forces and shock loads produced whilst machining, particularly relevant when intermittent cutting operations occur.

NB If an insert is not sufficiently tough, then when unwanted vibrations are induced, this can result in either premature failure, or worse, a shattered cutting edge. Cutting tool manufacturers, by careful balancing of these five factors for the ideal cutting tool, can produce grades of inserts which distinctly vary, allowing a wide range of workpiece materials to be machined through the selection of the correct insert grade for a particular material. In recent years, tooling manufacturers have produced wider ranges of workpiece-cutting ability from fewer types of inserts, across a diverse range of speeds and feeds, allowing tooling inventories to be reduced even further. This brief introduction showing how and in what manner correct tooling can be used to increase production output, needs to be considered against the current situation of advances in cutting tool materials and their selection – this will be the theme of the next section.

1.2 The Evolution of Cutting Tool Materials 1.2.1 Plain Carbon Steels Prior to 1870, all turning tooling materials were produced from plain carbon steels, with a typical composition of 1% carbon and 0.2% manganese – the remainder being iron. Such a tool steel composition meant that it had a low ‘hot-hardness’ (i.e, its ability to retain a cutting edge at elevated temperatures), as such, the cutting edge broke down at temperatures approaching 250°C, this in reality kept cutting speeds to approxi-

had quench cracks present which severely weakened the cutting edge, as a result of water hardening at quenching rates greater than 1000°C/second (i.e. necessary to exceed the critical cooling velocity – to fully harden the steel), upon manufacture. By 1870, Mushet (working in England), had introduced a more complex steel composition, containing: 2% carbon, 1.6% manganese, 5.5% Tungsten and 0.4% chromium, with the remainder being iron. The advantage of this newly developed steel was that it could be air-hardened, this was a significantly less drastic quench than using a water quenchant. Mushet’s steel had greater ‘hothardness’ and could be utilised at cutting speeds up to 8 m·min–1. This turning tool material composition, was retained until around 1900, but with the level of chromium gradually superseding that of manganese.

1.2.2 High-Speed Steels Around the turn of the century in the United States, fundamental metallurgical work was being undertaken by F.W. Taylor and his associate M. White and by 1901, these researchers had greatly improved the overall tool steel and slightly modifying its composition with a material that was to be known as High-speed steel (HSS), enabling cutting speeds to approach 19 m·min–1. High-speed steel was not a new material, but basically an innovative heat treatment procedure. The typical metallurgical composition of HSS was: 1.9% carbon, 0.3% manganese, 8% tungsten, and 3.8% chromium, with iron the remainder. Taylor and White’s tool steel mainly differed from that of Mushet’s by an increased amount of tungsten and a further replacement of manganese by chromium. By 1904, the content of carbon had been reduced, allowing for more ease in forging this HSS. Further rapid development of the HSS occurred over the next ten years, with tungsten content increased to improve its ‘hot-hardness’. Around this time, Dr J.A. Matthews found that vanadium additions had improved the material’s abrasion resistance. By 1910, the content of tungsten had increased to 18%, with 4% chromium and 1% vanadium, hence the wellknown 18:4:1 HSS had arrived, its metallurgical composition continued with only marginal modifications over the next 40 years. Of the modifications to HSS during this time, of note was that in 1923 the so-called ‘super’ HSS was developed, although this variant did not become commercially viable until 1939, when Gill reduced the tungsten content to enable the tool steel

Chapter 1

to be successfully hot-worked. Around 1950 in the United States, M2 HSS was introduced, having some of the tungsten content replaced by that of molybdenum. This gave the approximate M2 HSS metallurgical composition as: 0.8% C, 4% Cr, 2% V, 6% W and 5% Mo – Fe being the remainder. In this form, the M2 HSS could withstand machining temperatures of up to 650°C (ie the cutter glowing dull red) and still maintain a cutting edge. In 1970, Powder Metallurgy (P/M) by metallurgical processing via hot isostatic pressing (HIP), was introduced for the production of HSS, with careful control of elemental particle size; afterward the sintered product is forged then hot-rolled. This HSS (HIP) processing gave a uniformly distributed elemental matrix, overcoming the potential segregation and resulting non-homogenous structure that would normally occur when ingot-style HSS forging. Such P/M processing techniques enable the steel-making company to ‘tailor’ and specify the exact metallurgical composition of alloying elements, this would allow the newly-developed sintered/forged HSS tooling to approach that of the performance of cemented carbides, in terms of inherent wear resistance, hardness and toughness. In Fig. 3, a comparison of just some of the tooling materials is highlighted, here, fracture toughness is plotted against hardness to indicate the range of influence of each tool material and the comparative relative merits of one material against another, with some of their physical and mechanical properties tabulated in Fig. 3b. A typical sintered micro-grained HSS of today, might contain: 13% W, 10% Co, 6% V, 4.75% Cr and 2.15% C – Fe the remainder. One reason for the ‘keen’ cutting edge that can be retained by sintered micrograined HSS, is that during P/M processing the rapid atomisation of the particles produces extremely fine carbides of between 1 to 3 µm in diameter – which fully support the cutting edge, whereas HSS produced from an ingot, has carbides up to 40 µm in diameter. By way of illustration of the benefits of the latest micro-grained HSS – in the uncoated condition – when compared to its metallurgical competitor of cemented carbide, HSS has a bend, or universal tensile strength of between 2,500 to 6,000 MPa – this being dependent on metallurgical composition, whereas cemented carbide tooling has a bend strength of between 1,250 to 2,250 MPa. These metallurgical tool processing techniques have significantly improved sintered micrograined HSS enabling for example, high-performance drilling, reaming and tapping to be realised. Coating by either single-, or multiple-coating has been shown to significantly enhance any tooling mate-

rial, but this is a complex subject and more will be said on this subject shortly.

1.2.3 Cemented Carbide Possibly the widest utilised cutting tool materials today are the cemented carbide family of tooling, of which the group derived from tungsten carbide is most readily employed. Prior to discussing the physical metallurgy and expected mechanical/physical characteristics of cemented carbides, it is worth looking into the complex task of insert selection. In Fig. 4, just a small range of the material types, grades, shapes of inserts and coatings by a leading cutting tool company is depicted. Highlighting the complex chip-breaker geometries, necessary to both develop and break chips and evacuate them efficiently from the workpiece’s surface region. To give a simplified impression of just some of the tooling insert variations and permutations available from a typical tooling manufacturer, if 10 insert grades are listed, in 6 different shapes, with 12 differing chip breakers and five nose radii in the tooling catalogue, this equates to 10 × 6 × 12 × 5, or 3,600 inserts. In reality, there are a number of other important features that could extend this cutting insert permutation to well over five significant figures – for potential insert selection. When the permutated insert number reaches this level of complexity, selecting the optimum combination of insert characteristics becomes more a matter of luck than skill. Tungsten (synonym Wolfram, hence the chemical symbol W), is the heaviest metal in the group VIB in Mendeleev’s Period Chart (i.e. atomic number 74). It was named after the German word wolfram – from the mineral wolframite – as it was derived from the term wolf rahm, because the ore was said to interfere with tin smelting – supposedly devouring the tin. Whereas the term tungsten, was coined from the Swedish tung sten, meaning heavy stone. Hence, in 1923, the German inventor K. Schröter produced the first metal matrix composite, known today as cemented carbides. In these first cemented carbides, Schröter combined tungsten monocarbide (WC) particles embedding them in a cobalt matrix – these particles acted as a very strong binder. Cemented carbide is a hard transition metal carbide ranging from 60% to 95% bonded to cobalt, this being a more ductile metal. The carbides vary, ranging from having hexagonal structures, to a solid solution of titanium, tantalum and niobium carbides to that of a NaCl structure. Tungsten carbide does not dissolve


. Figure 3. Cutting tool materials – toughness versus hardness – and their typical material characteristics. [Courtesy of Mitsubishi Carbide]

any transition metals, but it can melt those carbides found in solid solution. Powder metallurgy processing route – liquid-phase sintering – is utilised to produce cemented carbides, as melting only occurs at very high temperatures and there is a means of reducing tung-

sten powder using hydrogen from chemically purified ore. Ore reduction can be achieved by the manipulation of the processing conditions, enabling the grain size to be controlled/modified as necessary. The uniform grain sizes of tungsten carbide today can range

10

Chapter 1

from 0.2 to 7 µm – enabling a final sintered product to be carefully controlled. Moreover, by additions of fine cobalt at a further processing stage, then wet milling the constituents, allows for precise and uniform control of the grain size – producing a fine powder. Prior to sintering, the milled powder can be spray-dried giving a free-flowing spherical powder aggregate, with the addition of lubricant to aid in its consolidation (i.e. pressing into a ‘green compact’). Sintering normally occurs at temperatures of 1500°C in a vacuum, which reduces the porosity from about 50% that is in the ‘green state’ , to less than 0.01% porosity by volume

in the final cutting insert condition. The low level of porosity in the final product is the result of ‘wetting’ by the liquid present upon sintering. The extent of this ‘wetting’ during liquid-phase sintering, being dependent upon molten binder metal dissolving to produce a pore-free cutting insert, this has excellent cohesion between the binder and the hard particles (see Fig. 5, for typical cemented carbide powders and resulting microstructures). It should be stated that most of the ‘iron-group’ of metals can be ‘wetted’ by tungsten carbide, forming sintered cemented carbide with excellent mechanical integrity.

. Figure 4. Cutting inserts indicating the diverse range of: shapes, sizes and geometries available, with compositions varying from: cemented carbide, ceramics, cermets, to cubic boron nitride derivatives. [Courtesy of Sandvik Coromant]


. Figure 5. Cemented carbide powders and typical microstructures after sintering. [Courtesy of Sandvik Coromant]

11

12

Chapter 1

The desirable properties that enable tungsten carbide to be tough and readily sintered, also cause it to easily dissolve in the iron, producing the so-called ‘straight’ cemented carbide grades. These ‘straight’ grades normally contain just cobalt and have been used to predominantly machine cast iron, as the chips easily fracture and do not usually remain in contact with the insert, reducing the likelihood of dissolution wear. Conversely, machining steel components, requires alternative carbides such as tantalum, or titanium carbides, as these are less soluble in the heated steel at the cutting interface. Even these ‘mixed’ cemented carbide grades will produce a tendency to dissolution of the tool material in the chip, which can limit high speed machining operations. Today, the dissolution tool material can be overcome, by using cutting insert grades based on either titanium carbide, or nitride, together with a cobalt alloy binder. Such grades can be utilised for milling and turning operations at moderate cutting speeds, although their reduced toughness, can upon the application of high feed rates, induce greater plastic deformation of the cutting edge and induce higher tool stresses. These uncoated cutting inserts were very much the product of the past and today, virtually all such tooling inserts are multi-coated to significantly reduce the effects of dissolution wear and greatly extend the cutting edge’s life – more will be said on such coating technology later.

1.2.4 Classification of Cemented Carbide Tool Grades Most cemented carbide insert selection guides group insert grades by the materials they are designed to cut. The international standard for over 30 years used for carbide cutting of workpiece materials is: ISO 5131975E Classification of Carbides According to Use – which has a colour-coding for ease of identification of sub-groups. In its original form, this ISO 513 code utilises 3 broad letter-and-colour classifications (see Fig. 6 for the tabulated groupings of carbides and their various colours, designations and applications):

The workpiece categories are arranged according to their relative chip production characteristics and certain metallurgical characteristics, such as casting condition, hardness and tensile strength. ISO 1832–1991 has clesignations: ‘P’ (Steels, low-alloy); ‘M’ (Stainless steels); ‘K’ (Cast irons); ‘N’ (Aluminium alloys); ‘H’ (Hardened steelas)

• P (blue) – highly alloyed workpiece grades for cutting long-chipping steels and malleable irons,

• M (yellow) – lesser alloyed grades for cutting fer-

rous metals with long, or short chips, cast irons and non-ferrous metals, • K (red) – is ‘conventional’ tungsten carbide grades for short-chipping grey cast irons, non-ferrous metals and non-metallic materials. Under this previous ISO system (Fig. 6), both steels and cast irons can be found in more than one category, based upon their chip-formation characteristics. Each grade within the classification is given a number to designate its relative position in a continuum, ranging from maximum hardness to maximum toughness. This original ISO 513 Standard, has been modified over the years by many tooling manufacturers, introducing more discretion in their selection and usage. Typical of this manufacturer’s modified approach, is that found by just one American tooling company, forming a simple colour-coding matrix, such as the three designated manufacturer’s chip-breaker grades (such as: F, M and R) and three workpiece material grades (i.e. Steel, Stainless steel and Cast iron) – producing a ninecell grid. While another manufacturer in Europe, has produced a more discerning matrix, based upon adding the ‘machining difficulty’ into the matrix, producing a 3 × 3 × 3 matrix – producing a twenty seven cell grid. In this instance, the tooling manufacturer uses the workpiece material to determine the tool material needed. The insert geometry is still selected according to the type of machining operation to be undertaken, while the insert grade is determined by the application conditions – whether such factors as interrupted cuts occur, forging scale on the part are present and the desired machining speed being designated as: good, average, or difficult. NB These manufacturer’s matrices for the tooling insert selection process will get a user to approximately 90% of optimum, with the ‘fine-tuning’ (optimisation) requiring both technical appreciation of information from the manufacturer’s tooling catalogue/recommendations from ‘trouble- shooting guides’ and any previous ‘know-how’ from past experiences – as necessary.


. Figure 6. Classification of carbides according to use. [Courtesy of Seco Tools]

13

14

Chapter 1

1.2.5 Tool Coatings: Chemical Vapour Deposition (CVD) Rather quaintly, the idea of introducing a very thin coating onto a cemented carbide cutting tool originated with the Swiss Watch Research Institute, using the chemical vapour deposition (CVD) technique. In the 1960’s, these first hard coatings were applied to cemented carbide tooling and were titanium carbide (TiC) by the CVD process (Fig. 7 shows a schematic view of the CVD process) at temperatures in the range 950 to 1050°C. Essentially, the coating technique consists of a commercial CVD reactor (Fig. 8a) with cutting tools, or inserts to be hard-coated placed on trays (depicted in Fig. 8b). Prior to coating the tooling situated on their respective trays, these tools should have a good surface finish and sharp corners should have small honed edges – normally approximately 0.1 mm. With the CVD technique, if these honed tool cutting edges are too large, they will not adequately support the coating, but if they are even greater, the cutting edge will be dulled and as a result will not cut efficiently. These tooling trays (Fig 8b) are accurately positioned one above another, being pre-coated with graphite and are then loaded onto a central gas distribution column (i.e tree). The ‘tree’ now loaded with tooling to be coated is placed inside a retort of the reactor (Fig. 8a). This contained tooling within the reactor, is heated in an inert atmosphere until the coating temperature is reached and the coating cycle is initiated by the introduction of titanium tetrachloride (TiCl4) together with methane (CH4) into the reactor. The TiCl4 is a cloud of volatile vapour and is transported into the reactor via a hydrogen carrier gas (H2), whereas CH4 is introduced directly. This volatile cloud reacts on the hot tooling surfaces and the chemical reaction in say, forming a TiC as a surface coating, is:

tooling, then the previously used methane is substituted by a nitrogen/hydrogen gas mixture. For example, if a simple multi-coated charge is required for the tooling, it is completed in the same cycle, by firstly depositing TiC using methane and then depositing TiN utilising a nitrogen/hydrogen gas mixture. As the TiN and TiC are deposited onto the tooling, they nucleate and grow on the carbides present in the exposed surface regions, with the whole CVD coating process taking approximately 14 hours, consisting of 3 hours for heating up, 4 hours for coating and 7 hours for cooling. The thickness of the CVD coating is a function of the reaction concentration, this being the subject of: various gaseous constituents and their respective flow rates, coating temperature and the soaking time at this temperature. The CVD process is undertaken in a vacuum together with a protective atmosphere, in order to minimise oxidation of the deposited coatings. However it should be noted that, in the case of high-speed steel (HSS) tooling such as when coating small drills and taps, the elevated coating temperatures employed, necessitate post-coating hardening heat treatment.

1.2.6 Diamond-Like CVD Coatings

TiCl4 + CH4 → + TiC + 4HCl

Crystalline diamond is only grown by the CVD process on solid carbide tools, because of the high temperatures involved in the process, typical diamond coating temperatures are in the region of 810°C. Such diamondlike tool coatings (Fig. 9), make them extremely useful when machining a range of non-ferrous/non-metallic workpiece materials such as: aluminium-silicon alloys, metal-matrix composites (MMC’s), carbon composites and fibreglass reinforced plastics. Although such workpiece materials are lightweight, they have hard, abrasive particles present to give added mechanical strength, the disadvantage of such non-metallic/metallic inclusions in the workpiece’s substrate are that

The HCl gas is a bi-product of the process and is discharged from the reactor onto a ‘scrubber’ , where it is neutralised. When titanium is to be coated onto the

Graphite shelves are most commonly employed, as it is quite inexpensive compared to either stainless steel, or nickel-based shelving, with an added benefit of good compressive strength at high temperature.

Some limitations in the CVD process are that residual tensile stresses of coatings can concentrate around sharp edges, possibly causing coatings to crack in this vicinity – if edges are not sufficiently honed – prior to coating. Additionally, the elevated temperatures cause carbon atoms to migrate (diffuse) from the substrate material and bond with the titanium. Hence, this substrate carbon deficiency – called ‘eta-phase’ is very brittle and may cause tool failure, particularly in interrupted-cut operations.


. Figure 7. A PVD-coating, with coated tooling, plus a schematic representation of the CVD and PVD coating processes. [Courtesy of Sandvik Coromant]

15

16

Chapter 1

. Figure 8. Modern insert/tooling coating plant. [Courtesy of Walter Cutters]


they become extremely difficult to machine with ‘conventional tooling’ and are a primary cause of heat generation and premature face/edge wear. Here, the high tool wear is attributable to both the abrasiveness of the hard particles present and chemical wear promoted by corrosive acids created from the extreme friction and heat generated during machining. Such diamond-coated tooling is expensive to purchase, but these coatings can greatly extend the tool life by up to 20 times, over uncoated tooling, when machining non-metallic and certain plastics, this more than compensates for the additional cost premium. Such diamond-like coated tools, combine the (almost) high hardness of natural diamond, with the strength and relative fracture toughness of carbide. The extreme hardness of diamond-like coatings enable the effective machining of non-ferrous/nonmetallic materials and, by way of an example of their respective hardness when compared to that of a PVD titanium aluminium nitride coated tool, they are three times as hard (see Fig. 3a). Although, these diamondlike coatings do not have the hardness properties of crystalline diamond, they are approximately half their micro-hardness value. Diamond-like coatings can range from 3 to 30 µm in thickness (see Fig. 9 – bottom), with the individual crystal morphology present measures between 1 to 5 µm in size (Fig. 9 – top). Recently, a diamond-coating crystal structure called ‘nanocrystalline’ has been produced by a specialised CVD process. The morphology has diamond crystals measuring between 0.01 to 0.2 µm (i.e. 10 to 200 nanometres), with a much finer grain structure and smoother surface to that of ‘conventional’ diamondlike coatings. This smoother ‘nanocystalline’ surface morphology presents less opportunity for workpiece material built-up edge (BUE) at the tool/chip interface, significantly improving both the chip-flow across the rake face of the tool and simultaneously giving a better surface finish to the machined component.

1.2.7 Tool Coatings: Physical Vapour Deposition (PVD) In 1985 the main short-comings resulting from the CVD process were overcome by the introduction of the physical vapour deposition process (Fig. 7), when the first single-layer TiN coatings were applied to cemented carbide. There are several differences between PVD and CVD coating processes and their resulting coatings. Firstly, the PVD process occurs at low-to-

17

medium temperatures (250 to 750°C), as a result of lower PVD temperatures found than by the CVD process, no eta-phase forms. Secondly, the PVD technique is a line-of-sight process, by which atoms travel from their metallic source to the substrate on a straight path. By contrast, in the CVD process, this creates an omni-directional coating process, giving a uniform thickness, but with the PVD technique the fact that a coating may be thicker on one side of a cutting insert than another, does not affect its cutting performance. Thirdly, the unwanted tensile stresses potentially present at sharp corners in the CVD coated tooling, are compressive in nature by the PVD technique. Compressive stresses retard the formation and propagation of cracks in the coating at these corner regions, allowing tooling geometry to have the pre-honing operation eliminated. Fourthly, the PVD process is a clean and pollution-free technique, unlike CVD coating methods, where waste products such as hydrochloric acid must be disposed of safely afterward. In general, there have been many differing PVD coating techniques that have been utilised in the past to coat tooling, briefly some of these are: • Reactive sputtering – being the oldest PVD coating method, it utilises a high voltage which is positioned between the tooling to be coated (anode) and say, a titanium target (cathode). This target is bombarded with an inert gas – generally argon – which frees the titanium ions, allowing them to react with the nitrogen, forming a coating of TiN on the tools. The positively-charged anode (i.e. tools) will attract the TiN to the tool’s surface – hence the coating will grow, • Reactive ion plating – relies upon say, titanium ionisation using an electron beam to meet the target, which forms a molten pool of titanium. This titanium pool then vaporises and reacts with the nitrogen and an electrical potential accelerates toward the tooling to subsequently coat it to the desired thickness. • Arc evaporation – utilises a controlled arc which vaporises say, the titanium source directly onto the inserts – from solid. As with the CVD process, all of the PVD coating production methods are undertaken in a vacuum. Further, the PVD coatings tend to have smoother and less

18

Chapter 1

. Figure 9. A vast array of differing cutting inserts, together with diamond coated cemented carbide. [Courtesy of Sandvik Coromant]


dimpled surface appearance, than are found by the ‘blocky-grained’ surface by the CVD technique. A typical tooling tungsten carbide substrate that has been PVD multi-coated is depicted in Fig. 10a. Such multiple coating technology allows for a very exotic surface metallurgy to be created, which can truly enhance tool cutting performance. In general and in the past, CVD coatings tended to be much thicker than their PVD alternatives, having a minimum coating thickness of between 6 to 9 µm, whereas PVD coatings tended to be in the range: 2000°C and applied pressures > 60 GPa, to enable the conversion to take place.

26

Chapter 1

order to speed-up the rate of sintering, additions of a solvent/catalyst are utilised (i.e. normally metals, or metal nitrides), but during sintering the whole mass must be held in the ‘cubic region’ of the respective pressure/phase diagram – to prevent these hard crystals reverting back to their original soft hexagonal form. By sintering these hard particles together, it is possible to form a conglomerate of CBN/diamond, in which randomly orientated crystals are combined to produce a large isotropic14 mass. A very wide range of polycrystalline products can be produced, utilising either CBN, or synthetic diamond as a base. For example, by changing the: grain size (see Figs. 12 c and d), solvent/ catalyst employed, degree of sintering and particle size distribution and, the presence/absence of inert fillers, this will have a profound effect on the mechanical and physical properties of the final product – Table 1 lists

14 Isotropic materials can be considered to have the same properties in different directions.

the physical properties of various comparable cutting tool materials. In order to produce the required tool geometries, both the polycrystalline layer of CBN and polycrystalline diamond (PCD) are bonded to a thick tungsten carbide backing layer, then cutting inserts are wirecut out of this large blank – obtaining the maximum number of insert shapes per blank (see Fig. 13a). These CBN/PCD inserts are either full-size, or smaller tips that are then brazed onto suitably-shaped blanks, to fit the desired tool holder (as illustrated in Fig. 13b). Both CBN and PCD cutting tools can successfully machine: super-alloys (ie with low iron content), grey cast iron and non-ferrous metals, but show distinct differences when other workpiece materials are to be productively machined – as depicted in Fig. 14. Polycrystalline diamond cutting tools are not utilised for machining ferrous workpieces, this is because when machining under the high temperatures and sustained pressures that occur during cutting, the diamond has a tendency to revert back to graphite, after only a few seconds in-cut. This reversion, does

. Figure 13. Cutting tool materials: Cubic Boron Nitride (CBN) and Polycrystalline Diamond (PCD) . [Courtesy of DeBeers – element 6]


. Figure 14. A diagram illustrating how Cubin Boron Nitride (CBN) and Polycrystalline Diamond (PCD) applications are grouped, by workpiece materials. Their effectiveness when ei-

27

ther machining highly abrasive components, or high temperatures in the cutting vicinity

28

Chapter 1

. Figure 15. Turning operations with Cubic Boron Nitride (CBN) and Polycrystalline Diamond (PCD). [Courtesy of DeBeers – element 6]


29

not take place when machining many non- ferrous and non-metallic workpiece materials. Although CBN is synthesised in a similar fashion to that of PCD cutting tool products, it is not as hard as PCD and is therefore less reactive with ferrous metals, as long as the cutting temperature is less than 1,000°C, it will not revert to its softer hexagonal form and oxidise in air. This means that CBN can machine many ferrous parts and cast iron grades. The complementary nature of both CBN and PCD is clearly depicted in Fig. 14, where the ‘cross-over’ between these ultra-hard cutting tool materials is shown. In both CBN and PCD machining applications, an excellent machined surface finish can be obtained (see Figs. 15a and b). In the case of many PCD operations, the cutting tool must not only machine widely differing materials that are situated adjacent to one another in many passes over such a diverse material workpiece, but produce an excellent machined surface finish, which really ‘challenges’ the tool. Tool life can be extended greatly by utilising either CBN, or PCD tooling, often tool lives can be increased by 50 to 200 times that of the previous cemented carbide alternatives. This boost in output, makes their additional purchase price irrelevant, when considered against the massive productive gains that are to be made by their adoption. Today, both CBN and PCD can often be found as either thin-coated layers on tooling (see Fig. 3 for their relative tool insert hardnesses/toughnesses), or as a ‘sandwich’ between metallic backing layers. These ‘sandwiched’ tool edges, permit brazing on both sides of the hardened product, which are then accurately positioned and held onto a tungsten carbide shank, making them an ideal alternative for many microdrilling operations. Such compound drilling edge technology, gives considerably improved edge retention and resistance to any abrasive particles present in the workpiece and its severely work-hardened swarf, typically found with the latest metal matrix composites (MMC’s). Such ultra-hard tooling, can be readily used on high-silicon aluminium alloys used in the automotive industries, while not discounting the wide range of workpiece composites employed by the aerospace industries and the resin-based components utilised in the furniture industry.

mond is used correctly in a very rigid machine-toolworkpiece setup for materials that require the best possible surface finish, then there is simply no alternative. By way of illustration of this fact, if production turning high-silicon content aluminium pistons with polycrystalline diamond (PCD) tooling, the best surface finish that can be obtained will be in the region of 0.4 µm, conversely using an SCD tool this will give a surface finish of better than 0.15 µm. If one really wants the ultimate surface finish currently obtainable by machining – in the ‘nano-range’ , then a monolithic diamond tool, mounted in a special-purpose diamond turning lathe is the only manner in achieving such superb ‘mirror-finish’ surfaces. SCD tool edges are produced as either razor sharp edges, or are made with a perfect radius being chip-free, imparting machined ‘mirror-finishes’ of just a few angströms (i.e. 10–10 m). The optical industries in particular find that the latest blemish-free ultra-sharp cutting edges of SCD, means that diamond (paste) polishing after machining has been virtually, if not completely eliminated, this fact in particular being a very big production cost for the final manufacture of large monolithic astronomical mirrors. A cautionary note, is that to use SCD tooling for anything other than as a finishing cut is totally uneconomic, as these precision components to be machined, should have been roughly configured to the desired shape, prior to diamond machining. Therefore, SCD tools should be employed for exceedingly light finish cuts of no deeper than 0.0008 m. Natural diamond is a truly remarkable material, that exhibit’s a diverse range of mechanical and physical properties. For example diamond has the highest known: bulk hardness, thermal conductivity, while having a very low coefficient of friction and will not corrode, these properties make it an ideal tool material for the highest precision and accuracy machined components. Of these properties, hardness is probably the most important characteristic in machining operations and, when measured by the Knoop indentor15. By way of comparison of ultra-hard cutting tool materials, the following two examples may prove informative:

1.2.11 Natural Diamond

15 Knoop indentors produce a wedge-shaped indentation in the form of a parallelogram, with one diagonal seven times longer than the adjacent one. The Knoop test method is generally considered the optimum technique, for crystalline solids – having crystallographic directionality (i.e. anisotropy).

Monolithic, or single-crystal diamond (SCD), is the hardest material available today. If such natural dia-

30

Chapter 1

• Natural diamond – has a hardness of 9,000 kg mm–2

(ie diamond orientation and test conditions): Diamond (111) surface, direction, 500g load, • Cubic boron nitride (CBN) – has a hardness of 4,500 kg mm–2, (111) surface, direction, 500g load. One of the main limitations of natural diamond is that it has distinct cleavage planes (111)16. This consistent cleavage plane makes it ideal for jewellery-makers to cleave the beautiful facets demanded of diamond jewellery, but this means that monolithic diamonds must be mounted in their respective tool holders in exactly the correct orientation/plane, so avoiding any potential cleavage in-cut. SCD tool cost is a draw-back, because these tools cost in the region of four times more than the equivalent PCD tool. However, despite this very high cost difference, SCD can reduce the overall operating costs and significantly improve productivity, when applied to the correct machining process. Expensive tooling such as SCD, must be handled with care, because although it is the hardest material known, it is also very brittle and subject to thermal shock, the problem being exacerbated with its very sharp tool edges. Therefore, it is essential that sudden impacts to the tool’s edge must be avoided, through either inappropriate cutting applications, or by rough handling.

References Journal and Conference Papers Boller, R. Crystal Clear – DLCoatings. Cutting Tool Engg., 36–40, May 2002. Craig, P. Behind the Carbide Curtain. Cutting Tool Engg., 26–41, Aug., 1997. Dzierwa, R. Slippery when Blue – Coatings. Cutting Tool Engg., 36–41, Jan., 2003. Eastman, M. Inserts Show their True Colors. Cutting Tool Engg., 30–36, April 1999. Feir, M. Post-treatment of PM Parts. Metal Powder Report, 28–30, Jan., 1981.

16 Miller indices determine the crystalline orientation for a plane in an atomic structure and for natural diamond it is normally on the (111) plane, although some cleavage has been observed on the (110) plane.

Fretty, P. Grade Wise. Cutting Tool Engg., 46–50, Feb., 2000. Gough, P. Tool Life Boosted by Titanium Nitride Coat. Machinery and Prod. Engg., 52–53, Feb. 1983. Gummeson, P.U. and Stosuy, A. Iron-carbon Behaviour during Sintering. In: Source Book on Powder Metallurgy, ASM Pub., 49–61, 1979. Hanson, K. Lowering your Grades. Cutting Tool Engg., 54–60, Jan., 2000. Heath, P.J. Ultra-hard Materials. European J. of Engg. Ed., Vol. 12 (1), 5–20, 1987. Israelsson, J. A Progress Report on Cutting Tool Materials. American Machinist, 39–40, Dec., 1992. Jindal, P.C. et al., PVD Coatings for Turning, Cutting Tool Engg., 42–52, Feb., 1999. Kennedy, B. Making the Grade – PCBN Applications. Cutting Tool Engg., 22–30, June 2002. Lewis, B. Fast Times in HSS. Cutting Tool Engg., 28–32, July 2001. Lewis, B. Confidence Game – Grades and Geometries. Cutting Tool Engg., 46–52, Dec., 2002. Mielert, W. Coating for Speed. Cutting Tool Engg., 40–44, Feb., 1996. Mirchandani, P.K. Making a Better Grade – Composite Carbide Substrates. Cutting Tool Engg., 58–61, Jan., 2005. Mitoraj, L. The Coating Edge. Cutting Tool Engg., 51–55, Feb., 2000. Novak, D. Single Minded – Single Crystal Diamond. Cutting Tool Engg., 38–41, June 2002. Raymond, M.K. Ceramics Ease Up the Machining of Highhardness Parts, American Machinist, May 1996. Raymond, M.K. Coatings Keep Cutting Tools Sharp. American Machinist, 40–42, May 1996. Richter, A. Raising Al – AlTiN Coatings. Cutting Tool Engg., 42–46, Jan., 2003. Richter, A. Top Coat. Cutting Tool Engg., 36–41, Dec., 2003. Sanders, E.H. Understanding Coated Carbides. Cutting Tool Engg. 3–7, Sept./Oct., 1977. Sprout, W. PVD Today. Cutting Tool Engg., 52–4057, Feb., 1994. Taylor, F.W. On the Art of Cutting Metals. Trans. of ASME 28, 31–350, 1907. Thalmann, R. Cracking the Code – Carbide Classifications. Cutting Tool Engg., 34–43, June 1995. Vasilash, G.S. The Superfard Coatings: More than Meets the Eye. 52–54, Production, Dec., 1995. Weiner, M. Coatings Move Forward. Cutting Tool Engg., 22–29, Feb., 1999. Woods, S. Coat, Please. Cutting Tool Engg., 50–56, Oct., 2004.


Books, Booklets and Guides Balshin, M.Y. and Kiparisov, S.S. General Principles of Powder Metallurgy. MIR (Moscow) Pub., 1980. Chin, G.Y. Advances in Powder Metallurgy. ASM Pub., 1982. Dieter, G.E. Mechanical Metallurgy. McGraw-Hill Kogakusha 2nd Ed., 1976. Dowson, G. Powder Metallurgy – The Processes and its Products. Adam Hilger (Bristol) Pub., 1990. Kalpakjian, S. Manufacturing Processes for Engineering Materials (3rd Ed.). Addison Wesley, 1997.

31

Lenel, F.V. Powder Metallurgy, Principles and Applications. Metal Powder Inds. Princetown (NJ), April 1980. Metals Handbook – Powder Metallurgy. ASM Pub. 8th Ed., Vol. 3, 1967. Modern Metal Cutting – A Practical Handbook. AB Sandvik Coromant Pub., 1994. Reed-Hill, R.E. Physical Metallurgy Principles. Van Nostrand Reinhold (NY) Pub., 1973. Trent, E.M. Metal Cutting. Oxford: Butterworth Heinemann (3rd Ed.), 1991.

2

Turning and Chip-breaking Technology ‘Machines are the produce of the mind of Man; and their existence distinguishes the civilized man from the savage.’

WILLIAM COBBETT

(1762–1835) [Letter to the Luddites of Nottingham]

34

Chapter 2

2.1 Cutting Tool Technology In the following sections a review of a range of Turning-related technologies and the importance of chipbreaking technology will be discussed.

2.1.1 Turning – Basic Operations Turning can be broken-down into a number of basic cutting operations and in effect, there are basically four such operations, these are: 1. Longitudinal turning (Fig. 16a), 2. Facing (Fig. 16b), 3. Taper turning – not shown, 4. Profiling – not shown. NB These turning operations will now be very briefly reviewed. In its most simple form, turning generates cylindrical forms using a single-point tool (Fig. 1.16a). Here, a tool is fed along the Z-axis slideway of the lathe (CNC), or a turning centre, while the headstock rotates the workpiece (i.e. the part is held in either: a chuck, on a mandrel, face-plate, or between centres – when overhang is too long), machining the component and thereby generating a circular and cylindrical form of consistent diameter to the turned part. Facing is another basic machining operation that is undertaken (Fig. 16b) and in this case, the tool is fed across the X-axis slideway while the part rotates, again, generating a flat face to the part, or a sharp corner at a shoulder, alternatively it can be cutting the partial, or finished part to length (i.e facing-off). Taper turning can be utilised to produce short, or long tapers having either a fast taper (i.e. with a large included angle), or slow taper (i.e. having a small included angle – often a ‘self-holding taper’ , such as a Morse taper). There are many different operations that can be achieved on a CNC lathe/turning centre,

The range of turning operations is vast, feedrates can be varied, as can rotational speeds.

Facing operations can also be used to produce either curved convex, or concave surface features to the machined part – here the surface is both generated and formed, requiring simultaneous programmed feeding motions to the Z- and Xaxes.

including: forming, while others such as drilling, boring, screw-cutting, of internal features, and forming and screw-cutting of external features, to name just a few of the traditional operations undertaken. With the advent of mill/turn centres, by having CNC control of the headstock and rotational, or ‘driven-/live-tooling’ to the machine’s turret, this allows prismatic features to be produced (i.e. flats, slots, splines, keyways, etc.), as well as drilled and tapped holes across and at angles to the major axis of the workpiece, or off-axis. Even this explanation of mill/turn centres is far from complete, with regard to today’s sophisticated machine tools. As machine tool builders today, can offer a vast array of machine configurations, including: co-axial spindles (ie twin synchronised inline headstocks), fitted with twin turrets with X- and Y-axes simultaneous, but separate control, having programmable steadies (i.e. for supporting long slender workpieces), plus part-catchers , or overhead gantries for either component load/unload capacity, to multiaxes robots feeding the machine tool. This type of machine tool exists and has multi-axes CNC controllers to enable the machine’s down-time to be drastically reduced and in this manner achieving high productive output virtually continuously.

2.1.2 Turning – Rake and Clearance Angles on Single-point Tools In order for a turning tool to effectively cut and produce satisfactory chips, it must have both a rake and clearance angle to the tool point (Fig. 17). Today’s single-point cutting tools and inserts are based upon decades of: past experience, research and development, looking into all aspects of the tool’s micro-geometry at the cutting edge. Other important aspects are an efficient chip-breaking technology, in certain instances critical control of the flexure (i.e. elastic behaviour) of the actual tool insert/toolholder combination for the latest multi-functional tooling is essential – more will be said on some of these topics later in the chapter. The rake angle is the inclination of the top face of the cutting edge and can vary according to the work-

Forming can be achieved in a number of ways, ranging from complex free-form features (externally/internally) on the machined part, to simply plunging a form tool to the required depth.

Turning and Chip-breaking Technology

35

. Figure 16. Typical turning operations with the workpiece orientation shown in relation to the cutting insert, for either: (a) cylindrical turning, (b) facing. [Source: Boothroyd 1975]

36

Chapter 2

piece material being machined. In general, for ductile materials, the rake inclination is a positive angle, as the shearing characteristics of these materials tends to be low, so a weaker wedge angle (i.e. the angle between the top face and the clearance angle) will suffice. For less ductile, or brittle workpiece materials, the top rake inclination will tend toward neutral geometry, whereas for high-strength materials the inclination will be negative (see Fig. 17), thereby increasing the wedge angle and creating a stronger cutting edge. This stronger cutting edge has the disadvantage of requiring greater power consumption and needing a robust tool-workpiece set-up. Machining high-strength materials requires considerable power to separate the chip from the workpiece, with a direct relationship existing between the power required for the cutting operation and the cutting forces involved. Cutting forces can be calculated theoretically, or measured with a dynamometer – more will be said on this subject later in the text. Both side and front clearances are provided to the cutting edge, to ensure that it does not rub on the workpiece surface (see Fig. 17). If the tool’s clearance is too large it will weaken the wedge angle of the tool, whereas if too small, it will tend to rub on the machined surface. Most tools, or inserts have a nose radius incorporated between the major and minor cutting edges to create strength here, while reducing the height of machined cusps, with some inserts having a ‘wiper’ designed-in to improve the machined surface finish still further – more will be mentioned on these insert integrated features later.

2.1.3 Cutting Insert Edge Preparations Often, a minute edge preparation (see Figs. 17 and 18b, c and d) is created onto the sharp cutting edge of the insert, this imparts additional strength to the outermost corners of the cutting edge, where the rake and clearance faces coincide. There are four basic manners in which the honed edge preparation is fashioned, these are:

Machined cusps result from a combination of the feedrate and the nosed radius of the tool. If a large feedrate occurs with a small nose radius then the resultant cusp height will be high and well-defined, conversely, if a small feedrate is utilised in conjunction with a large nose radius, then cusp height is minimised, hence the surface texture is improved.

1. Chamfer – which simply breaks the corner – not illustrated, 2. Land – stretching back negatively from the clearance side to various lengths on the rake face (see Fig. 18b), 3. Radius – around the actual corner (see Fig. 18c), 4. Parabolic – has unequal levels of honing on two faces (see Fig. 18d). Even here, more often than not, certain combinations of these four edge preparations are utilised, so that the cutting forces are redirected onto the body of the rake’s face, rather than directed down against the more fragile cross-section of the edge. The T-lands and hones are often actually incorporated into the insert geometry of the contoured surface. Typical T-lands range in size from 0.07 to 0.50 mm, having angles varying from 5 to 25° off of the rake face (Fig. 18b). Honing which is the ‘rounding’ of the cutting edge, can be performed in one of several ways. Probably the oldest technique for honing, utilises mechanical means, which employs a vibrating tub filled with an abrasive media, such as aluminium oxide – to ‘break’ the corner on these inserts. A variation in this design, uses an identical abrasive, except here the inserts are held by centrifugal force to the inside of a rotating tank. While yet another method of honing using an abrasive media, involves spraying the inserts with fine abrasive particles – to hone the edges of the inserts. Probably the most popular method for obtaining cutting insert honed edges, uses brushes made from extruded nylon impregnated with diamond (see Fig. 18a). The inserts to be honed pass by these brushes in individual carriers and rotate as they all revolve under the brushes, thereby applying equal hones to all insert edges. Depending upon the amount of desired honing, these brushes can be either raised, or lowered, or alternatively, the inserts can make multiple passes through the machine. All of the above honing techniques produce a hone that is roughly equal on both the flank and rake faces – what is termed a ‘round hone’ (Fig. 18c). Yet another honing profile termed the parabolic hone (i.e. sometimes this honed edge is known as:

The radius is sometimes termed ‘edge rounding’ (i.e. denoted by the letters ‘ER’) – often applied to most edge preparations, enabling the cutting forces to be directed on to the stronger part of the insert.


37

. Figure 17. Typical turning ‘finishing’ insert/toolholder geometry and the insert’s edge chamfering, in relation to the workpiece

38

Chapter 2

P-hone, oval, or waterfall), is produced by a machine with a soft, diamond-charged rotating rubber wheel. Therefore, as the abrasive material rubs across the inserts, it tends to extend slightly over the inserts sides, producing a hone of uneven proportions between the two insert faces (Fig. 18d). As in the case of the T-land

cutting insert edge preparation, the P-hone directs the cutting forces into the body of the insert. Honing can be specified in a number of sizes, usually being determined by the amount of time these insert spend in the honing device. The original Standard for honing was established in the United States by

. Figure 18. A honing machine (i.e. brush-style) and several types of honing edge preparations. [Courtesy of Ingersoll]


the American National Standards Institute (ANSI) in 1981, which included dimensions and expected tolerances for these three basic hones. Today, many cutting tool manufacturers have expanded upon this Standard, or adopted their own – specifying hone manufacturing and identification methods. Hones must be applied prior to the application of coatings. Inserts that are destined to receive a CVD coating, must have a minimum hone to strengthen the edge, in order to counteract the effects of this high temperature coating process. Conversely, PVD coatings, can be equally applied either over fully-honed insert edges, or on an unhoned cutting edge. In recent years, the cutting tool manufacturers have an emphasis toward providing honed edges of greater consistency and repeatability.

2.1.4 Tool Forces – Orthogonal and Oblique The cutting forces are largely the result of chip separation, its removal and chip-breaking actions, with the immense pressure and friction in this process producing forces acting in various directions. Stresses at the rake face tend to be mainly compressive in nature, although some shear stress will be present (see Table 2, by way of illustration of the machining shear stresses for various materials), this is due to the fact that the rake is rarely ‘normal’ to the main cutting direction. This compressive stress tends to be at its greatest closest to the cutting edge, with the area of contact between the chip and rake face being directly related to the geometry here, hence the need for tooling manufacturers to optimise the geometry in this region. There are two distinct types of forces present in machining operations concerning single-point cutting tools/inserts (see Fig. 19), these are: 1. Orthogonal cutting forces – two forces (ie tangential and axial – see Fig. 19b), 2. Oblique cutting forces – three forces (i.e. tangential, axial and radial – see Fig. 19a).

As well as the tool/chip interface temperatures being up to 1,000°C, the interface pressures can reach a maximum of 3,000 MPa, these being sterile smooth surfaces makes them ‘ideal’ conditions for the occurrence of ‘pressure-welding’/seizure.

39

. Table 2. Typical in-cut shear strengths of various materials Material:

Shear yield strength in cutting –2 (N mm )

Iron

370

0.13% C. steel

480

Ni-Cr-V steel

690

Austenitic stainless steel

630

Nickel

420

Copper (annealed)

250

Copper (cold-worked)

270

Cartridge brass (70/30)

370

Aluminium (99.9% pure) Magnesium Lead

97 125 36

[Source: Trent ( 1984)]

NB Both of these cutting force models are heavily influenced by the: cutting tool/insert orientation to workpiece, tool’s direction of cut and its applied feedrate.

Oblique Cutting Forces Fig.1.19a, can be seen a model of the three-dimensional cutting force components in an oblique turning operation, when the principal cutting edge is at an angle to the main workpiece axis (i.e. Z-axis). These component forces can be separated into the: • Tangential force (FT) – which is greatly influenced by the contact and friction between both the workpiece and tool, as well as the contact conditions between the chip and the rake face of the cutting edge. The magnitude of the tangential cutting force is the greatest of these three component forces and contributes to the torque, which in turn, influences

Feedrates play a major role in determining the axial force in single-point cutting operations, in association with the tool’s orientation to the part being machined.

40

Chapter 2

. Figure 19. The two- and three-force models of orthogonal and oblique cutting actions, with the component forces approximately scaled to give an indication of their respective magnitudes


•

•

the power requirement for cutting. Fundamentally, the product of the tangential force and the cutting speed represent the power required for machining. The specific cutting force is a unit expression for the tangential cutting force, being closely related to the material’s undeformed chip thickness and selected feedrate, Axial force (FA) – the magnitude of this force will vary depending on the selected feedrate and the chosen tool geometry and in particular, the ‘plan approach angle’ , or ‘entering angle’ , – more will be said on this topic later. Its direction is from the feeding of the tool, along the direction of workpiece machining, Radial force (FR) – is directed at right angles to the tangential force from the cutting point. The ‘plan approach angle’ and the size of the nose radius, will influence this force. NB These three component forces are significantly influenced by the rake angle, with positive rakes producing in general, lower cutting forces. The resultant force, its magnitude and angle, will be affected by all three component forces, in conjunction with the tool’s geometry and the workpiece material to be cut.

Orthogonal Cutting Forces In Fig. 19b the two-dimensional model for orthogonal cutting is depicted, once again, for comparison to the oblique cutting model, in a single-point turning operation. For simplicity, if one assumes that the point of the tool is infinitely sharp and that the tool is at right angles to the workpiece axis having no deflection present, then the two component forces are the tangential force and axial force (i.e. previously mentioned above). In this case, this tool geometry-workpiece configuration, allows long slender bars to be turned, as there is less likelihood of tool ‘push-off ’ (i.e. as the radial force

In reality, the specific cutting force is a better indication of the power requirement, as it is the force needed to actually deform the material prior to any chip formation. It will vary and is influenced by the: undeformed chip thickness, feedrate, and yield strength of the workpiece material. For example, if the cutting conditions are kept the same and only the material changed, then if a nickel-based alloy is machined, the initial chip forming force (i.e. specific cutting force) will be more than ten times greater than when cutting a pure aluminium workpiece.

41

has been neutralised – as indicated by the fact that the resultant force shows no X-axis offset). If any radial force was present, this would create either a ‘candlestick effect’ , or ‘barrelling’ to the overall turned length. In reality, there will always be some form of nose radius, or chamfer to the tool point, which will have some degree of ‘push-off ’ , depending upon the size of this incorporated nose feature – creating a ‘certain degree’ of radial component force affect.

2.1.5 Plan Approach Angles The manner in which the cutting edge contacts the workpiece is termed the ‘plan approach angle’ (Fig. 20a), being composed of the entering and lead angles for the selected tool geometry. In effect for singlepoint turning operations, the tool’s orientation of its plan approach, is the angle between the cutting edge and feeding direction. When selecting a tool geometry for turning specific workpiece feature – such as a 90° shoulder – it is important as it will not only affect the machined part geometry, but has an influence on consequent chip formation and the direction and magnitude of the component cutting forces, together with the length of engagement of the cutting edge (see Fig. 20b). In single-point turning (Fig. 20b), the depth of cut (DOC), or ‘cutting depth’ is the difference between an un-cut and cut surface, this being half the difference in the un-cut and cut diameter (i.e. the diameter is reduced by twice the DOC in one pass along the workpiece). This DOC is always measured at 90° to the tool’s feed direction, not the cutting edge. The manner in which the cutting edge approaches the workpiece is termed the ‘entering angle’ (i.e. plan approach angle), this being the angle between the cutting edge and feed direction (Fig. 20a – shown here in a cylindrical turning operation). Moreover, the plan approach angle not only influences the workpiece features that can be produced with this cutting geometry, it also affects the formation of chips and the magnitude of the component forces (Fig. 20b). The ‘entering angle’ affects the length of the cutting edge engaged in-cut, normally varying from 45° to 90°, as illustrated in the four cases of differing plan approach angles shown in Fig. 20b. Here, in ‘case I’ an

In single-point turning operations, the depth of cut (DOC) is sometimes referred to by the term: ‘undeformed chip thickness’.

42

Chapter 2

. Figure 20. Insert approach angle geometry for turning operations


entering angle of 45° and lead angle of 45° is utilised, giving rise to equal axial and radial component forces. In ‘case II’ , the entering angle has changed to 75° and lead angle is now 15°, these altered angles change the component forces, with an increase in the axial force while reducing the radial force. In ‘case III’ , an orthogonal cutting action occurs, with only a 90° entering angle (i.e. the lead angle reduces to zero), showing a large increase in the axial force component at the expense of the radial force component which is now zero10. In ‘case IV’ , an oblique cutting action has returned (i.e. as in ‘cases I and II’), but here the entering angle has changed to -15°, with the lead angle 75°, this produces a large axial component force, but the radial component force direction has now reversed. This last tool plan approach angle geometry (i.e. ‘case IV’), is similar to the geometry of a light turning and facing tool, allowing cylindrical and facing operations to be usefully undertaken – but the tool’s point is somewhat weaker that the others, with the tool points becoming of increased strength from right to left. Therefore, in ‘case I’ , for a given feedrate and constant DOC, the cut length/area is greater than the other ‘cases’ shown and with this geometry, it enables the tool to be employed for heavy roughing cuts. Returning to ‘case III’ , if this tool is utilised for finish turning brittle-based workpiece materials, then upon approaching the exit from a cut, if the diameter is not supported by a larger shoulder diameter, then the axial component force /pressure, will be likely to cause edge break-out (i.e. sometimes termed ‘edge frittering’), below the machined surface diameter at this corner (i.e. potentially scrapping the machined part). In mitigation for this orthogonal cutting tool geometry, if longer slender workpieces require cylindrical turning along their length, then with the radial force component equating to zero, it does not create significant ‘push-off ’ and allows the part to be successfully machined11. A single-point turning geometry is subject to very complex interactions and, as one geometric feature is modified such as changing the entering angle, or in-

10 In all of these cases, it is assumed – for simplicity – that there is no nose radius/chamfer on the tool and it is infinitely sharp. 11 In order to minimise the effects of the radial force component when cylindrically turning long slender workpieces with ‘Case I and II’ tool geometries, the use of a programmable steady, or a ‘balanced turning operation’ (i.e. utilising twin separately programmable turrets on a turning centre, with tools situated virtually opposite each other running parallel during the turning operation – see Fig. 41), will reduce this ‘push-off ’.

43

creasing the tool’s nose radius, this will influence other factors, which in turn could have a great impact on the: type of machined surface finish produced, expected tool life and the overall power consumption during the operation. In fact, the main factors that influence the application of tooling for a specific turning operation are: Workpiece material – machinability, condition I. (i.e. internal/external), mechanical and physical properties, etc., II. Workpiece design – shape, dimensions and machining allowance, III. Limitations – accuracy and precision requirements, surface texture/integrity, etc., IV. Machine tool – type, power, its condition and specifications, V. Stability – loop stiffness/rigidity (i.e. from the cutting edge to its foundations), VI. Set-up – tool accessibility, workpiece clamping and toolholding, tool changing, VII. Tool programme – the correct/specified tool and its tool offsets, etc., VIII. Performance – cutting data, anticipated tool-life and economics, IX. Quality – tool delivery system and service. In order to gain an insight into the complex and important decisions that have to be made when selecting tooling for the optimum production of either part batch sizes, or for continuous production runs, then the following section has been incorporated.

2.1.6 Cutting Toolholder/Insert Selection When deciding upon the correct selection of a toolholder/cutting insert for a given application, a range of diverse factors must be considered, as indicated in Fig. 21. As can be seen by the diagram (Fig. 21) and associated text and captions, there are many other variables that need to be considered prior to selection of the optimum toolholder/insert. Generally, the fixed conditions cannot be modified, but by ‘juggling’ with the variable conditions it is possible to accomplish the best compromise toolholder/insert geometry, to optimise these cutting conditions for the manufacture of a specific workpiece and its intended production requirements. Whenever toolholders and cutting inserts are required for a specific manufacturing process, it is important to view the tooling selection procedure as a logical progression, in order to optimise the best

44

Chapter 2

. Figure 21. The factors that must be considered prior to commencing a turning operation, when utilising indexable inserts


possible tools/inserts for the job in hand. Perhaps the following selection strategy for a ‘start point’ in choice and application of turning tools, can be undertaken according by the following step-by-step approach: Start Point →

Edge clamping system, ↓ Toolholder size and type, ↓ Insert shape, ↓ Insert size, ↓ Nose radius, ↓ Insert type, ↓ Tool material, ↓ Cutting data →

45

part access, as a toolholder is defined by its: effective entering and point angles12, together with the insert’s shape (see Fig. 22). Toolholders should be the largest possible size for the turning centre’s tool turret, this requirement is vital, as it reduces the ‘tool overhang ratio’ – providing rigidity and integrity to stabilise the insert’s cutting edge. NB Appendix 1a shows the ISO ‘Code Key’ – for External Toolholders. Appendix 1b shows the ISO ‘Code Key’ – for Solid Boring Bars. Appendix 1c shows the ISO ‘Code Key’ – for Cartridges.

Insert Shape

Final Tool holder and Insert Selection

Edge Clamping System Initially, the tool holder clamping system should be selected to provide optimum performance in different applications over a wide range of workpiece geometries. The type of machining operation and to a lesser extent, the workpiece size determines tool holder selection. For example, roughing-out operations on big components will make considerably different demands, to that of finishing passes on small components. NB Pin, clamp and lever are just three of the insert clamping systems available – consultation with the tool suppliers at this point might be beneficial.

The insert shape should be selected relative to the entering angle needed for the tool’s accessibility, or versatility. Here, the largest suitable point angle should be chosen for strength and economy (see Fig. 23). In Fig. 23, is illustrated a practical example of how changing only one variable – insert geometry (shape) – can influence an insert’s turning application. The shape of an insert will determine its inherent weakness, or strength, which is of particular relevance if roughturning operations are necessary. Furthermore, insert shape will influence whether it is prone to vibration, or not and its predictable tool life. Hence, if one is concerned about vibrations of either the tool, workpiece, or both, then a weaker insert such as a light turning and facing geometry with less cutting edge length exposed in-cut, might be more suitable. Variable conditions such as the selection of insert’s geometric shape can affect other machining parameters and, this is valid for other insert factors, so a compromise will always occur in any machining application.

Toolholder Size and Type Once the clamping system has been selected, the size and type of toolholder must be determined, with its selection being influenced by: feed directions (i.e. see Fig. 22 for turning insert shapes and feed directions), size of cuts, workpiece and toolholder situated in the machine for accessibility requirements. The workpiece’s shape plays a decisive role if surface contouring is necessary, this is particularly relevant for machining

12 Effective entering angles (κ1) must be carefully selected when the operation involves profiling, or copying. The maximum profiling angle (β) is recommended for each tool type – if ‘workpiece fouling’ is to be avoided. NB κ1 = κ + β (for plunging into a surface), whereas κ1 = κ – β (for ramping-out of a surface), κ1 = κ (β = 0°) for cylindrical turning, Where: effective entering angle (κ1), entering angle (κ), maximum in-copy angle (β). Always select the smallest entering angle that the part geometry will allow.

46

Chapter 2

. Figure 22. Tool paths in finish turning operations. [Courtesy of Sandvik Coromant]


47

. Figure 23. Selecting indexable inserts for turning operations. [Courtesy of Stellram]

NB Appendix 1d shows the ISO ‘Code Key’ – for Indexable Inserts.

Insert Size An indexable insert size is directly related to the toolholder selected for the operation, with the entering angle and insert shape having previously been established. Only the matching-shaped insert can be fitted into the seat of a particular toolholder, as its shape and size are predetermined by the seating dimensions.

In roughing-out operations, the largest cutting depth for a given toolholder, will influence the insert size. For any insert, the effective cutting length has to be determined (see Fig. 20b), as the entering angle will influence the size of the insert selected. If the effective cutting edge length is less than the depth of cut (DOC), a larger insert should be chosen, or the DOC should be reduced. Sometimes in more demanding turning operations, a thicker insert – of the same geometric shape – gives extra reliability.

48

Chapter 2

Nose Radius Of particular relevance in any turning operation is the insert’s tool nose radius (rε – see Fig. 17), as it is the key factor with regard to: • inherent strength in roughing operations, • the resulting surface texture from finishing operations. Further, the size of the nose radius affects vibrational tendencies (see Fig. 23) and in certain instances, the feedrates. The nose radius is the transition between the major and minor cutting edges, which determines the strength, or weakness of the point angle (see Figs. 16a and 17), therefore it is an imperative factor to get right. In general, roughing-out should be undertaken with the largest possible nose radius, as it is the strongest tool point (see Fig. 23). Further, a larger tool nose radius permits higher feedrates, although it is important to monitor any possible vibrational tendencies. Later in the relevant section, more will be said on the influence that the insert’s tool nose radius plays in the final machined surface texture, but it is worth mentioning here that the feedrate for roughing operations should be set to approximately half the size of the nose radius utilised. The size of the nose radius has an affect on the power consumed in turning in conjunction with the material’s yield strength and chip-forming ability, particularly in rough-turning operations. The maximum material removal rate (MMR) can be obtained by a combination of high feedrate, together with a moderate cutting speed, with other limiting factors, such as depth of cut (DOC), tool’s nose radius, under consideration. Often, the machine tool’s power (P) availability can sometimes be a limiting factor when mmR is the requirement and, in such circumstances the cutting speed is usually lowered somewhat. For a given nose radius and cutting insert geometry, the power can be derived, to ensure that the machine tool will be able to cope with this pre-selected mmR, in the following manner: Machine tool’s power requirement (P): P = tangential force (FT) x cutting speed (VC) P = FT × VC P = kC × A × VC ∴ P = kC × f × aP × VC (kW) Where: f = feed/rev aP = depth of cut

(mm/rev) (mm)

Cutting speed (VC) VC = πDN/1000

(m/min)

Where: D = workpiece diameter (mm) N = workpiece rotational speed (rpm) Specific cutting force (kC): kC = FT/A

(N/mm2)

Where: A = cutting area (mm2) For example, for finishing operations, with the nose radius in combination with the feedrate (i.e. pre-selected), this will affect the surface texture and part accuracy, in the following manner: Machined surface texture (Rt): (Rt, this parameter being: maximum profile height) (µm) Rt = f 2/8 × rε x 1000 Where: f 2 = feedrate per revolution rε = nose radius

(mm/rev) (mm)

NB The surface texture parameter ‘Rt’ , can be converted into other surface texture parameters – as necessary. By utilising either: larger turning insert tool nose radius, ‘wiper insert’ (yet to be discussed), a more positive plan approach angle, or in certain circumstances, a higher cutting speed, the surface texture can be improved. In general, the coordination of the tool’s nose radius and the pre-selected feedrate in finishing operations, indicates that the feed should be kept below a certain level to achieve an acceptable machined surface texture value.

Insert Type The cutting insert type is for the most part determined by the previously selected geometry – see Appendix 1d for the selection of indexable inserts. In reality, various cutting conditions and workpiece materials make different demands on the insert’s cutting edge. For example, when machining hardened steel parts, this will be completely different from that to the machining of aluminium components.


49

Once the insert shape has been established in connection with its plan approach angle together with the nose radius dimension, this just leaves the type of geometry to be found. In this instance, the type of insert geometry refers to the ‘working area’ (i.e. nominally found by its depth of cut and feedrate – more will be said concerning this topic later, when ‘chip-breaking envelopes’ will be discussed). Additional factors can influence the type of cutting geometry choice, such as: machine tool’s condition, its power, the stability of the workpiece-tool-machine set-up, other factors that could affect geometry selection include: whether continuous, or intermittent cutting occurs, any tendency toward vibration while machining. Turning operations can be separated into a number of ‘working areas’ , being based upon the removal of workpiece material and the generation of accurate machined component dimensions, in combination with specific surface texture requirements – as shown in Table 3. When establishing an insert type, the feedrate and depth of cut should be identified with one of the ‘working ranges’ (i.e. from Table 3), as the various insert types to be chosen relate to this chart. It should be borne in mind that the most suitable ‘working area’ selected, will vary, in combination with such factors as the insert’s: size, shape and nose radius.

Tool Material The penultimate evaluation to be made concerning tooling decision-making is the choice of insert mate-

. Table 3. Typical working areas for external turning operations Type of machining operation:

Feedrate (f ):

Extreme finishing

0.05 to 0.15

0.25 to 2.0

Finishing

0.1 to 0.3

0.5 to 2.0

Light roughing

0.2 to 0.5

2.0 to 4.0

Roughing

0.4 to 1.0

4.0 to 10.0

Heavy roughing

>1.0

6.0 to 20

Extremely heavy roughing

>0.7

8 to 20

(mm)

(mm)

[Courtesy of Sandvik (UK) Ltd]

Depth of cut (DOC):

rial, or combination of materials that constitute the cutter’s tool edge. Today, manufacturers of tooling have a strategy for continuous improvement with variations in both tool matrices and coatings being considerable. Not only are cutting tool material research and development an on-going intensive activity, but their application for wider ranges of machining applications are being considerably enhanced. A brief review of just some of the current tool materials and coatings have been previously mentioned in Section 1.2, with the main range of cutting tool materials being: cemented carbides, coated cemented carbides, ceramics, cermets, cubic boron nitride, polycrystalline diamond and monolithic (i.e. natural) diamond. NB A good ‘start-point’ for most machining operations, is to consider coated carbides initially, then if these grades prove unsatisfactory, for whatever reason, select one of the other materials – perhaps after consultation with a cutting tool manufacturer, or after a machinability testing procedure.

Cutting Data Once all of the physical, metallurgical and geometrical factors for the cutting tool have been established for the machining operation, then it is necessary to set, or calculate the cutting data – often these criteria can be found from tooling manufacturers recommendations and cutting data tables. Certain variable factors such as feedrate should have already been made, allowing the cutting speed to be calculated, from the well-known expression (below): VC = πDN/1000

(m min–1)

Where: VC = cutting speed (m min–1) D = Workpiece diameter (mm)13 N = rotational speed (rpm)

13 In the case of drilling, reaming and tapping operations, it is the diameter of the cutting tool that is used in the calculation. For any other internal machining operations – such as in boring, it is the initial hole diameter that is employed in the cutting speed calculation.

50

Chapter 2

Once again, manufacturers data tables are often useful ‘starting-points’ for estimating the initial cutting parameter information. Considerable care must be taken if the material has either a high work-hardening tendency, or intrinsic bulk (i.e. workpiece material) hardness, as this can influence the numerical data selected. Moreover, the plan approach angle also has an effect on the numerical value for the parameter, for example, oblique machining allows a higher value than for orthogonal machining.

2.2 History of Machine Tool Development and Some Pioneers in Metal Cutting 2.2.1 Concise Historical Perspective of the Development of Machine Tools Toward the end of the 1700’s, any high-quality machining at the time meant tolerances of 0.1mm being considered as ‘ultra-precision’ , with this level of tolerance having steadily improved from the beginning of the Industrial Revolution. Pioneers in machine tool development such as John Wilkinson (1774), developed the first boring machine, this being capable of generating a bored hole of 1270 mm in diameter, with a error of about 1 mm. A contemporary of Wilkinson, namely Henry Maudslay (1771–1831), invented many precision machine tools, but he was particularly noted for the design and development of the first engine lathe. Slightly later, Sir Joseph Whitworth (1803–1887), developed the first modern-day Vee-form screwthread and nut (i.e. 55° included angle – ‘Whitworth thread’), thereby enabling precision feed-motion to be achieved via suitable gear trains on such machine tools. These early fundamental advances in machine design, allowed others and in particular, Joseph R. Brown (1852) to design the ‘dividing engine’. This newly-developed equipment, allowed precision engraving of the hand dials on machine tool axes, enhancing them with much better machinist’s judgment in both rotary and linear control, in combination with consistent repeatability by the skilled operative. Shortly after these developments, Eli Whitney produced the original milling machine, which was refined still further

by the Cincinnati Screw and Tap Company in 1884. This ‘Cincinnati machine’ was a direct forerunner of today’s manual controlled knee-type milling machine tools. Of particular note was the ergonomic grouping of the controls centrally for a more efficient hand control by the skilled operator. At this time the machine tool still utilised the Vee-form screw thread, with the Acme-form (ie having the ability to take-up backlash) still someway off development. Steady development and refinement of a range of machine tools continued into the the first half of the 20th century until the next major ‘milestone’ occurred. This significant development was the ‘modern’ numerically-controlled (NC) machine. Around the late 1940’s, the ‘recirculating ballscrew’14 was designed so that it could take-up backlash in both directions of rotation for machine tool axes. These early ‘ballscrews’ were fitted to a converted Cincinnati Milling Machine Company’s ‘Hydro-Tel’ die-sinking machine tool, at MIT (Massachusetts Institute of Technology). This military research-funded project having been commissioned by the United States Air force – who required complex free-form aeronautical parts to be automatically machined for the latest aircraft. This research was undertaken by MIT, in association with ‘Cincinnati’ and the Parsons Tool Company. The binary-coded punched-paper tape, controlled the simultaneous machine tool axes using alpha-numerical characters (ie the forerunner of today’s programs using ‘G- and M-coded’ CNC controllers), through a

14 Who, when and where ‘recirculating ballscrew’ design and development took place is open to some debate. As proponents in the UK say it was Alfred Herbert and Sons, whereas in the United States, the Parsons Tool Company are often quoted as the originators. However, what is not in question, is that with its unique ‘Gothic’ arch’ (i.e. Ogival geometry), having point contacts between the screw and the adjacent recirculating balls, allows the assembly to be pre-loaded in-situ, thereby eliminating any appreciable backlash allowing accurate control of these axes. NB The previous Acme taper thread (i.e. 29° included angle) fitted to ‘conventional’ machine tools had an efficiency of no better that 40% – with backlash present, whereas today’s hardened ‘ballscrews’ have efficiencies of ~90%, coupled to an impressive rigidity (~900 N µm–1) and minimal ‘stick-slip’ , therefore minimising the so-called ‘ballscrew wind-up’ due to the action of torque-effects in combination with the cutting forces.


51

valve-driven hydraulically-servo controlled ‘computer’ called ‘Whirlwind’. In the late 1970’s, with the advent of microprocessor technology, these later NC machine tools were converted to Computer Numerical Control (CNC), offering a significant stride forward in operator-usability, via on-board editing – without the costly and timely re-punching of NC paper tapes each time a minor modification occurred to the NC program. Today, CNC machine tools have fast multiple-processor controls, with on-line computer graphics, enabling new programs to be written and ‘prove-out’ while the machine tool cuts other components, or the programs can be automatically down-loaded by a Direct Numerical Control (DNC) data-link from the CAD/CAM workstation, or via remote satellite-linkage from other sites either locally, or internationally. The design and development of some of today’s and the future machine tools, utilise ultra-fast CNC microprocessors, coupled to orthogonal multi-axes linear-induction motordriven slideways, that can be precisely monitored via laser-controlled positional encoders, with ultra-fast co-axial spindles. Moreover, non-orthogonal-axes controlled machine tools are under development, using simultaneous mulitple-axes slideway control, with hybrids having tool spindles that incorporate multiple angular orientation together with their linear slideways for truly sculptured free-form surface machining capabilities. Even today, operations carried out by several machine tools are now being incorporated into one hybid machine tool, with such as: turning, milling and grinding at one set-up. In the near future, the machine tools will have slideway acceleration/decelerations of faster >5g’s, with these machines having the ability to: rough-turn, mill, heat-treat, grind critical features, all remotely-controlled via satellite from the CAD/CAM designer, significantly speeding-up the product development process time-to-market.

In 1873, Hartig tabulated research into metal cutting in a book, which was the first authoritative work on the subject. A more practical metal cutting description was given by Tresca (1878), ustilising visio-plasticity models15. In 1881, a presentation at the Royal Society of London by Lord Rayleigh of Mallock’s metal cutting research findings was given. Mallock’s scientific study of carefully etched specimens of the workpiece and attached chip for both ferrous and non-ferrous metals, where he observed them using a microscope (magnification: x5). Mallock correctly surmised from his investigation of his ‘models’ that the cutting process was basically one involving shearing and, that friction occurred in forming the chip, emphasizing the importance of this friction along the cutting tool’s face – between the chip and the tool. The sharpness of the cutting edge was also mentioned and the reasons for instability of the cutting process, leading to unwanted vibrations, or ‘chatter’. Moreover, Mallock employed basic lubricants in this work, noting that the application of lubrication reduced chip/tool interface friction. These general observations by Mallock mentined above, offer a surprisingly close approximation to today’s theories on the ‘mechanics of metal cutting’ , although his equations for the work done in internal shearing and chip and tool friction were incorrect, surprisingly, he was unaware of the ‘plasticity models’ by Tresca and his theory of ‘plastic heating’. To compound the metal cutting problems still further, in 1900, an unfortunate ‘step backward’ in the understanding of the metal cutting process was taken by Reuleaux. He suggested that a crack occurred ahead of the tool’s point and likened the cutting action to that of splitting wood, regrettably having popular support for some years. In 1907, a seminal paper by the now-famous American researcher Taylor, who published his 26 years of

2.2.2 Pioneering Work in Metal Cutting – a Brief Resumé

15 Tresca’s visio-plasticity models, involved scoring a grid of accurate closely-spaced lines onto the edge of a specimen of metal to be machined, then partially cutting it at a preset depth of cut and leaving the chip attached. He then investigated the plastic deformation that had taken place as these grids were distorted and buckled by the action of machining. Both lighter and deeper cut depths were investigated in this manner, across a range of metal specimens. Tresca noted that finer depths of cut introduced greater plastic deformation than larger cut depths, stating that stiffer and more powerful machine tools were needed to benefit from these recommended greater depths of cut (i.e. undeformed chip thickness).

Basic research into metal cutting did not commence until approximately 70 years after the first machine tool was introduced. In 1851, early research by Cocquilhat was into the work required to machine a given volume of material by drilling. By 1870, the terms ‘chip’ and ‘swarf ’ were introduced by the Russian engineer Time, where he attempted to explain how chips were formed.

52

Chapter 2

. Figure 24. The formation of a continuous chip, based upon the ‘deck of cards’ principle. [After: Piispanen, 1937]

practical experience into investigation and research findings in metal cutting. Taylor, was fascinated by the application of time-and-motion studies that could be applied within the machine shop and in particular, ‘piece-work systems’16. In order to enable the progression through optimisation of these time-and-motion studies, new cutting tool materials were employed, in particular high-speed steels (HSS). Taylor investigated the effect that tool materials and in particular, cutting conditions had, on tool life during roughing operations, in order to assist in the application of these time-and-motion studies. His principal objective was to establish empirical laws that would enable optimum

cutting conditions to be attained. By establishing optimum cutting data for metal cutting operations and employing ‘piece-work systems’ at the company, Taylor was able to increase the Bethlehem Steel Company’s output by 500%. Of particular note, was the fact that the empirical law governing the cutting tool and its anticipated tool life17 is still used today, in the study of machining economics – more will be said on this topic later in Chapter 7 (Machinability and Surface Integrity). Notable in the years prior to World War Two, were the contributions made into the generation of data on cutting forces and tool life, initially by Boston (1926)

16 Piece-work systems are where a set time allowance is given for a particular job, or a batch and, a bonus is agreed if the worker performs this task within the allotted time.

17 Taylor’s machinability work produced a fundamental discovery, namely, that the interface temperature existing at the tool’s cutting edge controlled the tool-wear rate.


53

. Figure 25. Variations in chip morphological surfaces at different cutting speeds, giving an indication of the various shearing mechanisms. [Source: Watson & Murphy, 1979]

and later, by Herbert (1928). Around this time, the cutting speeds were steadily improving with the arrival of new cutting tool materials, such as cemented carbide. In 1937, Piispanen introduced his so-called ‘Deck of Cards’ principle as an explanation of the cutting process (see Fig. 24 for Piispanen’s idealised model, with Fig. 25 depicting sheared chips at a range of cutting speeds). Here, Piispanen’s model depicts the workpiece

material being cut in a somewhat similar manner to that of a pack of cards sliding over one another, with the free surface an angle, which corresponded to the shear angle (ϕ). So, as the tool’s rake face moves relative to that of the workpiece, it ‘engages’ one card at a time, causing it to slide over its adjacent neighbour, this process then repeats itself ‘ad finitum’ – during the remainder of the cutting process. Some important

54

Chapter 2

limitations are present with Piispanen’s model, namely that it: • exaggerates strain in homogeneity, • shows tool face friction as elastic rather than plastic in nature, • considers shearing takes place on a completely flat plane, • assumes that BUE does not occur, • takes an subjectively assumed shear angle, • takes no account of either chip curling, or prediction of chip/tool length. NB Piispanen’s model is easily understood and does contain the major concepts in the chip-forming process – admittedly for simple shear in the main. By way of further information concerning chip morphology: the micrographs of chip surfaces illustrated in Fig. 26 show in these cases, that the morphology indicates a semi-continuous chip form. These chip forms point towards the fact that noticeable periodic variations have occurred, perhaps as the result of the stress becoming unstable, rather than resulting from any vibrational effects produced by the machine tool. Any such instability, has the effect of causing minute oscillations (i.e. backward and forward motion) in the shear zone, while the machining takes place. The differences in segment shapes shown and their frequency occurring at differing cutting data in these micrographs, are thought to be dependent upon the frequency of the shear plane’s oscillation relative to the cutting speed. A considerable volume of fundamental work on machining research has been undertaken over the last few years, but during World War Two (i.e. from a European perspective), Ernst and Merchant (1941) produced another significant paper dealing with the mechanics of the machining process – some of these research findings will be briefly dealt with in the chapter on Machinability and Surface Integrity, along with other contributions to this subject.

picted schematically in Fig. 26. The overall machining process is well concealed behind a amalgamation of: workpiece material, high speeds and feeds, elevated temperatures and enormous pressures18. The actual cutting dynamics in contemporary machining operations, utilises just a few millimetres of physical contact between the tool and the chip of a precisely-shaped cutting edge geometry in an exotic mixture of tool material to efficiently machine the workpiece – this being an impressive occurrence worthy of note. In the early work on machining, it was thought that the chip was formed by deformation along a shear plane, elastically in the first instance, then plastically as the evolving chip passed through a stress concentration. The Piispanen model (i.e. Fig. 24) illustrates this point, where workpiece material is being cut by progressive slip relative to the tool point, an angle which corresponded to that of the shear plane. Here (i.e. Fig. 24), it shows how each chip segment forms a small, but very thin parallelogram, with slippage occurring along its shear plane. In an orthogonal cutting process19, as the workpiece material approaches this ‘shear plane’ it will not begin to deform until it reaches the ‘shear plane’. Here, it is transformed from that of simple shear, as it moves across a thin shear zone, with the minute amount of secondary shear being virtually ignored, as is the case for tertiary shear – this being the equivalent of a sliding friction but having a constant coefficient of friction. Chip deformation in reality, is produced over a zone of finite width, usually termed the ‘primary shear zone’ (see Fig. 26). As the chip evolves, the back of the chip tends to be roughened, due to the plastic strain being inhomogeneous in nature (see Fig. 25). This shearing action creates a particular chip morphology as a result of the either, stress concentrations, or by presence of points of weakness in the workpiece be-

2.3 Chip-Development

18 Interface pressures between the chip and the tool are normally exceedingly high, typically of the order of 1,000 to 2,000 N mm–1, with temperatures in certain instances at the tool’s face reaching approximately 1100°C.

Most metallic materials can be considered as relatively hard to machine and this is evident from all of the reported literature on the subject of metal cutting, indicating that shearing occurs in a concentrated region between the chip and tool, this effect being de-

19 Orthogonal machining, is when the cutting tool’s edge (i.e. rake face – see Fig. 19b) is presented ‘normal’ to the evolving chip and thus, to the workpiece, at 90° to the relative cutting motion. That is, little if any, side shearing action occurs, while the chip is being formed as it progresses up the tool’s rake face – effectively created by two distinct cutting forces: tangential and axial.


55

. Figure 26. Schematic representation of a sing-point stock removal process, during the continuous cutting of ductile metals

56

Chapter 2

ing machined20. Once the chip deformation begins, it will continue within this ‘zone’ , as though here in this vicinity, the workpiece material is exhibiting a form of negative strain-hardening. The oblique cutting process21 presents a different and much more complex analytical problem, which has been the subject of a lot of academic interest over the years. Even here, the whole cutting dynamics change, when the tool’s top rake surface is not flat, which is the normal status today, with the complex contoured chipbreaker geometries nowadays employed (typically illustrated in Figs. 4, 10 and 27a). Actual chips are normally severely work-hardened, in particular with any strain-hardening materials (for example: high-strength exotic alloys employed for heat-resistance/aerospace applications) as they evolve, by the combined action of: elevated interface temperatures, great pressures and high frictional effects. Such machined action of the combined effects of mechanical and physical work, produce a ‘compressive chip thickness’ 22, which is on average, dimensionally wider than the original undeformed chip thickness (see Fig. 26). The rake angle depicted in Fig. 26 is shown as positive, but its geometry can tend to the neutral, right through to the negative in its inclination. As the rake angle changes, so will the complete dynamic cutting behaviour also change, modifying the mechanical and

20 As the shear plane passes through a particular stress concentration point, it will deform more readily and at a lower stress value, than when one of these ‘points’ is not present. 21 Oblique machining, is when the rake face has a compound angle, that is it is inclined in two planes relative to the workpiece, having both a top and side rake to the face, creating a threeforce model (see Fig. 19a), where the cutting force mathematical dynamics are extremely complex and are often produced by either highly involved equations, or by cutting simulations. This latter simulated treatment is only briefly mentioned later and is outside the remit of this current book. However, this information on dynamic oblique cutting behaviour can be gleaned, from some of the more academic treatment given in some of the selected books and papers listed at the end of this chapter. 22 Compressive chip thickness is sometimes known as the: chip thickness ratio (r)* – being the difference between the undeformed chip thickness (h1) and the width/chip thickness of the chip (h2). *Chip thickness ratio (r) = h1/h2 ** (i.e. illustrated in Fig. 26). ** h2 = W/ρwl Where: W = weight of chip, ρ = density of (original) workpiece material – prior to machining, w = chip width (i.e DOC), l = length of chip specimen.

physical properties within the chip/tool region, as the various deformation zones are distinctly altered. In effect, due to rake angle modification (i.e. changing the rake’s inclination), this can have a profound affect on the: cutting forces, frictional effects, power requirements and machined surface texture/integrity. The chips formed during machining operations can vary enormously in their size and shape (see Fig. 35a). Chip formation involves workpiece material shearing, from the vicinity of the shear zone extending from the tool point across the ‘shear plane’ to the ‘free surface’ at the angle (ϕ) – see Fig. 26. In this region a considerable amount of strain occurs in a very short time interval, with some materials being unable to withstand this strain without fracture. For example, grey cast iron being somewhat brittle, produces machined chips that are fragmented (i.e. termed ‘discontinuous’), conversely, more ductile workpiece materials and alloys such as steels and aluminium grades, tend to produce chips that do not fracture along the ‘shear plane’ , as a result they are continuous. A continuous chip form may adopt many shapes, either: straight, tangled, or with different types of curvature (i.e. helices – see Fig. 35a). As such, continuous chips have been significantly worked, they now have considerable mechanical strength, therefore efficiently controlling and dealing with these chips is a problem that must be overcome (see the section on Chip-breaking Technology). Chip formation can be classified in a number of distinct ways23, these chip froms will now be briefly reviewed: • Continuous chips – are normally the result of high cutting speeds and/or, large rake angles (see Figs. 26 and 27b). The deformation of workpiece material occurs along a relatively narrow primary shear zone, with the probability that these chips may develop a secondary shear zone at the tool/chip interface, caused in the main, by frictional effects. This secondary zone is likely to deepen, as the tool/chip friction increases in magnitude. Deformation can also occur across a wide primary shear zone with

23 One of the major cutting tool manufacturer classifies chips in seven basic types of material-related chip formations, these are: Continuous, long-chipping – mostly steel derivatives, Lamellar chipping – typically most stainless steels, Short-chipping – such as many cast irons, Varying, high-force chipping – many super alloys, Soft, low-force chipping – such as aluminium grades, High pressure/temperture chipping – typified by hardened materials, Segmental chipping – mostly titanium and titanium-based alloys.


. Figure 27. Chip-breaking inserts and chip control whilst turning – in action. [Courtesy of Iscar Tools]

57

58

Chapter 2

. Table 4. Strength and hardness of chips when turning mild steel. Rake angle (γ°):

45

35

27

10

10

10

0.30

0.30

0.20

0.20

0.20

0.20

50

50

168

168

168

76

None

None

None

None

75

84

91

92

93

95

272

289

302

320

314

325

HV/UTS

3.6

3.4

3.3

3.5

3.4

3.4

Shear strain (Pa-s)

1.1

1.7

2.1

2.9

3.1

4.0

Feed (f - mm rev–1) Cutting speed ( VC - m min–1) Cutting fluid: Tensile strength (UTS – kg mm–2) Vickers hardness number (Hv)

Soluble oil

Soluble oil

[Source: Nakayama and Kalpakjian 1997]

curved boundaries, with the lower boundary being below the machined surface (Fig. 26), which may distort a softer workpiece’s machined surface – particularly with small rake angles and at low speeds. Strain-hardening of this type of chip, results in it becoming harder than the bulk hardness of the original workpiece material (see Fig. 28c, where the bulk workpiece hardness is 230 HK and the workhardened chip is ≈350 HK). This increase in the chip’s strength and hardness will depend upon the shear strain (see Table 4 for details of the Rheolological24 status, related to the inclination of the tool’s rake angle). Therefore as the rake angle decreases, the shear strain will increase, causing this continuous chip to become both harder and stronger – behaving in a similar manner to that of a rigid, perfectly plastic body. In order to satisfactorily deal with long continuous work-hardened chips, that could either wrap around the machined workpiece, potentially spoiling the surface texture, or become ensnarled around tooling, or even, reduce efficient coolant delivery to the cutting edge, with integrated tool chip-breakers having been designed and developed – see Fig. 27b.

24 Rheology is a branch of science dealing with both the flow and deformation of materials, with the shear strain rate, often termed just the shear rate. (i.e usually quoted in Pascals-seconds ‘Pa-s’).

• Continuous chips with a built-up edge (BUE) – when

machining ductile workpiece materials, a built-up edge (BUE) can form on the tool’s tip. This BUE consists of gradually deposited material layers from the workpiece, hence the term ‘built-up’ (see Fig. 28). As cutting continues, the BUE becomes larger and more unstable, eventually partially breaking away, with some fragments being removed by the underside of the chip, while the remainder is randomly deposited on the workpiece’s surface (Fig. 28a). This process of BUE formation, shortly followed by its destruction, is continuously repeated during the whole cutting operation. The BUE deposited on the workpiece will adversely affect the machined surface texture. The BUE modifies the cutting geometry, creating a large cutting tip radius (Fig. 28a and b). Due to the BUE being severely work-hardened by the action of successive deposits of workpiece material, the BUE’s hardness significantly increases by around 300% over the bulk component hardness (Fig. 28c). At this severely work-hardened level, the BUE becomes in effect a modified cutting tool. Normally, an unstable BUE is undesirable, conversely, a thin stable BUE is as a rule, regarded as desirable, as it protects the top rake surface. The formation mechanism for the BUE is thought to be one of adhesion of workpiece material to the tool’s rake face, with the bond strength being a function of the affinity of the workpiece to that of the tool material. This adhesion, is followed by the successive buildup of adhered layers forming the BUE. Yet another factor that contributes to the formation of a BUE,


. Figure 28. The development of a continuous chip with Built-Up Edge (BUE), its typical hardness distribution and its affect on the machined surface

59

60

Chapter 2

is the strain-hardening tendency of the workpiece material. Therefore, the greater the strain-hardening exponent, the higher will be its BUE formation. From experiments conducted on BUE formation, it would seem that the higher the cutting speed, the less is the tendency for BUE to form. Whether this lack of BUE formation at higher cutting data is the consequence of increased strain rate, or the result of higher interface temperatures is somewhat open to debate. However, it would seem that a paradox exists, because as the speed increases, the temperature will also increase, but the BUE decreases. The propensity for BUE formation can be lessened by: I. Changing the geometry of the cutting edge – by either increasing the tool’s rake angle, or decreasing the DOC, or both, II. Utilising a smaller cutting tip radius, III. Using an effective cutting fluid, or IV. Any combination of these factors.

• Discontinuous chips – consist of adjacent work-

piece chip segments that are usually either loosely attached to each other, or totally fragment as they are cut (Fig. 29). The formation of discontinuous chips usually occur under the following machining conditions: I. Brittle workpiece materials – these materials do not have the machining capability to undergo the high shear strains, II. Hard particles and impurities – materials with these in their matrix, will act as ‘stress-raisers’ and actively encourage chip breakage, III. Very high, or low cutting speeds – chip velocity at both ends of the cutting spectrum, will result in lack of adherence/fragmentation of the chip segments, IV. Low rake angles/large DOC’s – either small top rakes and heavy DOC’s will decrease the adherence of the adjacent chip segments, V. Ineffective cutting fluid – poor lubricity, combined with a meagre wetting ability, will encourage discontinuity of chip segments, VI. Inadequate machine tool stiffness – creating vibrational tendencies and cutting instability, leading to disruption of the machining dynamics and loosening of chip segments.

As mentioned in ‘Roman II’ above, the hard particles and impurities tend to act as crack nucleation sites, therefore creating discontinuous chips. Large DOC’s

increase the probability that such defects occur in the cutting zone, thereby aiding discontinuous chip formation. While, faster cutting speeds result in higher localised temperatures, causing greater ductility in the chip, lessening the tendency for the formation of discontinuous chips. If the magnitude of the compressive stresses in the both the primary and secondary shear zones significantly increase, the applied forces aid in discontinuous chip formation, this is because of the fact that the maximum shear strain will increase, due to the presence of an increased compressive stress. NB Due to the nature of discontinuous chip formation, if the workpiece-tool-machine loop is not sufficiently stiff, this will generate vibrational and chatter tendencies, which can result in an excessive tool wear regime, or machined component surface damage.

• Segmented chips – are sometimes termed: in-, or

non-homogeneous chips, or serrated chips. This chip form has the characteristic saw-toothed profile which is noted by zones of low and high shear strain (Fig. 30). These workpiece materials possess low thermal conductivity, as such, when machined their mechanical strength will drastically decrease with higher temperatures. This continuous thermal cycle of both fracture and rewelding in a very narrow region, creates the saw-toothed profile, being particular relevant for titanium and its alloys and certain stainless steel grades. For example, to explain what happens in realistic machining situation, the specimen Fig. 30a is displayed, for an austenitic stainless steel quick-stop micrograph. This micrograph being the result of a less than continuous machining process (1), utilising a 5° top rake-angled turning insert. Here, variations in the cutting process have created fluctuations in the cutting forces, resulting in waviness of the machined surface (2). Prior to the material yielding, then the shearing process occurring, the workpiece material has deformed against the cutting edge (3). To explain how changing the top rake angle influences the resultant chip formation for an identical stainless steel workpiece material, Fig. 30b is shown. Machining has now been undertaken with a 15° top rake, promoting a more continuous machining process than was apparent with the 5° tool (i.e. illustrated in Fig. 30a). This more efficient cutting process, results in smaller variations in the cutting forces (1 and 2). The chip is seen to flow over the rake face in a more


. Figure 29. Discontinuous chip formation. [Courtesy of Sumitomo Electric Hardmetal Ltd.]

61

62

Chapter 2

consistent manner (3). It was found with this workpiece material in an experimental cutting procedure, that the tangential cutting force component, was closer to the actual cutting edge than when similar machining was undertaken on unalloyed steel specimens. NB The cutting data for machining the stainless steel specimens in Figs. 30 a and b, were: 180 m min–1 cutting speed, 0.3 mm rev–1 feedrate, 3 mm DOC.

2.4 Tool Nose Radius The insert’s nose radius has been previously mentioned in Section 2.1.6, concerning: Cutting Tool holder/Insert Selection. Moreover, the top rake geometry of the cutting insert will significantly affect the chip formation process, particularly when profile turning. In Fig. 31a, a spherically-shaped component is being ‘profile machined’ using a large nose-radiused turning insert. Here, as the component nears its true geometric curvature, the cutting insert forces will fluctuate continu-

. Figure 30. Segmented chip formation, resulting from machining stainless steel and the work-hardening zone – which is affected by the sharpness of the insert’s edge. [Courtesy of Sandvik Coromant]


63

. Figure 31. The cutting insert’s tool nose radius when either profiling, or general turning, will modify both the profile and diameter as flank wear occurs. [Courtesy of Sandvik Coromant]

64

Chapter 2

ously as the insert progresses (i.e circular interpolates with the X- and Z-axes of the machine tool) around the curved profile. If the geometry of the tool was not itself of round geometry, then the ‘point-contact’ could not be maintained, leading to significant variations in chip formation. If this lack of tool-work contact were not to occur, then the machined profile would be compromised and due to insufficient chip control, the actual cut surface profile would not have a consistent and accurate surface texture. The machined surface texture generated by the passage of the cutting insert’s geometry, is to a large extent the product of the relationship, between the nose radius and the feedrate and, to a lesser degree the cutting speed and its tool wear pattern. The size of the tool nose radius will have quite an effect on the surface texture produced, if the feedrate is set, then a small nose radius will create a different workpiece surface texture to that of a larger one (see Fig. 31b). Moreover, if a large nose radius is selected for a lighter DOC, or if the feed is equal to the nose radius, then this larger nose geometry will be superior to that of a smaller tool nose radius. This is because the ‘larger nose’ offers a smaller plan approach angle, having the pressure of the cut distributed across a longer cut length, creating an enhanced surface texture. There are several disadvantages to utilising a larger tool nose radius geometry, these are that the: • Chip formed becomes more difficult to bend and effectively break, • Radial cutting forces are greater, • Power consumption increases, • Rigidity of the set-up is necessary – leading to possible vibrational tendencies on either weaker, or unstable workpieces. Tool wear (i.e. denoted by ‘∆’ in Figs. 31ci and cii) and in particular flank wear25, can significantly influence the resulting machined component dimensional accuracy (Fig. 31cii), which on a batch of components cut with the same insert, will result in some level of ‘tool

25 Flank wear is normally denoted by specific ‘zones’ – more will be said on this topic later – but, in this example, the tool’s insert wear ‘VB’ is shown in both Figs. 31ci and cii.

drift’ which could affect the process capability26 of the overall parts produced. This flank wear ‘VB’ can be calculated and utilised to determine the anticipated tool’s life (ie, in-cut), this important factor in production machining operational procedure, will be discussed in due course. Wiper blades (Fig. 32) are not a new insert geometry concept, they have been used for face milling operations for quiet a long time, but only in recent years are they being utilised for component finish turning. The principle underlying a wiper insert for turning operations, concerns the application of a modified ‘tool nose radius’ (see Fig. 32 – bottom left and right diagrams). When a ‘standard’ tool nose geometry insert is used (i.e. Fig. 32 – bottom left), it creates a series of peaks and valleys (i.e. termed ‘cusps’27) after the passage of the ‘insert nose’ over the machined surface. Conversely, a cutting insert with wiper blade geometry (i.e. Fig. 32 – bottom right), has trailing radii that blends – beyond the tangency point – with the tool nose radius which remains in contact with the workpiece, allowing it to wipe (i.e. smooth) the peaks, leaving a superior machined surface texture. In the past, wiper insert geometries were only employed for surface improvement in finishing opera-

26 Process capability denoted by ‘CP’ , is a measure of the quality of the parts produced, which is normally found by the following simple relationship: *CP = Drawing specification tolerance/6 σ Where: σ = a statistical measure, termed the ‘standard deviation’ for the particular production process. *CP values of 1.33 are termed as high process capability. NB Today, process capabilities of 2.0 are often demanded for high-quality machined parts for the automotive/aerospace sectors of industrial production, reducing likelihood of part scrappage. 27 Cusps are the product of the partial geometry of the tool nose radius geometry, positioned at regular intervals related to the selected feedrate. The cusp height (i.e. the difference in height between the peak and valley), will influence the machined surface texture of the component, in the following relationship: Rmax = fn2 × 250/rε (µm) Where: Rmax = maximum peak-to-valley height within the sampling length. fn = feedrate (m min–1) rε = tool nose radius (mm).


. Figure 32. The application of wiper insert geometry on the resulting surface texture when fine turning. [Courtesy of Iscar Tools]

65

66

Chapter 2

tions. With recent advancement in wiper28 geometry, this has allowed them to be used at double the previous feedrates for semi-finishing/roughing operations, without degrading the surface texture. The wiper geometry being in contact with the workpiece’s surface for longer than equivalent standard insert nose radius tends to wipe – hence its name, or burnish the machined surface, producing a smoother surface texture. Due to the fact that a ‘wiper’ has an extended edge, the cutting forces are distributed across a longer tool/chip contact region. The wiper portion of the insert, being somewhat protected, enables these wiper inserts to increase tool life by up to 20% more than when using conventional tool nose geometries. Wiper blades have their clearance lengths carefully designed, if they are too long, the insert generates too much heat, on the contrary, they need to be long enough to cope with relatively large feeds, while still smoothing over the surface cusps. Wipers with positive turning insert geometries, they can cope with feedrates of 0.6 mm rev–1 at DOC’s of up to 4 mm. For example, with steel component hardnesses of 65HRc, this often negates the need for any successive precision grinding operations. By designing wiper geometries with the cutting edge and nose radii to improve machined surface finish, while increasing tool life, can be considered as outstanding tool design.

2.5 Chip-Breaking Technology 2.5.1 Introduction to Chip-Breaking The technology of both chip-forming and chip-breaking has been one of the major areas of advancement in recent years. A whole host of novel toolholders and cutting inserts has been developed to enable the cutting process to be under total chip control, allowing some toolholder/inserts combinations to machine multiple component features with just one tool, removing at a ‘stroke’ the non-productive aspects of

28 Some tooling manufacturers have re-named wiper inserts as high-feed inserts, as they have demonstrated in production conditions to promote higher component output, without the recourse to expensive capital outlay.

tool-changing and setting, significantly increasing machine tool utilisation rates. Even when conventional turning inserts are employed, for heavy roughing cuts (Fig. 33a), where feedrates are high as are the large DOC’s, efficient control of the chip must be achieved. To enable excellent control of chip-breaking with roughing cuts (Fig. 33b), a similar overall insert geometry is shown to that in the previous example, but here the rake face embossed dimples/chip-breakers differ significantly. Finally, for light finishing cuts (Fig. 33c), chips are broken in a totally different manner to that of the previous examples. Hence, with all of these differing types of turning operations on workpieces, control of the chip is vital, as it can drastically impair the overall production rates and affect part quality, if not given due consideration. Chip formation is chiefly influenced by the following factors: • Workpiece material composition – its heat treatment (i.e. if any), which affects the chip’s strength, • Insert’s cutting geometry – rake and clearances, as well as any chip-formers present, the geometry being associated with the work piece material, • Plan approach angle – depending upon whether roughing, or finishing cuts are to be taken, • Nose radius – this being linked to the feedrate and here, to a lesser extent, the surface texture requirements, • Undeformed chip thickness (i.e. DOC) – this will affect the chip curling aspect of the chip’s formation – more will be said on this topic in the following section. Note: Another important factor that can also play a significant role in chip formation, is the application of coolant and its supply velocity. The shear angle has some effect on the contact length between workpiece and the rake face and, it is in this vicinity that cutting forces and machining-induced temperatures predominantly affect the cutting insert. Moreover, the insert’s rake is significant, in that as the rake angle increases the contact length decreases, the more positive the rake, the shorter the contact length. Actual chip formation is primarily dependent upon several factors: DOC, feedrate, rake angle, together with the workpiece’s mechanical strength, noting that the chip starts forming in the primary deformation zone (see Fig. 26). Thus, the chip is subsequently formed by the bending force of the cutting action, effectively ‘pivoting’ from the chip’s roughen ‘free top surface’ ,


. Figure 33. Turning cuts and associated insert geometries for forming and shearing of a chip. [Courtesy of Sandvik Coromant]

67

68

Chapter 2

this being a somewhat shorter length than that of the ‘shiny’ underside at the tool/chip interface. Many theories have been given for the actual ‘cause and effect’ of preliminary chip formation which is schematically illustrated Fig. 33d – ‘A’- one such, being that any formation is related to the cutting speed. A large insert rake angle normally means that there is less tendency for chip curling through a larger radius, but it will have lower cutting forces. In Fig. 33d – ‘B’ , is depicted a somewhat ‘idealised’ view of the actual cutting process, which can be expressed via the simple relationship of ‘λ’ and ∆X/∆Y. NB: In this schematic representation: ‘h1’ represents DOC and, ‘ϕ’ is the ‘shear plane angle’. When utilising CNC machine tools and in particular turning centres, a major problem is the variety of continuous chip forms created and the large quantity and volume of swarf29 produced. The manner to which swarf affects machining operations depends upon the operating conditions, but fundamentally there are several requirements in any form of swarf control, these are: • The swarf must flow freely away from the cutting zone, without impairing the cutting action’s efficiency, • Swarf must be of convenient size and shape to facilitate handling manually, or in swarf conveyors (i.e. if fitted), together with any future large-volume storage, then transportation and subsequent disposal, • Any swarf should drop away into the machine’s swarf tray, without snarling around, the workpiece, tool, or interfering with other functions such as: automatic tool-changing magazine/turret, in-situ touch-trigger inspection probes, component load-

29 Individual chips when in any great volume are generally termed swarf. It is important to be able to manage this swarf volume and, satisfactory chip control can be determined by ‘Lang’s chip-packing ratio’ , this being denoted by the letter ‘R’ , in the following manner: R = Chip volume (mm3)/Equivalent volume of uncut workpiece material (mm3) NB: ‘R’ ranges from values of 3-to-10, where an R-value of 4 gives satisfactory chip-breaking control, producing neatly curled ‘6 and 9-shaped’ chips.

ing equipment, such as overhead gantries, or dedicated robotic loading devices. In terms of priority for these swarf control factors, possibly the most important one is that the swarf should flow smoothly away from the cutting area, as with the latest chip-breakers fitted to today’s cutting inserts, chips can be readily broken and controlled30, this will be theme of the following section.

2.5.2 The Principles of Chip-Breaking In machining, the cutting edge’s primary function is to remove stock from the workpiece. Whether this is achieved by forming a continuous chip, or by the flow of elemental chips will depend upon several factors, including the properties of the workpiece material, cutting data employed and coolant type and its delivery. The terms ‘long-chipping’ and ‘short-chipping’ are utilised when considering the materials to be machined. Short-chipping materials such as most brasses and cast irons, do not present a chip-breaking problem for swarf disposal, so this section will concentrate on the long-chipping workpiece materials, with particular focus on ‘steel family’ grades. Steels are produced in a wide variety of specifications and this allows their properties to be ‘tailored’ to the specific industrial applications. In addition, these steels methods of primary processing, such as: casting, forging, rolling, forming and sintering, together with the type of subsequent heat treatment, creates still further metallurgical variations that may have an even greater influence on the workpiece’s chip-breaking ability. The workpiece’s strength and hardness values describe the individual material’s character to some extent, but it should be borne in mind that it is the chip’s mechanical strength that determines whether it can be broken with ease. No absolute correlation exists between a steel com-

30 Today, many high-volume manufacturing companies have realised the benefit of the value of clean and briquetted swarf, as opposed to oily scrap swarf, which sells at just ‘fractions’ of this value. At present, briquetted and cleaned aluminium swarf can be sold for approaching £1,000/tonne, moreover, the coolant/oil can be reclaimed, further driving down the overall machining costs. For other non-ferrous ‘pure’ metals and others, such as copper alloys and brasses, the economic savings are even greater.


ponent’s strength and the mechanical strength of the chip, illustrating that a complex metallurgical and cutting tool geometric relationship exists whilst machining occurs. In particular for turning operations, a conventionally-turned chip is a rather frail product of serrated appearance (see Figs 25 and 34a and b). In order to promote good chip-breaking tendencies, thus enabling short elements to be formed, it is necessary to encourage this basic character by causing these serrations to be as deep as possible and the chip sections in between to be rigid. This chip occurrence causes the chip to be inflexible, which can then subsequently be broken. There are several distinct ways in which chips can then be broken, these include: • Self-breaking – this is when the chip’s mechanical strength is not great enough to hold the chip segments together and they consequently break upon exiting the machining region (Fig. 31a), • Chip collision with the workpiece – as the chip is steered towards an obstacle such as the workpiece’s surface this provides the breaking force (see Figs. 33 and 34b), • Chip is stopped by the tool – here the chip-curling behaviour comes into play, this being a function of the: tool’s nose radius geometry, depth of cut and feedrate employed (see Fig. 34 bottom left-hand photograph), the latter two functions affecting the chip cross-section, or chip thickness31.

31 Chip thickness is influenced by the plan approach angle utilised and the DOC, in association with the selected feedrate. The chip thickness is measured across the cutting edge, perpendicular to the cut (i.e. along the main cutting edge). The chip width and thickness are the dimensions that define the theoretical cut of the edge into the workpiece material. Hence, the chip thickness will vary with the size of the plan approach angle according to the relationships involving: feedrate, DOC and the effective cutting depth. The chip thickness is related to the plan approach angle and this affects the amount of pressure bearing upon the cutting edge. Hence, the thinner the chip, the smaller the distributed pressure along the edge and the less power consumed, conversely, the thicker the chip, the greater will be the machine tool’s power consumption. A thicker chip is generally advantageous for an increased tool life, because of the improved contact between the chip and its cutting edge. Furthermore, if the plan approach angle is too small and chip thickness is thin, this will reduce tool life, however, this can be compensated for by increasing the feedrate, to produce a thicker chip.

69

NB The helical formation of this chip-curling behaviour will shortly be mentioned, but prior to this, chipbreakers/formers will be discussed.

2.5.3 Chip-Breakers and Chip-Formers Chip-breakers have been utilised on turning tools for many years, initially introduced in the 1940’s in the form of an abutment, or step, situated behind the rake face of the tool. Hence, with this type of early chipbreaker, as the continuous chip moves across the rake face it collides with this step and breaks. This original form of chip-breaker geomtery was relatively inefficient as the resultant force direction changed with the programmed tool path, this meant that the step would be approached by the chip from differing directions making chip-breaking less controlled. Such chip-breakers were superseded in the 1970’s by inbuilt ‘wavy-shaped’ chip-breakers sintered into the insert’s top face (Fig 34 bottom left-hand photograph). Recent developments in designing chip-breaker geometries by computer-generated (i.e. CAD) techniques, has shown a significant step-forward in both chipformer design enabling chip control and reduction in frictional forces across the rake face at a range of cutting data to be achieved. Such ‘automatic’ chip breaker geometry forces the chip to deflect at a narrower angle, causing it to break off, either immediately, or just after the free end of the chip has hit either the tool’s flank or, the workpiece before the first coil has formed. If such a collision does not take place, the result would be a smaller diameter spiral chip and, it can be anticipated that the chip would still break, but only when it became slightly longer – this later chip breakage is due to the increasing chip mass and the effect of gravity upon it, with, or without any further collision. Chip flow direction will depend upon several factors, such as the: chip-breaker profile, back rake and setting angles, nose radius, DOC and feedrate – these latter three factors require further discussion. The relationship between the nose radius, DOC and feedrate will often change during vectored tool paths in any machining operation. Even though the insert’s nose radius is preset, its influence on the chip direction differs for different DOC’s, depending on how much corner rounding is represented by the total engaged edge length (Fig. 34c). Further, the feedrate also affects the chip thickness: at different DOC’s and with a constant feedrate, the form of chip cross-section (i.e.

70

Chapter 2

. Figure 34. The principles of chip-breaking and chip-breaking envelopes for ‘coma-shaped swarf’ control and insert edge preparations


the ratio of chip width-to-thickness), will change and this has a deleterious effect on the insert’s chip-breaking ability.

71

2.5.4 Helical Chip Formation

the insert’s flank face (see Fig. 27b)32. Only today’s very complex chip-breaker designs can reduce these outward-curling helical chips. Although such chip helices produced by combinations of the feeds and DOC’s that result in the chip width being too small in relation to its thickness must be avoided.

Conventional Turning

Grooving and Recessing

For the general turning operations, such as sliding (i.e. Z-axis tool feeding) and facing (i.e. X-axis tool path motions), the chip is rolled into a helix, simply because the chip edges are formed from different rotation radii (Fig. 34d). Here, the two edges of the chip consume different quantities of workpiece material, creating differing edge lengths, coupled to the fact that a variation in cutting speed is present, these relationships will result in a helical chip formation. The appearance of the chip’s helix depends upon the workpiece’s diameter and its metallurgical specification/condition, which means the chip helices are extremely difficult to quantify. Most common types of helical chip diameters are determined either directly by the initial curvature from its origin, or are the result of additional bending, introduced by the chip-breaker. For example, the helical chip type shown in Fig. 34c (left), has its chip segments turned inwards, this being a desirable chip form when not fully developed, that is prior to the first coil being completed. Whether, or not the chip is of this form will already be determined even before it meets the chip-breaker, this being the result of its cross-section and the natural tendency to bend according to the ‘line of least resistance’. If the chips width is no larger than its thickness, for example, the resistance to bending in the segment-stiffened thickness direction is larger than in the width direction. In this case, unless this kind of chip is broken early, by colliding with either part of the tool, or the workpiece whilst it is still stiff and short – called ‘self-breaking’ – a helical chip will be formed. In this case, the barbed, or serrated edge is turned outwards causing additional bending, this being introduced by the chip-breaker. For example, the helical chip type shown in Fig. 34c (right), becomes difficult and awkward to control. This outward-curving helical chip also has weakened sections in the serrations between the chip segments, but applied loads on it are readily absorbed by the spring action of the chip. This type of chip will break as it hits

In conventional turning operations, it is significantly easier to form a manageable chip, than for features requiring either grooving, or recessing. The chip formed during plunge grooving counter-rotates in relation to the workpiece, whereby it does not experience the same twisting force as chips produced by either Z-, or X-axes turning operations. When grooving, ideally the chip resembles a ‘watch spring’ , indicating that the chip is curling back onto itself and will ultimately break in several distinct ways: such as at the completion of the grooving cycle, or due to friction between the chip and its groove side walls – as the chip diameter becomes greater. In grooving operations, three significant factors affect chip control, these are: (i) Insert geometry – applied to the rake face, can be classified into distinct groupings: • Radial-ground top rake (not shown), producing the desired ‘watch-spring’ chip formation. This grooving insert geometry will not thin the chip, therefore surface finish passes are necessary on both groove side walls. NB For long-chipping materials the chip-former does not provide enough resistance to produce chip curling, hence, a straight flat chip occurs, that may

32 One of the problems with this type of chip-breaking, is the potential for secondary wear on the insert’s non-cutting zone on the face, caused by the chip helix breaking locally against this face. Such an occurrence happens when the chip helix attains such a diameter and pitch that its free-end continually strikes the non-cutting portion of the insert’s edge – termed ‘chip-hammering’ – causing the edge to be locally weakened and to subsequently crumble. NB Chip-hammering can be alleviated by slightly increasing the helix diameter (i.e. by somewhat modifying the cutting data) causing the chip to break against the tool’s flank – below the insert’s cutting edge, this being one of the previously employed and favoured chip-breaking mechanisms, as shown in Fig. 27b.

72

Chapter 2

. Figure 35. The chip-breaking envelopes related to cutting data and chip-curling behaviour. [Courtesy of Sandvik Coromant]


wrap itself around either the tool, or workpiece, but such a geometry is perfect for machining aluminium, or non-ferrous materials.

• Radial top rake (illustrated in Fig. 4 middle and

to the left – three grooving insert sizes illustrated). This radial top rake is designed to thin the chip. Such chip thinning, eliminates the need to undertake finishing passes on the groove’s side walls. Furthermore, this type of grooving insert geometry being on-centre, enables axial turning of diameters for wide shallow grooves33, or recesses. • Raised bumps on top rake (see Fig. 27a – left). This sophisticated grooving geometry is utilised for materials where chip control is difficult, as it provides an ‘aggressive barrier’ to the curling chip. The raised bumps force the chip back onto itself, either producing a tightly curled watch-spring chip, or causes the chip to break. (ii) Surface speed of the workpiece – in order to obtain full advantage of a grooving insert’s chip-forming abilities, the chip must be allowed to flow into the chip-former. This chip-flow can be achieved by either decreasing the workpiece’s surface speed, or increasing the feed – more will be said on this shortly. The former technique of decreasing the surface speed, allows the material to move slower across the top rake of the cutting edge and as a result, has greater contact time to engage the chip-former. This slower workpiece speed, has the benefit of increasing tool life, through lower

33 A groove, or recess, can normally be considered as a straightwalled recessed feature in a workpiece, as illustrated in Fig. 40. Typical applications for grooves are to provide thread relief – usually up to a shoulder – so that a mating nut and its washer can be accurately seated , or for retaining O-rings. As the groove is produced in the workpiece, the tool shears away the material in a radial manner, via X-axis tool motion. The chip formed with insert geometries having a flat top rake, will have an identical width as the tool and can be employed to ‘size’ the component’s width feature. However, this chip action – using such a tool geometry, creates high levels of pressure at the cutting edge as a result of the chip wall friction, which tends to produce a poor machined surface texture on these sidewalls. Grooving with an advanced chip-former insert geometry, reduces the chip width and provides an efficient cutting action, this results in decreasing the cutting edge pressure somewhat. Chip-formers offer longer tool life and improved sidewall finishes with better chip control, than those top-rakes that have not incorporated such insert chip-forming geometric features.

73

tool/chip interface temperatures. The negative factors of such a machining strategy, are that the: • Part cycle times are increased and as a result, any batch throughput will be lessened, • As the cutting edge is in contact for a longer duration, more heat will be conducted into the tool, than into the chip, which could have a negative impact of inconsistent workpiece size control, • Due to the lower workpiece surface speed, the benefits of the insert’s coating will be reduced, as such coating technology tends to operate more effectively at higher interface temperatures. (iii) Increasing the feedrate – by increasing the feed allows it to engage the chip-former more effectively – this being the preferred technique for chip control. A heavier applied feedrate, produces a chip with a thicker cross-section. Further, a thicker chip engages the insert’s geometry with higher force, creating a greater tendency to break. Hence, by holding a constant workpiece surface speed, allows the faster feedrate to reduce cycle times.

Transversal, or Face Grooving Transversal grooving geometry has a curved tearshaped blade onto which, the insert is accurately located and positioned. The transversal insert follows the 90° plunged feed into the rotating face of a workpiece. These tools are categorised as either right-, or left-hand, with the style adopted depending upon whether the machine tool’s chuck rotates anti-clockwise (i.e. using a right-hand tool), or clockwise (i.e. left-hand). The minimum radius of curvature for such transversal grooving tooling is normally about 12mm, with no limit necessary on the maximum radial curvature that can be machined. For shallow face grooves, off-the-shelf tooling is available, but for deep angular face grooves they require specialised tools from the tooling manufacturers. If a relatively wide face groove requires machining with respect to the insert’s width, then the key to success here, is establishing where in the face to make the first plunge. This initial face plunge should be made within the range of the tool’s diameter, otherwise the tool will not have sufficient clearance and will ultimately break. Successive plunges to enlarge the face groove should be made by radially moving the insert 0.9 times the insert’s width, for each additional plunge. The rotational speed for face grooving is usually about 80% of the speed used for parting-off – soon to be

74

Chapter 2

mentioned. Feedrates are normally around 50% of parting-off values, with the proviso that for material which is subject to work-hardening, minimum feeds are necessary. In transversal grooving operations, a unique chip form occurs, because the chip is longer the further away it is from the workpiece’s centre line of rotation. This results in the chip which no longer flows in a straight line across the insert’s edge, instead it moves at an angle. Such a naturally curved chip is difficult to exhaust from the face groove, particularly if it is broken. Hence, no attempt should be made to break the chip. For deep and narrow grooves, the best solution is to retract the tool at short intervals, to check that the blade shows no signs of rubbing, this is to guard against any likely breakage that might occur when machining outside the blade’s range. Due to the fact that transversal grooving tooling is susceptible to chatter34, any excessive overhang of the tool should be minimised. The chip should never be allowed to become entangled within the transversal groove and should be ejected speedily, otherwise the tool is likely to break.

34 Chatter is a form of self-excited vibration and such vibrations are due to the interaction of the dynamics of the chip-removal process, together with the structural dynamics of the machine tool. Such chatter, tends to be at very high amplitude, which can result in either damage to the machine tool, or lead to premature tool failure. Typically, chatter is initiated by a disturbance in the cutting zone, for several reasons, such as: – Lack of homogeneity – in the workpiece material (i.e. typically a porous component, such as is found in a Powder Metallurgy compact), – Workpiece surface condition (i.e. typically a hard oxide scale on a hot-rolled steel component, utilsing a shallow DOC), – Workpiece geometry (i.e. if the component shape produces either a variation in the DOC – for example, because of uneven depth of casting material being machined, or light cuts on interrupted shapes, such as hexagon, square, or rectangular bar stock), – Frictional conditions (i.e. tool/chip interface frictional variations, whilst machining). Regenerative chatter is a type of self-excited vibration, resulting from the tool cutting a workpiece surface that has either significant roughness, or more likely the result of surface disturbances from the previous cut. These disturbances in the workpiece surface, create fluctuations in the cutting forces, with the tool being subjected to vibrations with this process continuously repeating, hence the term ‘regenerative chatter’. Self-excited vibrations can be alleviated by either increasing the dynamic stiffness of the system, or by increasing the damping. NB Dynamic stiffness can be defined as the ratio of the amplitude of the force to the vibrational amplitude.

For any face grooving of workpiece material that is subject to a continuous chip formation, always use copious amounts of coolant and at high-pressure – if possible, to not only lubricate the cutting zone, but to aid in chip flushing from this groove.

Parting-off The parting-off process is normally considered to be a separate machining operation, but it simply consists of cutting a groove to centre of rotation of the workpiece, to release it from the bar stock, or to ‘part-off ’ to a previously formed internal diameter (shown in Fig. 40 for left-hand side operations). Essentially in a parting-off operation, two time-periods are worthy of mention, these are: (i) At separation from the bar stock – a lower spindle speed than was previously used on the workpiece, will prevent the ‘released part’ from hitting the machine and potentially damaging its surface. Moreover, it allows an operator – if present – to hear the change in the lower spindle speed tone, as it is about to separate from the bar stock, avoiding the parting-off tool from getting ‘pinched’ between the stock and the soonto-be-released component. Often, ‘Part-catchers’ are utilised to reduce any surface damage to the falling component, once it has been parted-off. NB If the component to be parted-off is held in a coaxial/sub-spindle, at component release, the additional spindle supports the workpiece and under these conditions, the parting-off operation is virtually identical to that of found in a grooving cycle. (ii) Surface speed reduction – this effectively occurs when the machine’s spindle attains its maximum speed. For example, on a machine tool having a maximum speed of 3,000 rpm, 90 m min–1 would only be achievable until the parting diameter has reached about 8.6 mm. When parting to a smaller diameter than 8.6 mm, the surface speed would decrease at a fixed spindle speed. As the parting diameter reaches 5.8 mm the surface speed would be 55 m min–1, or 60% of the ideal, thus significantly increasing the chip loading as the tool approaches the workpiece’s centreline. In order to alleviate the increasing tool loading, lowering the feedrate by about 50% until separation is just about to occur, then finally dropping the surface speed to almost zero at this point, reduces the tendency for a ‘pip’ to be present on the workpiece. On a CNC driven spindle, it is not advisable for parting-off operations,


to utilise the ‘canned cycle’ such as the ‘constant surface speed’ 35 function. NB A more serious parting-off problem has been that in order to eliminate the pip formed at the centre of the ‘released component’ , some tools have been ground with the front edge angle of between 3° to 15°. Such a front edge geometry, can introduce an axial cutting force component, leading to poor chip control, which in turn, causes the tool to deflect. This parting-off tool deflection, can lead to the component’s face ‘dishing’ , creating a convex surface on one face and a concave surface on the other – so this tool grinding strategy should be avoided. Today, parting-off inserts normally consist of two main types with top rakes that are either of, negative, or positive cutting edge chip-forming geometries. The negative-style of chip-formers are possibly the most commonly utilised. These inserts have a small negative land at the front edge which increases the insert’s strength, giving protection in adverse cutting conditions, such as when interrupted cutting is necessary during a parting-off operation. The land width – often termed a ‘T-land’ , is relative to the breadth of the parting-off tool. This width of the insert’s land has a direct correlation to the feedrate and its accompanying chip formation. The feedrate must be adequate to force the workpiece material over the land and into the chipformer36. Notwithstanding the widespread usage of negative parting-off tooling, positive-style insert geometries have some distinct advantages. The chief one being the ability to narrow the chip at light feedrates, with mini-

75

mal tool pressure. If excessive tool pressure occurs, this can promote work-hardening of the ‘transient surface’ 37 of the workpiece. These abilities are important points when machining relatively low mechanical strength components, which might otherwise buckle if machined with negative-style inserts when subsequently parted-off. Positive cutting edge parting-off tooling having chip-formers, are ideal for applications on machine tools when either low fixed feedrates are utilised, or if the workpiece material necessitates lower cutting speeds. This positive-style of parting-off tooling, operates efficiently when machining softer workpiece materials, such as: aluminium-or, cooper-based alloys and many non-metallic materials, typically plastics. Feed rates can be very low with these positive-type parting tools, down to 0.0254 mm rev–1 with exceptional chip control and consistent tool life. One major disadvantage of using these positive tooling geometries for parting-off, is that the tool is much weaker than its equivalent negative geometry type. The concept of insert self-grip in its respective toolholder, was developed by the cutting tool manufacturer Iscar tools in the early 1970’s and has now been adopted by many other tooling manufacturers (Fig. 40 top left-hand side). These ‘self-grip’ tooling designs, rely on the rotation of the part and subsequent tool pressure to keep the ‘keyed and wedged’ insert seated in its respective toolholder pocket. Previously, doubleended inserts termed ‘dogbones’ , were often used but were limited to low DOC’s – due to the length of the secondary cutting edge, so have been somewhat overshadowed by the ‘self-grip’ varieties of parting-off tooling.

2.5.5 Chip Morphology 35 ‘Constant surface speed’ CNC capability as its name implies, allows the machine tool to maintain a constant surface speed as the diameter is reduced. The main problem with using this ‘canned cycle’ , is that as the maximum spindle speed is reached, the chip load will also increase. This is not a problem, so long as the maximum speed has not occurred, such as when parting-off a component with a large hole at its centre. 36 Parting-off operations that employ a negative-style insert (i.e. with a land and accompanying chip-former), normally have the feedrate determined in the following manner: by multiplying the width of the insert by a constant of 0.04. For example, for a 4 mm wide tool, it is necessary to multiply the insert’s width of 4 mm by 0.04 to obtain a feedrate of 0.16 mm rev–1. This will give a ‘start-point’ for any parting-off operations, although it might be prudent to check this feedrate is valid, from the tooling manufacturer’s recommendations.

The Characterisation of Chip Forms (Appendix 2) In the now withdrawn ISO 3685 Standard on Machinability Testing Assessment, of some interest was the fact that this Standard had visually characterised

37 Transient surfaces are those machined surfaces that will be removed upon the next revolution of either the: – Workpiece (i.e in rotating part operations), or – Cutter (i.e. for rotating tooling – drilling, milling, reaming, etc.).

76

Chapter 2

chip forms under eight headings, with several variations appearing in each groups (i.e. see Appendix 2 for an extract showing these chip form classifications). Although in the main, the chip forms were related to turning, some of these chip morphologies could be extrapolated to other manufacturing processes. The chip type that will be formed when any machining operation is undertaken is the product of many interrelated factors, such as: • Workpiece material characteristics – will the material that forms the chip significantly work-harden?, • Cutting tool geometry – changing, or modifying the cutting insert geometries38 and its plan approach angles will have a major influence on the type of chip formed, • Temperatures within the cutting zone – if high, or low temperatures occur as the chip is formed, this will have an impact on the type of chip formed, • Machine tool/workpiece/cutting tool set-up – if this ‘loop’ is not too rigid, then vibrations are likely to be present, which will destabilise the cutting process and affect the type and formation of chips produced, • Cutting data utilised – by modifying the cutting data: feeds and speeds and DOC’s, with the insert geometry maintained, this can play a significant role in the chip formed during machining operations. NB Chip formation has become a technology in its own right, which has shown significant development over the last few decades of machining applications. As has been previously mentioned, chip formation should always be controlled, with the resultant chips formed being broken into suitable shape formation, such as ‘spirals and commas’ , as indicated by the resultant chip morphology shown in Fig. 35a. Uncontrolled chip-steaming (i.e. long continuous workpiece strands), must be avoided, being a significant risk-factor to both the: machine tool’s operation and its CNC setter/operator alike.

38 Chip-breaking envelopes (see Fig. 34 middle right), are the product of plotting both the feedrate and DOC on two axes, with their relative size and position within the graphical area being significantly affected by the cutting insert’s geometry – as depicted by the three cutting insert geometric versions shown by types: A, B and C (Fig. 34).

For every cutting insert geometry, there is a recommended application area – termed its ‘chip-breaking envelope’ (i.e. see footnote 38 below) – with regard to its range of feedrates and DOC’s. Within this ‘envelope’ , chips of acceptable form are produced by the cutting insert’s geometry. Conversely, any chips that are formed outside this ‘envelope’ are not acceptable, because they are either formed as unbroken strands, or are too thick and over-compressed. When component profiling operations are necessary (Fig. 31a), this normally involves several machining-related parameters: variations in DOC’s, together with path vectoring of the feeds and as a result of this latter point, changes to the resultant chip’s path on the rake face. These factors are important as they can modify the chip morphology when profiling operations include: recessed/undercut shoulders, tapers and partial arcs, facing and sliding operations with the same tool, together with many other combined profiled features. All of these operations make significant demands on the adaptability of the cutting insert’s geometry to efficiently break the chip. In general, the cutting insert’s chip formation principles are concerned with the chip-breaker’s ability to create a chip form that is neither not too tight a curl, nor too open. If chip curling is too tight for the specific machining application, the likely consequences are for a chip form creating: • ‘Chip-streaming’ – producing long chip strands that are undesirable, wrapping itself around the machined surface of the workpiece with workhardened swarf and possibly degrading this machined surface, or may become entangled around the various parts of the machine tool, which could impede its operation, • Excessive heat generation – this can decrease tool life, or be conducted into the machined part and consequently may affect specific part tolerances for the individual part, or could lead to modifications in the statistical variability39 of a batch of parts,

39 Statistical variability in component production can cause variations from one part to another, as the standard deviation and mean changes, these important factors will be mentioned later in the text.


77

• Increased built-up edge (BUE) formation – which having a small angle when compared to the cutting through ‘attrition wear’ 40 may cause the risk of premature cutting edge failure.

When the chip curling is too open, this may result in the following negative tendencies: • Poor chip control – creating an inefficient chipbreaking ability by the cutting insert, • Chip hammering – breaking down the edge and causing it to crumble and as a result creating the likelihood of prematurely failing, • Vibrational tendencies – affecting both the machined surface texture and shortening tool life. Chip formation and its resultant morphology, is not only affected by the cutting data selected, but will be influenced by the plan approach (i.e. entering) angle of the insert. In most machining operations, they are usually not of the orthogonal, but oblique cutting insert orientation, so the affect is for the entering angle to modify the chip formation process. The insert’s entering angle affects the chip formation by reducing the chip thickness and having its width increased with a smaller angle. With oblique cutting geometry, the chip formation is both ‘smoother and softer’ in operation as the plan approach angle tends toward say, 10° to 60°, furthermore, the chip flow direction will also advantageously change with the spiral pitch increasing. As the nose radius is changed with different cutting inserts, this has the effect of changing both the direction and shape of the chips produced. This nose radius geometry is a fundamental aspect in the development of chips during the machining process – as depicted by Fig. 35b. Here, an identical nose radius and feedrate is utilised, but the difference being the DOC’s, with a shallow DOC in Fig. 35b (left), giving rise to a slow chip helix, whereas in Fig. 35b (right) the DOC is somewhat deeper, creating a tighter chip helix which is beneficial to enhanced chip-breaking ability. Shallow cutting depths produce ‘comma-shaped’ chip cross-sections,

40 Attrition wear is an unusual aspect of tool wear, in that it is the result of high cutting forces, sterile surfaces, together with chip/tool affinity, creating ‘ideal’ conditions for a pressure welding situation. Hence, the BUE develops, which builds-up rapidly and is the ‘swept away’ by the chip flow streaming over the top rake’s surface, taking with it minute atomic surface layers from the tool’s face. This continuous renewal and destruction of the BUE, enhances crater wear formation, eventually leading to premature cutting edge failure.

edge. Equally, a larger depth means that the nose radius has somewhat less affect from its radius and greater influence by the entering angle of the cutting edge, producing an outward directed spiral. Feedrate also affects the width of the chip’s cross-section and its ensuing chip flow41. Chip formation begins by the chip curving, this being significantly affected by combinations of the cutting data employed, most notably: feedrate, DOC, rake angle, nose radius dimensions and workpiece condition. A relatively ‘square’ cross-sectional chip normally indicates that an excessively hard chip compression has occurred, whilst a wide and thin band-like chip formation is usually indicative of long ribbon-like chips producing unmanageable swarf. If the chip curve is tight helix, coupled to a thick chip cross-section, this means that the length of the chip/tool contact has increased, creating higher pressure and deformation. It should be noted that excessive chip cross-sectional thickness, has a debilitating effect on any machining process. By careful use of CAD techniques coupled to FEA to construct the insert’s cutting edge, commashaped chips are the likely product of any machining, providing that the appropriate cutting data has been selected. In some machining operations, chip formation can be superior using a slightly negative insert rake angle, thereby introducing harder chip compression and self-breaking of the chip, particularly if utilising small feeds. Conversely, positive rakes can be give other important machining advantages, depending which chip form and cutting data would be the most advantageous to the part’s ensuing manufacture. Usu-

41 Chip-flow is the result of a compound angle between the chip’s side- and back-flow. The chip’s side-flow being a measure of the flow over the tool face (i.e. for a flat-faced tool), whilst backflow establishes the amount of chip-streaming into the chipbreaker groove. Detailed analysis of chip side-flow (i.e. via high-speed photography), has indicated that it is influenced by a combination of groove dimensions and cutting data. If the feedrate is increased, this results in a higher chip backflow angle, promoting chip-streaming into the chip-breaker groove. The ratio of feed-to-length of restricted contact has been shown to be an important parameter in the determination of chip- back-flow. Typically with low feedrates the corresponding chip back-flow is going to be somewhat lessened, resulting in poor chip-breaker utilisation. When the restricted contact between the chip and the tool is small – due to low feed – the chip-flow does not fully engage the chip-breaker and will as a result curve upward, with minimal ‘automatic’ chip-breaking effect.

78

Chapter 2

ally, for larger feedrates, a positive insert rake angle might optimise the chip-curving tendency, by not producing and excessively tight chip helix. Chip curve, its resultant chip flow direction, the chip helix and its accompanying shape are designed into each cutting edge by the tooling manufacturers. Tool companies ensure that a controlled chip formation should result if they are exploited within the recommended cutting data ranges specified. In Fig. 36a (left), effective chip-breaking decisionmaking recommendations are shown on a flow-chart,

indicating how to obtain the desired chip-breaking control. In the chart shown in Fig. 36a (right), the DOC’s indicate on the associated visual table the expected chip type showing that here types ‘C and D’ offer ‘good’ broken chips. Such chip morphology charts as these from tooling manufacturers, attempt to inform the user of the anticipated chip-breaking if their recommendations are followed. Whereas the flow-diagram illustrated in Fig. 36b, indicates that ‘good chip control’ improved productivity will result, if a manufacturing company adopts the machining

. Figure 36. Chip-breaking control and chip morphology and its affect on productivity. [Courtesy of Mitsubishi Carbide]


strategy high-lighted to the left-hand side. On the contrary, ‘poor chip control’ with an attendant decrease in productivity will occur, if the problems shown to the right-hand side transpire. Chip morphology can indicate important aspects of the overall cutting process, from the cutting edge’s geometry and its design, through to work-hardening ability of the workpiece. Many other factors concerning cutting edge’s mechanical/physical properties can be high-lighted, these being important aids in determining a material’s machinability – which will be discussed in more depth later in the text.

2.5.6 Chip-Breaker Wear Any form of tool failure will depend upon a combination of different wear criteria, usually with one, or more wear mechanisms playing a dominant role. Previously, it was found that the workpiece surface texture and the crater index act as appropriate tool failure criteria, particularly for rough turning operations. Moreover, tool life based upon these two factors, approximated the failure curve more exactly than either the flank, or crater wear criterion. In cutting tool research activities, it has been found that when machining with chip-breaker inserts, flank wear (i.e. notably VB) is not the most dominant factor in tool failure. In most cases, the ‘end-point’ of useful tool life occurs through an alteration of the chipgroove parameters, well before high values of flank wear have been reached. The two principal causes of wear failure for chip-breaker inserts are: • For recommended cutting data with a specific insert, the design and positioning of chip-breakers/ grooves may promote ‘unfavourable’ chip-flow, resulting in wear in the chip-breaker wall – causing consequent tool failure, • Alterations in the cutting data, particularly feedrate, affects chip-flow, which in turn, generates various wear patterns at the chip-breaker’s heel and edge (see Fig. 37). In the schematic diagrams shown in Fig. 37, are illustrated the concentrated wear zones on the: back wall (i.e. heel), cutting edge, or on both positions for a typical chip-breaker insert. Under the machining conditions for Fig. 37a, the chip-groove utilisation is very low, with the chip striking the heel directly. Thus, as machining continues, this results in abrasive

79

wear of the heel and ultimately this heel becomes flattened and chip-breaking is severely compromised. Conversely, when the cutting data produces a wear zone concentrated at the insert’s edge (Fig. 37b), then chip side-flow occurs and poor chip-breaking results, together with low tool life. This accelerated tool wear, resulting from an extended tool/chip contact region over the primary rake face, promotes a rough surface texture to the machined part. In the case of Fig. 37c, these are ideal conditions for optimum chipbreaking action and a correspondingly excellent and predictable tool life, because the wear zones at both the heel and edge are relatively uniform in nature, illustrating virtually a perfect chip-forming/-breaking action. Higher tool/chip interface temperatures can result as the heel wears, forming a crater at the bottom of the chip-breaker groove. Combination wear – as shown in Fig. 37c – generally results in significantly improved tool wear, in conjunction with more predictable tool life. In the photographs of chip-breaker grooves shown for an uncoated and coated Cermet cutting insert material in Figs. 38a and b respectively, the relative wear patterns can clearly be discerned. In the case of Fig. 38a – the uncoated insert – the predominant wear concentration is primarily at the edge, indicating that the cutting data had not been optimised. While in the case of the coated Cermet insert of identical geometry (Fig. 38b), the wear is uniform across the: edge, groove and heel. This would seem to suggest that ideal cutting data had been utilised in its machining operation. In both of these cases some flank wear has occurred, but this would not render the chip-breaking ability when subsequent machining invalid. NB A complex matrix occurs (i.e. Fig. 38c) with Cermets, this ‘metallurgy’ can be ‘tailored’ to meet the needs of specific workpiece and machining requirements.

2.6 Multi-Functional Tooling The concept of multi-functional tooling was developed from the mid-1980’s, when multi-directional tooling emerged. This tooling allowed a series of operations to be performed by a single tool, rather than many, typically allowing: side-turning, profiling and

80

Chapter 2

. Figure 37. Schematic representations of differing chip-breaking insert tool wear mechanisms – due to alterations in the cutting data. [Source: Jawahir et al., 1995]


. Figure 38. Improved wear resistance obtained with an uncoated and coated cermet, when turning ovako 825B steel, having the following cutting data: Cutting speed 250 m min–1, feed 0.2 mm rev–1, DoC 1.0 mm and cut dry. [Courtesy of Sandvik Coromant]

81

82

Chapter 2

grooving, enabling the non-productive elements42 in the machining cycle to be minimised. In the original multi-directional tooling concept, the top rake geometry might include a three-dimensional chip-former, comprising of an elevated central rib, with negative Klands on the edges. Such a top rake profile geometry could be utilised for efficient chip-forming/-breaking of the resultant chips. This tooling when utilised for say, grooving operations, employed a chip-forming geometry – this being extended to the cutting edge, which both narrowed and curled the emerging chip to the desired shape, thereby facilitating easy swarf evacuation. A feature of this cutting insert concept, was a form of effective chip management, extending the insert’s life significantly, thus equally ensuring that adequate chip-flow and rapid swarf evacuation would have taken place. When one of these multi-directional tools was required to commence a side-turning operation, the axial force component43 acting on the insert caused it to elastically deflect at the front region of the toolholder. This tool deflection enabled an efficient feed motion along the workpiece to take place, because of the elastic behaviour of the toolholder created a positive plan approach angle in combination with a front clearance angle – see Fig. 39a and b (i.e. illustrating in this one of the latest ‘twisted geometry’ insert multi-functional tooling geometries). Any of today’s multi-functional tooling designs (Figs. 39 and 40), allow a ‘some degree’ of elastic behaviour in the toolholder, enabling satisfactory tool vectoring to occur, either to the right-, or left-hand of the part feature being machined. These multi-func-

42 Non-productive elements are any activity in the machining cycle that is not ‘adding value’ to the operation, such as: toolchanging either by the tool turret’s rotation, or by manually changing tools, adjusting tool-offsets (i.e. for either: tool wear compensation, or for inputting new tool offsets – into the machine tool’s CNC controller), for component loading/unloading operations, measuring critical dimensional features – by either touch-trigger probes, non-contact measurement, or manual inspection with metrology equipment (i.e. micrometers, vernier calipers, etc.), plus any other additional ‘idle-time’ activities. 43 An Axial force component is the result of engaging the desired feedrate, to produce features, such as: a diameter, taper, profile, wide groove, chamfer, undercut, etc. – either positioned externally/internally for the necessary production of the machined part.

tional tools are critically-designed so that for a specific feedrate, the rate of elastic deflection is both known and is relatively small, being directly related to the applied axial force, in association with the selected DOC’s. At the tool-setting stage of the overall machining cycle, compensation(s) are undertaken to allow for minute changes in the machined diameter, due to the dynamic elastic behaviour of one of these tools in-cut. For a specific multi-functional tool supplied by the tooling manufacturer, its actual tool compensation factor(s) will be available from the manufacturer’s user-manual for the product. In-action these multi-functional tools (Fig. 39b), can significantly reduce the normal tooling inventory, for example, on average such tools can replace three conventional ones, with the twin benefit of a major cycle-time reduction (i.e. for the reasons previously mentioned) of between 30 to 60% – depending upon the complexity of features on the component being machined. Some other important benefits of using a multi-functional tooling strategy are: • Surface quality and accuracy improvements – due to the profile of the insert’s geometry, any ‘machined cusps’ 44, or feedmarks are reduced, providing excellent machined surface texture and predictable dimensional control, • Turret utilisation improved – because fewer tools are need in the turret pockets, hence ‘sister tooling’ can be adopted, thereby further improving any untended operational performance, • Superior chip control – breaks the chips into manageable swarf, thus minimising ‘birds nests’ 45 and entanglements around components and lessens automatic part loading problems, • Improved insert strength – allows machining at significantly greater DOC’s to that of conventional in-

44 ‘Machined cusps’ the consequence of the insert’s nose geometry coupled to the feedrate, these being superimposed onto the machined surface, once the tool has passed over this surface. 45 ‘Birds nests’ are the rotational entanglement and pile-up of continuous chips at the bottom of both trough and blind holes, this work-hardened swarf can cause avoidable damage in the machined hole, furthermore, it can present problems in coolant delivery for additional machining operations that may be required.


83

. Figure 39. Multi-functional cutting insert geometry for efficient stock removal and increased part productivity. [Courtesy of Iscar Tools]

84

Chapter 2

. Figure 40. Multi-functional tooling for the machining of rotational features such as: turning, grooving and profiling operations – having excellent chip control. [Courtesy of Sandvik Coromant]


85

. Figure 41. By employing twin turrets on a mill/turn centre, ‘balanced turning’ of the component can remove large volumes of stock at one pass. [Courtesy of DMG (UK) Ltd.]

serts, with improved insert security its toolholder location, • CNC programming simplified – many tooling manufacturer’s have specially-prepared software to significantly reduce CNC programming input times. NB This latter point utilises CNC ‘canned cycles’ to reduce program lengths. In Appendix 3, a guide for ‘Trouble-shooting for turning operations’ are listed, with possible causes and remedies to potential production problems. In the following chapters, many other important chip-forming production processes will be discussed, with hole-making techniques such as drilling being around 25% of all manufacturing techniques undertaken by machining-related companies – these and associated hole-production methods will be reviewed next.

References Journal and Conference Papers Boston, O.W. A Research into the Elements of Metal Cutting Trans ASME 48, 749–848, 1926. Cocquilhat, M. Experiences sur la Résistance utile Produite dans le Forage Ann. Trav. Publ.en Belgique 10, 199, 1851. Doi, S. and Kato, S. Chatter Vibration of Lathe Tools, Trans. of ASME, Vol. 78 (5), 1127–1134, 1956. Fabry, D. The Tool Channel. Cutting Tool Eng’g, 58–64, Sept. 2003. Gadzinski, M. Parting Know-how. Cutting Tool Eng’g, 52–57, March 2001. Galloway, D.F. Some Experiemnts on the Influence of Various Factors on Drill Performance. Trans. of ASME, 191–231, Feb., 1957. Humphries, J.R. Energing Technologies and Recent Advances in Multi-functional Groove/Turn Systems. Int.

86

Chapter 2

Conf. on Industrial Tooling, Shirley Press Ltd, 65–85, Sept. 1979. Isakov, E. The Mathematics of Machining. American Machinist, 37–39, Aug. 1996. Isakov, E. Reassessing Power Factors. American Machinist, 43–45, Dec. 1996. Jawahir, I.S. The Tool Restricted Contact Effect as a Major Influencing Factor in Chip-breaking: An Experimental Analysis. Annals of the CIRP, Vol. 31 (1), 121–126, 1988. Jawahir, I.S., Ghosh, R., Fang, X.D. and Li, P.X. An Investigation of the Effects of Chip-flow on Tool Wear in Machining with Complex Grooved Tools. WEAR, Vol.184 (2), 145–154, 1995. Kasahara, N., Sato, H. and Tani, Y. Phase Characteristics of Self-excited Chatter in Cutting. J. of Engg for Ind., ASME Pub, Vol. 114, 393–399, Nov. 1992. Kennedy, B. Facing Facts. Cutting Tool Eng’g, 29–37, Feb. 2002. Kennedy, B. Take a Bigger Bite. Cutting Tool Eng’g, 25–29, Aug. 2003. King, K. Added Functionality. Cutting Tool Eng’g, 52–55, Feb. 2005. Kondo, Y., Kawano, O. and Sato, H. Behaviour of Self-excited Chatter due to Multiple Regenerative Effect. J. of Engg. For Ind., ASME Pub., Vol. 103 (3), 324–329, 1981. Lewis, B. Turn your Wipers on. Cutting Tool Eng’g, 47–51, Jan. 2003. Mallock, A. The Action of Cutting Tools. Proc. of Royal Soc. Lond. 33, 127–139, 1881–882. Paterson, H. Strictly Boring. Cutting Tool Engg., 22–30, Oct. 1995. Pekelharing, A.J. Built-up Edge (BUE): is the Mechanism Understood? Annals of the CIRP, Vol. 23 (3) 207–211, 1974. Piispanen, V. Eripanines Teknilliseslä Aikakauslehdeslä 27, 315, 1937. Reuleaux, F. Uber den Taylor Whiteschen Werkzengstahl in Verein zur Beförderung des Gewerbefleisses in Preussen. Sitzzungsberichte, 79, 179, 1900. Smith, G.T. Fundamentals of Chip-breaking for Continuous Cutting Operations. Int. Conf. on Industrial Tooling, Molyneux Press Ltd, 72–82, Sept. 1999. Teets, B. Facing up to Grooving Problems. Machinery and Prod. Eng’g., 51–52, Oct. 1988.

Time, I. Soprotivlenie Metallov I Dereva Rezaniju St. Petersburg, 1870. Tipnis, V.A. and Joseph, R.A. Testing for Machinability, in: Influence of Metallurgy on Machinability, ASM Pub. 11–30, 1975. Tresca, H. Mémoire sur le Rabotage des Métaux Bull. Soc. d’Encourgement pour I’Industrie Nationale 585–685, 1873. Venkatesh, V.C. and Satchidanandam, M. A Discussion on Tool Life Criteria and Total Failure Causes. Annals of the CIRP, Vol. 29 (1), 19–22, 1980. Watson, D.W. and Murphy, D.C. The Effect of Machining on Surface Integrity. Metallurgist and Matls, 199–204, April 1979. Webzell, S. Wiping away Cycle Times. Metalworking Prod., Oct. 2003. Books, booklets and guides Armarego, E.J.A. and Brown, R.H. The Machining of Metals. Prentice-Hall Pub., 1969. Boothroyd, G. and Knight, W.A. Fundamentals of Metal Machining and Machine Tools. Marcel Dekker (NY), 1989. Finish Turning – Application Guide. AB Sandvik Coromant Pub., 1995. Hartig, E. Versuche über Leistung und Arbeitsverbrauch der Werkzengmaschine. 1873. Kaczmarek, J. Principles of Machining by Cutting Abrasion and Erosion. Peter Pregrinus Pub. (Warsaw), 1976. Modern Metal Cutting – A Practical Handbook. AB Sandvik Coromant Pub., 1994. Shaw, M.C. Metal Cutting Principles. Clarendon Press, Oxford, 1984. Smith, G.T. Advanced Machining – The Handbook of Cutting Technology. IFS/Springer Verlag, 1989. Smith G.T. CNC Machining Technology. Springer Verlag, 1993. Smith, G.T. Industrial Metrology – Surfaces and Roundness. Springer Verlag, 2002. Stainless Steel Turning, AB Sandvik Coromant Pub., 1996. Tlusty, G. Manufacturing Processes and Equipment. Prentice Hall, 2000. Trent, E.M. Metal Cutting. Oxford: Butterworth Heinemann (3rd Ed.), 1991.

3

Drilling and Associated Technologies ‘In all things, success depends upon previous preparation and without such preparation……there is sure to be failure.’

CONFUCIUS

(c550–c487BC) [Analects]

88

Chapter 3

3.1 Drilling Technology 3.1.1 Introduction to the Twist Drill’s Development Drilling operations are perhaps the most popular machining process being undertaken today, with their origins being traced back to cutting tool developments in North America in the 19th century. In 1864 toward the latter part of the American Civil War, Steven Morse (i.e. later to design the significant ‘Morse taper’ – for accurate location of the ‘sleeved drills’ into their mating machine tool spindles) founded the Morse Twist Drill and Machine Company in the ‘North’. Morse then proceeded to develop probably the most important cutting tool advance to date, namely, the ubiquitous twist drill. In Fig. 42, several of today’s twist drills are illustrated along with just a small range of ‘solid’ contemporary designs. Morse’s originally-designed twist drill has changed very little over the last 150 years – since its conception. In comparison to the somewhat cruder-designed contemporary drills of that time, Morse stated: ‘The common drill scrapes metal to be drilled, while mine cuts the metal and discharges the chips and borings without clogging’. Morse’s statement was at best, to some extent optimistic, whereas the ‘cold reality’ tells a different story, as a drill’s performance is influenced by a considerable number of factors, most of which are listed in Fig. 43.

3.1.2 Twist Drill Fundamentals The basic construction of a conventional twist drill is depicted in Fig. 44a. From this illustration two distinct cutting regions can be established: firstly, the main cutting edge, or lips; secondly at the intersection of the clearance and main cutting edge – termed the chisel edge. In fact for a twist drill, the cutting process can be equated to that of a left-hand oblique turning tool, where the rake and clearance face geometries are identical and the correlation between these two machining processes have been validated in the experimental work by Witte in 1982. Both of these regions remove material, with the cutting lips producing efficient material removal, while the chisel edge’s contribution is both inefficient and is mainly responsible for geometric errors in drilling, coupled to high thrust loads.

The main cutting edges are accountable for a relatively conventional chip formation, as shown in the ‘quick-stop’ photomicrograph in Fig. 44b. An oblique cutting action occurs to the direction of motion, being the result of an offset of the lips that are parallel to a radial line – ahead of centre – which is approximately equal to half the drill point’s web thickness and increases toward the centre of the drill. This obliquity is responsible for inducing chip flow in a direction normal to the lips in accordance with Stabler’s Law. The increasing chip flow obliquity can be seen in Fig. 45a, by observing the flow lines emanating from the chip’s interface along the lips and up the flute face. Such an oblique cutting action serves to increase the twist drill’s effective rake angle geometry. With the advent of ‘Spherical trigonometric computer software’ for obtaining direct three-dimensional calculations – previously described by Witte (1982) in two-dimensional formulae for cutting edge performance – these calculations have been enhanced. Under the chisel point, or web, the material removal mechanism is quite complex. Near the bottom of the flutes where the radii intersect with the chisel edge, the drill’s clearance surfaces form a cutting rake surface that is highly negative in nature. As the centre of the drill is approached, the drill’s action resembles that of a ‘blunt wedge-shaped indentor’ , as illustrated in Fig. 45b. An indication of the inefficient material removal process is evident by the severe workpiece deformation occurring under the chisel point, where such deformed products must be ejected by the drill to produce the hole. These ‘products’ are extruded, then wiped into the drill flute whereupon they intermingle with the main cutting edge chips. This fact has been substantiated by force and energy analysis, based on a combination of cutting and extruding behaviour under the chisel point, where agreement has been confirmed with experimental torque and thrust measurements. The chisel edge in a conventionally ground twist drill has no ‘true’ point, which is one of the major sources for a drilled hole’s dimensional inaccuracy.

Stabler’s Law – for oblique cutting, can be formulated, as below: Chip flow (cos η) = cos I (bc/b) Where: I = inclination of cutting edge, bc = chip flow vector, b = direction of cutting vector.

Drilling and Associated Technologies

The conventional twist drill chisel point geometry can be seen in Fig. 46, together with associated nomenclature for critical features and tolerance boundaries. From the relatively complex geometry and dimensional characteristics shown in Fig. 46, the ob-

89

tainable accuracy of holes generated whilst drilling is dependent upon grinding the drill to certain limits. Any variations in geometry and dimensions, such as: dissimilar lips and angles, chisel point not centralised, and so on, have a profound effect on both the hole di-

. Figure 42. A selection of just some of the many ‘solid’ and ‘through-spindle’ drilling varieties and ‘inserted-edge’ insert geometries currently available. [Courtesy of Seco Tools]

90

Chapter 3

. Figure 43. The principal technical drill performance criteria and factors associated with drilling operations in this case for example, on castings


. Figure 44. The twist drill geometry and associated chip shearing mechanism. [Source: C.J. Oxford Jr., 1955]

91

92

Chapter 3

. Figure 45. The twist drill shearing and extrusion mechanism at the bottom of a hole. [Source: C.J. Oxford Jr., 1955]


. Figure 46. Twist drill geometry

93

94

Chapter 3

mensional accuracy and roundness, with some ‘helical wandering’ as the drill passes through the workpiece. Hole accuracy and in particular the ‘bell-mouthing effect’ , is minimised by previously centre-drilling prior to drilling to ‘size’. The main cause of such this ‘bellmouthing’ is probably the inconsistency in the drill geometry. Such effects are exacerbated using Jobber drills, or even worse, by utilising longer-series drills, which tend to either slightly ‘unwind’ , or bend as a result of lessening rigidity promoting some drill bending/deflection. It is worth noting that the rigidity of a tool such as a drill will decrease by the ‘square of the distance’. Therefore it follows that the greater the drill penetration into the workpiece, the progressively larger the deflection and, the further from the ‘true axis of rotation’ will be the subsequent drill’s path. This deflected drilled hole slope angle ‘ϕ’ , can be defined in the following manner: Drilled hole slope angle ‘φ’ = 3/2 l × R/T (1 – I/k × tan k l) Where: l = length of deflected tool,

R = ratio of the transverse reaction at the drill point, T = thrust force, I = system’s ‘moment of inertia’ , k = √T/E I. As suggested above, this ‘axis slope error’ is initiated when the chisel edge begins to penetrate the workpiece and unless the feed is discontinued, or in some manner the error is corrected, the magnitude of deflection will increase as drill penetration continues. The drill’s magnitude of deflection can reach up to 60 µm, under exaggerated drilling conditions. The geometry of the point has been the subject of considerable research and development for many years, with some unusual departures from the ‘standard’ 118° drill point included angle. Typical of these extreme approaches were the so-called ‘Volvo point’ , having a negative 185° included angle – primarily utilised to avoid ‘frittering’ of drilled holes, or the highly positive geometries such as 80° included angle used for drilling some plastics. Not only can the point angle be modified, but the shape and profile of the chisel point, or web offers numerously-ground opportunities for detailed geometric modifications, with only some of which being shown in Fig. 47. Four of the most commonly-ground drill point geometries being: • Conventional – the ‘original’ Morse geometry, having a straight chisel edge, with poor self-centring drilling action (Fig. 46a), • Split-point – there are a range of point-splitting techniques available to alter the point profile, which has the effect of modifying the chisel point to allow a reasonable self-centring action (Fig. 47b),

‘Helical wandering’ is the result of the drill’s geometry being ‘unbalanced’ , resulting from of differing lip lengths, or an offset chisel point, causing the drill to ‘spiral-down’ through the workpiece, as it progresses through the part (see Fig. 70). ‘Bell-mouthing’ of the drilled hole is attributable to the chisel point and is produced by the line-of-contact, as the drill point initially touches the component’s surface, causing it to ‘walk’ until the feed/penetration stabilises itself at the outer corners (i.e. margins) entering the workpiece, whereupon, these margins guide the drill into the part.

‘Bell-mouthing effect’ is produced by the drill chisel point’s eccentric behaviour as it attempts to centralise its rotational motion as it enters, or exit’s the workpiece.

‘Jobber drills’ are considered to be ‘standardised drills’ that are normally utilised for most drilling general operations, unless otherwise specified.

‘Frittering’ refers to the break-out at the hole’s edge as the drill exit’s the part, on some brittle materials, such as on several Powder Metallurgy compacts.

‘Rigidity rule’: a drill, reamer, tap, or a milling cutter held in a spindle will have its rigidity decreased by the ‘square of the distance’ , namely, if a drill is twice as long it is four times less rigid.

‘Web’ refers to the internal core of the drill – which imparts mechanical strength to the drill. The web increases in thickness the further one gets from the chisel edge (i.e. shown in Fig. 47 – in lower diagrams and with cross-sections). Hence, if the drill is reground many times, the chisel point width will obviously increase, this necessitates that the chisel point must be ‘thinned’ , otherwise too high a thrust force occurs and an inefficient drilling action will result.

‘Split-point’ ground drills are sometimes referred to as ‘Multifacet drills’.

NB A cantilevered tool such as a boring bar has its rigidity decreased by the ‘cube’ or the distance – meaning that too much tool overhang, will seriously reduce tooling rigidity.


. Figure 47. A range of typically ground twist drill points

95

96

Chapter 3

• Web-thinning – as its name implies, the chisel point edges of the drill, increasing both tool life and improvis web-thinned/notched, by regrinding to reduce the width of the chisel point, while slightly modifying the profile, giving a partial self-centring action (Fig. 47c), • Helical – the chisel point is ground to an ‘S-shape’ , which modifies both the chisel point and its profiled shape, improving the drilling performance and self-centring action (Fig.47d). NB On some drills a sophisticated grinding action has imparted drills without a chisel point, which significantly improves their drill penetration rates into the workpiece, but requires a complex drill regrinding operation to re-sharpen them when the edge becomes ‘dulled’. Not only are drills supplied with appropriate point geometries, but for twist drills the twin spiral flutes of the drill can also be specified – from the tooling manufacturer, as this gives the drill its ‘equivalent of the rake angle’ as found on a single-point turning tool. On conventional jobber drills, the normal flute angle is 29° – giving a relatively ‘slow’ helix (Fig. 47a) and in the past, typically being utilised for drilling most plain carbon steel grades. Conversely, a drill with a ‘quick’ helix angle (Fig. 47d), might be employed to drill soft materials such as certain plastics. Brittle materials on the other hand, which might be utilised typically when drilling Cartridge Brass (i.e. 70Cu 30Zn composition), require a zero, or slightly negative helix. NB It is possible to temporarily modify the drill’s helix angle by re-grinding, termed ‘drill dubbing’ , which refers to lightly ‘flash-grinding’ the flutes at the lips to decrease the effective flute helix angle. The main strength of a drill is via its web, or its crosssection which can be changed and as a result, will modify the flute’s geometric profile (i.e. see Fig. 48). In general, drill cross-section are classified in three groups, namely: • Axe-shaped – having well-defined margins (Fig. 48 –top), • Rounded heel – with increased web, but small margins (Fig. 48 – middle), • Rhombic – incorporating a large web, with wide margins (Fig. 48 – bottom). NB Some twist drills feature oil/coolant holes to allow cutting fluid to reach right down toward the cutting

ing the hole’s ‘Surface Integrity’ .

3.1.3 The Dynamics of Twist Drilling Holes Introduction The term ‘drilling’ refers to all production techniques for the manufacture of cylindrical holes in workpieces using chip-making cutting tools, for short-hole10 and deep-hole drilling operations. The expression ‘solid drilling’ has been introduced in recent years – which is hole-making generation undertaken in a single operation, to differentiate it from that of the previous techniques of either: centre-drilling or, pilot-hole drilling (i.e, see Fig. 50b) prior to drilling to size. Drill technology includes a range of specialised hole-making tooling, including: twist drills, solid drills, counter-boring and trepanning tools and deep-hole drills11. In the first instance, mention will be made of twist drilling operations, then a review of these other drilling production methods will occur.

Twist Drills Twist drilling operations have been carried out for around 150 years, with a twist drill imparting ‘balanced cutting conditions’ , assuming that the drill’s geometry is symmetrical. It has been suggested that the work of drilling may be considered as two singlepoint lathe tools engaged in an internal straight turning operation. A twist drill produces both torque and thrust as it rotates and is fed into the workpiece. The main contribution to torque is through the lips, with a small amount of torque being generated by the chisel point as the drill rotates against the resistance of the

‘Surface Integrity’ has been coined to describe the ‘altered material zone’ (AMZ), for localised sub-surface layers that differ from those of the bulk material – considerably more will be said on this subject in Chapter 7.

10 ‘Short-hole drilling’ operations cover depth-to-hole-diameter-ratios of up to 6D (i.e. for diameters up to 30 mm), whilst larger drilled holes are limited to depths of 2.5D. Where: D = nominal drill diameter. 11 ‘Deep-hole drilling’ and ‘Gun-drilling’ operations are virtually the same, with the term Deep-hole drilling being the preferred term in this text.


. Figure 48. Symmetrical twist drill cross-sectional profiles [After: Spur and Masuha, 1981]

97

98

Chapter 3

workpiece (see Fig. 49). The thrust force (Fig. 49) is the result of the selected penetration rate (i.e. feed), in combination with the bulk hardness of the workpiece and its work-hardening ability and the efficiency of the coolant supply – if any – to the cutting edges (i.e lips). The resolution of the cutting resistance into their various components when twist drilling, is shown in Fig. 49 at a mid-point along the lips. The thrust force is just

one of the cutting resistances in a drilling operation, contributions to drill resistance are from the: • Lips – equal lip lengths and angles are important for a ‘balanced cutting action’ , this being considered an efficient cutting process, • Chisel edge – is highly negatively skewed and as it acts like a ‘blunt wedge-shaped indentor’ , extruding the workpiece material from this vicinity,

. Figure 49. The balanced cutting forces resulting from drilling holes utilising twist drill geometries. [After: Kaczmarek, 1976]


99

• Land, or margin – via a rubbing, or frictional ac- A = 2Az = szd = sd/2 = 2hzb tion.

NB The latter two are relatively inefficient processes, moreover, the resistance components of the lips and chisel edge are the product of resistance of the undeformed chip to plastic strain, in combination with resistance due to external friction. The land resistance occurs from the friction (i.e. rubbing) against the side of the drill’s hole. When symmetrical twist drilling (illustrated in Fig. 50), the undeformed chip can be characterised by its: • Cutting depth (a) – where ‘a’ = d/2, with ‘d’ being the drill’s diameter (mm), • Feedrate (s) – this being the distance the cutting edge moves in the drilling axis direction during one revolution. Normally two rigidly joined cutting edges are cutting at any instant, each one in its travel corresponds to feed ‘sz’ , which removes an undeformed chip whose size – in the direction of the drilling axis is: ‘s1’ and ‘s2’ respectively (i.e see Fig. 50a) and, as most drills are symmetrical in design, then: s = s1 = s2 = s/2

(mm),

• Undeformed chip thickness (hz) – to be removed by

each of the drill’s cutting lips, which can be determined from the following relationship:

hz = sz sinθ

(mm).

NB With a symmetrical drill, then: Hz = h1 = h2.

• Undeformed chip thickness (b) – can be found from the following relationship:

b = a 1/sinθ = d/2sinθ

(mm).

∴ It follows from these expressions, that the transverse cross-sectional area of the undeformed chip at each of the twist drill’s cutting lips, can be shown by the following relationship: Az = szd/2 = sd/4 = hzb

(mm2).

Hence, the total transverse cross-sectional area when drilling of the undeformed chip will be:

(mm2).

Conversely, in the case of ‘Pilot’ hole drilling (Fig. 50b), the undeformed chip elements are identical to ‘Solid’ drilling, but for the exception of the DOC, which can be expressed in the following manner: a = d-do/2

(mm).

Where: d = diameter of final hole (mm), do = diameter of primary hole (mm). Thus, for example in the case of ‘Pilot’ hole drilling, the total cross-sectional area of the undeformed chip, will be: A = 2Az = sz(d – do) = s(d – do)/2 = 2hzb

(mm2).

The calculation of cutting forces in ‘Solid’ hole drilling (Fig. 50a), can be found from the general formulae for axial force (F) and torque (M), in the following manner, respectively: F = CF d bF suF KH M = CM d bM suM KH

(kg) (kg mm)

Where: CF12 and CM13 = constants (i.e. derived from Kaczmarek‘s findings), d = nominal drill diameter (mm), bF and bM = exponents characterising the influence of the drill diameter, s = feed rate (mm rev–1), uF and uM = exponents characterising influence of feedrate, = workpiece material’s correction coKH efficient (i.e. concerning mechanical properties).

12 CF is derived from experimental data, typically: Carbon steel (construction) 84.7, Grey CI 60.5, Malleable CI 52.5, Bronze (medium hardness) 31.5 – with HSS drills, ranging from φ10 to 60 mm. 13 CM is derived from experimental data, typically: Carbon steel (construction) 33.8, Grey CI 23.3, Malleable CI 20.3, Bronze (medium hardness) 12.2 – with HSS drills, ranging from φ10 to 60 mm.

100

Chapter 3

. Figure 50. Drilling a hole with/without a ‘pilot’ hole and the cutting, rubbing and extrusion mechanism. [After: Kaczmarek, 1976]


101

Conversely, for ‘Pilot’ hole drilling (Fig. 50b), these mathematical formulae are modified in the following manner: F = CF d bF aeF suF KH M = CM d bM aeM suM KH

(kg) (kg mm)

Where: a = DOC (mm), eF and eM = exponents indicating the DOC’s influence. These axial force and torque formulae derived in the work by Kaczmarek, are concerned with ‘so-called’ average twist drilling values. These ‘averages’ are related to drill diameters between 15 to 35 mm, having feed ranges in the vicinity of 0.2 to 0.4 mm rev–1. Therefore, the entire axial force (F) and torque (M), comprises of contributions of the lips, land and chisel point, in the following manner: • Axial force (F) – lips (50%), land (10%) and chisel point (40%), • Torque (M) – lips (80%), land (12%) and chisel point (8%). NB These contributing factors to axial force and torque are for drill depths that do not exceed 2.5d. If the drilling force is significantly increased, then this has the effect of distorting the drill shaft. Such distortion, causes the drill’s cutting edge to move forward into the workpiece material, in this manner it jointly increases the DOC and the drilling force. Correspondingly, if the drilling force is reduced, the twist drill will recover its shape, with the cutting edge moving back from the workpiece, thus reducing both the DOC and cutting force. This stretching and compression of the drill’s shaft – somewhat like a spring14 – is unique to twist drilling, being an unstable element in the cutting process. By way of comparison, most cutting tool

14 ‘Lengthening effect’ is associated with the twist drill’s shaft being twisted by the application of torque, with elastically springs-back upon release of the drilling torque. Not only will the twist drill ‘spring’ , but it can also ‘bend’ due to the increased thrust loads produced by high penetration rates.

edges are normally deflected away from increases in the load. A common form of failure of twist drills in operation is from shattering, with such catastrophic failure being related to the dynamic nature of twist drilling. By way of illustration, a φ4.5 mm long-series twist drill is capable of withstanding a torque of approximately 6 Nm before it catastrophically fails. Normally, the torque for most drilling operations is around 1 Nm.

Temperatures in Twist Drilling The accumulation of heat in the vicinity of cutting is an important factor in the cutting process, with much of the mechanical energy necessary for machining being converted into heat, then conducted into the chip, workpiece and tool (Fig. 51). The consequential thermal phenomena are important, as they can affect the: • Mode of deformation – elastic/plastic behaviour of the chip, • Machined surface – for metals the ultimate metallurgical state of the material, • Tool wear rate – which depends upon a number of criteria, such as the tool’s coating, cutting data employed, work-hardening ability of the workpiece and coolant delivery and its efficiency. It is imperative to comprehend the factors that control both heat generation and its dissipation, together with the tool and work’s temperature distribution in and near the cutting zone. A drilling operation can be considered as a complex machining process, with specific and unique characteristics, not least of which, are the production of chips when drilling. These chips are in continuous contact with the drill flutes and the generated hole’s surface. Hence, any minute changes in the drill’s geometry, can cause enormous modifications to the either the drill’s wear rate and its predicted life. Heat generated whilst drilling will be transformed by a range of ‘states’ , including: • Conduction – through the chips, workpiece and drill, • Convection and radiation – via the ‘air-spaces’ in the hole as the drill penetrates deeper into the workpiece. The drilling temperature during a prolonged operation can approximate steady-state conditions, with the heat generated whilst cutting when employing a new drill

102

Chapter 3

. Figure 51. The drilling process and the asociated zones of heat generation whilst hole-making. [After: Trigger and Chao, 1951]


is associated with two distinct regions (i.e. see Fig. 51 – Section on X-X) at the: • Primary shear zone – where plastic deformation occurs, this being the major source of heat generation, • Secondary shear zone – from within the tool/chip interface, where pronounced friction takes place. NB The drill clearance surface temperature, is significantly affected by the rake face interface temperature. The bulk rise in the drill’s temperature is multifarious, due to the necessity to consider a range of factors, including: heat flow distribution, the geometric shape of the conducting bodies, together with any variation of thermal properties of both the drill and workpiece materials with temperature changes. The generated heat distribution when drilling depends upon the thermal properties of the tool, workpiece and chip. Therefore, the thermal diffusivity (K/ρc), will determine the rate at which heat transfers through the material, while also controlling the penetration depth of the surface temperature. While the absorption coefficient (Kρc), determines the quantity of heat being absorbed by a given mass of material. Drilling temperatures vary considerably in the research work undertaken over the years, being heavily influenced by a wide range of cutting-related parameters, making it extremely difficult to obtain meaningful comparisons of local temperatures in a real-time drilling operation. For example, the scatter of ‘bulk’ temperature values for say, a φ6mm twist drill, can vary between approximately 200 to 380°C, under steady-state drilling conditions15, for comparable workpiece materials, making it very difficult to obtain meaningful drill life comparisons. Coolant delivery is imperative when drilling and to this end, through-the-nose coolant operation enables the lubrication and cooling of the drill’s point (Fig. 52c – illustrating the coolant holes behind the lips). This efficient technique of ensuring that the coolant gets to the action of drilling, gives better chip control, helping

15 Twist drill interface temperatures have been reported to be over 870°C in the workpiece’s ‘plasticity region’ , which somewhat contradicts the ‘bulk’ temperatures, although in mitigation, it should be said that these very high temperatures at the interface at somewhat localised.

103

to reduce machining temperatures significantly and aid drill penetration rates, while increasing tool life. Coolant holes through-the-nose are not restricted to twist drills, as Spade-16 and Gun-drills17, together with Indexable drills also often incorporate this coolant delivery feature, to remove heat and lubricate the cutting edges.

3.1.4 Indexable Drills Indexable drills have some significant advantages over their twist drilling counterparts (i.e. a range of both indexable and twist drills are depicted in Fig.52a). These indexable drills – allowing the cutting inserts to be changed (see Fig. 52c), permit faster cutting speeds and enable a wider range of workpiece materials to be successfully drilled than when utilising conventional twist drills. Normally, indexable drills are limited to shorter hole depths of around ‘4D’ , than equivalent diameter twist drills. Indexable drills must be set up with care and in the correct relationship to the machine tool’s headstock/ spindle, ensuring that both the drill’s and the spindle’s centrelines are coincident, otherwise over-, or undersized holes may be produced (see Fig. 53a – top). Yet another problem that needs to be addressed when employing these indexable drills, is termed ‘radial runout’18, which affects the inserts centre height and should be limited to 5,000 components, to make the cost of the Special-purpose tooling viable. 26 Amortisation, refers to the ‘pay-back’ of the tooling over the ‘life’ of the production of the parts produced.

chamfering and counter-boring. In this case, not only does the usage of special-purpose tooling here seem the obvious solution, as it combines these individual operations in one, it has the advantage of meeting all three of the production criteria listed above, with the added advantages of both using fewer tools and utilising less space in the tool magazine. Some special-purpose tools are very complex in their design and quite sophisticated in operation, but their supplementary cost more than outweighs this by the production gains offered by their consequent implementation. Multi-spindle drilling 29 tooling is ideal when a series of hole patterns are required on a component, such as for specific configurations of: pitch circle diameters, hole grid patterns, line of holes, or a combination of these (i.e. see examples of specific patterns in Fig. 57 – top right). Hole pitch circle diameters can easily be accommodated, for large and small pitch diameters on the same tooling, Likewise, hole line and grid patterns can be quite diverse, within the diametral area of the ‘cluster plate’ (i.e. see Fig. 57 – top left). Multi-spindle drilling heads utilise a main drive gear which is engaged with an idler and then onto the drill spindle gear, this being attached to the individual drill (i.e. see Fig. 57 – exploded view of a typical system). The cluster plate orientates the individual drill spindles and their rotational speeds can be marginally increased, or decreased, by changing the driverto-driven gear ratio, moreover, their rotational direction can be changed by the introduction of another idler into the gear train. Therefore, if additional idlers are present, to change the drill’s rotation, then the an appropriate left-hand drill would be required here. By purposefully modifying each drill’s rotational direction, this has the advantage of minimising overall torque effects on the multi-spindle drilling head, allowing a large number of drills to be utilised for one particular operation (i.e. see Fig. 57 – lower left-hand photograph). An important point in utilising multispindle drilling heads, is presetting their respective drill lengths, so that they engage with the workpiece’s surface at the correct height. By the correct production application of both Special-purpose and Multi-spindle drilling tooling, then

27 Short cycle times, are considered to be the quickest time that the part can be produced, under ‘standard’ machining conditions. 28 Datum – the term refers to origin of the measurements for the particular component feature, which could be from a face, plane, or point.

29 Multi-spindle tools, refer to more than one individual tool rotating in its respective toolholder, enabling several holes to be manufactured in just one operation.


. Figure 56. Special-purpose multi-functional tooling can be designed and manufactured to machine many part features simultaneously

111

112

Chapter 3

. Figure 57. The application of multi-spindle drilling heads to increase productive throughput


significant economic savings can be made and their initial capital outlay will have been worthwhile. However, such complex and expensive tooling used inappropriately can be counter-productive, so considerable thought and care should be made into any future implementation of these tools.

3.1.7 Deep-Hole Drilling/ Gun-Drilling Deep-Hole Drilling – an Introduction Deep-hole drilling can be characterised by, high material removal rates, having excellent: hole straightness, dimensional tolerances and machined surface texture. Deep-hole drilling applications are utilised across diverse industrial applications, including: aerospace, nuclear power, oil and gas, as well as for steel and chemical processing industries. These industries place a high demand on all aspects of drilled hole quality and reliability, with components being very expensive, any failures will have severe economic consequences. The name Deep-hole drilling implies the machining of holes with a relatively long hole depth to its diameter. Typically at the lower length-to-diameter ratios they can be as short as x5 the diameter, conversely at the other end of the scale, ‘ratios’ of > x100 the diameter can be successfully generated, with close tolerances and a surface texture approaching 0.1 µm (Ra). There are a considerable number of deep-drilling production techniques, with each one having an appropriate usage for a particular hole generation method. A typical deep-drilling tooling assembly is essentially ‘self-piloting’ , in that the cutting forces generated are balanced, not with respect to the cutting edges – as is the situation with Twist drills – but invariably, by pads that are situated at 90° and 180° to that of the cutting edge. These pads rub against the bore’s surface being generated and therefore support the head, while burnishing30 the surface. The machine tools enabling these deep-drilled holes to be generated can be expensive, along with the appropriate tooling, but the production costs can be dramatically reduced, by employing such a machining strategy. One of the

30 Burnishing will improve the surface finish and dimensional accuracy, by plastically deforming the machined surface layers – cusps – without removing any additional workpiece material.

113

major problems of utilising Deep-hole drilling, is chip disposal, as the deeper the hole is drilled, the further the chip must travel from the cutting edge to the hole’s exit. This chip evacuation distance, can increase the probability of chip-jamming, or binding in the flute as the chip attempts to exit the deep-drilled hole. Notwithstanding the problems associated with BUE, which hinders the tool’s ability to break chips. Coolant control and its operational usage is important in any the Deep-hole drilling technique, as one of its main functions is – apart from lubricating/cooling the cutting edge and chip flushing – is to restrict frictional effects between the: drill, chip and hole wall. Moreover, if friction builds-up due to poor coolant delivery, this can result in higher torsional effects, which may cause the drill to snap.

Gun-Drills Gun-drills (i.e. see Fig. 58a), are normally utilised to machine small, straight diameters to high tolerances and having excellent finishes in a single operation. Drilled hole sizes can range from as small as φ1.5 mm to φ75 mm in a single pass, with depths equating to 100 times the tool’s diameter31. The ‘drilling system’ is a highly developed and efficient technique for producing deep holes in wide variety of workpiece materials, ranging from: plastics, fibreglass, to high-strength materials such as Inconel. This tooling usually consists of either a cemented carbide, or cemented carbide-tipped drill head fitted to a tube-shaped shank32. The former solid carbide drill head version allows the tooling to be reground as necessary, while the latter version is normally employed for larger diameter hole drilling operations. The drill head has two distinct designs, either having a ‘kidney-shaped’ , or a cylindrical hole present, for the delivery of cutting fluid, which provides: • Flow of cutting fluid – to create the maximum flow rate and chip-flushing,

31 ‘Special-purpose’ Gun-drills can be produced to generate drilled holes up to φ150 mm having 200:1 length-to-diameter ratios, at penetration rates of better than equivalent diameter Twist drills. 32 Gun-drills would as a rule, have their drill head’s brazed – via silver soldering – onto a tube-shaped shank, these in turn, are also brazed onto a ‘driver’ (i.e. of various designs) of the required length for the successful drilling of long slender holes in the workpiece.

114

Chapter 3

. Figure 58. Deep-hole drilling operations, such as: (a) gundrilling, (b) double tube ejector drilling and (c) single tube ejector drilling. [Courtesy of Sandvik Coromant]


115

• Minimal fluid-flow disturbance – giving consis- ure mode. Yet another Gun-drill failure situation may tent/regular flow-rate to drill-head,

arise if there is excessive clearance between the drill

be easily evacuated from the cutting region.

Gun-drill’s edge cuts a significant volume of workpiece material and, as this edge is not designed to cut – having a zero clearance angle (i.e. created by the circular margin at this edge) – the excessive cutting forces cause the edge to prematurely fracture. If insufficient coolant flow occurs, this is also a typical factor in subsequent Gun-drill failure. This lack of coolant causes the chips to pack in the V-flute, forming a plug, which then creates excessive torque in the Gundrill and, this plug allows the tip to separate away from the shank. Occasionally, end-users blame the Gun-drill tooling manufacturer for poor brazing, if the tool’s tip separates from the shank. However, when analysis of the brazed fractured surfaces occurs, invariably, small carbide particles are adhered to the shank, this being evidence of the fact that the braze was stronger than the tip, clearly demonstrating that the brazing was not at fault. In many circumstances, the Gun-drill tool manufacturer is blamed by the customer for its failure during machining, but when investigated, it is usually premature failure being the result of a poor tooling installation and operation. One of the major causes of Gun-drill failure, is via the coolant distribution system, where inconsistent delivery of the fluid can either ‘starve’ the Gun-drill’s cutting edge, or ‘over-flood’ the system. One of the major factors contributing to this over-/under-supply of coolant delivery, is due to the fact that in the main, coolant pressure is being monitored, rather than the measurement of coolant flowrate. If holes are Gun-drilled 0.02 mm, so any subsequent drill bush wear needs to be carefully monitored and controlled. It is usual practice to have a ro-

35 Asymmetric Drill Head design, refers to the fact that the cutting inserts are not only radially, but are angularly offset. Therefore, they normally require two support pads to counteract and sustain the radial cutting forces generated while drilling deep holes. By locating the cutting inserts on both sides of the drill head, the greater percentage of radial forces are negated at these pads. 36 Drill bushing tolerances between the drill and bush for both the ‘Ejector’ and Single-tube Systems, require a fit of ISO G6/ h6, equating to a minimum play of 0.006 mm. This drill bush is usually manufactured from a hardened material (i.e. 60 to 62 HRC) such as cemented carbide, as it has a longer service life, with bush wear normally limited to 0.03 mm.

tating workpiece and a stationary tool, with any centre divergence resulting in bell-mouthing at the hole’s entrance and a wavy hole surface. Once the support pads in the drill head have moved x5 their length down the drilled hole, then any further waviness is negligible, as they begin to press down on the hole’s curvature. Many deep-drilled hole profile and tolerance abnormalities result from centre divergence, which needs special attention to minimise such effects. Single-tube [Ejector] System drills (i.e. commonly referred to and abbreviated as simply ‘SST’) are schematically depicted in Fig. 58c. With this SST tooling assembly, the cutting fluid is pumped under pressure between the drill and the hole wall (i.e. normally this width of space is approximately 1 mm) and it exits with chips through the inside of the drill tube (Fig. 58c). The quantity of cutting fluid passing through the drill is twice as great and with higher pressure, than for an equivalent ‘Ejector’ tooling assembly. Hence, the SST set-up provides improved chip-breaking and minimises any potential chip-jamming, even when varying chip lengths occur. The drill head arrangement of cutting inserts will vary from two, three, or more, depending on the drill’s diameter, usually made of cemented carbide, often as brazed over-lapping tips, although disposable indexable pocketed inserts with chip-breakers are often utilised for larger diameter holes. SST tools can be used to drill small diameter holes, ranging from φ12.5 mm upward, with 100:1 depth-to-diameter ratios. The SST tooling system copes with difficult-to-machine workpiece materials, such as Monel, Inconel and Hastelloy and other ‘exotic materials’. In actual production machining trials, it has been found that SST tools can produce deep-drilled holes up to 15 times faster than is achievable by conventional Gun-drilling. This high production output level gives an 80% improvement in machining rates for this SST Deep-drilled hole production output and, it has been shown in several instances, to give a ‘Return on Investment’ (ROI)37 in about 6 months.

37 Return on Investment (ROI), for Deep-hole drilling operations (i.e. in % terms), is given (i.e. in simplistic terms) by the following formula: Cost of a -to- productivity gain % ROI = Total conversion cost


3.1.9 Deep-Hole Drilling – Cutting Forces and Power In Deep-hole drilling operations, the underlying theory for the calculation of cutting forces and for torque are similar to that utilised for ‘conventional’ drilling operations. The major difference between the hole production calculations for Deep-hole drilling to that of ‘conventional hole-making’ techniques, lies in the fact that support pads create a sizeable level of frictional forces, that cannot be ignored. These increased frictional effect contributions – by the pads – to the overall Deep-hole drilling cutting forces and torque values are somewhat difficult to precisely establish, however, an approximate formulae can be used to estimate them, as follows: Feed force (N): Fp + Fpµ = 0.65 × kc × ap × f × sinκr Where: = Feed force, or drilling pressure (N), Fp = Force and Frictional effects (N), Fpµ kc = Specific cutting force (N mm–1), = Depth of cut (mm), ap f = Feed per revolution (mm rev–1), sinκr = Entering angle (°). Torque, or Moment (Nm): Mc + M µ =

Where: M c M µ kc ap f D

kc � a p � f � D (. − ap �D) 

= Torque cutting (Nm), = Torque and Frictional effects (Nm), = Specific cutting force (N mm–1), = Depth of cut (mm), = Feed per revolution (mm rev–1), = Hole diameter (mm).

Relatively high speeds are utilised for Deep-hole Drilling operations, in order to achieve satisfactory chipbreaking, this necessitates having a machine tool with a reasonable power availability. The underpinning theory for calculating the power requirements, corresponds with that of ‘conventional’ drilling operations. However, the friction forces that are present, due to the employment of support pads, gives rise to a torque contribution (Mµ), which in turn produces an associated contribution ‘Pµ’ to the total Deep-

117

hole drilling power. Therefore, in order to estimate the machine tool’s power requirement (i.e. ‘P’ in kW ), an allowance must be made for any power losses in the machine tool. Hence, the gross power required can be established by dividing the Deep-hole drilling power (i.e. Pc + Pµ), by the machine tool’s efficiency ‘η’. This efficiency indicates what percentage of the power supplied by the machine tool, that can be utilised, while Deep-hole drilling. Power (kW): (Pc + Pµ ) =

kc � a p � f � vc (. − ap �D) , 

Where: Pc + Pµ = Power contributions of: cutting and friction respectively (kW), = Cutting speed (m min–1). vc ∴P

= Pc + Pµ/η

Where: η = Machine tool efficiency.

3.2 Boring Tool Technology – Introduction The technology of boring has shown some important advances in recent years, from advanced chip-breaking control tooling (i.e. see Fig. 59, this photograph illustrates just some of the boring cutting insert geometries that can be utilised), through to the ‘active suppression of chatter’ 38 – more will be mentioned on the topic and reasons why chatter occurs and its suppression later in the text. Probably the most popular type of boring tooling is of the cantilever type (Fig. 59), although the popularity of either ‘twin-bore-’ , or

38 ‘Chatter’ , is one of the two basic types of vibration (i.e. namely, ‘forced’ and ‘self-excited’) that may be present during machining. In the main, chatter is a form of self-excitation vibration.‘[It is]… due to the interaction of the dynamics of the chip-removal process and the structural dynamics of the machine tool. The excited vibrations are usually very high in amplitude and cause damage to the machine tool, as well as lead to premature tool failure’. [After: Kalpakjian, 1984].

118

Chapter 3

‘tri-bore-heads’ , with ‘micro-bore adjustment’ of the either the individual inserts, or having a simultaneous adjustment of all of the actual cutting inserts, is becoming quite common of late. Boring operations invariably utilise cantilevered (i.e. overhung) tooling, these in turn are somewhat less rigid than tooling used for turning operations. Boring, in a similar manner to Deep-hole drilling and Gun-drilling operations, has its rigidity decreased by the ‘cube’ of the distance (i.e. its overhang), as the following equation predicts: � π  � EI fo =  L (Mt + .Mb )

Where: fo = normal force acting on the ‘free end’ of the cantilever (i.e boring tool overhang), *EI = flexural stiffness (i.e. I = cross-sectional moment of Inertia) (Nm2), Mt = boring bar mass (kg), L = length of cantilever (mm), Mb = Modulus of elasticity of the boring bar (N mm–2). * E, relates to the boring bar’s ‘Young’s modulus’.

Boring a hole will achieve several distinct production criteria: • Enlargement of holes – a boring operation can enlarge either a single, or multiple series of diameters, to be either concentric to its outside diameter (i.e. O.D.), or machined eccentric39 (i.e. offset) to the O.D., • Correction of hole abnormalities40 – the boring process does not follow the previously produced

39 ‘Eccentric machining’ of the bore of a component with respect to its O.D., was in the past accurately achieved by ‘Button-boring’ – using ‘Toolmaker’s buttons’ (i.e. accurately ground and hardened buttons of ‘known diameter’) that were precisely offset using gauge blocks (i.e ‘Slip-gauges’). This technique might still be employed in some Toolrooms, but normally today, on CNC-controlled slideways, a simple ‘CNC offset’ will achieve the desired amount of bored eccentricity. 40 Correction of hole abnormalities, as Fig. 60 schematically illustrates, how boring can correct for ‘helical wandering’ of the drill as it had previously progressed through the workpiece. The drill’s helical progression would cause undesirable hole eccentricity, resulting from minute variations in its geometry,

hole’s contour, but generates its own path and will therefore eliminate drill-induced hole errors by the subsequent machining operation (i.e. see the schematic representation shown in Fig. 60), • Improvement of surface texture – the boring tool can impart a high quality machined surface texture to the enlarged bored hole. NB In this latter case, boring operations to previously drilled, or to any cored holes in castings, can be adjusted to give exactly the desired machined surface texture to the final hole’s dimensions, by careful adjustment of the tool’s feedrate and the selection of an appropriate boring tool cutting insert geometry.

3.2.1 Single-Point Boring Tooling ‘Traditional’ boring bars were manufactured as solid one-piece tools, where the cutting edge was ground to the desired geometry by the skilled setter/operator, which meant that their useful life was to some extent restricted. Later boring bar versions, utilised indexable cutting inserts, or replaceable heads (Fig. 61). Boring bars having replaceable heads are versatile, with the same bar allowing different cutting head designs and cutting inserts (Fig. 61a). Here, the insert is rigidly clamped to the tool post, with replaceable ‘modular tooling’ heads with the necessary mechanical coupling to be utilised (i.e. Fig 61b), offering ‘qualified tooling’ 41 dimensions.

necessitating correction by a boring operation. This ‘correction’ is necessary, because the drill’s centreline follows the path indicated, ‘visiting’ the four quadrant points as it spirally progresses through the part. Hence, hole eccentricity along with harmonic departures from roundness can be excessive, if the drill’s lip lengths and drill point angles are off-centre. The cross-hatched circular regions represent the excess stock material to be removed by the boring bar, where it corrects these hole form errors, while machined surface texture is also considerably improved. 41 ‘Qualified Tooling’ , refers to setting the tool’s offsets, with all the known dimensional data for that tool, allowing for ease of tool presetting and efficient tool-changing – more will be said on this subject later in the text.


. Figure 59. A selection of some tooling that can be employed for boring-out internal rotational features. [Courtesy of Seco Tools]

119

120

Chapter 3

In the case of the boring bar’s mechanical interface (i.e. coupling) example shown in Fig. 61a- top, the serrated V-grooves across the interface along with the

four clamping screws provide an accurate and secure fitment for the replaceable head, with internal tension adjustment via the interior mechanism illustrated.

. Figure 60. The harmonic and geometric corrections by a boring operation, to correct the previous helical drift, resulting from the drill’s path through the workpiece


. Figure 61. Interchangeable cutting heads for boring bars utilised in machining internal features. [Courtesy of Sandvik Coromant]

121

122

Chapter 3

Possibly a more adaptable modular system to the ‘serrated and clamped’ version, is illustrated in Fig. 61b, where the cutting head is held in place by a single rearmounted bolt and grub screws around the periphery of the clamped portion of the boring bar securely lock the replaceable head in-situ, enabling the cutting head to be speedily replaced. Some of these boring bar’s have a dovetail slide mechanical interface, with the dovetail coupling providing radial adjustment of the cutting insert’s edge. This ‘universal system’ (Fig. 61b), is normally used for larger bored diameters, that would range from 80 to 300 mm. Furthermore, it is possible to add spacers/shims to precisely control the boring bars overall length, this is particularly important when medium-to-long production batches are necessary, in order to minimise cycle time and its non-productive setting-up times. In Fig. 62a and b, are illustrated single-point interchangeable boring insert tooling, with Fig. 62a giving typical length-to-diameter (i.e. L/D) ratios for actual boring and clamping lengths. The amount of boring bar-overhang will determine from what type of material the boring bar will be manufactured. The most common tool shank materials are alloy steel, or cemented carbide, for L/D ratios of 31 kN, while fitting into the taper with a gauge-line of just 30 mm. The balllock mechanism used two balls that locked into the machined holes through the taper shank of the cutting unit (Fig. 120 and 121). This lock-up configuration, allows either a φ9 mm draw-rod, or disk-springs to be used to apply the necessary pull-back force. The holes in the tapered shank – into which the balls are seated, have a machined angle of 55°, this results in a mechanical advantage of 3.5 : 1. As the disk-springs – used in this method – are pulled back, it forces the two balls radially outward until they lock into the tapered machined holes, as depicted in Fig. 122 – where an Allen key T-bar is used to activate the lock-up sequence, via a series of back-to-back disk-spring washers. To release the cutting unit’s front-end, a force is applied by the T-bar, which pushes these disk-springs and releases the balls, while at the same time it ‘bumps’ the

Modular Tooling and Tool Management

225

. Figure 121. ‘Modular tooling concepts’ allow ‘qualified tooling’ to be set up with the minimum of adjustments, thereby significantly reducing downtime. [Courtesy of Kennametal Hertel]

. Figure 122. ‘KM’ modular quickchange tooling system being manually-fitted/changed – using the T-bar wrench, into a turning centre’s turret. [Courtesy of Kennametal Hertel]

cutting head and in so doing, releases it from its selfholding taper. Referring to the lock-up sequence once more. Once the cutting unit is inserted into the female taper (i.e. back-end), it makes contact at a stand-off distance of 0.25 mm from the face. Therefore, as the locking force is applied, a small amount of elastic deformation occurs at the front of the female taper. As the cutting tool is locked-up, there is a three-point contact that takes place: at the face, the gauge-line and at the rear of the taper. Finally, if one compares the coupling’s stiffness with that of a solid-piece unit which has been machined to identical dimensions, then when a 12 kN is applied – to simulate tangential cutting loads – the

difference in deflection between them, would be only 5 µm. Hence, this modular coupling tooling assembly, closes approximates to that of the ultimate rigidity found if a solid-piece cutting tool was utilised.

Tooling Requirements for Machining Centres Machining centres with their in-situ automatic load/ unload tool-changers and tool-storage carousels, or magazines, have reduced cut-to-cut times significantly, allowing faster response times to the next machining requirement of the CNC program. If a tooling-appraisal is made of the tool-storage facility of machining centres, it would soon be apparent that less-than-total

226

Chapter 6

capacity occurs. This noticeable under-storage tooling capacity may be due to one, or more of the following reasons: • Heavy tooling requirement in the tool-storage system – because of the tool storage system’s configuration – such as a chain-type magazine (Fig. 115) – tools have to be widely-spaced to allow the magazine to be kept evenly-balanced, • Large tools situated in the magazine – this normally requires that the adjacent pockets must be left empty, so avoiding them fouling each other upon magazine rotation (Fig. 115), • ‘Sister-tooling’ requirement – this allows for duplication of the most-commonly-used tools, as they are more susceptible to breakage, or wear, enabling longer overall machining time for the production run, prior to a complete tool changeover. NB This latter point of employing a ‘sister tool’ strategy, has the effect of significantly reducing the variety of tools that can be held in the finite amount of pocket-space available on many magazines, carousels, etc. In order to increase the capacity of a tool-storage system, while simultaneously expanding the range of tools that are available during a production run, modular tooling has been developed which further extends the machine’s capability and versatility. With today’s modular tooling all being of a ‘qualified size’10, they can be prepared from a centralised preparation/ storage facility, then transported to the machine tool automatically – more will be said concerning this level of sophisticated tool management toward the end of the chapter. So far, the relative merits of utilising a modular quick-change tooling system for machining centres has been discussed. Today, such systems can be used for both rotary and stationary tooling operations on machined workpieces. A ‘tooling exemplar’ , of such

10 ‘Qualified tooling’ , this refers to all of the tool’s offsets being known – this allows the tool to be fitted into its respective pocket in the tool storage facility, with the tool offset table updated, allowing the tools to be utilised, without the need for presetting on the machine tool, prior to use. NB Previously mentioned with regard to Boring operations in: Chapter 3, footnote 41.

tools, is the ‘Capto system’11, being an amalgamation of a self-holding taper and a three-lobed polygon (i.e. see Figs. 123 to 125). This novel tooling mechanical interface design, features a tapered polygon, which is an extremely difficult geometric shape to manufacture for both male and female couplings (Fig. 123-bottom left). However, this tapered polygon offers multiple-point contact in a robust and precision coupling, allowing high torques to be absorbed for both rotating and stationary tooling (Fig. 124). Complete ‘Capto’ systems – ranging in their available diameters – are presented for a variety of machine tool configurations, which are obtainable with a wide variety of ‘back-ends’ to suit many differing tool pocket styles (i.e. see Fig. 125 – e.g. ISO, VDI, ANSI, etc.). In order to enhance the use of say, the ‘KM-type’ of modular tooling still further and to ensure that a positive location between mating faces occurs, it is possible to utilise an electronically-activated backpressure device, coupled to the CNC controller. With this system in-situ, the tool-locking procedure, could be as follows: 1. ‘Old tool’ is removed from ‘front-end’ – this occurs by either activation of the tool-changer (i.e. on a machining centre), or a tool-transfer mechanism (i.e. on a turning centre), 2. Compressed air purges the female taper – this has the effect of cleaning-out the debris – fines12 – from the previous tool’s cutting operation, 3. ‘New tool’ is inserted into ‘back-end’ of toolholder – its male taper is cleaned, then it begins to seat itself in the female taper, 4. As it is pushed firmly home to register with its opposing taper – the back-pressure is electronically monitored and, a signal indicates that seating has taken place and this data is sent to the CNC controller, confirming coupling has been firmly locked,

11 ‘Capto system’ , was developed by a leading tooling company, its name is derived from the Italian word for: ‘I hold firmly’ – which seems somewhat appropriate for an excellent mechanical interface between the ‘front- and back-ends’ on a modular tooling system. 12 ‘Fines’ , are either minute particles resulting from the tool ‘recutting effect’ – in the form of small slivers of material, or is the result of dust/debris created when brittle-type material in particular, has been machined and these particles may electro-statically attach themselves to the machined mechanical interface coupling’s mating surfaces.


227

. Figure 123. Modular tooling ‘capto’ with tool security and precision location via face and lobed taper contact. [Courtesy of Seco Tools]

5. Tool is ready for use – this unmanned operation allows the next turning, or machining operation to commence. NB Quick-change tooling of this level of sophistication needs to be coupled to some form of tooltransfer mechanism, in order to gain the full benefits of its potential range of machining applications

and speed of operation, to minimise the pay-back period. The spindle nose taper fitment is an important factor in obtaining the necessary accuracy from modular quick-change tooling (Fig. 126a). Here, the ‘spindle cone’ must run true to the spindle’s Z-axis and the pull-stud pressure should be checked to ensure that it

228

Chapter 6

is within the machine tool manufacturer’s guidelines. Often when problems occur at the spindle taper, it is the result of several factors: • Pull stud pressure variation – this should be checked to ensure that it is within manufacturer’s specification,

• Spindle nose drift – this is the result of perhaps

running the spindle at continuously high rotational speeds, resulting in the spindle nose cone ‘thermally-growing’ , leading to the simultaneous: X-, Yand Z-axes drifting several micrometres (e.g. this thermal drifting can often account for around 10 µm . Figure 124. Modular tooling (Capto) illustrating stationary (turning) and rotational tooling (milling, drilling, etc.), with indentical lobed and tapered ‘back-ends’. [Courtesy of Sandvik Coromant]


of compound angular ‘spindle cone’ movement), which could present a problem for any close tolerance component features requiring machining. NB When these problems occur, the whole cutting tool assembly, can become ‘unbalanced’ , this is particularly true for high cutter rotational speeds.

229

Much more could be said concerning tool-changing techniques, where tool transfer arms are discarded in favour of the whole magazine, or tooling carousel being moved to the spindle to speed-up tool-changing even further. Alternatively, gantry-type tool/ work delivery systems are available, or complete turrets previously equipped with ‘qualified’ tooling can be delivered, for un-manned operations, in a ‘lights-

. Figure 125. The vast range of modular (capto) tooling available for: • machining centres, • turning centres and • mill/turn centres. [Courtesy of Sandvik Coromant]

230

Chapter 6

out’13 environment. The techniques for tool delivery to keep machine tools in operation virtually continuously is a vast topic, which goes way beyond the current scope of this existing tooling-up discussion. All of these rotational modular quick-change tools can be successfully utilised up to speeds of approximately 12,000 rev min–1, without any undue problems. However, once rotational tooling speeds increase above this rotational level, then invariably it is necessary to redesign the tool assemblies, allowing them to be dynamically balanced, this will be the theme of the next section.

6.4 Balanced Modular Tooling – for High Rotational Speeds When rotational spindle speeds are very high, the conventional ball-bearing spindles are limited and have an upper velocity of ≤80 m sec–1, this is where the balls lose contact with the journal walls and begin to promote ‘Brinelling’14 within the raceways. It is not usually the case, for a conventional milling spindle to be utilised at rotational speeds >20,000 rev min–1, without due regard for the: centrifugal force, frictional effects and spindle cone roundness levelling variations, that are likely to be present beyond these speeds. For any dynamic unbalance15 of the tooling assembly to occur, this will happen, if the mechanical interface is not se-

13 ‘Lights-out’ machining environments, refer to either completely un-manned machining, or minimal-manning levels. Some companies, run an fully-automated machining ‘nightshift’ without any personnel in attendance, allowing the lights to be turned out, thereby saving significant electrical power cost, when this factor is taken over the year’s usage. 14 ‘Brinelling’ , creates break-down and delamination of the raceways as the ‘unrestrained’ hardened balls strike both the internal and external races at high speeds, causing them to prematurely and catastrophically fail in-service. * Brinell hardness – uses a ø10 mm steel ball – hence the name. 15 ‘Dynamic unbalance’ , can occur in either of the two tooling planes, these are either radial, or axial movement, related to the high rotational speeds of the cutter assembly. In many cases, dynamic dual-plane balance can be achieved, using specialised tool assembly balancing equipment (i.e. see the chapter concerning high-speed milling applications).

cure – more will be said on this subject in the chapter describing high-speed milling operations. With balanced tooling in mind, cutting tool assemblies were developed that minimised rotational unbalance, being based upon the HSK taper fitment, shown in Figs. 126b and 127. The most important advantages of this exemplary mechanical interface with its tapered hollow shank, coupled to its axial-plane clamping mechanism (i.e. based upon: HSK-DIN 69893), is as follows: • High static and dynamic rigidity – the axial and radial forces generated in the tool shank, provide the necessary clamping force, • High torque transmission and defined radial positioning – the ‘wedging effect’ between the hollow taper shank and holder/spindle, causes friction contact over the full taper surface and the face (Figs. 127ci and cii). Two keys engage with the shank end of the toolholder, providing a ‘form-closed radial positioning’: thereby excluding any possibility of setting errors, • High tool-changing accuracy and repeatability – the circular form engagement of the clamping claws within the hollow tool shank, provides an extremely tight connection between the shank and holder/spindle (Fig. 127cii), • High-speed machining performance – improves in both locking/clamping power and effectiveness with increased rotational speed. The direct initial stress between the hollow shank and the spindle holder, compensates for the generated spindle expansion promoted by centrifugal force and, in so doing, negates any radial play. The face contact clamping, prevents any slippage in the axial direction (Fig. 127cii), • Short tool changing times – due to much lighter tooling, when compared to a conventional ISO taper: the shank is about 1/3 of its length and, approximately 50% lighter in weight, • Insensitive to ingress of foreign matter – the uninterrupted design of the ring-shaped axial plane clamping mechanism, simplifies coupling cleaning. During an automatic tool-change, compressed air purges mating surfaces and provides cleaning at the interface, • Coolant through-feed – via centralised coolant feed by means of a duct, which also excludes ingress of coolant, as the front- and back-ends are entirely sealed – preventing fouling of the mechanical interface, • Tool shank construction is both simple and economic to produce – as no moving parts are present,


. Figure 126. Milling cutter toolholder taper fitment. [Courtesy of Sumitomo Electric Hardmetal Ltd.]

231

232

Chapter 6

. Figure 127. HSK high-speed modular tooling, for machining applications on turning/machining centres. [Courtesy of Guhring]


thus significantly minimising any potential surface wear. These major tooling advantages for the HSK-type tooling design, has shown a wide adoption by companies involved in high-speed machining applications, throughout the world today. In the following section, a case is made for tool-presetting both ‘on’ and ‘off ’ the machine tool, with some of the important tooling factors that need to be addressed. The problems associated with tool-kitting and the area for undertaking such activities will be discussed, in order to ensure that the tools are efficiently and correctly assembled, then delivered to the right machine tool and at the exact time required – this is the essence of successful ‘Tool management’.

6.5 Tool Management Introduction Manufacturing industries involved in machining operations encompass a wide variety of production processes, covering an extensive field of automation levels. Not only will the cost of investment vary from that of simple ‘stand-alone’ CNC machine tools, to that at the other extreme: a Flexible Manufacturing Systems (FMS), but other factors such as productivity and flexibility play a key role in determining the tooling requirement for a particular production environment (Fig. 128). Each machine tool, operating either in isolation (ie. in a ‘stand-alone’ mode), or as part of a manufacturing cell/system, needs specific tooling (i.e. tool kits) to be delivered at prescribed time intervals. Such tooling demands are normally dictated by the devised sequence of production from some ‘simple’ form of manufacturing requirement, to that of a highly sophisticated computerised ‘Master Production Schedule’ (MPS). With the introduction of CNC machine tools in the late 1970’s, the drive has been towards smaller

233

batch sizes, this has meant that some form of tool management has become of increasing importance in machining operations, in order to keep down-time16 to a minimum. In an USA survey of tooling activities conducted some years ago into manufacturing companies involved in small-to-medium batch production using CNC machine tools, the tooling requirements and scheduling left a lot to be desired, in terms of efficient tool management – verging in some cases, on the chaotic! In Fig. 129 the diagram depicts the typical ‘fire-fighting’ concerned with this lack of tooling availability, highlighting the tool problems that were found. Here (Fig. 129), the diagram illustrates the actual time-loss constituents – in % terms, clearly showing that ‘line-management’ and operators spent considerable time and effort trying to find tools in the machine shop, or were simply looking for tools that did not exist! This chaotic state of affairs, meant that highly-productive machine tools were idle, while this ‘self-defeating activity’ was in progress. With the actual machine tool running costs being so high, this remedial action was somewhat futile and cost these companies considerable financial encumbrance, that would be difficult to estimate – in real terms. Today, some of these problems are still apparent in many machine shops throughout the world, so the tooling problems mentioned here are still valid. Had some form of ‘simple’ tool management system existed within these companies, then many, if not all of these tooling-related problems would have been eliminated. This fact was also confirmed in this tooling survey, by some of the more ‘enlightened’ companies that utilised tool management, either operating at the most elementary level, to that of a highly sophisticated computerised system, that encompassed: total tool control: servicing, presetting, delivery of kits, replenishment of tooling stock levels and monitoring of tooling and its utilisation level within the production operation in the machine shop. It is not unreasonable to assume, that tooling inventories can be vast within a relatively moderately-sized machine shop (i.e. see Fig. 130 as it visually indicates the problem of keeping some form

16 ‘Down-time’ , refers to the non-productive time that occurs when the machine tool is not actually involved in any machining operations. This ‘down-time’ might be the result of a range of individual, or inter-related factors, such as: unexpected machine tool stoppages, changing and adjusting tooling, settingup the fixtures/jigs/pallets, planned maintenance, or tools that are simply not available for the machine tool when they are needed!

234

Chapter 6

of control of the tooling). Not only is keeping track of individual tools and their identification, tool-building, presetting and kitting, together with other toolingrelated problems, becoming an almost impossible task, particularly when this is exacerbated by companies attempting to run a JIT17 philosophy, coupled to that of an MRPII18 production scheduling operation. In the past and, for many ‘traditional’ CNC production environments, any form of ‘tool management’ was generally the province of the machine tool operator. So, alongside each machine would be situated a limited kit of tools, these being maintained and replenished with spares and consumables, via the operator’s liaison with the tool stores. Hence, a skilled setter/operator’s main tooling responsibility was to select the correct tooling, then devise cutting techniques and utilise the appropriate machining data necessary to efficiently cut the parts. This ‘working-situation’ enabled a process planner, or part-programmer to treat the machine tool and operator plus the tool-kit, as a single, ‘self-maintaining system’ – with a well-established performance. Such production circumstances, allowed work to be allocated to specific machine tools, whilst leaving the detailed cutting process definitions: tool offsets, tool pocket allocation, tooling cutting data (i.e. relevant speeds and feeds), coolant application, machining operational sequencing, etc., to that of the operator’s previous skills and knowledge.

17 ‘JIT’ , refers to the manufacturing philosophy of ‘Just-in-time’ , where the system was developed in Japan (Toyota – in the main), to ensure a philosophy and strategy occurred to minimise time and production wastages. The JIT policy has essentially six characteristic elements, these are: (i) Demand call – the entire manufacturing system is ‘led’ , or ‘pulled’ by production demands, (ii) Reduction in set-ups and smaller batches – minimises time-loss constituents and reduces WIP*, (iii) Efficient work flow – thereby high-lighting potential ‘bottlenecks’ in production, *work-in-progress (WIP) levels, (iv) Kanban – this was originally based on a ‘card-system’ for scheduling and prioritising activities, (v) Employee involvement – using their ‘know-how’ to solve the ‘on-line’ production problems, (vi) Visibility – ensuring that all stock within the facility is visible, thereby maintaining ‘active control’. 18 ‘MRPII’ , Manufacturing Resource Planning (i.e. was developed from MRP) – essentially it is a computer-based system for dealing with planning and scheduling activities, together with procedures for purchasing, costs/accounting, inventories, plus planned-maintenance activities and record-keeping.

Today, with the increasing diversity of work that can be undertaken on the latest CNC machine tools, which has occurred as a result of the flexibility of manufacturing in conjunction with reductions in economic batch quantities, this has change the pattern of working. In order to cope with such work diversity, some ‘stand-alone machine tools’ 19 have acquired a very large complement of tools. However, a situation soon develops in which neither the operator, nor the part-programmer is sufficiently in control to accept responsibility for the range of tooling dedicated to any specific machine tool20. So, as a result of a full-deployment of CNC machine tools, the production organisation related to tooling applications, would normally change to one in which: • The production process is defined separately – being remotely situated from the shop floor, • Machining programs and associated tool list are produced – these being sent down to both the machine tool and tool-kitting area via a suitable ‘DNClink’21, with all of the process data and tooling ‘fullydefined’. NB There may be some element of doubt concerning the quality of the tooling definition and even the cutting data produced when the part was originally programmed.

• Batch sizes become smaller – the operator is under

increasing pressure to run the given program without alteration, which leads to ‘conservative cutting’ resulting in less-than-optimum machining, • Machine operator runs the program with the minimum of alteration – this means that the ‘finetuning’ of the operator’s past experiences are not

19 ‘Stand-alone machine tools’ , is a term that refers to highlyproductive CNC machines that are not part of an automated environment, such as either, a flexible manufacturing cell, or system (FMC/S). 20 If the company has not purchased a computer-aided manufacturing (CAM) soft-/hard-ware system, then it will not be in a position to take full advantage of the complex aids for tooling-selection criteria available with many of the more sophisticated CAM systems now currently available. 21 ‘DNC-link’ , is a term that refers to the direct numerical control, this being associated with a shared computer for the distribution of part program data, via data lines to remote CNC machine tools and other CNC equipment in a system.


235

. Figure 128. A comparison of manufacturing systems based upon the following criteria: automation level, productivity and investment costs. [Courtesy of Scharmann Machine Ltd.]

utilised, thereby creating inefficiencies in part cycle times. These factors, make the whole operation critically dependent on the ability of the tool-kitting area to supply and support the part programmer’s specific tooling requirements. This is an unsatisfactory and ineffective tool-management system, with the major problem being that there is no feed-back of experiences gained from machining specific components, which is obviously undesirable. This situation results in the part programmer being oblivious to any problems encountered during component machining, causing a further lack of awareness in the tool-kitting area, producing a critical loss of tool management support.

To minimise the problems associated with the lack of information received by the part programmer and the tool-kitting area, feed-back can be established from the operators, which can be for the whole shop, or for each section of machines. Normally, a centralised system based around an appropriate tool file is essential, this activity in turn, would usually be controlled and managed by a file editor. The tool file can be either a manual-, or computer-based system, but will in general, be accessible to the following personnel: process engineer, part programmer, machine operator, tool stores staff, file editor and management, as necessary – with certain levels of access-codes allowing some form of tooling interrogation (i.e. for security reasons). A typical tool file must contain all the information rel-

236

Chapter 6

. Figure 129. A cutting tool survey of companies in the USA – illustrating the tooling ‘fire-fighting’ solutions on the shop floor. [Courtesy of Kennametal Inc.]


237

. Figure 130. An efficient tool management system is vital if a company is to effectively monitor and control its supply to their production machining facilities. [Courtesy of Sandvik Coromant]

238

Chapter 6

evant to the needs of all the relevant personnel concerning every tool available – more will be said on this topic later.

6.5.1 The Tool Management Infrastructure Whenever a tool management system has been developed, an organised and well-planned tool preparation facility is vital to prepare the specific tooling requirements – off-line, so that tooling might be: • Built to pre-defined assemblies – from a range of standardised stocked parts, or from tool modules, • Replacing worn cutting inserts on used tooling assemblies – these tools being returned for rebuilding, or servicing, • Measuring tool offsets – then, when it is both timely and appropriate, sending tooling in the form of tool-kits to specified machine tools, • Inspecting tooling – normally undertaken on tool pre-setters and by visual means, to ensure that they are fit for immediate use, • Assembling: tooling, fixtures, gauges, etc., as a ‘complete tool-kit’ – to be issued to the appropriate machine tool at the correct time. In order to ensure that consistent and accurate tool preparation occurs, a documented ‘historical procedure’ covering all tooling-related aspects, is necessary, such as: tool inspection, servicing and building, is required for each tool. These factors can be controlled by utilising a computerised tool management system, as only the data files will need to be updated, together with tooling assembly instructions, with both servicing and inspection being undertaken by a step-by-step approach – if needed. Many of the more sophisticated tool management systems currently available, offer a link back to the original Computer-aided Design (CAD) software, allowing tools to be shown graphically assembled as tool parts. As the these tools travel around the machining facility, through various stages of preparation and measurement, then assembled as ‘qualified tool-kits’ visiting machine tools and then travelling back to the tool preparation area for re-servicing, each stage of the tool-kit’s cycle must be controlled. Information concerning the tool kit’s progress, must be available at any instant and, a means of exercising control is to link each tooling station to a central computer

via a DNC-link. As the unique data referring to any tool is stored within the central computer, its identity can be accessed allowing its ‘logistical progress’ 22 to be precisely tracked within the manufacturing facility. For some companies that are unable to justify such a complex tool management method of tooling control, then a much less costly and simpler ‘manual system’ using either printed labels, or bar-codes can be deployed for tool identification when delivering tooling to-and-from the required machine tool. A cautionary note concerning the use of paper labels for tool identification, is that they can more easily become detached during the machining cycle. In an automated machining environment, there is no real alternative but to have a ‘tooling requirement’ and in particular, employing some form of ‘intelligent/ tagged’ tooling, typically via permanent machinereadable tool identification. Such tool identification techniques, allow the necessary data to be interrogated and retrieved from critical areas around the production facility: machine tools, preparation area and storage, plus other peripheral areas – as required. Tooling equipped with ‘intelligent’ memory circuits embedded within them (i.e. typically shown in the case of the non-rotating ‘Block tooling’ in: Figs. 116, 117 and 118), can automatically perform the functions of: tool identification, tool offsets and cutting data up-dating on the machine tool. Other information complementing the tooling data-base pertaining to tool servicing needs can also be exploited by using these ‘tool-coded data chips’ , which are securely situated within the ‘front-end’ of each tool. So that ‘complete tooling control’ is maintained over all the items necessary relating to tool-kits, it is possible to extend stock control over all the tooling requirements out on the shop floor (Fig. 131). Such tool-tracking is important and certain logistical questions must be known, such as: what tooling is where, is it timed to be there now and, what is its present condition, together with other specific questions, which

22 ‘Logistical information and knowledge’ , in any production environment is vital and has been defined (i.e. by the Council of Logistics Management – CLM), in the following manner: Logistics is the process of planning, implementing and controlling the efficient, cost-effective flow and storage of: raw material, in-process inventory, finished goods and related information, from point of origin to point of consumption for the purpose of conforming to customer requirements.’


need to be addressed, indicating the complex task of monitoring all tooling, via a computerised tool management system (Fig. 132). Tool control software enables these physical transactions associated with the: tooling, servicing, kitting, recalibration, etc., to be

239

achieved, without loosing track of any individual tool items. The tooling software will also continuously monitor stock levels, allowing replenishments be actioned, once any itemised tool stock level falls below a certain pres-set value.

. Figure 131. Tooling and fixturing must be precisely controlled at the ‘focal-point’ of kit build-up/replenishment – at the tool preparation area. [Courtesy of Sandvik Coromant]

240

Chapter 6

. Figure 132. Efficient tool management of tool kits around the manufacturing facility, requires some form of ‘tool tracking and identification’ – as ‘kits’ are: serviced and built, measured, the sent to an awaiting machine tool. [Courtesy of Sandvik Coromant]

Obviously, it is important to create a suitable tool management system, that can operate successfully in a company’s machine shop and it needs to be customised to suit their particular tooling requirements from a relevant database. These tooling-related matters, will form the basis for a discussion in the following section.

6.5.2 Creating a Tool Management and Document Database Production Requirements Prior to any new machining activities being undertaken and, in order to establish the ‘true’ production requirements of a company, it is essential that co operation and information regarding the customer’s potential product occurs. More specifically, this de-

tailed dialogue should be between both the sales and manufacturing departments. The first requirement is an understanding of the manufacturing load, typically these being broken-down into the following batch and volume classifications23: • Job shop – one-, or two-off specialised workpieces,

23 Optimum/economic batch size, this will vary, but if batch size is graphically-plotted against cost , for values of set-up cost and holding cost, then the overall total variable cost can be derived, with the lowest plotted value representing the minimum cost batch size ‘Q*’ (i.e. derived from R.Wild’s book: Production and Operations Management, Chap. 14 – see References), as follows: Q* = √ 2Csr/C1 Where: Q* = minimum cost batch size, Cs = set-up, or preparation cost/batch, r = consumption rate, C1 = stock-holding cost/ item/unit of time.


241

• Small batch – up to perhaps 50 workpieces, Perishable and Capital Equipment Review • Medium batch – between 20 to 100 workpieces, • Large batch/Volume production – >100 work- In many cases, cutting tool manufacturers produce pieces.

NB These classifications of batch size are open to considerably much wider interpretation, obviously depending upon a specific company’s production requirements and the actual machined part’s: complexity, material cost, machining operations and its dimensional size and so on! At any workpiece quantity greater than the ‘Job shop’ levels having similar production processes undertaken, allows them to be grouped into ‘families’ , according to their: dimensions, tolerances, workpiece materials, etc. This technique of allocating components to be machined into similar groupings is often termed ‘Group Technology’ 24. It is vitally important that both the Sales and Marketing personnel are aware of the company’s patterns of manufacture and their capabilities, if the company is to be able to rapidly respond to their customer’s needs. The sales force will be able to relate a customer’s requirements to the standard range of parts produced, with the manufacturer being in a position to ‘fine-tune’ even small production runs for maximum efficiency. By comprehending the manufacturing process for the company’s standard-ranges, allows the optimum conditions of production to be utilised, even when ‘modified standards’ , or even ‘specials’ have to be produced. Flexibility here, plus the ability to cater for unique customer needs, may offer new market opportunities for the company.

24 ‘Group Technology’ (i.e. GT), is essentially utilised for ‘groupings’ in two distinct varieties: (i) Component geometry – the ‘closeness of shapes’ , (ii) Similar production processes – such as: Milling, Drilling, Turning, etc. The benefits of utilising a GT-approach to manufacture are: smoother logistical work-flow, simplified work control, more efficient plant layout and improved use of floor-space, contributing to enhanced manufacturing versatility and better response to variable workpiece shoploadings. NB The GT-approach to manufacturing lends itself to component coding systems, typically of the Opitz variety for a unique part-coding classification.

standard forms to enable companies to compile data on both their perishable and capital equipment needs. Therefore, it is necessary to gather the data together, because the performances of either categories are independent. If a tooling survey is approached in a methodical and step-wise manner, then the following sequence, may be of some help: • Collect data on perishable tooling, a company must analyse their entire tool-flow system, including tooling inventories: high-lighting the maximum and minimum levels, quantities of new and used tooling, together with their tool-storage requirements. As a preliminary data-gathering exercise, all the items in stock should be listed, plus the number currently in stock and, the quantity used in the last 12 months, with the last price paid for them, • Review the stock lists for tool obsolescence, by checking to see which items have not been used in the last 12 months and which can be replaced by say, an ANSI, or DIN standard item. Any tools falling into this category, can be considered obsolete, it would not be a surprising fact to find out that up to 50% of the current tooling inventory was obsolete – as has been shown in a survey in the USA. This level of obsolescence, can be regarded as money ‘tied-up’ and doing nothing for the company’s profitability, • Remaining ‘tooling items’ should now be reviewed, as these are not obsolete. For example, the cemented carbide insert grade of tooling, should be grouped according to their grade/coating, size, geometry, etc., then once they have been ‘grouped’ in this manner, it is now perhaps possible to order larger quantities of them, enabling the company to exert some ‘leverage’ over their suppliers and to obtain substantial cost advantages as a result. NB If tooling information of this nature is compiled and regularly up-dated, then reviewed, future tooling decisions can be speeded-up and decisions can be taken with some degree of confidence.

The compiling of information concerning capital equipment within the factory, usually commences with the preliminary identification of such machine tools and associated equipment, then numbering them (i.e. this activity is often termed: ‘brass-tagging’) and their spe-

242

Chapter 6

cific location within the company’s premises. A list is then compiled, allocating every machine tool’s: power capacity, spindle taper, number of spindles, its current operating condition25, present tooling utilised, plus the current and past operations performed on each machine tool. By organising information concerning the capital equipment capability and availability, produces a number of distinct benefits, including a knowledge of the machines basic characteristics, thus ensuring that the most advantageous machine tool can be selected and any machining operation is performed using the optimum parameters. Knowledge gained from such a study of perishable and capital equipment, allows for improvements in both: process planning – the action plan for the manufacturing of a certain part, together with production planning – the best use of a factory’s resources for a particular workload.

Building-Up the Tool File Probably the principal users of a tool file within any manufacturing organisation are the Process Engineer and Part-programmer, with perhaps the Tool-presetting operator and Stores-personnel, also making use of this ‘file’. It should be stressed that new tools are only added to the tool file after a proper investigation of the need for them, assuming that such tooling was not previously listed. By accepting this limitation on

25 ‘Operating condition’ , this machine tool activity is invariably of some concern, as although some form of periodic maintenance is likely to be undertaken, perhaps less attention is given to the machine’s current state of calibration. This situation can be diagnostically-achieved, both speedily and efficiently by the use of telescoping ‘Ball-bars’. This calibration equipment can undertake a quick ‘health-check’ and assess both the static and dynamic machine tool’s performance, indicating the following important characteristics: Servo-mismatch, Stick-slip, Reversal spikes, Scale uncertainty, Straightness, Squareness, Lateral play, Back-lash, Cyclic error. These machine tool related-factors are automatically prioritised by the software, then they can then be simply: analysed, diagnosed and then corrected, together with a machine tool ‘health-check’ report. NB Many companies perform these full diagnostic ‘healthchecks’ , periodically, or simply prior to a shift commencing, as they can be undertaken in just a few minutes for a ‘quick assessment’ , or perhaps a more thorough ‘Ballbar’ assessment can be achieved in just a few hours – when a convenient ‘maintenance window’ occurs.

the number of different tools in the tooling inventory, then a company can be assured that when the tools are ‘called-up for use’ in the manufacturing process they will be available and backed-up with spares, since the ‘stores’ has access to this ‘file’. An important feature of any tool file is the cutting data and machining times listed. These machining data are known to be achievable and will be those values expected to be employed during component production. More specifically, the data values are the ones utilised to calculate quotation prices for the product, for any future customer appraisal. The editor of the tool file has a key role to play in the acquisition of tooling data, so when building up the ‘file’ , they have to: • Scrutinise any reported deviations – from the recorded cutting data and the original tool file, • Investigate higher productivity ratings – some newly-available tooling may lay claim to be both faster and more efficient in comparison to either its predecessor, or competitors tooling products, • Obsolete tooling should be ‘weeded-out’ – particularly with the introduction and addition to the tool file of newer high-technology tooling, • Investigation of new tooling – to see if claims of new tooling products, with regard to their: geometry, coatings, performance, etc., are a genuine improvement over the previous versions used. The systematic accumulation of tooling knowledge in the tool file for each section of the manufacturing operation, ensures that the cutting performance will ‘continuously improve’ 26. Such improvements may be considered to be analogous to improving the skill of an operator on a conventional machine tool, but with more flexibility, as the tool file system software is able to cope with much more diverse and complex tooling

26 Continuous improvement programs’ , were originally developed in Japan and are now well-known and are often termed ‘Kaizen’ *, which is a philosophical and rigorous approach to process/product improvements, based upon: (i) Satisfying the customers – in order to keep the business alive and to be more profitable, (ii) Being both customer- and process-oriented – to promote vigorous improvements here, (iii) Requiring commitment and participation of a company’s personnel – using their knowledge and experience to achieve continuous improvements in both working practices and in the product’s quality. *In Japanese, Kaizen, approximates to: ‘Change to the better’.


situations. A ‘well-disciplined’ and ‘active’ tool file, completely eliminates the anticipated ‘hiccups’ that are likely to occur, whenever a new Part programmer is employed, or even when hiring new Stores personnel, or Machine operators, for that matter! With the software structure of the tool file – of necessity – being highly complex and interrelated in nature, it is not possible within the confines of this chapter to go into too much detail showing how the operating system works. However, an appreciation of a simpler, but still valid tool file format can be gleaned by describing how a ‘manual file’ is produced. Prior to constructing this manual tooling data-base, a separate record is produced for each tool, which is cross-referenced to separate cards for cutting inserts that can be utilised with the tool. In essence, there are four ‘fields of tooling information’ that are needed for a usable tool file, these are: 1. Tooling is built-up from modular elements – which are the ‘key’ to effective management and control, as they allow the widest range of tooling for an extensive assortment of machine tools available, from the minimum number of tooling elements in stock. Hence, a tool for a given machining application may be assembled from different modular elements, to suit a range of machines, 2. Materials Requirement Planning (MRP)27 – the system together with the tool stores should support the tooling from the tool file, with such items as: spares, consumables, plus back-up tooling. In order to achieve this objective, the tool file record would include details of the build-up for each tool as well as the stores location for each part, using a specific ‘key’ notation,

27 ‘Materials Requirement Planning’ (MRP*), is a softwaredriven system that enables manufacturing companies to calculate: how many materials are needed, what particular types that are required, plus at what times they are needed. To achieve this level of control, the system utilises a sales order book, which records known future orders and also a ‘forecast’ of what sales orders the business is reasonably confident it might have won. Then, the MRP software interrogates and checks all the ‘components/ingredients’ which are required to make these future orders and ensures that they are ordered in time. *MRP, was originally developed in the 1960’s and is sometimes termed MRPI, to differentiate it from the ‘derived’ MPRII system – see Footnote 18.

243

3. Certain ‘steering comments’ on tooling – normally these statements are based upon shop-floor experience, that are included to enable the Process-, or Planning-engineer to select the appropriate tool for the desired machining application, 4. Organisation of basic cutting data – this is normally produced so that the data can readily be included into the CNC program. NB This cutting data is organised according to the component to be machined and the optimum organisation of this machining data will vary from one company to another, depending on their needs. Of course, all of this data listed reflects the company’s actual experience, in particular, it includes the results of any ‘optimisation exercises’ (i.e. Machinability trials – more on this later) previously undertaken in the machine shop.

Practicalities when Starting-Up a Tool File Whenever a tool file system is initiated, the important point to observe is to: start small and keep the tooling information to be included ‘sound’. Having accepted this principle, a company may start to build-up the tool file steadily – over perhaps a month, including any practical test data for maybe a hundred, or so of the most popular tools utilised. This information now residing within the tool file, will flow through the toolmanagement system and, it will begin to highlight the requirements of the system users, driving forward the file’s further development. Conversely, a company may embark on a more comprehensive tool file system, incorporating all the machining data available on perhaps two thousand tools, but utilising provisional cutting data, instead of well-proven information. With this rather heavy-handed and rapidly-built tool file approach, the probable outcome will mean that the whole ‘file’ has incorporated many ‘dud tooling solutions’ , thereby ensuring that the system is discredited, even before it is correctly operating at anywhere near its optimum level. Obtaining meaningful test results and tool assessments does not necessarily demand extra effort from the company, merely the organisation of endeavour already being made by the company’s ‘Tooling-engineers’ and those from the tool suppliers. Often, most of the tooling trouble-shooting activities will dissipate once the current component batch is completed, sim-

244

Chapter 6

ply because there is no framework in which this vital information can be recorded – for future usage. So, all the time and pain’s-taking effort needed to collate ‘sound’ tooling data is disregarded and the information is discarded. So, when then a repeat batch order duly arrives, the whole tooling-related data-gathering process must once again begin, by ‘re-inventing the wheel’ – this being a total and unnecessary, but is costly waste of everybody’s time! Tool-kitting servicing to the machine shop must be based upon the assurance that the completed kits are dependable, whilst providing the maximum security from a limited budget for tooling stock. Usually there is a finite tool stock available, with the objective being to utilise, for example, the same modular tools across a range of machine tools. For this reason, modular quick-change tooling, has seen a widespread acceptance by machining-based companies of late. Yet another important factor in any tooling requisition for a specific machining operation, is the inherent quality of the tools used. One of the major function’s of the tool-kitting area, is to monitor and control the quality of delivery of tooling within the manufacturing environment, by accessing the ‘tool data flow’ for both out-going ‘new kits’ and in-coming tooling ‘old kits’ from completed production runs. As batch sizes become smaller, the ‘logistical-flow’ of kits speeds-up. The effectiveness of the tool-kitting personnel will be inversely-proportional to the number of tooling items on the inventory and the ‘standard’ they must control. This problem of effective tool control, is a further argument in favour of a factory-wide standardisation of the tooling inventory. Therefore, in summing up tooling-related activities within the production location, two main factors emerge, these are: 1. Linking every tool with its application technology – this is normally achieved in such a manner that it is the most productive tools that are chosen for new jobs and not the old ones – just because they have been previously used and are known to be supported by the tool stores. This tool selection strategy, will result in the optimum cutting conditions being selected, 2. Formulating a rationalised and optimised tool management ‘standard’ – this is essential as it supports tooling across the breadth of the whole factory. NB When purchasing any new tooling, or machine tools, reference to this ‘standard’ is of the essence for the overall system to operate effectively.

6.5.3 Overall Benefits of a Tool Management System By the correct implementation of a basic, but competent tool management control system, the following list highlights the ‘rewards’ that can be expected: • Manpower is conserved and training requirement minimised, • The number of tools lost, or misplaced is reduced, • Timely and up-to-date information on tool usage is produced, • Tool inventory shortages are identified and prevented, • The accuracy of the tooling inventory is improved, • Inventory levels and excess purchasing are minimised, • Time spent on re-ordering, etc., plus ‘piecemeal purchasing’ are reduced, • Record-keeping functions are consolidated, • Tool tracking and tooling availability within the machine shop is monitored, • Tools in rework can be tracked, • A record of scrapped tools can be kept, • Obsolete tooling can be identified and then eliminated, • The cost of the total tooling inventory can be critically-assessed, • The gauges and fixtures supplied with the tool kits can be identified and tracked, • Machine tool set-up, tool-return and withdrawal times are reduced, • Possibility of pin-pointing over-use machining problems, by specific personnel, • Improper charge-outs, losses, or pilferage can be minimised, • Space requirements and overheads are reduced, • Possibility of incorporating existing tool numbers and current mode of operation into an automated system, without making radical changes. Tool management systems provide all of the above benefits, by allowing the operations to be easily reported, analysed and corrected, enabling timely decisions to be made, concerning the tooling, with the minimum of manpower and operational changes necessary. So that the information required by a company can be obtained, the system should be organised to allow personnel responsible for the tools to record their activities. On the ‘shop floor’ , it is the usual practice to allow two basic groups of the workforce levels of responsibility/access to the system to provide both


vital and helpful tooling information, these are the: Tooling-supervisor and Stores personnel. So far, the information on Tool management systems has been principally concerned with the justification and benefits that accrue through the adoption by a company and the philosophy underpinning its practical application. In the ‘continuous circle’ of tool monitoring and control, the tool-kitting area is at the ‘heart’ of the overall tool management procedure. This vital day-to-day activity of tool preparation and setting, will be the subject of the following section.

6.5.4 Tool Presetting Equipment and Techniques for Measuring Tools

245

ing the drill’s length, once it has been suitably located and held in an appropriate chuck, 3. Qualified tools – are when all the tool offset data are known and this information can be readily input into the CNC controller’s tool table. Typically, ‘Modular quick-change tooling’29, can be considered under this category.

Presetting on the Machine Tool – Tool Contacting When setting an ‘unqualified’ tooling dimension – such as a drill’s length, on the machine tool, this being the crudest form of tool presetting*. It is achieved on say, on a vertical machining centre, in the following manner: the cutting tool’s tip is held in the machine’s spindle and is positioned over the table, being slowly ‘jogged-down’ 30 until its just touches a suitable ‘setting

Introduction Cutting tools that are to be utilised on CNC machine tools for the production of workpiece features, need to have exact measurement information regarding their offsets known, so that the CNC program can automatically displace (i.e. offset) the tool these dimensional distances, in order to perform the intended machining task. Otherwise, major errors in the machined component’s dimensional features would result. Hence, cutting tooling can be classified under three distinct headings, these are: 1. Unqualified tools – these are tools that do not have known dimensions, therefore they must be independently measured and these values can then be located and placed into a ‘suitable field’ within the CNC Controller’s tool table. Typical of such tooling, are special-purpose form tools that may be considered to fulfil this classification, 2. Semi-qualified tools – these are tools where not all of the tool measurement offset data are known. For example, a typical Jobber drill’s diameter would be normally be known – say, φ12 mm28, but perhaps its length for the purposes of utilising it immediately would not. Therefore, it would necessitate measur-

29 Modular quick-change tooling, such as the ‘front-end’ cutting units, fitted into the already machine tool-pocketed and located ‘back-ends’ , typified by the ‘KM tooling’ ranges (i.e. see Figs. 120 to 122), would give the following repeatability readings: • Axial tolerance ±0.0025 mm, • Radial tolerance: ±0.0025 mm, • Cutting-edge height tolerance: ±0.025 mm. NB All of these tooling manufacturer’s tolerances, limit the machining tolerances that can be held, unless they (i.e. already placed within quick-change tools in their respective holders) themselves are measured, which tends to negate the rationale for their original purchase! 30 ‘Jogging-down’ – sometimes referred to as ‘inching-down’ , is a manual means of slowly lowering the tool’s tip down onto a surface – in this case a known height ‘setting-block’. This linearly-controlled action is achieved, by employing the ‘handwheel’ , which allows the handwheels angular rotation to be equated to an operator preselected incremental amount. This incremental motion can be changed to a smaller value, as the block is slowly approached, to give a sense of ‘feel’ (i.e somewhat like using a ‘feeler-gauge’), as contact is made between the tool and the block. NB The tooling is usually kept stationary while this manual setting activity is undertaken.

28 Whenever a tool’s dimensional size is known, it is necessary to refer-back to the individual tooling manufacturer’s tolerance specification, in order to establish the limiting values when this data is utilised, when the tool is to be used without any form of pre-measurement being undertaken.

*

This is not strictly the most basic tool setting method, as the ‘cut and measure’ technique – then setting this measured value in the tool table, is the most primitive and time-consuming procedure of tool offset setting.

246

Chapter 6

block’ 31. The Z-axis position is then noted and its value is automatically entered into the tool table, giving a ‘semi-qualified’ tool offset, that can then be used for the important Z-axis motion – when coming down onto the workpiece’s surface to begin engaging in the first cut. If each tool length has to be input into the tool table’s ‘offsets’ , then this simple procedure has several disadvantages: it is labour-intensive, ties-up cycle considerable time, it is rather inaccurate and, it sets only one offset dimension. In the case of turning centres, the technique of determining offsets is different, but similar limitations still apply. A tool presetting device is often used on many of today’s machine tools, this technique is typified by the ubiquitous ‘touch-trigger probe’ 32. Hence, this type of tool-contacting presetting probe fulfils a number of ‘quoted benefits’ , such as: • Setting/re-setting of tool length and diameter (Fig. 133b) – automatically up-dating, or correction of the respective tool table offsets, even while the tool is still rotating, • Measurement of a complete tool station – automatically in just a few minutes,

NB A small vertical machining centre with a 12 to 15 tool station, would take at least 5 minutes per tool, with the traditional manual technique, mentioned above (i.e. see Fig. 133-bottom right, inset graph/description).

• Elimination of manual setting errors – tools that

• • •

•

are set manually, particularly tooling such as a large diameter face mill, it will be open to errors when setting both height and diameter offsets. This is because each cutting insert may ‘stand proud’ in its respective seating, giving a false offset reading – when stationary. Ideally, the whole tooling assembly needs to be rotated as its offset is set, No presetting of tools is necessary – as this is automatically undertaken on the machine tool, Accurate and precise ‘First-off machining’33 – this is the result of confidence in the tool offsets, set by the ‘probing system’ , In-cycle tool breakage detection – at convenient and programmed pre-selected intervals, the tool’s offsets can be checked for either: tool wear – to a prescribed level, or tool breakage, which will automatically stop the machine preventing either further workpiece damage, or part-scrappage, Improved confidence in unmanned machining – due to the fact that tool breakage detection periodically occurs, untended machining operations can be undertaken.

31 ‘Setting blocks’ , are usually manufactured from hardened steel, that have been accurately and precisely ground to a known dimensional size and tolerance, nominally to some conveniently ‘round figure’ , for example:100 mm in height. These ‘blocks’ are usually either rectangular, or round in cross-section. The rectangular ones are preferred, because different nominal dimensions can be utilised for each adjacently flat and square face. The tolerance for the ‘Setting block’ should be ‘very close’ , as any difference from the nominal size when input into the tool table, will impinge on the overall workpiece tolerance, in essence, somewhat reducing the tolerance’s ‘working range’.

These are ‘real benefits’ that occur when using ‘on-machine’ tool presetting equipment, but the ‘down-side’ of such systems is they do utilise some potential in-cycle cutting time. This negative effect using some of the cycle-time, can be significantly reduced for the following presetting system, employing non-contact laser-based tool setting techniques.

32 ‘Touch-trigger probes’ **, in the simplest form these ‘tool probes’ are omni-directional switches, that are sprung-loaded, which when the tool makes contact with either an attached setting cube, or a cylindrical ‘setting gauge’ (Fig. 133b), it immediately breaks the electrical circuit. This loss of electrical contact occurs when the three equi-spaced precision rods: each one seated on two precision balls (i.e. each rod being positioned at 120° to each other) in a simple kinematic seating mechanism, are lifted/pushed either individually, or ‘as one’ out of their respective seating(s), which triggers an ‘electrical pulse’ representing a nominal dimension and is automatically recorded as either a length, or radius – in the case of a rotating tool, which then automatically up-dates the tool table’s offsets for this tool. **On a turning centre (Fig. 133a), this tool setting touch-trigger probe, is termed a ‘tool eye’.

33 ‘First-off machining’ , this term is self-explanatory, in that it is the first component produced in a batch which is simply known as the ‘First-off ’ the machine. Invariably, this initial component produced, is subject to rigorous inspection procedures, being the ‘initiator’ for calculated data concerning the whole batch’s metrological and statistical variability/consistency.


. Figure 133. Cutting tool offsets being set on a turning and machining centre. [Courtesy of Renishaw plc]

247

248

Chapter 6

Presetting on the Machine Tool – Non-Contacting Tool Setting In recent years, laser systems for tool setting and broken tool detection on CNC machining centres have become popular (Fig. 134), as manufacturers realise the benefits of fast process set-ups and in-process feed-back on the tool’s current condition, particularly on diminutive tooling that cannot be easily measured by the more usual contact-type sensors. Laser non-contact tool setting systems, utilise a beam of laser light which passes between a transmitter and a receiver, located either on the bed of the machine, or on each side of it allowing the beam to pass through the ‘working volume’ (Figs. 134a and b). Hence, the tool’s passage through this beam, causes a reduction in light as seen by the receiver, which will then generate a ‘trigger-signal’. This ‘triggered-signal’ for the machine’s actual position, is instantly recorded and from which, the tool’s dimensional characteristic can be derived. Not only can the system measure the required tool’s dimensional parameters, it can also be used to detect broken tools. This tool breakage process involves rapidly moving the tool into a position where it can intersect the laser beam, so, if the light reaches the receiver, then the tool’s tip, or point, must be either missing, or broken. There are quite considerable benefits that accrue by the application of a non-contact laser tool setting system, these include: • Rapid measurement of both tool length and diameter – tools can be moved into the laser beam at high speed, without risk, or any attendant damage and the tool offsets are automatically up-dated (Fig. 134a), • Fast tool setting times can be achieved – tools can be measured at normal rotational speeds, allowing tooling assembly and taper fitment errors such as radial run-out, taper ‘pull-back’ to be identified, then compensated for by the system, • Minute, or delicate cutting tools can be conveniently measured – without any subsequent tool wear, or damage (Fig. 134b), • Tool breakage can be checked at very high feedrates – this efficient process minimises cycletime, while increasing confidence in untended machining applications, • Multi-point tooling can have each facet checked – this is automatically undertaken while the tool rotates,

• Monitoring tool settings on the machine – enables compensation for any ‘thermal movement’ 34 of the machine spindle.

Although the measurement process lasts for only a few seconds, this is long enough for the chance of a falling coolant drip to intersect the laser beam, possibly creating and attendant measurement error. Hence, the laser tool setting equipment, must be able to distinguish between reductions in light at the receiver, created by a ‘falling object’ (i.e. termed: ‘drip-rejection’) as compared to rotating tool, if it is to avoid ‘false-triggering’ producing tool measurement errors. This elimination of ‘false-triggers’ , is achieved by the filtering-out of signals by the electronic interface, this value being set at a pre-determined ‘trigger-threshold’. It should be noted, that the laser tool setting system cannot cope with following circumstances: the presence of ‘floodcoolant’ , cutter edge and profile checking, nor with radial broken tool rejection processes. The cutting edge laser measurement is quite a complex process, when the tooling assembly is both rotating and in linear motion simultaneously. If one considers the relative motion of just one of these cutter’s teeth, then, its edge moves in a circular path and superimposed onto which will be the axial feedrate, this motion being perpendicular to the laser beam. Hence, for each of the tool’s revolutions, the prominent edge approaches the laser beam by an increment, this value is the feed per revolution. Such incremental movement, introduces a potential error into measurement of the tool’s size. For instance, if a tool rotates at 1,000 rev min–1 while feeding toward the laser beam at 100 mm min–1, it will be seen to advance by 100 µm between intersections of its prominent cutting edge

34 ‘Thermal movement’ of the machine spindle, is important, as the whole tooling assembly can effectively ‘grow’* due to thermal effects, which may present problems – if not compensated for – when very tight machining tolerances have to be held, or maintained across either a high-quality machined component, or for consistency in a large batch run. *Tests undertaken, on a vertical machining centre equipped with a ball bearing spindle – utilising a special-purpose ‘Invar’ spindle analyser – with the tooling being rotated at 3,000 rev min–1 under ‘no-load conditions’ for one hour, have produced the following thermal results: Z-axis drift 9.2 µm, Y-axis drift 6.3 µm and X-axis drift 0.7 µm. Additionally, the bearing itself, had some radial error motion present – as indicated in a ‘polar-plot’ , this value being typically: 4.6 µm for the total radial error.


. Figure 134. Automatic cutting tool setting and tool breakage detection, utilising an ‘on-machine’ non-contact laser. [Courtesy of Renishaw plc]

249

250

Chapter 6

to that of the ‘stationary position’ of the laser beam – this being the maximum possible ‘feed per revolution’ error for any one particular reading. Conversely, an improved accuracy can be obtained by rotating the cutter faster, but advancing more slowly. For example, if one wants only a 1 µm rev–1 intersections, this level of accuracy can be obtained by rotating the tooling at 3,000 rev min–1,while advancing at only 3 mm min–1. In order to minimise cycle times, the tool measurement software, programs the machine tool to move the tooling into the beam initially from a ‘stand-off distance’ that is adequate to account for the uncertainty of tool assembly build-up – this is important when setting the tool’s length, if the tool is held in a collet, or similarly-designed toolholder. So, the initial move is a fast feedrate to gain an approximate position with respect to the laser’s beam, from which the tool is backed-off by a small linear distance. Here, the tool is ‘probed’ at a reduced reduced feedrate, this is necessary to more accurately find the tool’s location, from where a very short distance ‘back-off ’ move is executed. Finally, a measurement move is completed at a very low feed rate, so that an accurate measurement is tenable. This complete tool checking process is considerable quicker than approaching with the laser beam at a constant, but low feedrate, from a larger ‘stand-off distance’ – see Fig. 134c. While, yet another challenge to precise and accurate tool measurement, is the result of the presence of either coolant, or debris on the tool’s tip which is about to be measured. The most significant problem facing non-contact sensing, when compared to its equivalent contacting techniques – this latter method achieves ‘hard-contact’ with the tool and can thereby safely ignore any coolant films, or liquid drips – is that in the former case no actual tool contact occurs. This lack of contact in the presence of fluid media, can be overcome by rotating the cutting tool assembly at very high speeds, so as to dislodge any fluid residue, or perhaps another strategy is by utilising an air-blast on the tool for non-contact measurement. Yet another software technique that can be used, is the capacity to measure the tooling several times and apply a ‘scatter tolerance’ to check for any variation resulting from measuring ‘something’ , other than the tool itself (Fig 134d). This software routine will retake readings until it obtains several values within the required tolerance – these ‘tool-checking retries’ , plus the ‘scatter tolerance bandwidth’ can be pre-selected by the user. The detection of broken tools is somewhat less demanding than for tool measurement – in terms of ac-

curacy and precision, although the cycle-time tends to be more critical. The demands on a laser broken tool detection system require it to be ‘active’ at the instant it is required and, be able to operate under the prevailing conditions, instantaneously after machining stops. The laser transmitter for the non-contact tool detection system shown in Fig. 134ai, has been designed with a ‘MicroHole™’35, this ensures that the presence of coolant does not affect the integrity of the laser system. In practice, the laser system reliably operates under relatively ‘harsh’ workshop conditions and the broken tool detection system works in the following sequential manner: 1. Tool’s end is moved at rapid traverse into the laser beam by 0.2 mm, 2. Tool breakage cycle is activated via an M-code from the CNC controller, 3. End of tool dwells in the laser beam for between 0.1 to 0.3 seconds, 4. If laser light is received by the optical receiver unit for more than a specified time-period, typically 10 µs, then this distinguishes a broken tool is present, 5. If laser light is not received by the optical receiver unit, then the tool’s condition is satisfactory, 6. Tool is the moved rapidly to its respective home position – end of cycle. NB This detection cycle also enables small tools to be inspected, even when in the presence of ‘floodcoolant’ , thereby minimising cycle-times. These laser non-contact tool setting systems, offer many more software-based features, not described here, such as: cutter profile checking routines, together with many inspection/checking routines for either tool measurement, or broken tool detection. Furthermore, laser tool setting systems provide machining-based companies with a rapid, flexible, accurate and precise approach to control tooling dimensions and offer the techniques necessary to increase machining automation.

35 ‘MicroHole™’ – both for the laser transmitter and the optical receiver, incorporate an angled aperture of just 0.75 mm diameter, as this ensures that protection from: coolant, chips, swarf and other debris such as machined graphite fines occurs, because of a continuous stream of air that flows through and along the laser beam – protecting it as schematically illustrated in Fig. 134ai.


. Figure 135. An automated five axis CNC universal tool measuring machine for metrological and geometric inspection. [Courtesy of E. Zoller GmbH & Co. KG]

251

252

Chapter 6

Universal Measuring Machine – for Checking Tooling In many machining circumstances, the tool’s profile becomes part of the contoured form for the final machined component. Therefore, it is important after the milling cutter has been multi-axes ground to a desired profile, that this form is rigorously inspected, as the cutter’s ‘rotated-shape’ will become part of the final workpiece geometry. In order to establish this ground complex cutter profile, special-purpose universal tool measuring machines have been developed – see Fig. 135. Such multi-axes machines have a range of functions, from simply manually-checking elementary cutter forms, to that of fully-automatic assessment of a multi-faceted form cutters. The machine illustrated in Fig. 135, is based upon ‘sound’ kinematic principles, equipped with three linear axes and two rotary axes (Fig. 135c). The high-precision linear guidance motions are controlled by re-circulating ballscrews36, these being propelled by servo-motors. The incident light measuring technology associated with this type of machine is quite sophisticated (Fig. 135b), offering 3-D image processing, to permit threedimensional geometrical cutter elements to be fullyautomatically measured – using a ‘proximity method’ of assessment. The camera, lens and the LED incident light in combination with its automatically dimmable segments, have been designed to operate with: ground, or eroded PCD, cemented carbide and HSS tooling. Special-purpose ambient light filters and an automatic lighting calibration function, ensure that tool coatings, such as: ‘chemically-blackened’ , TiN-coating, or brightly-ground tool surfaces, can be scanned in 3-D, plus their respective profile geometries. The image processing software is enhanced, allowing a range of complex tool geometries and profile forms to be evaluated. To gain an understanding of this tool geometry complexity, some of ground tool forms are depicted in Fig. 137 where ‘program-

36 Re-circulating ballscrews, their geometry is based upon the ‘Ogival’ , or ‘Gothic arch’ principle. This geometry, ensures that point contact occurs between the ball, its nut and the screw, contributing low friction with better that 90% efficiency, at high-velocity slideway translations. Such ‘Ballscrews’ offer minimal backlash, with better than 5 µm accuracy/precision over 300 mm, typically having high stiffness values of up to 2000 N µm–1.

ming routines’ based upon an optical tool presetting machine are shown, for profile assessments. Typically, these universal tool measuring machines (Fig. 135) have image processing software, allowing for the following tooling-based metrological assessments: • Incident light image processing – with automatic illumination control, offering ‘search-and-run’ and auto-focus enhancements, NB For manual measurement of radial, or axial tool geometries, this is achieved at x200 magnification, having facilities for both image memory and log output.

• Contour-tracking tool/workpiece measurement – without the need to write complex programs which can be readily undertaken,

NB Thousands of tool geometry data points can be measured in just seconds, backed-up with nominal/actual comparison for ‘best-fit’ which can be speedily and efficiently achieved. This data can then be either printed out – in a ‘test log’ , or saved on a disk – for future reference.

• Fully-automatic measurement of contour radii

(i.e. see Fig. 135d) – giving vertically exaggerated graphical display of tool’s profile, with the specified tolerance range, allowing checking for ‘transitions’ on both the cutter’s end and along the tool’s shanks.

When a new tool requires measurement, it involves entering only the most important nominal tooling dimensions, while performing any necessary corrections during the automatic measuring procedure, afterward, this information is permanently stored in the relevant section of a tooling database. From this point, any further inspection procedure on the tool geometry, will be undertaken automatically – at the simple touch of a button! Such universal tool measuring machines have tooling-based software measurement programs, that permit, inspection of tools, such as: die-sinkers, and thread-milling cutters, etc., to be readily inspected. This automated cutter geometry inspection, allows the information to be down-loaded back to the CNC multi-axes cutter grinder – for further tool grinding enhancement, or it can be sent to the equipment in the tool presetting area.


253

. Figure 136. Optical tool presetting machine for sophisticated tool management control. [Courtesy of E. Zoller GmbH & Co. KG]

254

Chapter 6

Presetting off the Machine Tool High quality tool measuring equipment has been developed in order to eliminate the disadvantages of tool presetting on the machine tool. Presetting machines (Fig. 136), are usually designed so that they can accurately and precisely locate the toolholder and its respective cutter, in exactly the same orientation as it would be situated within the intended machine tool’s spindle. Once the tooling assembly has been securely located in the presetter, the tooling’s cutting edge(s), can then be measured by a range of means, including: a non-contact optical device, a contacting mechanical indicator, or more ‘primitively’ using some form of comparator gauges. Hence, by making the necessary tool adjustments whilst the tooling is located in the presetter, the operator can ensure that when this inspected tool is finally located in the machine tool, its respective tool offsets will be confidently known and applied to the cutting operation in hand. By utilising a tool presetter to measure and set tools off the machine tool, this has been shown to increase the shop floor productivity by >12% for every machine using preset tooling. Due to the demands for the highest ‘up-time’ possible in the automotive sector, virtually every production shop employs measured and preset tools. In fact, studies conducted at manufacturing companies using a presetting tooling facility, have noted that by utilising a presetter, this has been shown to save typical workshops >4.52 minutes every time tools are changed. In the following example on the use of presetters, it was noted that significant productivity time and hence cost-savings can be accrued, these calculations being based upon 20 tool changes per eighthour shift, this gave the following savings: • Minimum time saved for each tool = 3 minutes, • Total minutes saved per shift = 60 minutes, • Calculated productivity increase = 12.5%37.

37 This 12.5% productivity gain, meant that one hour was saved for every eight hours of shift operation. Hence, if the facility was run at the ultimate level of operation, such as in a mass production automotive machining facility, running a continuous three-shift system, seven days per week. Then, a total of three hours per day, or 21 hours per week would be saved, which would mean that the amortisation for the capital plant (i.e. the presetter and its presetting environment), would be very short indeed.

Significant tool setting and changing time-savings are only one major advantage from utilising a sophisticated tool presetter like the one shown in Fig. 136 and its associated screen displays in Fig. 137, other features include: integrated tool measurement and inspection and data-storage facilities. Down-loading this tool offset and other important data through a DNC-link to each relevant machine tool, making them a vital part of the overall tool-management system. A high-quality tool presetting machine can set tools to ‘micron-levels’ of accuracy and precision (i.e. typically ± 2 µm), holding these preset levels with confidence as soon as they begin cutting chips – so no ‘trial cuts’ are necessary. Moreover, a range of toolholding ‘backends’ can be accommodated in the machine’s spindle, by using special-purpose adaptors. The tool presetting software guides an operator through the measuring program and other tool management tasks. Within the presetter’s computer memory, an operator can store and retrieve tooling information as necessary, allowing for repeat setups, or replacement tools to be speedily and efficiently measured and set. On the presetter shown in Fig. 136, this machine allows typical tooling screen displays shown in Fig. 137, having a photo-realistic input screen, which guides the user through the measurement and setting program in easy-to-follow steps. This data is stored for further use and enables tooling repeatability, with very little variability, allowing each individual set tool to have almost identical offset dimensions. This repeatability ensures that the operator can load the machine tool’s spindle with confidence, allowing for tool-data optimisation to be achieved on the machine – when these tools are operating under batch, or mass production runs. For many of today’s presetting machines, they allow the operator to inspect the tooling with ‘videotechnology’ (Fig. 137) to assess for tool wear and its measurement. Flank wear in particular, is often a good guide as to the probable life left in the tooling, prior to a tool change. At a certain level of predetermined wear land, the tool is deemed to need replacing. Not only can a presetter be used for presetting tooling assemblies and for tool wear assessment, it can also be employed to monitor and inspect incoming tooling from suppliers in the ‘as-received condition’ , to


255

‘Vendor rate’ 38 and establish the tooling supplier’s quality levels – in terms of their tool geometry and in certain instances, dimensional tolerances. If a presetting machine’s tool set-up and inspection capabilities are combined with sophisticated software, its overall abilities are considerably enhanced. Here, it has the potential to perform both tool and component tracking, together with that of whole tooling assemblies within the production facility, whilst operating as a complete tool management system. With such a computerised-system in place, it can store data on individual components and when required, select the relevant information to assemble a complete tool. This data availability, can include the overall tooling inventory and the operator can monitor the workshop’s stock of tools and order replacements, based upon a ‘Just-in-Time (JIT)’ strategy (i.e. see Footnote 17) – by directly ordering from the tooling suppliers computerised-stocklists. In order to obtain maximum efficiency with the tool presetter, this can be achieved by linking it with the in-house computer network. Hence, a fullyintegrated presetting machine can exchange data with the company’s other peripheral-networked computer systems, enabling tool lists and other relevant information for specific production jobs to be down-loaded directly to the presetter. After the tooling assembly measurements are completed, the presetter can generate the data in a CNCcompatible format, then DNC down-load to the designated machine tool, removing the necessity for the machine tool setter/operator to input the tooling and cutting data into the controller, enabling production to begin as soon as the tools are loaded into the tool

38 ‘Vendor rating’ (VE), is a basic form of ‘Supply-chain management’ by an organisation and is normally used in purchase decision-making. In VE, this evaluation process is formalised to provide a quantitative measurement of ‘Vendor Quality’ (VQ). Therefore, VR is primarily meant to impart an overall rating of a particular vendor for use in: reviewing, comparing and selecting vendors – this procedure being an integral part of a rigorous purchasing process and in some instances is utilised instead of acceptance sampling. NB Often, it will be difficult to simply create a single numerical quality/rating score, due to the different factors which must be taken into account. Some form of ‘weighted-point’ VR plan, based upon the companies prioritised needs from individual suppliers, this allows for comparisons between different competitive suppliers.

. Figure 137. Typical cutter screen displays from an optical tool presetting machine. [Courtesy of E. Zoller GmbH & Co. KG]

256

Chapter 6

storage magazine. The latest tool presetting machines equipped with a full suite of tool management features and functions, can play a big role in improving shop: productivity/component quality, tool life, inventory control, whilst minimising down-time, reducing component cycle times and part scrappage.

Mounting and Adjusting Milling Cutters Possibly the most crucial cutter body to correctly mount and adjust, for the individual cutting inserts, is that of a side-and-face cutter (Fig. 138). The reason why it is important to set the cutter assembly up correctly, is that invariably the width of the slot in the machined workpiece is identical to that of the respective rotating face widths of the cutting edges. Moreover, whole cutter assembly must ‘run true’ as it rotates on its arbor39 – with no discernible ‘wobble’ – as this effective ‘wobbling’ will influence the machined slot geometry. At its most extreme, some of these special-purpose slotting cutters can be >2 tonnes in weight and larger than 1.5 m in diameter, having segmented cartridges that are precisely and accurately fitted onto the periphery of the cutter body. As a general ‘rule of thumb’ , most of these types of slotting cutters are used to machine component features to a depth of four times their slot width40. If a deeper slot is required, then the cutter has to be ‘optimised’ in some way. Perhaps by using a smaller width cutter than that required for the component’s slot width and, if possible, cutting each slot face separately and eventually taking it to the desired width/depth – arbor interference permitting. Mounting cutting inserts in the case of the staggered-toothed side-and-face cutter body shown in Fig. 138, is relatively straight forward, due to the lateral adjustment available by the splined cartridge seatings. Here, it is important to ensure that the insert seat is thoroughly cleaned prior to commencing fitment. Moreover, ensuring that the contact against the bot-

39 ‘Arbor’ , is the workshop term used for the extension from the machine tool’s spindle that the slotting-type cutter is located and driven from. It can be cantilevered – termed a ‘stub-arbor’ , or supported at its free-end, by an arbor-support – normally fitted with adjustable and suitable matched-bearing diameters. 40 When full slotting, using a side-and-face milling cutter at 40% of the maximum radial cutting depth, a typical feed per tooth would be around 0.25 mm tooth–1.

tom face of the seat occurs, prior to tightening the set screw – normally to a final torque value of 5 Nm (i.e. illustrated in Fig. 138b). Each set screw should be lubricated with the recommended lubricant before reuse. In order to ensure that each cutting insert runs true, the slotting cutter, or face mill assembly, should be correctly mounted – in the former case, onto the arbor, the latter into the correct spindle nose taper – being held on a suitable presetting machine. The whole assembly is then rotated to ensure that each cutting insert is both radially and axially positioned, thereby ensuring that no edges ‘stand-proud’ of each other and at the same time confirming that no discernible ‘wobble’ in the rotating assembly occurs (i.e see the deep-slotting cutter, held in a stub arbor with support, allowing the whole tooling assembly to be rotated and each cutting insert to be inspected/measured, in Fig. 139). Although cutter keyways are not strictly-speaking a mounting problem, the subject does need to addressed, as if the cutter’s diameter and its associated driving keys are not considered, this will limit the overall milling performance of the cutter. With most slotting, and side-and-face cutters fitted to arbors, they normally require a keyway/key for rotational driving purposes for the whole cutter assembly. Usually cutters that are φ140 mm with insert sizes of between 11 to 14 mm, they would frequently need two keyways41. Cutter diameter and driving key limitations, are determined by the cutter’s bore and its connected keyway, together with the DOC being limited by several factors: the arbor OD, its mechanical strength, plus any deformation of the driving key(s). For vertical slotting applications, mounting the cutter on an large diameter arbor with the minimum of overhang is desirable. If the feed per tooth can be reduced – assuming component cycle-times will allow – then this will reduce the tendency of key deformation during milling. Milling calculations and key strength, can be obtained from the following expressions and are valid for new cutting inserts:

41 Keyway positioning for two keys – is usually given by the distance between them as: 180° minus half the peripheral pitch of adjacent cutting inserts – as shown in the diagrammatic sketch in Fig. 138a.


Torque (T) = P [kW]/n [rpm] × 60,000/2π [Nm] [N] Force (F) = T/d [mm] × 1,000 Shear [keyway] stress (τ) = F/area = F/A × E [N mm–2].

257

NB As the cutting inserts wear, the above values will increase by approximately 30%, therefore, it is usual to add a ‘safety factor’ to the key(s) material shear strength, by multiplying this value by 1.5.

. Figure 138. The correct mounting and setting of a cutting inserts in a staggered-toothed side-and-face cutter body. [Courtesy of Sandvik Coromant]

258

Chapter 6

. Figure 139. Cutting inserts for large diameter cutters require pre-setting to minimise any run-out. [Courtesy of Starrag Machine Tool Co. and Sandvik Coromant]


If special-purpose applications are required, such as when form milling the ubiquitous ‘Vee-and-Flat’ configuration for an conventional engine-/centre-lathe bed, ‘gangs’42 of: side-and-face, angled- and helical-cutters are deployed to form and generate these slideways. Here, it is important to ensure that when presetting the cutters on the tool presetter, that the whole cutter assembly is held in the exact manner that they will be utilised when ‘gang-milling’. This ‘gang-milling’ setup, allows their dimensions and forms to be inspected/ measured, while slowly rotating the whole assembly. If two ‘helical cutters’ 43 are utilised in a ‘gang-milling’ operation, then their helices should be of the same pitch, but of different ‘hands’ (i.e. left-ward and right-ward respectively), as this arrangement will balance-out any end-thrust due to opposite cutter helices. Setting up ‘Long-edge milling cutters’ – these are sometimes termed ‘Porcupine cutters’ (i.e. see Fig. 124 – centre), which are normally required for the heavier and longer cutting applications, is quite a complex presetting process. As the individual cutting inserts must be slowly rotated to ensure that axial and radial run-out values are kept to a minimum. Otherwise those inserts ‘standing-proud’ of the remainder will suffer from greater wear rates, thereby prematurely reducing the cutter’s effective life quite significantly while milling an unwanted step into the machined sidewall. On standard face mills, ‘face run-out’ can be as high as >50 µm, so when close tolerances and good milled surface texture is mandatory, then extreme care must

42 ‘Gang-milling’ , is a complex forming process utilising two, or more milling cutters adjacent to one another. So, a sideand-face cutter, located directly together with a helical cutter, represents a ‘gang’ in its simplest form. This ‘gang’ of cutters, is normally permanently mounted together for re-grinding and tool presetting – this is assuming that the cutting edges are not made-up from a series of strategically-positions indexable inserts (see Fig. 76). NB ‘Straddle milling’ , should not be confused with ‘gangmilling’ , as here, it is normal to use two side-and-face cutters with spacing collars between them – of a specific and known dimensional size. Therefore, the cutters ‘straddle’ the part – hence its name – while they machine two faces at the required distance apart in one pass along the workpiece. 43 ‘Helical cutters’ , are sometimes known as ‘Slab-mills’ , having either a left-, or right-hand helix, which ensures that the length of cut and its shearing mechanism are reduced by a ‘quick-helix’ , which is necessary for the milling of more ductile materials.

259

be taken when presetting such tooling assemblies. In order to assist the presetting of such tooling on some face mills, ‘barrel screws’ allow fine adjustment to the cutting insert (Fig. 140a). Such ‘barrel screw’ designs are quite simply-designed, but surprising effective in both adjusting and retaining the cutting inserts, the following remarks explain how they are designed and their method of operation. ‘Barrel screws’ (Fig. 140a), are hardened to resist deformation and have a blackoxide finish to minimise corrosion. To prevent them from shifting during a face milling operation, a nylon pellet is embedded in the thread of the ‘barrel screw’. Right-hand cutting inserts use left-hand ‘barrel screws’ and vice versa, as this counter-acting rotation keeps the insert locked firmly in its pocket. The mating surface of a ‘barrel screw’ is reamed produce a minimum contact of 120° occurs, which ensures accuracy and precision, while minimising wear. The ‘barrel screw’ hole is off-set toward the reamed surface, to provide positive contact with the mating surface throughout the range of adjustment of this screw. It should be noted, that these ‘barrel screws’ cannot adjust the effective ‘gaugelength’ of the tooling, as the amount of adjustment is limited by the position of the cutting insert’s clamping screw. The face-milling cutting inserts shown in Fig. 140a are tangentially-mounted, offering considerable support and additional strength to the cutting edge. When presetting the face mill’s cutting edges when the cutter body is equipped with ‘barrel screws’ , the following procedure should be adopted: • To adjust the insert outward – leave the cutting insert tight and simply turn the ‘barrel screw’ to move the insert to the desired setting. NB Adjustment to the cutting insert’s position, should only be made in one direction only.

• To adjust the insert inward – loosen both the cut-

ting insert and ‘barrel screw’ , push the insert inward, then tighten the insert’s screw and adjust out again to the desired position.

In Fig. 140b, the simple ‘flow-chart’ highlights why it is important to keep any face milling cutter insert’s runout to a minimum. If the run-out of both the minor and peripheral cutting edges is large, then this can create several undesirable problems for the tooling assembly, including: • Poor surface finish – if a cutting insert ‘standsproud’ of the others in the face mill, then it will cut

260

Chapter 6

. Figure 140. Cutting inserts need to be precisely and accurately seated in their respective pockets of the cutter body, to eliminate potential run-out


•

•

in a similar fashion to that of a fly-cutter, creating a periodically scored surface – after each cutter revolution – degenerating the milled surface texture, Chipping due to vibration – as all of the inserts are not set the same, then the most prominent one will take the largest cuts on both the minor and peripheral cutting edges, causing shock loading as the cut is engaged, thereby increasing cutter vibration and potential thermal effects44 creating the likelihood of chipping here on the most exposed cutting inserts, Rapid growth of wear – because of a prominently set and poorly positioned cutting insert in relation to the others in the cutter body, it will absorb the greatest cutting loads, which will lead to shortened tool life, this being exacerbated by pronounced vibrational tendencies, resulting from unbalanced cutting forces and torque. NB All of these factors will contribute to a shortened cutter life.

Conversely, if the face milling cutter’s insert run-out is small, then a good surface finish and stable and predictable tool life will result.

Mounting and Adjusting Single-Blade Reamers The cutting head of a single-blade reamer was previously illustrated in Fig. 74a. The replaceable blade is positioned longitudinally by a blade end stop and

44 ‘Thermal fatigue’ , can be present when cutting is interrupted – as is the case for milling with a prominently exposed cemented carbide cutting insert. Numerous cracks are often observed at 90° to the cutting edge and are often termed: ‘Combcracks’ – due to their visual appearance to that of a typical hair-comb. These cracks, are the result of alternating expansion and contraction of the surface layers as the cutting edge is heated during cutting, then cooled by conduction into its body during intervals between cuts. This very fast alternating heating and cooling cycle, develops the cracks normally from the hottest region of the rake face – this being some distance from the cutting edge, which tends to spread across this edge and down the insert’s flank face. Once these cracks become quite numerous, they can join up and promote partial tool edging to break away – creating cutting edge chipping. NB Today, many cemented carbide tooling manufacturers use structures and compositions that are less sensitive to thermal fatigue, moreover, coatings also play a significant role in reducing thermal fatigue effects, when milling.

261

diametrically adjusted using the front and rear adjusting screws. The blade is micro-adjustable over a limited range of radial movement and can be preset in a special-purpose setting fixture (Fig. 141a and c), to ream the desired diameter that the tool can then consistently produce. This reaming blade normally has a back taper of: between 0.01 to 0.02 mm over a linear distance of between 10 to 25 mm, respectively – when positioned in the pre-setting fixture (Fig. 141b shows a three-guide pad designed single-blade reamer). A feature of the blade’s adjustment, is that it can be reset to compensate for any subsequent blade wear. A clamp, plus two clamping screws securely holds the blade in place, with the wedge-type clamp providing support along the entire blade length (Fig 74a). In the case of the single-bladed reamer design illustrated in Fig. 74a, the blade is located and positioned in the reaming head at an 12° positive rake angle. For this type of reamer design, additional standard blades can be fitted, offering both 6° and 0° rake angles. Taper reaming setting can be achieved by mounting the taper reamer (i.e a taper reamer is shown reaming a component feature in Fig 73b), into the special-purpose setting fixture (Fig. 141c). At least two dial-, or electronic-indicators are positioned along the blade’s length, then adjusted so that a very light pressure is applied to the cutting edge of the blade – to prevent it from inadvertently chipping. With the blade ‘semi-clamped’ , adjustment is made so that its is parallel along its length – relative to the tapered guide pads. Once the blade has been ‘fully clamped’ , adjustment occurs to position it higher than its guide pads’ diameter, by between 10 to 20 µm – all along the blade’s length, which achieves an accurate setting, but this setting will depend on both the workpiece material and the prevailing machining conditions45.

6.5.5 Tool Store and its Presetting Facility – a Typical System In the worst case scenario, for many of the ‘old-style’ traditional workshops, the tools are as often as not

45 Taper reamers – typical machining details: Cutting speed 4 to 20 m min–1 (Stainless steel 2 to 6 m min–1 ), Feed 0.2 to 0.8 mm rev–1, Machining allowance 0.2 mm and up to 0.5 mm – for large taper reamers, plus Coolant soluble oil @ 10% dilution.

262

Chapter 6

. Figure 141. Presetting equipment and ‘guidelines’ for the setting of single-bladed reamers


treated with almost contempt, until they are required for a repeat order, or utilised for a new machining operation. Here, when the tools are not in use, problems arise because of the following: • Wasted and poorly utilised space – tools are kept either simply ‘floating around’ in an ‘ad hoc’ system, or are just sitting on top of each other – damaging the precision-ground tooling surfaces – while being kept in the open, • Time is lost looking for tools – due to the lack of any form of tool management, tools must be looked for located, then assembled – under less than ‘ideal conditions’ – so the minimum of tool control occurs, • Lack of efficiency – as a result, more effort is required by personnel who must try to find tools, changing front- and back-ends to suit the selected machine tool for the production run causing leadtimes to lengthen. Considering that up to 10% of the machine tool cost is tied-up in the purchase of tooling, then tools need to be looked after with some degree of care and attention. An area adjacent to the workshop, should be set aside for the purposes of storing tools and associated equipment in purpose-built tooling cabinets (Fig. 142) and form a basis for tool presetting activities. There are several advantages in utilising telescoping drawer cabinets for tool storage (Fig. 137a and b), these include: • Increased tool storage density – allowing the drawer space to be completely efficiently filled-up in the minimum of space, • All tooling is kept under cover – thereby avoiding tool damage, while keeping tools dust and debrisfree, • All tools and ‘associated paraphernalia’ are readily to hand – tooling drawers can be completely open-up, so that the contents are both easy to see and to arrange, • Tooling components fit into their respective places – the drawers appropriately sectioned, to totally enclose the tools and their component parts, making it impossible for them to drop out and be damaged, • An organised and practical tool management system is achieved – within the tool store and presetting facility. Therefore, all of the tooling components can be classified and categorised in the respective drawers – with every tooling part clearly seen.

263

As a result of this efficient tooling component layout and tool-kitting facility, consumable tools can be ‘tracked’ within the tool store, but when these consumables leave the store vicinity (Fig. 142c), they are to all intents and purposes ‘scrapped’ as far as the tool stores are concerned. Returnable tools destined for eventual re-use can be ‘tracked’ around the machine shop by a number of tool-identification means (i.e as seen in Figs. 116 and 117 – using ‘tagged tools‘), or at the most basic of identification levels, by judicious labelling, or bar-coding of tools, with feed-back of information to the tool stores. Often, for the most frequently used tools, they are assembled into a composite forms, then issued as ‘grouped-tools’ in the form of kits for a specific job (i.e see Fig. 142c). Sometimes, tools are individually issued to machine tool setters/operators and are not assembled into ‘kits’ , under such circumstances, the stores will keep a record to show: to whom they were issued, the machine tool on which they will be used, the number and identification of these tools, together with the date of issue. A major benefit of creating an area set-aside close to the machine tools for dedicated tool management, is that tool kits can be made-up ahead of the time – this being dictated by master schedule, so that they are ready just-in-time before any machining commences. A result of this timely tooling strategy, the lead times46 are reduced, which is of prime importance to a company in a competitive fast-developing market. Often, it is the case that all tooling is assembled in the tool stores /preparation facility (Fig. 142b). This advanced preparation means that metrology-based

46 ‘Lead times’ , refer to the time taken before manufacture of the part, or prior to a production run beginning. These ‘times’ are dependent upon a range of interrelated factors, such as: component stock quantities and their availability, the machine tools that are available, plus ‘line-balance’* factors, etc. *‘Line of balance’ (LOB), refers to a technique which permits the calculation of the quantities of the particular activities, or components which must have been completed by a particular intermediate date, in order that some final delivery schedule might be satisfied. Therefore, in this instance, it can be considered as a machine tool scheduling and a control technique. In most of the LOB activities undertaken concerning machine tools, plant utilisation levels are paramount and if possible, a smooth and consistent LOB across all of the production machines is desirable – for both high efficiency and consistent work-throughput.

264

Chapter 6

supplies such as ‘limit gauges’ 47, together with jigs and fixtures , are also the responsibility of the tool stores personnel, being despatched with the tool kits for a particular production run. When this is the case, total packages are issued, containing: cutting-tool and work-holding kits, plus the limit gauges necessary for metrological checking/inspection. A tool store and prepartion facility with responsibility for all of these tooling-related aspects, becomes a ‘focal point’ for the machine shop for all matters relating to tooling, whether they are for breaking-down of previously used kits, calibration of gauges, or even purchase requirements for the latest tool available. Specialist personnel in the tool stores/preparation facility, have considerable responsibility in servicing tooling for the overall manufacturing resources on these machine tools and as a result, will have amassed a large working-knowledge of the production tooling requirements, so their opinions should be sought prior to any purchasing decisions on new tooling.

6.5.6 Computerised-Tool Management – a Practical Case for ‘Stand-alone’ Machine Tools If one considers the tooling requirement for a standalone machining centre, then today the market de-

47 ‘Limit gauges’ , are based upon the stated International Standards agreed for ‘Limits and Fits’ for component tolerancing. The use of limit gauges, is a form of ‘attribute sampling’* where no attempt is made to determine the size of the workpiece toleranced feature, but they are simply utilised to establish whether the component’s critical dimension is within the specified limits of size, or not. In practice, a component that has hole that has simply been drilled and reamed, might require a double-ended plug gauge, with one ‘Go’ end of the plug gauge being of full form and checking the maximum material condition and as many dimensions as possible, with the ‘Not go’ end checking the minimum material condition and only one dimension – which as its name implies, this latter end should not go into the reamed hole. This limit gauging technique fulfils ‘Taylor’s Theory of Gauging’. *‘Attribute sampling’ techniques, are a means of sampling imperfections that are not in the strictest sense, measurable quantities. For example, a mirror-surface that has been produced, might be scratch-free, or may have other blemish marks, these factors might be cause for its rejection. Hence, ‘attribute sampling’ can be considered as a two-way classification system for either acceptance, or rejection of the workpiece.

mands for its manufactured products has become much more diversified, with the number and multiplicity of tools required having also increased. As has been shown previously in this chapter in the preparation for workpiece machining, confirmation that all the tooling – including spares (i.e. ‘Sister-tooling’ – see Footnote 1 in this chapter), must be loaded into the tool magazine. A computerised-tool management system (Fig. 143a), eliminates the possibility of mistakenly selecting the wrong tool and exacerbating the situation of placing it in the incorrect tooling pocket in the tool magazine. With these ‘tagged tools’ (Fig. 143d) having non-contacting read/write embedded microchips in say, the pull-stud region of the assembled tool – allowing coolant-through-spindle applications. The following tooling data can be automatically registered into the CNC memory (Fig. 143b), such as tool: number and its ID number; name and the nominal diameter; length, plus its ‘working-diameter’; thrust and power coefficients; interference data, with ‘large diameter’ tool data – if required; life accumulated/actual usage time, wear and breakage flags. Many quite complex and sophisticated computerised-tool management systems exist, but essentially the practical system depicted in Fig. 143, can be usefully applied to a machining centre in an efficient production environment. Here, the system comprises of three modules, these are: • Tool module – this being the smallest component of the tool management system, where the tooling data is both read/written at the machine magazine tool loading/unloading position (Fig. 143c). The tool data is automatically registered in the CNC memory at the push of a button, with this toolingrelated data being continuously updated – as machining continues, • Tool Management module – once this is combined with the ‘tool module’ (above), then tool management is conducted on a much larger scale, allowing not only all of the previous tooling data to be monitored and controlled, but additional information on the: toolholder’s bill of materials, insert inventory, and its location, together with a graphic tooling display of its build-up and information regarding the correct procedure to ensure fast and error-free tool measurement on the presetting machine (Fig. 143a), • Tool transportation module – consists of a tool transporter robot having high positional accuracy (i.e. not shown), which automatically transports


. Figure 142. An integrated tool and data management system – for complete tooling deployment. [Courtesy of Susta – Tool Handling]

265

266

Chapter 6

. Figure 143. A computerised tool management system for an FMS facility. [Courtesy of Yamazaki Mazak]


the tools to-and-from each machine – changing the ‘stand-alone’ machine into a simple form of flexible manufacturing system (FMS). It consists of a ‘tool hive’ stocker, which acts to store tools in a centralised and convenient location within the shop, having a tool tarnsport controller – controlling the overall system’s operation. Such tool management systems are becoming quite common-place in many highly-utilised manufacturing environments around the world of late, which considerably increases the overall utilisation rate of these automated CNC machine tools.

References Journals and Conference Papers Chandler, B. Crib Control [Organising Tooling Workplace]. Cutting Tool Eng’g., 48–53, Sept. 1999. Gough, J. Going Hell for Leather [High-speed Milling]. Manufact. Eng’r., 13–15, 1991. Gust, C. Don’t Fret [How to Maintain Toolholders]. Cutting Tool Eng’g., 53–55, July 2003. Hanson, K. Ready, Preset, Go [Tool Presetting]. Cutting Tool Eng’g., 30–38, June 1999. Hanson., K. Not so Fast [Quick-change Tooling]. Cutting Tool Eng’g., 52–55, June 2002. Ingersoll Pub. Facing off with Stainless [Tangentiallymounted Face Mills]. The Cutting Edge, 8–9, No.1, 1989. Ingersoll Pub. Engineering Success [Large Special-purpose Slotting Cutters]. The Cutting Edge, 15–16, No.4, 1988. Ingersoll Pub. Special Tools solve Special Problems [Barrelscrew Adjustment of Cutting Inserts in Face Mills]. The Cutting Edge, 14–15, No.1, 1990. Ingersoll Pub. Milling Cutters: Milling Cutters [Production of Face Mills]. The Cutting Edge, 10–11, No.1, 1991. Johnson, B. Reaming Technology. Industrial Tooling Int. Conf., Southampton (UK), Molyneux Press, 132–152, Sept. 2001. King, D. If you’re not Presetting, you’re not Cutting It. Cutting Tool Eng’g., 32–34, Jan. 1994. Layne, M. On Balance [Milling Tooling/Spindles]. Cutting Tool Eng’g., 52–55, June 2002. Lewis, D.L. Factors for Successful Rotating Tool Operation at Hi-speeds. SME High-speed Machining Clinic (Raleigh, NC), April 1991. Linn. T. Fundamentals of Precision in High-speed Milling. Precision Toolmaker, 306–310, Oct. 1989.

267

Mason, F. Video Presetting: Reliable and Safe. American Machinist, 36–38, Aug. 1991. Mitchell, W.A. When and Where to Think High-speed Milling. Modern Machine Shop, 55–63, Oct. 1991. Park, H. and Little, T.A. Assessing Machine Performance. American Machinist, 39–42, June 1992. Renishaw plc. Innovative Laser Tool Setting Technology provides Accuracy, Flexibility and Robust Operation. Renishaw ‘White paper’ , 1–10, 2003. Richter, A. Squeeze Play [Shrink-fit Tooling]. Cutting Tool Eng’g., 34–38, Sept. 2000. Side and Face milling – Application Guide. Sandvik Coromant Pub., 1996. Sprow, E. High-speed Spindles for Milling? Tooling and Production, 32–35, Mar. 1991. Smith, G.T. Ultra-high Speed Machining – The Problems Associated with Spindle Designs, Tooling and Machined Parts when Milling. Proc. of FAIM Int. Conf., CRC Press Inc. (Florida), 947– 961, 1992. Smith, G.T. The Assessment of Machining and Turning Centres using Artifact-based Techniques, such as the Ballbar. Proc. of: Laser Metrology and Machine Performance, Computational Mechanics Pub., 307–316, 1993. Zoller, C. From Presetting to Tool Management. Cutting Tool Eng’g., 35–38, April 1998. Books, Booklets and Guides Ashby, M.A. Materials Selection in Mechanical Design. Butterworth-Heinemann Pub., 2001. Barnett, H. Operations Management (2nd Ed.). Macmillan Press Ltd., 1996. Bergman, B. and Klefsjö, B. Quality – from Customer Needs to Customer Satisfaction. McGraw-Hill Book Co., 1994. Blanchard, B.S. Logistics Engineering and Management – 4th Ed. Prentice-Hall Inc., 1992. Galyer, J.F.W. and Shotbolt, C.R. Metrology for Engineers (5th Ed.), Cassell Pub, Ltd, 1990. Haberberg, A. and Rieple, A. The Strategic Management of Organisations. Pearson Education Ltd, 2001. Kief, H.B. and Waters, T.F. Computer Numerical Control. Glencoe Pub. 1992. Modern Metal Cutting. Sandvik Coromant Pub., 1994. Oakland, J.S. Statistical Process Control – A Practical Guide. William Heinemann Ltd, 1986. Ránky, P.G. The Design and Operation of FMS. IFS Pub. Ltd., 1983. Renishaw Ballbar Diagnostic Manual. Renishaw plc Pub., 2007. Sandvik Pub. Modern Metal Cutting: Part 7 – Milling Tools. 1980.

268

Chapter 6

Shotbolt, C.R. Workshop Technology – Book 1. Cassell and Co. Ltd, 1973. Slack, N. et al. Operations Management. Pitman Publishing, 1995. Smith, G.T. Advanced Machining – The Handbook of Cutting Technology. IFS/Spinger Verlag, 1989. Smith, G.T. CNC Machining Technology. Springer Verlag, 1993. Timings, R.L. Manufacturing Technology – Level Two. Longman Group Ltd, 1979.

Trent, E.M. Metal Cutting: 2nd Ed. Butterworths & Co. (Pub.) Ltd, 1984. Valentino, J.V. and Goldenberg, J. Introduction to Computer Numerical Control (CNC) 2nd Ed. Prentice-Hall, 2000. Vickers, G.W., Ly, M.H. and Oetter, R.G. Numerically Controlled Machine Tools. Ellis Horwood Pub., 1990. Wild, R. Production and Operations Management (5th Ed.). Cassell Educational Ltd., 1995.

7

Machinability and Surface Integrity ‘It is common sense to take a method and try it. If it fails, admit it frankly and try another. But above all, try something. ’

FRANKLIN DELANO ROOSEVELT

(1882 – 1945) [32nd President: United States of America]

270

Chapter 7

7.1 Machinability Introduction – an Historical Perspective Today, greater emphasis is being placed on a component’s ‘machinability’ , but this term is an ambiguous one, having a variety of different meanings, depending upon the production engineer’s requirements. In fact, the machinability expression does not have an authoritative definition, despite the fact that it has been used for decades. In 1938, Ernst in his book on the ‘Physics of Metal Cutting’ , defined machinability in the following manner: ‘As a complex physical property of a metal involving: • True machinability, a function of the tensile strength, • Finishability, or ease of obtaining a good finish, • Abrasiveness, or the abrasion undergone by the tool during cutting.’ By 1950, Boulger had summarised these criteria more succinctly in his statement: ‘From any standpoint, the material with the best machinability is the one permitting the fastest removal of chips with satisfactory tool life and surface finish.’ This ‘Boulger definition’ leaves some unanswered questions concerning chip-forming factors, cutting forces and, has little regard for either the physical and mechanical properties of the material, nor potential sub-surface damage caused by the cutting edge. By 1989, Smith made the point that in fact machinability, had to address these properties and the word ‘metal’ should be substituted by the expression ‘material’ , in a combined general-purpose definition, as follows: The totality of all the properties of a work material which affect the cutting process and, the relative ease of producing satisfactory products by chip-forming methods.’ Even these definitions still lack sufficient precision to be of much practical use and by 1999, Gorzkowski, et al., in their powder metallurgy paper concerning ‘secondary machining’, entitled:

‘Secondary machining’ , is a term used to cover any additional post-machining operations (e.g. drilling, turning and milling, etc.), that has to be undertaken on powder metallurgy (i.e. sintered’) compacts, after compaction and sintering. Normally, these post-sintering production processes, are only carried out to ensure, say: a good turned registered diameter, a precision cross-drilled hole, precise and accurate screwthread, an undercut, or similar* – as this is a last resort, as it addsvalue to the overall component’s cost.

‘Machinability’ , stated that: ‘Machinability is a difficult property to quantify.’ Why is this so? It is probably is a combination of many inter-related factors, such as: chemical composition of the workpiece, its microstructure, heat-treatment, purity, together with many more effects which influence the overall machining operation. In Fig. 144, this diagram attempts to highlight some of the important factors that affect a component’s machined state – its ‘machinability’. Although even here, an important factor such as power consumption is missing, showing that this is by no means an exhaustive flow-chart of the complex mechanisms that exist when a material is subjected to machining. This is probably why it is virtually impossible to state that one, or another material after machining, was either a ‘good’ , or bad’ one to machine. By utilising some ‘impartial and objective testing program’ , it may be possible to ‘rank’ prospective or current materials, or production tools – in some way, perhaps by way of a ‘Design of Experiments’ (DoE), in combination with ‘Value Analysis’ (VA) approach to the production problem. This strategic technique to the problems of ‘machinability comparisons’ of differing factors will shortly be mentioned in more detail, after a brief resumé on just some of the machinability testing techniques favoured today.

7.1.1 Design of Machinability Tests and Experimental Testing Programmes Over the years, a range of machinability testpieces have been developed – more on this shortly – that are used to assess specific cutting conditions found when machining the actual production part. The assessment of a material’s machinability can be undertaken by two groups of tests, these are machining and non-machining testing programmes. The former machinability group, can be further sub-divided into either ‘ranking’ and ‘absolute’ tests and, it should be mentioned that the latter non-machining tests fall into the ranking category. Often, ‘ranking’ tests are termed ‘short

*Powders when they fill the dies and are compacted, cannot reproduce component features at 90° to the major pressing direction – hence, the powders cannot readily move sideways – as such, features, like: screwthreads, transversal features (i.e. undercuts, etc.), must be machined afterward, hence, the term ‘secondary machining’.

Machinability and Surface Integrity

. Figure 144. The major factors that influence a machined component’s condition

271

272

Chapter 7

tests’ , conversely ‘absolute’ tests are known as ‘long tests’. By their very nature, the ‘short tests’ merely indicate the relative machinabilities of two, or more different combinations of tool and workpiece. Whereas, the ‘long tests’ can produce a more complete depiction of the anticipated conditions for various combinations of tool and workpiece, but as their name suggests, they are more time-consuming and costly to develop and perform. Some of these test regimes are briefly reviewed below, but more information can be obtained from the listed references at the end of this chapter.

‘Ranking’ Machining Tests A series of these ‘ranking’ tests for fast assessment of actual production conditions has been devised over the years and some will be mentioned below, but this is by no means an exhaustive account of all such testing programmes, they merely indicate the relatively welltried-and-tested techniques, such as: • ‘Rapid facing test’ – this consists of a turning operation, requiring facing-off a workpiece, preferably having a large diameter, using an HSS tool . The machinability is assessed by the distance the tool will travel radially-outward, from the bar’s centre, prior to its catastrophic tool failure. This ‘endpoint’ as it is known, is compared with a similar trial, where the distance for tool failure by using a reference material was previously determined, NB Although the ‘Rapid facing test’ quickly assesses one particular test criterion that a machinability rating can be based upon, it suffers from a number of limitations. Firstly, the material’s diameter may be smaller than that which one would ideally prefer to use for the test. Secondly, if the workpiece material’s structure is not homogeneous, then this test only indicates properties over the diameter-range

‘HSS tool material’ is utilised, because under these extreme machining conditions, it will rapidly promote catastrophic tool failure as the forces steadily increase together with escalating tool interface temperature, as the tool’s edge is fed radially-outward during the subsequent facing operation.

‘Reference materials’ , are normally those workpiece materials that are considered to be ‘easy-to-machine’ , as their name suggests they, at the very least, give a ‘base-line’ , or datum, for some form of machinability comparison.

‘Homogeneity of material’ , refers to a uniformity of its microstructure and having isotropic properties.

used. This latter problem of lack of homogeneity of the workpiece material, can be somewhat lessened by boring-out the material at the workpiece’s centre, prior to commencing the test.

• ‘Constant-pressure test’ – this is quite a popular

testing technique and can be undertaken by a variety of methods of machining assessment. For example, in turning, machinability is measured by utilising predetermined geometry in association with a constant feed force. The technique has been used to some effect on the machining of free-cutting steels. This test is essentially a measure of the friction between the chip and tool, which is related to the specific cutting temperatures generated whilst machining, together with its effects on the tool’s wear-rate, NB Normally a turning centre has a constant feed force, in order to obtain relevant data. An engine-/ centre-lathe can also be employed to acquire identical data, but a tool-force dynamometer is used to measure this feed force, then plotting a graph of this feed force with its associated frictional effects, but this requires more effort and takes longer. Similar constant pressure tests can be employed for drilling processes.

• ‘Degraded tool test’ – consists of workpiece ma-

chining with a softened (i.e. degraded) cutting tool. The test’s ‘end-point’ is determined either: when a specified amount of tool flank/crater wear has been reached, or at catastrophic tool failure, NB If machinability testing is carried out on softer and more easy-to-machine materials – typically on various alloys of brass, then just a small variation in softening the tool steel prior to cutting, has a drastic effect on the results obtained, but for harder-tomachine materials this effect is significantly lessened.

• ‘Accelerated cutting-tool wear test’ – as an alterna-

tive to deliberately softening the tool (i.e. Degrading tool test), in order to speed-up the machinability process the cutting speeds are increased. If the cutting speeds are significantly increased, the tool will not behave according to the predictable tool life


273

NB Given the above limitations, these tests have proved to be quite valid and successful for screening a workpiece material prior to actual machining. Typical examples of this test type, rank materials using a V60 scale – giving cutting speeds in m min–1 and the machinability index of 100 (i.e. utilised by the ‘Volvo test’ – not shown). A close correlation between the chemical composition test and ‘absolute tests’ has been obtained with accuracies claimed to within 8%. For example, the relationship between chemical composition and cutting speed is: Cutting speed (V60) = 161.5 – 141.4 × %C – 42 – 4 × %Si – 39.2 × %Mn – 179.4 × %P + 121.4 × %S.

equation – due to the artificially-elevated cutting temperature generated. NB It is not prudent practice to extrapolate toollife data beyond that actually obtained during testing in order to obtain quantitative information about other ranges and conditions, with differing operations and parameters. As a result, this test is usually classified as a ‘ranking test’.

‘Ranking’ – Non-Machining Tests Whenever there seems to be a need to experiment with material cutting using perhaps one of the techniques just mentioned, it is important to establish whether any savings gained will be recouped in the actual production operation. If a company is unsure of the likely cost benefits of such testing, then a strong case can be made not to test the material at all! Fortunately, nonmachining tests exist that can be utilised in these doubtful situations, rather than ‘working blindly’ – with no relevant cutting data, to base the applied cutting conditions upon. Several of these ‘ranking’ non-machining tests can be employed, such as: • Chemical composition test – a variety of tests have been developed by which workpiece materials are ‘ranked’ according to their primary constituents. It is obvious that the results from such tests are only relevant when materials of similar type, having identical processing conditions/thermal history, are to be machined.

• Microstructure tests – are principally concerned

with the type of microstructure present in say, a steel workpiece, specifically: inclusion type, shape and dispersion. The test method gives a good indication of the likely machinability, but requires highly-specialised laboratory equipment for such a metallographical investigation although materials can only be ranked, as either: good, bad, or indifferent.

NB Early work here, primarily investigated lowto-medium carbon steel microstructures, notably considering the spacing between pearlite laminae achieved by heat treatment. The pearlite-to-ferrite proportions clearly influenced the materials hardness value (e.g. Brinell). When a cutting speed was selected (e.g. V80), a machinability rating could be obtained for either life at: a constant speed (minutes), or relative speed for a constant tool life (m min–1). It has been observed that when >50% pearlite was present, combined with a relatively high bulk hardness, then good machining characteristics occurred. In recent years, commercially-available steels have trace elements added to aid machinability, the so-called free-machining steels. Typically, sulphur and manganese additions, create manganese sulphide, with their shape, size and distribution within the steel’s matrix, playing a major role in aiding machinability factors.

Taylor’s tool life equation(s), has been utilised for many years, to determine the ‘end-point’ of a cutting insert’s useful life, under steady-state cutting conditions. The basis of the general formula: VcTα = C, has been modified and expanded to obtain an equation for the ‘economical cutting-edge life’ for a specified feed, as follows: Te = (1/α – 1)(C´t/C´m + tc) Where: Te = economical tool life, α = slope of the VT-curve (i.e. measured from a plotted graph), C´t = cutting-tool cost per cutting edge (i.e. see ‘Machining costs’ – later in the chapter), C´m = machine charge per minute (i.e normally established by the machine shop management), tc = tool-changing time for the cutting operation – this will vary according to whether the tooling is of the conventional, or quick- change type. ‘Thermal history’ , refers to the heat treatment thermal cycle that the component in question was processed, describing the time at temperature, with any modifications to the temperature-induced regime on the heat-treated part.

‘Bulk hardness’ , is a term that is used to state the overall hardness of the test specimen, not its micro-hardness – which only establishes localised hardness levels.

274

Chapter 7

• Physical properties test – requires specialist equipment in order to perform this test. The physical properties of the workpiece material are utilised in order to determine its machinability ranking.

NB Researchers, have produced a general machinability equation using a dimensional analysis technique and, by utilising conventional test methods to establish and measure its: thermal conductivity, harness (Brinell), percentage reduction in area, together with the test sample’s length. This ‘Physical properties test’ , gives close agreement with the V60 cutting speed for a range of ferrous alloys, although when brittle materials are assessed, the lack of a yield-point and the much smaller reductions in area – after tensile testing – may cause potential ranking problems.

‘Absolute’ Machining Tests As their name implies, the ‘absolute tests’ are utilised in order to obtain a comprehensive data-gathering machining-based activity, on particular types of workpiece and cutting tool combinations. Many of these ‘absolute testing’ techniques have been devised, with several of them listed below, including the: • Taper-turning test – being undertaken by turning a tapered workpiece. As a result of turning along the taper, the cutting speed will proportionally increase with increasing taper diameter – this also being in proportion to the cutting time. By originally establishing the cutting speed, the changing-rate of the

‘Yield-point’ , refers to the strain* at which deformation becomes permanent, when the material is subjected to some form of mechanical-working. The yield-point strain for ferrous and many ductile materials is well-defined, illustrating a ‘sharp’ transition from elastic-to-plastic deformation – where a permanent ‘set’ occurs. However, this is not the case for many brittle materials, here when say, a tensile test is conducted, an artificial ‘proof-stress’ value is used to intersect the stress/strain curve plotted, to establish its safe-working level of operation – see the relevant References for more in-depth details. *‘Strain’ , is a measure of the change in the size, or shape of a body – referring to its original size, or shape. For example, linear strain is the change per unit length of a linear dimension – after some form of mechanical working. For a tensile test specimen that has been subjected to a tensile test, it refers to its linear dimensional change from its original gauge length.

cutting speed in conjunction with the amount of tool flank wear – for two separate tests – allows the values of the constants (i.e.‘α’ and ‘C’) in Taylor’s equation for tool wear – see Footnote 5 – to be derived and, the tool life established for a range of future cutting tests. As the DOC must be consistently maintained throughout the test, either a CNC program must be written – using one of the standard ‘canned-routines’ available, or a taper-turning attachment is necessary on an engine-/centre-lathe, NB Some major advantages accrue from this comprehensive testing technique, not least of which is that results are valid for a range of pre-selected cutting speeds and, the test is of relatively short duration, but closely agree with many thorough and longer test methods. Although, the results obtained may not be representative of actual cutting conditions, owing to the fact that the cutting tool, machines at differing diameters throughout the taper turning test.

• Variable-rate machining test – achieves similar

results to the previously described ‘Taper-turning test’. In this case, the increase in cutting speed is obtained by turning a parallel testpiece axially, whilst simultaneously increasing the cutting speed as the tool traverses longitudinally along the workpiece. Once again, the constants are derived for the ‘Taylor equation’ after a minimum of two tests have been completed, NB The main advantages of this method over the ‘Taper-turning test’ , are that a standard testpiece can be used and the results probably reflect truer actual turning conditions – in that consistent diameters are being turned, although this argument is somewhat debased, if the turning of complex freefrom component geometry is demanded for the production part.

• Step-turning test – was developed to overcome

some of the problems associated with the two previously described testing techniques. In the ‘Stepturning test’ method, a range of discrete diameters and speeds are utilised to determine the ‘Taylor’s constants’. This test, shows close agreement with results obtained from the two previously-mentioned ‘absolute test’ methods, • HSS tool wear-rate test – this test assesses machinability by measurement of the tool’s flank wear, pro-


duced when machining free-cutting steels, with the major parameters being the elemental additions to the metallurgical composition of these steel grades. NB These tests are undertaken in a similar manner to the: ISO 3685:1977 Standard, for a long ‘absolute test’ , but it was withdrawn in mid-1984. All of the above ‘absolute testing’ programmes, relate to turning operations, principally due to the fact that the tool is engaged in the workpiece test sample for a reasonably lengthy period of time. This tool/workpiece engagement, allows for ‘steady-state’ conditions to be developed, having the additional benefit of producing relatively consistent ‘Taylor constants’. From a more practical viewpoint, the author has developed some other testpieces, which have proved somewhat useful in actual industrial machining applications, where a more representative machinability situation was demanded. Just some of these testpieces, along with a discussion of their relative merits, will now proceed.

Practical Testpieces – for CNC Applications The premise behind the development of the testpiece depicted in Fig. 145, was to attempt to ‘mirror’ the actual production operations and to a lesser extent, the physical geometry of a particular component part. Here, the component geometry was devised to be machined on either a machining centre, or a turning centre with the facility of driven tooling and at the very least, having an indexing workholding spindle/chuck. With this testpiece, the part is preferably a thick-walled tube that can be bored out, OD turned, circular interpolated (i.e. milled), drilled and tapped – as the drilling size, is also an M6x1 tapping size. This allows the component’s geometric features to be inspected ‘Onmachine’ – using metrological inspection routines in association with touch-trigger probes and, ‘Off-machine’ employing a CNC Co-ordinate Measuring Machine (CMM). These identical parts were from a series of exhaustive tests undertaken on both ferrous metals and aerospace-grade aluminium stock. Of particular note, was that when a milled circular interpolated feature – the boss, was assessed on the machining centre, it gave more accurate readings than its equivalent inspection routine on the CMM. This perceived difference in accuracy and precision, was the result of part changes caused by both relaxation of the clamping forces – upon release – and the greater temperature differential between these workpieces when inspected

275

on the CMM. However of note, was the fact that in general for the inspection of part features, the CMM showed a four times improvement in repeatability, to that of the touch-trigger probing undertaken on the machine tool, as the following Table 9 indicates: The above type of practical ‘testing regimes’ are generally termed: ‘Production Performance Tests’ (PPT). Typically, these PPT’s can be utilised to determine the maximum production rate – in parts per hour. Although it must be said, that with shifts normally consisting of between 6 to 8 hours duration of potential ‘in-cut time’ , this to a certain extent, limit’s the achievable machined surface finish requirement, particularly if a ‘Sister tooling strategy’ is not operated. One of the main problems connected with PPT’s, is that invariably free-cutting metals are usually selected for longterm testing, meaning that any wear-related data takes awhile to accrue. Despite this slight reservation, actual cutting data can be employed, which represents almost optimum machining conditions, leading the way to

. Table 9. A comparison of the machined component testpiece accuracies by either: ‘On-’ , or ‘Off-machine’ inspection procedures PARAMETERS:

MACHINES*: - equipped with Renishaw touchtrigger probes: Machining Centre (Vertical)

CMM (LK CNC Micro4)

Scope

Full range of: X-, Y- and Z-axes

Direction of test

Uni-directional

Positional Accuracy

±13 µm

X-axis ±8 µm Y-axis ±5 µm Z-axis ±6 µm

Repeatability

±10 µm

±2.5 µm

* Machine tools here, are part of a fully-industrial Flexible Manufacturing Cell (FMC), comprising of Cincinnati Milacron equipment: 200/15 Turning Centre, 5VC Vertical Machining Centre, T3 776 Robot- equipped with twin back-to-back grippers – for component loading/unloading, LK Micro4-CMM, DeVlieg Tool Presetting Machine, Component workstation, Cell Controller, all equipped with Sandvik Coromant quick-change tooling (Block Tools and Varilock Tooling), plus DNC-link to a CAD/CAM workstation – being designed and developed by Cincinnati Milacron and the Author, when acting as an Industrial Engineering Professor at the Southampton Solent University.

276

Chapter 7

. Figure 145. General machinability test piece for CNC machine tools. NB Holes marked ‘A, B and C’ are machined at different cutting speeds, as are the turned, bored and milled dimensions.


‘full’ production operational machining, meaning that with some degree of confidence, manufacturing dictates and objectives will be met. In Fig. 146, a commercial (PPT) testpiece has been developed showing typical machining data employed, based upon the secondary machining operations demanded by many companies on Powder Metallurgy (P/M) components – where light finishing cuts, or accurate and precise screwthreads are demanded. Here, the cutting insert can turn three different diameters – usually in some form of arithmetic progression, so that feedrate longitudinally can be metrologically assessed. Moreover, the insert’s passage over the surface can be metallographically-inspected and a microhardness ‘footprint’ across a tapered section can be undertaken, to see if any surface/sub-surface modifications have occurred. More will be said on this subject later in the chapter, when discussing the effects of ‘machined surface integrity’. This design of using a thickwalled tube (Fig. 146), that can be produced from either wrought stock, or P/M compact processing – the latter, giving a controlled ‘density’10 across and along the part, makes it particularly ‘ideal’ for any secondary machining machinability trials. Boring operations can also be conducted on such a testpiece geometry, allowing roundness parameters and its associated ‘harmonic profile’ to be metrologically assessed, in conjuction with any ‘eccentricity’ with respect to the OD and

‘Arithmetic progressions’ , are normally utilised for many applied machining (PPT) trials as they give a ‘base-line’ for the research work and increase at a controlled amount. For example, a feedrate, could begin and increase as follows: 0.1, 0.4, 0.7, 1.0, 1.3, … mm rev–1 – with the ‘common difference’ being 3. As a mathematical expression, this simple arithmetic progression, can be written as follows: a, a+d, a+2d, a+3d, a+4d, a+5d, … where the ‘common difference’ is ‘d’ , giving the: n´th term as: a+(n–1)d.

10 ‘P/M Density’ , refers to either the uncompacted, or free-particulates and is termed its ‘Apparent density’ (AD). This term AD, is used to refer to the loose material particulates prior to PM compacting, to describe the density of a powder mass expressed in grammes per cubic centimetre of a standard volume of powder. This AD differs from that of its ‘compacted density’ – which will vary depending upon the consolidation (i.e. compacting) technique utilised. For example, double-compaction – pressing the powder in the dieset from both ends, or using ‘floating diesets’ – to simulate double compaction, in this latter case, pressing from one end only, will produce a more uniform bulk density throughout the ‘green compact’ as it is known – prior to its subsequent sintering process.

277

ID – these machined surfaces both being produced in a ‘one-hit machining’ operation – then inspected by a suitable roundness testing machine. The main advantage of using industrial-based (PPT) testpieces similar to that shown in Fig. 146, is that ‘canned-cycles’11, can be used to produce the undercuts, turning passes, or screwcutting operations on each part. Moreover, optional ‘programmed-stops’ can be written, allowing the research-worker/operator, to have the facility to stop machining at a convenient point as desired, at the press of a button – giving a measure of control to the automated CNC machining processes. If a series of testpieces are to be machined, it is important that all of the parts machining sequences are known and that they are laid-out in a consequtive logical fashion. This allows one to measure the deterioration with machining time for the sequence of testpieces produced. To this end, not only should some unique and logical part numbering system be used, but in the case of P/M testpieces, the top and bottom for each compact should be established. As when each one was initially compacted, its local density have varied and, for consistency for all machining undertaken with each test piece, it needs to be held in the same orientation. Often it is possible to amalgamate two previous ranking machining test regimes into one, this is the case with ‘Accelerated Wear Test’ (AWT) illustrated in Fig. 147, this test being a combination of both the: ‘Rapid Facing’ and ‘Degraded Tool’ tests – previously described. In the case of the AWT technique, this hybrid test’s aim is to assess the relative machinability of either wrought, or secondary machined P/M compacts

11 ‘Canned-cycles’ , this is a preset sequence of events that is executed by issuing a single command, which may remain active throughout the program, or in this case will not, for a particular ‘canned-cycle’ *. For example, once the preset values/ dimensions together with the required tool offsets have been established, then a preparatory function entitled a ‘G-code’ can be used, such as a G81 code, which would initiate a simple drilling cycle, in association with the following G84 code which would then specify a tapping cycle on this drilled hole, or alternatively, a G32 code commences a threading cycle and so on. – which considerably reduces both the complesaty and overall length of a CNC program. *G-codes fall into two categories, they are either ‘modal’ , or ‘non-modal’. A ‘modal’ G-code, remains ‘active’ for all subsequent programmed blocks, unless replaced by another ‘modal’ G-code. Conversely, a ‘non-modal’ G-code will only affect the programmed block in which it appears.

278

Chapter 7

. Figure 146. A turning and boring surface texture test piece


279

. Figure 147. Machinability testing utilising an ‘accelerated testing procedure’ – a combination of the rapid facing and degraded tool tests

280

Chapter 7

on a moderately short timescale. Normally in many previous testing programs, an uncoated cemented carbide P20, or P10 grade would have been used, since these grades withstand both higher speeds and have better tool wear resistance to that of previously utilised cutting tool materials. However in this case, an P25 grade was chosen, which is a degradation from the optimum P20 grade, but it should still perform satisfactorily. Furthermore, the cutting speed was raised by >2.5 times the optimum of 200 m min–1, with all facing operations being conducted at a ‘constant surface speed’12 of 550 m min–1. Typical tool-life curves produce by the AWT technique are illustrated in Fig. 148, showing the expected three stages of flank wear. This flank wear being a function of: the initial edge breakdown, steady-state wear – as the insert’s flank progressively degenerates and finally, catastrophic insert edge breakdown – as the edge completely fails. Detailed metallurgical analysis can be made as to the reasons why some P/M compacts performed better than others, by reference to the literature on the metallurgical interactions between the tool and the compact – this subject being outside the scope of the present discussion. The facing-off secondary machining operation meant that after 10 facing passes, a pre-programmed ‘optional stop’ can then be applied, to allow both tool flank wear and compact surface texture to be established. The faced-off surface texture results can then be superimposed onto the same graph – for a direct comparison of flank wear and for that of the machined surface texture parameter. Without going into too much detail of the specific aspects of the processing and metallurgical interactions present here on the composite graph, some compacts abraded the cutting insert more than others, while the ‘faced’ surface texture, generally seemed to get worse, then improve and finally worsen again. However, this is a complex problem which goes to the ‘heart’ of the vi-

12 ‘Constant surface speed’ , this can be achieved by employing the appropriate ‘canned-cycle’ G-code accessed from the CNC controller, which allows the testpiece’s rotational speed to increase as the faced diameter decreases*. * Normally there is a restriction on the rotational speed limit – created by the maximum available speed for this machine tool, which would normally be reached well before the cutting insert has coincided with that of components centre line, but because in this instance, the compacted testpiece is hollow, the rotational restriction does not present a problem.

sual aspect of machined surfaces – wherein the real situation is that surface texture continuously degenerates, and it is only the burnishing (i.e.‘ironing’) of the surface that ‘masks’ the temporary improvement in machined surface – more on this topic will be made in the surface integrity section. What is apparent from using the AWT technique is that on a very short timescale, considerable data can be generated and applied research assessments can be conducted both speedily and efficiently. This topic of exploiting the minimum machining time and data-gathering activities to gain the maximum information, will be the strategic message for the following dialogue.

Machinability Strategies: Minimising Machining Time, Maximising Data-Gathering Prior to commencing any form of machinability trials, parameters for cutting data need to be ascertained in order to minimise any likelihood of repetition of results, while reducing the amount of testpieces to be machined to the minimum. Data obtained from such trials must be valid and to ensure that the cutting parameters selected are both realistic and significant a disciplined experimental strategy based upon the ‘Design of Experiments’ (DoE) approach is necessary – see Fig. 149. Here, a flow-chart highlights the step-bystep approach for a well-proven industrial technique, to maximise the labour-intensive and costly exercise of obtaining a satisfactory conclusion to an unbiased and ranked series of machinability results. There are a range of techniques that can be utilised to assess whether the cutting data inputs, namely: feeds, speeds, DOC’s, etc., will result in the correct inputs to obtain an extended tool life, or an improvement in the machined surface texture from the testing program. One such method is termed the ‘Latin square’ – which assesses the significance of the test data and its interac-


. Figure 148. Graphical results obtained from the accelerated machinability test, illustrating how flank wear and surface texture degrades, with the number of facing-off passes

281

282

Chapter 7

tions. For a practical machinability trial employing a ‘Latin square’ , it uses a two-way ANOVA13 table, with a limited amount of ‘degrees of freedom’ , typically: feedrate, cutting speed, DOC, plus surface finish – these parameters can be changed/modified to suit the ‘programme of machining’ in hand. By using a very limited group of cutting trials, a two-way ANOVA table can be constructed and their respective ‘F-ratio’ for each interaction can be determined. This calculated ‘Fratio’ should be greater than the 5% ‘confidence limit’ of the statistical distribution to be significant. If the Fratio falls below –5% (i.e. for the calculated F-ratio), then the interactions are not significant, which necessitates increasing the ‘factor strength’ (e.g. increasing the: cutting speed, feedrate, etc.), to generate data which is >5% confidence limit – as shown by the ‘feedback loop’ in Fig. 149, or alternatively, using a different factor. By such means, ANOVA tests for significance of machining data, ensures that the processing parameters utilised for the prospective machinability trial are both valid and the correct ones to use in the proposed machining programme.

13 ‘Analysis of variance’ (ANOVA), or as it should be more appropriately termed the ‘analysis of variation about the means’ , consists of portioning the total variation present in a data set into ‘components’. Each ‘component’ is attributed to an identifiable cause, or source of variation; in addition, one ‘component’ represents the variation due to uncontrolled factors and random errors associated with the response measurements.Specifically, if the data set consists of ‘n’ measurements ‘y1.…,yn’ and their mean is denoted by: ‘y ’ , the total variation about the mean is embodied in the ‘sum of squared deviations’ , as following diagram depicts, for the ‘partitioning scheme’ for ANOVA: Total Sum of Squares about the mean: n

� (y  − y¯)

i=

↓ Sum of squares – due to Source1


↓

Sum of squares – due to Source3


↓ Error, or residual Sum of Squares

The technique of analysis of variance decomposes this total ‘sum of squares’ into the parts shown above, for a case in which four identifiable sources of variation are present – in addition to the ‘error component’. The number of identifiable causes of variation and the formulae for the ‘component sums of squares’ are intrinsically connected to the specific experimental design utilised, in the data collection and to the statistical model deemed appropriate for this analysis.

Rather than spending considerable time, effort and indeed exorbitant expense, on a large and complex machining testing programme, which more often than not, produces numerous machined components that are almost indistinguishable from each other. It might be more prudent, to conduct a ‘condensed’ series of trials, based upon a rigorous statistically-designed methodology. Therefore, experiments based on the socalled ‘orthogonal arrays’ can be beneficially engaged in this regard. Many applied researchers and engineers have utilised a range of factorial-designed experiments, typified by the ‘Taguchi-approach’. The main problem with these ‘arrays’ is that in many situations the large number of ‘interactions’ (i.e. factors) have been shown to interfere with the overall results – introducing ‘secondary effects’ , which will not have been anticipated for, when the original strategic programme was devised14. Such spurious data, could seriously affect future machining recommendations and influence the outcome in a negative manner. The ‘interaction problem’ can have these affects considerably reduced by incorporating a more ‘truncated-approach’ to the experimental design strategy for the machinability trials, rather than using a ‘full’ Taguchi orthogonal array (Fig. 150). For example, if all of the experiments are conducted in for example one of ‘standard’ the Taguchi L8(27) orthogonal array, depicted in Fig. 150, then the ‘total outcomes’ (i.e. components machined), would be: 27 = 128 × 8 = 1,024 individual components machined. Here, in the Taguchi orthogonal array seven factors have been employed and with the vast amount of components produced from such a long-running and very costly machining programme, many of the pertinent details will be lost on those engineers/researchers attempting to de-code the vast assortment of machinability data collated. However, it is possible to utilise a much simpler-approach to the overall massive data-collection and analysis problem, yet still providing statistical significance, this can be achieved by adopting a ‘Fractional factorial-designed experiment’. Here, instead of the virtually ‘mindless task’ of producing 1,024 almost identical components,

14 ‘Orthogonal array factors’ – when utilising a ‘full’ Taguchidesigned orthogonal array for a complete picture of all of the interactions, then it has been shown (Shainin, 1985 – see references), that if many factors are employed (i.e. normally >5), this results in unwanted ‘secondary effects’ which cannot be accounted for, leading to spurious results from any machinability trials.


283

. Figure 149. Flow chart indicating the desigh philosophy for unbiased and ranked machinability trials

by using a ‘Fractional factorial-designed experiment’ with an identical matrix to that given in Fig. 150, only 8 components are produced! This testing regime is both significantly quicker and much less costly to perform, obtaining a ‘snap-shot’ of the overall machinability problem, but because considerably less testpieces are produced, the ‘interaction-problem’ and its ‘secondary effects’ are not an issue, even when seven factors are utilised. Obviously, this machinability data has to be collated and investigated in a disciplined and controlled fashion. One tried-and-tested method of establishing an unbiased and ranked interpretation of

these results, is to use the much misunderstood and maligned technique of ‘Value Analysis’15 (VA). This VA when used to show trends in competitive functions

15 ‘Value Engineering and Analysis’ (VE/VA), with VE being principally concerned with an overall improvement of designbased details on engineering components, while a more limited form of this technique is termed VA – being particularly relevant for detailed interpretation of recorded data from experimentation. Here, in this case, from the wide-ranging and often seemingly unrelated output of machinability trials.

284

Chapter 7

. Figure 150. A fractional factorial-designed experiment, based upon a Taguchi L8(27) – orthogonal array


and operations, can be successfully utilised from the comparisons of cutting fluids, through to complex and difficult-to-machine aerospace machinability trials. If a more sophisticated technique is required, then it is also possible to utilise ‘Quality Function Deployment’16 (QFD), to obtain a complete picture of the outcomes from machining trials. QFD is often used by industry as a means for its ‘Continuous-improvement programmes’17. Here for ‘simplicity’s-sake’ , the more basic and somewhat less complex VA tabulated data-collation approach, will be briefly reviewed. The application of VA to a series of collated and compiled massed-data is not new. In fact, it was widely-used during the 1960’s, but fell into disfavour, partly because its function and operation were often not well-defined – this being exacerbated by poor implementation of its recommendations. However, VA techniques are useful, allowing one to interpret data trends both quickly and objectively – without undue bias – at a glance of a spreadsheet. Not only can significant trends be readily seen, but the spreadsheet shown in Fig. 151 – shows a typical machinability data for P/ M compacts drilled by two differing drill-point geometries. By using the spreadsheet, not only can overall trends be readily seen, it also can depict sub-set trends as well, giving a complete picture (i.e. globally) of the important criteria in assessing machining data. As a simple ranking system is used, considerable objectivity can be gained and with little undue influence – bias, affecting the outcome from these tabulated results. In employing the ranking of the results, it is normal practice to decrement down and if two values are ranked identically, then they are given the same rankings, followed by the next lower ranking, being two numbers lower, as following example shows:

16 ‘Quality Function Deployment’ (QFD), is a general term that means the: ‘Deployment of quality through deployment of quality functions’ (Akao, 1988). It is often known as the ‘House of Quality’ , because the tabulated graphical representation looks similar to that of a house – when all the interacting factors for subsequent analysis have been included on the chart. This QFD technique, is a wide-ranging philosophy for the complete analysis of both simple and intricate designs and can be successfully exploited for machinability trials. 17 ‘Continuous-improvement programmes’ , can be defined as an: ‘Operational philosophy that makes the best use of resources in order to increase product, or service quality and result in more effective satisfaction of customers’ (Swanson, 1995).

285

For example, in Fig. 151 – for the values shown in column two (i.e. left-hand side: Jobber drill, Thrust Force 0.254 N): Compact type:

1

2

3

4

5

6

7

8

Ranking:

6

8

5

1

5

7

3

2

NB Here, two 5’s were ranked, meaning that the next decremented value would rank as 3. Hence, in this case the Low compaction Compact type No. 2 this was best and Low compaction No. 4 worst – as jobber drilled. This ‘truncated approach’ the elementary and easily comprehended VA tabulation (Fig. 151) , enables nonspecialists, together with knowlegdible experimenter, to recognize the influence various machining parameters have on the potential performance of the trials undertaken. By judicious use, the VA technique in conjunction with a strictly controlled and limited machining strategy – based upon some form of ‘orthogonal array’ , in combination with the ‘strength’ (i.e >5% ‘F-ratio’) of parameters by ANOVA, this will enable a researcher to conduct a speedy, compact, realistic, yet meaningful machinability assessment.

7.2 Machined Roundness Roundness is a condition of a ‘surface of revolution’ , which can take the form of a: cylinder, cone, or sphere, where all the peripheral data points (i.e. measurements) intersect. In reality, the radius of say, a nominally round workpiece tends to deviate – from the ‘true circle’ – around the periphery of the part, making these variations the theme to subjective interpretation of the measured results. In fact, in the past, the simplistic technique for the assessment of roundness was usually measuring three diameters on a workpiece, to determine the diametrical variations, then ‘averaging’ to give its overall dimensional size. Moreover, for variations in a workpiece’s radius about an axis of rotation, this was often found by positioning the part between a ‘bench-’ , or sine-centres’ – the latter equipment is employed for turned tapered features, then rotating and monitoring it with dial gauges both at and along its length. In the past, this rather superficial metrological workpiece assessment was supposed to inform the inspector as to its potential in-service performance. If some radial variations occurred, this geometrical

286

Chapter 7

. Figure 151. Value analysis – tabulation of the performance of two drilling points and a typical range of drilling data, when machining powder metallurgy compacts


lobing18, or elliptical state, may have not have proven to be detrimental to its prospective overall in-service performance. In reality, there might be a whole host of reasons for a machined part to vary in its radius – for a stated cross-sectional plane. The following list attempts to show where and why radial differences occur: • Machine tool and its production processes – inducing some form of rotational imperfections from either the machine/tool/workpiece system, • Release of strain, or that induced into a workpiece – the former case may be the result of releasing the part from its clamping pressure, while the latter may result from plastic deformation promoting localised surface residual ‘hoop-stresses’19, • Induced radial vibration – potentially resulting from cutting forces and its effect on rigidity, in association with both tool geometry and cutting edge displacement (i.e. see Fig. 152), • Circumferential surface texture – created by the lasting effect resulting from the recent production process. It has been alluded to above that the machine tool and particularly its spindle, can create machine-induced inaccuracies of various kinds onto the machined

18 ‘Lobing’ , has a constant diameter if measured in a single plane. When attempting to measure lobing with a ubiquitous micrometer calliper, this is not possible, as a constant micrometer reading will result. Conversely, an ‘elliptical’ workpiece has both a major and minor diameters, allowing this diametraldifference to be determined using a micrometer calliper. NB A ‘lobed-shape’ can be established, by either placing the workpiece in a Vee-block, then carefully rotating the part and, if any pointer motion appears on the touching dial gauge, this represents the lobed-harmonic difference. To obtain much more detailed information on a ‘lobed’ workpiece, it is necessary to inspect the part on a roundness measuring machine. 19 ‘Hoop-stress’ , this can be defined as: ‘The circumferential stress in a cylinder wall under pressure, or in a rotating wheel [i.e. mass]’ (Carvill, 1997). The maximum hoop stress can be found using the following expression:

σ hmax = p

(rb + ra ) (rb − ra )

i.e. at the inner radius: σL = 0) (rb2 – ra2) Where: r = radius, p = pressure.

287

workpiece’s ‘harmonic roundness’ 20, some of these factors include: • Spindle imbalance – introducing dynamic lowerfrequency harmonics on the part, • Cutting forces – can dynamically affect the machining process, causing a series of high-frequency harmonics to be superimposed on the lower-frequency harmonic, resulting from imbalance (Fig. 152), • Thermal growth effects – changing both the spindle’s growth (axially) and causing modifications of an elastic nature to the relative ‘axis-orthogonalities’ 21 of the machine tool – which in turn, creates harmonics on the machined part, • Working clearances and motor-drive configurations – this is necessary to allow for relative thermal growth and beaing component ‘running-fits’ – within the spindle/headstock assembly, which are exacerbated by the type of motor drive system, specifically belt-driven systems (see Fig. 153). This latter feature of spindle inaccuracy, is present in many belt-driven CNC Lathes and turning centre headstocks being minimised by having a machine tool with a direct-drive spindle. In the case of these beltdriven headstocks, the working clearances and beltdrive, have the belts-tensioned on one side only. This arrangement, causes an irregular harmonic rotational motion to the spindle and hence, its work-holding equipment – chuck, etc., which when translated onto the resultant machined roundness and to a lesser extent the surface texture, creates harmonic problems

20 ‘Harmonic roundness’ , refers to the departures from roundness of a workpiece, with harmonic effects – often termed ‘undulations per revolution’ (upr) – being instigated by any number of external sources, such as those described in Table 10. NB On Roundness testing machines, the various harmonics are superimposed onto each other. For example, the 1st harmonic of the workpiece, may have say, the 5th harmonics superimposed onto it, followed in a similar fashion by 60th harmonics. This composite harmonic behaviour can be ‘filtered-out’ by the judicious use of double Gaussian filters, in order to see the effects of individual harmonic behaviour on the machined part. 21 Machine tool ‘axis-orthogonalities’ , relate to the fact that most of today’s 3-axes machine tools have each axis positioned either on top of each other and at 90° with respect to each other (i.e. X- and Y-axes), or normal/right angles to these axes (i.e. with respect to the Z-axis) – hence the term ‘orthogonality’.

288

Chapter 7

. Figure 152. The harmonic departures from roundness of a component, resulting from a lack of rigidity/damping effects whilst turning


(Fig. 153a–c). The influence of this irregular harmonic rotational belt-driven rotation can be gained from the schematic representation shown in Fig. 153d, where a repeating-series of ‘tumbling three-lobed harmonic’ geometric shapes are reproduced on the workpiece. The irregular, but periodic nature of the rotational action of the belt-driven headstock is reproduced on the workpiece by a series of kinematic combinations of headstock rotation and linear motion supplied by the longitudinal feed of the cutting tool along the part (Fig. 153d). If a direct-drive headstock configuration is utilised (Fig. 153e), then there is virtually no harmonic influence associated from the machine, so more consistent turned components result. Returning to Fig. 152, the overall machine-toolworkpiece system, can be isolated to consider the simple effect of a cantilevered cutting tool that is inadequately supported, or the unlikely occurrence of too small a cross-sectional area – making it somewhat ‘under-strength’. The main cutting force in turning operations is the tangential force (i.e. see Fig. 19), it results from several factors, such as: • Resistance to rotation – caused by the workpiece material’s inherent shear strength, • Undeformed chip thickness – resulting from the radial DOC selected,

289

• Orientation of cutter rake angle geometry – this

being a combination of either a positive, neutral, or negative rakes, plus to a lesser degree, the effect of shape and size of the tool nose radius, • Feedrate – in combination with DOC, will heavily influence the size of the effective chip thickness and play a dominant role in the resulting surface texture. In the upper diagram in Fig. 152, the tangential force is simplistically shown contacting the cutting insert at the point. The application of the cutting force here, causes a large bending moment to occur at the pivot point – as shown. The resultant dynamic action of this effect, is depicted in the lower diagram of Fig. 152, where the tool has been elastically deflected in a downward manner by this bending moment. Moreover, as the resistance to deflection increases with the tool’s downward direction, this intensifies the pressure from the inherent tool-body mechanical strength, enabling a certain degree of recovery, therefore there is a partial upward motion of the tool. This cyclical upward, then downward tool point motion is repeated at a periodic medium-frequency, causing a sinusoidal motional effect with this being harmonically reproduced on the turned surface. High-frequency harmonics can also be

. Table 10. The harmonic behaviour related to either the component manufacturing process, or its measurement Harmonic:

Cause:

1st (1 upr)

Function of measurement – only caused by the setting-up error on the instrument being used to measure the departures from roundness. The amplitude of this harmonic is equal to the eccentricity of the part, relative to the spindle axis of the roundness instrument.

2nd (2 upr)

Function of measurement, or manufacture – this aspect of harmonics is generally termed ovality and can be caused either by a setting-up error of the roundness instrument, or the part being machined out-of-square to its axis of rotation.

3rd–7th

Function of manufacture – these harmonics are normally introduced by the work-holding technique during manufacture. By way of illustration, if a three-jaw chuck were used to hold a relatively delicate part and excessive clamping force was employed, then upon machining and subsequent workpiece removal a three-lobed part would be the result.

15th-upwards

Function of material and manufacture – this aspect of harmonic behaviour is usually introduced to the part by either machine tool instability (i.e. self-excited vibration – chatter), or by the reaction of the materials used in the component’s manufacture – cutting insert/toolholder, lubricant – if any used.

Upr: undulations per revolution NB Higher harmonics may be the result of instrument noise, or vibration. [Courtesy of Taylor Hobson]

290

Chapter 7

. Figure 153. By utilising turning centre headstocks with direct-drive spindles – for ‘harmonic supression’, a significant improvement in machined roundness will result. [Courtesy of Yamazaki Mazak]


superimposed onto the medium-frequency harmonics, this aspect can be shown to good effect by a ‘power spectrum analysis’ 22 of the harmonic behaviour during machining. For a simple turning operation, the resultant cutting forces occur from the consequential combination of: a workpiece material’s shear strength, its undeformed chip thickness, the cutting insert geometry and accompanying nose radius, which has a significant affect on the harmonic ‘departures from roundness’ of the turned part. So that the effect of these variables in the cutting generation process can be seen, while simplifying the discussion, only external-diameter operations will be mentioned concerning these process-based roundness relationships – in the following section.

7.2.1 Turned Roundness – Harmonics and Geometrics A typical operation on a either an engine-/centre-lathe, or a turning centre, is schematically illustrated in Fig. 154. This involves a longitudinal turning process – the workpiece being shown as partially completed – using a ‘light-turning and finishing cutting insert’ , as it progresses along the turned part. Here, the turning application has a long and slender workpiece this being held in a work-holding device: chuck, or collet – at the headstock end, with further support23 supplied by

22 ‘Power spectrum analysis’ , is a useful aid in process monitoring of the cutting capabilities – giving a good interpretation of the anticipated surface topography (i.e its ‘micro-terrain’). A major advantage of utilising the ‘power spectrum’ as a diagnostic aid, is that it can separate-out any process-related tool problems. NB More details concerning the application of ‘power spectrum analysis’ can be obtained from the References by either Whitehouse (1997), or Smith (2002). 23 ‘Programmable and fixed steadies’ are often used to give additional support to the long and slender parts, to minimise ‘barrelling effects’ – created by increased tool push-off the further away the tool’s longitudinal distance becomes from the influence of the tailstock/headstock. Hence, the part has smaller turned diameters toward its ‘supports’ steadily increasing in diameter toward its centre, then reducing again – creating a ‘barrel-like profile’ along the entire turned bar’s length.

291

either a ‘dead-’ , or ‘rotating-centre’24 – at the tailstock end. As the orthogonally-oriented cutting insert (i.e. having a zero-plan approach angle) turns along the workpiece, a ‘moving step’ is seen to be present as the ‘emerging diameter’ occurs – to its set dimensional size (Fig. 154). If a very high quality toleranced part is to be turned, then it is desirable to review the operation more critically, as some unexpected and unwanted features may be present in the final machined com ponent. As the turning insert has an orthogonal orientation to the axis of rotation of the part (Fig. 154), it might be thought that no radial force component occurs, but this is not the case, as the tool nose radius can create a radial force affecting the turned surface. The radial force has little effect on the part’s harmonics when close to the tailstock as shown by the crosssectional harmonic effect indicated in section ‘C-C’ (Fig. 154). Once the cutting insert has progressed some distance along the workpiece, the contributing and supporting influence by the tailstock is lessened and the effect of this radial force component increases, as exhibited by section ‘B-B’ , this being amplified still further in section ‘A-A’. Here (i.e. section ‘A-A’), the harmonic departures from roundness are significant, a fact that has been recognised by precision turners for many years. Experienced machinists when turning parts having long length-to-diameter ratios, will fit either a ‘fixed-steady’ , or more preferably a ‘movingsteady’ close to the tool cutting zone – on the opposite side of the workpiece – to counteract ‘push-off ’ by the radial force, while minimising component eccentricity/run-out. If twin-turrets (i.e. upper and lower) are fitted to turning centres, then ‘balanced turning’ 25 can be utilised as an alternative machining strategy. There is a direct link between cutting forces and the geometric shape of the insert, this effect being illus-

24 ‘Rotating centres’ , can introduce their own eccentric error into the turning process, as they are less rigid than their ‘dead-centre’ counterparts, but the latter, has a rotational speed restriction – otherwise ‘dead-centre burn-out’ is likely and is therefore not practicable for high-production volume demands. 25 ‘Balanced turning’ , situates one cutting edge slightly ahead of the other in their respective opposing turrets. In this manner, the radial force components for each cutting insert have the effect of ‘virtually’ cancelling each other out, allowing long and slender workpieces to be successfully turned. A production bonus being the removal of greater workpiece material stock per pass.

292

Chapter 7

trated in Fig. 155. In these diagrams a simplistic representation for a range of cutting insert profiles is shown and for clarity, the tangential force has been excluded, with just the axial and radial force components indicated for each type of cutting insert shape. Assuming that the overall cutting data is identical in each case

(i.e. the same: rotational speed, feedrate, DOC, insert rake angle, plus workpiece material), then the only variable in the longitudinal turning process here will be the cutting insert shape its orientation. The component cutting forces – axial and radial, will vary for each tool profile in their respective magnitudes, due

. Figure 154. Machined roundness is influenced by a number of factors: unbalanced cutting forces, non-integral headstock and lack of support on slender/long workpieces


to the variation in plan approach angles. In the case of the orthogonal insert (0°) plan approach angle, the axial force dominates with virtually no radial force component present, this axial force being directly linked with the feedrate. The displayed profile chart for this harmonic roundness trace (i.e. section on ‘A-A’), for

293

the 0° plan approach angle, shows virtually negligible harmonic effects. If a triangular-shaped cutting insert geometry was selected, in this case having an 15° plan approach, here there is a slight reduction in the axial force component and a corresponding increase in its radial counterpart. This slight increase in the radial

. Figure 155. Turned roundness can be significantly affected by the insert shape, its approach angle – which affects cutting forces – resulting in harmonic out-of-roundness

294

Chapter 7

force, in combination with a marginally longer cutting edge being in contact with the workpiece’s ‘transient surface’ 26, leads to a slight increase in the harmonics on the displayed profile chart (‘B-B’). As the obliquity of the insert’s plan approach angle increases, as depicted by the square-shaped cutting insert, being inclined at an angle of 45°, then the axial and radial force components equalise. Here, the considerable radial force component has a significant effect on the displayed profile trace, as illustrated by the section at ‘C-C’ , where the harmonics have increased, but with a notable vibrational tendency superimposed onto this primary harmonic. This increase in vibration during turning, is the result of two noteworthy factors: firstly, the length of the transient surface has increased – with this square insert’s greater plan approach angle; secondly, as a result of this first condition of increased obliquity, the radial force affects workpiece rigidity, which is compromised, leading to an exacerbated turned part surface roundness and accompanying chatter-marks. Finally, when turning with a round insert geometry, this will lead to a vast increase in the radial force component, which in turn may cause significant harmonic out-of-roundness, if a very rigid set-up is not utilised. In this case, the round insert’s displayed profile trace, shows evidence of a significant increase in vibration – chatter – present, which has a dramatic effect on the profile of primary harmonic (section on ‘D-D’) – more will be said on the subject of ‘chatter-marks’ on the component’s surface shortly. The rationale for this notable harmonic amplification when using round inserts is the product of several interrelated factors. The transient surface in contact with the round/curved profile has been markedly increased, together with the plan approach at the tangency position – with respect to the workpiece – being at a maximum; thus, the combination these two factors leads to a momentous deterioration in workpiece rigidity and as a result here, the vibration will especially increase.

26 ‘Transient surface’ , can be defined as: ‘The part of the surface formed on the workpiece by the cutting edge and removed during the following cutting stroke, [by the next] revolution of the tool, or workpiece’ (Boothroyd, 1975). NB In the case thread-turning operations, the transient flank’s surface only remains until the next pass of the screw-cutting insert obliterates it, or until the final thread depth is reached.

If a large volume of workpiece stock has to be removed in a series of roughing cuts, a strong insert is necessary, therefore the problem associated with the harmonic behaviour of cutting insert geometry becomes of less importance. This latter fact allows either a square, or round cutting insert to be utilised due to their intrinsic strength and if vibration is a problem, then a ‘finishing cut’ with an insert having an 0° plan approach geometry would remove any probable surface chatter-marks. The case of turning with a round insert geometry is worthy of a closer investigation, as several factors influence the harmonic roundness of a workpiece turned with its curved profile. For example, let us consider several conditions for employing a standardised round insert and its subsequent affect on the harmonics of the turned part. If one assumes that an identical: rotational speed, feedrate and workpiece material was used, but having differing DOC’s – to isolate variability in the turning process. In the first example using a small DOC, the radial force component is large in comparison to the axial force, but the harmonic workpiece roundness is not compromised here – as these forces are minute. Conversely, an extreme example of using a round insert might be when employing a larger DOC. Here, the radial force component has been reduced in comparison to the axial component, although the magnitude of these forces will be considerably greater than in the former case. The pressure exerted on the very long contact region at the transient curved surface, creates potential harmonics in the turned part as the resultant force has now significantly increased. So, the influence of an increased DOC, in combination with the round insert profile can create a very long contact region at the transient surface, thereby causing unwanted harmonic effects on the turned surface. In Appendix 9, a visual impression is given of the principal techniques for roundness measurement and its assessment, together with some of the filtering effects are highlighted.

7.3 Chatter in Machining Operations In the machining of metals, chatter (Fig. 156) is a form of self-excited vibration introduced by the closed-loop force-displacement response to cutting. The plastic deformation during machining operations is always pro-


ceeded by elastic deformation – the situation is akin to that of it acting somewhat like a ‘big spring’27. Moreover, the mechanism by which a cutting process dissipates energy is termed chatter and vibration, also this being a function of the workpiece’s rotational speed. Any chatter/vibration can clearly be heard as an unwanted machining noise by an experienced machinist, who would then modify the speed accordingly. There are a wide variety of causes for chatter, including the process-induced effects from the cutting forces, which may be the result of changes in: the cutting velocity; chip cross-sectional area; tool/chip interface friction; BUE, variations in the workpiece composition; or the most common factor being process modulation resulting in regeneration of vibration. If greater energy is input into the dynamic ‘machining-loop’ than can be readily dissipated by the following: mechanical work; damping, or friction, then an ‘equilibrium status’ is required and this output is via the somewhat superfluous effect of the generation of chatter/vibration. Vibration is a debilitating process affecting both the machined surfaces and reducing tool life in any machining operation, consequently it must be appropriately identified – classified, then one has the potential to find the actual cause of this unwanted effect and resolve it accordingly. In essence, in machining operations there are three types of vibration that may transpire, these are: • Free vibration – this being the response to sudden change, or to any initial condition, where the vibrational amplitude decreases with time, occurring at the system’s natural frequency, NB An interrupted machining operation, or workpiece feature can create this vibrational effect and it frequently appears as shadows, or lines following a surface discontinuity.

295

amplitude stays constant for a set of input conditions, being non-linearly related to speed. NB The most common examples of this effect are caused by: cutter imbalance, cutting teeth impacting on workpiece, tooling misalignments, plus the occurrence of any form of rotational system resonance.

• Self-excitation vibration, or chatter – occurs

through the system’s periodic response to a constant input, which may intensify in amplitude – becoming unstable, often occurring – regardless of the input, but close to the natural frequency of the system. NB Chatter is due to waviness regeneration in the machined surface, it commonly occurs during metal cutting operations (i.e. see Fig. 156a).

What is chatter and how might can be characterised? Degarmo, et al. (2003), has produced a list of the following factors that can indicate the onset of chatter, these being characterised by: • Sudden onset of vibration – whose amplitude will rapidly increase until a maximum threshold – saturation – is reached (i.e sounding like either: a screech, whine, or buzz), • Chatter frequency is near to that of the machining system’s natural frequency (i.e critical frequency) – changing only slightly with any process parameter variations. The largest force-displacement response occurs at ‘resonance’ 28 enabling the maximum dissipation of energy, • Chatter produces unacceptable surface texture (Fig. 156a) – normally highlighted by either an angular, or helical pattern (i.e. the visual appearance

• Forced vibration – can be regarded as a response to

a periodic – repetitive timing – input that occurs at an identical frequency. At this point, the vibrational

27 ‘Spring-cuts’ , are always present in any ductile component machining operation, resulting from the relaxation of the forces and the elastic recovery of the tool and workpiece after the cutting insert’s passage along the part. In fact, if the tool is repositioned once more at the beginning of the original cut, then simply fed along the component, it will take a minute cut – assuming that the tool’s edge is still sufficiently sharp, this is termed the ‘spring-cut’.

28 ‘Resonance’ , is of practical importance in many engineering applications, because relatively few oscillatory forces can result in large vibrational amplitudes that can cause damage, or interfere with the functioning of the system. NB The classic example of this phenomenon was found when ranks of soldiers marched across a bridge in unison (i.e. ‘instep’) which, if it coincided with one of the bridge’s resonant frequencies, could create damage to the structure – despite the fact that the bridge could safely support their overall weight. Hence, the order to ‘break-rank’ (i.e randomising both their pacing and steps) when proceeding over a structure such as a bridge was mandatory.

296

Chapter 7

. Figure 156. Vibration and chatter in machining operations, with their machine tool damping characteristics


is either ‘pearled’ , or ‘fish-scaled’) superimposed over the normal cutting insert’s feed marks, • Visible surface undulations – these effects are reproduced in the direction of feed, being the product of either serrated, or wavy chip formations, of variable thicknesses.

7.3.1 Chatter and Chip Formation – Significant Factors Influencing its Generation The stability of the cutting process and the onset of regenerative chatter is influenced by a range of factors, such as the: cutting stiffness (Ks)29 of the workpiece material – related to its machinability; parameters of the machining process (e.g. speed, feed, DOC, chip width – total); insert cutting geometry (e.g. rake and clearance angles, edge preparation, insert shape and size); cutting process dynamic characteristics (e.g. machine-tooling-workpiece/fixturing). Hence, during machining operations on the workpiece, the chip is formed by shearing over the chip area, producing the cutting, or tangential force (FT). The magnitude of this tangential force is heavily influenced by the product of the workpiece material’s stiffness (Ks) and the chip area, as follows: FT = Ks × t × w Where: FT = tangential force (N), Ks = workpiece material’s stiffness (N mm–2), t = chip thickness (mm), w = chip width (mm). The direction of the tangential force (FT) is predominantly affected by the cutting insert’s rake and clearance angles, together with the edge preparation on the insert. In many single-/multi-point machining operations used to generate for example a milled surface, there is a requirement to overlap the adjacent cutting paths (Fig. 84c). For most single-point machining op-

29 ‘Cutting stiffness’ (Ks), is closely associated with that of ‘flow stress’*, but is more simple to calculate and can be thought of as a workpiece material property, being dependent on its hardness. *‘Flow stress’ , can be defined as: ‘The stress required to sustain plastic deformation at a particular strain’ ( Kalpakjian, 1997).

297

erations, this former over-lapping of tool paths does not take place in the same manner, but will only occur after one complete revolution of either the workpiece, or tool. In operations by either milling (Fig. 85), or drilling (Fig. 50), an overlap takes place in a fraction of a revolution, this being dependent upon how many cutting edges are present on the tool. In the Degarmo, et al. (2003) machining model shown in (Fig. 157a), the cutting or tangential force (Fc)30 generation may cause a relative displacement ‘X’ between the cutting insert and the workpiece, affecting the uncut chip thickness (t), this results in changing the cutting force. This coupled relationship between displacement in the ‘Y’ direction – modulation direction – and the resultant cutting force, creates a closedloop response system. Here, the modulation direction is normally at 90° to the machined surface, so defines the chip thickness. As a consequence of these interrelated factors, there is a phase-shift (ε) between the subsequent overlapping machined surfaces, resulting in a variable chip thickness and modulation of the displacement, causing chatter vibration to take place. Accordingly, this phase-shift between overlapping cutting paths is accountable for the production of chatter (Fig. 157b). Moreover, there is a favoured speed corresponding to a phase-locked condition (e.g. when ‘ε=0’), resulting in a constant chip thickness (t). By obtaining a constant chip thickness, this results in a ‘steady-state’ cutting force generation with it and, the eradication of the feed-back mechanism for regenerative chatter. In essence, this is the goal for all machining operators, as they attempt to achieve this effect by vary the cutting speeds for a given set of conditions for a particular machining operation.

7.3.2 Chatter – Important Factors Affecting its Generation In the previous sections, a brief discussion was made concerning just some of the causes of regenerative chatter mechanisms. It is worth looking in greater detail at the reasons why this superfluous chatter occurs, explaining how and why it is generated in the hope of

30 In the Degarmo, et al. (2003) model diagrammatically shown in Fig. 157a, they use the term and nomenclature of: ‘cutting force’ and ‘Fc’ , whereas previously in the text, this has been referred to as the ‘tangential force’ , denoted by ‘FT’.

298

Chapter 7

. Figure 157. A chatter model, with potential chatter conditions and the application of the ‘stability lobe diagram’. [Source: Degarmo, Black & Kosher, 2003]


either entirely eliminating it, or at the very least, min imising its affect on the overall machining process. Chatter during machining can result from a range of multifarious and often linked-factors, they include: • Depth of cut (DOC) – can be considered as the principal cause and, for the prospective control of chatter. The DOC delineates the chip width, acting as the feed-back gain31 within the closed-loop cutting process, NB The machining processes ‘stability limit’ – being the threshold between stable cutting and chatter – can be determined from trial-and-error by simply incrementally increasing the DOC until the commencement of chatter, then‘backing-off ’ at this level. The prediction of chatter’s onset can be found analytically, this value being based upon thorough knowledge of material stiffness and cutting system dynamics.

• Rotational speed – is probably the simplest parameter to modify, thereby altering chatter and its associated amplitude,

NB The peripheral speed of either the rotating tool, or workpiece, affects the phase-shift between overlapping surfaces and its associated vibration regeneration.

• Feed – for milling operations the feed per tooth de-

fines the average uncut chip thickness (t), influencing the magnitude of the cutting process. Chatter is not unduly affected by the feedrate selected, but feed does have an effect on the predictable severity of vibration during machining, NB As no cutting force exists if the vibration occurs in the ‘Y’ direction – resulting in loss of contact between the tool and workpiece – the maximum amplitude of chatter vibration will be limited by its feed.

299

• Cutting stiffness (Ks) – is a material property con-

nected to: shear flow stress; hardness, as well as work-hardening characteristics of the workpiece, this factor often being referred to in a metaphorical sense of its material’s machinability characteristics,

NB Materials that might offer poorer comparative machinability, for example titanium, require considerably higher cutting forces leading to a greater displacement in the ‘Y’ direction and as such, offer a less stable cutting action.

• Width of chip (total) – is equivalent to the product

of the DOC multiplied by the number of cutting edges engaged in the cut. Furthermore, the total cut width will influence the stability of the cutting process, NB At a preset DOC corresponding to that of the ‘stability limit’ , increasing the number of engaged cutting edges, will result in chatter, or vice-versa.

• Cutting tool geometry – influences both the direc

tion and the magnitude of the cutting force, in particular the quantity of the force component in the modulation direction ‘Y’. So, an increased force occurring in the ‘Y’ direction, causes amplified displacement and vibration at 90° to the surface, creating ideal conditions for chatter. Other cutting insert geometrical factors that can influence the cutting stability include the following: – Back rake angle (α) – as it is inclined to a more positive angle, the length of the commencement of the shearing zone decreases, this in turn, reduces the magnitude of the cutting force (Fc). As the back rake inclination becomes larger, then this directs the cutting force in a more tangential manner, thereby reducing the force component in the ‘Y’ direction – creating improved stability at higher speeds,

NB An insufficient feedrate in comparison to the insert edge radius produces a less efficient cutting action, with more tool deflection and reduced machining stability.

– Clearance angle – reduction (γ) – has the effect 31 ‘Gain’ , can be practically defined in the following way: ‘The ratio of the magnitude of the output of a system with respect to that of the input – the conditions of operation and measurements must be specified’ (Smith, 1993, et al.).

of increasing the frictional contact at the interface between the tool and workpiece, possibly having a process damping effect. This potential stabilising effect could be the result of energy

300

Chapter 7

dissipation – heat transformation, which could result in decreased tool life, with the superfluous effect of thermal distortion of the machined part, or an increase in the workpiece’s heat-affected zone (HAZ), NB On a newly-fitted cutting insert, if initial wear occurs, this can sometimes have a stabilising effect for the onset of chatter.

– Nose radius – size, insert shape – diamond tri-

angular, square, round, plan approach angle – positive, neutral, negative – all influence the area of the chip shape and its corresponding ‘Y’ direction. The orientation of the modulation direction ‘Y’ toward a dynamically more-rigid direction angle, allows a decrease in vibrational response, giving greater overall process stability – having notably less chattering tendencies.

As machining process stability is a direct result of characteristics of dynamic force displacement between both the workpiece and the cutting insert, all of the various factors of a machining system: machine tool; spindle; tooling; workpiece; workholding – in varying degrees, can influence chatter. To increase process stability of the machining system, it is necessary to maximise the dynamics, this being the overall product of its static stiffness and damping capacity. Further, machining stability can be increased by utilising tooling with the greatest possible diameter with the minimum of tool overhang. By way of a caution concerning chatter frequency, this normally occurs near the most flexible vibrational mode of the machining system.

independence’ , or an ‘unconditional stability’. Hence, at relatively slow speeds an increased stability can be achieved within the process damping region – as shown. The ‘conditional stability’ lobe regions of the diagram, permit an increased total cut width (i.e the DOC x number of cutting edges, these being engaged in the cut) at dynamically preferred speeds, at which the phase-shift ‘ε’ between overlapping, or consecutive cutting paths approaches zero. In Fig. 157c, stability lobe number ‘N’ refers to the complete vibration cycles existing between overlapping surfaces. Moreover, the higher speeds correspond to lower lobe numbers, providing the utmost potential increase in the total cut width and material removal rate – this being due to the greater lobe height and width. If the total cut width exceeds the stability threshold – even assuming that the cutting process is operating at the desired speed, chatter will occur. So, the larger the total cut width above the ‘stability limit’ , the more unstable and aggressive the chatter vibration becomes. Referring to the diagrammatic representation of the SLD on the graph in Fig. 157c, if a chatter condition arises, such as that found at point ‘a’ , the rotational speed is attuned to the initial recommended speed (i.e. when ‘N=1’), resulting in stable machining at point ‘b’ on this diagram. The DOC can be incrementally increased until the onset of chatter again – as the threshold stability is crossed at point ‘c’. By utilising a hand-held ‘speed analyser’33 whilst the chatter continues – under the previously-selected operating conditions, this will result in the ‘analyser’ giving a modified speed recommendation that corresponds to point ‘d’. Now, if required, the DOC can be progressively incre-

7.3.3 Stability Lobe Diagrams In Fig. 157c, a ‘Stability lobe diagram’ (SLD) is depicted, which relates to the: total cut width that can be machined, to the tooling’s rotational speed, for a specified number of cutting inserts. For example referring to the: Degarmo, et al. (2003) diagram, suppose the total width of cut was maintained below a minimum level32, then the process stability would exhibit ‘speed

32 If the total cut width was maintained below a minimum level, in practical terms this would be of limited value for many machining systems.

33 ‘Speed analysers’ , are normally hand-held devices that produce dynamically-favoured speed recommendations and are commercially available. Such ‘speed analyser’* equipment when utilised for a cutting process, can show the relative motion between the tooling and the workpiece and recommends the appropriate speed to avoid chatter-effects. *‘Speed analysers’ can be successfully used for many industrial applications, such as those involving: High-speed; Thin-chip, hardened-die machining; multi-point cutting operations – milling, etc.; Turning and boring operations. These ‘speed analysers’ can also be employed for workpiece compositions ranging from ductile metals (i.e. aluminium and steel grades) and brittle materials (i.e. cast irons and brasses, etc.), together with some non-metallics (plastics, etc.) and composite materials (carbon fibre, etc.).


mentally increased to point ‘e’34 – this being a ‘safelimit’ for the optimum machining operation.

7.4 Milled Roundness – Interpolated Diameters Circular features such as bosses, circular rebates, etc., can be CNC milled by utilising a specific word-address ‘circular interpolation’ 35 command. This CNC function creates precise and accurate circular control in two slideways simultaneously, while the milling cutter mills around the workpiece, as depicted in Fig. 158. Here, the milling cutter’s rigidity plays an important role in the quality of the final machined feature, this being based upon the ‘rigidity square rule’ 36. The deflected milling cutter illustrated in Fig. 158-right, having lack-of-rigidity will produce some unwanted effects on the final milled part. Cutter deflection not only introduces the potential for chatter vibration, but if used to mill up to square shoulder, its deflection distorts the component geometry and introduces harmonic variation to the circular interpolated feature. So that minimal change takes place in a milled profile, it is advisable to keep to cutter lengths having short

34 Generally-speaking, it is not advisable to attempt to maintain both the DOC and the total cut width at the stability threshold , because any variation in the: workpiece affecting its cutting stiffness ‘Ks’; speed errors; or perhaps small changes in the overall dynamic characteristics of the machining system, could result in crossing the stability limit, creating severe chatter. For example, in a milling application, the amplitude of chatter vibration can be limited by a provisional feed per tooth reduction , until an established and desired speed has been achieved offering a stable DOC. 35 ‘Circular interpolation’ , is a block of entered information directing the CNC system to cut, either an arc, or a circle, (e.g. G02 – in a clockwise, or G03 anti-clockwise direction). 36 ‘Rigidity square rule’ – for milling cutters states: ‘Cutter rigidity decreases by the ‘square’* of the distance from the holder’ (Smith, 1993, et al.). *For example, if a cutter ‘stood-out’ from its respective toolholder by 50 mm to mill a circular feature (Fig.158 – left), then, if all other machining conditions remained the same and, then cutter was replaced by one of 100 mm long (Fig. 158 – right), it would now be 4 times less rigid, causing serious tool deflection.

301

stand-off distances, conducive with correct and current operational practices. There are several distinct problems involved in the milling high-quality circular interpolated features and, a slight digression into basic machine tool induced-errors is necessary to clarify the circumstances for the problems exhibited in Fig. 159. Most of today’s machine tools have what is termed ‘orthogonally-orientated axes’37 and in the case of the popular three-axis vertical machining centre configurations, if the axes have not been recently calibrated, then considerable ‘error’38 can be introduced into the final milled part features. It has been well-proven that a machine tool equipped with three orthogonal sideways: ‘X-axis’; ‘Y-axis’ – in the horizontal plane, together with the ‘Z-axis’ – in the vertical plane, can introduce up to 21 kinematic ‘errors’ into the cutting process. The kinematics for any machine tool are quite complex, when it has the ability to provide motion to all its axes simultaneously, although these errors are often small, they are

37 ‘Orthogonally-orientated axes’ , (is briefly mentioned in Footnote 2) refers to the fact that each axis is positioned at 90° with respect to each other, often situated on top of another axis. For example, on a typical 3-axis vertical machining centre, the ‘Y-axis’ sits on top of the ‘X-axis’ , but at right-angles to it, conversely, the ‘Z-axis’ is situated at 90° to these axes – hence the term ‘orthogonal’. NB Non-orthogonal machine tools exist, often having complex ‘kinematics’* between five and six axes. Therefore with these machine tools, in order to machine (i.e. mill) a straightline. all the axes must be in synchronised control to achieve this linear action.

*Kinematics, comes from the Greek word ‘Kinesis’ , which means ‘Motion’. It can be defined as: ‘The study of motion without regard for the cause‘ (Lombardi, 2001). In machine tool terminology, it refers to the translational effects of both linear and angular motions. It is principally concerned with the effects of the ‘degrees of freedom’ for a ‘free-body’ in three-dimensional space (also see: Footnote 47, in Chapter 3).

38 ‘Error’ is now not considered as an appropriate metrological term for any form of calibration, the recommended term today, is: ‘uncertainty’*. *‘Uncertainty’ , has been simply defined as: ‘The doubt that exists about the result of any measurement’ (Bell/NPL, 1999). This is why today, uncertainty in measurement is a combination of many factors, some physical, while others are induced. Hence, another term, along with all of these uncertainty factors has been coined, which is its ‘Uncertainty budget’ – this being a simple mathematical calculation, based upon a summary of these uncertainty calculations.

302

Chapter 7

. Figure 158. The effect of increased milling cutter length on the resultant circular interpolated profile on the workpiece


. Figure 159. The generated errors produced when circular interpolating at high feedrates when high-speed machining

303

304

Chapter 7

but significant ‘errors’ , which can be said to be simplistically produced as a result of: • Linear motions (six) – created by the displacement of the forward-and-backward motion of the X-, Yand Z-axes slideway movements, introducing particular non-linearities into the slideway positioning, • Rotational motions (three) – yaw, pitch and roll for each axis. All of these partial rotational motions can be practically-described in the following manner: • Yaw is the side-to-side ‘crabbing-motion’ along the slideway, NB ‘Yaw’ is normally the result of too much clearance (i.e. ‘slop’) in the adjacent slideway members.

• Pitch introduces a backward-and-forward rocking (pitching) action normal to the slideway, as the moving element traverses along the axis,

NB ‘Pitching’ is probably due to the ‘profile/waviness’ (i.e. long-frequency effects) in its respective slideway.

• Roll is the clockwise-and-anticlockwise rotational motion along the slideway.

NB ’Roll’ could be introduced by two ‘adjacent ways’ situated on each slideway, but not being coincident with respect to each other (i.e. laying in the same respective plane), causing a limited pivoting action – along the ‘line-of-sight’ of the axis as it traverses along its length.

the machined part surface and the machine tool’s profiling abilities. Moreover, ‘error-mapping techniques’ and sophisticated in-process control by an associated ‘dynamic error compensation system’ , have been shown to extensively reduce the effects of the ‘variety of errors’ that can be present on the machine tool, but once again, these topics are mentioned only for further research applications – as necessary. The circular interpolated milled profile shown in Fig. 159, shows significant departures from roundness of the milled workpiece, which is a function of most of the previously discussed kinematically- and thermallyinduced machine tool ‘errors’ , together with the possibility of some ‘load-induced errors’. This diagrammatic representation (i.e. Fig. 159), indicates that several ‘errors’ on the milled circular interpolated profile are present. At relatively slow simultaneous feeding-motions of the two axes (‘X-’ and ‘Y-axis’), it will generate a reasonable facsimile of the required circular feature. However, then by somewhat increasing this milled interpolation speed, the apparent roundness will appreciably degrade, the reasons for this degradation, might be the result of: • Servo-spikes – these unwanted effects occur at the ‘axis transition points’ 39 at their respective 90° angular intervals, often termed ‘quadrant-points’ , • Back-lash – possibly resulting from any form of axis reversals, originating from the recirculating ballscrews40, creating a slight ‘off-set’ , or ‘mismatch’ at the axis transition points, • Servo-errors – when both axes are simultaneously moving, their respective linear speed should be

• Squareness (three) – these ‘errors’ occur due to the

fact that each axis may not be at 90° (i.e. square) to one another.

These types of 21 ‘kinematic machine-induced errors’ can be appreciably reduced by the application of calibration through laser-based techniques. To a lesser extent, these ‘errors’ can be minimised via ballbar artifact-based methods, offering a quick ‘health-check’ by either static, or dynamic assessment techniques. The results of either the laser, or ballbar, can be fed back into the machine’s CNC controller for dynamic corrections as cutting takes place, offering a considerable improvement in the machine’s subsequent accuracy and precision. The above machine tool calibration techniques are somewhat beyond the scope of the present discussion, the same could be said for ‘thermally-induced errors’ , however, they can also influence

39 ‘Axis transition points’ , are where the ‘servo-spikes’ occur. They result from a reversal of one of the axes at this angular position and, its associated motor power-surge creating this ‘spike’. Normally, the ‘spike’ is associated afterward by a corresponding, but very small localised slack here, as axis take-up begins once more at these ‘quadrant-points’ on the circularinterpolated feature (i.e. see the inset and magnified diagram in Fig. 159). 40 ‘Recirculating ballscrews’ , are not supposed to have any appreciable back-lash present, as they are normally pre-stressed by applying loads by the application of either: tension-, or compression-shimming. However, as the pitch of any the screw has minute errors present, these are usually ‘mapped-out’ by the original machine tool builder – using the recognised International Standard laser-calibration techniques. Although, once the machine tool has been operating for sometime and either local ballscrew-wear occurs, or perhaps the machine has had the occasional ‘tool-crash’ , this can introduce and affect both its pitching- and back-lash-errors.


perfectly matched, allowing either a partial arc, or circular feature to be reproduced. If non-synchronised motion occurs, often termed ‘servo-mismatch’ 41 between these two axes, then an elliptical profile – usually inclined at an 45° angle occurs, • Squareness – when orthogonal (squareness) is not maintained between the two interpolating axes, then the net result will look similar to that of a milled angular elliptical profile shape, which is unaffected by the selected circular interpolation rotational direction. Considerably more machine tool-induced factors can affect a milled circular interpolated profile. These ‘errors’ can be found, isolated and then reduced by diagnostically interrogation by using dynamic artefacts, such as the ballbar. Ballbars and their associated instrumentation can not only find the sources of error, they can prioritise their respective magnitudes – to show where the main ‘error-sources’ occur, then instigate any feed corrections into the CNC controller to nullify these ‘machine-induced errors’. As a result of eliminating such ‘error-sources’ , this enables the milled circular contouring and overall performance to be appreciably enhanced.

7.5 Machined Surface Texture

305

with it, resulting from a combination of several interrelated factors, such as the: • Influence of the workpiece material’s microstructure, • Surface generation method which includes the cutting insert’s action, associated actual cutting data and the effect of cutting fluid – if any, • Instability may be present during the production machining process, causing induced chatter, resulting from poor loop-stiffness between the machinetooling-workpiece system and chosen cutting data, • Inherent residual stresses within the workpiece can occur, promoted by internal ‘stress patterns’ 42 – causing latent deformations in the machined component. From the restrictions resulting from a component’s manufacture, a designer must select a functional surface condition that will suit the operational constraints for either a ‘rough’ , or ‘smooth’ workpiece surface. This then raises the question, posed well-over 25 years ago – which is still a problem today, namely: ‘How smooth is smooth?’ This question is not as superficial as it might at first seem, because unless we can quantify a surface accurately, we can only hope that it will function correctly in-service. In fact, a machined surface texture condition is a complex state, resulting from a combination of three distinct superimposed topographical conditions (i.e. as diagrammatically illustrated Fig. 160a), these being:

Introduction to Surface Tex ture Parameters When a designer develops the features for a component with the requirement to be subsequently machined utilising a computer-aided design (CAD) system, or by using a draughting head and its associated drawing board, the designer’s neat lines delineate the desired surface condition, which can be further specified by the requirement for specific geometric tolerances. In reality, this designed workpiece surface condition cannot actually exist, as it results from process-induced surface texture modifications. Regardless of the method of manufacture, an engineering surface must have some form of ‘topography, or texture’ associated

41 ‘Servo-mismatch’ , can often be mistaken for a ‘squareness error’ , but if the contouring interpolation direction is changed, from G02 (clockwise) to G03 (anti-clockwise) rotation, then an elliptical profile will ‘mirror-image’ (‘flip‘) to that of the opposite profile – which does not occur in ‘squareness errors’.

42 ‘Stress patterns’ , are to be expected in a machined component, where: corners, undercuts, large changes in cross-sections from one adjacent workpiece feature to another, etc., produce localised zones of high stress, having the potential outcome for subsequent component distortion. ‘Modelling’ a component’s geometry using techniques such as: finite element analysis (FEA), or employing photo-elastic stress analysis* models or similar simulation techniques, will highlight these potential regions of stress build-up, allowing a designer to nullify, or at worst, minimise these potential undesirable stress regions in the component’s design. *Photo-elastic stress analysis displays a stress-field, normally a duplicate of the part geometry made from a thin two-dimensional (planar) nematic liquid crystal, or more robustly from a three-dimensional Perspex model, which is then observed through polarised light source. This polarised condition, will highlight any high-intensity stress-field concentrations in the part , which allows the ‘polarised model’ to be manipulated by applying either an un-axial tension, or perhaps a bi-axial bending external stress to this model, showing dynamically its potential stress behaviour during its intended in-service condition.

306

Chapter 7

. Figure 160. Surface texture comprises of: ‘long-’, ‘medium-’ and ‘short-components’, together with the ‘direction of the dominant pattern’ – superimposed upon each other. [Courtesy of Taylor Hobson]


1. Roughness – comprising of surface irregularities occurring due to the mechanism of the machining production process and its associated cutting insert geometry, 2. Waviness – that surface texture element upon which roughness is superimposed, created by factors such as the: machine tool, or workpiece deflections, vibrations and chatter, material strain and other extraneous effects, 3. Profile – represents the overall shape of the machined surface – ignoring any roughness and waviness variations present, being the result of perhaps the long-frequency machine tool slideway errors. The above surface topography distinctions tend to be qualitative – not expressible as a number – yet have considerable practical importance, being an established procedure that is functionally sound. The combination of roughness and waviness surface texture components, plus the surface’s associated ‘Lay’ 43 are shown in Fig. 160a. The ‘Profile’ is not depicted, as it is a long-frequency component and at best, only its partial affect would be present here, on this diagram. The ‘Lay’ of a surface tends to be either: anisotropic, or isotropic44 in nature on a machined surface topography. When attempting to characterise the potential functional performance of a surface, if an anisotropic ‘lay-condition’ occurs, then its presence becomes of vital importance. If the surface texture instrument’s stylus direction of the trace’s motion over the assessed topography is not taken into account, then totally misrepresentative readings result for an anisotropic surface condition occur – as depicted in Fig. 160b. This is not the case for an isotropic surface topography, as relatively uniform set of results will be present, regardless of the stylus trace direction across the surface (i.e.

307

see Fig. 161a – for an indication of the various classifications for ‘Lay’). Returning once more to Fig. 160b, as the stylus trace obliquity changes from trace ‘A’ , inclining toward trace ‘E’ , the surface topography when at ‘E’ has now become flat, giving a totally false impression of the true nature of the actual surface condition. If this machined workpiece was to be used in a critical and highly-stressed in-service environment, then the user would have a false sense of the component’s potential fatigue45 characteristics, potentially resulting in either premature failure, or at worst, catastrophic failure conditions. In Fig. 162, the numerical data (ISO 1302:2001), has been developed to establish and define relative roughness grades for typical production processes. However, some caution should be taken when utilising these values for control of the surface condition, because they can misrepresent the actual state of the surface topography, being based solely on a derived numerical value for height. What is more, the ‘N-number’ has been used to ascertain the arithmetic roughness ‘Ra’ value – with more being mentioned on this and other parameters shortly. The actual ‘N-value’ being just one number to cover a spread of potential ‘Ra’ values for that production process. Nevertheless, this single numerical value has its merit, in that it ‘globally-defines’ a roughness value (i.e.‘Ra’) and its accompanying ‘N-roughness grade’ , which can be used by a designer to specify in particular a desired surface condition, this being correlated to a specific production process. The spread of the roughness for a specific production process has been established from experimental data over the years – covering the maximum expected ‘variance’ 46 – which can be modified

43 ‘Lay’ , can simply be defined as: The direction of the dominant pattern’ (Dagnall, 1998).

45 ‘Fatigue’ , can be defined as: ‘The process of repeated load, or strain application to a specimen, or component’ (Schaffer, et al., 1999). Hence, any engineering component subjected to repeated loading over a prescribed time-base, will normally undergo either partial, or complete fatigue.

44 ‘Anisotropic, or isotropic surfaces, either condition can be individually represented on all machined surfaces. Anisotropy, refers to a surface topography having directional properties, that is a defined ‘Lay’ , being represented by machined feedmarks (e.g. turned, shaped, planed surfaces, etc.). Conversely, an isotropic surface is devoid of a predominant ‘Lay’ direction, invariably having identical surface topography characteristics in all directions (e.g. shot-peening/-blasting and, to a lesser extent a multi-directional surface-milling, or a radiallyground surface, etc.).

46 ‘Variance’ , is a statistical term this being based upon the standard deviation, which is normally denoted by the Greek symbol ‘σ’. Thus, variance can be defined as: ‘The mean of the squares of the standard deviation’ (Bajpai, et al., 1979). Thus, σ = √Variance, or more specifically for production operations: n � �s = ¯  �n −  ċ � (xj − x) j= *s = the standard deviation of a sample from a production batch run.

308

Chapter 7

depending upon whether a fine, medium, or coarse surface texture is obligatory. Due to the variability in any production process being one of a ‘stochastic output’ 47, such surface texture values do not reflect the likely in-service performance of the part. Neither the surface topography, nor its associated integrity has been quantified by assigning to a surface representative numerical parameters. In many instances, ‘surface engineering’ 48 is utilised to enhance specific component in-service condition. It was mentioned above that in many in-service engineering applications the accompanying surface texture is closely allied to its functional performance, predominantly when one, or more surfaces are in motion with respect to an adjacent surface. This close proximity between two mating surfaces, suggests that the smoother the surface the better, but this is not necessarily true if the surfaces in question are required to maintain an efficient lubrication film between them. The apparent roughness of one of these surfaces with respect to the other, enables it to retain a ‘holdingfilm’ in its associated topographical ‘valleys’49. While another critical factor that might limit the designer’s choice of the smoothness of an engineering surface’s selection, is related to its production cost (i.e. see Fig. 161b). Therefore, if the designer requires a very smooth machined surface, it should be recognised that its manufacturing time is considerably longer – so its respective cost will be greater to that of a rough surface, this being exacerbated by a very close dimensional tolerance requirement.

47 ‘Stochastic processes’ , are defined as: ‘A process which has a measurable output and operating under a stable set of conditions which causes the output to vary about a central value in a predictable manner’ (Stout, 1985). 48 ‘Surface engineering’ , is applying suitable discrete technologies to create surface films (e.g. 10 to 100 nm thick), or by manipulating the surface atomic layers (e.g. 2 to 10 atomic layers, approximately 0.5 to 3 nm), to enhance the ‘engineered’ surface condition (i.e. Source: Vickerman, 2000). 49 ‘Surfaces’ , are recognised to have topographical features that mimic the natural world. So a regular/irregular engineering surface can exhibit both peaks and valleys, not unlike mountainous terrain.

7.5.1 Parameters for Machined Surface Evaluation In order that a machined workpiece’s surface texture can be determined using stylus-based (two-dimensional) instrumentation, three characteristic lengths are associated with this surface’s profile (i.e. see Fig. 163a), these are: 1. Sampling length50 – is determined from: the length in the direction of the X-axis used for identifying the irregularities that characterise the profile under evaluation. Therefore, virtually all surface descriptors (i.e. parameters) necessitate evaluation over the sampling length. Reliability of the data is enhanced by taking an average of the sampling lengths as depicted by the evaluation length shown in Fig. 162a. Most of today’s stylus-based surface texture instruments undertake this calculation automatically, 2. Sampling length – can be established as: the total length in the X-axis used for the assessment of the profile under evaluation. From Fig. 163a, this length may include several sampling lengths – typically five – being the normal practice in evaluating roughness and waviness profiles. The evaluation length measurement is the sum of the individual sampling lengths (i.e. it is common practice to employ a 0.8 mm sampling length for most surface texture assessments), 3. Traverse length – can be defined as: the total length of the surface traversed by the stylus in making a measurement. The traverse length will normally be longer than the evaluation length (i.e. see Fig. 163a), this is due to the necessity of allowing ‘run-up’ and ‘over-travel’ at each end of the evaluation length. These additional distances ensure that any mechanical and electrical transients, together filter edge effects are excluded from the measurement.

50 ‘Sampling length’ , is often termed ‘Meter cut-off ’ , or simply the ‘cut-off ’ length and its units are millimetres. The most common cut-offs are: 0.25, 0.8, 2.5, 8.0, 25.0 mm. The 0.8 mm sampling length will cover most machining production processes. In any surface texture evaluation, it is essential that the cut-off is made known to the Inspector/Metrologist reviewing this surface topographical data.


309

. Figure 161. ‘Lay’ indicated on drawings, plus the relative cost of manufacture for different production processes

The number of two-dimensional surface profile parameters that have been developed over the years for just the stylus-based instruments – discounting the three-dimensional contact and non-contact varieties, has created a situation where many users simply do not understand, nor indeed comprehend the intrinsic differences between them! A term was coined some years ago to show exasperation by many metrolo-

gists’ with this ever-increasing development of such parameters. Many researchers and companies were totally disenchanted with their confunsion and plight, so they simply called the predicament: ‘parameterrash’. However, here we need not concern ourselves with this ‘vast expanse of surface descriptors’ , as only a few of the well-established parameters and discuss just the most widely-utilised ones. It is worth making

310

Chapter 7

. Figure 162. Anticipated process ‘roughness’ and their respective grades. [Source: ISO 1302, 2001]


. Figure 163. Surface texture data-capture, with techniques for the derivation of the arithmetic roughness parameter: Ra

311

312

Chapter 7

the point, that all of these two-dimensional surface parameters can be classified into three distinct groupings and just some of these parameters are: 1. Amplitude parameters (Ra, Rq, Wa, Wq, Pa, Pq)51 – with Ra52 being universally recognised for the ‘international’ parameter’ for roughness. It is: ‘The arithmetic mean of the absolute departures of the roughness profile from the mean line’ (i.e. see Fig. 163b and c). It can be expressed as follows:

lr

Ra = �lr � � z(x) � dx 

(units of m)

NB The Ra parameter is often utilised in applications to monitor a production process, where a gradual change in the surface finish can be anticipated, making it seem to be ‘ideal’ for any form of machinability trials, but some caution is required here, as will be seen shortly in further discussion concerning this ‘surface descriptor’. By way of comparison, another previously used amplitude parameter is given in Appendix 10 and is the ‘RZ(JIS)’ (i.e. 10-point height) parameter. Other useful parameters of the assessed profile, to be shortly discussed in more detail, include: ‘Skewness’ (Rsk, Wsk, Psk), which is often utilised in association with ‘Kurtosis’ (Rku, Wku, Pku), producing the socalled: ‘Manufacturing Process Envelopes’ – as a means of ‘mapping’ and correlating machined surface topographies. 2. Spacing parameters (Rsm, Wsm, Psm) – can be defined as: ‘The mean spacing between profile peaks at the mean line, measured within the sampling length’ (i.e. depicted along a machined cusp – at differing

feedrates in Fig. 169a and b). It can be expressed in the following manner:

Rsm = �n � si =

n = number of peak spacings.

i=n i=

Where:

NB The Rsm parameter needs both height and spacing discrimination and, if not specified the default height bias utilised is 10% of: Rz, Wz, or Pz, – where these are the ‘Maximum height of profile’. As can be seen from the ‘idealised’ machined surface topography given in Fig. 169a and b, the spacing parameters are particularly useful in determining the feed marks. Moreover, the Rsm parameter relates very closely to that of the actual programmed feed rev–1 of either the cutter, or workpiece – depending on which production process was selected. See also, Appendix 10 for a graphical representation of the previously utilised ‘High Spot Count’ (HSC) parameter. 3. Hybrid parameters (Rmr, Wmr, Pmr, R∆q, W∆q, P∆q, Rpk, Rk, Rvk) – each of these ‘hybrids’ will now be briefly mentioned. Rmr, or its alternative notation Mr is the ‘Material ratio curve’ , which can be defined as: ‘The length of the bearing surface (expressed as a percentage of the evaluation length ‘L’) at a depth ‘p’ below the highest peak (i.e. see Fig. 165). – Rmr: It is often known as the ‘Abbott-Firestone curve’ , the mathematics of this Rmr-curve can be expressed in the following manner:

51 The designation of the letters follows the logic that the parameter symbol’s first capital letter denotes the type of profile under evaluation. For example, the: Ra* – calculated from the roughness profile; Wa – derives its origin from the waviness profile; with the latter in this logical sequence, namely the Pa – being derived from the primary profile. Here, in this discussion and for simplicity, only the first term in the series – e.g. ‘Ra’ notation – will be used. *Ra is today shown in the International Standard (i.e. ISO 4287: 1997) as being denoted in italics, while in the past, it was usually shown as follows: ‘Ra’ , but even now, many companies still use this particular notation. 52 Historically, the classification of the relative roughness of surfaces was initially developed in England and was then termed: ‘Centre Line Average’ (CLA), while in the USA its equivalent term was the ‘Arithmetic Average’ (AA).

XS+XS+XS...+XSn n

Rmr =

b+b+b=B...+bn n

�  =

 n

i=n

� bi

i=

NB This ‘Material ratio curve’ represents the profile as a function of level. More specifically, by plotting the bearing ratio at a range of depths in the profile trace, the manner by which the bearing ratio changes with depth, provides a method of characterising differing shapes present on the profile (i.e. see Fig. 165 and Appendix 10).

– R∆q:

The R∆q parameter, can be defined as: ‘The root mean square (rms) slope of the profile within the sampling length’ (i.e. see how its angle changes at differing machining feedrate conditions shown in Fig. 169b and c), it can be mathematically expressed as follows:


�

lr

¯  dx R � q = �lr � [θ(x) − θ]

Where:



lr

θ¯ = �lr � θ(x)dx 

θ = slope of the profile at any given point.

• Rpk, Rk, Rvk:

These parameters (i.e. see Appendix 10 for graphical representations of the parameters), were originally designed for the control of potential wear in automotive cylinder bores in volume production by the manufacturing industry. Today, Rpk, Rk and Rvk are employed across a much more diverse-field by a range of industries. Such hybrid parameters are an attempt to explain – in numerical terms, the respective form taken from the profile’s trace of the ‘material ratio curve’ (Rmr), hence: – Rpk parameter – is the ‘reduced peak height’ , illustrating that the top portion of a bearing surface will be quickly worn-away when for example, an engine initially begins to run, – Rk parameter – is known as the ‘kernal roughness depth’ , therefore the long-term running – ‘steady-state wear’ of this surface will influence for example, the performance and life of the automotive cylinder(s), – Rvk parameter – is the ‘trough depth’ this indicates that the surface topography has an oilretaining capability, specifically via the ‘trough depths’ which have been purposely ‘crosshoned’53 into the bore’s surface.

Arithmetic roughness parameter (Ra) Although the Ra ‘amplitude parameter’ has been widely quoted ‘Internationally’ , there are a few provi-

53 ‘Cross-honing’ , uses either: (fine) Abrasives/CBN/Diamond – ‘stones’ , that are fitted into a ‘honing head’ which then rotates and oscillates within a hole, or an engine’s bore. The critical parameters are the rotational speed (Vr) oscillation speed (Vo), the length and position of the honing stroke, the honing stick pressure (Vc). The inclination angle of the cross-honing action, is a product of the up-/down-ward head motion (Vo) and the rotational speed for the head (Vo). This complex action of rotating and linear motion, generates the desired cross-honed ‘Lay-pattern’ within the bore – for improved oil retention.

313

sos, or conditions that must be met, if it is to be utilised satisfactorily, these are: • The Ra value over one sampling length represents the average roughness. The effect of a spurious nontypical peak, or valley within the profile’s trace being ‘averaged-out’ so will have only a small influence on the Ra value obtained; • The evaluation length contains several sampling lengths (Fig.163a), this ensures that the Ra value is representative of the machined surface under test; • An Ra value alone is meaningless, unless quoted with its associated metre cut-off (λc) length. Repeatability of the Ra value will only occur at an identical length of metre cut-off; • If a dominant surface texture pattern occurs (Lay), then the Ra readings are taken at 90° to this direction; • That Ra does not provide information as to the shape of either the profile, or its surface irregularities. Different production processes generate diverse surface finishes, for this reason its is usual to quote both the anticipated Ra numerical value along with the actual manufacturing process; • Ra offers no distinction between peaks and valleys on the surface trace. The most confusing argument concerning the use of an Ra value alone, is that its numerical value is not only meaningless, but it can have catastrophic consequences if interpreted incorrectly. These opinions can be substantiated by close observation of Fig. 164a, where an identical numerical Ra value produces widely divergent surface topographies. In addition, if a designer’s engineering application called for a ‘bearing surface’ (Fig. 164ai), rather than a ‘locking surface’ (Fig. 164aiii), then the numerical value of 4.2 µm in isolation, becomes pointless, as it tells the designer nothing about the ‘functional’ surface topography. This problem is exacerbated when the wrong surface topography is selected for a specific engineering application. For example, a ‘locking surface’ applied to a bearing industrial application in a harsh environment, can be expected to catastrophically fail after very little in-service time.

Skewness (Rsk, Wsk, Psk) and Kurtosis (Rku, Wku, Pku) Parameters These surface descriptors of ‘skewness’ and ‘kurtosis’ are often derided as simply ‘statistical’ amplitude parameters, that can introduce spurious results and as a consequence, having little use in engineering applications. However, when used in the correct context, they can provide a valuable insight into the overall shape of

314

Chapter 7

. Figure 164. Arithmetic roughness parameter (Ra) can give a misleading representation of the surface topography, so skewness (Rsk) and kurtosis (Rku) may provide help in the interpretation of the surface. [Courtesy of Taylor Hobson]


the surface topography, if used in conjunction with the Ra value (Fig. 164). This is particularly true for many machining applications and where a comparison of ‘Manufacturing process envelopes’ is required – more on this subject shortly. Both skewness and kurtosis can be mathematically defined, as follows: • Skewness parameter (Rsk): The measure of the symmetry of the profile about the mean line’. Skewness has the ability to distinguish between asymmetrical (i.e. having ‘biased tails’) of identical Ra numerical values (i.e depicted in Fig. 164bii). Skewness can be expressed as follows: lr

Rsk = �R  q[�lr � z  (x)dx]. 

• Kurtosis parameter (Rku):

Measure of the ‘sharpness’ of the surface profile’. Kurtosis can be expressed in the following manner (i.e shown in Fig. 164bi):

Rku = �R  q[�lr � z  (x)dx] .

lr



Discussing both ‘skewness’ and ‘kurtosis’ in turn, implies that they are separate parameters, but this is not the case, when one observes Figs. 164bi and bii. However, to begin with and for ease, each of these parameters will be individually mentioned. The Rsk parameter cannot distinguish if a profile’s trace has peaks54 that are relatively evenly distributed above, or below the mean line (i.e. Fig. 164aiii), or being influenced by any isolated peak, or valley (i.e. this topography is shown to good effect in Fig. 164ai) – within the sampling length. The Rsk parameter of an amplitude distribution curve as illustrated in Fig. 164bii, indicates a certain amount of bias that could be either in an upward, or downward direction (i.e shown either as: left- and right-ward, in this example). The amplitude distribution curve’s contour can be very informative as to the overall structure of the surface topography. If this curve is symmetrical in nature then it indicates regularity of the profile trace (Fig. 164ci), conversely, an asymmetrical surface’s trace will be indicative of a ‘skewed’ amplitude distribution curve (Fig. 164cii). Utilising

54 ‘Peaks’ , are often known by a variety names, such as ‘spikes’ , or to use a more scientific term this would be: ‘asperities’.

315

the skewness parameter distinguishes between profile traces having if not similar, or identical Ra values. Machined surfaces can exhibit a broad range of surface topography-related conditions. For example, a boring operation with a relatively long length-to-diamter ratio may result in bar deflection (i.e. elastic deformation) and occasion the cutting insert to deflect, producing large peak-to-valley undulations along the bore (waviness). Super-imposed onto these longer wavelengths are small-amplitude cyclical peaks – periodic oscillations, indicating vibrations resulting from the boring process. Thus, the consequential surface profile for the bored hole, would portray the interactions from the boring bar deformations and any harmonic oscillations. The likely outcome of such a boring operation and the bar’s relative motion, would be reflected in the profile trace, exhibiting a low average profile height, but with a large range of height values. Moreover, a highly negative skewness is indicative of a good bearing surface, particularly if some valleys are present to allow for subsequent oil-retaining abilities (Fig. 164cii). The shape of the amplitude distribution curve in terms of its relative ‘flatness’ , or ‘spikeness’ can also relay useful information concerning the ‘dispersion’ , or ‘randomness’ of the surface profile, which can be quantified by means of the surface descriptor known as kurtosis (Rku). However, unlike skewness (Rsk), kurtosis can detect if the profile peaks are distributed in an even manner across the sampling length’s trace (Fig. 164ci), or vice-versa. This latter case of producing both a ‘spiky’ and ‘skewed’ distribution having either a positive, or negative skew to its resultant distribution with its associated surface topography – is shown in Fig. 164cii. As a consequence of this ability to differentiate the variations of the actual surface topography, Rku is a beneficial parameter in the prediction of in-service component performance, with particular respect to any potential lubrication-retention issues, or for any succeeding industrial wear behaviour circumstances.

Material Ratio Curve (Rmr) The material ratio curve (Rmr) represents the profile as a function of level. Specifically, by plotting the bearing ratio at a range of depths for the profile, the manner by which the bearing ratio changes with depth, provides a method of characterising different shapes present on the profile (i.e. see Fig. 165). The bearing area fraction

316

Chapter 7

. Figure 165. Hybrid surface texture parameters – spacing and depth. [Courtesy of Taylor Hobson]


can be defined as: ‘The sum of the lengths of individual plateaux at a particular height, normalised by the total assessment length’ – with this parameter designated by Rmr (Fig. 165ai and aii). In the majority of circumstances mating surfaces demand specific ‘tribological conditions’ 55: these are the direct result of particular machining operational sequences. Normally, the initial production operation will establish the general shape of the machined surface – by ‘roughing-out’ – providing a somewhat coarse finish, with subsequent operations to improve and enhance the finish, resulting in the desired design properties. This machining strategy provides the operational sequence that will invariably remove surface peaks from the original machining process, but often leaves any deep valleys intact. This standardised industrial machining practice of ‘roughing and finishing’ , produces a type of surface texture known as a ‘stratified surface’. When these ‘stratified surfaces’ occur, the height distributions tend to be negatively skewed making it somewhat difficult for an ‘averaging parameter’ like Ra to represent the surface effectively to the designer’s specification, or in matters concerning quality control. In the diagrammatic representation for the derivation of the Abbott-Firestone curve56, or ‘Material ratio curve’ (Rmr) shown in Fig. 165b, this enables the user to select differing slices, or depths through the profile, with these ‘slices’ having a specific ratio for the proportions of air-to-material. The top of the highest peak within the profile trace having been evaluated, establishes the reference, or zero percentage line for the Rmr curve. Calculation of this curve is influenced by the largest peak’s height in relation to the others, although in reality, the effect of a single peak on a surface’s in-

55 ‘Tribology’ , was a technology that originated about 40 years ago, its name was derived from the Greek ‘τριβοσ’ translated: ‘Tribos’ – meaning ‘rubbing’ so that the literal translation would be the ‘science of rubbing’. Today, tribology can be more accurately defined as: ‘The science and technology of interacting surfaces in relative motion and of related subjects and practices’ (Williams, 1996, et al.). 56 ‘Abbott-Firestone curve’ , was named after two researchers in the USA working in the early days of surface topography – circa 1933. They defined the ‘Bearing area fraction’ at a given height above the mean line as: ‘The proportional length of all plateaux which would result if the surface were abraded away, down to a level plane at that height’ (Thomas, et al., 1999).

317

service function has little significance. In order to minimise the effect of a single peak on the Rmr curve, an artificially-induced reference line is chosen to shift this line below the highest peak – as illustrated in Fig. 165b – this value now being expressed as a material ratio percentage (i.e. in this example the Rmr being at 45%). Specifying for example, an Rmr at the 5% reference line (Fig. 165d), testifies that the top 5% of the profile is not included as part of the calculation for the ‘material ratio’. The selection of the zero line beneath the highest measureable peak will be dependent on the topography of the associated peaks in the profile trace, but industrial practice suggests the reference level is usually set between 2% to 5%.

7.5.2 Machined Surface Topography Tool nose insert geometry (Fig. 166a) plays a important role in the resultant machined surface topography of a turned component’s surface. Longitudinal turning operations will leave the residual effects of this partial tool nose geometry on the workpiece surface as ‘machined cusps’ forming the dominant profile on the turned surface topography (Figs. 166b and 169). This geometry is a complex relationship of curved and linear inter-connected portions whose insert profile is ‘fashioned’ onto the turned surface – being appreciably influenced by the pre-selected feedrate (Fig. 166a). Therefore, according to the ‘Shaw-model’ (1984, et al.) – being a somewhat ‘simplified geometry’ without chip-breakers present, with an enlarged view of the turning insert’s nose region (Fig. 166a and b), its resultant surface topography can be defined by three distinct insert-related factors, these are: 1. Nose radius – r = OT (i.e. illustrated in Fig. 166b), 2. End-cutting edge angle – denoted by Ce (i.e. shown in Fig. 166a), 3. Side-cutting angle – denoted by Cs (Fig. 166a). In effect, there are three discrete ‘turning-cases: I, II and III,’ that may occur when utilising the tool nose cutting insert geometry shown (Fig. 166a), when computing the theoretical peak-to-valley (Rth) surface texture. When a very light DOC and its associated feedrate is imparted onto the workpiece’s surface, then ‘Case I’ conditions will be met – illustrated by the high-density cross-hatched portion of the insert‘s geometry depicted in Fig. 166a. The mathematics of the theoretical cusp height (Rth) in Fig. 166b, is given by the following expression:

318

Chapter 7

. Figure 166. How the ‘residual influence’ of the turning insert’s partial geometry combined with feedrate, affect the subsequent machined component surface topography


319

Rth = OT – OU = r – (r2 – t2/4)½ ∴Rth ≅ t2/8r

Case I

When medium values of the turned feedrate are utilised – as is the situation in ‘Case II’ (Fig. 166a), then both the tool nose (r) and the partial effect of the ‘End-cutting angle’ (Ce) must be considered, when estimating the deeper ‘Rth’ value. Assuming that ‘t’ is equal to ‘JL’ (Fig. 166a and b), then the following geometrical conditions are met: t = JK+HE+NL ∴ t = [r2 – (r – Rth )2]½ + r sinCe + r(Rth/r – 1 + cosCe)cotCe

(Fig. 166a) Case II

NB This equation is valid, just as long as the ‘t’ length lies between positions: ‘SE’ and ‘FR’. If even larger values of ‘t’ are utilised (i.e. higher feed rates), then the three geometrically curved and linear portion’s of the cutting insert’s profile, will affect resultant turned surface topography. Namely, this surface topography after machining, will now be comprised of the sum of the portions of the: end-cutting angle ‘Ce’; nose radius ‘r’; together with the side-cutting angle ‘Cs’ , as follows: Rth

= AB – AG + r

� Rth =

t tan C s +cot C e

−

(Figs 166a and b) r cos(−C e �−C s �) sin(−C e �+C s �)

+ r

Case III

The equations given in ‘Cases: I, II and III’ , can be utilised to produce a ‘non-dimensional’ graphical plot for various cutting insert approach angles (i.e. see Shaw, 1984 – p 513) enabling the estimated value of ‘Rth’ to be found – for any combination of: ‘t’; ‘r’; ‘Ce’; plus ‘Cs’. Historically, it has been well-known that there exists a ‘minimum undeformed chip thickness’ 57 and if a value

57 ‘Minimum undeformed chip thickness’* was proposed some years ago by Sokolowski (1955), where it was suggested that a ‘limiting value’ occurred for a chip thickness, below which a tool simply rubs. *In the experiments undertaken at the time by Sokolowski, using a very sharp and honed single-point turning tool with a cutting edge nose radius of 12 µm at a cutting speed of 210 m min–1, the smallest cut that could be taken had a depth of 4 µm. For additional information on succeeding research work on this sbject, also see Footnote 58.

is smaller than this actual chip-thickness, it is not possible to form a successful chip, as a result, only rubbing will occur – this being a combination of many interrelated factors. Applying this ‘minimum undeformed chip thickness’ concept to the insert’s secondary cutting edge, it was found that a small triangular portion of workpiece material which should have been removed, is normally left ‘in-situ’ (Fig. 167). This fraction of the workpiece left behind, has been the subject of intensive interest by previous machinability researchers over a number of years and is known as the ‘Spanzipfel’ 58. In Fig. 167b, the raised portion (i.e. Spanzipfel) occurs at increments equivalent to that of the feed rev–1 and is illustrated in Fig. 167a. An expression has been derived to take account of this Spanzipfel on the theoretical surface texture, as follows: R′th = t2/8r + tm/2 (1+ rtm/2) NB In this equation, the 2nd term represents the contribution of the Spanzipfel. Both the theoretical values for the surface texture and actual ones from the trials are in close agreement, which was not expected, as the Spanzipfel will be subject to some plastic deformation – as it comes into direct contact with the tool’s clearance face. In both end- and face-milling operations (Fig. 168a), the machining process is one of interrupted cutting, tending to impart an isotropic surface topography to the milled surface (Fig. 168b). If in Fig. 168b (i.e. top diagram) the excess stock material is removed by face-milling, the resultant surface exhibits quite a complex surface topography. This milled surface complexity results from the cutter’s trailing edge, as it moves over the previously cut surface at the periodic and pre-selected feedrate. This periodic surface topog-

58 ‘Spanzipfel’ , was initially investigated and analysed by Brammertz (1960). He was particularly interested in the Spanzipfel’s affect of the resulting machined surface topography/texture. Later work by Pahlitzsch and Semmler (i.e. between 1960 to 1962), looked at the ‘minimum undeformed chip thickness’ and the Spanzipfel’s influence when fine-turning AISI 1045 steel workpieces with specially-sharpened ceramic tooling. Here, they found that machining with this much more abrasive-resistant ceramic tooling, the ‘minimum undeformed chip thickness’ height could be dramatically-reduced to just 1 µm. Cutting speed utilised in these tests was 400 m min–1 and the machinability trials used an in-cut time of just 6 seconds – in order to maintain a sharp cutting edge on the tools.

320

Chapter 7

. Figure 167. The affect of the ‘minimum undeformed chip thickness’ on the machine topography, producing the socalled ‘Spanzipfel effect’


raphy will not occur, if either a ‘slight’ milling spindle camber is utilised (i.e. see Fig. 86), or some form of ‘torque-controlled machining’ (TCM) system is employed, that usually incorporates some form of ‘adaptive control constraint’ (ACC) engagement of the feeding-system – additional information on the subject of TCM will be given in the final chapter. The major reason for TCM feeding variability, is because the torque is monitored and as the stock height varies along the workpiece’s length, the torque is either lessened by slowing the feedrate, or increased, thereby maintaining relatively constant cutting forces on the tooling. This form of ‘adaptive control’ (ACC) by constraining the cutter’s feed, will impart a variable surface topography to the milled surface. In the previous milling scenario where there was no necessity for an TCM requirement, the milling cutter’s rotation in combination with the feed for a given ‘cut-off ’ slightly changes the milled surface topography. It introduces some variability to the respective cusp heights along the workpiece’s milled surface (Fig. 168b). Here, the periodic nature of the surface topography is regular (i.e. a constant ‘Sm’), but its periodicity marginally changes according to whether the surface is measured at the centre of the cut, or offset across the face-milled surface, which has an some effect on the relative cusp heights. Conversely, across the milled surface at arbitrary positions denoted in these examples as: ‘X-X’ and ‘Y-Y’ (Fig. 168b), the topography fluctuates somewhat at a predetermined and quantifiable interval, depending upon where the surface trace was taken. Hence, any milled surfaces with a non-directional, or indefinable Lay – as is the case for most re-cutting effects introduced by either end-, or face-milling operations, should not simply have an arbitrary Ra quoted on the engineering drawing, as this has been shown (Fig. 168b) to be somewhat meaningless. Milled surfaces having these latter characteristics, requiring the need to indicate the maximum tolerable Ra value for a given Lay direction – in a similar fashion to that of an anisotropic machined surface topography. Returning to the longitudinal turned surface topography once more. If consideration is given to the ‘idealised’ turning surface (Fig. 169a and b), then for a constant tool nose insert geometry and undeformed chip thickness, as the feed rev–1 is increased, the surface texture will be degraded. The residual cusps that periodically occur on the turned surface after the tool’s passage along the part, are a product of two previously described inter-related phenomena (Fig. 154),

321

namely: the ‘moving-step effect’; in conjunction with the ‘emerging diameter’. This relationship diminishes the notable height of the turned cusps with a reduced feed rev–1, while it increases with larger feed rev–1 – this aspect being depicted in Figs. 169a and b, respectively. If a proportionally larger feed rev–1 is selected, this increases the residual influence of the tool nose contact region on the surface – as formerly mentioned when discussing Fig. 166. As a result of higher feeds the RSm increases, which inevitably heightens the cusps, promoting a larger recorded value in Ra, in association with greater angles for ∆q (Fig. 169b). Of course, the opposite is true in the case of reduced feeds (Fig. 169a). Explicitly, as a smaller tool nose contact region occurs – with reduced feed rev–1 (i.e ‘Case I’ in Fig. 166a), this has the effect of producing a smaller cusp height (Rth) and its accompanying Ra, giving a more shallow value of ∆q; due to the partial curvature of the tool nose radius tending toward zero as it approaches tangency with that of the workpiece’s axis (Fig. 169a). The dominant factor here is the feedrate, as it has an enormous influence on the resultant turned cusps, affecting their: height; profile shape; periodicity; in association with the pre-selected tooling geometry; these factors radically influencing both the measurement and magnitude of the machined surface topography, which in turn, affects the surface texture parameters. If just the Ra parameter had been specified, it could not adequately describe the nature and condition of the surface topography in any consequential manner. Assuming that standardised cutting conditions are selected for a turning operation: workpiece composition; rotational speed; feedrate, undeformed chip thickness, with only the tool nose geometry change, then the resulting surface topography will be markedly different. This divergence in machined topography is illustrated to good effect in Figs. 169c and d, where turned ferrous P/M compacts are depicted. Here, two extremes of cutting insert nose radii are utilised: Fig. 169c the nose radius was 0.8 mm; whereas in Fig. 169d a button-style insert (φ12 mm) having the equivalent of 6 mm nose radius was used. The turning insert with the 0.8 mm nose radius produced visually-apparent regularly-spaced cusps and despite the fact that a new turning insert was employed, there is evident signs of tears, laps and burrs present around the turned periphery. In contrast in Fig. 169d, the turned surface topography appears appreciably smooother in profile, although even here, the surface topography is marred by similar tears, etc., which might be a cause for its potential rejection. This smoother surface was due to

322

Chapter 7

. Figure 168. In face- and end-milling operations, due to the ‘re-cutting effect’ of the trailing insert cutting edges, they impart an ‘almost’ isotropic milled surface topography to the part


. Figure 169. How the feedrate influences the machined cusp/surface roughness value ‘Sm’ and its affect on the waviness parameter ‘Δq’, plus surface topography of actual longi-

323

tudinal CNC turned P/M ferrous compacts – cut with differing nose radii. [Courtesy of Joel (UK) Ltd.]

324

Chapter 7

the considerably larger effective tool nose radius, acting like a ‘wiper-blade’ blending-out and obliterating the surface’s cusps. This technique of utilising a large tool nose geometry has traditionally been used by precision turners to improve the overall surface finish.

7.5.3 Manufacturing Process Envelopes The principal features of manufacturing process envelopes and indeed, for many amplitude distribution curves is that they can be approximated by the socalled ‘beta-function’ – ‘β’ (Fig. 170a). Here, the function has two parameters that are independent of one another enabling them to be used as a means of surface characterisation. The notation ‘a’ is the allocated weighting for the profile ordinates measured from the lowest valley and above, with notation ‘b’ being given to weighting the profile from the highest peak down. Hence, peaks and valleys have accordingly different weights. One of the problems that has arisen from utilising this technique for a topographical profile, which has somewhat discredited them for certain applications, is how and in what manner can one determine ‘a’ and ‘b’. The ‘beta-function’ is normally defined within a set range of: 0→1, being expressed in the following manner: 

β(a, b) = � z a− (a − z)b− dz . 

If by changing the range of the ‘beta-function’ equation above, from: 0→1 to Rp + Rv, or indeed with that of Rt, then substituting σ (i.e. the standard deviation of the distribution) with Rq, the beta-function parameters ‘a’ and ‘b’ become: a = Rv (Rv Rp – Rq2/Rt Rq2) b = Rp (Rv Rp – Rq2/Rt Rq2). The fact that any dominant peak, or valley within the assessment length is only raised to a unit power, infers additional stability over the ‘skewness/kurtosis approach’. The problem with this ‘beta-function’ method is in accurately determining ‘sound’ results from the Rv and Rp, which confirms the difficulty that obtaining information from peak/valley measurement and then deriving valid information is fraught with complications. In Figs. 170: ‘ai’ it is symmetrical; with ‘aii’ being asymmetrical; for their respective ‘beta-func-

tions’ , these relationships are based upon a class of statistically-derived ‘Pearson distributions’ 59. In the symmetrical case (Fig. 170ai) the skewness equates to zero; conversely, for an asymmetrical series of results (Fig. 170aii), skewness can be either positively, or negatively skewed (i.e see Fig. 164bii). Nevertheless, even allowing for these limitations, an example of the groups of manufacturing process envelopes for a range production processes is illustrated in Fig. 170b. Here, the production processes can be simplistically classified and grouped into either a ‘bearing’ , or ‘locking’ surface topography. The ‘bearing-/locking-groupings’ indicate that certain production processes can achieve specific functional surfaces for particular industrial applications. These ‘groupings’ (Fig. 170b), also indicate that the general classifications are less distinct than might otherwise be supposed, as certain processes can be provide either a ‘locking’ , or a ‘bearing’ surface condition – termed ‘intermediate groupings’. Typical of such an ‘intermediate group’ are the P/M drilled compacts. One reason for this is that any P/M ‘secondary machining’ often utilises twist drills, which may produce

59 ‘Pearson product moment correlation coefficient’* – to give it its full title, is a statistical association utilised when a ‘relationship’ exists between several quantities and it is a measure of the extent of this affiliation, thus producing its ‘correlation coefficient’ which can then be utilised. *For example, having a sample of pairs of observations ‘x’ and ‘y’ , the value ‘r’ of this ‘Pearson coefficient’ , is given by the ‘generalised formula’ , below: r = Σ(x – x ) (y – y) /√ Σ(x – x)2 Σ(y – y)2 (Bajpai, et al., 1979): The calculated ‘Pearson coefficient’ should lie either close to –1, or +1. If the calculated value is close to –1 then the ‘straight-line trend’ is in a downward direction, conversely, if the calculated value is close to +1 then the ‘straight-line trend’ direction is upward. If the value is close to zero (0), then no correlation exists, so the pairing of the data is disparate and cannot be utilised. With calculated data that is reasonably close to either –1, or +1, a ‘regression line’ (i.e. using linear regression) can be calculated based upon the general straightline formula: Y = a + bX Where a ‘regression line of Y on X’ for the two constants ‘a’ and ‘b’ respectively are: a = Σy – bΣx/n b = nΣxy – (Σx)(Σy)/n(Σx2) – (Σx)2 Where a ‘regression line of X on Y’ for the two constants ‘a’ and ‘b’ respectively are: b = nΣxy – (Σx)(Σy)/n(Σy2) – (Σy)2 a = Σx – bΣy/n NB The ‘regression line’ being the equivalent of the ‘least squares line’ , allows data on each axes to be compared – with some degree of confidence.(Wild, et al.,1995)


. Figure 170. The ‘beta-function’ and typical ‘manufacturing process envelopes’

325

326

Chapter 7

a ‘saw-toothed profile’ to the hole’s surface, along with pores in the compact that are open to the ‘free-surface’ of the hole. The hole topography may have this ‘sawtoothed effect’ present, it being a combination of the drill’s partial lip and margin occurring at the feed rev–1 periodicity, formed by the drill spiralling-down and around the hole’s periphery. Hence, the drill’s passage creates a positive skewness via drilled ‘saw-toothed cusps’ , while the pores can introduce negative skewness – creating a potential ‘intermediate group’ to the manufacturing process envelope groupings.

7.5.4 Ternary Manufacturing Envelopes (TME’s) In machining operations the dominant factor that influences surface topography has been shown to be the tool’s feedrate. In Fig. 171, the feedrate, in conjunction with the principal factors such as surface texture (Ra) and roundness (i.e least squares circle – LSC), are utilised to define the limits for these ‘Ternary manufacturing envelopes’ (TME’s). By using such diverse factors as: surface texture, roundness and processing parameters (feedrate), for the major axes on the ternary graph, enables the surface to be characterised in a unique manner. Such TME’s differ quite considerably from the more usual and restricted ‘manufacturing process envelopes’ alluded to in the previous section – the skewness and kurtosis axes of the manufacturing envelopes, might otherwise mask crucial information. The ‘TME approach’ gives a psuedo three-dimensional representation on its ternary axes, which can be exploited to illustrate how the influence of changing a parameter – such as feedrate – modifies the relationship of the associated surface texture and roundness values for the final machined result. As an example of the effectiveness of this TME approach to the complex problem of machining data analysis, Fig. 171 has been drawn from an actual machinability trial. If one observes this TME graph closely for a pre-selected range of turning and boring processes, indicated in Fig. 171, with specific reference here, to turning operations – by way of illustrating the TME’s expediency. The TME shows how – for the turning operations – at low feedrate (0.10 mm rev–1) the surface texture is closely confined to a relatively small spread of values – nominally around 0.5-1.5 µm Ra, whereas its associated roundness lies between approximately 5 and 50 µm LSC. As the feedrate in-

creased in an arithmetic progression to 0.25 mm rev–1, the range of the surface texture bandwidth proportionally expanded to 1.5 at approximately 5-6.5 µm Ra, with a corresponding roundness ranging from 8 to 48 µm LSC, giving a proportional bandwidth of 1.6. As the feedrate was raised even higher, to 0.40 mm rev–1, it was not surprising to note that this also produced increases in both the surface texture and its proportional bandwidth, with similar values with respect to its roundness. These ‘machinability and metrology trends’ allow examination of both the bandwidth variability and the affect of different feedrates on other disparate factors – such as its machined roundness. Similar trends occurred for the boring operation, but here only two feedrates were employed, by application of this analysis technique via the ‘TME-approach’ to a concise machinability trial, complex analysis of the TME is possible. The pseudo three-dimensional graph, offers perhaps an unusual insight into the multifaceted inter-relationships that exist after workpiece machining. The TME shows that simply examining one metrological parameter in isolation to those that could affect it, may mask vitally important relationships and trends that would otherwise remain unseen. By careful selection of the parameters for the respective axes, perhaps based upon the feedrate (i.e. here, normally situated along the X-axis), allows an appreciation of the whole surface at any instant along the three graph’s axes.

7.6 Machining Temperatures Ever since Taylor in 1907, recognised that elevated tool and workpiece temperatures in metal cutting played a crucial role in influencing tool edge wear rates, the subject has been one of intensive study. Moreover, that the tool/chip interface temperature has a controlling influence on the rate of crater wear and the fact that tool life can be drastically curtailed by these induced machining temperatures, as such, the topic has received considerable research attention. Here, space will only allow a brief resumé of this complex temperature-induced machining problem. During metal cutting in particular, there are several temperature effects that need to be considered. In Fig. 51, an orthogonal single-point cutting operation is schematically illustrated, indicating the distribution of heat sources within the three deformation zones. In


. Figure 171. ‘Ternary manufacturing envelopes’ for the production processes of turning and boring, axes: feedrate, roundness and surface texture

327

328

Chapter 7

particular, the heat generated in the main ‘body’ within the cutting region via both the primary and secondary zones is a result here, of the workpiece’s plastic deformation. Still more intensive heat is generated at the tool/chip interface – along the rake face, with the majority of heat being swept away with the chips, while the remainder of heat is either conducted through the tool, or conducted/convected into the workpiece. Assuming that no coolant application is present in the machining operation, then any heat loss to the ambient air becomes insignificant. An equation has been developed that governs the temperature distribution – via its isothermal gradients – in machining (Fig. 172), this being an ‘energy-based equation’ as follows: ρC(∂T�∂t + V � ∇T) − k∇ T − q˙ = 

(Source: Tay, et al., 1993).

Therefore, in steady-state machining operations, the transient term will disappear and at the region of the tool (i.e. insert) only the conduction term remains. ∴ Rate of heat generation (˙q) = σ ε˙ Where: ‘σ’ was obtained from an emprical function of: ‘ε’ , ‘˙ε’; plus ‘T’. NB This rate of heat generation only exists within the primary and secondary deformation zones. By way of example of how the temperature generation/distribution occurs in orthogonal cutting, in the more-easily understood ‘Boothroyd machining model’ (i.e. being in a slightly modified form – by the author), this workpiece material is in a state of ‘continuous motion’ during cutting (Fig. 172). If specific points are selected to show how temperatures occur as they pass along/through these deformation zones then, the points: ‘X, Y and Z’ can be considered for special observation. So, as the workpiece material enters the cutting region at point ‘X’ , it begins to move toward the cutting insert. It approaches and passes through the primary deformation zone where it is heated-up until it leaves this zone, it is then swept-away by the formed chip. Equally point ‘Y’ , passes through both the primary and secondary deformation zones (i.e see Fig. 51 for these deformation zones) and continues to heat-up until it leaves the secondary deformation zone. In both of the above cases, these points (i.e.

namely: X and Y) are cooled as heat is conducted into the chip’s body (as it exit’s the cut), where it eventually achieves a uniform temperature right the way through. Prior to this occurring, the maximum temperature occurs along the cutting insert’s rake face, some distance from the actual cutting edge (i.e see Fig. 172). Conversely point ‘Z’ , which remains attached to the workpiece, is heated by conduction from the primary deformation zone and some heat is also conducted from the secondary deformation zone into the body of the cutting insert, while the tertiary deformation zone will also impart some heat into the machined surface of the workpiece. Many thermal and thermographical techniques have been developed over the years to obtain accurate isothermal temperatures within the: cutting zones: tool/insert interface plus rake face vicinity; together with the machined surface region of the workpiece. Moreover, ‘indirect methods’ have been utilised to obtain similar thermal historical data from within these dynamic and harsh environments, but only one of these techniques will be mentioned in the next section.

7.6.1 Finite Element Method (FEM) The popular approach today, to obtaining ‘simulated’ thermal data is by employing the ‘Finite element method’ (FEM), to calculate temperature distributions in the vicinity of the cutting regions (Fig. 173). Typical of this approach and worth mentioning in some detail, was that conducted and described by Tay (1993), where he experimentally-obtained information regarding: velocity, strain and strain-rate distributions, by utilising a printed-grid and quick-stop technique. The rate of heat generation within the primary deformation zone was determined from the equation: (˙q) = σ ε˙. From the deformed grid pattern (Fig. 173a), the actual dimensions of the triangular deformation zone, as well as the velocity distribution along the tool/chip interface can be established and analysed. By this FEM technique, it is possible to determine the shear-strain rate within the secondary deformation zone at the tool/chip interface: (˙γint) has been found to be approximately constant and equal to: Vc/δt2. The shear strain-rate within the


329

. Figure 172. Typical temperature distributions (isotherms) during machining, illustrated across the: chip, insert and workpiece; at relatively low cutting speed

330

Chapter 7

secondary deformation zone tends to be linear in nature from: (˙γint) – at the interface, → zero – at the boundary of the triangular secondary zone. The frictional stress along the tool/chip interface can be assumed to be constant along the first half of the contact region, then linearly decreasing to zero at its end. The frictional heat source distribution at this interface, can be obtained from stress and velocity distributions at this location. In Fig. 173a, the basic ‘FEM mesh’ is shown, with typical temperature distributions obtained from this being illustrated in Fig. 173b. The accuracy of this particular example for the ‘Tay-model’ for the total sum of all heat sources was within 2.6% of actual measured power consumption (FcU). Moreover, the values of ‘β’ calculated from the temperature distributions closely-agreed to those obtained some years earlier by Boothroyd (1963). The FEM approach to machining data capture and analysis covers these and other related parameters and clearly indicates the power of simulation – more will be mentioned on this subject later in the chapter.

7.7 Tool Wear and Life Introduction The working environment for most machining processes is extremely harsh, with pressures exerted onto a minute area of tool tip being of the order of >1600 MPa, with localised temperatures reaching over 750°C creating a sterile surface at the tool/chip interface, making this an ideal state for a pressure-welding condition. In attempting to minimise this affinity between the work-hardened chip – often this plastic deformation making the chip >5 times harder than that of the parent workpiece material, means that there are several ways of relieving this tool/chip affinity. The obvious one is to use a cutting tool material that is inert to the workpiece such as a either a: ceramic, or mixed-ceramic cutting insert composition, or something similar, but this may not prove to be satisfactory, particularly if interrupted cutting conditions are anticipated. In this situation above, perhaps by utilising a multi-coated cemented carbide insert this may reduce this ‘adherence-tendency’. Lastly, the correct grade of ‘flood-coolant’ may: lower the interface temperature,

reduce friction here, while somewhat improving the machined surface texture. When only partial success is achieved by employing the above tooling strategies, the last resort may be to adjust the cutting data to enhance and provide a ‘less-abusive machining regime’ , while simultaneously improving the ‘steady-state’ wear conditions. So far, no mention has been made here concerning frictional effects in the cutting process. Friction is very complex subject which relates not only to: chip flow-stress and ‘stiction’ 60 problems at the chip/tool interface, but concerns the tribological conditions along this interface. Cutting tool rake and flank faces are never perfectly smooth, as even when faces and edges have been either been ground, or super-finished the abrasive nature of the super-finishing process produces an abraded surface that approaches the grit size of the abrasive medium. Therefore, to the naked eye the insert’s surface looks smooth, but at the ‘micronlevel’ of surface magnification (i.e. 1 × 10–6 m), the cutting insert’s surface has localised ‘high-spots’ , or asperities present. These asperities significantly reduce the contact area produced between the forming chip and its contact at the interface on the tool’s rake face. Not only can these asperities considerably decrease the ‘real area of contact’ and as a result increase the coefficient of friction here, but the asperities may be either ‘plastic’ , or ‘elastic’ in nature61. In Table 11 (i.e. experimental data extracted from: Childs, et al., 2000, concerning surface texture assessment of cutting insert faces), comparison is made between a small sample of

60 ‘Stiction’ , is sometimes confused with its ‘close alternative’ this being: ‘stick-slip’. These terms are worth stating, to explain their respective differences and have been defined in the following manner: ‘Stiction’ is: ‘The phenomenon at an interface where the frictional stress is equal to the shear yield stress of the softer material.’ ‘Stick-slip’ is: ‘A jerky motion between sliding members due to the formation and destruction of junctions.’ (Kalpakjian, 1984) 61 ‘Plastic asperities – on a plastic chip’ , these are ‘high-spots’ that will sink into the chip and how they achieve this action, does not depend on local conditions at interface contact, but on the bulk plastic flow field. Specifically, the lower the hydrostatic stress in the bulk flow field, the less effort is required for these asperities to sink.‘ Asperities – on an elastic foundation’ , this situation is extremely complex phenomena and put simply, in conditions of low contact stresses, the chip beneath these asperities is elastic. (Childs, et al., 2000)


331

. Table 11. Cutting insert surface texture and contact stress severity data. ← 10klocal/E* [°] → Tool finish:

Surface texture data:

Al/HSS:

Cu/HSS:

Brass/WC

Steel/WC

Ra [µm]

∆q [°]

CVD-coated

0.2-0.5

3-7

1.2

1.9

2.8

1.8

Ground

0.1-0.25

2-4

1.2

1.9

2.8

1.8

Super-finished

0.03

0.4

1.2

1.9

2.8

1.8

* When s/k is 1 MHz in range, this being a usual aid for any form of inprocess monitoring operations. By using a combination of AE and force monitoring, this has been shown to be a means of condition monitoring of the cutting tool’s state – more will be said on this topic later. In metal cutting operations AE occurs due to plasto-mechanical processes of crack formation and chip removal, in combination with surface friction. Any form of tool wear alters the contact surfaces between the tool and workpiece, influencing and increasing the AE signal intensity. Hence, advanced warning of potential tool breakage sometimes results in the appearance of micro-fissures in the tool, which cause an escalation of the AE signals – allowing a basic form of tool and process monitoring to be achieved. AE generation in metallic machining operations, can extend over frequencies of several MHz, although the signal intensity is normally very low and diminishes with increasing distance from its source. Any form of machine vibrations and interferences from the local environment introduce signals from a low frequency range, meaning that any form of significant analysis is normally only possible above 50 kHz. Machine tool interference sources are usually the result of either electrical, or hydraulic main and feed drives, as well as from bearing noise, spindles and gears. These unwanted interferences can be suppressed by utilising suitable high-pass filters, or alternatively a well-designed AE sensor(s), with inherent high-pass frequency characteristics. 77 ‘Root mean square’ (rms), is a measure of the effective mean current of an alternating current. Its actual rms value is derived from the power dissipation by an ac current.

347

nals respectively, on the ‘over-turning’ of transversal holes present in the external turning of the workpiece. Hence, the interrupted cut can clearly be seen periodically in the resultant force traces. In Fig. 179biii, the AE rms signal shows this interference, albeit not very well pronounced, unlike that of the force trace produced in Fig. 179biv, where a definite noise spike can be seen. This combination of two complementary sensing elements and their sensor signals, allows the reliable detection of a process fault, such as tool breakage detection. Until approximately the mid-1990’s, commercial versions of cutting force monitoring equipment for the measurement of a rotating cutting tool, or an edge was not readily available for: drilling, reaming, tapping and milling applications. A major advantage of these rotating cutting force dynamometers, is that they can be used for multi-axes contour milling applications, or simply for an investigation of a discrete tool’s cutting edge geometry and its anticipated machining performance. An early version of such a rotating cutting force dynamometer, is depicted in Fig. 180a. In Fig. 180b, graphs have been produced showing the cutting force and torque results respectively, produced by the rotating cutting force dynamometer. Interest frequently centres on the forces and moments acting on the rotating tool. A rotating cutting force dynamometer (Fig. 180a), allows measurement of three orthogonal forces: ‘FX’ , ‘FY’ , ‘FZ’ , together with the moment ‘MZ’. The data measured by the rotating dynamometer occurs via miniature charge amplifiers, which are then transferred by telemetry to an appropriately positioned stationary antenna. The telemetry involves a bi-directional transmission, with measured data being transmitted to the ‘stationary side’ of the monitoring system and any control commands for the integral charge amplifiers transmitted to the appropriate section of the rotating dynamometer. The power supply to the electronics in the rotor, occurs by the same antenna, but having a different carrier frequency to that of the data transfer. Typical resultant signals produced by the rotating dynamometer are shown in Fig. 180b and have been ‘zoomed’ for the investigation of a single drill’s cutting edge. Cutting force dynamometers of various configurations, are invaluable tools for any form of in-depth machinability study, as they indicate the precise conditions at the cutting tool’s edge(s), in a truly dynamic situation. All dynamometers that are purchased from the manufacturer must come with an appropriate cali-

348

Chapter 7

. Figure 179. A Rotating Cutting-force Dynamometer (RCD), utilising piezoelectric sensor systems. [Courtesy of Kistler Instrumente AG]


. Figure 180. A Rotating Cutting-force Dynamometer (RCD), utilising piezoelectric sensor systems. [Courtesy of Kistler Instrumente AG]

349

350

Chapter 7

bration certificate, to ensure that the results obtained are both valid and sound. A cautionary note: if the dynamometer has been inadvertently dropped, or it has possibly collided with an obstruction when in use on the machine tool, it should be sent back to the manufacturer for servicing and recalibration, otherwise, spurious cutting force data may be the result.

7.9 Machining Modelling and Simulation Introduction Previously, it was been widely accepted that most cutting tool modelling technqiues are somewhat incomplete, in both their analysis of the process and their accompanying derived mathematics. Early, but worthy attempts at analysing the chip formation mechanics of the orthogonal cutting process were undertaken initially by Ernst and Merchant (1941) – shown schematically depicted in Fig. 181, followed by further work concerning the analytical graphical interpretation of the orthogonal cutting action which was presented by Merchant and Zlatin (1945) and later work by Lee and Shaffer (1951) – not shown. This earlier work was then followed by Zorev’s (1963) interpretation of an ‘idealised cutting model’ (Fig. 182). In all of these above modelling cases and others not mentioned – for brevity’s sake!, the very complex nature of the cutting process, is a vast subject ‘straddling’ many engineering and physical disciplines. Such modelling involves aspects of: the tool’s geometry, chip/tool contact lengths and pressures, chip formation, cutting forces, frictional and thermal factors and so on, making it virtually impossible to obtain close agreement between with any truly meaningful results between each proposed model. This lack of correlation of these modelling processes, is to be expected, as in reality a shear zone, rather than a shear plane exists, but for mathematical treatment, a shear plane allows some degree of geometrical association. Due to the complex nature of the inter-related variables that occur in any dynamic cutting situation for just simply the orthogonal cutting process, let alone for oblique machining modelling, this has meant that the ‘optimum modelling solution’ has as of now, not yet been fully addressed. Many ‘learned tomes’ have been written in the past, concerning the ‘mechanics of machining’ and it is not

the intention to fully discuss them here, in this book which is principally concerned with ‘current practice’ concerning machining applications. However, a brief resumé of just one of these ‘orthogonal models’ shown in Fig. 181 will be mentioned below, together with a concise review of friction in metal cutting operations (Fig. 182) will be presented, to attempt to show why the subject of ‘theoretically modelling’ the cutting process is so complicated.

Ernst and Merchant’sComposite Cutting Force Circle If a continuous chip formation is produced when machining ductile materials in an orthogonal cutting process, such as that found when cylindrically turning a component’s periphery with an undeformed chip thickness (‘t1’), this will cause a chip compression (‘t2’) – Fig. 181(top). The cutting forces can be obtained by employing a cutting force dynamometer as discussed in the previous section, to typically measure the forces ‘FC’ and ‘FT’ and so on. By utilising such cutting tool dynamometry, Ernst and Merchant (1941) were able to classify the forces acting in the vicinity of metal cutting which gave rise to both local plastic deformation and frictional effects. In Ernst and Merchant’s theory which is often termed the so-called ‘shear-angle solution’ , it is assumed that the cutting edge is always perfectly sharp and that a continuous-type chip without BUE occurs, this former assumption in practice does not actually occur. Moreover, another assumption in their analysis it was that the chip would behave like a ‘rigid body’ , which is held in equilibrium by the action of the applied forces transmitted across the tool/ chip interface and transversely over the shear plane. Boothroyd (1975), offered a reasonably ‘elegant solution’ to Ernst and Merchant’s numerical and geometrical analysis and, this has been somewhat modified and further simplified below. In order to abridge the ‘shearangle solution model’ shown in Fig. 181, the resultant force ‘R’ is depicted acting at the tool’s cutting edge, being resolved into components ‘N’ and ‘F’ in directions along and normal to the tool’s face respectively, as well as into components ‘FN’ and ‘FS’ – once again, along and normal to the shear plane correspondingly. Further, the cutting force ‘FC’ and the thrust ‘FT’ components of the resultant force, are also shown. Here, it can also be assumed that the entire resultant force is transmitted across the tool/chip interface and that on both the tool’s flank and edge no force occurs, meaning that a zero ‘ploughing-force’ is present – see Fig. 184 which illustrated this ‘nose-rounding effect’.


The foundation purported by Ernst and Merchant’s theory, was the proposition that the shear angle ‘φ’ would acquire such a value, thereby reducing the actual work done to a minimum. In view of the fact that

. • • • • • • • • • • • • •

351

that for preselected cutting conditions, the work done during cutting was comparative to that of ‘FC’ in terms of ‘φ’ , hence allowing one to obtain the value of ‘φ’ when ‘FC’ is at a minimum. Thus, from Fig. 181:

Figure 181. ‘Merchant’s’ composite metal cutting circle for an orthogonal cutting model, where: FT = thrust force component, FC = cutting force component, FN = normal force component on the shear plane, FS = shear force component on shear plane, R = resultant tool force component, N = normal force component on tool face, Φ = shear angle, α = working normal rake angle, τ = mean friction angle on tool face, t1 = undeformed chip thickness, t2 = deformed/compressed chip thickness, A0 = cross-sectional area of uncut chip, Ac = cross-sectional area of deformed/compressed chip.

NB: Force arrow vector directions have been reversed and some terms have been modified from the original work. [Source: Ernst & Merchant, 1941]

352

Chapter 7

FS = R cos(φ + τ – α)

(i)

and FS = τSAS = τSAC/sin φ

(ii)

Where: τS = Workpiece material’s shear strength – on the shear plane, AS = Shear plane’s area, AC = Uncut chip’s cross-sectional area, τ = Mean friction angle, between too and chip [i.e. arctan (FT/FN)], α = Normal rake angle (i.e. working). From equations (i) and (ii): R=

τ s Ac  sin ϕ cos(ϕ + τ − α)

(iii)

By geometry:

FC = R cos(τ – α) τs Ac cos(τ − α) sin ϕ cos(ϕ + τ − α)

(iv)

π 

(v).

(vi).

In comparative analysis undertaken by Merchant (1945), he found close correlation with experimental results was obtained when machining synthetic plastics, but a somewhat poor theoretical correlation occurred when steel had been machined with cemented carbide tooling. It needs to be mentioned that when Merchant differentiated with respect to ‘φ’ , it was assumed that ‘AC’ , ‘α’ and ‘τS’ would be independent of ‘φ’ , but on further consideration, Merchant decided to offer in a ‘modified theory’ the relationship of : τS = τSo + kσS

(viii)

and: σS AC FNs = σS AS = sin ϕ

(ix).

From equations (viii) and (ix):

σS =

sin ϕ R sin(ϕ + τ − α) AC

(x).

τS = σS cot(φ + τ – α)

(xi)

and, from equations (vii) and (xi):

Now, equation (v) can be differentiated with respect to ‘φ’ , then equated to zero to obtain a value of ‘φ’ when ‘FC’ is at a minimum. Hence, the requisite value is specified by: ϕ + τ − α =

FNs = F sin(φ + τ – α)

Combining equations (iii) and (x):

Hence, from equations (iii) and (iv): Fc =

equal to ‘τSo’. This new assumption by Merchant was confirmed by previous work undertaken in the literature published by Bridgman (1935, ’37 and ’43), where the shear strength when experimental machining of polycrystalline metals was shown to be dependent on the normal stress on the ‘plane of shear’. Now from Fig. 181, we can obtaing the following relationship:

(vii).

Here, Merchant brought in a modification to the shear strength of the workpiece material ‘τS’ which now increased linearly with an increase in normal stress ‘σS’ on the shear plane, where zero normal stress ‘τS’ is

τS =

τSo  − k tan(ϕ + τ − α)

FC =

τSo AC cos(τ − α) (xiii). sin ϕ cos(ϕ + τ − α)[ − k tan(ϕ + τ − α)]

(xii).

This equation (i.e. xii), explains why the value of ‘τS’ may be influenced by modifications in the shear angle (‘φ’), which is now inserted into equation (v), to obtain a new equation for ‘FC’ in terms of ‘φ’ , therefore, the resulting expression becomes:

Now, it can be assumed that both ‘k’ and ‘τSo’ are constants for the specific workpiece material, further, that ‘AC’ and ‘α’ are also constants for the machining operation. Hence, equation (xiii) can now be differentiated to obtain a new value of ‘φ’ , with the resulting expression: 2φ + τ – α = C Where: C = Can be obtained from ‘arccotk’ , which is a constant for the workpiece material. NB In further more recent experimental work, it has been shown that ‘τS’ will remain constant for a specified workpiece material, across a diverse range of cut-


353

ting conditions, as such, the value of ‘k’ can be equated to zero. When the shear-angle values are compared for the plotted linear relationships of the earlier theoretical and experimental work undertaken by Ernst and Merchant, to that of the later comparative work by Lee and Shaffer, then these shear-angle relationships diverge somewhat. However, one area that these researchers both agreed upon, was the fact that friction on the tool’s face was the most important factor during metal cutting. In the following discussion, the influence that frictional behaviour has at the tool/chip interface will be briefly mentioned.

Frictional Effects During Machining For simplicity’s sake, friction between dry sliding surfaces will be concisely reviewed. In 1699 Amontons ‘laws of friction’78 were formulated, then verified by Coulomb79 in 1785, with Bowden and Tabor (1954) contributing greatly to an explanation of these empirical laws. If two apparent flat surfaces are placed together, the larger asperities (i.e. peaks) on each mating face will only establish contact. With just normal loading, the top of these asperities will yeild, creating a real area of contact until such a time, that they are capable of supporting an applied load. In the main, for the vast majority of engineering applications this real contact area (‘Ar’), is normally only a very minute portion of the apparent real of contact (‘Aa’) – even after the contacting asperities of the softer material has plastically deformed (i.e. yielded). Thus: Ar = Fn /σy

(i)

78 ‘Law(s) of friction’ , Amontons in 1699 stated in the main, that: ‘Friction is independent of the apparent area of contact and proportional to the normal load between two [mating] surfaces.’ 79 Coulomb (1785), confirmed these ‘frictional laws’ with the observation that: ‘The coefficient of friction* is substantially independent of the speed of sliding.’ *Coefficient of friction (µ) = F/N Thus, the friction force is proportional to the perpendicular force between contacting surfaces and is independent of the surface area, or its ‘rubbing speed’.

Where: Fn = Normal force, σy = Yield pressure of the softer material. Thus, for most metallic materials, this close asperity contact between mating surfaces creates localised ‘cold welding’. Prior to any sliding with respect to these two contacting surfaces, a force is necessary to continually shear potentially contacting and reforming asperity tips of these ‘welded junctions’. Hence, the total frictional force ‘Ff’ is given by: Ff = τf Ar

(ii)

Where: τf = Shear strength of softer contacting material, Ar = Real contact area. From equations (i) and (ii), the actual coefficient of friction (‘µ’) between these contacting surfaces will be: µ = Ff/Fn = τf/σy

(iii).

In majority of machining operations typified by the continuous turning of metallic workpieces, the coefficient of friction (‘µ’) at the chip/tool interface can vary quite considerably. Its frictional variation, being influenced by any changes in: cutting speed, feedrate, rake angle contact regions – this latter factor is particularly relevant for multi-functional tooling, where the cutting insert’s top face is not planar (i.e flat). In modern machining practice, the normal pressures exerted at the tool/chip interface are exceedingly high, typically when machining commercial grades of medium carbon plain steels these pressures are >3.5 GPa. This high normal pressure at the ‘interface’ causes the real contact area to approach that of the entire area, where: Ar/Aa = unity! This means that under these circumstances the ordinary laws of friction no longer apply, and the frictional force ‘Ff’ is still represented by equation (ii), but is now independent of the normal force ‘Fn’. This acute change in the frictional behaviour during machining, means that the shearing action is not confined only to the interface asperities, but includes workpiece material in the local substrate. In work published by Zorev in 1963, his ‘model’ considered the frictional behaviour in continuous metal cutting operations where no BUE was present (Fig. 182). Under these machining conditions, Zorev

354

Chapter 7

noted that the normal stresses at the tool/chip interface were sufficiently high enough to cause ‘Ar/Aa’ to approach unity, over the zone denoted given by length ‘lst’ (Fig. 182) – this being termed the ‘sticking region’ 80 – also see magnified quick-stop photomicrograph in Fig. 184(top) showing the affect on the resultant chip. The ‘sliding region’ , extends from from the end of the ‘sticking region’ to a point where the chip loses contact with the rake face (i.e. ‘lf’-‘lst’), here, the ‘Ar/Aa’ ratio is less that unity – meaning that the coefficient of friction is constant. In 1964, Wallace and Boothroyd produced evidence that the ‘sticking manner’ observed on the underside of the chip had abruptly stopped. They observed that in an adjacent vicinity to that of the tool’s cutting edge, grinding marks on the rake face were imprinted onto the chip’s underside. This phenomenon indicated that no relative motion between the chip and the tool had occurred, suggesting that here, the ‘real’ and ‘apparent’ areas of contact were ‘equalised’ in this region. Therefore, under ‘sticking friction’ conditions, the ‘mean angle of friction’ on the rake face, depends on the: 1. Form of the normal stress distribution, 2. Tool/chip contact length (i.e. ‘lf’), 3. Mean shear strength of the chip material – in the ‘sticking region’ , 4. Coefficient of friction – in the ‘sliding region’. NB It seems apparent that a simple numerical value for the mean angle of friction, is inadequate when attempting to completely describe the frictional circumstances on the rake face. In Fig. 182, is schematically depicted the ‘frictional model’ purported by Zorev (1963). Here, Zorev suggested that the normal stress distribution (‘σf’) on the tool’s rake face, could be represented by the following simple expression: σf = qxy

(iv)

Where: ‘x’ = Distance along the rake face – from the position where the chip losses contact with the tool’s face, ‘q’ and ‘y’ = are constants. The maximum normal stress ‘σfmax’ occurs when ‘x’ equals ‘lf’ , so: σfmax = qlfy or, transposing with respect to ‘q’: q = σfmaxlf–y

(v).

Substituting for ‘q’ in equation (iv), we get: σf = σfmax (x/lf)y

(vi)

Thus, in the ‘sliding region’ , ranging from: x=0 to x = lf – lst the coefficient of friction ‘µ’ is constant and, the distribution of shear stress ‘τf’ along this region, is represented by: τf = σf µ = µ σfmax (x/lf)y

(vii).

So, in the ‘sticking region’ , the shear stress becomes a maximum (‘τst’), therefore from: x= lf – lst to x=lf: τf = τst

(viii).

By integrating equation (vi) to obtain the normal force ‘Fn’ acting on the rake face (i.e. from Fig. 181), gives: lf

Fn = aw � σ f max (x�l f ) y dx 

= σfmaxaw lf/1 + y

(ix)

Where: aw = Chip width (i.e. width of cut). The friction force ‘Ff’ on the tool’s rake face, is obtained by:

80 ‘Sticking region’ at the tool/chip interface is often termed the ‘stagnation zone’ , but in this case it appears on the formed/ sheared chip – depicted in Fig. 184 (top). Here, is shown a magnified and etched view of a quick-stop for an insert cutting carbon steel at 150 m min–1. A ‘stagnation zone’ is present, that follows the insert’s profile, with ‘softened’ workpiece material protecting the tool, by a sticking/sliding action. A ‘flow zone’ occurs after ‘shear plane’ , visibly dividing undeformed/ deformed material.

l f −l s t

F f = aw [τs t ls t + � 

µ σ f max (x�l f ) y dx]

= τst aw lst + µ σfmax aw (lf – lst)1+y/lfy(1+y)

(x).

At position: ‘x = lf-lst’ , the normal stress ‘σf’ is given by: ‘τst /µ’. Moreover, from equation (vi), it is given by:


355

σfmax (lf-lst/lf)y

stress will be reduced. Moreover, from equation (xvi), an increase in ‘α’ could be expected to increase the Therefore: mean friction angle. This observation, has been confirmed by experimental work undertaken by Pugh (xi). (1958), where an increase in ‘α’ was shown to result in τst = µ σfmax (lf – lst/lf)y a a complementary increase with repect to the mean By substituting equation (xi) in equation (x), it simpli- friction angle – across an extensive range of workpiece fies the expression for ‘Ff’ , as follows: materials. Since the passage of time when both of these ‘theo(xii). retical models’ for the mechanics of machining and the Ff = τst aw lst + τst aw (lf – lst)/1+y frictional work produced by Ernst and Merchant and Hence, the mean coefficient of friction on the rake face, that of Zorev, respectively, were produced. Considercan now be found from both equations (ix) and (xii): able effort and progress has been made into advances in our present understanding of: ‘slip-line field modelling’ , Mean friction angle = Ff/Fn = τst/σfmax (1+ylst/lf ) (xiii). ‘thermal and frictional modelling of the cutting process’ , ‘chip-flow analysis’ , ‘examination of the primary and From equation (xi), the mean normal stress on therake secondary shear zones’ , non-orthogonal (three-dimensional) machining processes, advances in tool geometries face (σfav) is given by: with their accompanying force and shear relationships σfav = Fn/awlf = σfmax/1+y and the advent of applications to machining operations employing finite element analysis (FEA). If all of these subjects and other not listed, were only concisely Thus: (xiv). mentioned, then the book would be of epic proporσfmax = (1+y)σfav tions! defeating the object of reviewing cutting appliSubstituting for ‘σfmax’ in equation (xiii), produces: cations and trends in current practice. However, the role of FEA, in both machining research and for the Mean friction angle = arctan{τst/σfav [1+y(lst/lf )]} (xv). ‘dynamic and geometric modelling’ of applied residual stresses occurring in both the tool and workpiece in However, from Zorev’s experimental work, he found the anticipated cutting processes is worthy of discusthat the term: sion. The topic of ‘computer simulation’ utilising an FEA approach, has become of increasing importance τst [1+y(lst/lf )]/1+y of late and several companies have produced products in this field. However, to obtain some degree of conwill remain relatively constant for a specified work- sistency in the dialogue only one particular company’s piece material, across a diverse range of unlubricated FEA product will now be described. cutting conditions and, as a result, the equation for the mean friction angle becomes: Computer Simulation of Machining Mean friction angle = arctan K/σfav

(xvi)

Where: ‘K’ = a constant. What Zorev’s equation (xvi) indicates, is that the mean friction angle is somewhat dependent on the mean normal stress on the rake face, allowing the result to explain the effect of modifications in the working normal rake to that of the mean friction angle. Thus, as the value of ‘α’ (i.e. working normal rake angle) increases, the resultant component tool force being normal to the rake face will decrease, so that the mean normal

Processes – an Introduction

By the application of computers to simulate and analyse machining processes, this provides tool manufacturers and users alike, assistance in improving machining efficiency and will also predict the likely cutting response to real-time machinability environments. Most of the early two-dimensional simulation models for the mechanics of machining, focussed attention on shear-zone modelling. These previous ‘models’ ignored vital factors such as the frictional conditions along and between the rake’s tool/chip interface, furthermore, any potential work-hardening and temperature effects in this vicinity were also tended to be disregarded.

356

Chapter 7

Over the last two decades, the application of finite element analysis (FEA) to that of three-dimensional machining applications, has been successfully applied to the cutting process. Major advantages of using FEA, are its ability to accurately compute: complex material property definitions; tool/chip interactions; non-linear geometric boundary conditions – typified by the chip’s surface; prediction of local variables – such as stress and temperature distributions in the cutting locality.

Several approaches for the numerical modelling of machining operations and in particular, for that of metal cutting have been employed, such as ‘Lagrangian and Eulerian’ techniques. In the former case which has been utilised for over two decades, ‘Lagrangian methods’ utilise the tracking of discrete material points. Here, the technique is uses a predetermined line of separation at the tool’s point, which propagates a fictitious crack ahead of the tool’s tip. Previously, this

. Figure 182. An idealised orthogonal cutting model of chip-tool friction, where: • σf = normal stress, • τf = shear stress, • τst = shear strength of chip material in the sticking region, • If = chip-tool contact, • Ist = length of sticking region. [Source: Zorev, 1963]


routine precluded the resolution of the cutting edge radius and an accurate resolution of the secondary shear zone – due to severe mesh distortion. In an attempt to alleviate any form of element mesh distortion ‘adaptive remeshing techniques’ have been employed to resolve the tool’s cutting edge radius. Whereas the ‘Eulerian approach’ tracks volumes rather than material particles, having the advantage of not needing to rezone any distorted meshes. Moreover the ‘Eulerian technique’ , requires ‘steady-state free-surface tracking algorithms’ and relied upon a particular bur unreasonable assumption that a uniform chip thickness occurred, further this method precluded the modelling of either a segmented chip formation, or that of the milling process. The former technique of a ‘Lagrangian FEA machining model’ will be reviewed (Fig. 183), as it has the integrating ability to achieve ‘adaptive remeshing’ with explicit dynamics and tightly coupled transient thermal analysis, allowing it to ‘model’ the complex interactions between the cutting tool’s geometry and that of the workpiece.

Lagrangian FEA Simulation Machining Modelling In Fig. 183, just a few images of the Lagrangian FEA machining model are depicted for several applications of machining operations. This simulation modelling technique contains a tightly coupled thermomechanical material response capability, this being a vital factor for any elevated temperatures that occur at the tool/chip interface, furthermore, having a fully adaptive mesh generation ability. Hence, the material modelling facility forms an intergral part when attempting to predict the workpiece’s material behaviour, under high strain and increased stress conditions. This advanced modelling capability ensures the accurate capture of any strain-hardening, or thermal softening effects coupled to rate sensitivity properties, for a given set of material conditions. Many of today’s ferrous and non-ferrous materials, together with ‘exotic materials’ such as nickel and titanium alloys can also be successfully simulated, across a diverse range of single- and multi-point cutting tools (e.g. turning, milling, drilling, broaching, sawing, etc.).

Machining Simulation – Validation Not only can simulation techniques of the type shown in Fig. 183 be utilised for workpiece material machin-

357

ing modelling: work-hardening; thermal-softening effects; machining-induced residual stresses; induced temperature effects and heat-flow analyses with the ‘adaptive’ ‘Lagrangian FEA machining model’. The ‘model’ can also reasonably accurately predict: both two- and three-dimensional cutting force magnitudes. Invariably, the individual cutting force components can be closely validated to actual machining working practice, the same can also be said for a comparison of a ‘dynamically-modelled chip’ to that of an actual chip’s morphology – including both its chip-curling tendency and any chip segmentation occurring. These validated simulation capabilities enable the cutting process to be improved, by using the: • Force and temperature information – to reduce overall production cycle times, • Temperature and thermal effects – can be utilised to improve tool life and part quality, • Tool wear analysis – predicts the effects of tool flank wear and how this wear land influences subsequent: temperatures, pressures and forces, • Chatter and vibration prediction – indicating the onset and magnitude of these unwanted effects, • Residual stress information – helps alleviate potential machined component fatigue and aids in part deformation analysis. NB The influence that tool coatings have on the dynamic machining efficiency can also be reviewed, plus the capability to customise the tooling with chip-breakers to improve chip-curl, or chip evacuation abilities. Computer machining simulation of the type illustrated in Fig. 183, can be integrated into an overall CAD/CAM package, enabling a range of significant advantages to accrue, without having to operate a costly and time-consuming task of undertaking an extensive machinability trial. This off-line machining simulation facility, allows: realistic cycle-time calculations including cut-and non-cut timings; visualisation of tool paths to high-light and then avoid any localised ‘power-spikes’ occurring during a machining operation and many other useful production features. The use of a dynamic FEA machining simulation package similar to the one mentioned and illustrated in Fig. 183, adds a scientific element to the understanding of the overall machinability of specific workpiece materials and associated tooling. Such machining simulation offers not only a visual interpretation to

358

Chapter 7

. Figure 183. Simulated machining processes. [Courtesy of Third Wave Advant Edge]


359

. Figure 184. The insert’s cutting edge: illustrating the ‘rounding effect’ (exaggerated) or, a manufacturer’s ‘edge preparation’ and the material flow conditions that arise as a result

the machining operation, but it also provides the user with validated and relevant data analysis. These positive benefits enable tool designers and users alike, to design and develop advanced cutting tools and to undertake efficient and optimised machining operations. Beyond the positive advantages of tool optimisation,

simulation can significantly reduce tooling development costs and lead times to bring a newly-developed product to market. The role of machining simulation is likely to rapidly grow, as more tooling and production engineers become aquainted with these software packages.

360

Chapter 7

7.10 Surface Integrity of Machined Components – Introduction

fects when discussing the machined surface condition in the following section.

Previously in Section 7.5 concerning machined surface texture, the discussion was principally concerned with the resultant surface topography where the topographical information was valid, but disguised the fact that potential sub-surface material layers might have compromised and altered the machined component. The concept of the overall functional performance of a surface and its accompanying sub-surface condition was recognised by Field and Kahles (1971), where they used the term ‘Surface Integrity’ to describe its potential state. The overall concept of surface integrity ant its various generating mechanisms in conjunction with the production process is known as the ‘unit event’ 81. This unit event has now been reclassified into five discrete generating mechanisms: chemical, mechanical, mechano-thermal, thermo-mechanical and thermal – the order they are listed reflects their respective power density per unit area. For example, increases in the power density from the chemical end of the series, results in an augmented level of thermal energy entering the surface leading to greater thermal damage and poorer part surface integrity. The chemical mechanism is dominant across all classes of production process to some degree and that surfaces react with their immediate environment, via absorbates, oxidation, etc., as illustrated in Fig. 185 – more will be said on these ef-

A machined surface is the product of either ‘abusive’ , or ‘gentle’ machining regimes, these being the direct result of the cutting process and its chosen machining data. Thus, machining being a complex relationship of many interrealated factors, affects the outcome of the production process – see Fig. 144. Here, a simplistic schematic diagram attempts to show the complexity of a machining operation, with the surface integrity grouping indicating for a turning operation the following features: • Surface condition – surface texture and its associated roundness, • Micro-structural changes – micro-cracks, dislocations and fissures, etc., • Surface displacement – bulk transportation of material and residual stresses, • Surface/sub-surface micro hardness – plastic deformation and localised residual stress layers.

81 ‘Unit event’ , is a complex interrelated series of reactions with the potential for distinct zones to be present within the surface vicinity, including a: – Chemically affected layer (CAL) – resulting from chemical surface changes by the production process, or from postproduction exposure to a local environment, – Mechanically affected layer (MAL) – this may be due to factors such as material bulk transportation: deposits; laps; folds and plastic deformation, – Heat affected layer (HAL) – principally concerned with factors such as: phase transformation; thermal cracking and retempering, – Stress affected layer (SAL) – is in the main, the result of residual stresses being a combination of the above. (Field and Kahles, 1971)

7.10.1 Residual Stresses in Machined Surfaces

Machined surfaces are even more complex than seemingly at first glance, as their performance can be influenced by either external layers (chemical transformations and plastic deformations) and internal layers (metallurgical transformations and residual stresses). By way of example, the anisotropic – periodic – longitudinally turned surface illustrated in Fig. 185, is affected by the cutting insert’s tool tip geometry and the regularity of the cusps (i.e. peaks and valleys) – the surface topography being dominated by the pre-selected feedrate. A series of other micro-technological features can also occur, these often being superimposed onto the machined surface, typically the result of: tool wear, vibrational influences and to a lesser extent, machine tool-induced errors. In the circumferential direction the ‘Lay’ is both periodic and regular, albeit this round generated surface by the turning operation, will probably have some form of harmonic effects present: departures-from-roundness characteristics (i.e. a combination of harmonic influences present). The exposed sterile surface (Fig. 185), is the result of highly localised temperatures and transients, which when turned the machined surface will be instantaneously oxidised and adsorb contami-


361

. Figure 185. The cross-section of an anisotropic (i.e. periodic) surface, illustrating surface contaminants (oxides and adsorbates), together with some sub-surface plastic deformation (the residual stress zone) and an unaffected substrate

362

Chapter 7

nants. The outermost adsorbate layer is often termed the ‘Beilby layer’ 82: ≈1 µm in thickness and consisting of many complex factors. Notably, this ‘layer’ would more than likely have hydrocarbons present and water vapour, that originated in the coolant, or the atmospheric environment, respectively. Underneath this metallic surface for work-hardening materials, there is normally a plastically-strained region that has usually been metallurgically altered. The depth of this strain-hardened layer will vary somewhat, but it is in the region of 10 µm, its actual thickness is dependent upon the amount of plastic deformation induced by the tool’s passage over the surface and is influenced by the metallic substrate’s composition. The plastic deformation and work-hardening depths83, can penetrate to fractions of a millimetre this is particular true, if a ‘wiper-insert‘, or roller burnishing tools is employed to purposely create this localised hardened region to the component’s surface.

Residual Stress Deformations For any residual stresses acting within a body (i.e. component), they will occur without any external forces, or moments. Internal forces form a system that is currecntly in a state of equilibrium and if portions are removed – by machining, the equibrium status is normally disturbed, resulting in potential component deformation. This effect of machining distortion is well-known to practising industrial engineers, when, for example, machining just one side of a thin component, this operation will cause a partial release of local residual stresses causing it to bend and bow. If either a casting, or forging has not been heat-treated for stress relief and its needs asymmetrical machining (i.e. on one side only), it is likely to deform after unclamping restraint from its work-holding device on the machine tool. In an attempt to minimise this distortion created by residual stress release, an experienced machinist will release the clamping forces after roughing cuts so that

82 ‘Beilby layer’ , on the machined surface is ‘practically amorphous’ – this condition being proposed by Sir George Beilby around the beginning of the 20th century. 83 As an approximation, the depth of hardness penetration is approximately 50% to that produced by residual stress penetration, whereas the observational plastic deformation is about 50% greater than this penetration.

the stressed surfaces are equalised, prior to reclamping and taking a finish pass. If this unclamping and then re-clamping activity is not possible, components clamped in-situ on the machine tool are occasionally vibrated at their natural frequency, to minimise these induced residual stresses. Component deformation is roughly proportional to the removed cross-section of workpiece material. Any further finishing is usually concerned with just a light cut to minimise any detrimental effects resulting from residual stresses by a previous production processing operation, or route. The release of internal residual stresses must not be confused with the input of such stresses by machining, as indicated in Fig. 186b. The machining process generates residual stresses by plastic deformation (Fig. 187a), or from localised metallurgical transformations. In Fig. 186a, the residual stress effects influence a range of mechanical and physical properties of the workpiece material, such as: • Deformation – this point has been alluded to above and can create problem with small workpiece crosssections, • Static strength – is affected by the yeild point of the workpiece material, which in turn, is influenced by the presence of residual stresses, • Dynamic strength – of the part in-service can often have its fatigue strength and life affected by the influence of residual stresses present, • Chemical resistance – if certain metals are subjected to induced residual stresses on exposure to atmosphere over a period of time, then stress corrosion may occur, • Magnetism – residual stresses present, can affect a component’s magnetic properties, creating disturbances of the crystalline structure.

Taper-Sectioning and Micro-Hardness Assessment So that an improvement of metallographical inspection of a sectioned machined surface can be made without unduly affecting any form of surface distortion, ‘tapersectioning’ has often been utilised. A tapered-section (Fig. 187b), allows such sub-surface features as: phase transformations; plastic flow zones; localised cracking; bulk transportation and redeposit of material; to be investigated which would otherwise have been missed, if only profilometry (i.e. surface topography assessment) had been undertaken. As its name implies, a taper-section overcomes the limitation of perpendicular sectioning. By taking an


363

. Figure 186. The effects of residual stress and deformations of a workpiece by machining. [After: Brinksmeier et al., 1982]

angular planar slice through the components crosssection, this modified cut angle enhances the substrate magnification, without unduly distorting exposed surface features – giving greater discretion when observing, or testing the surface topography. In Fig. 187b, an 11° sectional cut improves surface discrimination by increasing the vertical section magnification by around five times. The taper-section angle (TSA) will thus be 79°, with the vertical magnification being obtained from the following expression:

TSM = secant (TSA) Where: TSM = taper-section magnification, TSA = taper-section angle. Often, the exposed sub-surface feature of interest that has been plastically deformed, or mechanically altered is in the main quite small, somewhat less than 0.1 mm in width. If a micro-hardness indentor such as either

364

Chapter 7

the Vickers84, or the Knoop85 is utilised (Fig. 187c) to establish hardness readings in the vicinity of this residual stress zone, then more indentations are possible using the Knoop, rather than the Vickers indentor, giving, more discrimination to the ‘foot-printing’ assessment. A note of caution here when originally attempting to take the taper-section, is that it is quite possible to metallurgical alter the sub-surface features, if when taking the section too much heat is induced when cutting it from the parent component. This comment is also a valid statement for the subsequent grinding and polishing of the removed taper-section, prior to metallographical/hardness assessment.

Surface Condition – Being Affected by Cutting Speed Prior to discussing the surface and sub-surface modifications to the machined part – shortly to follow, it is worth taking a closer look at the series of photomicrograph images shown in Fig. 188. Here, a group of identical metallurgical composition ferrous workpieces was machined, but at various cutting speeds. It can be demonstrated that the role played in affecting the machined surface condition, is significantly influenced by the cutting speed, with its accompanying amplification of induced temperature effects as ‘speeds’ are increased. Moreover, it can also be said,

84 ‘Vickers indentor’ , has a square-based dymond pyramid with and indentor included angle of 136°. Its indentation is defined as: ‘The load divided by the surface area of the indentation’. The Vickers hardness [i.e. penetration] number (VPN), may be determined from the following expression: VPN = 2Psin(θ/2)/L2 Where: P = applied load (kg), L = average length of diagonals (mm), θ = angle between opposite faces of diamond (136°). 85 ‘Knoop indentor’ , has complex facets to its diamond indentor, having angle of 130° (Short diagonal) and 172.5° (Long diagonal), respectively. This facet geometric indentor arrangement (i.e. having a diagonal ratio of 7:1), leaves a significantly narrower and longer surface indentation, to that of the Vickers – mentioned in Footnote 84. Thus, the Knoop hardness number (KHN) has been defined by the National Bureau of Standards (USA), as: ‘The applied load divided by the unrecovered projected area of the indentation’. The following expression relates to the Knoop’s surface indentation: KHN = P/Ap = P/L2C Where: P = applied load (kg), Ap = unrecovered projected area of indentation (mm2), L = length of long diagonal (mm), C = constant – supplied by indentor manufacturer.

that a material’s properties are dependent on the strain rate, with the type and magnitude of tool wear changing according to the cutting speeds, so simplistically speaking: • Low cutting speeds – wear is normally characterised by attrition (i.e. mechanical removal of surface layers), • High cutting speeds – here, attrition gives way to diffusion type wear and ‘Fick’s laws’ dominate the cutting regime. NB Such ‘broad classifications’ of tool wear mechanisms occurring, affects the type of: surface produced; chip formation and strain behaviour. In some interesting trials undertaken by Watson and Murphy (1979) – which highlight the disguised nature of the underlying factors in surface integrity investigations. In this practically-based experimental work, they used a cemented carbide insert on an alloy steel (Fig. 188). It was found that the feedrate and DOC have only marginal effects on the sub-surface damage to a machined workpiece, with the cutting speed being the most influential in this situation. This fact has been established in Fig. 188, when a range of similar workpiece specimens was machined with the only variable being the cutting speed, as follows: • Photomicrograph a – the machined specimen was machined at a very low cutting speed (2.6 m min–1) The chip formation was discontinuous and the surface shows an alternating effect of both chip formation and fracture, with some evidence of deposited residual BUE. Here, the surface topography is the result of complex interactions by various effects, such as changes in shear angle in the contact area between the tool and chip, plus ‘straining’ causing increases in the chip thickness. These phenomena produce a variety of conditions, from strain-tocracking and visually introduces an irregular and an alternating surface topography, • Photomicrographs a to d – cutting speeds in the range from 11 to 59 m min–1, generate a continuous chip formation. It is evident from these photomicrographs (b, c and d), that the surface texture was gradually improving as the cutting speed increased, although even at 59 m min–1, there was some indication of debris from re-deposited BUE here (i.e. in ‘d’), • Photomicrograph e – once the ‘optimum’ cutting speed had been reached (112 m min–1 – for this ce-


. Figure 187. The tribological action of machining and its affect on induced residual stresses and the microhardness ‘foot-printing’ technique

365

366

Chapter 7

. Figure 188. Some photomicrographs of component surfaces machined at different cutting speeds – otherwise with identical cutting data – illustrating the surface, but not sub-surface steel’s condition. [Source: Watson & Murphy, 1979]

mented carbide insert grade), the surface texture appears to be in the main, ‘good’ , with only isolated areas of the topography exhibiting marginal workpiece side-flow effects, • Photomicrograph f – when the cutting speed was increased to 212 m min–1, then in these trials, greater cutting insert wear-rate occurred and was attributed to appreciable carbide edge breakdown, although the surface topography indicated that an excellent surface texture was present.

The machined surfaces produced at the lower range of cutting speeds indicated in Figs. 188 a to d, shows evidence of some re-deposited BUE material to greateror-lesser extent: having broken away from original ‘BUE mass’ , then being re-deposited over several adjacent machined feed cusps (i.e. see Fig. 28a, fullyappreciate this effect). To obtain a better and deeper understanding of these machined surface and subsurface effects at the extreme conditions of either very low, or high cutting speeds: Figs. 188 a and f, respec-


tively, the following comments can be made. When longitudinal taper-sections were taken through these specimens’ cross-sections, the ground, polished and etched surfaces reveal their true substrate damage. In the case of Fig. 188a, BUE was presents on the surface, moreover, there was a cutting/fracture sequence indicated with confirmation of work-hardening having ‘layered scales’ of with cracks and crevices beneath them. Conversely, the test specimen machined at high cutting speed (Fig. 188f), there is some verification of a ‘white-layer’ formation – which is a complex metallurgical phenomena found in certain ‘abused’ ferrous workpiece situations – more will be said on this condition shortly. In fact, the ‘good’ machined surface topography disguises the fact that an underlying ‘whitelayer’ condition was present, having a local recorded hardness of 860 HVPN. By way of comparison, if this same alloy steel composition had received a ‘conventional’ hardness heat-treatment process: heated and water-quenched from 1200°C, then the bulk hardness would only be approximately 700 HVPN – see Appendix 12 for Hardness Comparison Tables. From these examples of cutting speed investigative results and the previously mentioned discussion, it is evident that the ‘optimum’ machined surface texture is obtained when the cutting speed is closely aligned to that of the tooling manufacturer’s recommendations, so here in this case it is ≈112 m min–1, with a correspondingly ‘good’ surface topography/integrity. If the cutting speeds had been employed at the ‘higher’ cutting data (i.e. 212 m min–1), then one could have been fooled into accepting this apparently ‘improved’ surface topography. Nevertheless, underlying this machined surface would be an unstable sub-surface condition, which if used in a stressed and critical in-service environment, it might potentially fail, by a reduced fatigue-life – this is why the topic of surface integrity is so important in today’s climate of potential industrial litigation, when component failure occurs!

Surface Cracks and White-Layers If any cracks are present at the free surface which extends into the material’s substrate, they are potential sites for premature component failure – for highly stressed in-service components. It has been reported in the findings of industrial enquiries into the UK railway industry of late, that despite these railroad tracks being precision machined and then occasion-

367

ally inspected by non-destructive (NDT)86 techniques – according to the maintenance schedule, instances have occurred when these rails and particular on high-speed banked corners – have delaminated. This catastrophic rail delamination has caused several passenger trains to lose contact with the rails and crash, resulting in significant loss of life. Hence, the method of machining – ‘abusive’ – can contribute poor surface integrity and to the susceptibility of these machined surfaces to prematurely fail. In the case of milling operations, it has been recognised for a number of years that up-cut milling – alternatively termed ‘conventional milling’ (Fig. 190a), can introduce a surface tensile residual stress into the surface layers of a milled workpiece. If this machined component is then subjected to both an arduous and potentially fatigue-inducing environment, then the cyclical nature of continuous stressing followed by its immediate stress release, can initiate surface crack sites causing them to open-up, which could result in premature part failure. Conversely, an identical machined component that has been ‘down-cut’ – otherwise termed ‘climb-milling’ (Fig. 190b), will induce surface compressive residual stresses. This surface layer with its residual stress compression, has invariably been shown to remain closed and thus, avoiding crack propagation and growth, when machined under identical cutting data and environmental circumstances. Moreover, for many years, it has been recommended that for CNC milling applications ‘climb-milling’ not only generates this favourable machined surface compressive stress effect, but is a more efficient cutting process and as a result, draws less spindle power. In Appendix 13a and b, two useful ‘nomographs, are given to determine either the cutting data (Appendix 13a) this is related to the workpiece’s diameter and, a diagram (Appendix 13b) to obtain the spindle power from the anticipated chip area, respectively. In a machined surface, both craters and pits do not pose too great a fatigue problem, as they cannot achieve the ‘critical radius’ (i.e see Footnote 67) necessary to instigate a site for crack initiation at a poten-

86 ‘Non-destructive testing’ (NDT), is a range of ‘non-invasive’ sub-surface inspection testing techniques, typically: Eddycurrent testing, Ultrasonics tests, X-ray investigation, etc., that can, in many cases be automated for the detection of otherwise hidden flaws in the component(s).

368

Chapter 7

tial stress concentration location. Furthermore, craters and pits normally exhibit shallow depth-to-width ratios and are normally only present a problem from the ‘cosmetic appearance’. Cracks in the surface are normally classified as either ‘micro-’ , or ‘macro-cracks’ , with these cracks having depth-to-width ratios of >4, typically they can promote: • Reductions in: mechanical strength; fatigue life; plus creep resistance87, • Increases in the susceptibility to stress-corrosion88, • Probability increase in a surface material break-out and generation of debris, • Surface delamination and fatigue. Cracks may be considered as either separations, or narrow ruptures that interrupt the surface continuity and normally include sharp edges, severe directional changes, or both. Macro-cracks can usually be visually inspected with the naked eye, conversely microcracks obviously require microscopic examination. Often these cracks are complex metallurgical interactions which are exacerbated by an ‘abusive’ machine regime, leading to an unacceptable surface condition. A crack’s origin can be the result of several multifarious phenomena, typically they can be an inter-granular attack that might be degraded by surface dissolution, via chemical processes. Whenever preferential intergrannular attack takes place, it can additionally promote a grain boundary network of micro-cracks that can extend beneath the surface, tracing-out and following the underlying grain boundaries. Even micro-cracks should not be ignored, as they can affect the component’s functional performance, because they act as a potential source for macoscopic crack fatigue. Hence, once a crack has been generated it cannot be successfully resealed, owing to subsequent contamination and continuous chemical reactions. In fact, the process of fatigue failure (i.e. see Fig. 190 bottom-right for photomicrographs of a crankshaft’s fatigue failure

87 ‘Creep’ , is: ‘The time-dependent plastic deformation of materials that occur under constant load at relatively high temperatures and low stresses’. 88 ‘Stress-corrosion cracking’ (SCC), is: A combined mechanical and chemical failure mechanism in which a non-cyclic tensile stress [below the yield strength] leads to the initiation and propagation of fracture in a relatively mild chemical environment’.

mechanism) can be characterised by three discrete steps: 1. Crack initiation – where a minute crack forms at a particular site, such where a high stress concentration occurs, 2. Crack propagation – during which time at which the crack incrementally advances with each stress cycle89, 3. Final failure – rapidly occurs, once the advancing crack has reached a critical size being close to ‘speed of sound’: Mach 1 – and is a catastrophic failure mechanism.

White-Layers The so-called ‘white-layers’ 90 that can appear when ‘abusive machining’ certain ferrous work-hardening materials, are a result of microstructural and metallurgical alterations to the machined sub-surface layers of a workpiece (Fig. 189c). This undesirable and unwanted ‘white-layer’ condition is visually apparent (i.e. when a taper-section through the machined surface has been taken), as it resists standard etchants and the consequence is a visible ‘white-layer’ – when viewed under an optical microscope.

89 ‘Striations’ , (also known as ‘Beach-’ , or ‘Clamshell-marks’ – see Fig. 190bottom-right), are concentric ridges that expand away from the initial crack site(s), frequently appearing in a circular, or as a semi-circular radial pattern. NB This ‘striation effect’ is analogous to that of a stone being dropped into a still pond – with the stone entry being the equivalent of the initial crack site, while the radial/circular waves generated, are akin to the cumulating concentric stress ridges – until they intersect with the pond’s bank [ie free-surface]. 90 ‘White-layers’ , are a metallurgically unstable sub-surfaces exhibiting a very hard localised state, with a supplementary heat-affected zone (HAZ) beneath it, which is softer than the overall bulk hardness of the workpiece’s matrix – hence, this metallurgical instability. ‘White-layers, can be classified depending upon whether it resulted from: mechanical; chemical; or thermal events, which also directly relates to machined workpiece factors such as: strain; strain-rate; heating/cooling rates; plus environmental conditions. NB In the past, ‘white-layers’ were known by several terms, such as: ‘white-phases’; ‘white-etchings’; ‘hard-etchings’; etc. – depending upon the variety and type of ‘white-layering’ production.


. Figure 189. The influence of the cutting edge’s condition on the resultant machined surface integrity

369

370

Chapter 7

In Fig. 189c, a ‘white-layer’ (i.e. for this ferrous drilled part, being a localised untempered martensitic phase of 63 HRc91) exists beneath the recast and redeposited layer, in this case produced by a ‘dull’ drill’s cutting lips and margins. Due to the fact that the recast layer (i.e. heat-affected zone – HAZ) has a similar metallurgy to that of the ‘white-layer’ , with the delineation of these ‘white-layers’ regions and their accompanying HAZ’s are not clearly defined. This latter HAZ is a complex metallurgical condition, comprising of some: untempered martensite (UTM); over-tempered martensite (OTM), while beneath these layers, the bulk substrate material remains unaffected. The thickness of these ‘white-layer’ zones is strongly influenced by both the actual plastic deformation created here and, to a lesser degree, by the thermal influence of the passage of the tool’s edge over the machined surface as heat penetrates into the locality of the component’s surface. Probably the worst ‘abusive machining’ conditions that can exist, are when drilling holes in workhardening materials having long length-to-diameter ratios (i.e. L/D ratios of >12:1) with inadequate coolant supply, creating high levels of friction, this condition being exacerbated by an inefficiency produced by a ‘dulled’ drill’s cutting lips. Virtually all tooling even the most sharp – the notable exception here being monolithic faceted natural diamond cutting edges, have a finite tip radius of ≈8 µm (i.e. see Fig. 184 – high-lighting the tool tip ‘rounding effect‘), this results in increased forces and tool wear, which can transform the surface metallurgy by thermo-mechanical generation. The case has already been made concerning the fact that machining processes impart residual stresses into the surface layers, as indicated in the schematically-represented milling conditions shown in Fig. 190 and graphically, in Fig. 191 for a series of milling operations where preset ‘wear lands’ were generated on the cutter’s teeth prior to workpiece machining. This latter case (Fig. 191) of artificially-inducing a controlled ‘wear land’ onto the face-milling cutter’s individual tooth (i.e. with the other teeth removed, hence, acting as ‘Fly-cutter‘), then after

91 By way of comparison of this untempered martensitic ‘whitelayer’ phase, a conventional high-speed steel (HSS) milling cutter’s teeth would have had a maximum hardness after heattreatment of 62 HRc, which clearly signifies the true local hardness of these ‘white-layers’.

several milling passes plotting the residual stress levels from the surface and into the 4340 steel workpiece’s substrate under standardised cutting data (i.e the steel specimens having previously been quenched and tempered to a bulk hardness of 52 HRc). Hence, the effect of these different induced tool wear rates and their influence in terms of their respective magnitudes and depths, can clearly be seen. Even when the cutting edge has ‘sharp tooth’ , a certain degree of tensile residual stress was apparent in the immediate surface region. Here, directly under this tensile stress zone, the stress concentration changed to one of compression (i.e. to a depth of ≈50 µm). As each milling cutter tooth flank became steadily more worn, the substrate compression layer also increased in magnitude, which could lead to considerable workpiece distortion, once the clamping forces had been released – particularly if only one-side of the part was milled (i.e. see Fig. 186b). If the forces involved in the machining process exceed the flow stress, plastic deformation occurs and the structure is deformed. In the case o ductile materials, the plastic flow can create a range of degenerative surface topography characteristics, such as: burrs; laps; BUE residue; plus other unwanted debris deposits. If this deformation becomes severe as a result of excessive plastic flow, any grains adjacent to the surface may become fragmented to such an extent that little, or no metallic structure can be metallographically resolved, therefore ‘white-layering’ will result. Normally, a ‘white-layer’ region extends to quite a small depth beneath the surface, in the region of 10 to 100 µm, depending upon the severity of the ‘abusive regime’ of surface generation. Considering Fig. 191 once again, as can be seen, the residual stress is indicated along the vertical axis, here instead, it is alternatively possible to superimpose a micro-hardness axis – see Fig. 191 circular inset graph. A note of care is required when changing the vertical axis from residual stress to that of micro-hardness, as they are two distinct quantitative values. As mentioned the hardness profile closely approximates that of the residual stress curve, however in the latter case, instead of tensile stress at the in the surface region, the sub-surface layer could equally be compressive in nature. ‘White-layers’ must be avoided under all occasions, because of the unstable metallurgical condition, compounded by the fact that the these regions act as potential stress-raisers for any critically-engineered component and can lead to premature failure, or at worse, catastrophic failure in-service.


. Figure 190. Typical fatigue characteristics within the component’s surface region, being influenced by the mode of milling: up-cut or down-cut

371

372

Chapter 7

. Figure 191. Comparison of the residual stresses in some milled surfaces, obtained with artificiallyinduced tooth wear lands. [After: Field & Kahles, 1971]


373

Altered Material Layers So that an impression of the altered material layers (AMLs) that can occur for a diverse range of: surface and sub-surface topographical features; different metallurgical processes; mechanical applications and uses; Table 13 has been constructed, to high-light their particular influence on functional performance. In the majority of cases given in Table 13, the influence of

these sub-surface defects tends to be of significance, especially with respect to an ‘abusive regime’ producing a machined ‘white-layer’. In some instances, the ‘altered material zone’ (AMZ), can affect component inservice performance in a variety of ways. For example, where thein-service tribological situations produce either re-deposited, or recast layers in the surface region, it has been known that such defects will influence wear and affect reliability. This often undetected sub-surface

. Table 13. The influence of substrate features on function Surface integrity: sub-surface features Function:

Metallurgy UTM or WL

Wear



Strength



Chemical attack



Fatigue



OTM rev





Deformation

Deposits

Stress

Aust

IGA

WL

Plastic defn

Burrs

Cracks

Tears and laps

Tool frags

Redp matl

Res stress







































































Magnetism







Bearings



















Seals



















Friction













Forming















Bonding and adhesion





















Key: : strong influence on function; : some influence on function; : possible influence on function Abbreviations: UTM: untempered martensite; OTM: over-tempered martensite; Aust rev: austenitic reversion; IGA: intergranular attack; WL: white-layer; Plast defn: plastic deformation; Tool frags: tool fragments; Redp matl: re-deposited material; Res stress: residual stress. [After: Griffiths et al., 2001]

374

Chapter 7

condition degrades the functional performance, due to the fact that they are the product of hard, brittle and unstable layers, with tensile residual stresses present. These factors, combined with an acute alteration to the bulk substrate, are likely to ‘spall’ (i.e. delaminate and break-away). Conversely, if a sub-surface feature produces severe plastic deformation, evidence has shown in particular for the die and tool industry, that some dies benefit from increased life due to enhanced abrasion resistance. From Table 13, the design engineer can see that by simply selecting a production process without an intimate knowledge of how components are to be manufactured will inevitably affect the subsequent part’s in-service application. Moreover, due regard must be given to the machined workpiece’s potential sub-surface state, as this condition will inexorably lead to problems in terms of potential impairment of its servicing needs and reliability.

Surface integrity Manipulation – Burnishing Part’s for Surface Improvement Burnishing and in particular roller burnishing (Fig. 192) is a very fast production technique for improving both the finish and dimensional accuracy of either an internal, or external surface, by pressure rolling without removal of workpiece material. Roller burnishing is a cold-working process, that produces a fine surface texture by the application of the planetary rotation of hardened rolls over the previously machined bored, or turned surface (Fig. 192c). Moreover, unlike the primary forming process of cold-rolling which normally produces large sectional changes, roller burnishing involves cold-working just the surface layers of the workpiece, to improve the surface structure. Roller burnishing tooling (Fig.192a) can be used for minute diameter adjustment down to 25 µm, allowing component dimensional accuracies of ±0.006 mm to be obtained. The action of roller burnishing causes plastic deformation of the workpiece’s previously machined surface. At a given depth below the burnished surface, the material is elastically deformed and attempts to spring back. This action, gives rise to compressive stresses at the surface and tensile stresses in the elastically-deformed zone. This complex stress interaction increases the resistance of the material to fatigue failure, because any external forces must firstly overcome these residual stresses.

The potential for cracking that can occur due to the interaction between the static and tensile stresses in the metal and a corrosive medium is termed ‘stress corrosion cracking’. During roller burnishing, these tensile stresses are eliminated when the burnising tool compresses the workpiece surface. Likewise, any pits, scratches and porosities in the surface, which might otherwise collect reactive substances and contaminants, are eliminated, hence, roller burnishing increases the corrosion resistance of the material. Crystalline materials typified by their metal lattices, are never completely without flaws. The atomic lattice will always contain built-in irregularities of various types. These so-called atomic dislocations reduce the strength of the material, as less force is necessary to alter the atomic lattice. Dislocation motion of atoms is a complex subject, which goes beyond the scope of the present text, however, it can be said that upon the application of an external load (i.e. burnishing tooling), because the lattice is invariably not perfect, less force is necessary to defrom the structure. Here, an attempt is made to inhibit the movement of dislocations by means of differing hardening procedures. Coldworking increases the number of dislocations and one would expect the material to become softer, but in fact, the opposite effect transpires. This increased hardness takes place, because there are so many dislocations as a result of cold-working, that they prevent and restrict each other’s motion, as a result the surface hardens. This is what occurs in roller burnishing, as the material is displaced and the net result is that it becomes both harder and stronger – due to dislocation obstructions. By way of a cautionary note, both Rockwell and Brinell hardness testing methods cannot realistically obtain surface hardnesses readings satisfactorily, therefore it is recommended that the Knoop test (Fig. 192b) should be used, then converted with a suitable ‘hardness comparison chart’ – see the appropriate table in Appendix 12. This completes a brief synopsis of a discussion on certain aspects of both machinability and surface integrity, which hopefully conveys the importance of the machining activities and the resulting machined surface condition. Considerably more space could have been devoted to a comprehensive review of these topics, but space was limited, this is the reason for a reasonably comprehensive list of references – for a more in-depth discriminating reading on these important machining and related issues.


375

. Figure 192. Roller burnishing improves the metallurgical properties of the previously machined surface. [Courtesy of Sandvik Coromant]

376

Chapter 7

References Journals and Conference Papers Albrecht, P. New Developments in the Theory of Metal-cutting Process – Part 1: The Ploughing Process in Metal Cutting. ASME J. of Engg. For Ind., Vol.82, 348–358, 1960. Bayoumi, A.e., Xie, Q. and Hamdan, M.N. Effect of Cutting Conditions on Dynamic Properties and Surface Integrity of Work Material. WEAR, Vol. 146, 301–312, 1991. Bellows, G., Field, M. and Kohls, J.B. Influence of Material and its Metallurgical State on Surface Integrity. In: Influence of Metallurgy on Machinability, ASM Pub. (Ohio), 272–295, 1975. Bonifácio, M.E.R. and Diniz, A.E. Correlating Tool Wear, Tool Life, Surface Roughness and Tool vibration in Finish turning with Coated Carbide Tools. Wear, Vol. 173, 137–144, 1994. Boothroyd, G. Temperatures in Orthogonal Metal Cutting. Proc. of IME., Vol. 177, 789, 1963. Bridgman, P.W. In: Phys. Rev., Vol. 48, 825, 1935; ibid. Proc. American Acad. Arts Sci., Vol. 71, 386, 1937; ibid, J. Appl. Phys., Vol. 8(5), 328, 1937; ibid, J. Appl. Phys. Vol.14(6), 273, 1943; and ibid, The Physics of High Pressure, G. Bell & Sons Ltd. (London), 1949. Brinksmeier, E., Cammett, J.T., König, W., Leskovar, P., Peters, J. and Tönshoff, H.K. Residual Stresses: Measurement and Causes in Machining Processes. Annals of the CIRP, Vol. 31(2), 1982. Chou, K.Y. and Evans, C.J. White Layers and Thermal Modelling of Hard Turned Surfaces. Int. J. of Mach. Tools & Manufact., Vol. 39, 1863–1881, 1999. Dagnell, J. Machinability Test by a Drilling Method. Annals of the C.I.R.P., Vol. XV, 301–308, 1967. Dillion, O.W., DeAngelis, R.J., Lu, W.Y., Gunasekera, J.S. and Deno, J.A. The Effects of Temperature on the Machining of Metals. J. Mater. Shaping Technol., Vol. 8, 23–29, 1990. Dornfeld, D.A. Monitoring of the Machining Process by Means of Acoustic Emission Sensors. In: Acoustic Emission: Current Practice and Future Directions, ASTM STP 1077, W. Sachse, J. Roget and K. Yamaguchi, ASTM (Philadelphia), 328–344, 1991. Dornfeld, D.A. König, W. and Ketteler. Aktueller Stand von Werkzeug- und Prozess-überwachung bei der Zer spanung. Neuentwicklungen in der Zerspantechhnologie. Proc. Of the CIRP/VDI Conf., Düsseldorf, 363–376, 1993. El-Wahib, A.I. and Kishawy, H.A. A New Method to Improve the Surface Quality During CNC Machining. Int. J. Prod. Res., Vol. 38(16), 3711–3723, 2000. Eyada, O.S. Reliability of Cutting Forces in Machinability Evaluation. Proc. of FAIM’92, CRC Press, Inc. (Florida), 937–946, 1992.

Ernst, H. and Merchant, M.E. Chip Formation, Friction and High Quality Machined Surfaces. In: Surface Treatment of Metals, ASM Pub. (NY), Vol. 29, 299, 1941. Fathailal, M., Danai, K. and Barber, G. Effect of Flank Wear on the Topography of Machined Surfaes. Tribology Trans., Vol. 36(4), 693–699, 1993. Feather, J.J. Using Value Analysis to Target Customer Service Process Improvements. Ind. Engg. Solutions, 33–39, May 1998 Fick, A.E. [Laws of Diffusion], Annals of Physics (Leipzig, Germany), Vol. 170, 59, 1855. Field, M. and Kahles, J.F. Review of Surface Integrity of Machined Components. Annals of the CIRP, Vol. 20(2), 153–163, 1971. Gorzkowski, E. and Sathyanarayanan, E. Machinability. Cutting Tool Engg., 54–58, Feb., 1999. Griffiths, B.J. Problems in Measuring the Topography of Machined Surfaces Produced by Plastic Deformation Mechanisms. WEAR, Vol. 109, 195–205, 1986. Griffiths, B.J. Manufacturing Surface Design and Monitoring for Performance. Surface Technology, Vol. 1, 61–69, 1988. Griffiths, B.J. and Furze, D.C. Tribological Advantages of White Layers Produced by Machining. Tans. of ASME – J. of Tribology, Vol. 109, 338–342, April 1987. Griffiths, B.J. Deficiencies in Surface Specifications. Proc. of Lamdamap III, Computational Mechanics, 465–474, 1997. Griffiths, B.J. Mechanisms of White Layer Generation with Reference to Machining and Deformation Processes. Trans. of ASME – J. of Tribology, Vol. 109, 525–530, July, 1987. Gugger, M. Getting to the Bottom of Chatter. Cutting Tool Engg., 54–60, April, 2000. Hirao, M. Determining Temperature Distribution on Flank Face of Cutting Tool. J. Mater. Shaping Technol., Vol. 6, 143–148, 1989. Kackar, R.N. Off-line Quality Control Parameter Design and the Taguchi Method. J. of Quality Tech., Vol. 17(4), 176–188, Oct., 1985. Kasahara, N., Sato, H. and Tani, Y. Phase Characteristics of Self-excited Chatter in Cutting. J. of Engg, for Ind., 393–399, Vol. 114, Nov. 1992. Kennedy, B. Calming Chatter – Strategies for Minimising Tool Chatter. Cutting Tool Engg., 28–35, July, 2004. Kirchheim, A., Schaffner, G. and Wolfer, P. Piezoelektrische Sensoren zur kombinierten Messung von Kräften und Acoustic Emission für die Prozessüberwachung. Int. CIRP/VDI Konferenz: Überwachung von Zerspan- und Umformprozessen, Düsseldorf, 30–31.3, 1995. Kohls, J.B. Metallurgical Damage in Drilling and Hole Quality. In: Influence of Metallurgy on Hole Making Operations, ASM Pub. (Ohio), 145–158, 1978.


Kops, L. Gould, M. and Mizrach, M. Improved Analysis of the Workpiece Accuracy in Turning, based on the Emer ging Diameter. ASME Pub., J. of Engg. for Ind., Vol. 115, 253–257, Aug. 1997. Koster, W.P., Field, M., Fritz, L.J., Gatto, L.R. and Kahles, J.F. Surface Integrity of Machined Structural Components. Airforce Matls. Lab. Tech. Report: AFML-TR-7011, mmP Project No. 721-8, Metcut Research Associates Inc., Cincinnati (Ohio), March, 1970. Kramer, B.M. Tribilogical Aspects of Metal Cutting. Proc. of ASME, PED-Vol.54/TRIB-Vol.2, 1991. Kumar, S. and Hoefler, B. Simulated Action [FEA Modelling of Machining]. Cutting Tool Engg., 44–49, March, 1999. Lamb, A.D. Some Aspects of the Character of Machined Surfaces. Metals and Matls., 75–79, March, 1987. Lee, E.H. and Shaffer, B.W. The Theory of Plasticity Applied to a Problem of Machining. J. Appl. Mech., Vol. 18 (4), 405, 1951. Lombardi, G.G. Kinematics and Dynamics. [Internet address can be found as follows]: www.drphyscis.com/syllabus/kinematics/kinematcis.htl Lorenz, G. Measurement of Machinability … A Survey of Testing Methods. IAAE Journal 70–81, June 1966. Marusich, T.D. and Ortiz, M. Modeling and Simulation of High-speed Machining. Int. J. Num. Met. Eng’g., Vol. 38, 3675–3694, 1995. Mechant, M.E. Mechanics of the Metal Cutting Process. J. Appl. Phys., Vol. 16 (5), 267 (a), (6) 318 and 324(b), 1945. Merchant, M.E. and Zlatin, N. In: Mech. Eng’g., Vol. 67, 737, 1945. Obikawa, T., Sasahara, H., Shirakashi, T. and Usui, E. Application of Computational Machining Method to Discontinuous Chip Formation. J. of Manufact. Science and Engg., Vol. 119, 667–674, 1997. Pahitzsch, G. and Semmler, D. Z. fur wirtschaftlich Fertigung. In: Vol. 55, 242, 1960; Vol. 56, 148, 1961; and Vol. 57, 45, 1962. Painter, P.R., Smith, G.T. and Hope, A.D. Performance Evaluation of a Machining Centre using Laser Interferometry and Artifact-based Techniques. Proc. of FAIM’96, CRC Press, Inc. (NY), 962–974, 1992. Pawar, K., Forrester, P. and Glazzard, J. Value Analysis: Integrating Product/Process Design. Integrated Manuf. Systems, Vol. 4(3), 14–21, 1993. Pekelharing, A.J. and Hovinga, H.J. Wear at the End Cutting Edge of Carbide Tools in Finish and Rough Turning. Proc. of 8th Mach. Tool Des. Res. Conf., 6430651, Sept. 1967. Pekelharing, A.J. Built-up Edge (BUE): Is the Mechanism Understood? Annals of the CIRP, Vol. 23(3), 207–211, 1974.

377

Pontius, K. Split Personality [Machining Dissimilar Metals]. Cutting Tool Engg., 41–43, Feb., 2002. Pugh, H.D. Mechanics of the Cutting Process. Proc. of IME Conf. Tech. Eng’g. Manufact. (London), 237, 1958. Rakhit, A.k., Sankar, T.S. and Osman, M.O.M. The Influence of Metal Cutting Forces on the Formation of Surface Texture in Turning. Int. J. of Mach. Tool Des. Res., Vol. 16, 281–292, 1976. Shainin, D. Better than Taguchi Orthogonal Tables. ANSQC Quality Congress Trans., Anaheim, USA, concerning: The Fundamentals and Future of Quality, 446–451, 1986. Shiraishi, M. and Sato, S. Dimensional and Surface Roughness controls in a Turning Operation. Trans. of ASME, Vol. 112, 78–83, 1990. Smith, G.T. Surface Integrity Aspects of Machinability of FeC-Cu Powder Metallurgy Components. Powder Metallurgy, Vol. 33(2), 157–164, 1990. Smith, G.T. The Surface Integrity of Turned Ferrous Powder Metallurgy Components. Powder Metallurgy, Vol. 33(2), 155–165, 1990. Smith, G.T. Secondary Machining Operations and the Resulting Surface Integrity: An Overview. Surface Topography, Vol. 3, 25–42, 1990. Smith, G.T. Some Aspects in the Surface Integrity and Tool Wear Associated with Turning Powder Metallurgy Compacts. WEAR, Vol. 151, 289–302, 1991. Smith, G.T., Hope, A.D., Painter, P.R. and Blackshaw, D.M.S. The Assessment of Machining and Turning Centres, using Artifact-based Techniques. Proc. of Lamdamap I, Computational Mechanics, 275–286, 1993. Smith, G.T. What is Machinability and How Can it be Assessed? Int. Conf. on Industrial Tooling, Shirley Press (Southampton, UK), 105–114, Sept. 1995. Smith. G.T. Inspection of Components Manufactured on a Machining Centre, by On- and Off-line Techniques. Proc. of FAIM’96, Begell House, Inc. (NY), 216–224, 1996. Smith, G.T. Ternary Manufacturing Envelopes (TME’s): A New Approach to Describing Machined Surfaces. Int. J. of Mach. Tools and Manufact., Vol. 40(2), 295–305, 2000. Smith, G.T. Fight to the Finish [Tool Geometry and its Affect on both: Roundness/Surface Texture]. Metalworking Production, 49–52, June, 2001. Spur, G., Pöllmann, A. and Kirchheim, A. Kombi-Sensorik beim Scherschneiden, Zeitschrift für wirtschaftliche Fertigung und Automatisierung. In: Zwf, Vol. 90, 1–2, 1995. Stefanov, Y.P., Makarov, P.V., Burkov, P.V. and Matveev, V.S. Dynamic Simulation of Chip Generation and Formation in Metal Cutting. Theoret. and Appl. Fracture Mechanics, Vol. 28, 117–124, 1997. Stout, K., How Smooth is Smooth? Surface Measurements and their Relevance in Manufacturing. Production Engr., 17–22, May 1980.

378

Chapter 7

Strenkowski, J.S, and Athavele, S.M. A Partially Constrained Eulerian Orthogonal Cutting Model for Chip Control Tools. J. of Manuf. Science, Vol. 119, 681–688, 1997. Tay, A.A.O. A Review of Methods of Calculating Machining Temperature. J. of Matls. Processing Tech., Vol. 36, 225–257, 1993. Taylor, F.W. On the Art of Cutting Metals. Trans. of ASME, Vol. 28, 31, 1907. Thomsen, E.G., Application of the Mechanics of Plastic Deformation to Metal Cutting. Annals of the CIRP, Vol. 14, 113–123, 1966. Tlusty, J. and Polacek, M., The Stability of Machine Tool against Self-excited Vibrations in Machining. ASME Prod. Res. Conf. (Pittsburgh), 454–465, 1963. Tönshoff, H.K. Eigenspannungen und plastische Verformungen im Werkstück durch spanned Bearbeitung. Dr.-Ing. Dissertation, TH Hannover, 1965. Turley, D.M., Doyle, E.D. and Samuels, L.E. A Structure of the Damaged Layer on Metals. Proc. of the Int. Conf. on Prod. Engg., Tokyo, Japanese Society of Prec. Engrs.: Part 2, 1974. Vajpayee, S. Analytical Study os Surface Roughness in Turning. WEAR, Vol. 70, 165–175, 1981. Wallace, P.W. and Boothroyd, G. Tool Forces and Tool-chip Friction in Orthogonal Machining. J. Mech. Eng’g. Sci., Vol. 6 (1), 74, 1964. Watson, D.W. and Murphy, M.C. The Effect of Machining on Surface Integrity. The Metallurgist and Matls. Technol., 199–204, April, 1979. Wells, J. Material Matters [Optimisation of Workpiece Matls]. Cutting Tool Engg., 58–64, June 2005. Williams, J.E., Smart, E.F. and Milner, D.R. The Metallurgy of Machining, Part 1: Basic Considerations and the Cutting of Pure Metals. Metallurgia, 3–10, Jan., 1970. Williams, J.E., Smart, E.F. and Milner, D.R. The Metallurgy of Machining, Part 2: The Cutting of Single-phase, Twophase and some Free Machining Alloys. Metallurgia, 51–59, Feb., 1970. Whitehouse, D.J. Beta Functions for Surface Topologie? Annals of the CIRP, Vol. 27, 491–497, 1978. Whitehouse, D.J. Conditioning of the Manufacturing Process using Surface Finish. Proc. of Lamdamap III, Computational Mechanics, 3–20, 1997. Whitehouse, D.J. Surface Measurement Fidelity. Proc. of Lamdamap IV, WIT Press, 267–276, 1999. Young, H.T. An Experimental Investigation of Temperatures in Chip Formation. Proc. of 29th Int. Matador Conf. (Manchester), The Macmillan Press Ltd., 559–565, April 1992. Zorev, N.N. Interrelation between Shear Processes Occurring along Tool Face and on Shear Plane in Metal Cutting. Int. Res. in Prod. Eng’g., 42, 1963 (Presented at Int. Prod. Eng’g. Res. Conf., Pittsburgh, 1963).

Books, Booklets and Guides Akao, Y. Quality Function Deployment (QFD) – Integrating Customer Requirements into Product Design. Productivity Press (Portland, Oregon), 1988. American National Standard (ANS) B211.1 – Surface Integrity. 1986. Amontons, G. Histoire de l’Académie Royale des Sciences avec les Mémoires de Mathématique et de Physique. Paris, 1699. Armarego, E.J.A. and Brown, R.H. The Machining of Metals. Prentice-Hall Pub., 1969. Avner, S.H. Introduction to Physical Metallurgy (2nd Ed.). McGraw-Hill Int. Book Co., 1974. Arvill, J. Mechanical Engineer’s Data Handbook. Butterworth-Heinemann, 1997. Bajpai, A.C., Calus, I.M. and Fairly, J.A. Statistical Methods for Engineers and Scientists. John Wiley Pub., 1979. Bedford, a. and Fowler, W. Engineering Mechanics – Statics and Dynamics (3rd Ed.). Prentice-Hall, Inc., 2002. Bell, S. A Beginner’s Guide to Uncertainty of Measurement – Measurement Good Practice Guide. NPL Pub. No. 11, Aug. 1999. Bhattacharyya, G.K. and Johnson, R.A. Statistical Concepts and Methods. John Wiley Pub., 1977. Birch, K. Estimating Uncertainties in Testing – Measurement Good Practice Guide. NPL Pub. No. 36, March 2001. Bird, J. Engineering Mathematics. Newnes – Elsevier Science Pub., 2003. Boulger, F.W. Machining – Theory and Practice. ASM (Cleveland, OH), 1950. Boothroyd, G. Fundamentals of Metal Machining and Machine Tools. McGraw-Hill, New York, 1975. Bowman, K. Mechanical Behavior of Materials. John Wiley & Sons, Inc., 2004. Bowden, F.P. and Tabor, D. Frictionand Lubrication of Solids. Oxford University Press (London), 1954. Brammertz, P.H. Ursachen für Form und Massfehler an Feinbearbeiten Werkstüchken. Dissertation (T.H. Achen), 1960. Carvil, J. Mechanical Engineer’s Data Handbook. Butterworth-Heinemann, 1994. Childs, T.H.C., Maekawa, K., Obikawa, T. and Yamane, Y. Metal Machining – Theory and Applications. Arnold Pub., 2000. Collett, C.V. and Hope, A.D. Engineering Measurements. Pitman Pub. Ltd., 1979. Coullomb, C.A. Mémoires de Mathématique et de Physique de l’Académie Royale des Sciences. Paris, 1785. Cross, N. Engineering Design Methods – Strategies for Product Design (3rd Ed.). John Wiley & Sons Ltd., Nov. 2003.


Cutting Tool Materials. American Soc. of Metals (OH), 1981. Dagnall, H. Let’s Talk Roundness. Taylor Hobson Ltd. Pub., Nov. 1996. Dagnall, H. Exploring Surface Texture (3rd Ed.). Taylor Hobson Ltd. Pub., March 1998. Degarmo, E.P, Black, J.T and Kosher, R.A. Materials and Processes in Manufacturing (9th Ed.). John Wiley & Sons, Inc., 2003. Dowson, G. Powder Metallurgy – The Process and its Products. Adam Hilger – IOP Pub. Ltd., 1990. Ernst, H. Physics of Metal Cutting. ASM (Cleveland, OH), 1938. Griffiths, B. Manufacturing Surface Technology – Surface Integrity and Functional Performance. Penton Press, 2001. Höganäs AB Pub. Machining Guidelines – Book 5. Sept., 1998. Higgins, R.A. Properties of Engineering Materials. Hodder and Stoughton Educ., 1979. Influence of Metallurgy on Hole Making Operations. American Soc. of Metals (Ohio), 1978. Influence of Metallurgy on Machinability. American Soc. of Metals (Ohio), 1975. Preliminary Study of Variations in Machinability of Carbon Steel. PERA Pub. No. 179, April 1968. Kaczmarek, J. Principles of Machining by Cutting, Abrasion and Erosion. Peter Peregrinus Pub. (Warsaw, Poland), 1976. Kalpakjian, S. Manufacturing Processes for Engineering Materials (3rd Ed.). Addison Wesley Longman, Inc., 1997. Komanduri, R. Symposium on US Contributors to Machining and Grinding Research in the 20th Century. American Soc. for Mech. Engrs.: Applied Mechanics Reviews, Vol. 46(3), March 1993. Kronenberg, M. Machining Science and Application. Pergamon Press (NY/London), 1966. Kruszynski, B.W. and Cuttervelt, C.W. Plastic Deformation in the Surface Layer after Cutting. Advanced Manufacturing Engineering, Vol. 1, 1989. Mills, B. and Redford, A.H. Machinability of Engineering Materials. Applied Science Pub., 1983. Modern Metal Cutting – Part 12: Machining Economics. Sandvik Coromant Pub., 1981. Mummery, L. Surface Texture Analysis – The Handbook. Hommelwerke GmbH Pub., 1990. Ramsey, D.C. Engineering Instrumentation and Control. Stanley Thornes Pub. Ltd., 1981. Reason, R.E. The Measurement of Surface Texture. CleaverHume Press Lt., 1960.

379

Roller Burnishing, Sandvik Coromant Pub., No.: HV-5300:012-ENG. Rowe, G.W. An Introduction to the Principles of Metalworking. Edward Arnold (Pub.) Ltd, 1968. Sander, M. A Practical Guide to the Assessment of Surface Texture. Feinprüf GmbH Pub. (Göttingen), 1991. Sandvik Coromant Pub. The Cutting Process. No.: CMSP 900–038 ENG, 1989. Schaffer, J.P, Saxena, A., Antlovich, S.D., Sanders, Jr., T.H. and Warner, S.B. The Science of Engineering Materials (2nd Ed.). McGraw-Hill Comp. Inc., 1999. Shaw, M.C. and Cook, N.H. Metal Cutting Principles. MIT Pub. 3rd Ed., Chapter 4, 1954. Shaw, M.C. Metal Cutting Principles. Clarendon Press (Oxford), 1984. Smith, G.T. Advanced Machining – The Handbook of Cutting Technology, IFS Pub./Springer Verlag, 1989. Smith G.T. CNC Machining Technology, Springer Verlag, 1993. Smith, G.T. Industrial Metrology – Surfaces and Roundness. Springer Verlag, 2002. Sokolowski, A.P. Prazision in der Metallbearbeitung. VEB Verlag Technik, Berlin, 1955. Stout, K. Quality Control in Automation. Kogan Page Ltd., 1985. Surface Technology – Wear Protection. Verlag Moderne Industrie AG & Co., 1990. Swanson, R.C. Quality Improvement Handbook – Team Guide to Tools and Techniques. Kogan Page Ltd., 1995. Thomas, T.R. Rough Surfaces (2nd Ed.). Imperial College Press, 1999. Trent. E.M. Metal Cutting (2nd Ed.). Butterworths & Co (Pub.) Ltd., 1984. Taguchi, G., Experimental Designs (3rd Ed.). Vol. 1, Maruzen Pub., 1976. Valentino, J.V. and Goldenberg, J. Introduction to Computer Numerical Control (CNC) (2nd Ed.). Prentice Hall Int., 2000. Vickerman, J.C. Surface Analysis – The Principal Techniques. John Wiley and Sons Ltd., 2000. Vieregge, G. Zerspanung der Eisenwerkstoffe [Machining of Ferrous Metals], Vol. 16 of Stahleisenbücher [Iron and Steel Series – 2nd Ed.], Verlag Stahliesen (Düsseldorf, 1970. Whitehouse, D. Surfaces and their Measurement. Hermes Penton Science., 2002. Zorev, N.N. Metal Cutting Mechanics. Pergamon Press (Oxfrod), 1966.

8 Cutting Fluids

‘Everything flows and nothing abides.’

HERACLITUS

(540 – 480 BC) [An early Metaphysician from Ephesus (Asia Minor), in: On Nature]

382

Chapter 8

8.1 Historical Development of Cutting Fluids General speaking, metalworking mass production techniques can be traced back to the 16th century, but it was really not until the late 18th century that engineers in the industrialised countries paid close attention to increasing production, due to the vast rise in their populations and significant industrial growth. In Europe at that time, two countries where important areas of applied machining and fluid research were being pursued was: in France, where the machining of metals was being investigated and developed into a science – specifically in terms of the effects of tool feeding and lubrication and its affect on surface finish; also in the mid-19th century in England, where the effects of water as a coolant to enhance tool performance was also studied. Thus, as these research activities progressed, complementary advances were taking place into the study of tool materials their heat treatment and in particular, tool hardening techniques. As has happened on many occasions in the past, considerable advances took place as a result of the enormous demands for armament manufacturers and their production needs during times of war, significantly adding advancements and refinements to the: machine tools; tooling; as well as for lubricants. As these research programmes developed, it soon became clear that for cutting fluids, while water may have had the optimum specific heat capacity of all available fluids, it brought real problems due to corrosion of the machined components and to the exposed surfaces of machine tools. Frequently, such related losses far outweighed the benefits of increased production throughput and the improvement in tool life that it imparted to the overall manufacturing process. A simple solution was at hand in the form of corrosion inhibition, via the use of: animal oils; fatty acids; soda; which when combined with water to form a ‘soap’ , offered an improvement in product protection against rusting, while effectively retaining the overall cooling properties of water. In particular in the North of England during the mid-19th century, notably around Manchester and the Huddersfield areas – where the world’s major cotton and woollen industries were now in full production. These areas, had for centuries used the benefits of ‘soft-water’ , which unfortunately had a tendency for the coolant solutions to generate large quantities of

foam, hence the term ‘suds’ which is sometimes used to the present day, although this name is hardly relevant to modern-day cutting fluids. Almost by accident and as an incidental benefit of these ‘soap solutions’ , they were found to impart improved lubricating properties between the tool and the component, through a ‘machining mechanics’ and ‘chemical relationship’ that was at the time, not fully comprehended. In the meantime, mineral oil: which had advanced from being simply thought of and used as just an aliphatic additive to vegetable oil, to that of becoming recognised as a useful lubricant, which was at the time currently and widely available. This mineral oil was demonstrably shown at the time, to offer improvements to both the machined surface finish quality and enhancing the tool’s lubrication. At the beginning of the 20th century, experimental studies into topics such as the initial studies into: boundary lubrication; lubricity; plus its relative viscosity; for the newly-developing engines in the automotive industries were being rapidly developed. Moreover, when general manufacturing industries started the mass production of consumer goods, the resultant quality was of prime importance and basic water lubrication was now no longer sufficient. By the mid-20th century, the preliminary forays into the basic development of today’s modern-day cutting fluids occurred. At this juncture, it soon became evident that it was essential to combine the properties of several dissimilar fluids to produce an early, but ‘workable’ form of cutting fluid, these ‘ingredients’ were: • Oil – to act as a lubricant between the chip, tool and machined workpiece, • Water – for cooling, to extract the heat from the cutting process, • Detergent – to break down the ‘surface tension’ between the oil and water,

‘Surface tension’ , this is often generally defined as the: ‘Interfacial tension between two phases, one being a liquid, while the other is a gas’. More specifically, surface tension is a physical force in the surface of the liquid that arises as a result of the liquid’s atoms pulling their neighbours in all directions. While atoms deep in the liquid have no net force applied to them, conversely, surface atoms have no neighbours above them, as a result they experience a net inward force from the bulk of atoms below them. Hence, this net inward force is known as its surface tension, with the greater the radius of curvature, the higher the surface tension (i.e see Fig. 208a). Hence, a ‘droplet of water’ sitting on a flat surface – termed a ‘spherical cap’ has

Cutting Fluids

383

• Sulphur – to act as an ‘extreme pressure’ (EP) ad- ternal turning processes, an insert’s chip-breaker will ditive to reduce frictional effects at the various cutting interfaces.

NB Sulphur was soon to prove unpopular as a satisfactory EP additive, as it had a tendency to stain, or erode certain decorative machined finishes for specific metals and alloys.

8.2 Primary Functions of a Cutting Fluid In the previous section, it was recognised that two of the primary functions of a cutting fluid was to cool and lubricate both the workpiece and cutting tool’s edge. In addition, one could add the improvement of machined surface quality and an increase in tool life (i.e. see Fig. 193b). Further, it has been shown that a reduction in spindle power is an added bonus to many machining processes, which offers considerable savings when this reduction in electrical demand is accrued per annum. If a problem occurs where work-hardened swarf disposal from the cutting vicinity presents an obstacle to efficient cutting, then flushing this zone with flood coolant, may eliminate this difficulty. Effective chip removal by the application of flood coolant (Fig. 194a: showing a twist drilling operation, 194b: milling with the periphery using a ‘porcupine cutter’), can minimise an otherwise serious problem on machining centres where large volumes of densely-packed swarf can impede the cutting process. Even on continuous cutting operations such as when undertaking external/in-

a high contact angle (i.e. the angle of tangency that the spherical cap, or a bubble makes with the surface). When this angle is considered on a ‘wetability scale’ it has a high contact angle and as such, is not considered as ‘wet’!, due to its high curvature (Fig. 208a), as indeed does a typical lubrication oil. Conversely, a liquid detergent does not have particularly a high contact angle and as such, will chemically react with both the oil and water and breaks-down this surface tension between the concentrations of an unmixed oil and water. This loss of surface tension between these two ‘products’ thereby produces a basic mixture, or suspension and it then becomes somewhat ‘milky’ in appearance, thus it is often termed a (basic) ‘emulsion’.

break the swarf into convenient shapes and sizes, but these chips may still necessitate flushing-away – being deposited into a swarf conveyor and then onward into an adjacent skip. Swarf removal has become of significant importance as material removal rates have increased with latest tooling advances and high-production machine tools, where they may be continually fed wrought material allowing them to operate untended for 24 hours per day. Possibly the most stringent test for any cutting fluid is in deep-hole drilling applications (Fig. 58), where coolant is delivered under highpressure through suitable coolant holes and is forced up to the cutting edge to not only cool the drill, but provide lubrication and flush any swarf back and away from the cutting vicinity. In fact, with extremely highpressure coolant delivery systems having pressures >300 MPa, such as when using a through-the-nose indexable insert short-hole drill as illustrated in Fig. 195, the advantages are: increased speeds; penetration rates; more holes per insert edges are achievable – see the following section for more details on this highpressure coolant delivery topic, with particular reference to turning operations.

8.3 High-Pressure JetAssisted Coolant Delivery Probably the most important criteria in many metal cutting operations is an acceptable chip control, with respect to its: chip form; chip-flow; plus its chip-breaking ability. It has been mentioned earlier in Chapter 2: Section 2.5, that good chip control will have an affect on: tool life, machined surface texture; cutting forces; reliability; etc. Productivity is strongly influenced by poor chip control, as the machine tool must be frequently stopped to manually remove the vast quantities of swarf present in the working area. This problem becomes especially acute when turning smaller internal diameters on products, since limited space soon becomes filled and compacted with work-hardened chips, that can damage the recently machined surfaces. Reasonable chip control can often be achieved by indexable inserts with an appropriate cutting geometry that is having chip-formers, these being developed to meet the requirements for specific machining operations.

384

Chapter 8

. Figure 193. Heat dissipation during machining can be lessened by utilising appropriate cutting fluids

Cutting Fluids

. Figure 194. ‘Standard-pressure’ (i.e. 200%, operating with lower cutting forces because of the improved frictional conditions between tool/chip interface, with an attendant reduction in machining-induced vibration levels. All of these advantages will improve the machined surface texture and offer better and more consistent dimensional accuracy, by a reduction in component process variability.

Cutting Fluids

When utilising high-pressure jet-assisted machining at cutting fluid pressures of >110 MPa with a velocity of >122 ms–1, some precautions need to be considered, prior to applying this cutting fluid strategy. Caution should be made, when using certain types of cutting tool geometries and grades, as they may not have been designed for this increased level of coolant delivery, which could if inappropriately applied, actually lower productivity. The above mentioned effects do not only depend on effective heat dissipation, but require the contact length between the chip and the rake face to be reduced. Since the application of a coolant by a highpressure jet, partially penetrates between the tool/chip interface, via a ‘hydro-wedge’ which here is created and, then provides hydrodynamic lubrication at this position in the ‘friction zone’. Hence, the shorter the contact length the lower the friction, causing a larger shear angle, which in turn lowers the chip compression factor (Fig. 196). This ‘hydraulic wedge’ – as a result of HPC, influences the chip formation in several ways, it affects both the ‘up- and side-curl’ , thus breaking them into manageable pieces as well as vectoring the chips. By aiming the HPC cutting fluid jet to either the main, or secondary cutting edges this will influence and affect chip-curling behaviour. This chip-curling action in turn, affects the resultant tool life, as it is thought that a reason for this difference in respective tool life is due to the temperature distribution on the rake’s face – as a result of the vectoring angle of the jetassisted coolant application.

8.4 Types of Cutting Fluid

387

down of these cutting fluid groupings). In most technological countries, relevant Standards for both the chemical and technical requirements are published concerning their: storage, usage and disposal, along with their pertinent operator health needs. The ‘Aqueous-based’ cutting fluids can be divided into either: ‘emulsifiable’; or ‘water-soluble’ types (Fig. 197). The former ‘Oil-based’ cutting fluids are supplied as readyto-use products, while the ‘Aqueous-based’ products are normally offered in the form of a concentrate, which must be admixed with water to the desired concentration, prior to use. Once these latter products have been mixed with water, the ‘emulsifiable’ versions form an ‘emulsion’ , whereas the ‘soluble’ type forms a ‘solution’ . In both cases, the resulting cutting fluid is termed ‘water-mixed’.

The ‘Ideal’ Cutting Fluid Having accepted the fact that a cutting fluid is a requirement for the machining of many of today’s engineering materials, be they either metallic, non-metallic in composition and are necessary for various production processes. Then one must ensure that the selected ‘fluid’ achieves its intended purpose, moreover, that it does not create additional problems. These conditions imply that there are many and varied specific characteristics that an ‘ideal’ cutting fluid should possess, such as: • Optimum cooling and lubrication – clearly, the ‘ideal’ cutting fluid would have the most favourable cooling and lubricating properties, to ensure paramount cutting performance as measured by: production rate; tool life; surface texture,

Introduction Modern cutting fluids can be sub-divided into two major classifications: ‘Oil-’ , or ‘Aqueous-based’ , with further sub-division into ‘Semi-synthetic’ , or ‘Synthetic’ fluids (i.e. see Fig. 197, for a ‘family-tree’ and break-

‘Hydrodynamic wedge’ , as its name implies, cannot actually penetrate into the chip/tool interface, as the separation pressures here – at the interface – are simply far too high. However, this hydrodynamic wedge acts as a sort of ‘lever’ (Fig. 196) on the emerging formation of the curling chip, changing its contact length, which in turn, modifies the shearing zone and as a result, influences the chip compression factor.

‘Emulsions’ , are a disperse system (consisting of several phases), which arises through mixing of two liquids which are not soluble in each other. Hence, one liquid forms the inner, or disperse phase, distributed in droplet form in the carrier liquid (the outer, or continuous phase). NB The emulsifiable metalworking fluids are what is known as ‘oil-in-water’ emulsions, that is the oil forms the inner phase, conversely, its counterpart is formed by emulsifying metalworking fluids, which are ‘water-in-oil’ emulsions. (Source: Cincinnati Milacron/Cimcool, 1991)

‘Solutions’ , are a metalworking fluid solution, these are watersoluble fluids mixed with water. (Source: Cincinnati Milacron/ Cimcool, 1991)

388

Chapter 8

. Figure 196. The effect of rake angle on chip thickness – with and without coolant supply

Cutting Fluids

. Figure 197. The main types of cutting fluids for machining operations

389

390

Chapter 8

• Acceptability to the operator (i.e. when machine

tool is manned) – as all operators have some degree of exposure/contact with the cutting fluid. Operators will consider the lubricant’s overall performance, but even when the fluid is ‘perfect’ in every other respect, complaints are likely if, for example, the smell was unpleasant. The following features are likely to be of particular interest to the operator: – Smell – ideally, the cutting fluid should have no perceivable odour, but if present, it certainly should not be objectionable, – Colour and clarity – most operators prefer products which are perceived to be ‘clean and fresh’ throughout their life and, some operators prefer dye-coloured translucent products for this reason, – Misting – high-speed cutting operations tend to generate a mist. Occasionally these mists may be associated with operator health problems: dry-throats; stinging eyes; etc.; leading to complaints. Although misting is largely dependent on the: machine tool; its operation; atmospheric ventilation; etc.; different fluids have diverse misting characteristics and, ‘ideally’ the fluid should be non-misting – more will be said on the operator’s health issues later in the chapter, – Irritation to the skin and eyes – these operator issues have been associated with physical contact from cutting fluids, such as: skin and eye irritation; itching; rashes, swellings; stinging; etc. Once again, fluid formulations that are ‘kind and gentle’ are preferred. As mentioned above, more will be mentioned on these health issues shortly, • Long ‘sump-life’- with all machining fluids having a finite life, at some point, the machine’s cuttingfluid system must be completely emptied, cleaned, flushed through and refilled with new fluid. There are numerous reasons why the cutting fluid might be regarded as ‘dead’ – these points will be raised when investigating the ‘problems’ with cutting fluids later in the chapter. The fluid’s life is an important economic consideration in terms of: fluid usage; labour costs; down-time; etc. Some leadingmanufacturer’s cutting fluid formulations are capable of achieving significantly better overall performance and have an extended ‘sump-life’ over their cheaper contemporaries. The increased ‘sump-life’ will enable better use of a company’s maintenance department’s manpower resources, thereby enabling it to be more efficient in their anticipated

•

•

•

•

‘planned maintenance scheduling’ over prescribed shut-down periods, or ‘maintenance windows’ , Corrosion protection – all cutting fluids are formulated to provide corrosion protection to the machine tool and the workpiece during and, for a short time after the cutting operation. In the main, these fluids should preserve their corrosive-protection properties throughout their useful life, to avoid the potentially expensive problems of rusting of the machine and machined components alike, the latter being rejected by the customer. More information will be given on corrosion protection later in the chapter, Low foaming properties – on some of today’s machine tools, they incorporate fluid systems that agitate the cutting fluid to such an extent that foam spills out of the coolant tank and onto the floor. The ‘ideal’ fluid will withstand: swarf-washing jets; high-pressure fluid delivery; centrifuges; etc.; even when prepared with the softest quality water supply. The subject of foaming will be addressed specifically later in the chapter, Machine tool compatibility – no self-respecting engineer wants to see their newly purchased, or well-maintained older machine being attacked by its cutting fluid. The optimum fluid should create no detrimental effects on the machine tool’s: paint finishes; seal materials; screens and guarding; etc., Workpiece compatibility – means that the widest possible range of workpiece materials should be machined with a particular, but versatile grade of

‘Planned maintenance scheduling’ , many companies adopt either a ‘Total Productive Maintenance’ (TPM), or ‘Reliability-centred Maintenance’ (RCM) philosophical and practical approach to their overall maintenance organisational needs. NB TPM is defined as: ‘A system of maintenance which covered the entire life of every piece of equipment in every division including planning, manufacturing and maintenance. (Source: Japan Institute of Plant Maintenance – JIPM, 1971). RCM is defined as: ‘A process used to determine the maintenance requirements of any physical asset* in its operating context’. (Source: Moubray, 1996) Often, companies run both a TPM and RCM strategy together, to achieve an overall high level of maintenance planning discrimination, coupled to plant security – in association with individual asset reliability.

* A physical asset is any piece of operating plant, or equipment that requires a maintenance function to be undertaken upon it at some prescribed time period, or requiring various modifications to it for its operating context.

Cutting Fluids

•

•

•

cutting fluid, although certain non-ferrous metals may have a susceptibility to staining, so here, it is prudent to discuss the problem with the cutting fluid manufacturer, Water-supply compatibility – a water-soluble cutting fluid should ‘ideally’ be capable of being diluted with any water supply. Geographical locations can create variations in water supply and its condition, this latter factor is especially true for water hardness (i.e see Fig. 199b), where its hardness can vary quite considerably. Thus, the ‘ideal’ cutting fluid would not cause the typical problems of: foaming in soft waters; or forming insoluble soaps in hard waters, Freedom from tacky, or gummy deposits – as water soluble fluids dry out on a machine, or component’s surface, the water content evaporates to leave a residue which is basically the product concentrate. This residue should ideally be light and wet, allowing it to be easily wiped-off. However, any gummy, or tacky deposits collect swarf and debris, necessitating increased machine and component cleaning, ‘Tramp oil’ tolerance – is a lubricating, or hydraulic oil which leaks from the machine tool and contaminates the cutting fluid. Most modern machines are equipped with ‘total-loss’ slideway lubricating systems which can contaminate the cutting fluid with up to a litre of oil per day – on a large machine tool. The ‘ideal’ cutting fluid would be capable of tolerating this contamination without any detrimental effects on its operating performance. Some cutting fluids are formulated to emulsify the ‘tramp-oil’ , while other fluid formulations reject it, allowing

‘Total-loss’ fluid systems, are as their name implies in that they purposely leak oil to the machine’s bearing surface, requiring periodic tank replenishment. When this oil leaks-out of the machine tool it is termed: ‘tramp-oil’ , therefore the oil will eventually end up in the machine tool’s coolant tank, where it is either tolerated by the coolant product, or is separated-out, requiring periodic ‘tramp-oil skimming’. NB ‘Tramp-oil’ losses are invariably not accounted for in many production shops, which invariably means their ‘economic model’ for such losses are habitually not considered, or not even thought about by the company. It has been reported that on a quite ‘large-sized’ horizontal machining centre, it can lose up to 365 litres of ‘tramp-oil’ per annum, which is an on-going cost that needs to be addressed. Multiply this individual machine tool loss by the number of machines in the manufacturing facility and this will represent considerable unaccounted for expenditure!

391

the residual ‘tramp-oil’ to float to the surface for removal by physical ‘skimming’ , • Cost-effectiveness – but what does this term mean? There was a time when the cost-effectiveness was simply judged in terms of the price per litre of the product concentrate. Fortunately, there are only few engineering companies who still take this view, with most recognising that there are many interrelated factors that contribute to cost-efficiency. Some of these factors might be the: dilution ratio; sump-life; material versatility; tool life; machined component quality; health and safety aspects; plus many others. Having identified the ‘ideal’ cutting fluid features, one must unfortunately face reality, as there is no such product that encompasses all of these desirable characteristics – at the optimum level in just one cutting fluid product. However, all cutting fluids are not equal and even apparently similar products may well perform in quite different ways! Therefore, it is for the machineshop supervisors/managers – in conjunction with other interested parties: purchasing; health and safety; unions; etc., to select a reputable supplier who is prepared to undertake the necessary survey and ‘troubleshooting’ exercise to recommend the best fluid(s) for a particular manufacturing environment. Today, there are many different types of cutting fluids available they can be classified according to widely varying criteria, although some unified system of terminology exists in various countries guidelines and Standards. This commonality of ‘language’ reflects both the chemical and technical requirements of the users. On the basis of the various countries publicised cutting fluid literature, the following classification is perhaps the most useful – from the user’s point of view. Broadly speaking, it was previously shown in Fig. 197, that cutting fluid groups are of two main types, either ‘oil-’ , or ‘aqueous-based’. The ‘aqueous’ cutting fluids can be divided into ‘emulsifiable’ and ‘water-soluble’ types. As has already been mentioned, the former ‘oil-based’ cutting fluids are supplied as ready-for-use products, while ‘aqueous’ types are normally found in the form of a concentrate, which must be mixed with water, prior to use. Once mixed with water, the ‘emulsifiable’ cutting fluids form an emulsion, conversely, the ‘soluble’ variety forms a solution. In both of these cases, the resultant cutting fluid product is termed: ‘watermixed’. In the following section, the various types of cutting fluids currently available will be briefly mentioned.

392

Chapter 8

8.4.1 Mineral Oil, Synthetic, or Semi-Synthetic Lubricant? Mineral Oil In order to manufacture cutting fluids the raw materials are naturally occurring oils, such as: mineral oils; animal and vegetable oils; or fats. Of these oils, the former mineral oils are probably most commonly utilised by the manufacturing industry. These mineral oils, in a similar fashion to naturally occurring oils, tend to be complex mixtures of widely varying compounds. Such compounds consist of carbon and hydrogen and as such, are usually referred to a ‘hydrocarbons’. In addition, they will contain: sulphur; nitrogen; plus various trace elements. So that the mineral oil can be separated out to form a ‘stock oil’ – with natural lubricating properties, thermal processes are employed by the fluid manufacturer. These partly-refined ‘stock-oils’ are still chemically complex mixtures of hydrocarbons, with widely varying characteristics. By way of an example of the diverse nature of ‘crude oil’ , it is a mixture of more than one thousand hydrocarbons, with different chemical structures. Such widely varying characteristics make it impossible to supply mineral oil to closely defined specifications, which limits its uses and performance as a cutting fluid. The complex structure of a cutting fluid made up entirely from naturally occurring oils, is schematically illustrated in Fig. 198a.

Synthetic Lubricants The use of Synthetic lubricants cannot be compared with those lubricants that are extracted from naturally occurring oils, since the properties of the latter are always an aggregate of the properties of their many different components, as such, cannot be exactly predicted. While the former synthetic lubricants are made from two types of raw material: 1. Mineral oil – normally from: polyalpha olefin and alkali aromatics, 2. Polybutenes. At present (i.e. from around the late 1980’s, until now), synthetic hydrocarbons predominate, as they are not derived from mineral oils, they have become of increased importance. In particular, they include

derivatives from ‘fractioning’ of plant oils. The most significant classes of compounds are the esters and polyglycols. These synthetic lubricants being a solution of chemicals, which usually contain: corrosion inhibitors; biocides; dyes; in water. Moreover, they may contain such additions as synthetic lubricity additives and wetting agents. Synthetic lubricants form transparent solutions and as a result, provide good visibility of the cutting operation. In use, synthetic fluids require special attention in their application, because they contain no mineral oil, they tend not to leave a corrosion-protective oily film on machine surfaces. As a result, it is essential to lubricate exposed machine tool surfaces carefully. In addition to this lack of protection, there may be some effect on certain paint finishes and even degradation of the machine’s seals, as a result of this synthetic fluid entering the machine tool’s lubrication system. Normally, these problems of practical usage, limit these synthetic lubricants in the main, to grinding operations.

Semi-Synthetic Lubricants Today, the use of Semi-synthetic lubricants, or ‘Microemulsions’ – as they are sometimes known, has become much more widespread, because of certain advantages they have over mineral-soluble oils. By increasing the ratio of: emulsifier-to-oil in the formulation, either by reducing the oil content, or by increasing the level of emulsifiers, the product takes on different characteristics from those of mineral-soluble oils. Due to this increased ‘ratio’ , the oil particles formed, are significantly smaller than those found with the mineral-soluble oil types (i.e. see Fig. 201a). Hence, these ‘micro-emulsions’ , visually appear to be translucent, or even transparent, owing to the fact that the oil particles are smaller than the wavelength of light (i.e. 4 µm in size, leaving the critical sub-µm particles still present in the atmosphere. The earliest chemical interventions to reduce misting were high-molecular-weight polymer additives, that act to stabilise MWF’s and thus suppress mist formation. With conventional petroleum-based fluids, polyisobutylene has been the preferred anti-mist additive. While, for aqueous-based cutting fluids, polyethylene oxide (PEO) has been utilised. Due to the susceptibility of PEO’s to shear degradation, repetitive additions of the PEO polymer are needed to maintain mist reduction. Today, a newer class of shearstable polymers has been developed to overcome the shear degradation as indicated by PEO’s. These latest polymer products have been derived from complex: 2-acrylamido-2methlypropane sulphonic acid monomers, hence, providing longer-term performance in continuously recirculating aqueous-based MWF systems. So, very high concentration cutting fluid mists will over a short period of time cause: ‘smarting’ of the eyes; irritation of exposed skin; result in slight irrita-

Cutting Fluids

tion of the mouth and throat; by inhalation, will irritate the lungs; by ingestion, of the stomach – it may promote nausea; and affect other internal organs. If exposed to toxic mists over a long period of time, this could cause lasting damage to both external and internal bodily-parts, with at the extreme condition, promoting the growth of malignant tumours. In order to restrict misting and minimise operator health risks, then special-purpose filtering systems have been developed, which will be briefly reviewed below. The conventional mist-collection technology, such as: filters; rotating drums; or cyclones; will collect particles of >1 µm in diameter, but cannot cope with smaller sub-µm particles. Further, it has been reported that fibrous filters once they are wet, lose efficiency over time – see Fig. 212. Therefore, the optimum manner of removing sub-µm mists are by fitting one of the following: High-efficiency Particulate Air filters (HEPA); Electrostatic Precipitators (ESP’s); or Fibre-bed systems. Probably the two best systems for removal of sub-µm mist particles are the HEPA and ESP systems. Each one has its disadvantages, with HEPA filters being expensive and become clogged, thereby losing efficiency. So, when disposable filter replacements are needed this hidden replacements cost, will result in both costly maintenance and disposal. While, ESP’s need frequent maintenance and cleaning, thus representing a continuous on-going cost burden. Meanwhile, Fibre-bed systems offer high efficiency in mist collection, but with ease of maintenance, although they are larger requiring more electrical power to operate them.

Vegetable Oil-Based MWF’s Driven by the health and safety concerns of both workers and manufacturers alike, vegetable oil-based MWF’s have been developed, to substitute for the same machining operations as either the mineral-, or petroleum-based fluids, currently undertake. It has been reported that compared with mineral oil-based cutting fluids, the alternative vegetable-based MWF’s, enhance cutting performance by extending tool life while improving machined surface texture, with the additional benefit of being an environmentally-friendly MWF. In particular, Soybean oils have shown considerable promise as a practical alternative to ‘traditional’ MWF’s, where they have improved component surface texture and reduced tool chatter. One of the principle reasons for these surface texture and machining improvements, is that the vegetable oil-based MWF’s have enhanced lubricity, coupled with a slight

423

‘polar-charge’ – which acts to attract the vegetable oil molecules to the metallic surface being tenacious enough to resist any subsequent wipe-off. The opposite is true for a mineral-based oil, where there is no molecular charge, so offers little improvement in lubricity. Mineral-based MWF’s are just straight hydrocarbon, while their vegetable oil counterparts contain oxygen, which is tenaciously-attracted to the sterile elevated temperature of the recently-machined workpiece’s metallic surface, thus it bonds more strongly – acting as a result as a better lubricant. Yet another performance benefit of utilising vegetable-based oils over their mineral-based equivalents, is that they have a higher ‘flash-point’ 34, which reduces both the tendency for smoke formation and fire-risk. Yet another reason for selecting a vegetable-based MWF over its mineral-based counterpart, is that it has a high natural viscosity35. Hence, as the machining temperature increases, the viscosity of the vegetable oil drops more slowly than for that of a mineral oil. Conversely, as the temperature falls, the vegetable oil remains more fluid than its counterpart mineral oil. Thus, facilitating more efficient and quicker drainage from both the swarf and workpiece. The high viscosity index36 of vegetable oil ensures that it provides more lubricity-stability, across the operating temperature range being found during a range of machining operations. High viscosity allows vegetable oils to be used as a slideway lubricant and for gear lubrication in gearboxes, acting as a so-called: ‘multi-functional fluid’ (i.e. see Section 8.9). Along with the above stated benefits, there is also a down-side to vegetable-based fluid applications, the limitations are that they lack sufficient oxidative sta-

34 ‘Flash-point’ of oils, is the instantaneous ignition of the oil at a specific temperature, without the aid of a flame. So, in the case of a Soybean oil it has a flash-point of 232°C, while a typical mineral oil has a flash-point of just 113°C. 35 ‘Viscosity’ , can be defined* as: ‘The resistance of a fluid to shear force.’ Therefore, the shear force per unit area is a constant times the velocity gradient, with the constant being the coefficient of viscosity. SI units are: Newton-seconds per square metre (Ns m–2), denoted by the Greek symbol: ‘µ’. [Source: Carvill, 1999] *While another definition for a fluid’s viscosity is: ‘The bulk property of a fluid, semi-fluid, or semi-solid substance that causes it to resist flow.’ [Kalpakjian, 1984] 36 ‘Viscosity index’ , can be defined as: ‘A measure of a fluid’s change of viscosity with temperature: the higher the index, the smaller the relative change in viscosity.’ [Kalpakjian, 1984]

424

Chapter 8

bility for many machining applications. Thus, a low oxidative stability means that the oil will oxidise quite quickly during use, becoming thick as it polymerises to a plastic-like consistency. Once the oil has become too thick, or even too thin for that matter, the cutting tool’s edge(s) will quickly wear-out. Vegetable oils become oxidised and as a result, will chemically change,

along with their viscosity and lubricating abilities. There is some concern among users of vegetable-based cutting fluids, that this oil reacts with the environment (i.e. oxygen and metals), thus breaking-down, which is not the case for petroleum-based products. Both of these fluid products oxidise with heat, but vegetable oils are more susceptible to oxidation. While another

. Figure 212. At the filter some droplets and volatiles are removed from the atmosphere, but the remainder pass through and are re-entrained. Other particulates are ‘indefinitely’ retained, but with time reduce filter efficiency. Optimum filters

maximise droplet removal, while minimising evaporation and re-entrainment – at a reasonable pressure-drop. [Source: Raynor P. & Leith, D. – Univ. of North Carolina, 2003]

Cutting Fluids

drawback to utilising vegetable-based oils, are its lack of hydrolytic stability37. Typically, when making an emulsion; obviously oil and water are present; so if oxygen and some form of alkaline component is at hand, it may cause certain ester linkages within the vegetable oil to break down. These broken-down components act in a different manner to that of the original vegetable oil, thereby affecting its cutting fluid performance. Conversely, mineral-based cutting oils are resistant to hydrolytic reactions. Vegetable oils can support microbial growth more readily than the equivalent mineralbased cutting fluids. Although this vegetable oil’s biodegradability is ideal for subsequent waste treatment, conversely, when this product is ‘festering’ in a machine’s sump, it becomes both smelly and sour, via its bactericide and fungicide reactions. Finally, for many companies, probably the biggest limitation in changing over to vegetable-based products in machining operations is its purchase cost. For example, canola oil, can cost up to 300% more than its equivalent petroleumbased product and to compound the financial problem still further, costly ingredients are necessary to control oxidation and enhance its biological stability – considerably adding to the finished product’s cost.

8.12 Fluid Machining Strategies: Dry; Near-Dry; or Wet So far, this chapter has been principally concerned with all aspects of flood/wet coolant applications to the overall machining process. Several other complementary cutting strategies can be adopted, these include: dry; near-dry; together with wet machining; thus, in the following sections a discussion of these important issues and concerns will be briefly mentioned.

37 ‘Hydrolytic stability’: ester molecules consist jointly of condensed fatty acids and alcohols; so the vegetable oils will naturally exist as esters – often termed ‘triglycerides’ , these being a condensation of fatty acid, plus glycerine. Under the right conditions, the triglyceride can split and revert back to a fatty acid and glycerine, which acts differently from that of the original ester. In the case of mineral-based oils, they do not contain these ester linkages and as such, will not break down, nor ‘hydrolise’.

425

8.12.1 Wet- and Dry-Machining – the Issues and Concerns In the past twenty-five years the cost of cutting fluids has risen from just 3% of the overall cost of the machining process, to that of >15% of a production shop’s cost. Cutting fluids and especially ones that are oil-based products have become something of a liability of late, this is due in the main, to many countries ‘Environmental Protection Agencies’ , strictly regulating their ensuing disposal – at the end of their natural life. In many countries ‘spent’ cutting fluids have been re-classified as either ‘toxic-’ , or ‘hazardous-waste’ , moreover, if they have been found to have machined certain alloyed and exotic material workpieces, they are under even harsher disposal regulations. Due to the increasing popularity today of high-speed machining (HSM) – more will be said on this subject in the following chapter – coupled to increased cutting data and the application of coolants via high-pressure systems, these factors have significantly contributed to the creation of air-borne mists within the workshop environment (i.e. see Fig. 212). Such coolant mists now have even stricter permissible exposure levels (PEL’s) imposed in the working environment, to regulate and control these air-borne particulates, thereby minimising workers health risks. Thus, the cost of: fluid maintenance; record-keeping; with strict compliance to current and proposed regulations, have rapidly increased the overall price of cutting fluids. In many manufacturing companies involved in a significant amount of machining operations, they are considering the strategy of cutting dry, to overcome the cutting fluid-based costs and disposal concerns during and after their subsequent usage. For many companies involved in significant workpiece machining operations, they are unsure if they could cut all their components ‘dry’. Furthermore, they are under the impression that to achieve higher cutting data and ‘hard-part’ machining, then cutting fluids are essential in achieving these objectives. Moreover, many companies also believe that the cost of changing from a ‘wet-’ to ‘dry-machining’ operations would be prohibitively high. Neither of these impressions are true. So, by machining ‘dry’ it can be considered as a standard operational procedure for most metal-cutting operations, including: turning, drilling and milling operations (i.e see Figs. 39, 49 and 168a, respectively). Moreover, it is not only possible to ‘hard-part’: turn (Fig. 15) and bore (Fig. 65b); or mill (Fig. 172); etc.; but these can now be classified as highly-profitable ‘dry-machining’ activities.

426

Chapter 8

Probably the chief obstacle to dry-machining acceptance, is that conventional wisdom dictates that MWF’s are vital in attaining acceptable machined finishes and will considerably extend the tooling’s life. In many circumstances these are valid points, but with some of the advanced cemented carbide grades and high-technology coatings, such tooling can be operated at higher cutting data than was previously the case and, cope with their elevated machining temperatures. In fact, if interrupted cutting occurs, the hotter the cutting zone becomes, the more unsuitable will be the application of a cutting fluid – as the thermal shock38 becomes greater with a ‘wet-machining’ strategy. Present tool coating technologies are vital to drymachining applications, as they both control the temperature fluctuations, while restricting heat transfer from the cutting vicinity to the insert, or tool. Multiple coatings act as a heat barrier because they offer a lower thermal conductivity to that of the tool’s substrate and the workpiece material. Thus, coated inserts/tooling absorb less heat and as a result, can tolerate higher cutting temperatures, allowing more aggressive cutting data, whilst not debilitating the tool’s life. Coating thickness is also important, as the thinner the overall coatings, the better they can withstand temperature fluctuations, that might otherwise arise, if thicker coatings were utilised. The main reason for this improved thermal shock performance of thinner, rather than thicker tool coatings, is that a thinner coat is less likely to incur the same stresses, hence, they are less susceptible to cracking as a result. So, by running thin coatings in ‘dry-machining’ operations, normally extends tool edge life by up to 40%, over thicker coatings39.

38 ‘Thermal shock/fatigue’ , the cyclical nature of both rapid heating followed by immediate cooling – in for example facemilling (i.e see Fig. 213 – top), or when interrupted turning (e.g. when eccentric turning, or OD/ID machining with either splines and keyways present), promotes potential tool edge fracturing – resulting from the cyclic thermal stresses and increased temperature gradients, being exacerbated by the application of a cutting fluid. 39 ‘Thin coats-v-thick coats’ , the former coating offers longer life than the latter coating process. Today, it is normal to utilise the coating process of: physical vapour deposition (PVD) as this type of coating is thinner and will adhere/bond more strongly, than the alternative chemical vapour deposition process. For example, a TiAlN PVD coated insert/tool can have a hardness of 3,500 Hv, withstanding cutting temperatures up to 800°C.

‘Dry-machining’ – some Factors for Consideration

• Adopting a ‘dry machining’ strategy will only make • •

• • • • •

sense, if all the cutting processes in the part’s manu facture can be performed without coolant, Only by utilising specialised cutting tool geometries, can ‘dry-machining’ be possible and effective, Tooling typically having special hard multi-layered, or diamond-like coatings, etc., to isolate heat and create minimal thermal conduction across the tool/ chip interface, Employing cutting tool materials producing sharp edge geometries – to reduce heat, For drilling operations, utilise ‘soft-glide’ coatings – for lubrication, with the necessary and appropriate efficient chip transportation geometries, Speedy and efficient removal of both chips and associated steam – by suction – are important factors in ‘dry-machining’ , Utilise new machining concepts, plus the latest fully-enclosed machine tools – whenever possible, Employ faster, rather than slower cutting data, to allow the majority of heat to be confined to the evacuated chips.

8.12.2 Near-Dry Machining The strategy of ‘near-dry’ machining is not a new concept, it has been in existence for more than 50 years. However, this machining and lubricating approach is still not a universal practice, which is surprising when one considers the real benefits that accrue from the practice over its ‘wet-machining’ counterpart. As its name implies, in ‘near-dry’ machining little lubricant is used – normally vegetable-based oils, meaning that both cutting fluid treatment and its disposal are eliminated. Further, instigating a ‘near-dry’ machining strategy means that there are fewer worker health risks from resultant mists, which might otherwise create: re-

NB From a metallurgical/materials science viewpoint, the: TiAlN – PVD tool coating can attribute its superior mechanical/physical properties to an amorphous aluminium-oxide film that forms at the tool/chip interface, as some of the aluminium of the coating surface oxidises at these elevated machining temperatures. While, even more exotic multiple-type diamond-like coatings can be applied and their like, which offer even greater cutting performance – in certain machining circumstances, when applied to the tool’s cutting edge(s).

Cutting Fluids

spiratory problems: skin irritations; etc. The ‘near-dry’ cutting approach can be exploited across a wide range of either ferrous, or non-ferrous workpiece metals. Most machine tools are equipped with the capabi lity of supplying flood coolant to the cutting process, together with ‘through-coolant’ tooling systems, meaning that the cost to reconfigure for that of a ‘near-dry’ technology is not prohibitive. Assuming the worsecase scenario of requiring a through-coolant tooling system, then probably just over $5,000 at today’s prices should prove sufficient capital to complete the task. Some re-tooling to complement the ‘near-dry’ machining production techniques may be necessary, allowing the precise application of lubricant to the cutting edge(s). Further, the user must consider a method for efficient chip removal from the cutting area. Usually, with external ‘near-dry’ cutting operations, the lubricant is transported within the media of a compressed air application, via the correct-sized aperture nozzle – pointed toward the cutting zone. Control of the volume of lubricant delivery to the tool and workpiece area is critical, with the common misconception being that more lubricant is better! The optimum arrangement for ‘near-dry’ lubricant application, is when the minimum of over-spray and resultant misting does not occur. With external ‘near-dry’ operations, dispensing systems usually consists of reservoir metering pumps and valves, being mounted on the machine tool’s exterior – at some convenient location. While the nozzles are strategically-mounted so that they can easily be directly aimed at the tool’s cutting edge(s). Normally, the nozzles are a manufactured from either copper, or plastic and ‘snap-together’ – being much smaller in size than their ‘wet-machining’ counterparts. For internal machining operations, having tooling with ‘through-the-nose’ delivery, the lubricant is mixed with compressed air prior to delivery to the cutting zone. The admixing of compressed air and lubricant keeps the lubricant in suspension, with these oil particles being broken-down into minute particles prior to being fed into the compressed air jet stream – on their way to the tooling. For ‘conventional’ flood coolant delivery the systems, the coolant channels are filled with cutting fluid, which inevitably finds its way to the cutting zone. If however, in a ‘near-dry’ machining configuration, a heavy mist of lubricating oil floats through the compressed air, attempting to negotiate all of the twists and turns on its way to the cutting zone, this may present a potential lubrication clogging/starvation prob-

427

lem. Hence, for a successful ‘near-dry’ delivery system, the lubricant channels need to be smooth and even, with direct flows from the coolant pump to the cutting zone. A basic misapprehension by some machine tool designers and manufacturers, is that copious volumes of flood coolant are necessary to remove large quantities of swarf. In fact, just the opposite can occur, as wet chips will not only pack tightly together but have a surface tension property to them, tending to make them adhere to machine tool surfaces (i.e. see Footnote 29, ‘Lang’s chip-packing ratio’ in Chapter 2). This is not generally the case for ‘near-dry’ lubrication, as the chips here, have a thin layer of non-oxidising lubricant surrounding them and with their evacuation velocity – after being machined, coupled to gravitational effects, means that they will fall to the bottom of the swarf tray, or into the chip conveyor. It is good working practice to use the external air-only supply’s blow-off nozzles to clear away chips form the cutting area40, however, it is not recommended to use the oil/ mist to achieve chip clearance, as it will simply blow the lubricant straight past the cutting edge(s), while probably creating an unwanted oil-misting problem. It is possible to incorporate both ‘wet-’ and ‘near-dry’ lubrication systems onto the same machine tool. It has been reported that for external/internal work the change-over from one system say, from ‘wet-’ , to the other – ‘near-dry’ , takes about 3 minutes to complete. For ‘near-dry’ machining to be successful, it depends upon various factors, including: workpiece material to be machined; tool geometry and its coating(s); speeds and feeds selected; plus other important factors. If applied correctly, ‘near-dry’ machining has significant direct and indirect benefits to the machining process as a whole.

Economics of: ‘Dry-’; ‘Near-Dry’; and ‘ Wet-Machining’. For any tool and workpiece lubrication strategy to operate effectively, a range of cost factors need to be considered, regardless of the method of machining

40 ‘Chilled compressed air’ , has been successfully utilised in the past for not only removing chips from the cutting vicinity, but on certain materials, the continuous application of chilled compressed air acts simply as a form of ‘basic lubricant’ for the cutting process in hand.

428

Chapter 8

undertaken. In Fig. 213, a table has been constructed to show the relative merits of the three machining strategies previously discussed, namely: ‘dry-’; ‘neardry’; or ‘wet-machining’. The cost component for each of these lubrication strategies has been broken down into its relevant parts, with some of them not being ap-

plicable to every lubrication application. If one ignores the individual cumulative factor in the overall cost and simply looks at the ‘bottom-line’ , namely, the total relative costs for each process, then a clear message is being given here! Explicitly, that ‘wet-machining’- in certain cases, when compared to ‘dry-machining’ is

. Figure 213. Indicates the comparative costs for utilising either: ‘dry-’, ‘near-dry-’ or ‘wet-machining’ strategies

Cutting Fluids

>330% more expensive overall, this being a good reason to look carefully at employing ‘dry-machining’ techniques – when applicable! In Appendix 14, a MWF ‘trouble-shooting guide’ has been included, to help establish the relative causes and remedies for certain fluid-related problems – as they arise.

References Journals and Conference Papers Antoun, G.S. The Pressure’s On to Improve Drilling [HighPressure and Volume Coolant Supply]. Cutting Tool Eng’g., 59–68, Feb., 1999. Batzer, S. and Sutherland, J. The Dry Cure for Coolant Ills. Cutting Tool Eng’g., 34–44, June, 1998. Benes, J. Life Support – for Cutting Fluids, American Machinist, 42–46, Feb., 2007. Benes, J. Engineering Metalworking Fluids. American Machinist, 48–51, March, 2007. Boyles, C.M. Managed Fluid Care. Cutting Tool Eng’g., 40–45, Jan., 2002. Brutto, P. Water Hazard [Leaching Cobalt Binder from WC Tooling]. Cutting Tool Eng’g., 50–56, Dec., 1996. Da Silva, M.B., Machado, A.R. Wallbank, J. On the Mechanism of Lubrication in Single Point Cutting. Int. Conf. on: Behaviour of Materials in Machining, Straffordupon-Avon (England), Pub. by: IOM Communications Ltd, 79–89, Nov. 1998. Davidian, S. Setting Standards [Metalworking Fluid Stand ards]. Cutting Tool Eng’g., 52–55, Sept., 2001. DeChiffre, L. Testing the Overall Performance of Cutting Fluids. Lubric. Eng’g., Vol. 34 (5), 244–251, 1978. DeChiffre, L. Mechanical Testing and Selection of Cutting Fluids. Lubric. Eng’g., Vol. 36 (1), 33–39, 1980. Deodhar, J. Soluble Cutting Fluids: Chemistry, Management and Control. Int. Conf. on Industrial Tooling Southampton, UK, Shirley Press Ltd., 151–158, Sept. 1995. Dunlap, C. Should you try Dry? [Dry Machining]. Cutting Tool Eng’g., 22–33, Feb., 1997. Eade, R. Are Metalworking Fluids a Threat to Health? Cutting Tool Eng’g., 80–86, Aug., 1998. El Baradie, M.A. Cutting Fluids: Characterisation Part I. Int. Conf. On Adv. In Matls and Process. Technol., Vol. III, Dublin City Univ. Press, 1999–2010, Aug., 1993. El Baradie, M.A. Cutting Fluids: Recycling and Clean Machining Part II. Int. Conf. On Adv. In Matls and Process. Technol., Vol. III, Dublin City Univ. Press, 2011–2019, Aug., 1993. Excell, M. Well Oiled Machines [Machine Tool Lubrication]. Metalworking Production, 53–56, Dec., 1995.

429

Folz, G. Treat MWFs [Metal Working Fluids] Right. Cutting Tool Eng’g., 36–42, Oct., 2003. Graham, D. Dry Out [Dry Machining]. Cutting Tool Eng’g., 56–65, Mar., 2000. Gugger, M. Putting Fluids to the Test [Testing Methods]. Cutting Tool Eng’g., 54–62, Aug., 1999. Howard, R. Types of Cutting Fluids used on Turning/Machining Centres. Int. Conf. on Industrial Tooling (Southampton, UK), Shirley Press, 140–150, Sept. 1995. King, N. Ah, the Smell of It [Microbiological Colnat Effects]. Manufact. Engr., 7–9, Dec., 1991/Jan. 1992. Krahenbuhl, U. Lightening [Coolant Lubricity]. Cutting Tool Eng’g., 28–34, Sept., 2002. Leep, H.R. Investigation of Synthetic Cutting Fluids in Drilling, Turning and Milling Processes. Lubric. Eng’g., Vol. 37 (12), 715–721, 1981. Manfreda, J. and Elenteny, D. Cool Savings, Cutting Tool Eng’g., 42–51, Feb., 2006. Phillips, D. Under Pressure [High-Pressure Coolant Delivery]. Cutting Tool Eng’g., 48–53, Jan., 2000. Phillips, D. Dry Run [Dry Drilling]. Cutting Tool Eng’g., 34–40, Feb., 2000. MCCabe, J. Dry Holes – Dry Drilling Study. Cutting Tool Eng’g., 40–50, Feb., 2002. Protch, O. Fluid Cutting. Cutting Tool Eng’g., 40–43, Dec., 2001. Rahman, M., Senthil Kumar, A. and Salam, M.U. Experimental Evaluation on the Effect of Minimal Quantities of Lubricant in Milling. Int. J. of Mach. Tools and Manufact., Vol. 42, 539–547, 2002. Shanley, A. Mist Collection: The Pressure is On. Cutting Tool Eng’g., 46–51, June, 2003. Sluhan, W. Don’t Recycle – Keep your Coolant. Cutting Tool Eng’g., 53–55, Oct., 1993. Sluhan, W. Selecting Coolants: Why and How. Cutting Tool Eng’g., 56–65, Oct., 1995. Sluhan, W. Does Your Machine Like Its Coolant? [Watermiscible Cutting Fluid Concerns]. Cutting Tool Eng’g., 62–69, April, 1996. Sluhan, W. Humans and Cutting Fluids [Health Issues]. Cutting Tool Eng’g., 88–92, June, 1996. Sluhan, W. Pure Water – Isn’t Hard to Find [Hard Water Problems]. Cutting Tool Eng’g., 28–32, Dec., 1996. Sluhan, W. Bubble Trouble [Foaming Effects]. Cutting Tool Eng’g., 24–28, April, 1997. Sluhan, W. Itching for a Solution [Causes of Dermatitis]. Cutting Tool Eng’g., 36–43, Dec., 1997. Sluhan, W. The Good, the Bad and the Smelly [Bacterial Effects]. Cutting Tool Eng’g., 24–32, Mar., 1998. Smith, A. Lagerberg, S. Dahlam, P. and Kaminski, J. Highpressure [Coolant] Jet-assisted Turning, Int. Conf. on Industrial Tooling Southampton, UK, Molyneux Press Ltd., 62–71, Sept. 1999.

430

Chapter 8

Smith, G.T. Managing and Controlling Cutting Fluids in a Flexible Manufacturing Environment. Proc. of Managing Integrated Manufacturing, Keele Univ. Press, Vol.1, 491–503, 1993. Smith, G.T. Wet and Dry Machining – Understanding the Key Issues. Metalworking Production, 22–24, Oct., 2000. Smith, P.L. Coolants and Cancer: Fact or Fiction? American Machinist, 46–50, Dec., 1996. Threadgill, J. Animal, Vegetable, or Mineral [Sump-cleaning Strategies]. Cutting Tool Eng’g., 30–35, Sept., 1993. Walz, T. Fine Ideas [Ultra-filtration]. Cutting Tool Eng’g., 48–52, May, 2002. Woods, S. Going Green [Vegetable Oils], Cutting Tool Eng’g., 48–51, Feb., 2005. Yang, C.C. The Effects of Water Hardness on the Lubricity of a Semi-synthetic Cutting Fluid. DeChiffre, L. Testing the Overall Performance of Cutting Fluids. Lubric. Eng’g., Vol. 35 (3), 133–136, 1979. Woods, S. Going Green [Alternative: Vegetable-based MWF’s]. Cutting Tool Eng’g., 48–51, Feb., 2005. Woods, S. Nearly Dry [Nearly-dry Machining]. Cutting Tool Eng’g., 58–64, Mar., 2006. Books, Booklets and Guides Byers, J.P. Metalworking Fluids. Marcel Dekkar (NY), 1994. Carvill, J. Mechanical Engineer’s Data Handbook. Butterworth-Heinemann Pub., 1997. Hartley, D. Cutting Fluids – Their Care and Control. Reprint from: Maintenance Eng’g., May, 1979. Mang, T. Wassemischbare Kühlschmierstoffe für die Zerspanung [Water-miscible Metalworking Fluids]. Vol. 61, Kontakt and Studium, Tribotechnik. Expert Verlag, Grafenau, 1980.

Kronenberg, M. Machining Science and Application. Pergamon Press (NY/London), 1966. Metalworking Fluids. Pub. by: Verlag/Cincinnati Milacron, 1991. Metal Working Fluids Guide. Kuwait Petroleum Inter national Lubricants Pub. Moubray, J. Reliability-centred Maintenance, ButterworthHeinemann , 1996. Nachtman, E.S. and Kalpakjian, S. Lubricants and Lubrication in Metalworking Operations. Marcel Dekkar (NY), 1985. Robertson, W.S. Lubrication in Practice. Esso Petrol. Co. Ltd.,/Macmillan Educ. Ltd., 1987. Schey, J.A. Tribology in Metalworking – Friction, Wear and Lubrication. ASM Pub. (Ohio, USA), 1983. Sluhan, C. Cutting and Grinding Fluids: Selection and Application. Soc. Of Manufact. Engrs., (Dearborn, Mich., USA), 1992. Smith, G.T. CNC Machining Technology: Vol. 2: Cutting, Fluids and Workholding Technologies. Springer-Verlag, 1993. Smith, G.T. Condition Monitoring of Machine Tools (Chapter 9). In: Handbook of Condition Monitoring, 1st Ed. (Edited by: Rao, B.K.N.), Elsevier Advanced Technology, 1996. Talking About – Cutting Fluids. Castrol (U.K.) Ltd, Pub. No.: IND 71b/9/91, Sept., 1991. Talking About – Health and Safety (3rd Ed.). Castrol (UK) Ltd., 1990. Yust, M.S. and Becket, G.J.P. Microbiology Theory and Practice in Metalworking Fluids. Reprint from: Ind. Lubric. & Trib., Nov./Dec. 1980.

9

Machining and Monitoring Strategies ‘Inque brevi spatio mutander saecla animantum Et quasi cursores vitai lampada tradunt.’ TRANSLATION: ‘The generations of living things pass in a short time and like runners hand on the torch of life.’

Titus LUCRETIUS Carus

(94 – 55 BC) [In: On the Nature of the Universe, II]

432

Chapter 9

9.1 High Speed Machining (HSM) Background to HSM Possibly the first work of note on HSM was that of Salomon who ran a series of applied experiments from the period 1924 to 1931, when a German Patent was granted for this work. The Patent was founded upon a series of cutting speed curves plotted against machining temperatures for a range of materials (Fig. 214). In these tests Salomon achieved peripheral cutter speeds of 16,500 m min–1, utilising either fly-cutters – for chip morphology data, or helical milling cutters – notably when cutting aluminium. Salomon contended that the cutting temperature ‘peaked’ at a specific cutting speed, which he termed the ‘Supercritical speed’, further, when the speed was increased still faster the temperature he noted, decreased. Moreover, either side of this ‘supercritical speed’ zone, it was suggested that ‘unworkable regimes’ occurred where the HSS cutters could not withstand the severe forces and temperatures generated. As mentioned, when the cutting speed increased beyond this ‘zone’ the temperature reduced to those expected by ‘normal data’ cutting conditions, permitting practical cutting operations to be carried out. The problem being with this early HSM work is that any theoretical rationale is not available and the experimental procedures are somewhat unclear, but Salomon can be considered the founder of ultra-high speed machining (UHSM) – taking cutting

‘Supercritical speed’ mentioned by Salomon when HSM, has never been truly substantiated. The practical data was based upon chip morphology* experiments with ‘fly-cut‘ climb-milling of: non-ferrous alloys; ‘soft’ aluminium; red cast brass; the latter being ‘un-machinable’ with HSS cutters between 60 to 330 m min–1 (i.e. see Fig. 214). As the machining parameters and experimental details only exist as a partial fragment of the original German work. This fragmented machining information, is because during the Second World War, unfortunately, the vital details were lost, moreover, none of the participants in this work also surviving to explain how this applied research data was collated. *Chip morphology was achieved by situating a heavily waxcoated board, this being strategically positioned to allow the fly-cut chips to stick to the board – during the peripheral climb-milling experimental operation, ready for future analyses.

data beyond that considered in the so-called ‘Taylorequations’. Effectively it can be established that there were four distinct periods of advancement in the field of UHSM, with the first one being from the early 1920’s to the late 1950’s, with each period during this time and there after, being separated by a significant event. Obviously during the first period, the work instigated by Salomon (i.e. in the 1920’s – alluded to above), was followed by the originally-commissioned United States Air Force (USAF) major research contract, being from 1958 to 1961. Previous to this contract award, little in the way of UHSM work in the preceding decades had occurred, apart from that in the States by Vaughn (1958). Vaughn, shortly became aware of the Salomon Patent, acquiring limited information on this work through the United States Consul in Berlin. Vaughn’s (Lockheed) group were also familiar with the many technical references concerning the ‘art’ of oil well tube perforation utilising perforated cutters, these ‘cutting actions’ being employed to perforate oil well casings at explosive speeds. Such background work, now meant that Vaughn had ‘set the scene’ for the second period of UHSM development. The USAF Materials Laboratory commissioned study – mentioned above, which was awarded to Lockheed (i.e. Vaughn’s group), with the objective of evaluating the ‘machining response’ for a selected range of high-strength materials to that of (surface) cutting speeds of up to 152,400 m min–1. The ‘principal aim’ of this ‘USAF-commissioned work’ was to increase producibility, while improving both the quality and efficiency of the fabrication of aircraft/missile com ponents. Vaughn’s experimental apparatus included the use of modern-day ex-military cannons, which were positioned on ‘railed-sleds’ to minimise subsequent recoil upon firing, while obtaining the desired ballistic exit velocity of the ‘material-slug’ for the cutting speeds. Unfortunately, these ‘machining results’

‘Ballistic cutting speeds’ , were obtained as the cannon fired the projectile (i.e. a specific material-slug) at ultra-high speed out of the cannon. At the projectile’s exit from the cannon’s barrel, a very robust cutting tool arrangement was situated and held in an accurate position to take a linear cut along the exiting slug’s periphery. The evidence of this ballistic cutting action was then recorded by very high-speed photographic equipment – this being both electronically-timed and strategically positioned, for visual dynamic recording and future reference and analysis.

Machining and Monitoring Strategies

433

. Figure 214. Graphical relationship of high speed machining of metals – according to Salomon’s machining trials. [Source: Salomon, circa 1920’s–30’s]

434

Chapter 9

did not indicate how such ballistic speeds might be incorporated into a production application, also, an analytical model of this high-rate cutting phenomenon was not developed. In the 1960’s and early 1970’s some consolidation of important machining data occurred, with notable work on UHSM cutting mechanisms, etc., that occurred in various countries being led principally by the USA in work from: Coldwell and Quackenbush (1962), plus Recht (1964); Okushima et al. (1966) and Tanaka et al. (1967) in Japan; Fenton and Oxley (1967) in the United Kingdom; and Arndt (1972) in Australia. However, although there was a general increase global research activity during this period, it was somewhat disparate and mainly of a fundamental, rather than applied nature. The third period of HSM development was instigated in the mid-to-late 1970’s by the United States Navy, in conjunction with the Lockheed Missiles and Space Company, Inc., who contracted a series of machinability studies on marine propellers. Here, the Lockheed group headed-up by King, were mainly concerned with the feasibility of utilising HSM in a production environment, primarily for aerospace aluminium alloys, then later, working on nickel-aluminium-bronze. King’s team at Lockheed, demonstrated that it was economically feasible to introduce ‘highspeed-machining’ procedures into the production environment, thereby realising the significant improvements in productivity with this HSM application. Such applied research work, promoted significant activity and interest in this HSM field and, it soon became clear that attention needed to be focussed for all of the subsequent small and diverse research programmes.

Finally, in these formative years of experimental work into HSM, the fourth period of development began in 1979, when the USAF awarded a contract to the General Electric Company, in this instance, to provide a scientific basis for faster metal removal via HSM and Laser-assisted machining. A further contract by the USAF in 1980, was also awarded to ‘General Electric’– the group also being headed by Flom (1980). With this new HSM contract, also being granted to Flom’s group, with the objective to evaluate the production implications of the previous contract. Both of these contracts being supported by a consortium of industrial companies and selected universities in the USA – initiating the fourth HSM period of development. At around this time (i.e. in the late 1970’s), the introduction of computer numerical control (CNC) systems occurred and as a result, they were immediately being fitted to a new range of machine tools, significantly enhancing both their usability and programming capabilities, acting as a catalyst in the development of HSM strategies. As these CNC controllers became more sophisticated and processing speeds substantially increased, this meant that the potential for HSM could now be fully realised. In recent years, HSM machine tools are just about everywhere in machine shops around the world, where ever there is a need for highly productive part production with very fast cycle times. Obviously, on HSM machines as the spindle speeds have increased in association with both tool and their respective workpiece materials (i.e. see Fig. 215), this has meant that there are now considerable design implications on these machine tools, these topics will now be succinctly discussed.

9.1.1 HSM Machine Tool Design Considerations

‘Marine propeller manufacture’ , whether from a wroughtsolid material, or finishing-off a casting, is probably the most difficult and complex multi-axes milling operation that can be undertaken – due to the fact that the propeller surfaces to be machined are continuously changing their geometry as they are essentially parabolically-curved. Invariably, the part geometry is typified by say, the NACA/NASA Standard 16-021 ‘aerofoil cross-sectional profile’ , which for high-performance propellers are exacerbated by normally having both considerable rake and skew to each blade – creating an exceedingly complex geometry and fillet where the boss and blades intersect*. * See: Smith and Booth (1993) paper – in the references, which goes some way to explain the propeller manufacturing subject, and for more detailed multi-axes machining and for subsequent machined propeller measurement information.

Prior to a discussion concerning machine tool design factors that must be addressed, before to fully-implementing this HSM technology, one might ask the question: ‘Why do we need to rotate cutters at such high spindle speeds?’ There are a number of advantages that can accrue from adopting such a progressive produc-

USAF contract to the General Electric Company in 1979, was: F 33615-79-C-5119 – for a feasibility study into the fast metal removal operations by HSM and Laser-assisted machining.


435

. Figure 215. The chronological development of cutting tool material introductions, which had an influence on high-speed cutting trends. [Courtesy of Yamazaki Mazak Corporation]

tion machining strategy (i.e. see Fig. 216) and, they can be succinctly summarised as follows: • Direct benefits – improved machining efficiency, – reduction in cutting tool varieties, – reduction in distortion of workpiece, – effectiveness of swarf removal. • Indirect benefits – quality of finish improved, – increased cutter life, – reduced changes in material properties, – capability of machining thin walls/sections. These production improvements are by no means all that occur, as invariably, due to the superior machined surface texture, the final part surface may not need to be deburred – a significant real saving on complex

component geometries. Moreover, by employing an HSM strategy, more simple fixturing can be utilised, as the actual tool forces are significantly lessened. It is an established fact that with higher rotational cutter speeds the resulting cutting forces and tool push-off are considerably reduced. In order to benefit from these improved cutting practices, the machine tool’s axes must have both faster acceleration and deceleration – see Fig. 221, more will be mentioned concerning these important dynamic aspects of a machine tool’s operational performance shortly. Many of today’s conventional and HSM machine tools, are based upon a modular design concepts (Fig. 217a). This modular design philosophy, allows the machine tool builder the opportunity to standardise certain features over a range of machine tools. Such practice benefits the manufacturer and consumer alike, by reducing design

436

Chapter 9

. Figure 216. Diagram illustrating the main benefits to be gained from adopting a high-speed machining strategy. [Source: Smith, CNC Machining Technology, 1993]

and development costs, while minimising the customer’s purchasing costs, yet still allowing more attention to be given to the detailed design of each ‘module’ in the machine. So, an identical column, or table may be common to a variety of machines and this trend can often be seen across a whole product range of machine tools. Not only are the major castings, or large fabrications manufactured by employing modular concepts, but this design philosophy also allows any other smaller components to be standardised and fitted accordingly. Such as: the recirculating ballscrews;

servo-motors; linear scales – if fitted; etc.; plus other standardised items to be built into the constructed machine tool ranges. In order to minimise ‘stick-slip’ in the slideway motions on heavy moving cast and fabricated members –

‘Stick-slip’ or ‘Stiction’ , is the jerky-motion between sliding members due to the formation and destruction of junctions [due to localised pressure-effects]. (Kalpakjian, 1984)


437

. Table 14. Comparison of different drive systems for machine tools CONTRIBUTIONS:

Leadscrew:

Ballscrew:

Belt-drives:

Linear motor:

Noise

Quiet

Noisy

Quiet

Moderate

Back-driving

Self-locking

Backlash

Increases with wear

Constant

Increases with belt wear

Negligible

Repeatability

±0.005 mm

±0.005 mm

±0.004 mm

Best (2,000 N µm–1, so if greater kinematic and dynamic performance is required, then it might be necessary to utilise linear-motor drives. A tabulated table for suitable comparisons of the various types of motional drive systems available today, is presented in Table 14. It has been widely-reported that either a high-quality lead-, or ball-screw having rotary encoders, will have a unidirectional repeatability to within 6-to-

Where: L = overall length of the ‘elastic member’. (Galyer and Shotbolt et al., 1990) Example: If a machine tool’s structural moving member is 1,000 mm in length and it is situated on a much longer ‘bedway’ , then the ‘Tychoways’ should (ideally) be symmericallypositioned (i.e. fixed to the moving member) at a distance of 554 mm apart – in order to obtain the minimum of elastic distortion as it moves backward and forward along the hardened bedway.

438

Chapter 9

. Figure 217. Typical ‘modular design’ and construction of machining centres, with ‘ballscrews’ and ‘tychoways’. [Courtesy of Cincinnati Machines]


7 µm. However, if linear encoders are fitted this minimises any potential ‘ballscrew wind-up’ , although the problem of an ‘Abbé -error’ is still present. Yet another source of machine tool-induced problems are more specifically termed ‘uncertainties’ , while ‘hysteresis’ can also contribute to the overall ‘error budget’. Thus, hysteresis may result when the same position has been commanded by the CNC controller,

‘Abbés Principle’ – was derived by Professor Ernest Abbé in 1890 (i.e. having studied and graduating from the University of Jena) and is still valid today. The Abbé principle simply states that: ‘The line of measurement and the measuring plane should be coincident’. An example of this ‘Abbé Law’ wellknown to engineers the world over, shows that a conventional micrometer calliper ‘virtually-obeys’ the ‘Abbé principle (i.e. there is a small ‘cosine error’ present – which can usually be ignored), whereas, the Vernier calliper has a much larger offset between the fixed and moving jaws – where the component being measured is situated, to that of the beam – where the scale’s reading for this measurement is taken. Ideally, both measurement and reading should be lined-up, without an offset. [Sources: Busch, 1989; Whitehouse et al., 2002]

‘Uncertainty of machine tool positioning’ – the question often asked in calibration-related tasks, is: ‘What is measurement uncertainty?’ Uncertainty of measurement refers to the doubt that exists about any measurement; there occurs a margin of doubt for every measurement. This expression of measurement uncertainty raises other questions: ‘How large is the margin?’ and ‘ How bad the doubt?’ Hence, in order to quantify uncertainty two numbers are required: (i) being the width of the margin – its interval, (ii) plusits confidence level. NB This latter value states how sure one is that the actual value occurs within this margin.For example:On a CNC machine tool, the command may be to move the X-axis slideway 1000 mm plus, or minus 0.05 mm at the 95% confidence level. This uncertainty could be expressed, as follows: X-axis slideway motion = 1000 mm ± 0.05 mm, at a level of confidence of 95%. In reality, what this statement is implying to either the programmer, or calibration engineer is that they are 95% sure that the actual X-axis position now will lie between: 999.95 mm and 1000.05 mm. Many factors can contribute to the overall ‘error budget’ as it is sometimes known, but this is beyond the scope of the present discussion – see references for further information. (Smith, 2002)

‘Hysteresis’ , can be defined as: ‘The difference in the indicated value for any particular input when that input is approached in an increasing input direction, versus when approached in a decreasing input direction.’ (Figliola and Beasley, et al. 2000)For example:Hysteresis usually arises because of strain energy stored in the system [i.e. in this case, within the machine tool], by slack bearings, gears, ballscrews, etc. (Collett and Hope, 1979)

439

but from opposite directions, causing the motion system to creep by an amount larger than the backlash alone (i.e. the hysteresis). This effect is the result of unseen working clearances and compounded by the machine’s elastic deformations, although by pre-loading the ballscrews it will minimise some of these effects. In HSM machining applications, all ballscrew and indeed any screw-driven systems have some additional limitations, such as its ‘critical speed’ of rotation. At the critical speed, a ballscrew starts to resonate10 at its first natural frequency (i.e termed ‘whipping’). Hence, the critical speed is proportional to the ballscrew’s diameter and is inversely proportional to the distance between the screw’s supports – squared. For a very long and slender screw-driven machine tool application with wide supports, here, most recirculating ballscrews would have a critical speed of approximately 2,500 rev min–1. It should also be said, that for many ballscrews assemblies they can be rotated at higher rotational speeds than the 2,500 rev min–1 previously mentioned, before they reach their critical speed, but for very fast accelerations and decelerations, then they become increasing challenged. In fact, on some HSM machine tool configurations, multi-start ballscrews have been employed to increase linear response, but here the ‘critical speed’ will probably be lessened – due to reduced inherent ballscrew stiffness. Even ‘matched’ twin ballscrews have been fitted to HSM machine tools – to minimise any potential ‘yawing motions’ at high linear speed by the moving member along the machine’s bedway. These ballscrew limitations are probably why linear-motional drives are becoming a realistic alternative, but they are only fitted at present, to high capital cost HSM machine tools. One of the bi-products of HSM’s greater stock removal rates, is the excess volume of hot swarf which must be speedily and efficiently removed from the vicinity of the machine – which is more readily achieved for horizontal machine tool configurations. Thermal effects in general on any machine tool become a problem, particularly as many milling spindles utilise direct-drives, with the motor being mounted in-situ with the spindle. Here, the spindle motor creates heat, the thermal effects of which can be analysed by either

10 ‘Resonant frequency’ , can be defined as: ‘The frequency at which the magnitude ratio reaches a maximum value greater than unity.’ (Figliola and Beasley et al., 2000)

440

Chapter 9

a three-, or five-probe spindle analyser, as depicted in Fig. 218 along with a typical case-study for a spindle’s thermal drift graph. Not only can such spindle analysis equipment determine a spindle’s thermal growth, it can also detect: thermal distortion; spindle error; machine resonance; vibration; plus repeatability. In the casestudy graphically-depicted in Fig. 218b, the variation in the machine tool’s temperature is the major cause of positioning error. This thermal drift graph offers-up a number of significant questions that need to be addressed, such as: ‘How long does it take for the machine tool to stabilise?’; ‘How much Z-axis growth does the machine produce at full spindle speed?’; ‘How far has the machine’s displacement become, because of distortion in the machine tool structure?’. Spindle analysers are able to operate at rotational speeds varying from 0 to 120,000 rev min–1, making them an ideal tool for condition monitoring diagnostics on machine tool’s equipped with HSM spindles for subsequent analyses. In Appendix 15, a trouble-shooting guide has been produced to high-light: problems; causes; tests; etc.; that can be obtained from analyses by employing machine tool spindle analysers. In Fig. 219, the polar plots illustrated show how the spindle analyser has the ability to evaluate both a new and rebuilt machine tool spindle’s condition. In the spindle error indicated on the polar plot depicted in Fig. 219a, it illustrates here, that a very badly rebuilt spindle is simply not acceptable in this state. This poor spindle performance, is in the main, largely the result of significant radial variability (i.e. total error: 12 µm @ 4,005 rev min–1), which would severely compromise any cutting tool’s machining performance. Conversely, in Fig. 219b, a well-worn spindle assembly is shown prior to rebuild, having a total error of 4.6 µm @ 1,702 rev min–1, after its ‘correct’ rebuilding (Fig. 219c), the total error has been drastically reduced to a total error of 1.9 µm @ 1,700 rev min–1. Visually, the differences between these two polar plots is quite astounding, in that both the asynchronous and average errors present have virtually disappeared in the latter case, making it ‘as-new’ and, ready to perform significant machining service. It is well-known fact by machinists familiar with their older and heavily-utilised machine tools, that certain machine spindles will run smoother and produce better machined roundness on workpieces if they are run at the so-called ‘sweet-spot’ , equally, the quality of the parts produced will be affected if the spindle is run at its ‘sour-spot’. By utilising a spindle analyser, significant information can be gleaned from such rotational analyses, allowing speed

ranges to be selected which would currently optimise the present status of the spindle’s condition, prior to its potential rebuild – when called for at a due date in the future. For most machine tools today that are involved in HSM activities, in general the spindle cartridges are of three distinct design configurations – which incidentally do not normally include ball bearing spindle types11, such as: • Magnetic ‘active’ spindles (Fig. 220a) – typically might have a cartridge with a speed range from: 4,000 to 40,000 rev min–1, delivering 40 kW continuous power, peaking at over 50 kW . Such ‘active’ magnetic spindles can maintain 1 µm maximum run-out, by digital control of the series of specifically-orientated magnetic currents – being initiated by radial and axial sensors, that continuously monitor the spindle’s rotational axis position 10,000 times second–1, NB Further refinements to the spindle occur, with these temperature-controlled milling spindle cartridges maintaining dynamic balance, regardless of the milling cutting loads imposed, this latter factor being an important criterion when attempting to reduce cutting tool vibrational effects. However, these ‘active’ spindles are not cheap to purchase and run, with another negative effect being that they are normally rated for only several thousand hours operational running time. Such cartridges come with a variety of rotational speed ranges and spindle power outputs.

• Pneumatic spindles – have been available for many years, with aerostatic bearings equalising the forces exerted while cutting and remaining centralised within the spindle’s housing, yet still achieving dynamic rotational balance,

11 ‘Ball bearing spindle designs’ , are not normally specified for HSM operations, because at around 20,000 rev min–1 – this being ‘mechanically-set’ as the upper rotational velocity of such spindles. So, due to these high rotational speed effects and of the combination of centrifugal forces, it means that at approximately 80 m sec-1 rotational speed, the balls will lose contact with the journal walls. As a result of this loss-of-contact the hardened balls and raceways will rapidly wear out (i.e. the results of so-called: ‘Brinelling’ in the raceways, creating both poor circumferential wear patterns, delaminations and associated frictional effects – leading to major debilitating spindle roundness modifications.) (Smith et al., 1992)


. Figure 218. Machine tool spindle analysis system. [Courtesy of Lion Precision]

441

442

Chapter 9

. Figure 219. Machine tool spindle error plots, illustrating spindle condition. [Courtesy of Lion Precision]


. Figure 220. A typical UHSM spindle cardridge listing some of factors affecting such a spindle’s design and its operation

443

444

Chapter 9

. Figure 221. Attainable cutting parameters – with differing milling spindles, plus HSM is affected by the feedrate and distance to be traversed, prior to the desired velocity being achieved – for conventional slideway motions


NB Pneumatic spindles can be rotated at exceptional speeds, by virtue of the ‘ideal condition’ of minimal metal-to-metal contact, although one serious disadvantage being they suffer from a low power output, or to be more specific – torque.

• Hybrid spindles (Fig. 214-top) – have been de

veloped to answer the major drawback to utilising pneumatic spindles. The hybrid spindle as its name suggests, is a combination of conventional ballbearing and pneumatic spindles. Here, the spindle design incorporates an aerodynamic thrust bearing with transversal spiral grooves (Fig. 214-top right) thereby creating an intense pressure wave profile, which can withstand up to 300% greater static loads to that of a conventional aerostatic bearing. A typical hybrid aerodynamic spindle bearing allows the assembly to achieve rotational speeds ranging from: 20,000 to 40,000 rev min–1, with >15.5 kW power at peak speed. NB Hybrid spindle cartridges are significantly less expensive than the magnetic ‘active’ spindles, but more expensive than pneumatic spindle cartridges. In both of these latter versions, they have a relatively long in-service life, as wear-rates are minimised, but do not have the stock-removal capability of the former cartridge.

In Figs. 220 and 221a, are shown some of the principal factors that affect UHSM spindle performance. In Fig. 220, these factors are represented in an Ishikawa (i.e. ‘Cause-and-effect’) diagram. Here, the many of the inter-related effects can be seen, although other factors can also be added, depending upon the local conditions of usage: cutting data; workpiece material; wet, dry, or near-dry cutting; together with the machine tool’s overall condition.

9.2 HSM Dynamics – Acceleration and Deceleration If the HSM spindle cartridges – mentioned above – are fitted to conventional CNC machine tools, or the more likely low-cost scenario would be to simply fit a

445

echanically-driven speed-increaser. Then the result m of this HSM spindle fitment, will enable high rotational speeds to be produced, but it leaves the CNC-processors somewhat compromised in its ability to produce the desired acceleration and deceleration capabilities. As a practical example of the problems likely to be encountered, the graphs produced in Fig. 221b and c, were drawn from an industrial HSM machining experience at a precision metrology company’s premises, using several of the latest vertical machining centres with the spindle of each machine, being fitted with a mechanically-driven speed-increaser (i.e. see Fig. 243a). By utilising the inboard CNC clock – having a resolution of 0.0001 seconds, the elapsed times for slideway motion over varying distances was established. In Fig. 221b, an exponential relationship is depicted for the X-axis, this being a typical situation for the other axes on the machine tool. By a determination of the required motional distance to attain specific velocities, it was possible to illustrate the restrictive nature of both acceleration and deceleration for small slideway motions. In Fig. 221c, this illustrates the effect of the required distances to be executed at various velocities. From Fig. 221c – by way of an example, if a feedrate of 8,000 mm min–1 was utilised, then it would be necessary for a minimum movement of the slideway to be 16 mm to momentarily achieve the desired feedrate, which is typical for a machining centre having an acceleration of 1.08 m sec2. This physical problem in actual positioning to the required component’s dimensional feature is not too great a problem for long linear feeding distances – as the slideway velocity could be reached, but acceleration and deceleration becomes exacerbated by the smaller more intricate prismatic features normally found on the more minute, or smaller parts often produced by HSM milling operations, leading to potential scrappage problems. From the results of inspection procedures conducted on the HSM over a range of standardised testpieces, it was concluded that virtually all of the detrimental dimensional effects introduced by the HSM milling operations, could be attributed to severe ‘servo-droop’ – more will be mentioned on this subject shortly. Prior to manufacturing the HSM testpieces, they were designed on a CAD system and their respective tool paths were post-processed by the integrated CAM software. Hence, with regard to machining cycle times, they were either calculated by the CAD/CAM, or were the actual in-cut times – see Table 15. With re-

446

Chapter 9

. Table 15. A comparison of the theoretical and actual machining times for the manufacture of testpieces, by various production routes Machining Method:

Theoretical Cad/ Cam Time:

Actual In-cut time:

Conventional

3.86

3.71

High-speed

1.26

1.50

High-speed (G61)*

1.16

1.55

NB All times in minutes * G61 is the ‘Exact-stop’ mode of machine command and, when activated, the machine tool will not initialise another movement until the previous axis command has been completed (i.e the targetpoint), thus ensuring an accurate and final slideway positioning. [Source: Smith and Maxted, 1995]

gard to these cycle-times for testpiece manufacture, a significant improvement accrues when utilising HSM milling techniques. Although the increase of actual cycle-time from that of the theoretical high-speed CAD/ CAM estimation, can be due to a number of factors as previously noted by Smith and Hanson (1993). Not least of which was found in this case, where the CAM system tended to under-estimate the actual time to machine a component feature. This time-difference is marginally compounded by the ‘servo-droop’ effects. If the machine tool’s G61 (‘Exact-stop’) mode was employed, the actual cutting time in general showed only a marginal increase, over the normal HSM cutting time, although the dynamics of motion tended to be somewhat jerky in action as the command ensured it reached its targeted positions. At normal HSM milling performance, another exacerbating reason for the increase in cutting time over theoretical, was attributable to the axis acceleration/deceleration parameters (i.e see Figs. 221b and c). Thus, these machine tools basically failed to reach the required slideway acceleration/deceleration then maintain these velocities for about 20% of the total in-cut times, this was for a component of somewhat moderate dimensional size and pocketing intricacy (i.e. the overall testpiece dimensions were approximately: 150 mm in squareness, by 50mm deep). From the testpiece results, various remarks can be made concerning the advantages of employing an HSM strategy over conventional milling practices, these are:

• Despite a reduced cut depth, the HSM cycle times • • •

are a 66% improvement over conventional milling production techniques, Using HSM it will significantly reduce burr formation – although not entirely eliminating it, when compared to that of conventional practice, Distortion of the thin wall features was minimised by HSM, When employing a speed-increaser (Fig. 243a), its bearing’s stiffness is critically important in order to obtain an acceptable milled surface texture.

Finally, by utilising even the most elementary form of HSM approach – using a speed-increaser, highlights the production advantages to be gained from adopting this strategy, albeit for limited periods of continuous cutting time, which normally dictates such ‘increaser’s’ practical usage.

9.2.1 HSM Dynamics – Servo-Lag Most of today’s CNC machine tools use ‘proportional servo-systems’ , where the axis velocity is proportional to the difference between the actual position and the command position (Fig. 222a). This ‘error signal’ is utilised by the system to determine any acceleration/ deceleration necessary as well as the steady-state velocities. As one can visualise from Fig. 222a, the distance between the actual and commanded positions is commonly termed ‘servo-lag’. This explanation can be taken a stage further in Fig. 222b, where the illustration depicts how a ‘proportional servo-system’ is used to mill a sloping line. In this example, DX and DY are the total programmed changes in position on both the X- and Y-axes, respectively, to go from point ‘A’ to point ‘B’. Conversely, DXL and DYL are the amount of lag on each axis at point ‘C’ along the tool’s path from ‘A’ to ‘B’. Furthermore, in such a system the lag on the X-axis must be proportional to a similar lag in the Yaxis, in order to accurately follow the slope of the line. This affect can be mathematically-represented by the following relationship: DX L DX = = Slope of the line. DYL DY

In Fig. 222c, we can gain an appreciation of just what happens when the servo-lag on both axes is not proportional. As the machine tool’s axes travels from point ‘A’ to point ‘B’ , the lag on the X-axis is proportionally


. Figure 222. The CNC control problem of servo-lag and its affect on the associated HSM motional kinematics

447

448

Chapter 9

less than the lag on the Y-axis. This error might be the result of the ‘servo-gains’12 between the X- and Y-axes not being properly synchronised. Normally, ‘servogain’ can be expressed in units of: mm min–1 [i.e. velo city / mm (i.e. distance in 0.001)] of lag. Thus, lag can be determined using the following relationship: Lag L (mm) =

,  Feedrate F = = . mm. Gain G �.

For example: If a machine tool’s moving axis is travelling along its slideway at 2,500 mm min–1 and the servo has a gain of 2, the lag will be 1.25 mm, as indicated in the following calculation: L (mm) =

F ,  = = .mm. G �.

9.2.2 Effect of Servo-lag and Gain on Corner Milling If two axes with correctly matched servo-lags can move in a straight line from point ‘A’ to point ‘B’ , then to comprehend the effect of gain, let us consider what occurs when milling a right-angled corner at a constant feedrate without stopping (Fig. 222d). Whilst milling the corner from ‘A’ to ‘B’ and the onward to ‘D’ , the servo develops a steady lag (DXL), until sufficient command signals have been generated to reach point ‘B’. It is at this position that the control begins to generate commands toward point ‘D’ , although the actual slideway has not yet reached point ‘B’ , owing to the servo-lag (DXL). At this point the Xaxis will begin to decelerate and, simultaneously, the Y-axis begins to accelerate, that is the velocity is proportional to the distance between the command signal and the actual position. Acceleration factors affect the slideway motions producing the result that the distance from ‘B’ to ‘C’ is always greater than DXL. Furthermore, this curved path is not a circular arc, but an exponential curve, with the amount of variance from the sharp right-angled corner being dependent on the

12 ‘Gain’ or to be more specific: ‘servo-gain’ , in this instance, is a measure of the servo’s responsiveness. Thus, the higher the gain, the lower the lag.

magnitude of servo-lag13, which itself depends upon the affect of feedrate and gain – according to the previous formula.

9.2.3 Effect of Servo-Lag and Gain Whilst Generating Circular Paths For one to fully understand just what happens when milling complex contours, it will be helpful to consider the simple case of a milled path where two straight lines are joined by a semi-circle (Fig. 222e). In this situation, the milling operation occurs at a constant feedrate moving from point ‘A’ in a straight line until the command dimension reaches point ‘B’ . However, at this point, because of the effect of servo-lag, the axis motion will have only reached point ‘BL’. Therefore, as the control command is moving forward at a constant rate, it begins to generate commands toward point ‘C’. This action results in the axis motion beginning to move away from the desired path at point ‘BL’. The dotted line depicted in Fig. 222e, shows the actual path taken by the cutter and as one can visually observe, from points ‘BL’ to ‘CL’ , the deviation from the desired path is shown as ‘e’. In this example, the magnitude of ‘e’ is determined as a function of the: feedrate; gain; plus the desired radius. When the radius error approaches the programmed radius, the resulting machined profile appears distorted and is hence, impracticable. Specifically, if one needed to mill a 25 mm radius at a feedrate of 2,500 mm min–1 with a machine tool gain being: 25 mm/min/0.001, then the error generated would be approximately 0.125 mm, equally, if the gain was increased to 100 mm/min/0.001, the maximum error ‘e’ will be considerably reduced to approximately 0.008 mm. A machined curve is an approximation on CNC machine tools, in that the profile is constructed from a series of short connected segments, or chords. The controlling factor on the length of such segments is the deviation between the centrepoint of any chord

13 ‘Servo-lag’ , is sometimes referred to in the literature as: ‘Servo-droop’ – due to its ‘rounding-effect’ at the corners, this being particularly prevalent when fast feedrates are selected, creating fast tool path velocities, particularly when normally undertaking high-speed milling operations.


and a point at right angles on the programmed curve. The linear distance between these two points is usually termed the ‘maximum allowable chordal deviation’ and is a function of the CNC controller’s executive software. So, the resultant machined curve is a combination of the chordal deviation and the servo-lag for a particular machine tool. To illustrate this condition, Fig. 220f shows the culmination of servo-lag when following a contour, with the curve ‘C1’ being the desired contour, ‘C2’ a linear approximation (i.e. the programmed path), ‘C3’ is the actual generated path resulting from servo-lag utilising a high gain servo and finally, ‘C4’ being the path generated by a low gain servo. Through servo-lag, a smoothing of any contour occurs owing to the lagged cutter path, this causes severe contour problems with respect to part accuracy and precision for the simple arc geometry depicted in Fig. 222e. Clearly then, servo-lag and gain promote a variety of effects on complex shapes, depending upon their geometry and tolerance, with these errors becoming still more complicated when one considers three-dimensional milling contouring. In many circumstances the cutting of three-dimensional profiles may necessitate utilising four, or more axes with either one, or two of them being rotary axes being necessary to create the required tool paths to produce the component. The servo-lag and gain on all axes must be considered when manufacturing complex part geometries. Regardless of the workpiece’s geometry, or the number of axes utilised, there is one factor that should be emphasised concerning potential errors created by servo-lag. If servo-lag is extremely large, then this ‘lag’ can easily exceed the positioning errors in the machine tool’s basic specifications.

449

be read, interpreted, the activated to obtain dynamic slideway motions. This CNC exercise is usually referred to as the ‘block processing time’. The maximum allocated time for block processing of information is dependent on the length of slideway stroke (i.e. chord length) and its associated feedrate. It is possible to calculate the maximum block processing time (Tb), as follows: Tb =

Maximum stroke length Feedrate

Tb =

. . = = . s, or  ms , � 

For example: if we require a profile’s chord length (i.e. stroke length) of 0.50 mm, in order to maintain contouring accuracy whist milling at 3,000 mm min–1, or 50 mm s–1, with a maximum block processing time, then this ‘time’ should be less than:

Possibly the main factor limiting contouring speed is the CNC’s inherent processing speed, with each programmed-block14 generated for every axis having to

Many CNC’s have block processing times typically within the range of 30 to 60 ms, as can be seen from the above example, the CNC program would suffer from ‘data starvation’ , whilst the controller attempts to catch up on its data processing. Such ‘starvation’ would cause hesitation in the slide motions, slowing down the cutting time and leaving ‘dwell marks’15 on the machined workpiece’s surface. Since this ‘data starvation’ effect is unacceptable, a lower feedrate must now be programmed to overcome the problem and as a result, the cycle-time increases. In the above example, if an older CNC was fitted to the machine tool with the controller’s block processing time being 60 ms, the cut would have taken six times longer to generate the profile, than a more modern CNC controller having a processor capable of 10 ms. So that we can fully-comprehend the CNC processing speed problem, let us now consider two widely differing machining applications: 1. Complex three-dimensional milling of a hob – to manufacture a die utilised in the production of intricate and expensive military metal buttons. Such

14 ‘Programmed blocks’ , these are basically the ‘G-’ and ‘M-’ and ‘Auxiliary-codes’ which make up each individual block’s line, with successive blocks in a logical sequence containing the whole CNC program. Generally speaking, the smaller the number of blocks – for the successful production machining of the part, the more efficient and refined has been the programming. (Smith et al., 1993)

15 ‘Dwell marks’ , here are the result of an ‘untimed delay’ in the program’s execution, created in this instance, by data starvation (i.e. block processing speed was simply not fast enough). These untimely delays in the activation of programming blocks cause the rotating cutting tool to rest and press against machined surface and thus, generate minute ‘gouging-effects’ in the surface. (Source: Seames, 1990; Smith, 1993)

9.2.4 CNC Processing Speed

450

Chapter 9

a hob will more than likely have very fine detailed work on its surface, perhaps with radii as small as 0.25 mm, requiring a tool tip radius of 0.025 mm. In order to machine the button’s elaborate features with such a small milling cutter, the spindle speeds might need to reach 40,000 rev min–1, utilising a feed per revolution of 0.008 mm, giving a feedrate of 320 mm min–1. Many production engineers would not consider this as an example of high-speed milling, but let us look more closely at this particular machining problem. If the controller has a servogain of 4, with a feedrate of 320 mm min–1, this means that the servo-lag would be 0.75 mm min–1, which is consistent with milling radii of 0.25 mm. However, if the servo-gain was 1, this would cause a servo-lag of 0.320 mm min–1 and in this case, it obviously could not machine that button’s intricate detailing. In such circumstances, it would be necessary to appreciably reduce the feedrate to say, 75 mm min–1 to generate the button’s contours, leading to the cycle-time increasing by 400%. Let us also now consider the impact of block processing time under these conditions. To mill a radius as small as 0.25 mm, we would need to produce linear stroke lengths of just 0.075 mm – to reproduce acceptable button detailing. This intricate contouring work requires a block processing time of 15 ms. If the CNC controller has a block processing time of just 60 ms, then the feedrate must be limited to 75 mm run–1 which again, increases milling time by a factor of four. 2. ECM pattern electrode for a Turbine fan (i.e large aluminium casting) – here, the electrode’s geometry has very gentle three-dimensional curves. In this situation the chosen CNC machine tool’s milling spindle has a 250,000 mm min–1 capability, coupled to adequate power to cut at a feed of 0.25 mm rev–1. This production requirement produces a feed rate of 62,500 mm min–1 (i.e. being the product of: 250,000 x 0.25) would be possible. For accuracy and precision, a chordal deviation (Cd) of 0.005 mm would indicate a stroke length of 0.75 mm – if the minimum radius of curvature for the Turbine fan’s geometry was 25 mm. Assuming that the servogain of 1 was available, then we would obtain errors as large as 0.125 mm and with such errors, the machine tool would not produce an acceptable part. Further, at 62,500 mm min–1, if the block processing time (Tb) was 60 ms, this ‘timing’ would require stroke lengths of 2.5 mm instead of the 0.75 mm we needed for the required accuracy and precision. Therefore, in order to eliminate the effects of low

gain, or slow processing time, it is necessary to depress the feedrate, resulting in the cutting time being increased up to 400%. When considering these two practical examples from a metaphorical sense, the former method can be compared to that of racing a go-kart on a small tight track, while the latter method is similar to a highly tuned sports car racing on a longer and smoother track. The go-kart may only reach speed a of 30 km h–1, whereas the sports car may hit speeds of >200 km h–1. The corner forces and reaction times are similar for both methods, even though the speeds are vastly different. Looked at from yet another viewpoint, we can say that the frequency of response of both the drive and car, that is their servo-gain and processing time, are very similar in both examples even though the speeds (feed rates) are radically different. In the day-to-day production environment, the duplication of specific and precise contours is the end result of a combination of many inter-related factors. As the number of machine tool axes required to produce sculptured part surfaces increases, the difficulty of obtaining the desired profile also becomes proportionally problematic. So, machine tools that would normally produce excellent general-purpose machining work, may not be either accurate, nor efficient enough to manufacture complex part contouring geometries. That is, unless their CNC processors can achieve block processing speeds of 1g while providing >3 tonnes of continuous thrust in the Z-axis. The machine tool is compliant with Standard: ASME B5.54, obtaining accuracies of 175 Nm µm–1. Pallet change time is 10 seconds, with a tool change time of 6 seconds (chip-to-chip), having a tool stor-

age capacity of 50 tools (i.e. expandable). The standard tool spindle cartridge is rated at 22 kW , with spindle speed ranging from 100 to 16,000 rev min–1, having an angle of tool spindle inclination of up to 25°. In demonstrations and in-house trials at the manufacturer’s premises, the initial prototype ‘Variax’ machine was operated and machining components for three years. On one large aerospace machining application of a critical component, the original cycle-time was 19.3 hours on a conventional machining centre, but when this same part was cut on the ‘Variax’ , it took just 3 hours to complete, with the additional benefit of being both more accurate and precise. Equally, when a smaller aerospace component – a landing bracket – was originally machined it took 1.65 hours on the conventional machine tool, but when the same part was placed on the ‘Variax’ , it took only 0.55 hours cycletime to complete.


. Figure 225. Variax/Hexapod offers a unique and rigid design solution, with many production benefits. [Courtesy of Giddings & Lewis]

455

456

Chapter 9

The capital outlay for a ‘Variax/Hexapod’ machine tool, costs about the same as a similar specification – component size capacity, to that of a conventional fiveaxis machining centre, but with the above additional performance benefits.

Robotic Machining Robotic machining applications have been utilised for some years, currently up to a thousand such installations are to be found world-wide. Probably the biggest user of robotic machining is the aerospace industries, although the automotive industry is catching up fast. Most of the current research work into robot machining is undertaken with an anthropometric type of robot configuration, usually having either five, or six axes (i.e. see Fig. 226). If a six axis robot is employed (Fig. 226), it gives several benefits over say, a five-axis machining centre, with probably the greatest production advantage being access to the workpiece’s surface features (Fig. 226b). This extra degree of freedom, allows the alternative wrist positions to achieve identical tool positional orientation – regardless of the workpiece contour, enabling the robot axis the ability to rotate the wrist about the tool’s axis. Moreover, where speed and acceleration are important, robots normally out-perform the more traditional machine tool structures, mainly due to both their low weight and minimal inertial effects. Within the robot’s programming language, parameters exist that allow a balance to be made between robotic arm speed and accuracy. So, when roughingout higher speeds can be employed – at the expense of accuracy, then the parameters can be changed to a slower speed, but with greater finished machined part accuracies. Limitations also occur when using robots for machining, with perhaps the most obvious one being their intrinsic lack of rigidity, when compared to that of virtually any machine tool. If a robot is employed at conventional spindle speeds the tool’s cutting forces are simply too great, creating both vibrations and deflections in the robotic arm, which badly impacts on the component’s machined surface texture and dimensional accuracy. In order to mitigate against any cutting force and accuracy deficiencies, it is essential to utilise HSM spindles, in combination with taking small DOC’s, to minimise these effects. The robot’s kinematic structural complexity also causes positional problems, due to innate manufacturing tolerance effects to the location of the joints in the

robotic arm. This dimensional and geometric tolerancing build-up, means that the mathematical model used to control individual joint positions in space, will minutely differ from the actual reality to that of the joint placements. This positional difference being most apparent and exaggerated around the perimeter of the robot’s working envelope. Often, it is impossible to calibrate the mathematical model for these complex inter-related errors, as the overall complexity of robot’s control algorithms then become such that they cannot execute the kinematic motions fast enough to dynamically control the robot at the required axis trajectories. In calibration trials on a typical six-axis anthropometric robot undertaken by Young (1999) while working in cartesian co-ordinates for the robot’s positional accuracy – when assessed with laser interferometer instrumentation, the results produced linear errors of ± 0.8 mm in each cartesian co-ordinate (i.e. X-, Y- and Z-axes) – across the complete working range. The error curves produced were in fact, symptomatic of the kinematic structure of these robots. Characteristics included a decrease in backlash to almost zero – toward the perimeter of the working envelope (i.e. due to gravitational effects), while a combination of linear and sinusoidal effects combined to produce the total error. Even though these robot errors are large with respect to those found on a machine tool, they should not deter robotic usage for suitable HSM applications. The repeatability shown by most robots is usually far superior to that of its accuracy. In practice, any multiple axis robotic arm utilised for sculptured machining applications, needs to have their axes biased and offset in order to eliminate the ‘paradox’ that might overwhelm them when all axes are attempting to keep the end-effector (i.e. tool) normal to the contoured surface (Fig. 226b). For example, on a three-roll wrist (Fig. 226), the combination of the kinematic linear and rotational build-up, may result in instead of one of these axes angularly moving just 1°, it causes it to actually move 359° instead – thereby scrapping the workpiece in the process! To alleviate this problem, if the workstation/stand, is offset to one side and angled, this offset and compound angle will minimise the axis predicament that can afflict the robot’s subsequent programmed motions. The workstation position in its ‘known space’ with respect to the robot axes datums must be known and accurately calibrated, thus ensuring that the angled grid-plate for resulting workpiece fixturing is both ‘fixed and qualified’. By slowing down the robotic axis trajectories in the fi-


. Figure 226. Robotic high-speed machining with a multi-axis anthropometric robot, equipped with HSM milling head. [Courtesy of Southampton Solent University/Westwind Air Bearing Ltd./Smith, G.T.]

457

458

Chapter 9

nal machining passes over a sculptured workpiece surface, a certain degree of accuracy can be achieved, but even here, it does not approach that of any moderately accurate and precise machine tool. Tool centre-point programming is the preferred option, as when utilising cartesian co-ordinates, the programmed points represent the cutter’s end – once the required offset has been established. This tool centre-point program allows the programmer to effectively disregard the tool’s length in any subsequent machining applications. However, care must be taken with any flexible coupling connections, such as pneumatics, water, or electrical wiring to the HSM spindle – allowing sufficient slack in the piping (i.e see Fig. 226), thus ensuring that the robot will not inadvertently de-couple these services, as it attempts to manufacture the complex part.

9.4 HSM – Toolholders/ Chucks Introduction Rotational speeds for tooling assemblies subjected to HSM applications must of necessity be very high, this can create several problems for any tooling utilised in such machining activities. Notably, due to the centrifugal force the toolholder could swell and slacken its gripping force on the tool’s shank. In an extreme situation due to the application of cutting forces this might cause the tool to speedily exit from its holder, in so doing either scrapping the workpiece, or become a severe safety hazard to any operator in attendance at the machine tool. Therefore, both a good fit and connection is an essential requirement of any HSM machining application. Thus, the mechanical interface between the toolholder and the machine’s spindle, together with that of the tool’s shank and its respective toolholder are the prerequisites for a successful HSM application.

9.4.1 Toolshank Design and Gripping Pressures Tool and Sleeve – Mechanical Interface The latter point briefly mentioned in the previous section concerning the tool shank’s fitment in the holder, is an important criterion in any HSM applica-

tion. Nonetheless, of greater interest and note, is the actual gripping pressure exerted at the toolholder’s mechanical interface with its mating tool shank. Some interesting toolholder designs have been attempted to increase the gripping pressure here, with the level exerted at the free-end of the holder not being too great a problem, due to elastic compliance of the sleeve in this region. The notable difficulty arises as the shank is being gripped toward the flange end of this sleeve, where with most conventional toolholder sleeve designs, the gripping pressure drops-off significantly at this locality. At high rotational speed and under the application of the cutting forces, the cutter will tend to become unstable and present a distinct ‘wobbling motion’ due to the lack of gripping pressure and support here, being exacerbated by the higher imparted centrifugal forces. In order to try to alleviate this lack of grip problem, significant design effort has been expended over the years to attempt to increase elastic behaviour and hence amplify gripping pressure at this sleeve’s region. In the graph shown in Fig. 227b, the elastic behaviour and its associated contraction are plotted from the tool sleeve’s free-end in linear steps along the sleeve and toward the flange. In most prior designs the sleeve contraction on the tool’s shank near the flange was approximately 50 µm (i.e. shown by plot ‘B’). By introducing a radially-plunged undercut to the sleeve’s outside diameter at the juncture of the flange and sleeve (i.e. see Fig. 227a), this creates significantly increased elastic behaviour and hence improved gripping pressure at this region of the sleeve, as illustrated by the plotted relationship of the test results – producing a sleeve contraction of 145 µm at the flange, as indicated by ‘A’ , in Fig. 227b. This vastly improved gripping pressure with the ‘undercut flange’ is of the order ≈150%, when compared to other design techniques, which restricts any attempt at the tool’s ‘wobbling behaviour’ as a result of the improved elastic contraction on the shank. In order to obtain a good overall elastic gripping pressure along the length of the sleeve, multiple hardened needle rollers are positioned – at a slight angular inclination, both around and along the male shallowangled tapered sleeve’s periphery (i.e. items 2 and 4 in Fig. 227a). As the female tapered sleeve (item 3) is rotated, the caged inclined needle rollers rotate and begin to steadily move up the male tapered sleeve toward the flange (item 1) and, in so doing, elastically compress the sleeve (item 2), which in turn will tighten on a tool’s shank. As the tapered female sleeve (item 3) is fully tightened with its ‘C-spanner’ , it will com-


459

. Figure 227. High-speed milling toolholders (chucks), holding cutter’s shank along its whole gripping length. [Courtesy of Diashowa Tooling]

460

Chapter 9

press its face against the rubberised contact seal and the flange (item 1), thereby acting as both a vibration damper and sealing against particle/debris ingress into the mechanism. The front face already being sealed against such potential particle ingress (i.e. see Fig. 228 sectional detail diagram). In order to increase the elastic behaviour of the male tapered sleeve (item 2 – Fig. 227) and improve its gripping pressure here, the end of the sleeve is partially slit along its length and around it at eight equally-spaced positions (i.e. see Fig. 228). This mechanical design solution for the tightening of the male tapered sleeve, gives possibly the optimum gripping pressure for such a mechanical interface on the tool’s parallel shank.

The disadvantages to hydraulic toolholders are few, but may prove a significant obstacle to their introduction, including the fact that their purchase costs are higher than their mechanical counterparts, with the other limitation being a new hydraulic sleeve is required for different tool shank diameters. However, this latter point can be mitigated against, by conducting a programme of: rationalisation; consolidation; and optimisation on the various production requirements for toolholder varieties and shank diameters (i.e. see Chapter 1, Sections 1.1.1 to 1.1.3 for details of such tool survey).

Tool and Sleeve – Hydraulic Designs

Shrink-fit systems require specially made tool holders, being designed for a specific tool shank diameter, although they can accommodate any style of shank. Once in-situ the tool in its thermal sleeve behaves in a very similar way to that of a one-piece tool. The high gripping pressure coupled to excellent concentricity (i.e. 20%, when compared to most of their mechanically-designed counterparts. Due to greater rotating concentricity of shrink-fit toolholders, there is a better wear pattern developed on the milling cutter’s teeth, or drill’s lips, etc., which it is claimed, increases tool life by >30% over conventional toolholders. In a similar fashion to that of hydraulic toolholders, thermal contraction occurs both around the periphery and along the whole mechanical interface, which will automatically centre the tool’s shank within its mating bore. This complete toolholder-to-tool fitment, minimises centrifugal force when operating in an HSM mode. Thus, the contraction of the toolholder rigidly locks the tool in-situ, this gripping pressure is at least 500% greater than for conventional toolholders. In fact the pressure exerted here, is even greater than that of the pull-stud (i.e. retention knob), meaning that in the presence of high forces, the whole assembly would be pulled out of the spindle before the tool would be released by its mating holder. The significant component of a shrink-fit toolholding system is the induction heating unit, as schematically depicted in Fig. 229a. Typically, the solid-state and self-contained unit is relatively compact and in operation to change tools, the user positions the tool holder in a receptacle built into a shelf at the front of the unit. An induction coil is located beneath the shelf and encircles the toolholder’s sleeve (i.e. collar), which

For an alternative tool-gripping pressure design concept, the hydraulic toolholder (i.e. see Fig. 228 – top right – for a section through the tool sleeve), offers an ideal alternative to the mechanical tool interface previously mentioned above. Here, the hydraulic toolholder manufacturer guarantees a 0.005 mm tolerance-inroundness (TIR) at 100 mm from the front face, with 690 kPa (i.e. 1,000 psi), coupled to perhaps, micro-filtration of the coolant with special pipes and couplings. The CAT system of dual-contact offers reasonable rotational control of the tooling assembly at moderate-to-high rotational speeds, as the mechanical interface system of face-and-cone provides a certain security against

24 ‘Dual-contact 7/24 taper system’ , refers to the taper being to the 7 inches of taper per 24 inches of length. This 7/24 system incorporates several Standards: CAT and BT 40- and 50-taper tooling.

466

Chapter 9

the onset of imbalance. Typical applications for these HSM dual-contact systems include: aerospace part production; precision die and mould making; automotive component production; as well as medical component manufacturing. It is worth digressing somewhat, to explain the situation of why the single-cone mechanical interface is simply not effective for HSM production applications. When rotational speeds begin to approach 20,000 rev min–1, it is not an unusual occurrence for the singlecontact conventional, or standard CAT V-flange tooling assembly to be effectively sucked into the spindle (i.e. as there is no mechanical contact at the flange), this being the result of a combination of the pull-stud pressure and the machine’s spindle ‘taper swelling’ – due to the very high centrifugal force acting at such high rotational speeds. In fact, this minute amount of ‘taper swelling’ can cause the tool holder to separate from the spindle’s surface and as a result cause considerable damage to both the cone’s male and female surfaces. In order to alleviate this HSM problem and run the tooling assemblies at even faster rotational speeds, the HSK dual-contact toolholders were developed, which will now be briefly mentioned.

Hsk Dual-Contact Tooling There are a number of toolholder designs that are alternatives to the conventional steep-taper spindle connection. Probably the most popular version for HSM is the HSK-designed tooling connection (i.e see pertinent HSK tooling details in Fig. 126c). HSK toolholder connections offer simultaneous fitment on both the taper and face, at the front of the spindle. The reason for their acknowledged popularity amongst the HSM machining companies, is because the increased rigidity of the joint, coupled with their inherent reduction in dimensions, compared to the equivalent conventional steep-taper connection. In Fig. 126c, the HSK 8° (included angle) short taper with its gauge face contact and simultaneous taper interference can be seen, which was designed in Germany to Standard: DIN 69893, being introduced in 1993. HSK is a German acronym that translates into English as: ‘Hollow short taper’. Thus, the HSK connection provides: • both high static and dynamic stiffness, • offering great axial and radial repeatable accuracy, • with low mass and stroke, • having inner clamping.

Therefore, with all these proven design advantages over conventional spindle connections, it allows the HSK tooling assemblies to utilise the increased rotational speeds necessary for an HSM strategy.

Triple-Contact Tool/Spindle Design The triple-contact connection is being offered by a few toolholder manufacturers (i.e. shown in Fig. 230). The triple-contact design relies on an inner expanding sleeve which maintains uniform contact between the machine tool spindle and the: toolholder’s top taper; bottom taper; and flange; this being regardless of the spindle speed employed. Of particular note is the inner expanding sleeve which functions particularly well at high spindle speeds. So, as the centrifugal forces increase – with higher rotational speeds, it causes the spindle to grow (i.e. ‘swell’), the toolholder’s spring mechanism forces the split-cone sleeve to proportionally-expand with the spindle. Further, the expanding sleeve also acts as a vibration-dampening device. The expanding sleeve extends the tool’s life on average by between 300 to 500%, by virtually eliminating vibration. As a result of this ‘vibration-free interface’ between the tool and workpiece, it provides smoother machining of: tool steels; aluminium alloys; plus other metallic alloys. This triple-contact connection system, also performs efficiently with extra-long tools (i.e see Fig. 231), notably when utilised on horizontal machining centres. The main reason for the enhanced triplecontact tool’s cutting performance with extended tooling assemblies, is the result of the ‘floating’ inner sleeve (Fig. 230) which acts to minimise any potential Z-axis deflection, thus maintaining its rotational concentricity. Such triple-contact tooling is not inexpensive to purchase, but these toolholders really do amortise their cost, by significantly extending cutter life, while improving part production rates. Further, it is claimed by the tooling manufacturer that the toolholder is ‘maintenance-free’ , while its spring-mechanism in ‘life-testing’ has achieved upward of one million tool changes. With the advent of either the double- and triple-contact systems, enabling contact between the machine tool’s spindle and the toolholder’s mechanical interface: top-taper; bottom-taper; plus flange; while ‘eliminating vibration’; this has been achieved under the unique conditions that arise with today’s HSM and high-accuracy and precision manufacturing needs.


9.5 Dynamic Balance of Toolholding Assemblies Introduction Balancing tools that are intended for HSM applications is vitally important and there are quite a few In-

467

dustrial/Manufacturing engineers and users who do not really understand the concept of how to achieve balanced tooling, or why it is really necessary. Either very long extended tooling required for say, for deeppocketing (Fig. 231), or tooling that is out-of-balance, will more than likely produce: chattering effects; gouging of a step, or face; loss of workpiece accuracy and precision; not to mention uneven and premature cutter wear. Whenever a new tooling assembly is destined

. Figure 230. Triple-contact tool connection system is ideal for any potential HSM operations. [Courtesy of Heartech Precision Inc. (HPI)]

468

Chapter 9

. Figure 231. Tool runout (≥10 µm) should be of prime importance when machining deep pockets. [Courtesy of Sandvik Coromant]


for HSM applications on a workpiece, a balancing operation needs to be undertaken, this statement is also true for many sub-HSM applications, particularly when extended tooling is used for whatever reason (Fig. 231). In fact, every rotating object (i.e. chuck, or tooling assembly, etc.), will generate vibration. As has been explained in the previous section, this vibration results from a number of sources, but principally here, from centrifugal forces produced by the rotation of an unbalanced mass. There are several types of unbalance that could arise, but here, we are mainly concerned with what is termed dynamic unbalance, which increases by the square of the rotational vel ocity. For example, any vibration produced by a tooling assembly at 3,000 rev min–1, is × 100 greater than an identical tooling configuration that is rotating at 300 rev min–1. Moreover, what is often either misunderstood, or indeed overlooked, is that any change to the tooling assembly – no matter how small it might seem, requires re-balancing! These tooling modifications include any occasion when a cutting tool is adjusted, or changed, or similarly if the toolholder is also either adjusted, or changed. Such changes to the ‘status quo’ of the tooling, will directly affect its ensuing balance, even minutely when just a ‘few microns’! So that, these miniscule changes to the tooling’s dynamic condition, causes a degree of tooling oscillation, hence an out-of-balance condition – with the likely problems that this creates. With the wide variety of tooling that is held in: tool storage carousels; magazines; turrets; etc.; they must all be ‘balanceable’ by some means. A range of balancing techniques can be employed here for either single-, or dual-plane balancing – more will be said concerning these effects will be made in the following section. The techniques utilised in achieving tool balance could include: • ‘Hard-balancing’ (i.e. see Fig. 234b) – when the complete assembly either has to have material removed, or added at a certain part of its assembly. NB The major problem associated with ‘hard-balancing’ is that if the tooling setup changes, so will the likely rotating mass change, which will mean modifying the amount of material to be either added, or subtracted from this newly-distributed mass,

• ‘Adjustable balancing rings’ (i.e. see Fig. 232) – by rotating the twin lower and higher balance rings

469

either clockwise, or anti-clockwise they minutely modify the balance-condition, allowing singleplane balance to be achieved. NB These matched pair of balance rings are in a symmetrical state of unbalance (i.e. they are both ‘unbalanced’ to the same degree). Letting the user adjust the pair to counter any unbalance in the cutting tool/toolholder assembly and locking them into place – usually achieved on commerciallyavailable balancing machines (i.e. see Fig. 234a). The state of unbalance is not merely a subject to the ‘caprice’ of the machine tool operator, a tool assembly’s balance is given by various quality Standards, such as ISO 1940/1, or ANSI S2.19 – being basically exact reflections of each other. In the following related sections, they deal with how and in what manner rotating cutter assembly balance is achieved, utilising such HSM balance calculations and associated graphical details as necessary, from these Standards.

9.5.1 HSM – Problem of Tool Balance Unbalance of a rotating body (i.e. here we are concerned with a complete tooling assembly), can be defined as: ‘The condition existing when the principal mass – axis of inertia – does not coincide with its rotational axis’ (i.e. shown schematically in Fig. 232). For example, such an undesirable state of affairs can be comprehended by considering the following situation: if a φ50 mm face mill assembly is rotated at 15,000 rev min–1, it will produce a peripheral speed >240 km hr–1, which may prove to be disastrous if it is unbalanced! Basically there exists, three types of unbalance conditions for rotating assemblies – such as tooling, these are: 1. ‘Static unbalance’ – single-plane. This type of unbalance occurs when the mass does not coincide with the rotational axis, but is parallel to it and the force created by such unbalancing, is equal to the magnitude at both ends of the rotating body. Thus, if some relief – metal removal (i.e see Fig. 234b) – on the toolholder body equal to the out-of-balance mass that occurs, then a nominal static unbalance is achieved, 2. ‘Couple unbalance’ – Under these circumstances, the cutter assembly – mass axis – does not coincide

470

Chapter 9

. Figure 232. The taper fitment against runout/eccentricity for a milling cutter and its associated balanced toolholder


471

with the rotational axis, but intersects it at the centre of gravity of the ‘assembly’s body’. Under such conditions the force vectors equalise, but are 180° apart. 3. ‘Dynamic unbalance’ – dual-plane. Such a condition of the toolholder assembly arises when the axis does not coincide with the rotational axis and is not either parallel to, nor intersecting this axis (i.e. see Fig. 232). For any rotating tooling assembly, estimating the cutter unbalance is possible using the following variables: M = cutter/holder mass, S = mass centre, e = displacement of mass centre, r = distance from centre of tooling, to the centre of gravity of mass (m), ω = angular velocity, m = mass unbalance, U = cutter unbalance, 9549 = a constant. Determining the relative unbalance (U) of a rotating tooling assembly, can be found by the following expression(s): U = M × e or, alternatively: U = m × r

(i).

It is usual to express unbalance in terms of the product of the mass times distance, typically using the units: ‘g-mm’. Finding the magnitude of centrifugal force produced by the rotating tooling assembly with a given unbalance, can be established as follows: F = U × ω2

(ii).

Where: ‘ω’ is the angular velocity in units of radians sec–1. The formula to find ‘ω’ is expressed by: ω=

 � π � r pm 

(iii).

Therefore, by combining formulae: (i) and (iii), in (ii), we can obtain the magnitude of centrifugal force ‘F’ , as follows: F = m × r × ( 2 × π × rpm/60)2

(iv).

As established in equation (iv), the centrifugal force caused by tooling unbalance will increase by the

‘square of the speed’ , in a similar manner to the spindle nose taper swelling (i.e. growth) previously mentioned. Nonetheless, assuming that this specific toolholder initially has a low unbalance, this will become a problem if the rotational speeds are increased beyond 10,000 rev min–1. For example, with most toolholders exhibiting single-plane unbalance25, research experimentation has shown that the initial unbalance of a typical tooling assembly will be of the order: 250 gmm. When such tooling is rotated at 15,000 rev min–1, this 250 g-mm of out-of-balance develops a continuous radial force of 642.6 N. Unbalanced tooling can introduce considerable detrimental effects on not only the machine tool – this high centrifugal force causing internal bearing stresses leading to premature spindle failure, but affects cutter life and degrades workpiece surface texture. Much of the principal tooling unbalance problems can be traced-back to several sources, such as: • Toolholders of the V-flange type, which might have different depth of drive/slots, these toolholder features being part of the inherent design, • Toolholders for some end mills and slot-drills, having set screws for locking the cutter securely in place, so due to necessary clearance and the radial application of the set screw, this creates minute cutter eccentricity – causing unbalance, • Out-of-balance caused by an unground V-flange base, • Collet and its collet nut tend to be recurring sources of unbalance in HSM tool holders. NB Most of these tool holding-related issues can be eliminated by simply modifying the tooling design. As can be seen from Fig. 232, the marginally eccentric adjustable balance rings can be rotated to adjust the degree of single-plane balance, with several of the tooling manufacturers offering differing adjustment methods for HSM toolholders. Finally, consideration needs to be given to the level of balance-quality required and in HSM applications for example, a milling cutter is expected to withstand

25 ‘Single-plane unbalance’ , relates to the type of unbalance that occurs in either one of two planes. Namely, the tooling assembly’s single-plane unbalance will be in either its axial, or radial directional plane.

472

Chapter 9

both high rotational speeds and associated cutting forces, thus here it can be considered as a ‘rigid rotating body’. This assumption allows one to use the ANSI S2.19-1989 Standard, for achieving balance – see Fig. 233, which defines the permissible residual unbalance of a rotating body relative to its maximum speed. This Standard and its equivalents (e.g. ISO: 1940:1; ISO: 1290 G), assigns different balance-quality grades termed: ‘G-numbers’ , related to the grouping of rotating bodies (i.e. not shown), these groupings being based upon the experienced gained with a variety of: sizes; speeds; and types. Thus, the balance-quality grade ‘G’ , equals the specific unbalance ‘e’ times the rotational speed ‘ω’ , as follows: Balance-quality G = e × ω (mm sec–1). Furthermore, the equation was described earlier, thus: e=

U M

(i)

∴ solving for ‘U’ , we obtain:

U=

 � M � G r pm

(v).

From the Standard, the balance-quality for machine tool drives is given as: G2.5, although in many instances the value utilised should ideally approach that of G1.0 – this being the specification for grinding machine tool drives, as today in HSM applications they are compatible. However, if for the purposes of clarification of the unbalance tooling condition the value of G2.5 is utilised, then the following worked example illustrates the balance-quality necessary using a toolholder weighing 3 kg, rotating at 25,000 rev min–1:

U (higher) =

 �  � . (g-mm) , 

∴ U (higher) = 2.85 g-mm.

As alluded to previously, this unbalance condition is the ‘worst case’ and the tooling should ideally approach G1.0, this balance-quality value, gives: U (lower) =

 �  � . (g-mm) , 

∴ U (lower) = 1.14 g-mm.

This then follows that the balance is between 1.14 and 2.85 g-mm, which is toward the ‘upper-end’ for the

maximum residual specific unbalance for the G2.5, while approaching this level for the G1.0 (i.e shown by the graph in Fig. 233). Even when the tooling assembly has been dynamically balanced in both planes (i.e. see Fig. 234a – more to be said on this topic shortly in Section 9.5.2), problems still exist, particularly in the fit of the spindle taper connection (Fig. 232). This is a result of the taper rate accuracy requirements between both the shank and taper socket. In fact, the situation is quite a confusing one, due to the relative cone ‘Angle Tolerance’ grades: AT-1 to AT-6, that are employed using the conventional fitment of: 7:24 taper. Not only do different countries often have their own connection Standards, but previously, even individual machine tool manufacturers within each country had adopted differing Standards! Today, many machine tool companies tend to utilise taper spindle connections that are compatible to an appropriate Standard and complement those of the tooling manufacturers.

9.5.2 HSM – Dynamic Balancing Machine Application It has been discussed in the previous sections that cutting tool assemblies when combined with an HSM strategy, can be a large contributor to dynamic unbalance. For instance, in the production and manufacture of say, the geometry of a face-mill, the tooling stock material is: externally/internally turned on one side; unclamped; flipped-over and rechecked; then turned on the other side; then located onto a milling machine tool for operations on the individual insert pockets that must be milled; and indexed26 – as appropriate for the number of cutting edges; this necessary clamping/reclamping workpiece (i.e. face-mill) procedure, will create a tool that is marginally-unbalanced. With HSM, the otherwise unnoticeable unbalance at conventional rotational speeds, becomes intolerable in these high-speed ranges. Often, the most economical technique for achieving balanced tooling for tooling

26 ‘Insert pockets’ , are sometimes ‘differentially-pitched’ which means they have unequal spacing of teeth around the cutter’s periphery. This pitching technique for cutting insert pockets, is quite effective as a means of reducing machining vibrational effects often encountered with coarse-pitched face-mills.


. Figure 233. A graph to determine high-speed cutter unbalance ‘U’ (ANSI S2.19–1989)

473

474

Chapter 9

designated for the HSM ranges, is by dynamically balancing the tools in an appropriate machine (Fig. 234a). Thus, during tool balancing, the cutter is clamped in a fixture that rotates in very rigidly supported bearings (Fig. 234a). Any unbalance in the rotating cutter is directly measured as centrifugal force, which is transmitted along with its actual rotational position to a specially-configured computer. The computer calculates the amount and location of the material to be removed from the tool’s body, in order to properly adjust the mass distribution. This unwanted tool material can then be removed, by either drilling holes, or milling flats (i.e. see Fig. 234b – illustrating that a small amount of material has been removed – milled – from its flange and is termed ‘Hard-balancing’). Usually, flats are preferred for removing larger amounts of tool material stock, because at high rotational speeds a tool with a hole can generate a unacceptable ‘whistling noise’: see Fig. 235 to gain an appreciation of the effect of these high peripheral speeds, here, a large face-mill is shown in an HSM automotive application. Since the ‘balancing operation’ can compromise the tool’s strength and performance, it is important to establish where any superfluous tool material can be safely removed from on the cutter’s body. Balance is a ‘zero-quantity’ , so it is customary to measure balance in the absence – within acceptable limits – of unbalance. Thus, the ‘balance tolerance’ is the maximum residual unbalance (g-mm) allowed for a particular tool’s weight and rotational speed. For example, the ANSI Standard quality-grades for balance tolerance range from G0.4 to G6.30, with the lower the ‘G-number’ the closer the balance tolerance. It should be emphasised, that only when a balanced tool and its balanced toolholder are balanced together as a complete unit, are they truly dynamically-balanced27.

27 ‘Dynamic-balancing’ postscript: the cutting inserts, screws wedges that are retained in the cutter’s pockets must be securely locked into position. If these small items become detached when HSM, this may cause disastrous consequences for any operator in the vicinity. Therefore, machine tool guarding of more than adequate protection is vital here, to minimise potential safety hazards to the personnel when in use.

9.6 HSM – Research Applications Introduction Possibly the foremost reason for conducting applied research programmes at various universities and similar research-based organisations, is to ‘push the boundaries’ of our theoretical and practical understanding of current, or novel machining concepts. Rather than attempt to give a perhaps less-than-informed account of what is transpiring at other ‘learned organisations’. It was decided just to deal with some recent work that was undertaken by the author, in association with industrial and academic colleagues, both in the UK and abroad. As these particular research projects were either undertaken, or instigated by the author, often in close union with others, mainly concerned with industrial-sponsored doctoral programmes, it was felt that here at least, some continuity concerning HSM-related research themes could be achieved.

9.6.1 Ultra-High Speed: Face-Milling Design and Development Introduction In order to achieve a cutting speed of say, 1,000 m min–1, a φ10 mm milling cutter has to rotate at approximately 32,000 rev min–1. In fact, this is quite feasible for smaller diameter cutters, as there are quite a few of today’s machine tools having HSM spindles that can exploit these speeds. However, it is worth a sight digression here, prior to continuing with the theme concerned with special-purpose UHSM by face-milling, to ask the pertinent question: ‘What do we mean by HSM?’ Is it: • High rotational speed machining? • High cutting speed machining? • High feed machining? • High speed and feed machining? • High productivity machining? Even these five potential contenders for what amounts to HSM, are by no means an exhaustive list, one could also add more obscure factors such as: the power demand on the spindle; tool assembly balance speeds;


. Figure 234. Dynamic dual-plane (i.e. radial and axial) cutter balancing. [Courtesy of Ingersoll]

475

476

Chapter 9

. Figure 235. High-speed large diameter face (finish) milling on a grey cast iron engine block. [Courtesy of Sumitomo Electric Hardmetal Ltd.]


the taper size in relation to its rotational speed; thinwalled machining capability; etc.; the list will grow, depending upon what we consider to constitute as going either very fast, or what speed allows us to machine certain types of part features! Although even here, the major benefits of an ultra-high speed machining strategy are somewhat lost if a ‘working definition’ is not clearly stated. In this current discussion on the subject, one measure of HSM could be if the peripheral speed of the cutter, or workpiece is >1,500 m min–1. Hence, this could represent a base-line for the transition from conventional to HSM, so for comparison, as in the case for previously mentioned small diameter milling cutter of φ10 mm, it would need to rotate >47,750 rev min–1, or conversely, a larger face-mill of say, φ300 mm, would have to rotate at approximately 1,600 rev min–1 – in order to sustain a peripheral cutter speed of at least 1,500 m min–1. This latter rotational speed although considerably slower, is a much greater problem that that presented for the former small diameter cutter. The reasons for this are three-fold: firstly, has the machine tool got enough spindle power to achieve the necessary stock removal rates required, or will it be likely to stall? Secondly, is the spindle taper fitment robust enough to cope with the torque effects and bending moments imparted during machining? Thirdly, will the cutting inserts still be retained at the high centrifugal forces generated in association with and exacerbated by the imparted cutting forces? These and other lesser important questions and decisions need to be addressed, if the large diameter face-mill is to successfully mill across a wide workpiece surface with any degree of efficiency. This former point of the manner in how workpiece surface stock is removed is important, for example, two markedly differing machining strategies could be adopted, such as: I. Shallow depths of cut combined with rapid traverse rates and small step-overs, utilising smaller diameter cutters at high peripheral speeds, II. Deep and wide cuts with a large diameter facemill with slower traverse rates, having much lower rotational speeds. NB Both machining strategies will remove similar amounts of part stock!

477

range, a number of critical features need consideration, such as: cutter-body material; its weight and rigidity; taper fitment; dual-plane balancing; as well as its aerodynamic behaviour – at fast peripheral velocities. If one attempts to design and develop a large facemilling cutter with an insert cutting circumference designed to rotate at 3,000 m min–1, which at first glance, may not seem that fast. However, if we equate this cutting speed to that of the same φ10 mm milling cutter previously mentioned, then this smaller tool would have to rotate at ≈95,500 rev min–1, but for the larger diameter cutter, it would also require to be dual-plane balanced28. Without dynamic balancing, the large cutter may be prone to a disastrous series of vibrational problems, which may ultimately lead to premature cutter failure – with all its attendant safety hazards. The cutter design in Fig. 236 for this applied research programme, was dual-plane balanced to Standard defined in ISO:1940/1, being to a specific ‘G-number’. This Standard was initially conceived for the rotational balancing of impellers and similar high-speed equipment, across a large speed range. The large face-mills had been dual-plane balanced to G2.5 @ 10,000 rev min–1. As mentioned in Section 9.5.1, this ‘G-number’ refers to the maximum tolerable imbalance for the complete tooling assembly, being based upon the previously described formula (i.e given below again for clarification), resulting in the following calculations: Unbalance: U =

 � M � G (g-mm) N

[or] Force: F = U �

N  � (N) 

Where: U = allowable unbalance (g-mm); 9549 = a constant; M = mass, or weight of the total cutter assembly (kg); G = pre-selected balance tolerance number from ISO: 1940/1;

UHSM: Face-Milling Cutter Design When designing large diameter face-milling cutter assemblies for production applications in the UHSM

28 ‘Dual-plane balanced cutters’ , are those cutters that have been dynamically balanced in two planes: having both radial and axial balance.

478

N F

Chapter 9

= maximum rotational speed of tooling assembly (rev min–1); = force (N).

When the above equation for cutter unbalance is utilised, for these tooling assemblies, utilising the following values: G = 2.5; M = 4.294; N = 6,000; which then gave an allowable imbalance U = 17.9 g-mm. This level of imbalance means that the cutter’s mass cannot rotationally shift by more than 17.9 g-mm, if it is to maintain dual-plane balance at a peripheral speed of at least 3,000 m min–1. In fact, two identical cutter assemblies were designed and manufactured, having a BT40 taper fitment – for the vertical machining centre (i.e. Cincinnati Milacron Sabre 500). In order to maintain both structural rigidity and integrity, the complete cutter bodies and their associated tapers were each produced from single stock of EN24T steel. After precision turning and milling the complete bodies and insert pockets, these cutters were nitridehardened29 to HRC 52, prior to a ‘very light-grinding’ process and then balancing. The four cutting insert pockets were equally-spaced (pitched) and the button-style cemented carbide inserts were: φ12 mm by 4mm thick, single-sided and TiN-coated (Stellram: RPET 1204 DFZ). The insert pocket geometry had a 11° toe angle with a neutral geometry. Button inserts were selected as they give the strongest shape cutting geometry available (see Fig. 23), producing an infinite approach angle to the workpiece (see Fig. 83b), thus minimising impact load at entry to the cut while offering multiple cutting edges – when subsequently turned in their seatings. As one might expect, insert security is vitally important, due to the great centrifugal effects and applied cutting forces. Due to the previous nitride-hardening process, hardened insert seats were unnecessary, once the retaining screws had been ‘torqued-up’ locking and then sealing them – for se-

29 ‘Nitride-hardening’ , produces a very hard surface with a softer and tougher matrix. The UHSM cutters were held in a pressure-tight furnace and heated to between 500-550°C for some hours in an ammonia gas*, allowing the nitrogen atoms to diffuse into the surface and to form fine stable nitride precipitates with aluminium constituents, allowing the nitridedsteel surface to be precipitation- hardened. No subsequent heat-treatment is necessary. → *Approximately 30% of the ammonia disassociates ( NH3 ← 3H+N) and part of the nascent nitrogen is absorbed by the surface layers of the steel. (Source: Cotrell et al., 1979)

curity. The actual seatings for the inserts had considerable body-support around their periphery, just having a working-clearance at the insert’s cutting region. These UHSM face-mills were extremely compact, with the minimum stand-off height from the cutter’s gauge line (i.e. see Fig. 236), which reduced the effects of the previously mentioned ‘rigidity rule’. Due to the relatively large diameter and weight of these face-mills and the fact that the machining centre had limited spindle power, these cutters could, if used appropriately, exploit the ‘mass’ , or ‘flywheel-effect’ of their weight in conjunction with rotational speed to ‘store inertia’. So, when the spindle power is restricted, cutters with high mass must be taken up to their desired rotational speed in a progressive manner, otherwise they are likely to ‘trip’ a ‘spindle over-load’ in the CNC controller. This steady and progressive increase in the cutter’s rotational speed occurred at 500 rev min–1 increments – dwelling for several seconds to minimise inertial power overload, between increases to the desired peripheral speed. Due to the machine tool having a maximum spindle speed of 6,000 m min–1, this equated to a peripheral cutting speed of 3,000 m min–1, with the face-mill having 0.5 m cutting circumference. Once the cutter has reached its top speed, it can then be rapidly progressed (i.e. fed) across the workpiece at a rate of 20 m min–1. In this case the workpiece materials were a range of stainless steel alloy testpieces (Fig. 236). After rapidly face-milling these ‘stainless testpieces’ , the cutter’s rotation was decremented in 500 rev min–1 intervals until stationary. The cutter once stationary, could have its edge wear (inserts) assused and workpiece milled surface texture and surface integrity could be inspected and investigated. One factor to bear in mind concerning UHSM with large face-milling cutters being utilised for their ‘inertial effect’ , is to design them without driving dogs. If these ‘dogs’ were fitted, not only can they introduce out-of-balance effects, but tend to significantly disrupt the air-flow and introduce alarming and high noise factors. This aspect of cutter design is important, if the cutter cannot ‘cleave through the air’ with aerodynamic efficiency, turbulent air flow will result and operational noise becomes excessive. One problem that these particular cutters did not suffer from (i.e despite the conventional taper-cone angle) – unlike many of their higher rotational speed counterparts, was ‘spindle nose swelling’ , which can cause a lack of register if the taper fitment connection is not of either the double-, or triple-contact face-and-cone types. One unexpected aspect of employing such large face-mills


. Figure 236. A specially-designed dual-plane (radial and axial) face mill, for ultra-high-speed milling. [Source: Smith, Wyatt & Hope, 1998]

479

480

Chapter 9

in UHSM, was found to be due to the very high peri pheral speed. A ‘suction-effect’ resulted from the underside clearance of the cutter’s body, created a ‘lowpressure region’. This ‘virtual vacuum’ here, meant that conventional-pressure flood coolant application was not possible, as it simply vaporised to a mist!

UHSM: Cutting Trials The neutral geometry enabled the cutting inserts to present a strong cutting edge to these stainless steel testpieces, enabling an undistorted cut path and machined cusp to be generated. This insert geometry feature, allowed the milled surface topography to be unaffected by insert inclination angles. A stringent test for any cutter is to machine stainless steel by UHSM (ie. being least × 10 faster than any work previously undertaken) and here, tests were conducted on various grades (Fig. 236). The subsequent milled surface analysis showed little in the way of sub-surface plastic deformation – after UHSM. The surface layers exhibiting only marginal increases in the vicinity of the surface when tapered sections were micro-Knoop ‘foot-printed’ , over these stainless steel’s substrate (i.e. see Footnotes: 15, Chapter 1; 85, Chapter 7; and Fig. 187c – concerning Knoop indentors usage). Milled surface topography can be viewed visually and surface parameters taken by either: 3-Dimensional contact; or non-contact instruments; but in this case, by utilising an SEM30 with its unique ‘Stereo-imaging and topog-

30 ‘Scanning Electron Microscopes’ (SEM’s), operational principle is relatively simple. In that, at the top of the SEM’s column an electron gun resides having a tungsten filament held in a strong electrical field. This results in the electron gun emitting electrons (i.e. negatively-charged atomic particles), which accelerate to very high speeds. These high speed electrons – held in a vacuum – travel down the column, being influenced by lenses lower in the column, which squeeze them together to form an electron beam of very small diameter. This minute diameter electron beam is then focussed prior to colliding with the test specimen in the microscope’s specimen chamber, now as a diminutive spot. This minuscule spot will then scan both to the left and to right as well as up and down over the test surface, the information from which is then brought to a screen as an image. Prior to this, as the electron beam strikes the test sample’s surface, many different processes occur, such as: secondary electrons; backscattered electrons; Auger electrons; X-rays; Cathodo-luminescence; etc.; these being emitted, collected and counted, then utilised for further analyses.

raphy software’ – for 3-D visual assessment coupled to its height-to-depth profiling application. For the milled testpieces produced from 316-austenitic stainless steel, subjected to UHSM by this facemill at 3,000 m min–1, the surface topography showed the influence of the wear land flat produced by the four φ12 mm TiN-coated inserts, although the remainder of periodic surface offered little sign of any surface modification. The 303 stainless steel grade testpieces, indicating a slight improvement over the former 316 grade. While, 416 martensitic stainless steel testpieces under identical cutting data generated no appreciable surface blemishes, with the additional benefits of: an extended cutting insert life; significant reductions in both cutting forces and power requirements, over the 303 and 316 stainless steel grades.

9.6.2 Ultra-High Speed: Turning Operations Introduction As has been shown in the previous sections with reference to HSM by milling, considerable applied and fundamental research effort has occurred, conversely, little endeavour has been made regarding high-speed turning operations. Possibly the major reason for the lack of interest here into HSM by turning operations, is because a different approach to the workholding issues needs to be taken. In that, on a CNC turning centre, or lathe, the ‘bursting-pressures’31 resulting from significant centrifugal forces with conventional work-

NB The depth of fields from an SEM are considerably deeper than that produced by conventional microscopes, allowing some exacting surface topography analysis to be undertaken. (Source: Smith et al., 2002) 31 ‘Bursting-pressure’ problems, have been well-known in traditional turning activities for many years. This aspect of safe-working practice, was particularly relevant for large cast iron face-plate work, where a rotational speed limitation is imposed by a machine tool builder. If this restricted speed is exceeded, then the cast iron – being poor in terms of tensile strength, will literally fragment (i.e. ‘burst’), due to the excessive centrifugal forces imposed. While, the problem is not as severe for chuck and collet work, the lack of gripping-pressure on the part – at high rotations, will affect the workpiece if long slender parts supported on one end only are turned – possibly causing a ‘whipping-effect’ and attendant safety hazard.


. Figure 237. An ultra-high-speed turning operation undertaken on a vertical machining centre. [Source: Smith, Littlefair, Wyatt & Berry, 2003]

481

482

Chapter 9

holding: chucks and face-plates; in combination with a reduction in gripping-force at high rotational speed is a potential safety hazard. In order to achieve the desired UHSM rotational speeds of >1,500 m min–1, which can be considered as the ‘threshold’ for such a machining strategy, then the workpiece must be held in a machining centre spindle – with the tool stationary, in a similar manner to vertical turning (i.e. see Fig. 237). Relatively recently in pioneering work by Yousefi and Ichida (2000), they utilised rotational speeds up to 15,000 m min–1 by this technique, showing that the turned surface texture parameter ‘Ra’ values (i.e see Section 7.5.1 in Chapter 7) steadily reduced with increases in cutting speed. Moreover, this Japanese machining study found that the cutting forces remained relatively constant throughout the test range of: 1,200 to 15,000 m min–1. This trend of reasonably constant turning forces with increased speed is contrary to that normally found for UHSM by milling operations (Komanduri, 1995, et al.), where a significant reduction in milling forces results from high cutter rotations, allowing large length-to-diameter cutter ratios to be used (Gough et al., 1991). In any HSM operations, it is essential that the rotational mass of either the workpiece, or cutter – depending upon which is the rotating item, is dynamically balanced, in order to minimise out-of-balance effects which would otherwise impede both the cutting process and affect machined surface texture. Ideally, workpieces, or a cutter should be rigidly held in-situ during machining and at the very least, be single-plane balanced.

UHSM: Turning Strategy Prior to undertaking the UHSM turning operations (Fig. 237), the machining centre was checked for diagnostic errors by a ‘Telescoping Ballbar’ 32 assessment

32 ‘Telescoping Ballbar’ (Fig. 242a), is a powerful instrument for machine tool error diagnostics. The Ballbar as its name implies, is a ball-ended length transducer (i.e. an LVDT- measuring element, is positioned between the fixed and telescoping balls). This LVDT has a range of ± 0.75 mm with a resolution of 0.1 µm and accuracy of 1 µm. It is held in kinematic (magnetic) seatings between the machine spindle and its base. Extension bars can increase the radial length up to 300 mm, covering a large volumetric sweep for the two axes being diagnostically monitored. Any kinematic plane can be rotationally swept by the Ballbar. In operation – from the software program, the machine’s CNC will move the Ballbar to the start position (i.e. radially offset the required distance for the orien-

set at the radial turning distance in the plane of the cut (i.e. see the Ballbar configuration in Fig. 242a). A range of rotational Ballbar speeds were utilised, albeit at considerably lower peripheral speeds than those which were employed for the UHSM turning trials. For this current UHSM work, a machining strategy was adopted utilising a ‘Variable quasi-pilgrim stepped arithmetic progression’ (i.e the progression is schematically depicted in Fig. 238), for the selection of ‘turned testpiece’ rotational speeds, being based upon the following general progression case criteria: Sn1 → Sn4 = 2000 + {n1/2 [2a1 + (n1 – 1) d1 ]} – 1000 + {n2/2 [2a2 + (n2 – 1) d2 ]} – 500 + {n3/2 [2a3 + (n3 – 1) d3]} – 250 + {n4/2 [2a4 + (n4 – 1) d4]} – 125 + … Snn Where: a1 = 2,000, a2 = 4,000, a3 = 5,000, a4 = 5,500; d1 = 1,000, d2 = 500, d3 = 250, d4 = 125; n1 = 4,000, n2 = 2,000, n3 = 1,000, n4 = 500. Such an unusual ‘progression’ mathematically described above (i.e also being shown schematically in Fig. 238), enables significant discrimination of rotational results coupled to data analyses toward the upper limit of the UHSM turning process, while giving ‘traceability’ to rotational speeds within the conventional range of the rotational turning process at the beginning of the turning process. A special-purpose workholding device – being dual-plane balanced to G2.5 @ 10,000 rev min–1, was fitted into the spindle of a vertical machining centre (Fig. 237). This BT40 tapered workpiece holder was constructed from one-piece of EN24T steel hardened by nitriding to >50 HRC, weighing ≈2.5 kg with the aerospace-grade aluminium disk-shaped testpiece in-

tated planes), then it slowly rotates CW for 180° – to pick up uniform rotational velocity, where it rotates through a further 360° – for polar measurement, finally rotating another 180° – to slow down. This complete cycle is then repeated CCW. Then polar plots are generated with a diagnostic printout, which ‘ranks’ these errors, so that they can then be eliminated, or significantly reduced, accordingly. This is a speedy, efficient diagnostic ‘health-check’ of the machine tool errors in the two measured planes, providing significant information, which can be utilised to improve the machine tool’s overall performance. (Source: Renishaw Ballbar Training Manual) NB Typical ‘polar plots’ are shown in Appendix 16, together with a diagnostic print-out of the results.


. Figure 238. A ‘variable quasi-pilgrim stepped arithmetic progression’ – being utilised for UHSM (turning). [Source: Smith, Littlefair, Wyatt & Berry, 2003]

483

484

Chapter 9

situ. These pre-shaped testpiece disks were: φ300 mm by 6 mm thick, made from aluminium 2017F. The tool with various cemented carbide tooling insert grades, was held on a platform dynamometer (Kistler model: 9257B) – having complemntary charge amplifiers coupled to suitable data analysis software. A range of DOC’s were utilised: 0.5, 1.0 and 1.5 mm, with a constant and rapid feedrate of 30 m min–1. Apart from cutting force analysis, turned surface texture, harmonic roundness and micro-hardness results were obtained, together with metallographical inspection of the subsurface regions. Hence, from this testpiece setup and utilising the ‘progression’ for peripheral workpiece speed strategy described above, the speeds ranged from the conventional, through to UHSM.

UHSM: Turning Trends Unlike the previous findings of Yousefi and Ichida (2000), where they suggested that the cutting forces remained relatively constant across a broad spectrum of UHSM – for turning operations. This UHSM turning work indicated that there was a decrease in mean cutting forces between 2,000 to 6,000 m min–1, with a corresponding improvement in turned surface texture (i.e. ‘Ra’) across this range. The harmonic departuresfrom-roundness were influenced by the sinusoidal effect of the fluctuating tangential force as it progressed around and along the turned surface’s periphery. This harmonic behaviour was evident in the cutting force data, where the analysis software showed both a rising and falling relationship, as the turning insert passed over the rotating workpiece’s surface at great peripheral speed. Such cutting force traces occur in high-speed interpolation by the milling process, where there is a general undulating increase/decrease in force generation, this being related to the axis transition cross-over during cutter interpolation around the workpiece (i.e. see Fig. 159). This cutting force undulation is the result of the machine tool’s servo-motors reversing direction at these transitions, albeit, at significantly slower speeds than utilised for this UHSM turning work. At such high turning speeds, chip-streaming was apparent at peripheral speeds >4,000 m min–1. Chipstreaming in UHSM by turning is the preferred chipform, as it exhausts the work-hardened swarf away from the cutting vicinity, thereby minimising entanglement around the newly-formed turned workpiece surface. At such high turning peripheral velocities, the chip-streamed swarf is directed radially-away

from the work surface. Conversely, at lower rotational workpiece-to-insert velocities, there was a marked tendency for ‘chip-curling’. As a result of the influence of the insert’s geometry: nose radius and DOC relationship, the so-called ‘theta-effect’ in conjunction with the feed-per-revolution occurs (i.e see Figs. 34c and d). There is a direct and predictable relationship to chipcurl tendency when certain conditions arise at the lower peripheral turning speeds, which is not apparent at the UHSM turning range. This UHSM turning applied research work, has shown that it is feasible to employ ultra-fast turning practices to the relevant components if the correct tooling, workpiece and machine tool relationships can be met.

9.6.3 Ultra-High Speed: Trepanning Operations Intoduction Trepanning has been a well-recognised production process for many years, it is principally utilised to produce large hole diameters, since this technique does not require as much spindle power as solid drilling. Moreover, in the ‘conventional’ approach to trepanning, it is undertaken in one operation, but instead of all the workpiece material being removed in the form of a large volume of swarf, a cylindrically-shaped core is left behind at the centre of the hole. Thus, this method must be utilised for through-hole applications, assuming that the internal feature – hole manufactured is the scrap material. Conversely, in the UHSM trepanning work shortly to be discussed, the two cutting edges are externally set against the workpiece’s periphery, making the ‘slug’ the product of the machining operations (Fig. 239c).

UHSM – Trepanning Fix ture Design As in the case of vertical turning, UHSM by trepanning was undertaken on a vertical machining centre utilising the same workholding arrangement (Fig. 239). Here, a special-purpose trepanning fixture 600 mm in overall length, was designed and manufactured with twin-opposing tools (Fig. 239a). The tooling was conventional TiN-coated cemented carbide turning inserts – having straight toolholders these being positioned on their sides in opposing directions


485

. Figure 239. An ultra-high-speed trepanning fixture and dual-plane balanced workpiece holder: utilised for an UHSM research programme of work. [Source: Smith, Hills & Littlefair, 2005]

486

Chapter 9

(Fig. 239b). So, by the simple action of turning a hand wheel at its end, the tools could be simultaneously opened and closed – for the required trepanned diameter. This simultaneous tooling action was achieved, by the singular rotating action of both the φ20mm by 4 mm pitch left- and right-hand (i.e. M 20 × 4) squarethreaded leadscrews (Fig. 239c). One of the major advantages of an UHSM trepanning operation over its equivalent turning counterpart, is that the cutting forces are virtually ‘cancelled-out’ , in a similar fashion to a conventional ‘balanced turning’ operation (Figs. 41 and 238 – top right). Here in this instance, one tool is set and positioned slightly ahead of the other, thereby not only reducing the overall DOC, but allowing the ‘trailing tool edge’ to effectively act as a ‘finishing tool’. This tooling positioning strategy produced an improved trepanned surface texture, while it significantly reduced the harmonic departures-fromroundness, as metrologically assessed later on the roundness testing machine. Moreover, by effectively ‘halving’ the DOC, this allowed for an improvement in the chip-streaming behaviour to be attained. In a later modification to the trepanning fixture (i.e. not shown), a large micrometer drum with its integrated vernier scale was fitted in place of the knurled adjustable hand-wheel (i.e see Fig. 239a), allowing for some considerable discretion over the linear tooling’s diametral adjustment. With such a large trepanning fixture – having the opposing tooling widely-spaced, it is vital that these tools are centralised directly beneath the machine’s spindle. Otherwise, there is a possibility of both sine and cosine errors being present, creating ‘Abbé-type errors’ , when adjusting and setting these tools for their diametral in-feed.

UHSM – Trepanning Operation This preliminary work on UHSM by trepanning, has shown that with a suitably robust tooling fixturing and allowing a large (indirect) range of tooling diameter adjustment – via the twin leadscrews, then not only is the process feasible, but it offers considerably improved machining performance and an inherent improvement in trepanned surface and roundness characteristics, over vertical turning processes. Possibly in a later modification to a heavily-revised tooling adjustment system, it might be possible to employ twin coaxial ballscrews, with CNC servo-control, allowing automatic control for machining tapers and profiling to the workpiece – by utilising the supplementary rotary axis control in the machine’s CNC controller. Moreover, one limitation to this UHSM trepanning technique is

the length of longitudinal cut that can be taken, prior to the Z-axis motion causing the rotating part to foul on the central portion of the trepanning fixture. This problem can be mitigated against, by increasing the relative stand-off height of the twin-tooling from the top of the fixture by mounting each toolholder in an extended tool block, so allowing greater Z-axis feeding to be undertaken. Moreover by rearranging the tools in relation to the workpiece, it would be possible to ‘turn’ shallow, depth internal trepanned features. UHSM by trepanning offers significant advantages over ‘conventional’ vertical turning, in that, in this current work, if was found that the trepanned workpiece surface and roundness were significantly improved from the previously discussed UHSM by vertical turning, described in Section 9.6.2.

9.6.4 Artefact Stereometry: for Dynamic Machine Tool Comparative Assessments Introduction The use of machinable artefacts for the assessment of machine tools such as machining centres, has been utilised for some of years (i.e typically: NAS Standard 979: 1969; ISO Standard: 10791-7: 1997; Knapp, 1997), being developed just for this purpose. Both the NAS and ISO Standard testpieces incorporated notable prismatic and rotational characteristics, manufactured to specific geometric and dimensional tolerances, such as: at the top, an φ110 mm circular feature; 6 mm below this round shape, an 110 mm diagonal feature is cut; a central φ30 mm though-running hole is produced; with a series of counter-bored holes at four equi-spaced quadrants are generated these being situated 6 mm below the diagonal shape. Taken in cross-section, the geometry of the machinable artefacts resembles a stepped component, having an overall height of 50 mm. In fact, this type of artefact has long been employed by industry to establish the overall machining performance capabilities of a particular machine tool under test. However, although this prismatic and rotational featured machinable artefact achieves some measure of conformance and indicates the likely operational performance of the machine tool, it does tend to have several significant limitations, such as the: • Overall dimensional size of the artefact is quite small – when compared to that of the volumetric envelope of typical industrial machining centres,


. Figure 240. Artefact stereometry, illustrating its integrated volume geometries, for a: 1. (right) conic frustum, 2. (right) cylinder, 3. rectangular volume of machine tool’s axes. [Source: Smith, Sims, Hope & Gull, 2001]

487

488

Chapter 9

• Circular feature cannot be directly compared to ric relationships were intrinsically set and datumed to that of diagnostic instrumentation – such as the Ballbar, as the diameter of this rotational feature differs from that of the standard Ballbar sizes, • Weight of the artefact does not realistically compare to any workspaces normally placed on the machine tool in its ‘loaded-state’ , meaning that ‘true’ machine tool loading-conditions are not directly comparable. With these machinable testpiece limitations in mind, it was thought worthwhile developing a new calibration strategy for such machine tools, but here, under more realistic ‘loaded conditions’ , also this new arte fact being more directly comparable to diagnostic instrumentation (i.e. such as the Ballbar), but having considerably larger volumetric size and weight, with the capacity for reuse of the expensively-produced precision part of the machinable artefact’s assembly.

Stereometric Artefact – Conceptual Design Stereometry has been a concept that has often been over-looked, but it deals with the volumetric content of a range of geometric shapes. However, if this ‘volumetric concept’ is carefully integrated into a single artefact, it could be employed for calibration work on machine tools such as machining centres (i.e see Fig. 240). Here, the cylinder was represented by three machinable aerospace aluminium disks (grade: 2017F – produced from 6 mm sheet, to nominally slightly >φ300 mm) each one being set 100 mm apart in height (i.e. disks: 1, 2 and 3) and after machining, the disks were exactly φ300 mm (i.e see Fig. 241). The conic frustum included angle was 22.5°, this being the result of producing 4 equispaced holes in each disk. Starting on the bottom (disk 1), then stopping the machine and fitting the middle disk (disk 2) and drilling the 4 holes and likewise upward to the top disk (disk 3), while simultaneously producing a 3-dimensional Isosceles triangle33 (Fig. 241). Each disk had these individual holes being set at an angular relationship of 90° equi-spaced apart, so, when they are taken as a ‘volume’ , a conic frustum is produced (Fig. 242b). These geometric and volumet-

33 ‘Isosceles triangle’ , has two sides with two angles being equal, but in this case, with the geometry of a right-angled triangle. NB These side lengths and associated angles can be varied, so long as they both (i.e lengths, or angles) remain of identical proportions.

a centrally-machined slot in the base of the precision mandrel. This fact, meant that the exact angular and volumetric relationships remained in-situ, when the stereometric artefact was then taken off the machine tool for subsequent analyses.

Stereometric Artefact – Machining Trials Prior to the stereometric artefact having its machinable disks milled, the initial test machine tool (i.e. in the initial trials on a Cincinnati Milacron Sabre 500 equipped with a Fanuc OM CNC controller) was fully diagnostically calibrated by: Laser interferometry; long-term dynamic thermal monitioring of its dutycycles in both a loaded and unloaded condition; together with Ballbar assessment. Prior to discussing the actual machining of the disks, it is worth taking a few moments to consider the precision mandrel that accurately and precisely locates each disk in the desired orientation, with respect to each other and the machine tool’s axes. This mandrel body was produced from a eutectic steel34 (0.83% carbon), which after through-hardening to 54 HRC, was precision cylindri-

34 ‘Eutectic steel’ or ‘Silver-steel’ as it is generally known, due to its almost ‘shiny appearance’ when compared to other grades of plain carbon steels. In brief, this 0.83% carbon content steel is so-called a eutectic* steel as it relates to the eutectic composition derived from the iron-carbon thermal equilibrium diagram. Producing an 100% pearlitic structure (i.e. hence its ‘metallographic-brilliance’ , or its ‘irridescence’) when viewed under a microscope, exhibiting fine alternate layers of: Fe3C and Fe. To harden eutectic steel, its temperature is raised slightly above the ‘arrest point’ (i.e. arrest point here, equals 723°C, so hardening could be undertaken at ≈765°C) into ‘γ-solid solution’ (i.e. austenitic region), then rapidly quenched and agitated in water to prevent carbon atomic diffusion (i.e undertaken at greater than the ‘critical cooling velocity’), with the carbon atoms now being effectively ‘fixed’ – though not intrinsically part – of the atomic lattice structure. This carbon entrapment, creates intense local strains that block dislocation movement. Hence, the resulting structure is both hard and extremely strong, but also very brittle. Microscopically, the hardened structure appears as an array of random needles, being completely different from the original pearlitic structure. This needle-like structure formed by trapped carbon atoms in an iron crystal lattice is termed, ‘martensite’. Thus, the degree of hardness – after quenching, being proportional to its lattice strain. After hardening, the mandrel needed to be tempered. Tempering is a controlled heat-treatment process to allow some of the trapped carbon to escape from the interstitial spaces between the iron atoms distorted lattice structure, where they eventually form particles of cementite.


cally-ground on the three register diameters, with the top and bottom faces being surface ground. Previous to this heat-treatment and the grinding processes, dowelling datums (i.e. φ6 mm) were drilled and reamed, then 3 equi-spaced tapped clamping holes were produced for each disk, along with a ground tenon groove in the base – all these features being orientated to the geometry of the machines axes (Fig. 241). Several unique features are introduced within the machinable portions of the disks, such as: • These aerospace-grade aluminium disks were milled to φ300 mm diameter, which directly corresponded to the radial path of the Ballbar (i.e see Fig. 242a) – used previously for diagnostic machine tool assessment, ensuring that some degree of correlation occurred between them, • The three Z-plane disk heights of: 70, 170 and 270 mm (i.e. modified from the original design Fig. 241), coincided with both the X-Y plane table position and vertical heights utilised for the Ballbar plots, creating a reasonably large cylindrical volumetric envelope (Fig. 242b). Moreover, the stereometric artefact was both designed and orientated to coincide with the start and finish positions of the Ballbar’s polar traces, • The 4 circular interpolated holes (φ10 mm) on each disk (i.e see Fig. 241), were geometrically positioned to form a three-dimensional Isosceles triangle at the three Z-axis heights for each quadrant of these disks – with the 1st an 3rd holes relating to the axes transition points in the X-Y planes. Thus, each of the interpolated milled holes in the face of separate disk’s, produced the geometric stereometry of a conic frustum, having an included angle of 22.5° – when the angular orientation of the middle disk is ‘software-realigned’ to produce a straightline relationship (i.e see Fig. 242b),

NB The temperature at which tempering is undertaken is critical, thus between 200–300°C, atomic diffusion rates are slow with only a small amount of carbon being released, thereby the component retains most of the hardness. So if higher ‘soaking-temperatures’ are employed (i.e between 300–500°C), then this creates greater carbon diffusion forming cementite, with a corresponding drop in the component’s bulk hardness.

* A eutectic structure is a two-phase microstructure resulting from the solidification of a liquid having the eutectic composition: the phases exist as fine lamellae that alternate with one another. (Sources: Thelning, 1981, Alexander et al., 1985; Callister, Jr. et al., 2003)

489

• Overall weight of the mandrel and three disk assembly was 38 kg, consequently, this could be considered as a realistic ‘loaded condition’ for the machine tool to operate under, from a practical sense.

In order to minimise the milling forces on the machinable disks, HSM was employed using a spindlemounted ‘Speed-increaser’35 (Fig. 243a) equipped with a φ6 mm slot drill. The HSM speed-increaser was operated under the following conditions: 18,000 rev min–1; at a circular interpolation feed of 750 mm min–1; with the disks having 1 mm of excess stock for each machinable disk – to be milled by circular interpolation. In Fig. 243a, the last machinable disk has been located and clamped and the whole mandrel-and-disk assembly was nearing completion, having previously had its φ10 mm quadrant-positioned holes for each disk machined by small circular interpolated motions by the slot drill (i.e. see the sectional details of the φ10 mm hole geometry in each disk’s quadrant co-ordinates, as illustrated in Fig. 241).

Stereometric Artefact – HSM Results After HSM by milled interpolation on the vertical machining centre, the complete artefact with its machinable disks in-situ, was carefully removed from the machine tool, then automatically-inspected for its quadrant hole positions and disk diameters, on an Eastman bridge-type Co-ordinate Measuring Machine (CMM). This CMM having previously been thermally error-mapped, then checked with a ‘Machine Checking Gauge’ 36 (MCG) – prior to artefact inspection. The CMM utilised a specially-made and calibrated

35 ‘Speed-increasers’ , are a means of multiplying the rotational speed of the machine’s spindle, by utilising a fixed relationship geared head. Here, this actual speed-increaser had a 3:1 gearing ratio, equating to a top speed of 18,000 rev min–1, when it is operating at the top speed for this particular machine tool (i.e. 6,000 rev min–1). NB Normally, these HSM milling/drilling geared heads are limited to a certain proportion of running time per hour at its top speed, as they could over-heat and thereby damage the bearing/gearing mechanism. 36 ‘Machine Checking Gauge’ (MCG), is utilised to check a CMM’s repeatability and accuracy and to detect for any potential ‘lobing-type errors’ from the ‘triggering-positioning’ mechanism of the touch-trigger probe, these being invariably used on such machines.

490

Chapter 9

. Figure 241. Artefact stereometry was designed for the volumetric and positional uncertainites on machining centres, by: HSM interpolation of machinable disks. [Source: Smith, Sims, Hope & Gull, 2001]


491

. Figure 242. HSM (milling) of three machinable disks in-situ on a stereometric artefact, on a vertical machining centre. [Source: Smith, Sims, Hope & Gull, 2001]

492

Chapter 9

cranked-probe – with its calibration obtained from the ‘reference measurement sphere’ being located on the CMM’s table, utilised to inspect the φ10 mm hole geometry and there respective co-ordinate positions. This probe arrangement was swapped for a conventional ‘touch-trigger probe assembly’ to measure the machinable disk diameters – while holding the same cartesian co-ordinate relationships as when it was originally UHM. Later – without ‘breaking-down’ , while still maintaining the same angular orientation, this stereometric artefact assembly was inspected on a roundness testing machine (Taylor Hobson: ‘Talyrond 265’) for individual disk parameters of roundness and cylindricity37 assessment – for the ‘three-disk relationship’. The results of all of these ‘averaged’ roundness measurements and Ballbar polar plots are graphically depicted as histograms in Fig. 243c. When a comparison is made of these results from three individual and completely differing inspection procedures, namely: Ballbar; CMM; and Talyrond, they show some degree of measurement consistency individually, but less so when each disk data is grouped. For example, in the case of the Ballbar, it indicated a 1 µm variation (i.e. range) from the top-tobottom disks, while having a mean value of 17.5 µm. The Talyrond polar plots (i.e. ‘Least Squares Reference Circle’ 38: departures from roundness) also produced consistent roundness results, ranging from 40 HRC). In the USA some industrial trials were conducted into milling such hardened workpieces and, it was reported that by not using flood coolant then tool life was increased by 500% – on average. These trials including

various methods of coolant delivery, via: throughthe-tool coolant holes; coolant grooves; coolant hoses; and for normal- and high-pressure coolant applications; in all cases the tool life was reduced. The main problem with the various forms of coolant delivery it would seem, is the result of the cemented carbide tooling suffering from ‘thermal-shock’ , creating by the high tool/chip interface temperatures and the immediate ‘quenching-effect’ of the coolant application – this ‘thermalcycling behaviour’ occurring at very fast rates. Nonetheless, work-hardened chips in the cutting vicinity must still be evacuated from deep recesses and pockets to avoid the ‘recutting effect’. By using an air-and-mist application – close to the tool’s edge, this will provide a means of swarf removal, while producing some ‘token’ cutting edge lubrication – assuming that ‘coolant-effect’ permissible exposure levels (PEL’s) can be safely dealt, thereby with minimising potential health hazards. Any decisions concerning tool replacement will depend on the users machining needs, with the tool failure generally being apparent by the naked eye, or under low optical magnification – simply observing the cutting edges to determine the ‘wear-patterns’. In-cut, a worn tool’s edges will tend to emit a dull ‘red glow’60, this indicates that excessive forces and heat are being generated in the cutting zone, shortly leading toward a rapid and catastrophic tool failure condition. This ‘vis ual glowing effect’ is initially usually confined to the

60 ‘Tooling – glowing red’ , this temperature-induced machining condition has been widely reported. Trent, 1984, stated: ‘Under very exceptional conditions, when cutting fully hardened steel, or certain nickel alloys at high speed, chips have been seen to leave the tool red hot* – i.e. a temperature of over 650°C’. *This term ‘red hot’ – relating to temperature is somewhat vague, as shown in the chart, for: Variation of colours with temperatures – tempering, stress relief and hardening: Colour:

°Fahrenheit:

°Celsius:

Colour:

°Fahrenheit:

°Celsius:

Colour:

°Fahrenheit:

°Celsius:

Straw yellow

430

220

Light blue

590

310

Faint red

950

510

Light brown

465

240

Grey

615

325

Dark red

1150

620

Brown

520

270

Grey-purple

660

350

Dark cherry

1175

635

Purple

545

285

Grey-blue

705

375

Cherry red

1300

705

Dark blue

565

295

Dull Grey

750

400

Bright cherry

1470

800

NB Temperatures above are either slightly rounded-up, or -down. Conversion: °Celsius = 5/9 (°F - 32)

(Sources: Avner, 1974; Bofors, 1981)

516

Chapter 9

cutting edge corners – where high stresses and temperatures are generated, which can be precisely temperature-monitored by thermographic equipment61, or simply rather crudely, by naked eye observation – with the machine tool’s lights turned out! So, by applying the correct tooling in a consistent and repeatable manner, becomes a vital factor for any form of predictability with all hard-metal machining applications. This is also particularly true for any form of hard-part: drilling; reaming; and tapping operations; where these production processes offer serious challenges to the cutting edges, as the bulk hardness of the workpieces increase to >40 HRC.

9.10 Ultra-Precision Machining Introduction In the last few years, there has been a momentous drive toward producing components and indeed assemblies, significantly more minute than was previously the case. The demand might be to locate and align mechanical parts together in much closer proximity, or perhaps, offering improved functionality and providing enhancement of power-to-weight ratios necessary for electronic micro-circuitry. In fact, the adjacent circuit dimensions for nanometric electronic devices can have a proximity to each other of: 0.000006 mm (i.e 6 nm ≡ 6 × 10–9 m).

61 ‘Thermography’ , utilises the infra-red radiation emitted by a temperature-induced body. These thermographic cameras, offer considerable benefits to any form of actual temperaturemonitoring applications. Thermal gradients, hot-spots and heat losses can be observed – by ‘line-of-sight’ measurements – during actual machining operations. They can also be used to assess temperatures in electrical cabinets, servo-motors, ballscrews, etc., for actual condition monitoring of the machine tools. (Source: Smith et al., 1996) NB Typical temperature ranges for thermographic cameras are: –40°C to >2,000°C with a thermal sensitivity of 0.08°C, making then ideal for some forms of tool temperature monitoring – assuming that the chip-stream is away from the camera’s lens. (Source: Flir Systems™)

The challenges for the whole of the ultra-precision manufacturing industries, are to be able to make and supply miniscule devices that will meet these latest design objectives. Before, discussing the tooling requirements and machining techniques necessary to produce these diminutive parts, often with a high volume demand. It is worth trying to comprehend the ‘true dimensional size and scale’ of these miniature components and assemblies. Previously, the term ‘hair’s breadth’ was often quoted as a very minute dimensional size, but if one looks at Fig. 253a, here, the large circle is supposed to represent the diameter of an actual hair – for comparison. Although even here, a hair is not of uniform diameter. In some very simple comparison tests undertaken about 6 years ago (Smith, 2002). He plucked one of his own hairs from his head – that he could not really afford to miss!, plus four more from several other people in the vicinity! Then, he located this group of hairs within an scanning electron microscope (SEM) chamber and simply measured them. The surprise was that they varied quite considerably, ranging from the smallest hair: at φ30 µm to that of the largest hair at: φ100 µm. So, the diagram (Fig. 253a) indicating that the hair’s size was ≈φ89 µm is somewhat misleading as a form of measurement criteria, as we will begin to appreciate, that the difference of a few ‘microns’ can be excessively out-of-tolerance in some ultra-precision components and assemblies. Even the previous high-accuracy value of a ‘micrometre’ – often simply termed the ‘micron’ this dimensionally-being 10–6 m (i.e. illustrated against the hair for comparison in Fig. 253a), which is not considered and exceptional dimensional size to ‘hold’ in today’s ultra-precision machining world. In fact, the technical challenge now and into the future, is not one of the actual manufacture of these parts (Fig. 253b), but measuring them, as the old statement, that: ‘We make it, then measure it, at ten times this accuracy’ , does not hold true anymore. We are ‘almost routinely’ of late, making ultra-precision components at the ‘atomic levels of resolution’ , so how can one measure sub-atomic sized components at the ‘absolute limits’ of today’s metrological instrumentation? This form of ‘infinitesimal measurement’ , is where the term ‘uncertainty of measurement’ really does become and important factor. There are so many variables in the actual manufacturing process that can influence the overall dimensional sizes of critical features with these miniscule components.


. Figure 253. Micro- and nano-machining of parts is a big challenge for both today and tomorrow

517

518

Chapter 9

The term ‘ultra-precision machining’ – that is its accuracy and precision, really does need to be more clearly stated62. Many people wrongly surmise that ‘ultra-precision manufacture’ only refers to very minute components, but one can also relate to a large-scale dimensional workpiece, that has certain features for example, its machined diameter, or face (Fig. 257b), together with its overall manufactured width, or length (Fig. 254 – top), to these ultra-precision dimensions. In recent years, the distinct challenges with respect to ‘micro-machining’ manufacture have been reasonably successfully overcome, but ‘nano-machining’ (10–9 m) and indeed, ‘pico-machining’ (i.e. 10–12 m), is now distinctly on the horizon (i.e see Fig. 254, for an indication of the relative ‘sizes and scales’ demanded – of late. These latter machining operational strategies, offering simply massive challenges in terms of the: • Production environments – controlling and monitoring the temperature, humidity atmospheric pressure, cleanliness and dust ingress, together with any floor- and air-borne vibrational effects,

62 ‘Accuracy’ , here and to make it more readily understood, will be stated in distinct two ways, thus: – ‘Accuracy’ , can be established by the difference between the actual position of the machine’s slide and the position demanded by the CNC, or: – ‘Accuracy’ , is the conformity of an indicated value to a true value (i.e. an actual, or accepted standard value – this being a qualitative term only).So, the accuracy of a control system can be expressed as the: Deviation, or difference between the ultimately controlled variable and its ideal value – usually in the steady-state, or sampled instants. – ‘Precision’ , on the other hand, is: The degree of discrimination with which a quantity is stated – e.g. a three digit numerical, discriminates among 1000 possibilities. NB Precision* is often contrasted with accuracy. For example, a quantity expressed with 10 decimal digits of precision, may only have one digit of accuracy. (Source: Smith, 1993) *Bell (1999) states that: ‘Precision’ , is a term meaning ‘fineness of discrimination’ , but is often misused to mean ‘accuracy’ , or ‘uncertainty’. Its use should be avoided if possible – according to Bell. So if this latter statement is the actual situation, then some singular confusion reigns, when we use the word ‘precision’. In effect, we should always say: ‘accuracy and precision’ , as uniquely and simply metaphorically-depicted in the well-known Archery: ‘Target analogy’ – for arrows hitting a target, thus: – Wide arrow scattering, but centred (on average) on the ‘gold’ – accuracy, – Arrows off target centre (i.e. from the ‘gold’), but closely grouped – precision, – Close arrow grouping on the ‘gold’ – accuracy and precision. (Source: Oakland, 1986, Smith et al., 1993)

• Work-holding security, part restraint; manufac-

•

•

turing with its potential for component distortion – perhaps achieved by some form of ‘cryogenic machining’ , together with monitoring tool wear the component, Retrieval of parts once manufactured – this collection operation of minuscule machined parts after manufacture is not without problems, as at best, they will be almost invisible to the naked eye (i.e see some of these ‘larger parts’ manufactured in Fig. 253b), ‘True’ dimensional measurements – of such machined part’s actual features, need to be measured perhaps when in-situ as it is being manufactured, or later, when any distortions through subsequent handling and temperature effects have been nullified.

These major manufacturing and measurement problems will have to be addressed in the relatively nearfuture, if these latter ‘invisible to the naked eye’ parts some of which being beyond the visible spectrum for our sight are to be dealt with outside the ‘research environment’ and into actual ultra-precision production.

9.10.1 Micro-Tooling Introduction The question often posed when considering micro-tooling, is: ‘What constitutes a micro-tool?’ For example, some automotive engineering companies might consider micro-tooling to be

Cutting tool technology: industrial handbook

Handbook of Industrial Membrane Technology

Handbook of Industrial Membrane Technology

Machine Tool Vibrations and Cutting Dynamics

Automotive Embedded Systems Handbook (Industrial Information Technology)

Industrial Burners Handbook (Industrial Combustion)

Advanced Industrial Control Technology

Handbook of Machine Tool Analysis

Handbook of Industrial Biocatalysis

Industrial burners handbook

Handbook of industrial organization,

Handbook of industrial organization,

Handbook of Industrial Mycology

Handbook of Industrial Robotics

Maynard's industrial engineering handbook

Industrial Communication Systems (The Industrial Electronics Handbook)

Industrial Power Engineering Handbook

Industrial Waste Treatment Handbook

Handbook of Industrial Catalysts

Handbook of industrial organization,

Handbook Of Industrial Automation

Industrial Solvents Handbook

Industrial burners handbook

Maynard's Industrial Engineering Handbook

Handbook Of Industrial Automation

Industrial Solvents Handbook

Industrial Refrigeration Handbook

Handbook of Industrial Crystallization

Industrial Waste Treatment Handbook

Handbook Of Industrial Automation

Cutting tool technology: industrial handbook

Handbook of Industrial Membrane Technology

Handbook of Industrial Membrane Technology

Machine Tool Vibrations and Cutting Dynamics

Automotive Embedded Systems Handbook (Industrial Information Technology)

Industrial Burners Handbook (Industrial Combustion)

Advanced Industrial Control Technology

Handbook of Machine Tool Analysis

Handbook of Industrial Biocatalysis

Industrial burners handbook

Handbook of industrial organization,

Handbook of industrial organization,

Handbook of Industrial Mycology

Handbook of Industrial Robotics

Maynard's industrial engineering handbook

Industrial Communication Systems (The Industrial Electronics Handbook)

Industrial Power Engineering Handbook

Industrial Waste Treatment Handbook

Handbook of Industrial Catalysts

Handbook of industrial organization,

Handbook Of Industrial Automation

Industrial Solvents Handbook

Industrial burners handbook

Maynard's Industrial Engineering Handbook

Handbook Of Industrial Automation

Industrial Solvents Handbook

Industrial Refrigeration Handbook

Handbook of Industrial Crystallization

Industrial Waste Treatment Handbook

Handbook Of Industrial Automation

Recommend Documents