Fatigue Failure of Textile Fibres (Woodhead Publishing Series in Textiles)

Fatigue failure of textile fibres The Textile Institute and Woodhead Publishing The Textile Institute is a unique org...

Author: M. Miraftab

664 downloads 3106 Views 5MB 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

Fatigue failure of textile fibres

The Textile Institute and Woodhead Publishing The Textile Institute is a unique organisation in textiles, clothing and footwear. Incorporated in England by a Royal Charter granted in 1925, the Institute has individual and corporate members in over 90 countries. The aim of the Institute is to facilitate learning, recognise achievement, reward excellence and disseminate information within the global textiles, clothing and footwear industries. Historically, The Textile Institute has published books of interest to its members and the textile industry. To maintain this policy, the Institute has entered into partnership with Woodhead Publishing Limited to ensure that Institute members and the textile industry continue to have access to high-calibre titles on textile science and technology. Most Woodhead titles on textiles are now published in collaboration with The Textile Institute. Through this arrangement, the Institute provides an Editorial Board which advises Woodhead on appropriate titles for future publication and suggests possible editors and authors for these books. Each book published under this arrangement carries the Institute’s logo. Woodhead books published in collaboration with The Textile Institute are offered to Textile Institute members at a substantial discount. These books, together with those published by The Textile Institute that are still in print, are offered on the Woodhead website at: www.woodheadpublishing.com. Textile Institute books still in print are also available directly from the Institute’s website at: www.textileinstitutebooks.com. A list of Woodhead books on textile science and technology, most of which have been published in collaboration with The Textile Institute, can be found on pages xi–xv.

Woodhead Publishing in Textiles: Number 86

Fatigue failure of textile fibres Edited by Mohsen Miraftab

Oxford

Cambridge

New Delhi

Published by Woodhead Publishing Limited in association with The Textile Institute Woodhead Publishing Limited, Abington Hall, Granta Park, Great Abington Cambridge CB21 6AH, UK www.woodheadpublishing.com Woodhead Publishing India Private Limited, G-2, Vardaan House, 7/28 Ansari Road, Daryaganj, New Delhi–110002, India Published in North America by CRC Press LLC, 6000 Broken Sound Parkway, NW, Suite 300, Boca Raton, FL 33487, USA First published 2009, Woodhead Publishing Limited and CRC Press LLC © Woodhead Publishing Limited, 2009 The authors have asserted their moral rights. This book contains information obtained from authentic and highly regarded sources. Reprinted material is quoted with permission, and sources are indicated. Reasonable efforts have been made to publish reliable data and information, but the authors and the publishers cannot assume responsibility for the validity of all materials. Neither the authors nor the publishers, nor anyone else associated with this publication, shall be liable for any loss, damage or liability directly or indirectly caused or alleged to be caused by this book. Neither this book nor any part may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, microfilming and recording, or by any information storage or retrieval system, without permission in writing from Woodhead Publishing Limited. The consent of Woodhead Publishing Limited does not extend to copying for general distribution, for promotion, for creating new works, or for resale. Specific permission must be obtained in writing from Woodhead Publishing Limited for such copying. Trademark notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation, without intent to infringe. British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library. Library of Congress Cataloging in Publication Data A catalog record for this book is available from the Library of Congress. Woodhead Publishing ISBN 978-1-84569-327-5 (book) Woodhead Publishing ISBN 978-1-84569-572-9 (e-book) CRC Press ISBN 978-1-4398-0210-6 CRC Press order number N10048 The publishers’ policy is to use permanent paper from mills that operate a sustainable forestry policy, and which has been manufactured from pulp which is processed using acid-free and elemental chlorine-free practices. Furthermore, the publishers ensure that the text paper and cover board used have met acceptable environmental accreditation standards. Typeset by SNP Best-set Typesetter Ltd., Hong Kong Printed by TJ International Limited, Padstow, Cornwall, UK

Contents

Contributor contact details Woodhead publishing in textiles

ix xi

Part I: Principles and types of fatigue in textile fibres

1

1

3

1.1 1.2 1.3 1.4 1.5 2 2.1 2.2 2.3 2.4 2.5 2.6 2.7 2.8 2.9 2.10 3 3.1 3.2

Basic principles of fatigue M Miraftab, University of Bolton, UK Introduction Fibre fatigue Fatigue data representation Future trends References

3 4 6 7 8

Tensile fatigue of textile fibres A Bunsell, Ecole des Mines de Paris, France Introduction Principles of tensile fatigue The fatigue failure of thermoplastic textile fibres produced by melt spinning Mechanisms involved in fibre fatigue Fibre fatigue failure at temperature and in structures Tensile properties of organic fibres Fatigue of liquid crystal fibres High modulus polyethylene fibres Conclusions References

10

Flex fatigue of textile fibres M Miraftab, University of Bolton, UK Introduction Methods of flexing fibres

34

10 11 13 19 23 27 29 31 32 33

34 35 v

vi

Contents

3.3 3.4 3.5 3.6 3.7

Kink bands Effect of temperature on flex fatigue Effect of temperature and humidity on flex fatigue Theoretical aspects of flex fatigue References

38 40 44 49 52

4

Torsional fatigue failure in fibres B S Wong and X Wang, Nanyang Technological University, Singapore Introduction: principles of torsional fatigue Types of fibres affected Methods of testing torsional fatigue Factors affecting fibre torsional fatigue Ways of reducing torsional fatigue Sources of further information and advice References

53

Biaxial rotation fatigue in textile fibres B S Wong and X Wang, Nanyang Technological University, Singapore Introduction: principles of biaxial rotation fatigue Types of fibres which can be tested Methods of testing biaxial fatigue Factors affecting biaxial fatigue Rotation over a pin (single end drive) New developments and future trends Advice on ways of reducing biaxial fatigue Conclusions References

73

Part II: Factors affecting fatigue life and fatigue case studies

93

4.1 4.2 4.3 4.4 4.5 4.6 4.7 5

5.1 5.2 5.3 5.4 5.5 5.6 5.7 5.8 5.9

6

6.1 6.2 6.3 6.4 6.5 6.6

Effect of structure–property relationships on fatigue failure in natural fibres L Wang, RMIT University, Australia and X Wang, Deakin University, Australia Introduction Natural fibre structure and morphology Fatigue of natural fibres Methods of controlling fatigue in natural fibres Conclusions References

53 53 53 56 71 72 72

73 74 74 75 82 90 90 91 91

95

95 97 99 121 129 129

Contents 7

7.1 7.2 7.3 7.4 7.5 7.6 7.7 8 8.1 8.2 8.3 8.4 8.5 8.6 9

9.1 9.2 9.3 9.4 9.5 10

10.1 10.2 10.3 10.4 10.5 10.6 10.7 10.8

vii

Effect of textile processing on fatigue J Militky and S Ibrahim, Technical University of LIBEREC, Czech Republic Introduction Fatigue of materials and structures Fatigue of textile structures Prediction of fatigue during wearing Conclusions Acknowledgements References

133

Environmental aspects of fatigue K Slater, University of Guelph, Canada Introduction: aspects of importance Effects of environment on fatigue fracture Effects of fatigue fracture on the environment Overcoming environmental effects Future trends References

169

Fatigue of polymer-matrix textile composite materials Y Gowayed, Auburn University, USA Introduction Experimental evaluation of fatigue response Modeling of fatigue behavior Conclusions References Fatigue damage in structural textile composites: testing and modelling strategies W Van Paepegem, Ghent University, Belgium Introduction Materials Fatigue testing methods Typical fatigue damage in structural textile composites Modelling strategies for fatigue damage in textile composites Future trends and challenges Sources of further information and advice References

Index

133 134 145 163 165 165 165

169 170 178 180 182 185

188 188 190 194 199 199

201 201 202 205 218 223 228 230 231

242

Contributor contact details

(* = main contact)

Chapters 1 and 3

Chapter 6

Dr M. Miraftab Centre for Materials Research and Innovation University of Bolton Bolton BL3 5AB UK E-mail: [email protected]

Dr L. Wang* School of Fashion and Textiles 25 Dawson Street RMIT University Brunswick Vic 5026 Australia E-mail: [email protected]

Chapter 2

Professor X. Wang Centre for Material and Fibre Innovation Deakin University Geelong Vic 3217 Australia E-mail: [email protected]

A.R. Bunsell Ecole des Mines de Paris Centre des Matériaux BP 87 91003 Evry Cedex France E-mail: [email protected]

Chapters 4 and 5 Associate Professor B. Stephen Wong* and Dr Xin Wang School of Mechanical and Aerospace Engineering Nanyang Technological University Singapore E-mail: [email protected]

Chapter 7 Professor Dr J. Militky* and Professor Dr S. Ibrahim Department of Textile Materials Textile Faculty Technical University of LIBEREC Studentska Street No 2 46117 Liberec Czech Republic E-mail: [email protected] [email protected] ix

x

Contributor contact details

Chapter 8

Chapter 10

Professor K. Slater School of Engineering University of Guelph Ontario, N1G 2W1 Canada E-mail: [email protected]

Professor Wim Van Paepegem Ghent University Department of Mechanical Construction and Production Sint-Pietersnieuwstraat 41 9000 Gent Belgium E-mail: Wim.VanPaepegem@ UGent.be

Chapter 9 Yasser Gowayed Department of Polymer and Fiber Engineering Auburn University Auburn AL 36849 USA E-mail: [email protected]

Woodhead publishing in textiles

1

Watson’s textile design and colour Seventh edition Edited by Z. Grosicki

2

Watson’s advanced textile design Edited by Z. Grosicki

3

Weaving Second edition P. R. Lord and M. H. Mohamed

4

Handbook of textile fibres Vol 1: Natural fibres J. Gordon Cook

5

Handbook of textile fibres Vol 2: Man-made fibres J. Gordon Cook

6

Recycling textile and plastic waste Edited by A. R. Horrocks

7

New fibers Second edition T. Hongu and G. O. Phillips

8

Atlas of fibre fracture and damage to textiles Second edition J. W. S. Hearle, B. Lomas and W. D. Cooke

9

Ecotextile ’98 Edited by A. R. Horrocks

10

Physical testing of textiles B. P. Saville

11

Geometric symmetry in patterns and tilings C. E. Horne

12

Handbook of technical textiles Edited by A. R. Horrocks and S. C. Anand

13

Textiles in automotive engineering W. Fung and J. M. Hardcastle

14

Handbook of textile design J. Wilson

15

High-performance fibres Edited by J. W. S. Hearle

16

Knitting technology Third edition D. J. Spencer

xi

xii


17

Medical textiles Edited by S. C. Anand

18

Regenerated cellulose fibres Edited by C. Woodings

19

Silk, mohair, cashmere and other luxury fibres Edited by R. R. Franck

20

Smart fibres, fabrics and clothing Edited by X. M. Tao

21

Yarn texturing technology J. W. S. Hearle, L. Hollick and D. K. Wilson

22

Encyclopedia of textile finishing H-K. Rouette

23

Coated and laminated textiles W. Fung

24

Fancy yarns R. H. Gong and R. M. Wright

25

Wool: Science and technology Edited by W. S. Simpson and G. Crawshaw

26

Dictionary of textile finishing H-K. Rouette

27

Environmental impact of textiles K. Slater

28

Handbook of yarn production P. R. Lord

29

Textile processing with enzymes Edited by A. Cavaco-Paulo and G. Gübitz

30

The China and Hong Kong denim industry Y. Li, L. Yao and K. W. Yeung

31

The World Trade Organization and international denim trading Y. Li, Y. Shen, L. Yao and E. Newton

32

Chemical finishing of textiles W. D. Schindler and P. J. Hauser

33

Clothing appearance and fit J. Fan, W. Yu and L. Hunter

34

Handbook of fibre rope technology H. A. McKenna, J. W. S. Hearle and N. O’Hear

35

Structure and mechanics of woven fabrics J. Hu

36

Synthetic fibres: Nylon, polyester, acrylic, polyolefin Edited by J. E. McIntyre

37

Woollen and worsted woven fabric design E. G. Gilligan

Woodhead publishing in textiles 38

Analytical electrochemistry in textiles P. Westbroek, G. Priniotakis and P. Kiekens

39

Bast and other plant fibres R. R. Franck

40

Chemical testing of textiles Edited by Q. Fan

41

Design and manufacture of textile composites Edited by A. C. Long

42

Effect of mechanical and physical properties on fabric hand Edited by Hassan M. Behery

43

New millennium fibers T. Hongu, M. Takigami and G. O. Phillips

44

Textiles for protection Edited by R. A. Scott

45

Textiles in sport Edited by R. Shishoo

46

Wearable electronics and photonics Edited by X. M. Tao

47

Biodegradable and sustainable fibres Edited by R. S. Blackburn

48

Medical textiles and biomaterials for healthcare Edited by S. C. Anand, M. Miraftab, S. Rajendran and J. F. Kennedy

49

Total colour management in textiles Edited by J. Xin

50

Recycling in textiles Edited by Y. Wang

51

Clothing biosensory engineering Y. Li and A. S. W. Wong

52

Biomechanical engineering of textiles and clothing Edited by Y. Li and D. X-Q. Dai

53

Digital printing of textiles Edited by H. Ujiie

54

Intelligent textiles and clothing Edited by H. Mattila

55

Innovation and technology of women’s intimate apparel W. Yu, J. Fan, S. C. Harlock and S. P. Ng

56

Thermal and moisture transport in fibrous materials Edited by N. Pan and P. Gibson

57

Geosynthetics in civil engineering Edited by R. W. Sarsby

58

Handbook of nonwovens Edited by S. Russell

xiii

xiv


59

Cotton: Science and technology Edited by S. Gordon and Y-L. Hsieh

60

Ecotextiles Edited by M. Miraftab and A. Horrocks

61

Composite forming technologies Edited by A. C. Long

62

Plasma technology for textiles Edited by R. Shishoo

63

Smart textiles for medicine and healthcare Edited by L. Van Langenhove Sizing in clothing Edited by S. Ashdown

64 65

Shape memory polymers and textiles J. Hu

66

Environmental aspects of textile dyeing Edited by R. Christie

67

Nanofibers and nanotechnology in textiles Edited by P. Brown and K. Stevens

68

Physical properties of textile fibres Fourth edition W. E. Morton and J. W. S. Hearle

69

Advances in apparel production Edited by C. Fairhurst

70

Advances in fire retardant materials Edited by A. R. Horrocks and D. Price

71

Polyesters and polyamides Edited by B. L. Deopora, R. Alagirusamy, M. Joshi and B. S. Gupta

72

Advances in wool technology Edited by N. A. G. Johnson and I. Russell

73

Military textiles Edited by E. Wilusz

74

3D fibrous assemblies: Properties, applications and modelling of threedimensional textile structures J. Hu

75

Medical textiles 2007 Edited by J. Kennedy, A. Anand, M. Miraftab and S. Rajendran

76

Fabric testing Edited by J. Hu

77

Biologically inspired textiles Edited by A. Abbott and M. Ellison

78

Friction in textiles Edited by B .S. Gupta

Woodhead publishing in textiles 79

Textile advances in the automotive industry Edited by R. Shishoo

80

Structure and mechanics of textile fibre assemblies Edited by P.Schwartz

81

Engineering textiles: Integrating the design and manufacture of textile products Edited by Y.E. El-Mogahzy

82

Polyolefin fibres: Industrial and medical applications Edited by S.C.O. Ugbolue

83

Smart clothes and wearable technology Edited by J. McCann and D. Bryson

84

Identification of textile fibres Edited by M. Houck

85

Advanced textiles for wound care Edited by S. Rajendran

86

Fatigue failure of textile fibres Edited by M. Miraftab

87

Advances in carpet technology Edited by K. Goswami

xv

1 Basic principles of fatigue M MIRAF TAB, University of Bolton, UK

Abstract: The notion of fatigue phenomena and their distinctive difference from other forms of failure is briefly explained in this chapter. Different methods by which fatigue failures are manifested and collected data are represented and discussed. Effects of physical and environmental conditions on ultimate failure forms are highlighted and finally the need for further research in emerging areas and application spectrum of ultra-fine fibres is stated. Key words: failure, fatigue, tensile, survival diagram.

1.1

Introduction

Flexibility and versatility of textile materials ultimately lead to failure of one kind or another. Failure, as in appearance of holes in a garment or edge fraying in a work uniform, has been the subject of much study and specialised debate in the past. However, their importance as an engineering malfunction has been substantially marginal due to their limited use and practical requirements. With the advent of technical and performance textiles, where durability and material resilience override fabric appearance and normal wear and tear, an in-depth understanding of fracture and failure of textile materials has become an essential ingredient in design and manufacture of textile-based products. To understand fabric or structural failure, a thorough investigation trailing back to the most fundamental level, i.e. ‘fibres’ representing the building blocks of any textile-based assembly must be undertaken. Fibres’ innate or adopted specifications determine fibre/yarn strength, uniformity, flexibility, durability and ultimately product performance. It is therefore only logical to try to understand failure from a single fibre point of view. Because of the fineness of fibres, it is possible to impose on the material very fast and controlled sequences of temperatures, chemical environment, stress and so on. This enables the development of the fine structure of the material to be closely engineered in a way which is not possible with materials in bulk. Fibres can be heated and cooled at 1000 ºC/sec or faster, whereas 3

4


in bulk materials, the limits of thermal conductivity slow down changes, cause differences between different parts of the material and builds in internal stresses. A given material can be made in a higher strength form as a fibre than as a large piece.(1) Application of such advanced engineering techniques would inevitably affect the economics of production and the subsequent marketing potential. Durability and resistance to large deformation would therefore be a fundamental requisite of any fibre assembly. Hence the study of ‘fatigue’ as possibly the single most important cause of fibre failure has become vital. The term ‘failure’ is an umbrella definition of what fibre breakdown is and does not differentiate between different modes and causes of failure. Most commonly perceived types of failures are those caused by pulling or over-stretching a fibre, yarn or a fabric to the point of breakdown, i.e. when these materials undergo tensile failures. In the literature much investigative study has been devoted to these kinds of failures relating to both natural and man-made fibres. However, materials can also fail by rubbing or frictional interaction with the same or different media causing what is commonly referred to as breakage due to abrasion. Abrasion itself is a well-understood phenomena where many factors including temperature, humidity, type and nature of materials influence its mode and form of failure. A third and maybe the least obvious type of failures are those caused by repeated or cyclic action of a force that is always far below the normal force required for stretch or tensile breaks. These are failures induced by the phenomena of fatigue.

1.2

Fibre fatigue

Cyclic loading and off loading generally leads to fatigue failures. However, the cyclic action is not limited to tensile or compressive modes where lateral tension or compression eventually leads to failure. Repetitive bending (flexural) and twisting (torsional) are other means by which fatigue failures occur. Figure 1.1 shows the principal differences between straight tensile failures and that of a tensile fatigue failure. When a single fibre is stressed at a relatively fast speed, tensile failure eventually occurs as shown by curve (a). When stress is applied at relatively slow speed, the fibre undergoes gradual extension or creep and reaches the breaking point at a lower stress level over a longer period (b). When cyclic loading in the form of reduction and increase in the applied load is experienced, as shown by the sinusoidal profile (c), failure never occurs. Fatigueresistant structures, i.e. suspension bridges, specialised conveyor belts and so forth, are designed and erected based on this principle. However, when cyclic loading and off loading (or thereabouts) is applied, failure occurs

Basic principles of fatigue

5

(a) Not fatigue (b) Stress

Not fatigue ...continues (c)

Fatigue (d) Time

1.1 The principal differences between straight tensile failures and a tensile fatigue failure.

rather quickly. This is indeed what is commonly referred to as a fatigue failure. Fibre fatigue is a relatively new concept compared to metal fatigue where unforeseen disastrous accidents in the early days of engineering led to serious consideration of metal fatigue.(2–4) Since only small cyclic forces lead to fatigue failures, failure always occurs within the elastic or fully reversible regions of materials, i.e. along the Hookean region not far from the origin of stress–strain axis. Fibres differ from metals in one major way and that is fibres can not undergo compression as they tend to buckle under compressive stresses, thus giving the rather desirable aesthetic property of flexibility and drape, which is absolutely unacceptable in metals. Fibres being semi-crystalline materials display visco-elastic behaviour, i.e. a bit of viscous liquid and a bit of elastic behaviour. This translates into a classical stress–strain curve that is common to most fibres. Elastic and plastic regions in visco-elastic materials corresponding to strain reversibility and bond breakage respectively are not clearly defined in fibrous materials. Figure 1.2 depicts a typical stress–strain curve for fibres. Fibres are also anisotropic in nature, which means that there are directional restrictions for molecular alignments and thus crystallinity within a fibre. This inherent property directly influences the failure mode of the fibre depending on the directional application of the cyclic stresses. Hence fibre fatigue failures can be far more complex than metal failures. Furthermore, given the inherent flexibility of fibres, under normal practical conditions and use, no single failure mode could be held responsible as the sole cause of the ultimate failure. Under these circumstances the fibrous materials could be undergoing tensile stretch/failure as well as tensile/flexural and/or


Stress

6

Strain

1.2 A typical stress–strain curve for fibres.

torsional failures whilst also experiencing some form of abrasion and this is not even taking temperature and humidity variations into account. Hence a very confused and complex state of affairs could be the outcome of what might otherwise credulously be taken as a straightforward failure. In order to isolate and hence study failure and fracture characteristics caused by single failure mode, a number of laboratory designed instruments have over the last thirty years or so been made and many of these findings have been published in various journals and books.(5–8) Generally, up to twenty different failure patterns have been identified which provide a comprehensive insight to the forms and types of failures.(11–13) These findings, often captured in the form of three-dimensional micrographs, have since become an indispensable design guide for textile material engineers and diagnostic investigators. Where once individual experience and trial and error methodologies were the dominant approach in design and product development, today almost exclusively all performance materials are designed and manufactured based on their functional requirements and their predictive behaviour supported by in-depth understanding of their failure characteristics. Although mathematical modelling and accurate theoretical predictions masterminding performance criteria continues to remain an ongoing challenge.(9–11)

1.3

Fatigue data representation

One common way of assessing fatigue life is to plot stress against number of cycles to failure. Given that the numbers of cycles to failure are usually high, a log scale is normally selected to represent cycles to failure. To make S-N relationships meaningful, a single loading amplitude is often selected whilst these variables are recorded. When fatigue testing fibres or yarns, coefficients of variations in fatigue lives could be quite high, under these

7

Stress

No. of survivals


log N

No. of cycles x 103

1.3 Typical S-N and survival diagrams.

circumstances it is not uncommon to use histograms or even survival diagrams where numbers of fibres/yarns surviving a fixed number of cycles are recorded at each stage. Figure 1.3 depicts typical S-N and survival diagrams. Survival diagrams can be particularly helpful where comparative fatigue failure behaviour is being studied under different conditions, e.g. temperature and humidity.

1.3.1 Factors affecting fibre fatigue data Other than the chemical make-up of the material and the fatigue mode under which fatigue testing is carried out, there is a whole range of other factors that affect fatigue failure results. Testing under standard laboratory conditions is by no means a true reflection of what might actually happen to the material in use but it is a good comparative indicator, all things being equal. For example, a cyclic fatigue phenomenon experienced by a tyre cord may be less than 1 millisecond and yet an industrial fabric experiencing a similar phenomenon may take days if not months for cyclic recurrence. Furthermore variables such as physical dimensions, impurity contents, temperature, humidity, environmental conditions; i.e. acidity/alkalinity, presence or absence of UV, bacteria and so forth all influence and alter fatigue behaviour of fibres. Some of these important factors are dealt with in the course of this book and references to relevant literature for further studies are made.

1.4

Future trends

Increasing application of fibrous matrices in polymer-based composite structures where, standard mechanical and physical properties apart, structural safety and material integrity are of utmost importance, knowledge of

8


fatigue behaviour and its consequential effects are crucial. Unlike metals, composite structures are anisotropic in nature and therefore experience stress and mechanical deformation in a different way. Failures in composites do not necessarily occur by propagation of a single crack, factors such as matrix cracking, de-bonding, de-lamination, fibre breakage, etc., acting independently or interactively could all contribute to the final failure. In-depth attention to this topic as well as other less conventional application areas is given in the following chapters of this book. Following on from production and processing know-how of a wide variety of microfibres and their proven advantages over standard and/or conventional fibres, there now exists an enormous interest in nanofibres, where boundaries of applications are intertwined with material delicacy and required precision in performance. Their utilisation potential ranges from protective and wearable clothing, filtration, solar sails, and pesticide delivery in plants to wound dressings, drug delivery, surgical implants and scaffolds. Despite this rather broad spectrum of applications, there exists very little or no reported work in the literature in the area of nanofibre degradation, durability and, in particular, fatigue behaviour of these fibres. Shortterm benefits of using such materials may undermine or mask their possible harmful effects in the long term. In the human body, for instance, finely fragmented fibres and/or fibre additives could migrate into the blood stream and the respiratory system causing permanent damage to the overall health of the patient and could even be fatal.(14) Hence further work in these new areas will enhance understanding and potentially the means of providing better overall control.

1.5

References

1. UMIST Symposium, J.W.S. Hearle, High Performance Fibres, Textiles and Composites. 2. Carlson, R., Kardomateas, G.A., Introduction to Fatigue in Metals and Composites, ISBN 0412572001, Kluwer Academic Publishers, November 1995. 3. Stephens, R.I., Fatemi, A., Stephens, R.R., Henry, O. and Fuchs, H.O., Metal Fatigue in Engineering, 2nd edition, ISBN: 978-0-471-51059-8, Wiley-Interscience publication October 2000. 4. Bannantine, J.A., Comer, J.J., Handrock, J.L., Fundamentals of Metal Fatigue Analysis, June 1989, Prentice Hall, ISBN-13: 9780133401912. 5. Bunsell, A.R., Hearle, J.W.S. and Hunter R.D. (1971) An apparatus for fatigue testing of fibres, J. Phys. E: Sci, Instrum. 4, 868–872. 6. Calil, S. F. and Hearle, J.W.S. (1977) Fatigue of fibres by biaxial rotation over a pin, Fracture 1977, ICF4 Conference, Waterloo, Canada 2, 1267–1271. 7. Clark, I.E. and Hearle, J.W.S. (1979), The development of the biaxial rotation test for fatiguing fibres, J. Phys. E: Sci. Instrum. 12, 11–14. 8. Clark, I.E., Hearle, J.W.S., and Taylor, A.R. (1980) A multi-station apparatus for fatiguing fibres in various environments, J. Phys. E. Sci. Instrum. 13, 515–519.


9

9. Hearle, J.W.S. and Wong, B.S. (1977) A fibre fatigue tester, based on rotation over a wire for use in different environments, J. Phys. E. Sci. Instrum. 10, 448–450. 10. Miraftab, M., The influence of temperature and humidity on the flex fatigue lives of nylon 6, nylon 6.6 and polyester fibres, PhD thesis, UMIST, 1986. 11. Hearle, J.W.S., Lomas, B. and Cooke, W.D. (1998) Atlas of fibre fracture and damage to textiles, 2nd edition, Woodhead Publishing ISBN 1 85573 3196. 12. John W.S. Hearle, The challenge of changing from empirical craft to engineering design, International Journal of Clothing Science and Technology, 2004, Volume:16, Issue:1/2, Pages:141–152, Emerald Group Publishing Limited. 13. Leech, C.M., Banfield, S.J., Overington, M.S., Lemoel, M., The prediction of cyclic load behaviour and modulus modulation for polyester and other large synthetic fiber ropes, OCEANS 2003. Proceedings Volume 3, Issue 22–26 Sept. 2003 Page(s): 1348–1352. 14. Nanotubes: The Next Asbestos?, Environmental Health, Nanotechnology, Liz Borkowski, May 2008. http://thepumphandle.wordpress.com/2008/05/21/ nanotubes-the-next-asbestos/.

2 Tensile fatigue of textile fibres A BUNSELL, Ecole des Mines de Paris, France

Abstract: Nylon and polyester fibres find many uses which subject them to cyclic tensile loads. Under specific cyclic loading conditions these fibres can fail by a fatigue process, which can be identified from their fracture morphologies. Fatigue failure occurs only when the minimum cyclic load is below a threshold level. Fatigue failure can therefore be avoided by increasing the overall loading on the fibre. Fatigue crack initiation has been observed to be associated with small particles within the fibres which are added to aid the manufacturing process. Crack initiation becomes generalised throughout the fibres and failure morphologies more complex as the temperature increases. Key words: thermoplastic fibres, fatigue, fracture morphologies, effect of structure, effect of temperature.

2.1

Introduction

Fine diameter organic fibres are used for traditional textile applications as well as in advanced technical structures. The failure of the fibres due to the repeated loading of the fabric or structure can, in some cases, have serious consequences and the mechanisms leading to unexpected fatigue failure have to be understood and taken into account for many applications. The fibres which will be discussed in this chapter will be of a wide variety of types; however, emphasis will be given to polyester (polyethylene terephthalate, PET) and nylon (polyamide 6 and 66) fibres. PET fibres are the most widely used and produced fibres (including natural fibres) throughout the world. Polyamide (PA) fibres were the first synthetic fibres to be produced and are the second most used and produced type of synthetic fibre. Both types of fibre find wide use in apparel and in high performance technical structures such as tyres, cables such as mooring ropes, parachute cords and belt drives. There are many other examples of course. The desirability of avoiding unexpected failure, due to fatigue, in the latter structures should be obvious. These types of fibres fail by fatigue under certain types of cyclic tensile loading. The distinctive fracture morphologies, which occur when fibres fail in fatigue, can be used for diagnostic purposes and allow an insight 10

Tensile fatigue of textile fibres

11

into the mechanisms controlling this behaviour. The understanding of the fatigue processes in these fibres can suggest ways of eliminating or reducing the probability of unforeseen failures. PET and PA fibres are drawn from the melt and the act of drawing aligns the fibre molecular structure, enhancing the fibres’ properties and making them anisotropic. This process of making the fibre, from the molten polymer, does not lead to a perfect alignment of the molecular structure. The molecules are not all arranged parallel to the fibre axis. If this were the case, the rigidity of such fibres would be much greater. For this reason other manufacturing processes have been used to produce fibres with their macromolecules aligned more perfectly parallel to the fibre axis and the results are truly amazing increased mechanical and often heat resistance. This type of fibre is used for technical apparel and fibre reinforced composites. These fibres are usually made by a liquid crystal process in which molecular alignment occurs intrinsically due to atomic bonding within, usually, a solution and the locally aligned molecules are then arranged naturally parallel to the fibre axis during passage through a spinneret, much as tree trunks can be become aligned when transported down a river. In order for liquid crystals to be created it is necessary for the macromolecules to be themselves very rigid and this is achieved by the use of cycles of carbon atoms in their make-up. The chemistry is difficult and expensive so that there exist other fibres based on very simple and flexible polymers, which, by using very dilute solutions of the polymer in a solvent, can be drawn with aligned structures. These very high performance fibres represent a small part of the overall textile fibre market but a disproportionally large part of the advanced technical textile fibre market. All of these fibres will be mentioned and their fatigue failure discussed.

2.2

Principles of tensile fatigue

Fibres are long fine structures. Textile fibres have diameters usually in the range of 5 to 40 μm and the technical fibres which will be discussed have diameters around 10 to 25 μm. This can be compared to the diameter of a human hair, which is about 80 μm. Their fineness means that even the stiffest fibres in tension are very flexible in bending. The bending stiffness is a function of the reciprocal of the fibre diameter to the fourth power. Although there are ways of getting around the buckling of fibres in compression, most evaluation of fibres is in tension.1 The tensile fatigue evaluation of fibres presents particular difficulties as the tests are not quick to undertake and the fibre properties change throughout the test. The study of the fatigue behaviour of materials, in general, began in earnest in the 1950s, due to major problems encountered by the first civilian jet airliners, although the failure of metal structures under cyclic loading

12


Deformation (a)

Load

Load

Load

had been encountered since early in the industrial revolution. Metals, however, have an elastic region in which a cyclic strain of the material induces a cyclic stress. The two are in phase as the material is in its elastic domain and this means that the standard way of fatigue testing metals is to apply a cyclic deformation which induces an in-phase cyclic load, related by Hooke’s Law, and neither is varied throughout the test. It can be noted that in most applications it is a cyclic load to which a material is subjected rather than a cyclic strain. Attempts to evaluate the fatigue characteristics of organic fibres initially used the same type of test as was being used to test metals. That is to say, a cyclic deformation was applied to the fibre. As the fibres were not purely elastic, the plastic deformation produced during each cycle led to an everincreasing length, resulting in a reducing load amplitude experienced by the fibre and eventually its complete buckling. This type of simple extension cycling is shown schematically in Fig. 2.1a. In this test the fibre either fails in the first cycle or not at all as the maximum loading levels quickly fall. A more complicated version of this test, designed to avoid the accumulation of plastic deformation is accumulated extension cycling and the results are shown in Fig. 2.1b. In this test the fibre is held vertically and the plastic deformation produced each strain cycle is removed by opening the bottom grip. A small weight attached to the bottom end of the fibre, which passes through the lower grip, pulls the fibre taut. The grip closes and the fibre is taken through another strain cycle. This means that the volume of fibre being tested is progressively decreased so that, although the maximum displacement imposed does not change, the fibre is taken progressively up its stress–strain curve. With this second type of test, the fibre ultimately fails but it can never be clear if the break is due to a fatigue process or just because the end of the stress–strain curve has been reached. The optimal way of conducting a fatigue test on an inelastic fibre is to monitor the maximum cyclic load and maintain it constant.2 This requires a machine

Deformation (b)

Deformation (c)

2.1 Different ways of conducting tensile fatigue tests: a) simple strain cycling; b) cumulative extension cycling; c) maximum load cycling. The dotted line shows how the maximum cyclic load on the fibre varies throughout the test.


13

capable of adapting the loading conditions on the fibre as it creeps and deforms plastically.3 The fibre does continue to deform by creep but this can be evaluated by constant load tests. Failure by creep under cyclic conditions, during which the fibre is subjected to the maximum load for only a brief part of the cycle, would be expected to occur after longer times than that observed if the maximum cyclic load were applied constantly. It is this maximum load cycling technique which has revealed the fatigue process in organic fibres.4

2.3

The fatigue failure of thermoplastic textile fibres produced by melt spinning

2.3.1 Fatigue fracture morphologies The failure by fatigue of the two largest groups of synthetic fibres, polyamide (PA) and polyester (PET) will be described under the above heading with mention of the failure of polyethylene naphthalate (PEN) which behaves very similarly to PET fibres. The most striking feature of the fatigue failure of this class of fibres is their fracture morphology, which is very distinctive and different from tensile or creep failure. Figure 2.2 shows the complimentary ends of a PA66 fibre broken in tension. The arrows show the region of crack initiation. The type of failure shown in Fig. 2.2 is characteristic of tensile and creep failure of PA 6, PA 66, PET and PEN fibres. The two complementary broken ends are very similar in appearance. There are two obvious regions of crack propagation.5 From the region of initiation the crack there is a bevelled zone, resulting from a phase of slow crack growth during which the plastic deformation ahead of the crack leads to an opening of the crack. The loadbearing cross-section of the fibre is reduced by the advance of the crack and finally fails in an uncontrolled manner resulting in a fracture zone normal to the fibre axis direction. The fatigue failure of the same type of PA66 as shown in Fig. 2.2 is shown in Fig. 2.3. In fatigue, the difference in fracture morphology from that obtained in tension or creep is very clear. Crack initiation, as in tensile failure, is in the region of the fibre surface but instead of progressing across the fibre, the crack begins to run along the fibre at a slight angle to the axial direction. If can be seen in Fig. 2.3 that the break leaves a concave impression on the end from which the tongue of material is removed. This is different from the convex surface seen in the case of peeling of the fibre. When the loadbearing cross-section is sufficiently reduced the PA fibre fails from the root of the fatigue crack by a tensile mechanism, as can be seen from the circled right image in Fig. 2.3 in which the two regions of tensile failure can be observed at the point of final failure.

14


2.2 Complimentary ends of a PA66 fibre broken in tension, at room temperature. The arrows show the region of crack initiation. The diameter of the fibre was 27 μm.

2.3 Complementary ends of a PA66-A fibre broken in fatigue, at room temperature, after cycling from zero load to 80% of simple tensile strength at 50 Hz.


15

Exactly similar tensile and fatigue behaviours are seen with PA6 fibres with indistinguishable fracture morphologies seen in both types of nylon fibre. Analogous behaviour is seen with polyester (PET) fibres both in tension and fatigue; however, some differences, in the angle of crack penetration into the fibre and final failure, occur in the latter case. The failures of PET fibres in tension or creep give very similar fracture morphologies as those shown for PA fibres in Fig. 2.2. In room-temperature fatigue the same scenario of initiation near the surface followed by propagation along the fibre, gradually reducing the load-bearing cross-section is again observed. However, the angle of propagation is smaller than in the PA fibres, leading to a longer crack before the load-bearing section is sufficiently reduced to cause failure as can be seen in Fig. 2.4. The final failure stage of a PET fibre, which breaks in fatigue, occurs behind the fatigue crack tip by a creep process as shown in Fig. 2.5. The initiation of the final creep failure phase does not occur from the fatigue fracture surface but from near the apparently undamaged fibre surface as shown in Fig. 2.6. The failures of polyethylene naphthalate (PEN) fibres in tension and fatigue seem identical to those of PET fibres and the same long breaks are

2.4 The tongue end of a PET fibre broken at room temperature at 50 Hz revealing the very long crack development before failure. The diameter of the fibre was 18 µm and the length of the crack was 2.4 mm.

16


2.5 Final failure by fatigue of a PET fibre occurs behind the fatigue crack tip by a creep process.

2.6 The final failure stage of room temperature fatigue failure in PET fibres occurs from the surface of the fibres, as shown circled, and not from the fatigue fracture surface, as in PA fibres.

seen under cyclic loading which leads to fatigue.6 However, the PEN fibres do show a mechanism which may exist in PA and PET fibres but is less easily observed in these latter fibres. Under certain cyclic loading conditions PEN fibres fail with a superficially creep fracture morphology; however, a closer inspection of the slow crack growth zones reveals that a step by step growth of the crack has occurred, shown as a series of striations.6 The striations can be seen in Fig. 2.7. Close inspection of fracture morphologies of PET fibres, originally interpreted as being due to creep under cyclic loading conditions has revealed that faint striations, similar to those seen in PEN fibres can sometimes be observed. It is not known why the striations are so much more obvious in PEN fibres but their identification suggests the possibility of another type


17

2.7 Striations showing step by step advancement of a fatigue crack in a PEN fibre subjected to loading at 50 Hz at room temperature leading to a superficially creep type failure morphology.6

of fatigue crack growth in fibres. Another explanation would be that the striations are due to arrested slow tensile cracks which stop as the load falls, due to the cyclic form of loading, only to continue propagation at the next load cycle. Fatigue tests at varying frequencies would allow a fuller understanding of this type of failure but have yet to be carried out.

2.3.2 Loading conditions leading to fatigue failure There are loading criteria which must be fulfilled if the fibres are to fail by fatigue. It seems likely that the fibres have to be subjected to a certain load amplitude for fatigue failure to occur, although a minimum amplitude level has not been determined. However, what is clear in all of the fibres so far discussed is that the minimum cyclic load has to be below a certain level, but not necessarily zero, for fatigue failure to occur. Figure 2.8 shows the effects of increasing the maximum cyclic load and also on increasing the minimum load on PET fibres subjected to fatigue loading at 50% at room temperature. It can be seen that increasing the maximum load, from 70% of simple tensile breaking load to 80%, reduces median lifetimes, as would be expected. However, if the maximum load is kept at 80%, increasing the minimum load from 0 to 10% of breaking load increases the median lifetime to the same as when the fibre is cyclically loaded from 0 to 70% of breaking load.7 The loading levels shown in Fig. 2.8 are rather high and have been used so as to obtain failures in reasonable times. The fatigue process has been seen to be related to the internal damping which occurs during cycling and Fig. 2.9 shows how this energy dissipation is affected by changes of both the

Fatigue failure of textile fibres 100%

1

80%

0.9 0.8

Surviving fibres

18

0–75% 0.7 0–80% 0.6

60%

0–70%

0.5 40%

0.4 0.3

20%

0.2 0.1

0 0.1

0–70% 0–75% 0–80% 5–80% 10–80%

10–80% 5–80%

0 1 Lifetime (h)

10

2.8 The survival graphs of PET fibres subjected to different maximum and minimum cyclic loads. The median lifetime is defined as that which produces 50% survival rates.7

Dissipated energy (1e–6J

)

8 PET

6 4 2 0

80 0

70

10

Min

60

20

imu ms

50

30

tres

s (%

)

40

40

xim Ma

um

)

s es str

(%

2.9 Energy dissipation during cyclic loading of PET fibres as the minimum and maximum stresses are varied.8

minimum and the maximum load levels.8 It can be seen that, for any given loading condition, as the minimum stress is increased the dissipated energy falls quickly but as the maximum stress is reduced the dissipated energy reduces much less quickly. This suggests that although there is a minimum load cut-off level above which fatigue is inhibited, reducing the maximum stress only increases lifetimes but does not prevent ultimate fatigue failure.


19

The effects of changing loading parameters are also seen when testing PA fibres.

2.4

Mechanisms involved in fibre fatigue

The long fatigue cracks developed in the fibres which have so far been described is a reflection of the anisotropy of their molecular structures, although the distinctive angles of penetration seen between the two main families of fibres, PET and PA, are not fully understood. It is likely that this angle is related to the long periods of the molecular morphology found in each type of fibre. The structure of a fibre is complex, as is shown schematically for a PA 66 fibre (Fig. 2.10). The structure of a PET fibre is thought to be very similar although perhaps showing greater groupings of the nanofibrils into larger fibrils, which may explain the longer breaks observed in fatigue. These fibres are spun and drawn from the melt at very high speeds, 3000 to 7000 m/min. On leaving the spinneret the material is near its melting point, around 260 ºC and is quickly cooled. Cooling is most intense at the surface of the fibre, which is the region which first solidifies. Shortly afterwards the core of the fibre cools and contracts. This results in residual stresses across the fibre section with the surface being put into compression with respect to the core. The residual stresses have been measured by Raman Spectroscopy and can be very significant.9 Cooling also produces a skin which can be observed by transmission optical microscopy and on scanning electron micrographs of fracture surfaces. At the molecular level the macromolecules are generally thought of as being folded in compact

Core Crystallite

Skin

Microfibril Oriented amorphous domains

Amorphous domains

Ø ≈ 30 μm 7.5 / 8 dtex

2.10 Macro and nano-strucure of a PA 66 fibre.9

20


crystalline regions, making the structure semi-crystalline. The molecular structure is arranged in fibrils and possibly bundles of fibrils which must influence fatigue crack growth. The fibres also contain other materials than the polymer. These are added to the polymer before drawing for a variety of reasons. For example, antimony is added as a catalyst to aid polymerisation and bromine is added, often carried on flake glass, as an antioxidant or flame retardant material. These add small inclusions usually less than one micron in size but they are very significant in initiating crack growth in fatigue and possible under simple tensile or creep loading. Occasionally large particles can initiate failure from within the fibres when tested at room temperature. In this case the failure morphologies are conical as can be seen in Fig. 2.11.10,11 Figure 2.12 reveals a particle, which has been identified as antimony, still in place in one of the two complimentary fracture surfaces of a PET broken apparently in creep. Figure 2.13 shows the tip of the tongue obtained after the fatigue of a PA66 fibre together with the complementary initiation point. The break has clearly been initiated just under the fibre surface by a particle. An examination of the fibres before testing reveals the presence of the particles and there seems always to be a particle or several particles in the initiation regions of fatigue cracks.12 Figures 2.14 and 2.15 show such particles at the crack initiation point in, respectively, a PA66 fibre and a PET

2.11 Both broken ends of a PA 66 fibre broken at room temperature in fatigue at 50 Hz showing crack initiation inside the fibre. The small crater at the tip of the cone reveals the origin of the crack as an inclusion.11

2.12 Both ends of a PET fibre, showing a classical tensile or creep fracture morphology, the crack having been initiated by an antimony particle of about one micron in size.11

2.13 Complementary initiation points of a fatigue break of a PA 66 fibre showing that the initiation was by a particle which has left a crater in the tongue end of the break.11

2.14 A particle revealed by transmission optical microscopy at the point of initiation of a conical fatigue crack in a PA66 fibre tested at room temperature.

22


2.15 A particle is shown at the crack initiation point in a PET which has failed in fatigue at room temperature.

fibre, which had failed in fatigue at room temperature. At room temperature, it is usually only the particles situated at the interface between the fibre skin and the core of the fibre, about one micron under the surface, which initiate cracks. Clearly the interface represents a weakened boundary within the fibre and the particle further weakens it. The polymer, when drawn from the melt, undergoes considerable extension but the hard inclusions do not. This results in a region before and after each particle in which the polymer experiences different deformation from that of the rest of the fibre. This must create a weakened zone, which when it is at the skin-core boundary can initiate fatigue cracks and possibly other types of failure. Figure 2.16 shows two successive sections of the initiation region in a PA66 fibre. The presence of a particle at the skin–core interface initially induces a debonding, which may be visible on the fibre surface by the presence of some irregularity. The skin is then broken and this is seen as the beginning of the longitudinal fatigue crack. It should be noted that surface damage or irregularities are not the causes of the crack initiation but rather the disturbance inside the fibre due to the presence of a particle. When the fatigue crack has begun to propagate, its path can be influenced by other particles and it can be seen to be deviated so as to pass preferentially through regions near particles.


Skin/core separation without crack breaking the fibre surface

23

Fatigue crack breaks through fibre surface revealing start of longitudinal fracture.

2.16 Microtomed sections of the initiation region of a fatigue crack in a PA66 fibre of 26 μm in diameter, revealing that initially the skin becomes separated from the core (left) and then the fracture breaks through the skin to appear as the initiation point of the longitudinal fatigue crack.

2.5

Fibre fatigue failure at temperature and in structures

The appearance of the fracture morphologies of PA or PET fibres is seen to change as the temperature of the environment in which they are tested is increased. In tensile tests the fracture morphologies become less crisp and sometimes more complex. Around the glass transition temperature the fatigue fracture morphologies of both PA and PET fibres show two types. The familiar long fatigue fractures found at room temperature can be seen but increasingly, as the temperature is increased, another complex, truncated fatigue morphology, comes to dominate the majority of the fibre breaks as shown in Fig. 2.17. Above Tg only the truncated fatigue breaks are found. These types of fracture morphologies are also found in fibre bundles which are cycled at 50 Hz at room temperature. The temperature of the bundles has been found to increase above Tg due to the poor heat dissipation of the fibres, so that even if the surrounding environment is at room temperature, the fibres inside the bundle experience large increases in temperature and fail by the truncated fatigue process.7 For the same reason these breaks have also been observed with fibres removed from rubber which had been reinforced with PET and PA66 fibres and subjected to fatigue tests.

24


PA66

PET

2.17 Examples of high temperature fatigue breaks found in PA66 and PET.

2.18 Optical micrograph of a truncated fatigue fracture of a PET fibre, tested at 80 °C, revealing several inclusions associated with the break.

An examination of these complex breaks by transmission optical microscopy reveals that multiple breaks are initiated at high temperature and throughout the body of the fibre. Figure 2.18 shows such a micrograph which again reveals that crack initiation is associated with the presence of particles. At high temperature however the fractures can be initiated throughout the volume of the fibre. Figure 2.19 shows a section of a PET fibre fatigued at 120 ºC and reveals an internal crack which, successive microtomed sections show, does not exit at the fibre surface.10 The processes involved are an extension of the mechanisms seen at room temperature as initial damage is still seen at the fibre–core interface but the presence of particles within the body of the fibre initiates crack propagation. It has been seen that these internal cracks do not necessarily exit at the fibre surface but clearly weaken the fibre. That the cracks do not reach the fibre surface means that there is no significant shear stress generated at the crack tip so that propagation along the fibre is limited. This explains


25

2.19 At temperatures around Tg and above a majority of fatigue breaks in PET and PA66 show complex truncated fatigue breaks. Some cracks can be seen to occur inside the fibre and do not exit at the surface.12

the shorter conical breaks seen by Herrera-Ralmirez et al.10,11 As has been demonstrated it is likely that several independent cracks can be initiated within any given length of fibre. Eventually these cracks coalesce and the complex truncated fatigue break occurs. It seems likely that the increased temperature reduces transversal bonds between microfibrils making up the fibres so that the weak interface, provided by the skin–core boundary, is no longer unique and failure can be initiated throughout the body of the fibre encouraged by the presence of inclusions. The truncated fatigue breaks observed in this study resemble exactly the fibre breaks taken from fatigued composite disk specimens consisting of the fibres embedded in a rubber matrix, as reported by Yabuki et al., Winkler and Naskar et al.13–15 It is clear that in this type of test, the fibres are subjected to large temperature increases due to internal damping and the poor heat transfer properties of the fibres and the rubber. The appearance of the fatigue breaks is then that of the truncated fatigue morphologies rather than those observed with single fibres tested at room temperature, for which heat exchange with the surrounding environment has been calculated to be of the order of only several degrees Celsius. The simple tensile failure stress and strain to failure of the PET fibres tested in fatigue vary as is shown in Fig. 2.20.

26

Fatigue failure of textile fibres 45

1.1

Failure strain (%)

Failure strain Failure stress

35

1 0.9

30

0.8

25

0.7

20

0.6

10 15

0

50

100 Temperature (°C)

Failure stress (GPa)

40

0.5 200

150

2.20 The failure strains and stresses of PET fibres as a function of temperature.7

1 Fraction of surviving fibres

0.9

0.1

0.8 0.7 0.6

0–80% 0–75% 0–80% 0–75% 0–75%

20 °C 20 °C 80 °C 80 °C 120 °C

0.5 0.4 0.3 0.2 0.1 0

1 Lifetime (h)

10

2.21 Fatigue lifetimes of PET fibres as a function of temperature and stress level.7

It can be seen that the failure stress is markedly different at 80 and 120 ºC from that at room temperature and this can be of great importance if the fibres are thought to be at a lower temperature but because of internal heating, which may occur in fibre structure, are really at a much higher temperature. This should be borne in mind when examining the fatigue results shown in Fig. 2.21 for which the percentages given are those of the maximum cyclic load compared to the simple tensile breaking load at that temperature. It


27

should be noticed that the lifetimes decrease with temperature, although raising the minimum load is still found to increase lifetimes at any given temperature. The principal reason why lifetimes are reduced is that the final failure process in the fatigue of PET fibres is governed by the creep of the reduced load-bearing section of the fibre and creep rate is increased with temperature. Truncated fatigue failures have been observed at room temperature in polyamide 66 fibres which had been immersed in hot water and then tested at room temperature.16 Significantly these fibres absorb water and as they do the Tg falls and can fall below room temperature. In this case it seems that the fibres were above their glass transition temperature and the fatigue failure resembled those obtained at high temperatures.

2.6

Tensile properties of organic fibres

Table 2.1 shows the molecular repeat unit for a variety of organic fibres. This table shows clearly that there are two families of organic fibres. The melt spun fibres, such as PET and PA fibres, which have low Young’s moduli and the others, which can have tensile stiffnesses greater than that of steel combined with densities which are about one-fifth that of steel. Polyamide and PET fibres are composed of flexible molecules and can be spun from the melt. The molecular morphology of these fibres is not completely drawn out so that loads are only supported by the co-valent bonds of the macromolecules’ backbones. The results are an initial elastic modulus far from that which could be achieved if the molecules were stretched out parallel to the fibre axis. This can be achieved by using molecules which are very stiff and which in solution can be aligned by secondary bonds to form liquid crystals. During spinning the molecular structure is aligned so that, even without drawing, the resulting fibres have an aligned molecular structure and a very high initial modulus. The stiffness of the molecules is ensured by the introduction of cyclic groups of carbon atoms. The aromatic groups in the poly(p-phenylene terephthalamide) fibres, produced under the names of Kevlar and Twaron, comprise part of the aramid family of fibres. More recently the Zylon fibre has been produced using the PBO chemistry and this completely straight molecule results in a fibre with a Young’s modulus 40% stiffer than steel with a density only about 20% of that of steel. The cyclic groups of atoms also lead to great thermal stability of the fibres. The impressive stiffness of the fibres made by liquid crystal technology is achieved by aligning the stretched out molecules parallel to the fibre axis. There is a price to be paid, however, and that is in the increased anisotropy of the fibres. The radial strengths of such fibres depend on rather weak secondary bonds so that these fibres split very readily.

Poly(p-phenylene benzobisoxazole) PBO [Zylon]

Poly(p-phenylene terephthalamide) [Kevlar, Twaron]

Polyethylene terephthalate [Polyester]

Polyamide 6/6 [Nylon 6.6]

Polyamide 6 [Nylon 6]

Fibre type

HN

O CO

NH CH2

NH CH2

O

N

N

CH2

CH2

CH2

O

NH CO

CO

CH2

CH2

O

CH2

CH2

CO

CH2

CH2

CH2

n

CH2

CO

Repeat unit in the macromolecule

n

NH CO

CH2

CH2

CH2

CH2

CO n

550

650

280

260

260

230

Melting or decomp. temperature °C

135

15

5

4

Maximum elastic modulus (GPa)

Table 2.1 Properties and structures of organic fibres spun from the melt and those spun from liquid crystal solutions


2.7

29

Fatigue of liquid crystal fibres

There are few studies on the fatigue of fibres made by the liquid crystal route. One reason is that the fibres are very anisotropic and split readily in almost any form of failure as can be seen from Fig. 2.22, which shows a Kevlar fibre broken in simple tensile loading. The fibre has a diameter of 12 μm. Kevlar was the first high modulus commercialised fibre produced in the aramid family and became available around 1973. The name aramid comes from the combination of aromatic and polyamide as the elemental composition is that of nylon. Kevlar fibres were originally produced with tyre reinforcement in mind but have found many applications such as for cables, belt drives, flack jackets, etc. Another, very similar fibre is called Twaron. These fibres possess a molecular structure which is aligned parallel to the fibre axis, with little or no chain folding. The result is a high Young’s modulus and little capacity to creep or to deform plastically in tension. The transversal bonding normal to the fibre axis is Van der Waal’s and hydrogen bonding, which are secondary bonds and account for the anisotropy and splitting. The splitting and

10 μm

2.22 A Kevlar fibre broken in tension.

30


plastic deformation in compression means that this type of fibre is difficult to cut and finds use in structures which protect against ballistic impacts. The fibres have to be cycled to very high fractions of their tensile breaking load, typically above 85%, for failure to occur by fatigue.17–18 That an effect of the cyclic loading exists can be seen by a comparison of creep failure data and failure data obtained with cyclic loading, up to the level used in the creep tests. It is found that the cycled specimens fail earlier than those only supporting static loads equal to the maximum fatigue load. If creep were the dominant mechanism, failure should take longer under cyclic conditions, as the fibre is not subjected to the maximum load for more than a fraction of the loading time. As was observed with the melt spun fibres, fatigue lifetimes lengthened significantly if the cyclic maximum load was maintained constant and the minimum load raised. Median lifetimes for Kevlar 49 fibres cyclically loaded at 50 Hz, from zero load to 90% of their nominal tensile strength, were around 4 × 104 cycles but this increased to 6 × 106 cycles with a minimum load of 60% of breaking load. There seemed to be less of an obvious minimum threshold level, which would prevent fatigue failure, than in melt spun fibres. It has been observed that the splitting, which occurs during cyclic loading, seems to be, qualitatively, more extensive and longer than when the fibre breaks in tension. Figure 2.23 shows the failure of another liquid crystal fibre called Zylon. This fibre became available in the late 1990s and is the commercial fibre with the highest Young’s modulus. Its extraordinary tensile stiffness is due to the basic molecular building block being perfectly straight, unlike that in Kevlar, which contains a slight kink. However there are only Van der Waal’s bonds in the planes normal to the fibre axis and this means that the fibre also splits very readily.

2.23 A close-up view of the splitting of a Zylon fibre.


31

2.24 Regularly spaced compression bands seen on a Zylon fibre, broken in tension.

The Zylon fibre splits readily in tension but perhaps even more in fatigue, although there is little data on this recently produced fibre. The manufacturers claim that whilst there is no fatigue effect in tensile cyclic loading of Zylon, the tensile modulus is said to fall during loading. Like the Kevlar and Twaron fibres, unpublished results show that, cyclic tensile loading at 50 Hz, of Zylon, to high fractions of breaking strength, produces premature failure when compared to that which would be expected from creep lifetimes, which suggests that in this fibre as well there exists a tensile fatigue failure mechanism. Failure of these highly oriented fibres results in a compressive shock wave being transmitted along the fibre and this produces quite regularly spaced compressive kink bands in the fibre as can be seen in a Zylon fibre shown in Fig. 2.24.

2.8

High modulus polyethylene fibres

Stiffness in organic fibres, as has been shown above, comes from aligning the macromolecules parallel to the fibre axis. In the case of fibres made by liquid crystal technology this is achieved by developing very stiff molecules which can be aligned by the shear forces in the solution during manufacture. An alternative approach is to take a simple and very flexible molecule, dissolve it in a very dilute solution so as to allow maximum liberty of movement to the molecular structure and then optimise the spinning of the fibre so as to align the macromolecules parallel to the fibre axis. This has been done in the production of high modulus polyethylene fibres. About 2% of polyethylene is dissolved in the solution so that a great deal of solvent has to be removed during fibre manufacture. The result is,

32


2.25 The split and fused break of a high modulus polyethylene fibre showing the longitudinal separation of the structure which occurs during fatigue.

however, a fibre which can have a modulus greater than that of glass with a density lower than that of water. The fibres compete against the liquid crystal fibres for the same types of applications, particularly for ballistic protection and cables. The polyethylene is, however, much more sensitive to temperature, and creep can be a problem for structures, such as cables, which are loaded continuously for long periods. These high modulus fibres are also very aligned and fail by splitting as shown in Fig. 2.25. The low melting point of polyethylene, around 150 ºC, means that the ends of broken fibres can be seen to have fused during failure. The heat is produced by the stored energy released at break.19 The separation of the fibrillar structure, which can be seen behind the break is seen only after fatigue loading. These fibres also fail more quickly in cyclic loading than in steady loading. Unpublished results from the author’s laboratory show that lifetimes increase if the maximum load, say 65% of tensile breaking load, is maintained constant and the cyclic load amplitude reduced, so revealing the damaging effects of fatigue loading.

2.9

Conclusions

The most widely used and produced organic fibres, polyester and nylon fibres, fail by a tensile fatigue process which produces a distinctive fracture morphology very different from that obtained with other types of loading. Fatigue failure requires a sufficiently large load amplitude but a necessary criterion is that the minimum cyclic load must be lower than a critical load threshold. The mechanisms which control this behaviour are complex and the macrostructure of the fibre which has been shown to have a skin–core


33

structure is important in the initiation of fatigue breaks at room temperature. The role of small inclusions in the fibres has been shown to determine the point of initiation of the fatigue cracks which then run along the fibre, gradually reducing the load-bearing cross-section until failure occurs. At higher temperatures the crack initiation is found to occur at inclusions but throughout the thickness of the fibre. This produces truncated failures. High modulus organic fibres in which the macromolecules are stretched out parallel to the fibre axis break with considerable splitting so that interpretation of the fracture morphology is difficult. However, under cyclic load the fibres also break much quicker than in steady load tests which reveals that these fibres, also, fail by a fatigue process.

2.10

References

1. Bunsell AR, Schwartz P ‘Fiber Test Methods’ Comprehensive Composite Materials Vol 5. Ed L.A. Carlsson, (2000) pp. 49–70, Elsevier Science, Oxford. 2. Hearle JWS (1967) J Mat Sci 2, 474–488. 3. Bunsell AR, Hearle JWS, Hunter RD (1971) J Phys E: Scientific Instruments 4:868. 4. Bunsell AR, Hearle JWS (1972) J Mat Sci 6, 10,1303–1311. 5. Hearle JWS, Lomas B, Cooke WD (2000) Atlas of fibre fracture and damage to textiles Second edn., CRC Press. 6. Lechat C, Bunsell AR, Davies P, Piant A (2006) J Mat Sci 41, 6, 1745–1756. 7. Le Clerc Ch, Bunsell AR, Piant A (2006) J Mat Sci 41, 7509–7523. 8. Le Clerc Ch, Monasse B, Bunsell AR (2006) J Mat Sci 42, 9276–9283. 9. Marcellan A, Colomban Ph, Bunsell AR (2004) J Raman Spectroscopy 35: 308. 10. Herrera Ramirez JM, Bunsell AR (2005) J Mater Sci Lett 40, 5:1269–1272. 11. Herrera Ramirez JM, Bunsell AR (2006) J Mater Sci, 41, 22: 7261–7271. 12. Le Clerc Ch, Bunsell AR, Piant A, Monasse B (2006) J Mat Sci 41, 20, 6830–6842. 13. Yabuki K, Iwasaki M, Aoki Y (1986) Textile Research Institute 56(1):43. 14. Winkler EM (1991) Textile Research J 61(8):441. 15. Naskar AK, Mukherjee AK (2004) Poly. Degrad. Stabil. 83(1):173. 16. Nasri L, Lallam A, Bunsell AR (2002) Textile Res. J. 71, 5 459–466. 17. Bunsell AR (1975) J Mater Sci 10, 1300–1308. 18. Lafitte MH, Bunsell AR (1982) J Mater Sci, 17, 2391–2397. 19. Dessain B, Moulaert O, Keunnings R, Bunsell AR, (1992) J Mater Sci, 27, 4515–4522.

3 Flex fatigue of textile fibres M MIRAF TAB, University of Bolton, UK

Abstract: Flex fatigue, as possibly the single most important phenomenon by which cyclic bending under low stress takes place, has attracted the attention of many researchers where desire to reproduce such failures has resulted in a number of imaginative and original testing apparatuses. This chapter sets out to explore flex fatigue life cycle of synthetic fibres under specially designed environmental conditions where temperature and humidity are manually regulated. The results, based on careful collection of failed fibres and their subsequent microscopic analysis, reveal gradual development of fatigue failure not too dissimilar to those reported by other researchers. However, under certain identified environmental conditions, uniquely, a central crack running along the entire length of the flexing zone is observed. This is claimed to be due to the opposing shear forces generated by the back and forth movement of the fibre, reaching a maximum value towards the centre of the fibre. Partial flexing of fresh samples under the identified conditions followed by tensile testing confirms the central weakness/crack phenomenon by registering typical ductile tensile failure independent of the central crack. Key words: flex fatigue, shear forces, central crack, residual strength.

3.1

Introduction

Some physical properties of a fabric such as drape, flexibility, handle, creasing, and wrinkle-recovery are dependent on the flexural properties of yarns, which in turn are dependent upon the properties of the individual fibres.(1) Furthermore, since textile materials are subjected to not only axial tensions and twisting moments, but also bending, it is important in certain cases to study the flexural properties and fatigue life of single fibres under suitable conditions of flexing to determine their flexibility and their degree of endurance to bending. Khayatt and Chamberlain(2) and Chapman(3) have stressed that, for practical textile applications, the flexing properties of fibres are as important as the tensile properties because the strains to which the fibres are exposed in use are not entirely tensile in nature. Thomson and Traill(4) showed that, despite its low strength, wool has maximum ability among various natural fibres to withstand repetitive bending operations. Smith,(5) discussing the 34

Flex fatigue of textile fibres

35

characteristics of fibres, proposes that collectively; toughness, breaking strength, and stiffness are three indices of quality that provide a far more accurate general guide to the mechanical behaviour of a fibre than its breaking strength alone. For example, flax(6) fibres have high tenacity, but cannot withstand repetitive flexing on account of their brittleness and low toughness. Silk fibres, on the other hand, are very pliable, possess moderate tensile strength and are very tough. Consequently, they can withstand a large number of repetitive bending cycles before they rupture.

3.2

Methods of flexing fibres

Flexing, unlike other techniques described, involves tension on the outer side and compression on the inner side of the filament. The earliest technique of investigating resistance to bending dates back to Franz and Henning.(7) They applied bending by means of holding one end of a fibre in a suitably designed grip and hanging a weight at the other end. By rotating the grip in an arc of a circle they were able to bend their fibres up to and beyond 180º and recorded the number of bends to failure on a chart recorder. The arrangement is schematically shown in (Fig. 3.1a). Later Thomson and Traill(4) used the same principle to achieve bending fracture. They altered the original pointed grips into rectangular ones and reduced the minimum angle of flexion to 180º. They reported that the design of the grip’s head could have substantial influence on the final fracture and obtained different results depending on surface smoothness of the grips and radius of curvature. The highest number of bends before failure was recorded with wool and nylon, averaging up to 20,000 cycles. Their photographs of the fibre before failure indicated fibrillation in wool, but not in cotton. They further concluded that number of cycles to break increased as the bending angle decreased. Increase in tension in general decreased the fatigue life. The range of tension used by Thomson and Traill(4) was between ½ g and 5 g, cycled at a frequency of 1 Hz. Lincoln(8) also emphasised the importance of bending strain and fibre diameter on the final failure. An alternative technique involving a freely rotating pin was first adopted by Binode C Goswami at UMIST. Here the fibre was made to oscillate at one end while the other end went over a pin carrying a weight suspended vertically. A sketch of this apparatus is shown in Fig. 3.1b. He observed micro-deformations known as kink bands on the compressed surface of polyester and noted the failure along these bands. (Kink bands will be later discussed in some detail.) Goswami’s instrument ran at a low frequency of about 3 Hz and his tests lasted eight hours or more. Jariwala,(9) replacing the rotating pin by a fixed one, used Goswami’s tester to test his fibres at room and elevated temperatures. At high temperatures he placed a hot plate in the path of the test fibre and prevented

36

Fatigue failure of textile fibres Vibrator

Rotating pin Fibre

Clamp

Fibre

Weight (a)

(b) Fibre

Vibrator Clamps (c)

Vibrator

Fibre

Fibre

Rotating pin

Pulley (e) (d)

3.1 Methods of achieving flex failure.

the temperature from dropping by inserting an asbestos cylinder round the fibre. At 90 ºC he reported an increase in fatigue life in polyester and yet a significant drop in that of nylon 6.6. Lack of splitting in nylon 6.6 at elevated temperatures was another factor noted. His methods of generating heat and controlling temperatures were rather crude. Thus his results must be accepted with some reservation. Jariwala also replaced the weight, which was free to rotate, by a tensioned spring with a vase to check that the fibre remained in the same configuration,


37

with tension on one side and compression on the other. Parvizi(10) replaced the hot plate with a hot air blower positioned next to the point of flexure. He observed a gradual increase in fatigue life of nylon 6 tested up to 100 ºC and then a sudden drop. Polyester showed an apparent increase in fatigue life up to 60 ºC before dropping. His results remain unreported. Unavoidable abrasion has been a consistent problem throughout this kind of fatigue process. In an attempt to overcome this, Jariwala turned his attention to failure by means of buckling. That is to study the effect of repeated bending strain alone. In this method a filament is mounted between two jaws 2 mm apart. One jaw remains stationary while the other oscillates (Fig. 3.1c). His test periods are unacceptably high, averaging up to 20 hours per sample running at 50 Hz. He reported virtually infinite life with nylon 6.6. In a further attempt to remedy this problem Wong(11) designed the instrument shown in (Fig. 3.1d). This involved rotating the pin back and forth in unison with the fibre, whilst the test was performed. The pin was held in Teflon brushes and hence extremely friction free. The system was driven by a vibrator and a freely hanging weight providing the tension, as shown. Wong’s technique proved suitable for Kevlar-29, avoiding extensive abrasion, but extremely long lives were once again reported for both polyester and nylon. Later on Ellison et al.(12) used an alternative technique to test flex fatigue resistance under selected conditions of fibre temperature and moisture content. Figure. 3.1e shows the salient features of their flexing apparatus, first put forward by Lundgren.(13) They claim to have largely avoided abrasion and all other associated drawbacks prominent in other methods of flexing. With cotton, they have reported an inverse relationship between flex life and tenacity. Subjecting wool to different treatments, they concluded that increase in relative humidity increases flex life irrespective of the modification made to the wool. The influence of temperature increase from 22 ºC to 50 ºC has been noted to be less marked. Nylon and polyester were exposed to sunlight and Gamma radiation respectively and then tested under different environments. Generally, loss of life with exposure time has been reported. Flex life is once again noted to increase with humidity. Miraftab(14), using the methodology first adopted by B. C. Goswami, fixed the pin and made a multi-stationed system enclosed in an appropriately designed tank to allow testing under different temperatures and humidity. Miraftab also attached an elastic band to the free end of the fibre via a small piece of cardboard before hanging the weight. The hanging weight was subsequently removed once the elastic band had been securely clamped under tension. This setup avoided the unnecessary turbulence and friction caused as the vibrator moved the fibre back and forth at 50 Hz over a 2.5 mm length. Schematic and photographic images of the setup complete

38

Fatigue failure of textile fibres Vibrators Pins

Humidity probe

Heating element

Pump

Fibres

Elastic bands

Conc. sulphuric acid followed with series of filters, not shown

Clamps

3.2 Schematic and photographic image of multi-station flex fatigue tester complete with temperature and humidity control.

with humidity probe and passage of heated air via drying medium are shown in Fig. 3.2.

3.3

Kink bands

Under certain conditions of strain, polymeric materials produce highly localised deformations which are visible under a microscopic. These types of deformations are generally known as micro-deformations.


39

(a) Optical image of kink bands

(b) SEM image of kink bands(9)

(c) SEM image of kink band leading to central split(9)

(d) SEM image of an advanced flexed failure

(e) SEM image of wedge-shape failure(9)

3.3a to e: show typical kink-band formation leading to final failure.

40


Micro-deformations play an important role in the physical characteristics of materials, particularly in metals. Micro-deformations are of various kinds: slip bands, twin bands, kink bands and transverse bands. The initial formation of micro-deformations in polymers involves changes in the orientation or in the shape of the crystalline or amorphous regions without any rupture of the structure. The continuity of structure is maintained, except perhaps for some formation of micro-voids. The appearance of the marking is suggested(9,12,15,16) to result from shearing deformations and it has been demonstrated that tensile or compressive load introduces a shear stress that has a maximum value one-half that of the axial stress at a direction 45 ºC to the axis. Thus, if the yield point in shear is less that half the yield point in tension (or compression), shear deformation would occur, usually at nearly 45 ºC. Kink bands in polymer are commonly observed on the compressed surface. Jariwala(9) has made an extensive study of kink bands by arranging the filaments in a loop configuration. He measured the magnitude of ‘critical’ strain required to allow kink bands on the compression side of filament. Distinct differences were reported in appearance of kink bands on drawn and undrawn filaments. Nylon 6.6 apparently does not indicate the presence of kink bands until and above the glass transition point, whereas polyester shows kink bands even at room temperature. Taking this a step further, Jariwala used the flex fatigue over a pin technique and explained the mechanism of flex fracture and its propagation along these bands. It is generally agreed(9,17) that during repetitive bending strain, cracks occur along the kink bands, but do not form beyond the neutral axis, and hence a wedge shape cut proceeds along the neutral axis leading to final failure. Figure 3.3 illustrates the progressive development of kink bands on the compressed surface of a filament followed by a central split and the subsequent propagation of fracture along these bands. Further kink bands also appear beyond the neutral axis. The presence of micro-deformation on the compressed surface of polyester is clearly visible under an optical microscope (Fig. 3.3a). SEM studies of the compressed surface have shown (Figs 3.3b, c, d) pronounced forms of kink bands and their gradual growth into deep cracks reaching into the neutral plane. Finally the formation of a wedge shape cut as a typical flex failure is presented in Fig. 3.3e.

3.4

Effect of temperature on flex fatigue

The ultimate fatigue failure of a filament is governed by the number of parameters which either individually or otherwise influence the manner in which fracture will occur.(16) Ideally the most desirable technique of


41

monitoring such variations is to alter one variable at a time while retaining all other parameters constant, but this is not usually a practical possibility due to the complex nature of the forces involved and the physical limitations imposed by the testing apparatus. However, in the work reported by Miraftab,(14,18,19) an attempt was made to adjust some of these variables in an effort to minimise undesirable deviations from the expected fatigue failures when using a fixed pin. The following are some of the variables considered to be important when flex testing techniques are used: • bending curvature (pin diameter) • tensile stress • nature of pin’s surface • frequency of vibration • amplitude of displacement of fibre over the pin. Friction at the point of contact is said(9,17,20,21) to lead to severe abrasion on the compressed surface of the bent section. This drastically reduces the effect of fatigue as the dominant cause of failure (see Figs 3.4a and b). To study the effect of temperature and/or humidity of flex fatigue life, Miraftab fixed all the above variables and initially raised the testing temperature from 20 ºC to 120 ºC in 20 ºC intervals in an enclosed environment. Fatigue lives associated to each fibre under these circumstances are shown in Fig. 3.5. In this work, polyester interestingly showed an increase in flex fatigue life up to around 60 ºC before it began to drop. Similarly, nylon 6 showed a huge increase in fatigue life up to 100 ºC before dropping off. Fatigue life of nylon 6.6, on the other hand, did not appear to change with increase in

(a)

(b)

3.4 a Severe surface abrasion; b Abrasion and crossing kink band marks.

42


X 103 Fatigue life (cycles)

Polyester 500 400 300 200 100 0 20

40

60

Nylon 6

Nylon 6.6

80 100 120 Temperature °C

140

160

3.5 Effect of temperature on flex fatigue lives of polyester, nylon 6 and nylon 6.6.

(a) Room temperature

(b) 40 °C

(c) 40 °C

(d) 40 °C

(e) 60 °C

(f) 100 °C

(g) 120 °C

(h) 120 °C

3.6a–h demonstrate flex fatigue failures of nylon 6 at various temperatures.

testing temperature. Under these elevated temperatures no attempt had been made to control the humidity. SEM micrographs associated to these tests are presented in Figs 3.6a–h. At room temperature nylon 6 (Fig. 3.6a) shows a typical flex fatigue in progress where fracture appears to have


43

started from the compression side and propagated at an angle into the body of the fibre until the neutral axis. This is then met by a central crack leading to final failure along the axis. Much surface abrasion is also evident under these conditions. Consistent evidence shows that the kink band failures do not necessarily occur at one point, in fact usually more than one kink band failure is in progress at any one time. For example Fig. 3.6b is a broken end of a fibre which has failed along a kink band but a prospective kink band failure is in progress further down the broken zone. This is even better illustrated in Fig. 3.6d where apart from the catastrophic kink band failure there appears to be at least three other prospective failure zones where failure could have occurred. The spacing between each intended kink band failure is also noted to be curiously very regular. A consistent axial split along the centre of the fibres in all cases could be an indication that the crack might occur first followed by an outward or inward failure along kink bands. The axial split characteristic is maintained with increasing temperature but increasingly most of the surface definition disappear (Figs 3.6e–f). At 100 ºC, melting largely dominates and eventually at 120 ºC the free and separated ends coil on their own axis in a unique fashion (Figs 3.6g–h), presumably due to snapback after breakage. Examinations of nylon 6.6 fibres also indicate failure along kink bands with a number of prospective failure regions along the length of the fibre (Figs 3.7a–e). The central crack is also well evident in the majority of the cases examined.

(a) 40 °C

(b) 60 °C

(c) 80 °C

(d) 100 °C

(e) 120 °C

3.7a–e demonstrate flex fatigue failures of nylon 6.6 at various temperatures. (a) 40 °C; (b) 60 °C; (c) 80 °C; (d) 100 °C; (e) 120 °C.

44


(a) Room temperature

(b) 40 °C

(c) 40 °C

(d) 100 °C

(e) 100 °C

3.8a–e demonstrate flex fatigue failures of polyester at various temperatures.

Polyester appears very different from the two nylons. The fibre splits up into many fibrils along the axis with no apparent indication of failure along kink bands. Each layer appears to tear off from the adjacent one producing yet more fibrils. Increase in temperature initially appears to encourage fibrillation in polyester (Figs 3.8a–b) but this tendency is somewhat reduced at 80 ºC and beyond. At 100 ºC and 120 ºC melting tends to fuse together most of the fibrillated sections (Figs 3.8c–e).

3.5

Effect of temperature and humidity on flex fatigue

Whilst operating within the same temperature regime as before, humidity within the tank was maintained at one of the following settings, i.e. 5%, 30%, 70% or 95%. These humidity settings were arrived at by allowing the air to pass through concentrated sulphuric acid and water in carefully regulated manner. The results associated to temperature and humidity variation for each fibre are given in Figs 3.9 to 3.11. As relative humidity increases from 5% to 95%, nylon 6 displays a general shift in peak fatigue life from 100 ºC to room temperature (Fig. 3.9). The shift is such that the highest fatigue life is recorded at around 80 ºC when the relative humidity is at 70%. A six-fold increase in fatigue life under these circumstances appears unusually high, suggesting, maybe, some failure mechanism



95% R.H.

70% R.H.

30% R.H.

45

5% R.H.

700 600 500 400 300 200 100 0 20

70 120 Temperature °C

170

3.9 Effect of temperature and relative humidity on nylon 6 fibres.


95% R.H. 350 300 250 200 150 100 50 0 20

70% R.H.

40

30% R.H.


5% R.H.

100

120

3.10 Effect of temperature and humidity on flex fatigue life of nylon.


95% R.H. 500 400 300 200 100 0 20

40

70% R.H.

30% R.H.


5% R.H.

100

120

3.11 Effect of temperature and relative humidity on flex fatigue life of polyester.

other than the conventionally accepted kink-band failure is at work. Indeed, SEM examination of some nylon fibres prior to failure shows this to be the case. Figure 3.12 shows a single nylon 6 fibre subjected to flex fatigue failure at 30% r.h. and 60 ºC where a central crack runs along the whole length of the fibre with no impending kink band failures in sight. The insert shows where the fibre would eventually separate, i.e., at an angle possibly along a kink band initiated from the inner/outer side inward or outward.

46


3.12 Nylon 6 at 30% r.h. and 60°C where a central crack is dominant.

3.13 Polyester at 5% r.h. and 80°C where multiple parallel splits are dominant.

Similar observations are made for nylon 6.6 and polyester. These observations clearly indicate that a different failure mechanism must be at work between 30 and 70% relative humidity and temperatures ranging from 60 to 80 ºC. However, with polyester, multiple splitting rather than a single split is observed from 5% relative humidity up to 80 ºC. Although undergoing the same mechanism of failure as those of the nylons, the multi-splitting is probably due to somewhat weaker cohesive forces present between the neighbouring polyester molecules than that of fatigue mechanism, Fig. 3.13. It is also important to note from Fig. 3.11 that the fatigue life behaviour of polyester appears to be opposite to those of nylons, i.e. 5% relative humidity at all temperatures gives the highest fatigue life. These findings suggest that under ‘ideal’ conditions, i.e. certain temperatures and humidity


(a)

47

(b)

3.14a–b flex failures overwhelmed by surface abrasion.

where no abrasion is present, flex fatigue occurs due to opposing shear forces within a fibre as it moves back and forth over the pin. Kink band failures appear to be the dominant cause of failure only when the fibre is not under these ‘ideal’ conditions. Figure 3.14a–b show what are typically regarded as flex fatigue failures, obtained under ‘non-ideal’ conditions where considerable amount of abrasion is also present. To prove that the central crack appears well before the ultimate kink band failure, a selected range of fibres were partially flexed under ‘ideal’ temperature and humidity conditions and subsequently tested for their residual strength, and their SEM micrographs were examined post tensile failures. These sets of results are most revealing. Figures 3.15a–d show typical ductile tensile failure of unflexed fibres along with those of partially flexed-tensile tested fibres. These micrographs clearly show that an unflexed fibre initiates failure from a weak spot on the fibre surface and comes apart in a characteristic V notch manner typical of most ductile fibres. The developing notch extends up to a third of the way into the body of the fibre before tearing across, as the load can no longer be tolerated. In partially flexed fibres under ideal conditions, where abrasion has been avoided, upon tensile testing, they behave as if they were not flexed at all and display similar tensile strength to unflexed fibres following the same V notch failure profile despite the fact that a central crack has already appeared. Figure 3.15c–d clearly show the V notch development on at least two sites, irrespective of the central split/crack. This phenomenon is less obvious with polyester as it undergoes multi-axial splitting, Figure 3.16a–d. Figures 3.17a–c are SEM and optical images of early stages of central splitting of nylon fibres captured at their infancy under ‘ideal’ conditions. These observations clearly illustrate that in cyclic flexing where abrasion is nonexistent, the prime mechanism of flexing is possibly due to shear forces acting in opposite directions rather the common kink band approach although these also play a part.

48


(a)

(b)

(c)

(d)

3.15a–d Tensile failure of unflexed and partially flexed nylon 6 fibres.

(a)

(b)

(c)

(d)

3.16a–d Tensile failure behaviour of partially flexed polyester fibres.


(a)

(b)

49

(c)

3.17a–c Early stage development of the central crack in nylon 6 fibres.

3.6

Theoretical aspects of flex fatigue

During pure bending, compressive and tensile stresses are simultaneously produced within the fibre. When a fibre contacts a curved surface, the fibrils on the upper side are lengthened or extended, and the stress acting on them is tensile in nature. The fibrils on the lower side are shortened or compressed, and the stress acting on them is a compression. For an ideal material, the neutral plane is central and so the tensile and compressive stresses and strain are symmetrically distributed, but opposite in sign. Tensile stresses due to tension applied to the fibre would be superimposed on the stresses due to bending. If a fibre is held in contact with a pin under tension from two distant points which are in directions at right-angles to each other, it would have simple curvature rotor geometry where fibre is in contact with the pin and zero curvature where it is not. However, discontinuity in curvature, as the fibre moves, causes an unsupported discontinuity in stress, thus changing the original geometry such that the length in contact with the pin is reduced and the curvature changes gradually from a maximum value to zero. Standard bending theory(22–23) shows that changes in curvature give rise to shear stresses. Assuming that Hooke’s Law is obeyed and that fibre/pin contact produces no friction, i.e. ideal condition, strain due to bending would be: εb =

y Ro + r

Where, y is the distance from fibre axis, Ro the radius of the pin and r is fibre radius. Since: r 3

High-performance fibres are very strong fibres, have high modulus and have good resistance to high temperature. In use, they may be formed into yarns and ropes or woven into fabrics. Hence during their lifetimes these fibres will experience oscillation loads which may be bending or torsional in nature. Despite their good tensile properties, it is not obvious that their resistance to oscillating loads will be good. Torsional oscillations will be a major part of the varying loads these fibres may experience during their use. Hence it is important to study their torsional fatigue characteristics and obtain quantitative data. As has been described, longitudinal cracking is a prevalent mode of deterioration in fibre fatigue caused by the shearing stresses and strains at the weaker interfaces between fibrils. The structure of high-performance fibres such as Kevlar and Twaron fibres is also fibrillar similar to lowerstrength fibres but the fibrillar structure is much finer (Hersh et al. 1994). Liu and Yu (2005) investigated the torsional fatigue behaviour of highperformance polymers using the apparatus shown in Fig. 4.2. The fibres they tested were Twaron 2000, AKZO, Kevlar 129, Dupont, Kevlar 29, Dupont and ultrahigh molecular weight polyethylene (UHMW-PE), Dyneema SK65, DSM. Some properties of the fibres are shown in Table 4.2. In common with other workers (Goswami et al. 1980; Hearle & Wong 1977a), considerable variance was found to exist between results when attempting to determine the fatigue lifetimes under a specific set of conditions. This indicated the variability of the fibre structure under fatigue conditions. Hence they made 60 measurements for each set of conditions. The effect on torsional fatigue life of pre-tension and torsion angle Liu and Yu (2005) investigated the effects of pre-tension (s 0) and torsion angle (b) on the torsional fatigue lifetime (Nt) of Twaron 2000 fibres. It was found that that linear relationships occurred between (ln Nt) and both (s 0) and (b). This implies that the fatigue lifetimes showed an exponential decrease with increase of (s 0) and (b). The twist angle was changed from 10° to 25° and the pre-tension changed from 1.2 to 2.0 dN/dtex.

Torsional fatigue failure in fibres

69

Typical equations which were developed were: For 25° torsion angle: ln Nt = 6.36 − 2.56 s 0, r = −0.99 For 2.0 cN/dtex pre-tension: ln Nt = 4.63 − 0.13 b, r = −0.99 The regression result for a multivariate linear model with two variables was ln Nt = 6.406 − 0.002 b − 0.873 s 0 − 0.067 s 0 b Because, in this equation the coefficient of the pre-tension, 0.873, is much larger than the coefficient for the torsion angle, 0.002, the pre-tension influences the torsional fatigue life of the fibre much more than the torsional angle. Comparison of the torsional fatigue lifetimes of the high-performance fibres A comparison between the torsional fatigue lifetimes of the four fibres for s 0 = 1 cN/dtex and b = 25° is shown in Fig. 4.18. The UHMW-PE, SK65 performs considerably better than the other fibres which appear to perform similarly. It appears that as the shear modulus increases (Table 4.2), the torsional fatigue life increases for these highperformance fibres and hence a higher shear modulus would indicate that resistance to torsional fatigue would be good. Figure 4.19 shows the typical fracture morphologies of the fibres. It is interesting that extensive fibrillation is a common mode of degradation. The aramid fibres shown in (a), (b) and (c) are all similar and show the more common extensive fibrillation effect. This is of interest since their fatigue performances were also similar (Fig. 4.6). The extensive fibrillation shown

Number of cycles (N)

50 40 30 20 10 0

Twaron

K29

K129

SK65

4.18 A comparison of the torsional fatigue lifetimes of the high performance fibres.

70


(a)

(b)

(c)

(d)

4.19 Fracture morphologies of high performance fibres (a) Twaron 2000; (b) Kevlar 29; (c) Kevlar 129; (d) UHMW-PE, SK65.

by the fractured Kevlar 29 fibre (b) bears a resemblance to the extensive fibrillation found by Hearle and Wong et al. (1977b) in their biaxial rotational fatigue tests. The better performing fibre, UHMW-PE, SK65, appears to show a more macro plastic deformation effect. Due its different molecular structure to the aramid fibres, it is more flexible and plastic. It appears that the ability of the latter to avoid extensive fibrillation has assisted its fatigue performance. These fracture ends appear similar to those in Chapter 5, ‘Biaxial rotational fatigue’, indicating a similar degradation process which is not unexpected due the similar mechanical bending stresses occurring during the torsional and rotating bending procedures. Furthermore, these torsional fracture morphologies resemble those of fibres in use indicating that torsional rotational fatigue is a mode of degradation of fibres in use.

4.4.4 Microstructure investigations of torsional fatigue of fibres. Bakra et al. (1995) investigated microstructure changes during the torsional fatigue of PET fibres using a scanning electron microscope (SEM) and a

Torsional fatigue failure in fibres 10

cycles vs. tenacity (gpd) cycles vs. strain

9

71

0.7 0.6

8

Tenacity (gpd)

6

0.4

5 0.3

4 3

Strain

0.5

7

0.2

2 0.1 1 0 0

2000

4000

6000

8000

0.0 10000 12000

Cycles

4.20 Tenacity and breaking strain of PET fibres after a certain of torsional fatigue cycles. The twist strain amplitude was 20°.

transmission electron microscope (TEM). Also fatigue damage was quantified using a new procedure which will be described. Figure 4.20 shows the not unexpected reduction in strength and also the reduction in strain at fracture after a certain number of fatigue cycles. Again considerable longitudinal splitting was observed with an SEM. A new procedure to assess the quantity of fatigue damage was developed. This involved determining the quantity of energy absorbed by a fibre during fatigue.

4.5

Ways of reducing torsional fatigue

Torsional fatigue can be reduced by lessening the effect of the variables discussed in the previous section on factors affecting torsional fatigue. The basic way is to prevent the twist angle from becoming too large. Since it is governed by the equation: tanα = πdn/l, where d is the diameter of the fibre, n is the number of turns, and l is the length of the fibre. Other measures that could reduce the effects of torsinal fatigue (as discussed in this chapter) are listed below: • • •

increase the draw ratio of the fibre reduce the strain amplitude reduce the pretension on the fibre.

72


4.6

Sources of further information and advice

A major source of information are the papers written by B. C. Goswami, many of which are referenced in this chapter.

4.7

References

Bakra S K, Ellison M S, Goswami B C and Davis H A (Aug 1995), National Textile Centre Annual Report. 211–220. Bunsell A R and Hearle J W S (1971), ‘A Mechanism of Fatigue Failure in Nylon Fibers’, J. Mater. Sci. 6, 1303. Duckett K E and Goswami B C (1984), Text. Res. J., 54, 43–46. Goswami B C and Hearle J W S (1976), ‘A Comparative Study of Nylon Fibre Fracture’, Textile Res. J. 46, 55–70. Goswami B C, Duckett K E and Duckett K E (1980), ‘Torsional Fatigue and the Initiation Mechanism of Pilling’, Textile Research Journal, Vol. 50, No. 8, 481–485. Hearle J W S and Wong B S (1977a), J. Textile Inst. 68, 89–94. Hearle J W S and Wong B S (1977b), ‘Flexural fatigue and surface abrasion of Kevlar-29 and other high-modulus fibres’, J Materials Sci, 12, 2447–2455. Hersh S P, Ellison M S, Goswami B C, Batra S K, Mccord M G and Davis H A (1994), ‘Annual Report – National Textile Center’ 989, 305. Lu Fu-Min, Goswami B C, Spruiell J E and Duckett K E (1985), ‘Influence of fine structure on the torsional fatigue behaviour of poly(ethylene terephthalate) fibres’, J. Appl. Polym. Sci., 30, 1859–1974. Liu X Y and Yu W D (2005), ‘Static torsion and torsion fatigue of UHMW-PE and aramid filaments’, High Perform. Polym., 17, 593–603. Peterlin A (1975), Int. J. Fracture, 11(5), 761–780. Van derVegt, A K (1962), Torsional Fatigue of Fibres, Rheol, Acta 2, 17–22.

5 Biaxial rotation fatigue in textile fibres B S WONG and X WANG, Nanyang Technological University, Singapore

Abstract: Textile fibres used in clothes are very likely to suffer some form of bending and twisting. Also the fibres are more likely to suffer fatigue rather than steady tensile failures. Hence it is essential that bending fatigue techniques are investigated. Biaxial rotation is one of these techniques. In this chapter, types of fibres affected, methods of testing biaxial fatigue and factors affecting biaxial fatigue are investigated. The procedure of fracture is usually by axial splitting indicating the significant strength of the fibre along its axis due to its long aligned molecules. The axial splitting and morphology of the fracture breaks look similar to many fracture breaks that occur in actual use, implying that rotational biaxial tests are useful for predicting fibre degradation in many applications. Furthermore, various severe fatigue parameters and environments can influence fibre fatigue lifetimes. Key words: textile fibres, biaxial fatigue.

5.1

Introduction: principles of biaxial rotation fatigue

Textile fibres used in clothes are very likely to suffer some form of bending and twisting. Also the fibres are more likely to suffer fatigue rather than steady tensile failure. Hence it is essential that bending fatigue techniques are investigated. Biaxial rotation is one of these techniques. In this technique a fibre is bent, usually through 90º, and then rotated. Figure 5.1 shows various methods in which this arrangement can be achieved and these are described in detail in Section 5.3. The cyclic action induces cyclic tension and compression on the outside of the fibre. Progressive breakdown of the fibre occurs and multiple splits are produced along the fibre axis leading to ultimate failure (Hearle and Wong 1977 a,b,c,d). These failure morphologies resemble those which occur in common use of the fibres. Hence it is possible that some form of biaxial fatigue occurs on fibres during use. Therefore experiments using biaxial fatigue are useful methods in simulating the type of fatigue which a fibre suffers during actual use as in clothes and other applications. 73

74


5.2

Types of fibres which can be tested

Biaxial rotation fatigue tests are suitable for synthetic fibres such as nylon with moderate rigidity. These fibres are usually monofilaments of circular cross-section and they can be bent to take fairly severe curvatures required to produce a biaxial rotation fatigue failure. The test is unsuitable for very stiff fibres, such as carbon, which will fracture easily when bent to the curvatures required for failure by fatigue. Many natural fibres, such as cotton or wool with their hollow soft structures, are more difficult to test, particularly when a pin is used to create the curvature, they will flatten when placed over the pin. Tests conducted with these latter fibres indicate the failed fracture ends are due mainly to abrasion and not fatigue (Wong 1975).

5.3

Methods of testing biaxial fatigue

Three basic types of apparatus have been used (Calil et al. 1980). For coarse monofilaments, the fibre can be bent and clamped in jaws so that both ends can be rotated together as in Fig. 5.1(a). To obtain reasonably rapid fatigue, the radius of curvature should be about ten times the monofilament radius. It has not been possible to produce apparatus to obtain a small enough radius of curvature to test typical textile fibres which may be of the order of 10 μm. Consequently, an effective procedure is to force a small radius of curvature by passing the fibre over a pin under some tension. In an initial form of the technique, shown in Fig. 5.1(b), the fibre is rotated from one end and tensioned by a hanging weight (Hearle and Wong 1977c). In a later

90°

(a)

(b)

(c)

5.1 Techniques for fatigue testing by rotating a bent fibre: (a) free biaxial rotation; (b) rotation over a pin, with a single drive and hanging weight; (c) biaxial rotation over a pin.

Biaxial rotation fatigue in textile fibres

75

form, Fig. 5.1(c), the fibre is driven from both ends and tensioned in other ways (Calil 1977).

5.4

Factors affecting biaxial fatigue

5.4.1 The development of axial splitting Synthetic fibres are made of highly oriented long linear polymer molecules and often they are highly crystalline. The covalent bonds along the molecular chains make the fibre very strong along the length of the fibre. However, the weaker intermolecular bonds, e.g. Van der Waals, make the fibre weak, transverse to its axis. Often fibres will fracture through the development of long axial splits in the filaments. The tensile stress along the fibre length could be a single load or cyclic in nature as in fatigue. If a fibre is perfectly uniform with long linear molecules aligned parallel to each other and it is subjected to tension, there would be no stress across the planes parallel to the fibre axis and hence there exists reason for axial splits. However, in practice there may be small discontinuities or defects inside the fibre or on its surface. As the load (whether it is constant or oscillating as in fatigue) on the fibre increases, small shear stresses will occur at these defects as shown in Fig. 5.2(a). These shear stresses will eventually overcome the relatively weak transverse cohesive forces and hence axial cracks are likely to form, Fig. 5.2(b). If the crack moves slightly off axis, it will cross the fibre and lead to failure, Fig. 5.2(c). Failure occurs in this manner because even a small shear stress will overcome the weak intermolecular bonds between the polymer molecules before the large tensile stress breaks the covalent bonds within the chain molecules as shown in Fig. 5.3. The difference may be increased by structural discontinuities.

(a)

(b)

(c)

5.2 (a) Shear stresses at discontinuities in a fibre under tensile stress; (b) shear stresses which are causing single or multiple cracks; (c) crack at an angle to fibre axis, moving across the fibre and producing fracture. (Hearle et al. 1989).

76


5.3 An illustration of strong covalent bonds of the polymer molecules which are parallel to the fibre axis and the weak bonds between the molecules. A defect is shown on the left which leads to axial cracking.

T

T

F

Const. Mt

F

max Zero Mb

F

F F

Mt

0 Const. Mb

Const. Mt

F max

Zero Mb

5.4 Fibre rotating over a pin, indicating frictional forces F, torque Mt, bending moment Mb, and tension T (Calil et al. 1980).

5.4.2 Torque in biaxial testing The forces on the fibre in biaxial fatigue will now be discussed. The technique used by Calil (1977) of biaxial rotation fatigue where the fibre is driven from both ends will be described first. The torque on the fibre which arises from friction on the pin or hysteresis is zero at the centre point on the pin as shown in Fig. 5.4. The forces and moments would vary in the way indicated in Fig. 5.4 with the torque being zero at the centre point and rising to a constant value where the fibre leaves the pin. This variation in torque would explain the fact that in coarse fibres the splitting occurs where the torque is higher. In fatigue testing by rotation of a bent fibre, the multiple splits which result (Fig. 5.5) indicate a high level of twist has developed. Figure 5.5 (a) to (d) show the axial splitting occurring where the twist is a maximum, i.e. from where the fibre leaves the pin. The fracture eventually occurs mainly due to this splitting. Figure 5.5 (e) and (f) show the fractured ends.


77

2054

2011 cycles

(b)

(a)

2327

2538 (c)

2606

(d)

2606 (e)

(f)

5.5 A polyester fibre (4.2 tex) tested using biaxial rotation over a pin, driven from both ends. (Calil et al. 1989). Number of cycles given in the captions. (a) to (d) before failure; (e) and (f) fractured ends after failure.

78


When the drive is from one end and a rotating weight is supported at the other end (Fig. 5.1(b)), the torque on the fibre is similar to that in Fig. 5.4, in that the maximum torque occurs between where the fibre leaves the pin and the driven end. The torque will gradually reduce to zero or a low value as one moves around the pin to where the fibre leaves the pin and then extends to the supported weight. Between where the fibre leaves the pin to the supported weight there will be a low torque value if the weight is rotating in a liquid environment due to the viscous drag on the weight. Single drive rotation over a pin results are shown in Fig. 5.7 and described in Section 5.4.4. When rotation is over a pin, whether driven from one or both ends, surface friction will be a source of torque. The other source of torque is the need to overcome the energy dissipated through bending hysteresis. In biaxial testing, using freely bent monofilaments with no frictional forces, as in Fig. 5.1(a) splits in fibres are produced indicating that high levels of twist have occurred. High angled splits are produced for this testing situation. It is important in the rotation over pin arrangement (whether drive is from one or both fibre ends) to determine the contribution of the internal hysteresis and surface friction to the torque developed in the fibre to assess how much the frictional forces contribute to the final failure. Calil et al. (1980) developed a formula to obtain the torque, MH, due to the bending hysteresis: MH =

η EC 2θ f 4π 2 ρr

(5.1)

where h = shape factor, E = specific modulus, C = linear density, r = density, r = minimum radius of curvature. For a particular experiment, the numerical values were:

η = 1 ( circular cross-section ) E = 9 N tex

(N.B. N tex

−1

= N (gm km −1 )

−1

C = 4.2 tex

ρ = 1.38gcm −3 r = 12.7 × 10 −3 cm

θ =π 2 f= Therefore: M H = f × 36 × 10 −6 Nm Table 5.1 shows different values of MH for various values of f.


79

Table 5.1 Values of MH for various values of f f

0.01

0.1

0.5

π/4

MH (Nm)

3.6 × 10−7

3.6 × 10−6

3.6 × 10−5

3.6 × 10−5

With a coefficient of friction of 0.2 and a fibre tension of 7.1 × 10−2 N, the torque due to this external friction would be 3 × 10−7 Nm. If the fibre is nearly elastic (f = 0.01, i.e. low loss factor), then the torque due to external friction would be about the same as that due to internal hysteresis. However, most fibres are viscoelastic and since f = 0.1 or greater, and then relatively high hysteresis and high loss factors result. The resulting torque due to internal hysteresis would be an order of magnitude or more than that due to the external friction, indicating that friction would have a small effect in the rotation over a pin test procedure. The combination of cyclic bending with torsional shear stresses is likely to be the major source of the breakdown of the fibres.

5.4.3 The movement of the neutral plane The movement of the neutral plane, when a fibre is bent, may have an effect on fatigue characteristics and on the eventual fracture morphologies. The effect is illustrated in Fig. 5.6. A fibre yields easily in compression because the aligned long chain molecules do not support a heavy load effectively, as shown in Fig. 5.6(a). Because of this, when the fibre is bent as in the rotation around a pin test, the neutral plane will tend to move towards the tension side as shown in Fig. 5.6(b). The part outside the neutral plane will suffer more severe tension–compression cycling effects as indicated in Fig. 5.6(c). It is possible that the outer part will break up during a first stage and the central region will break up separately during a second stage.

5.4.4 Morphologies of fibre fractures of single end driven rotation of a pin tests Figure 5.7 (a) to (d) show fibres tested using the rotation over a pin procedure with drive from one end (Wong 1975). The nylon and polypropylene fibres show the typical multiple splitting characteristics. Tests (a) to (c) were conducted in air at room temperature and test (d) was conducted in a water environment at room temperature. Figure 5.7 (a) and (b) appear to show that the multiple splitting occurred in the outer region of the fibre, implying that this may have been an initial stage of the failure due to the more severe tension/compression action caused by the movement of the neutral plane as described in Section 5.4.3.

80

Fatigue failure of textile fibres Stress

Strain (a) T C C NP (b)

C (c)

5.6 (a) A typical fibre stress–strain curve which shows that yield is low in compression; (b) neutral plane location, NP. The dotted line is the central plane, which would be the neutral plane in an ideal material obeying Hooke’s law; (c) cross-section, showing the rotation of the neutral plane. The outer zone is in tension compression (T and C) and the inner zone is always in compression.

The definition of strain amplitude, ε, used in Fig. 5.7 and throughout this chapter is:

ε=

r r+R

(5.2)

where r = fibre radius R = pin radius This calculated oscillating strain essentially occurs at the outer diameter of the circular cross-section of the fibre. This strain reduces linearly with distance towards the centre of the fibre. However, this definition assumes the neutral plane stays central and hence can be considered only an apparent strain and will not be quantitatively accurate. The supported weight, which is rotating during the test, is essentially a mean load and the applied strain amplitude produces an oscillating load about this mean load.


81

(a)

20 mm

(b)

20 mm

(c)

10 mm (d)

20 mm

5.7 Fibres tested using rotation over a pin, drive from one end. (a), (b) Nylon 66, 1.7 tex, failed at 3690 cycles, opposite ends, strain amplitude 13.6%, supported weight 12.8 gm; (c) Nylon 66, 1.7 tex, unbroken after 5000 cycles, strain amplitude 13.6%, supported weight 12.1 gm; (d) Polypropylene, failed at 3435 cycles, strain amplitude 16.5%, supported weight 5.8 gm.

Figure 5.12 shows fracture morphologies of fibres in actual use, with the axial splitting being a predominant feature. Hence it is reasonable to assume that the biaxial rotation fatigue test procedures described in this chapter will have some similarity to the actions undergone by fibres in actual use. The axial splitting is likely to start from the surface of the fibre where the strain amplitude is the greatest and where small defects are like to occur initiating the shear stresses which result in the axial splitting as in Fig. 5.7(c). The cracks are likely to penetrate towards the centre before final fracture occurs (Fig. 5.7(a) and (b).

82


5.5

Rotation over a pin (single end drive)

5.5.1 Disadvantages and advantages Experiments conducted by Hearle and Wong (1977 a,b,c,d) using the single end driven, rotation over a pin technique, will now be described. The technique will be called ‘rotation over a pin’ from here on. It should not be confused with Calil’s biaxial drive rotation over a pin. A schematic drawing of the technique is shown in Fig. 5.1(b). A disadvantage of this rotation over pin technique is that friction between the fibre and pin would contribute to the failure. In the application of fibres, e.g. in clothes, friction is not likely to be combined with the fatigue action. However, it was shown in Section 5.4.2 that the friction should only have a small effect in the failure of fibres which are usually viscoelastic in nature. The other disadvantage is that the freely hanging weight which is rotating will produce some torque on the fibre due to the viscosity of the medium in which the weight is hanging. This will be negligible when the weight is hanging in air. However, in some of the tests to be described the weight is surrounded by water or a similar liquid and failure of the fibre is therefore enhanced by this additional torque. An objective of some of the tests to be described is to do comparative measurements and hence precautions were taken to ensure that addition torque was at least constant in all tests. An advantage of the rotation over a pin is the ease with which a multiple station testing apparatus can be fabricated. This is essential in any fatigue test if meaningful quantitative results are to be obtained because of the extreme variation of results from fatigue tests and the difficulty in obtaining meaningful mean and median values. Another advantage is that with such equipment the fibres undergoing fatigue testing can be readily immersed in various environments to study corrosion fatigue effects. This is essential to provide useful data for fibres in use, since the fibres will often be used in many different environments during their lifetimes. Typical examples would be cold and hot water, washing liquid and even alkaline and acidic conditions. The effects of all these environments will be described in the following sections.

5.5.2 Apparatus The multiple ten-stage station apparatus is shown in Fig. 5.8 (Hearle and Wong 1977d). There is a single motor drive to the ten stations which are geared together which ensures consistent drives to each station during testing. A tank also exists which enables the location of the fibre over pin areas to be immersed in liquid for corrosion fatigue testing. When a fibre fails, the hanging weight drops into the cup below which activates a micro-


83

5.8 Ten-station fibre over a pin testing apparatus for corrosion fatigue tests.

Table 5.2 Rotation over a pin tests of medium tenacity nylon and polyester in air and water Fibre

Environment

Diameter (μm)

Strain amplitude (%)

Mass of supported weight (gm)

Median life (cycles)

Mean life (cycles)

Nylon Polyester Nylon Polyester

Air Air Water Water

43 39.5 43 39.5

14.5 13.4 14.5 13.4

10.8 10.8 10.8 10.8

6648 3766 4207 3947

6966 3883 4681 3876

switch which stops a counter which records the number of revolutions or cycles taken to fracture the fibre. A test frequency of 5 Hz was used in all tests reported unless stated otherwise.

5.5.3 Effect of water on nylon and polyester Table 5.2 shows rotation over a pin results comparing medium tenacity nylon and medium tenacity polyester fibres in air and water (Hearle and Wong 1977a). These fibres are commonly used and manufactured by a wellknown company. The tests for air were conducted with the supported weight immersed in water and the fatigued portion of the fibre was in air. The drag on the rotating weight would then be the same for the fatigue tests of the fibre in air and in water and the effect of water could be accurately evaluated. The median and mean life values shown in this table and subsequent tables in this chapter are derived from population measurements of 30 fibre

84


tests in all cases. Median values do have more meaning in the tests because of the many extreme values obtained in many tests. The results show that the water environment has negligible effect on polyester fibres. It is known that water also has no effect on the tensile strength of polyester fibres. In both cases this is because the polyester fibre has no hydrophilic groups in its molecular chains to absorb significant quantities of water to weaken it. There is slight water absorption into the amorphous regions of the polymer since its moisture regain is 0.4%. However, this is shown not to reduce its tensile and biaxial fatigue properties. For nylon, there is over 30% reduction in the median and mean fatigue lives by water environment. The tensile strength of nylon also reduces in a water environment. The reason for these factors is the hydrophilic CO-NH group in the nylon chain which is absorbing water. It was difficult to obtain the same fibre diameters for the two different types of fibre and hence it was impossible to make accurate quantitative comparisons between the nylon and polyester, since the strain amplitudes were not the same. However, since the diameters were not vastly different, it appears the fatigue life of nylon does look to be superior to polyester in the tests in air. However, once the two fibres are placed in water, the performances become more similar.

5.5.4 Comparison of fibres Table 5.3 shows comparisons between some smaller diameter fibres, high tenacity nylon and polyester fibres and a polypropylene fibre. The tests were conducted in water. Again quantitative comparisons are difficult because Table 5.3 Rotation over a pin tests of high tenacity nylon, polyester and polypropylene fibres Fibre

Environment Diameter Strain Mass of Median (μm) amplitude supported life (%) weight (cycles) (gm)

Mean life (cycles)

High-tenacity Water nylon High-tenacity Water polyester Regular Water polypropylene

25

14.5

2

6631

6726

22

13.4

2

3061

4034

28.7

16.5

2

Regular Water polypropylene

28.7

16.5

5.8

No No breaks breaks after after 30 000 30 000 4739 4712


85

of the fibre diameter differences. However, perhaps of significance is that the fact that polypropylene had no fractures after 30 000 cycles and required a mean supported weight of nearly three times nylon and polyester before comparable lifetimes were achieved. These results indicate the good fatigue resistance of polypropylene fibres.

5.5.5 Effect of strain amplitude and supported weight on fatigue lifetimes Sections 5.5.5 to 5.5.8 focus on specific tests on the medium tenacity nylon 6.6 fibres which are similar to those used in the previous sections (Hearle and Wong 1977b). Table 5.4 shows how the fatigue life varied significantly by varying the strain amplitude and weight. This is typical of varying the oscillating and mean stresses in any fatigue test. Water environment results are also shown. Very approximately, doubling of strain amplitude (change from 13.6% to 28.4%) or doubling of weight or mean load (from 5.8 gm to 12.1 gm) resulted in a near order of magnitude or more reduction in fatigue life.

5.5.6 Effect of pH The effect of pH on the fatigue of nylon 6.6 fibres was reported by Hearle and Wong (1977b). Hydrochloric acid and sodium hydroxide were used to produce various pH concentrations. The fibre diameter was 43 μm and the strain amplitude was 13.6%. A weight of 2 gm was used from 0 to 2 pH values and 12.1 gm for pHs from 2 to 14, because of the less corrosive effects of the latter values.

Table 5.4 Effect of strain amplitude and weight on fatigue lifetimes of nylon 6.6 Mass of weight (gm) Strain amplitude (%) In air 13.6 22.1 28.4 In water 13.6 22.1 28.4

5.8

8.3

9.6

12.1


31 090 8 832 4 424

19 855 4 740 2 151

9181 3430 1290

5148 1610 –

14 400 2 351 1 720

7 330 1 123 800

5009 1008 671

4042 – –

86


Figure 5.9 shows that there was an approximate linear increase of fatigue life with increasing pH from 0 to 2. For pHs from 2 to 14 there was little change in fatigue life, showing little degrading effect of the pH value on the fibres (Fig. 5.10). All points are median values of 30 readings. Heilwell (1973) reported similar shaped curves for pH on the tensile strengths of nylon 6.6 fibres, indicating similar chemical degradation effects in the two test procedures. Tests conducted on the tensile strengths of the nylon 6.6 fibres used in the fatigue tests described above showed that a change in pH from ‘7’ to ‘0’ resulted in a 12% reduction in tensile strength (Hearle and Wong 1997b).

Median life, cycles

8000

6000

4000

2000

0

0

0.5

1.0

1.5

2.0

pH

5.9 Effect of pH on the rotational fatigue life of nylon 6.6 fibres, load 2 gm.

Median life, cycles

3000

2000

1000 0

2

4

6

8

10

12

14

pH

5.10 Effect of pH on the rotational fatigue life of nylon 6.6 fibres, load 12.1 gm.


87

This compares with a load reduction from 12.1 gm at pH ‘7’ to 2 gm at pH ‘0’ to produce approximately the same fatigue lifetime. The effect on fatigue lifetimes appears to be considerably more significant.

5.5.7 Effect of temperature The effect of increasing the water environment temperature (Hearle and Wong 1977b) is shown in Table 5.5. A strain amplitude of 14.5% and a weight of 10.8 gm were used. The fibre was weakened significantly at 66 ºC. The glass transition temperature of nylon of approximately 50 ºC would contribute to this. Hence possibly washing clothes made from this fibre at the same high temperature may have a similar effect on reducing the fatigue resistance. A further result reported was that the addition of a common commercial washing powder in a high-temperature test had no additional significant effect on reducing the fatigue lifetime.

5.5.8 Effect of preconditioning fibres Another result reported was that leaving the nylon 6.6 fibre in the environment to be tested did reduce the fatigue lifetimes slightly. Leaving the fibre in tap water (pH 7) for three hours before the test, resulted in a life reduction from 4237 to 3639 cycles (strain 14.5%, weight 10.8 gm). Leaving the fibre in hydrochloric acid (pH 0) for three hours resulted in a life reduction form 1980 to 1670 cycles (strain amplitude 13.6%, weight 2 gm). Therefore the absorption of the environment liquid prior the test has a small effect. More significant were tests reported at 0.83 Hz in water in which a median lifetime of 1418 cycles was found. This is to be compared with the median lifetime of 4207 cycles found when the testing was conducted at 5 Hz. (These latter tests were also conducted in water and at the same strain amplitude as the tests at 0.83 Hz.) These results indicate that a major factor would be that the corrosive action of the environment on the fatigue crack growth has more time to act in the slower cycling tests indicating that the time of the total test is a major factor. Table 5.5 The effect of water temperature on the rotational fatigue life of nylon 6.6 fibres Water temperature (°C)


20 50 66

4207 3996 2356

88


5.5.9 High modulus fibres Using high modulus fibres creates a difficulty because the fibre may not follow the pin curvature when bent around it. A formula by Schoppee and Skelton (1974) can be used to determine the force, FB, needed to bend a fibre to follw closely a specific radius: FB =

πd4E 8D2

(5.3)

where: d = fibre diameter E = elastic modulus D = pin diamter Rotation over a pin tests were reported by Hearle and Wong (1977c) for high modulus fibres. Figure 5.11 shows the fractured end of a high modulus Kevlar-29 fibre. It is clear that the fibre had considerably more extensive splitting than the other fibres reported in this chapter. This reveals the stronger chain bonding, drawing and alignment of this fibre giving it its high strength and modulus. The strain on this fracture was given as an ‘apparent’ strain of 7.7%. This was because equation 5.3 shows that for a weight (force) of 0.48 gm which was used in the fatigue test, only at 2.64% will the fibre follow the pin curvature. So initially the fibre is unlikely to have followed the pin curva-

5 mm

5.11 Kevlar-29 fibre broken by the rotation over a pin technique: 10 788 cycles, apparent strain 7.7%, weight 0.48 gm. The fibre diameter was 12.5 μm.


89

ture, but as the structure weakens it may have followed the curvature during the later stages of its life. Hence Kevlar performs reasonably well in the rotational fatigue test because it is able to yield in axial compression unlike carbon and glass which tend to fracture at the high bending strains. Tests were also conducted on carbon and glass fibres. These will break if forced to follow a small curvature. However, tests performed on carbon with an apparent strain of 1.73% (the real strain would be smaller) and weight of 0.038 gm did not break after 20 000 cycles and a test on glass fibre with an apparent strain of 2.65% and weight of 0.071 gm lasted for a similar lifetime.

5.5.10 Failure of fibres in use Hearle et al. 1998 (Part VI – Case studies: clothing and domestic uses), shows numerous fibres which have fractured in use. Many have the characteristic splitting which showed that some torsional fatigue action is likely to have caused failure. Figure 5.12 shows an example of a fibre fracture resulting from general use that came from a pair of polyester trousers. The similarity to some of the fractures described in this chapter indicate that the rotational fatigue tests conducted may represent in part how fibres fail in use. The results generated in biaxial fatigue testing are hence of practical value.

5.12 Polyester fibres from used trousers showing extensive splitting characteristics similar to fibres which have undergone biaxial rotation testing.

90


5.6

New developments and future trends

There has not been a great amount of literature published in recent years concerning strictly biaxial rotation fatigue of single fibres as described in this chapter. Most new research and developments have been in flexural (bending) and torsional fatigue testing. These techniques show some morphological fracture features similar to biaxial rotational fatigue, such as axial splitting before the final fracture. These techniques are described in other chapters of this book. Hearle (2002a and 2002b) produced a comprehensive, review of all the various types of fibre fracture. These include brittle tensile failures in inorganic fibres, ductile failures in melt-spun synthetics, high speed breaks in melt-spun synthetics, twisting, bending and lateral pressure. Fatigue breaks are also reviewed including tension cycling, repeated bending and twisting. It is concluded again that the commonest form of failure in use is multiple splitting due to weaknesses in the transverse direction. The fracture of highly oriented, chain-extended polymer fibres is described in Hearle (2002c). These include the para-aramid and aromatic copolyesters. The failure in shear from flex fatigue and axial compression is described qualitatively and time and temperature dependencies are modelled by statistical mechanics. Axial splitting is again prominent. Surface abrasion is another prominent feature of these types of fibres. Chawla (2002) also produced a fibre fracture review. He stated that fracture in fibres initates at internal or surface flaws. Because of the high surface to volume ratio of fibres, the incidence of a surface flaw leading to fracture is higher than in bulk materials. These surface flaws initiating the fracture may be microvoids or inclusions. The fundamental processes leading to failure are chain scission and/or chain sliding. Metallic fibres are reviewed and their strengths are very much affected by surface condition and inclusions. Ceramic and silica-based fibres such as glass fibres also have their strengths limited by the same type of crack initiating flaws.

5.7

Advice on ways of reducing biaxial fatigue

A summary of the methods to reduce biaxial fatigue based on results reported in this chapter is now given. The obvious application for most of the methods is for clothing. 1. Reducing the strain amplitude. This was shown to have a significant effect in Section 5.5.5. A practical application could be in continual severe creasing of clothes. 2. Reducing the mean load (or weight in the rotation over a pin test). This was shown to have a significant effect in Section 5.5.5.


91

3. If the fibre has chemical groups which attract water like nylon, its rotational fatigue resistance in water is likely to be poorer than a fibre with no hydrophilic groups. If excessive use in water is required, polyester, with its lack of hydrophilic groups, may be a better choice (Section 5.5.4). 4. Avoid contact with acidic environments. The quantification shown in this chapter was that nylon, pHs lower than 2 caused by hydrochloric acid should be avoided (Section 5.5.6). 5. Where possible avoid use of fibres in higher temperatures, particularly near the fibres, glass transition temperature. The reduction in fatigue life with temperature for nylon fibres was shown in Section 5.5.7. 6. Fibres which are more elastic and less viscoelastic are more likely to be more resistant to biaxial fatigue. It was shown in Section 5.4.2 that less torque would be produced and the hysteresis effect due to bending would be less.

5.8

Conclusions

Biaxial rotational fatigue tests can usually be conducted on synthetic fibres of round cross-section using two main procedures. One procedure is to drive from both ends, a fibre bent through 90º. The radius of curvature can be formed without a pin or where more severe strain amplitudes are required by forcing the fibre to follow the radius of curvature of a pin. The other procedure is by driving the fibre from one end and hanging a weight which will freely rotate during the fatigue test. The procedure of fracture is usually by axial splitting, indicating the significant strength of the fibre along its axis due to its long aligned molecules. The axial splitting and morphology of the fracture breaks look similar to many fracture breaks which occur in actual use, implying that rotational biaxial tests are useful for predicting fibre degradation in many applications. Various severe fatigue parameters and environments can significantly reduce fibre fatigue lifetimes and these should be avoided to prolong the use of the fibres. For sources of further information, relevant references by Hearle J W S (Section 5.9) would be appropriate. Comprehensive descriptions of various types of fatigue failure, and causes of failure and fracture morphologies are detailed. In particular, the reader is referred to: Hearle J W S, Lomas B and Cooke W D (1998), Atlas of Fibre Fracture and Damage to Textiles (2nd Edition), Woodhead Publishing.

5.9

References

Calil S F (1977), PhD thesis, ‘A study on the fatigue and abrasion properties of fibres’, University of Manchester Institute of Science and Technology, UK.

92


Calil S F, Goswami B C and Hearle J W S (1980), ‘The development of torque in biaxial rotation fatigue of fibres’, J Phys D, 13, 725–732. Calil S F, Clark I E and Hearle J W S (1989), ‘The sequence of damage in biaxial rotation fatigue of fibres’, J Materials Sci, 24, 736–748. Chawla K K (2002), ‘Fibre fracture: An overview’, Elsevier Ltd. Hearle J W S and Wong B S (1977a), ‘A comparative study of the fatigue failure of nylon 6.6, PET polyester and polypropylene fibres’, J Textile Inst, 68, 89–94. Hearle J W S and Wong B S (1977b), ‘The effects of air, water, hydrochloric acid and other environments on the fatigue of nylon 6.6 fibres’, J Textile Inst, 68, 127–132. Hearle J W S and Wong B S (1977c), ‘A fibre fatigue tester based on rotation over a pin for use in different environments’, J Phys E: Sci Instrum, 10, 448–450. Hearle J W S and Wong B S (1977d), ‘Flexural fatigue and surface abrasion of Kevlar-29 and other high-modulus fibres’, J Materials Sci, 12, 2447–2455. Hearle J W S, Lomas B, Cooke W D and Duerden I J (1989), ‘Fibre failure and wear of materials, an atlas of fracture, fatigue and durability’, Chichester, Ellis Horwood Limited. Hearle J W S, Lomas B and Cooke W D (1998), ‘Atlas of fibre fracture and damage to textiles’, (2nd Edition), Woodhead Publishing. Hearle J W S (2002a), ‘Forms of fibre fracture’, Elsevier Ltd. Hearle J W S (2002b), ‘Fracture of common textile fibres’, Elsevier Ltd. Hearle J W S (2002c), ‘Fracture of highly oriented, chain-extended polymer fibres’, Elsevier Ltd. Heilwell H G (1973) ‘Notes on Research’ (Report No. 239), Editor, Textile Research Institute, Princeton, N.J., USA. Schoppee M M and Skelton J (1974), Textile Res J. 44, 968. Wong B S (1975), PhD thesis, ‘A comparative study of some synthetic fibres in various environments’, University of Manchester Institute of Science and Technology, UK.

6 Effect of structure–property relationships on fatigue failure in natural fibres L WANG, RMIT University, Australia and X WANG, Deakin University, Australia

Abstract: The fatigue behaviour of natural fibres is important for fibre processing efficiency as well as the properties and performance of products made from these fibres. This chapter presents some fibre fatigue results obtained from various experimental methods and testing conditions. It demonstrates that many factors, including fibre structure and chemical treatments, affect the fatigue failure in natural fibres. Examples of controlling fibre fatigue life for different purposes are also presented. Knowledge of the fatigue behaviour of natural fibres will help product design and care. Key words: natural fibres, fatigue, wool, cotton, fibre structure, fatigue control.

6.1

Introduction

Fatigue is defined as the failure or decay of mechanical properties after repeated application of stress or strain. Fibre fatigue occurs when a fibre is repeatedly stressed at a force level less than that needed to cause failure in a single application. This small force, which can be tension, compression, or both, may not be high enough to break the fibre; but the accumulation of stresses from each load/unload cycle will result in eventual fibre breakage. Textile fibres in their practical use are almost always subjected to stresses and strains. Repeated tensioning, bending, abrading and twisting of fibres are common during fibre processing and end use applications, which cause fibre fatigue. Fatigue is one of the main mechanisms by which fibres fail in use, hence it is important to test the fatigue performance of individual fibres. Accelerated fibre fatigue can be simulated in the laboratory by cyclically tensioning, bending, twisting, buckling a fibre, either individually or in a combined mode. An example of the combined mode is bending a fibre over a pin, and then rotating the fibre or oscillating it to-and-fro over a fine wire mandrel. Figure 6.1 shows some examples of fibre fatigue simulation. The 95


Oscillation

Oscillation

(b)

(a)

Fibre

Pin

Fibre

Fibre

96

Fibre

Tension (c)

(d) Alternation

Oscillation Pin

Tension (e)

Fibre

Fibre

Fibre

Pin Pin

Pin

Tension (f)

Tension (g)

6.1 Fibre fatigue tests, (a) cyclic tension fatigue (Lyons 1962a); (b) buckling flex fatigue (Dunlop and Barker 1973); (c) torsional fatigue (Hearle et al. 1998); (d) biaxial rotation over a pin (Lyons 1962b); (e) bending-abrasion over a pin (Hearle and Miraftab 1991); (f) bending fatigue (Chauhan et al. 1980); (g) cyclic bending fatigue over a pin (Hearle and Wong 1977).

ability of a fibre to non-destructively absorb the energy of this repeated deformation will define the fatigue limits of the fibre. Fibre fatigue has been extensively investigated in the last century. Fatigue tests on most manufactured fibres, in particular high-performance fibres, have been performed to understand the fibre fatigue mechanism, estimate their service life and provide scientific data for their technical applications. Since the uniformity of manufactured fibres is relatively good, their property variations are small compared to natural fibres. This means that fewer tests are required to evaluate the fatigue life of manufactured fibres. Natural fibres, on the other hand, are irregular, and more experimental data are required in order to accurately assess their fatigue performance. Cotton and wool are the main natural fibres today that are used in significant quantities commercially. This chapter examines the fibre fatigue performance of natural fibres, with a focus on wool and cotton fibres. The following section gives a brief account of the structure and morphology of cotton and wool.

Effect of structure–property relationships on fatigue failure

6.2

97

Natural fibre structure and morphology

6.2.1 Wool Keratin fibres, such as wool, alpaca and human hair, have complex structures and morphologies. The fibre surface is covered with scales (Fig. 6.2), which are collectively referred to as cuticles. The cuticle cells are attached at the root end and they point towards the tip end of the fibre, like tiles on a roof. Enclosed by the scales is the main body of the fibre, which is an assembly of closely packed cortical cells, known as the cortex. It accounts for approximately 90% by weight of a wool fibre (Anderson et al. 1972). There are three types of cortical cells: ortho-cortical, meso-cortical and para-cortical cells. Intercellular adhesion is provided by the cell membrane complex. Wool fibres have a bilateral structure with an ortho-cortex on one side and a para-cortex on the other. The ortho-cortex and para-cortex have slightly different compositions, structures, and properties, and the radially asymmetric cortical properties result in the unique property of natural crimp for wool (Morton and Hearle 1993). Fibre crimp, with para-cortical cell on the inside of the crimp and ortho-cortical cell on the outside, causes the major axis of elliptical cross-section fibre to rotate along the fibre length, which

6.2 SEM images of (a) wool, and (b) alpaca fibres.

98


may influence the evaluation of some fibre properties such as the fibre flexural rigidity. A loosely packed porous region called the medulla is located near the centre of the fibre and is usually found in coarse wool and some hair fibres. Each of the morphological components contains various structural elements which affect the fibre’s fatigue properties. Wool fibres consist of polypeptide chains held together by salt linkages, cystine linkages, and hydrogen bonds. These long-chain molecules are coiled in unextended fibres, and they are uncoiled as the fibres are stretched. When a moderate stress is applied, the cross-links remain unbroken and they prevent the chain molecules from slipping completely over each other. This assists greatly in the restoration of the initial structure after the stress is removed. Therefore, wool has high extensibility (normally >30%) and recovery capability.

6.2.2 Cotton Cotton fibres are anisotropic and have a complex morphological structure. They are single cell seed hair that grows around the seeds of the cotton plant. Cotton fibre quality is governed by numerous factors including fibre growth environment. Large variations in the shapes (particularly length and diameter) and maturities of fibres are inevitable. As a result, significant differences in fibre properties exist within bales, even within a given boll and on a single seed (Bradow and Davidonis 2000). Cotton fibre diameter width mostly ranges from 8 to 20 μm and its length ranges from less than 20 mm to 35 mm and beyond. A mature cotton fibre contains a primary and a secondary wall enclosed by a waxy cuticle (Fig. 6.3a) (Maxwell et al. 2003). The primary wall or outer wall is on the outer surface and less than 0.5 μm thick. It consists of a mixture of cellulose fibrils, waxes, and pectin. Once traces of pectin, waxes, proteins, etc., are removed, the remainder is a natural polymer of pure cellulose. The secondary wall is laid down on the inside of the primary wall next to the lumen. It is generally assumed that this deposition has a periodicity associated with day and night, which results in the formation known as growth rings. Between these two wall portions, a third layer of different helical pitch is often recognized, the so-called winding or S layer, which forms part of the secondary wall (Heyn 1966). The cellulose molecules in the cell walls are arranged in a helix around the central axis of the very long cylindrical cell. Each fibre is made up of twenty to thirty layers of cellulose coiled in a neat series of natural springs, which gives cotton unique properties of strength and durability. Maturity of cotton fibres is a measure of the extent of development of their primary and secondary walls. Both immature and over-mature cotton fibres are undesirable for textile applications.


99

6.3 (a) Schematic structure and (b): longitudinal section image of cotton fibres.

The lumen in the centre of the fibre is a narrow canal-like structure running the length of the fibre. It will be closed gradually by the growth of cell walls. Upon drying, the entire cell collapses and forms ribbon-like structures of somewhat irregular diameter with periodic twists or convolutions along the length of the fibre (Fig. 6.3). The twist forms a natural crimp that enables the fibres to adhere to one another, making it one of the most spinnable fibres. Fibres stretched in water and then dried are without convolutions and are stiffer (Hearle and Sparrow 1979). Cotton can be altered by using chemical treatments or finishes. Mercerization, for instance, is the process of treating cotton yarns or fabrics with sodium hydroxide for increasing absorbency, lustre, softness, and strength. During mercerization, fibre swells and its cross-section becomes rounder. The strength of the fibres increases markedly and their extensibility decreases due to the swelling effect, which causes untwisting of the fibre and creep, and further reduces its degree of crystallinity and its crystallite length (Haig Zeronian et al. 1990).

6.3

Fatigue of natural fibres

Fibre failure may occur because its maximum strength in a particular mode of deformation has been exceeded during its lifetime. Quite commonly, however, a fibre fails because it has been subjected to cumulative application of relatively low stresses, which are well within the elastic limit. When a fibre is subjected to repeated stress, the stress built up within the fibre increases continuously, and, when this stress becomes sufficiently large to overcome the cohesive forces acting within the fibre, rupture occurs. Hearle

100


et al. (1998) used four principal testing methods to investigate the fibre damage and breakage: 1) tensile fatigue; 2) flex fatigue; 3) bending and twisting fatigue and 4) surface abrasion. All of these methods include application of combined forces to the test specimens. The fatigue patterns observed from these tests have been found very similar to that most commonly found in worn textile materials. This section presents some fibre fatigue results and demonstrates that the structure of natural fibres affects fibre fatigue performance.

6.3.1 Fibre irregularity and fatigue In general, material non-uniformity affects the fatigue behaviour to a great extent. Natural fibres such as wool and cotton exhibit considerable variations in cross-section area along the fibre length. Characteristic abnormalities occur in natural fibres because their growing conditions constantly change. An interruption in the growth of the cotton hair, such as drought or premature harvesting, can prevent the secondary cell wall from developing, and produce an immature fibre. In silk, the fibre cross-section and cocoon length are influenced by the condition and diet of the silk larva. An illness in the animal or an absence of suitable grazing produces thin or tender places on wool. Weathering and ultra-violet radiation often degrades wool (tips in particular) prior to clipping, causing variations in fibre tensile properties. Figures 6.2 and 6.4 show examples of irregularity on wool. Because of the non-uniformity of the structure and variability from fibre to fibre, fibres rupture at different levels of stress according to the weak-link theory (Peirce 1926). Consequently, their fatigue life will be scattered over a fairly wide range, which is an inherent part of the fatigue behaviour of natural fibres. Fibre non-uniformity also exists at the microstructural level at which the fatigue originates. Ruptures are usually initiated at some structural imperfection on, or in, the fibre. Upon application of stresses, these imperfections soon develop into sub-microscopic cracks and grow to some critical size, which will eventually cause the fibre to break. Chauhan et al. (1980) studied the flexural bending fatigue of single cotton fibres both in the form of raw cotton and after slack mercerization, and concluded that the cotton fibre cell-wall thickness affects fibre flexural bending. There is a good correlation between the cell-wall thickness and the number of flexural cycles to rupture. The greater the cell-wall thickness of a cotton fibre (assuming other properties being equal), the greater its flexural-fatigue life will be. Slack mercerization increases the fatigue life of cotton because of the improvement in the fibre uniformity through the removal of some weak places in the fibres. Chauhan et al. (1980) also suggested that the distribution of fibre fatigue life can be described as a Weibull


101

26 (a) Fibre diameter (µm)

24 22 20 18 16 14 0

(b)

20 40 60 Fibre length from tip (mm)

80

Diameter: 30.6 µm

30 mm

6.4 (a): Diameter variation along the length of a fine wool fibre; and (b): 3D profiles of a coarse wool fibre (Deng et al. 2007).

distribution when the frequency of flexural life is counted at a certain interval. However, when the distribution of flexural life is classed at an interval of every 2000 cycles for all four types of cotton used in their report, the histogram as shown in Fig. 6.5 clearly does not follow the Weibull distribution. Nevertheless, this example does show that the fatigue life of natural fibres varies within a fibre type and between fibre types. Therefore, the geometrical and internal structure irregularities of natural fibres should be important factors for estimating early fatigue failures. It should be mentioned that most fibre fatigue tests require tensioning the test specimen, and the pre-tension should be governed by the fibre linear density so that test results can be comparable. This is difficult in practice, one way to minimize the test error is to increase the number of test specimens.

102

Fatigue failure of textile fibres 70 Cotton samples Giza 45 Giza 45 mercerized Sea island Hybrid cotton

Frequency (%)

60 50 40 30 20 10 0 0–2000

2001–4000 4001–6000 Above 6000 Range of flexural cycles to rupture

6.5 Frequency distribution of flexural bending fatigue life of raw and mercerized cottons (data source: (Chauhan et al. 1980)).

6.3.2 Tensile fatigue of some protein fibres Repeated stresses and strains on fibres can substantially change their tensile properties. Under a small constant extension force, the polymer chains in a fibre are extended and gradually drawn past each other as the polymer chain-to-chain bonding is broken. The polymer chains slip to a point where there is not enough chain-to-chain bonding to sustain the load applied and the fibre fails. If a load is cycled onto a fibre enough times then eventually the fibre will fail. The smaller the load, the larger the number of deformation cycles required for fibre failure. The cyclic stress test measures the fatigue performance of a fibre that is subjected to an alternating change between tension and relaxation. The measured variable is the number of load cycles to break the fibre. Animal fibres such as wool, show a stress–strain curve typical of keratinous fibres (Fig. 6.6). Transition of α-keratin to β-keratin in the yield region is the reason for the unique shape of the curve. As shown in Fig. 6.6, after a wool fibre has been stressed or strained then relaxed, it usually shows a lower extensibility. The part of the elongation that could not recover after a longer relaxation period is known as a permanent set. In a stress-recovery cycle, the total elongation is the sum of immediate elastic recovery, delayed recovery and permanent set. The proportions of immediate recovery, delayed recovery and permanent set are dependent upon the test conditions and fibre types. A lower extension speed, a delay in releasing the specimen after extension, or a shorter recovery period would decrease the recovery values and increase permanent set. Conversely, a higher extension speed would decrease the immediate elastic recovery portion and increase delayed recovery value. Such changes would also alter the tenacity and the entire stress–strain relationship.

250

Stress (MPa)

200 150

Stress–strain curve Loading curve Recovery curve Relaxation Delayed recovery Permanent set

103

Immediate elastic recovery


100 50 0 0

10

20 30 Strain (%)

40

50

6.6 Tensile recovery behaviour of a wool fibre.

Immediate elastic recovery occurs mainly in the amorphous region, and is associated with displacement of atoms or molecules from their positions of equilibrium and with their spontaneous and immediate return when the external force is removed. This recovery is predominantly below the yield point at low stresses and strains. Such recovery is possible when a sufficient number of strong cross-linkages are present to prevent the long-chain molecules from sliding over each other, thus facilitating the return of the deformed structures to their original arrangement. Immediate elastic recovery can also result from the straightening of flexible long-chain molecules or from the unfolding of folded molecules. Permanent set is associated with the structural rearrangements including slippage of some long-chain molecules along each other, and the alignment of linear chain molecules due to tensile stress and the breakdown of the secondary bonds between molecules. The molecular rearrangements obtained during the stretching process will remain through new cross-linkages between chain molecules and result in a permanent elongation after the force is released. Delayed recovery can be considered as an intermediate elastic recovery: some displaced molecules gradually return to their initial states after release of the tension. The stress-recovery behaviour affects, to a certain extent, the fibre mechanical properties and fatigue life. Susich and Zagieboylo (1953) compared the characteristics and stress– strain properties of mechanically conditioned fibres, yarns and filaments with their original counterparts. They cyclically extended the specimens 50 times to an elongation level corresponding to 80% of specimens’ respective elongation at break using a strain rate of 62.5%/min, then relaxed the fatigued samples for 1 hour. After the 50 repeated loading and unloading cycles of mechanical conditioning, there is a marked increase of fibre/yarn

104

Fatigue failure of textile fibres 20 (a)

Fineness reduction (%)

Unrecovered elongation (%)

25 20 15 10 5

(b) 15

10

5

0

0 Human hair Wool yarn

Human hair Wool yarn

Silk

Silk

6.7 Increasing length and decreasing fineness after relaxing the fatigued samples (data source: (Susich and Zagieboylo 1953)).

Tenacity (g/dtex)

5

60 Without fatigue Fatigued sample

(a) Elongation at break (%)

6

4 3 2 1 0

50

Without fatigue

(b)

Fatigued sample

40 30 20 10 0


Silk


Silk

6.8 Tensile properties of fatigued and non-fatigued fibre materials (error bar indicates standard deviation; data source: (Susich and Zagieboylo 1953)).

length (Fig. 6.7a) and consequently a decrease in the fibre/yarn linear density (Fig. 6.7b) for all materials examined. However, as shown in Fig. 6.8, the mechanical conditioning did not alter the fibre material tenacity, though it is evident that the elongation of the fatigued samples is significantly lower than that of their original counterparts. The non-recoverable elongation from mechanical conditioning is due to the changed fine structure (orientation of crystallites, slippage of long-chain molecules, and formation of new cross-bonds), and fibre rearrangement and elimination of crimp for the case of worsted wool yarn. On the other hand, the high elongation at break of fatigued materials recovered during the relaxation period suggests that the morphological and fine-structure changes attained by mechanical conditioning are partly reversible.


105

Elongation component (%)

70 Immediate elastic recovery Delayed recovery Permanent set

60 50 40 30 20 10 0

Mohair

Human hair

Wool

Silk

Wool yarn

6.9 Recovery behaviour of protein fibres tested at the breaking point in wet condition (relaxation time: 5 min; data source: (Susich and Zagieboylo 1953)).

The tensile properties of protein fibres are markedly affected by changes in temperature and humidity (Postle et al. 1988). Although permanent set is irreversible, in general, heat or liquids cause shrinkage of the material, and may remove the permanent set. Figure 6.9 shows, in wet tests, the three components of the total elongation at the breaking point: immediate elastic recovery (recovery in the first few seconds), delayed recovery (recovery after 5 min relaxation) and permanent set. Wool fibre, wool yarn, mohair fibre and human hair have similar recovery characteristics in that the permanent set almost disappears and delayed recovery is predominant when swollen in water. The wool single fibres appear entirely elastic because wool swells more than human hair and mohair due to the differences in their structure and chemical composition. The ratio between the swelling-resistant cuticle cells and the easily water attacked cortical cells is higher for hair than for wool. In addition, hair keratin has a higher cystine content (19.0%) than wool keratin (12.2%), and the sulphur cross-bonds of cystine increase the stability of protein molecules (Susich and Zagieboylo 1953). The exceptional recovery property of wool and wool yarn suggests that the crimp in individual wool fibres removed by stretching can be restored spontaneously during the relaxation period. This improves the elastic recovery of wool yarns. Hence, wool fabrics have excellent resilience, crease-retention, wrinkle-resistance, and handle. Water weakens the protein fibre structure, resulting in higher extensibility (comparing Fig. 6.8b with Fig. 6.9), lower yield strength, and lower initial and torsion moduli (Postle et al. 1988). The cuticle of wool, hair and other α-keratin fibres protects the cortex from physical and chemical damage. However, when these fibres are in water, the projection of scales from the fibres is pronounced. Upon agitation and abrasion, the fibres can easily become entangled, scales can be damaged, and the cortex may be

106


weakened. Although the permanent set can be largely recovered in liquids, animal fibres are likely to suffer fatigue damage faster when subjected to wet rather than dry processing. Therefore, care should be taken to minimize the loss of fibre strength and reduction of fatigue life during animal fibre wet processing. Mechanical action during washing could also cause fibre fatigue and damage. Figure 6.9 also shows that silk has a high permanent set and low delayed recovery compared to other protein fibres. This is because silk has a high crystallinity, and the swelling of fibres occurs mostly in the amorphous part of the fibre structure.

6.3.3 Flex fatigue Certain physical properties of a fabric, such as drape, flexibility, handle, creasing, and wrinkle-recovery, are dependent on the flexural or bending properties of yarns, which in turn are dependent upon the properties of the individual constituent fibres. Flex fatigue is usually associated with bending and twisting. The flexural properties and fatigue life of single fibres that are subjected to both longitudinal and transverse stresses are often employed to determine the fibre’s capacity to withstand repetitive loading. Flex fatigue of fibres can lead to different modes of rupture. There are many testing methods to estimate the flex fatigue life of a fibre. Axially reciprocating a fibre over a curved surface until the fibre breaks is a basic approach. Tests on natural fibres such as cotton, wool and hair often lead to multiple splitting of the test specimen (Chauhan et al. 1979; Hearle et al. 1998). Upon fibre bending, material at the inner side of the bend is compressed, while that at the outer side of the bend is extended. The outermost and innermost bend surfaces are at maximum tensile strain and compression strain respectively. This causes cracks at the outside of the bend, kinkbands on the inside of the bend and splitting along the fibre length. The cross-section of natural fibres is not usually circular; they bend about their longer axis (the easiest direction). Hence, as mentioned before, the geometrical irregularity of natural fibres is one of the key factors affecting their flexural fatigue behaviour. Kamal et al. (1984) investigated the flexural fatigue performance of cotton fibres using long and medium-long staples from nine raw Egyptian varieties and an American variety, Pima S-5. They claimed that the long staple cotton varieties have shorter flex lives than the medium-long staple types (Fig. 6.10). A general trend is that the average flexural fatigue cycles to break is inversely related to fibre brittleness and tenacity, and positively related to cotton fibre fineness, Micronaire. Cotton mercerization generally enhances the fibre fatigue life. Through examining the fracture morphology and patterns of flexurally fatigued

107

Long staple category (50% span length: 17.8–18.8 mm) Medium-long staple category (50% span length: 15.5–16 mm)

25000 20000 15000 10000

Dendara

Giza 80

Giza 75

Giza 69

Giza 67

Pima S-5

Giza 77

Giza 76

0

Giza 70

5000 Giza 45

Flexural fatigue cycles to break


6.10 Average number of flexural fatigue cycles to break raw single cotton fibres of different varieties (error bar indicates standard deviation; data source: (Kamal et al. 1984)).

single raw cotton and slack-mercerized cotton fibres, Chauhan et al. (1979) found that the fractured ends of raw cotton fibres showed extensive fibrillation due to repetitive bending stresses. The concurrent compressive and tensile stresses resulted from the flexural fatigue test weakened the lateral cohesion and fibrillar structure of the cotton fibre, resulting in many fibrils to loosen up and separate. After slack mercerization the flexural fatigue fracture patterns showed rather sharp breaks with very little fibrillation. This is because the lateral cohesion was improved from slack mercerization; hence very little slippage of fibrils took place and the fibrils bent to rupture in bundles. Ellison and Zeronian (1984) measured the flexural fatigue cycles of raw wool, scoured wool, phenylisocyanate treated scoured wool, and acetylated scoured wool at different relative humidity (RH) levels. As shown in Fig. 6.11, the mean flex life increased for all wool samples as the relative humidity was raised from 30% to 65%. The increased flex life is a result of hydrogen bond rearrangement in the material, which is facilitated by the higher fibre moisture content at 65% RH. It should be pointed out that although the fatigue life of raw and scoured wools shown in Fig. 6.11 was longer than that of phenylisocyanate-treated and acetylated wools, not all chemical treatments shorten the fibre fatigue life. Some treatments, such as glacial acetic acid, may increase the torsional fatigue life relative to untreated wool.

6.3.4 Twisting and flex fatigue Fatigue studies of fibres can be conducted in three major modes of deformation: tensile, torsional, and flexural. The breaking-twist-angle (BTA) test

Fatigue failure of textile fibres Mean number of flexes to break (×1000)

108

160 30%RH 65%RH

140 120 100 80 60 40 20 0 Raw

Scoured Phenylisocyanate-treated Acetylated Wool type

6.11 Effect of relative humidity on flexural fatigue life of wool fibres (temperature: 22 °C; load during flexing: 0.9 gram; fibres flexed on 42 micron diameter music wire) (data source: (Ellison and Zeronian 1984)).

may be considered as a combination of the three modes, at least tensile and torsional. A fibre with diameter of D is fixed between two clamps with a gauge length of l, then twisted until the fibre breaks. The BTA is obtained with the following equation (Ellison and Zeronian 1978): ⎛ l ⎞ BTA = tan −1 ⎜ ⎝ π Dr ⎟⎠ where r is the number of turns imparted to the fibre before break. Kamal et al. (1984) reported that among the strains of Gossypium Barbadense cotton, there is a negative linear correlation (r = −0.59) between BTA and flex fatigue life of the fibres, and the cotton fibre brittleness as indicated by the BTA was independent of the fibre fineness, wall thickness and immaturity ratio. Long staple cotton fibres had higher BTA values and higher strength, but lower elongation. A brittle material is usually characterized as having a relatively low breaking extension. The ability of a material to survive a twisting deformation is an indication of brittleness. Therefore, decreased BTA implies lower brittleness and higher fatigue life. In other words, among the cotton varieties they have examined, long staple cotton fibres are brittle and easy to fatigue compared to medium long staple cotton fibres. Ellison et al. (1989) examined the BTA and flex life of different fibre types. Table 6.1 shows results of wool and cotton. It can be seen that cotton fibre BTA values are higher than wool, which suggests that cotton is considered to be more brittle than wool. The mean flex life in Table 6.1 agrees with this.


109

Table 6.1 Mean flex life and BTA of wool and cotton fibres at 65% RH and 21 °C (data source: (Ellison et al. 1989)) Fibre

Mean flex life ± standard error (×1000)

Wool Unscoured pima S-5 cotton Mercerized pima Scoured deltapine cotton Mercerized deltapine

148.0 9.0 14.0 17.0 19.0

± ± ± ± ±

24.0 1.0 3.0 4.0 3.0

Diameter (μm)

BTA (°)

26 11 11 12 14

44.69 64.17 64.74 61.88 64.74

As BTA measurement is simple compared to flex fatigue testing, it is an easy technique for single fibre fatigue study.

6.3.5 Buckling fatigue Mechanical buckling fatigue is a particularly important factor in carpet wear as it is considered the principal cause of carpet fibre breakdown. For apparel, fibre buckling fatigue behaviour is related to fabric-evoked prickle. It is desirable that fibres have a low bending modulus and long buckling fatigue life to maintain fabric comfort, handle and serviceability. For carpet piles, fibres must be highly resistant to bending stresses in order to avoid premature breakage during heavy wear. The main difference between apparel wear and carpet wear is that pile fibres are relatively free from tension, whereas the fibres in fabrics are constrained under tension and can be flexed under tension during wear. The damage produced in wool fibres by flexing under compression is dependent on the severity of flexing, i.e. on the magnitude of the strains imposed and on the compressive loading of the fibre. Buckling tests on a wool carpet yarn by Hearle et al. (1998) revealed that after a period of cyclic buckling flex, fatigue from localized sharp kinks led to splits/cracks at cell boundaries, and eventual fibre breaks (Fig. 6.12). Figure 6.12a shows wool fibre buckling yields cracks at kinkbands, which lead to fibre rupture as the cracks become more pronounced. The tendency of wool fibres to split at cell boundaries is demonstrated in Fig. 6.12b.

6.3.6 Abrasion fatigue Fibres in use are often subjected to surface shear. Wherever two surfaces slide, roll or rub against each other, worn surfaces generate friction forces and frictional heat. When fibres slide on solid items, friction can lead to fibre surface damage, deformation and even breakage. This kind of resistance is

110


6.12 Buckling fatigue of fibres in a wool carpet yarn.

referred to as abrasion resistance. The extent of damage depends in part on the type of friction encountered: rubbing, rolling, abrasion etc. (Warner 1995). If a fibre is able to effectively absorb and dissipate these forces without damage, the fibre will show good abrasion resistance. The toughness and hardness of a fibre is related to its chemical and physical structure and morphology and will influence the abrasion of the fibre. To simulate surface abrasion fatigue, a tensioned fibre may be abraded against a rotating rod (Hearle et al. 1998). Anderson et al. (1972) investigated the role of yarn torsional fatigue in the wool fibre morphological breakdown and Martindale fabric abrasion resistance using undyed plain-weave worsted fabric (170 g/m2) made from 35tex 70’s Merino wool yarn. The fabric was treated with nine chemicals separately and the treated and raw fabrics were tested to determine the number of rubs required to produce a hole in the test specimens for abrasion resistance, and the number of torsional fatigue cycles required to break their constituent yarns and fibres. It was found that the Martindale abrasion resistance linearly increases as the resistance to torsional fatigue of the constituent yarns and fibres increases. Anderson et al. (1972) suggested that torsional forces predominate in fibre breakage during flat abrasion testing and in actual wear. Their work also demonstrated that the resistance to abrasion and torsional fatigue can be modified with appropriate chemical treatments. Some treatments increase the abrasion resistance and torsional fatigue lives relative to untreated wool while others markedly reduce them. In other words, the fibre flexibility and brittleness of wool can be chemically tailored for different applications. It should be pointed out that chemical


111

treatment may affect other properties of the fibre products, such as mechanical properties, moisture absorbency, handle and comfort. These properties should be taken into account when modifying the fibre flex fatigue performance. Extensive fibrillation occurs during animal fibre flat abrasion or twisting. Figure 6.13 shows an example due to flat abrasion. The weakening of the cell membrane complex under forces is a key mechanism of animal fibre abrasion fatigue. Once fibres are broken, some fibrils at the broken ends may show the dimensions and shape of cortical cells (spindle shaped cells 80–100 μm long and 3–6 μm diameter). Localized twisting and/or bending forces cause fibre fibrillation and further transverse fracture. Pills on a wool/alpaca fabric were observed to contain large numbers of fatigued fibres/zones, with a high degree of fibre interaction both around established pills and also in areas remote from the pills. As shown in Fig. 6.13, many of the fibres involved in such entanglements had suffered considerable fatigue damage, and some fibres contained multiple fractures from bending and twisting. The progressive introduction of fatigue zones into the pills led to the development of much tighter and more compact entanglements (Fig. 6.13), which caused the pill cores to increase in mass. This occurs because the fatigue zones enable the fibres to bend more easily, and so develop smaller bending radii within the pill structure. As a consequence, fibres that fatigue readily – such as cotton and fine wool – tend to form tight and hard pills.

6.3.7 Bending-abrasion fatigue of wool and alpaca fibres For practical textile applications, both bending and abrasion fatigue properties of fibres are as important as the tensile properties, because the strains to which the fibres are exposed in use are not entirely tensile in nature. Flex fatigue combined with surface abrasion, such as cyclic stress test over a wire (Fig. 6.1(e)), may be used to examine the fatigue damage for textile fibres. Flexing by pulling a fibre backwards and forwards over a pin under some tension is not simple cyclic bending or abrasion alone. The fibre is being bent, abraded and sheared; therefore, the fibre is subjected to flexural fatigue, surface shear stress and tensile fatigue. For apparel textiles, the fibre bending-abrasion and cyclic stress test provides a measure for the tendency of a fabric to pill. The smaller the cycle number to break, the lower is the tendency to pilling, since the pills that develop during textile wear will be shed within a short period. For carpet fibres, a long fatigue life is desired. Wool fabrics have a propensity to pill and wool fibre fatigue directly affects pill development, growth and wear-off. Understanding the fatigue and damage mechanism of animal fibres is important for pilling control. Liu

112


Fibre rupture and fibril end

20 µm

Fibre fracture and entanglement in the pill

100 µm

Fibre bending and twisting

20 µm

6.13 Pills sampled from fabric tested with Martindale pilling test method (fabric was made from 70% 17.9 μm wool and 30% 21.6 μm alpaca blend).


113

et al. (2005) studied the bending-abrasion fatigue of wool and alpaca fibres. They measured single wool and alpaca fibre diameter profiles along the length of each single fibre using a Single Fibre Analyser (BSC Electronics), and bending-abrasion fatigue of these fibres on a Fibrestress tester (Textechno), which employs the concept of Fig. 6.1(e). For all fatigue tests, fibre samples with a pretension of 0.8 g were cycled back and forth (10 mm amplitude) under a 90º angle over a wire (50 μm) at 5 cycles per second, and the number of cycles up to fibre breakage was measured. Their results indicated that differences in fatigue performance exist between wool and alpaca fibres, even though both fibre types are of animal origin. As shown in Fig. 6.14, both types of fibres follow a general trend that the bending-abrasion cycles at fibre break increase as fibre diameter increases. This reflects an increase in surface thickness and fibre bending rigidity as the fibre diameter increases for both types of fibres (Owen 1965). In addition, the 0.8 g pretension applies to all size fibres, which means thin fibres are likely to fail earlier than coarse fibres. Figure 6.14 also shows that it took longer to break the alpaca fibres than wool, and on average alpaca fibres recorded more abrasion cycles than the wool fibres of similar diameter. The mean fatigue cycles for wool and alpaca are 870 and 1209 respectively, which are statistically significant at the 5% level. When the mean numbers of cycles to bending-abrasion fatigue are divided by their respective mean fibre diameter, the alpaca fibres also withstand a significantly higher number of cycles than the wool. Reasons for this have been postulated based on their surface, mechanical property and structure differences. Liu et al. (2005) revealed that, for similar mean diameter fibres, the diameter variation along the length of wool fibres is higher than alpaca fibres, and wool fibre has a lower tenacity and elongation to break than the alpaca fibre. According to the weak link theory (Peirce 1926), fibres with high irregularity would have a lower tenacity and break at the weakest (finest)

Bending-abrasion cycles

5000

Linear regression R2 = 0.496 Linear regression R2 = 0.214

Alpaca Wool

4000 3000 2000 1000 0

15

20

25 30 Fibre diameter (µm)

35

6.14 Relationship between fibre diameter and bending-abrasion fatigue cycles (data source: (Liu et al. 2005)).

114


point. This means that the high breaking energy and low irregularity of the alpaca fibre will prolong its fatigue life compared to wool. The fact that the coefficient of friction of alpaca fibres is significantly smaller than that of wool fibres (Wang et al. 2005) could be one of the main reasons for the higher abrasion cycles for alpaca fibres. As natural fibre will flatten due to its soft structure when placed over the wire, friction will be considerable when testing the fibre because of its coarse outer surface. Alpaca fibre surface is smoother due to its higher scale frequency and lower scale height than the wool fibres (Liu et al. 2004). The lower coefficient of friction for alpaca fibres will reduce the frictional resistance during abrasion. On the other hand, high friction on wool will generate more heat than alpaca during abrasion-bending tests. The increased heat could contribute to the earlier failure of the wool cuticle and scale structure. Abrasion severely damaged the inner bend surface as shown in Fig. 6.15. At the time of fibre breakage, the fibre cuticle of the abraded section had been removed by wire abrasion, and the exposed cortical cells had split along the fibres due to the shear stress. It was observed that the abrasion between fibre and wire first caused scale cracking, lifting and peeling. The abrasion led to a progressive thinning of the fibre. As the flex test continued, the elastic cohesion was lost due to the separation of fibrils and the fracture of macro-fibrils. Fibre thinning and structural disintegrating (cortical cell splitting) eventually led to the fibre rupture. This fatigue mechanism can

Wool fibre

Alpaca fibre

6.15 SEM images of broken fibre ends from bending-abrasion tests (Liu et al. 2005).


115

be found in many wear situations, such as in the knee area of work pants. Such fibre fatigue can also happen during the early stages of fibre processing because of the interactions between fibres and carding/combing elements. Bending caused fibre cracks on the outer bend side of fibres, as shown in Fig. 6.16, due to the shear stresses. Once the cuticles were all sheared from the surface, the cortical cells were exposed to abrasion, leading to separation and sometimes fibrillation on a macro-fibrillar level. As the cuticle is stiffer than the fibrils, cracks develop and form large fibre axial splitting that is associated with bending. Figures 6.15 and 6.16 also show that, after bending-abrasion fatigue tests, the shape of the broken wool and alpaca fibres is curled, and the wool curled more than alpaca. Also, scale peeling started along the direction of curvature for wool, but almost only on one side for the alpaca fibre. This is simply because both fibres have natural crimp. The amount of crimp depends on the distribution of para- and ortho-cortical cells. Wool fibres have a distinctive bilateral structure, but such structure is less obvious for alpaca fibres, particularly in coarse fibres (>23 μm). Therefore alpaca fibres have

Wool fibre

Alpaca fibre

6.16 Fibre cracks due to bending (Liu et al. 2005).

116


6.17 Wool curvature effect on abrasion (Liu et al. 2005).

Table 6.2 Methods for Eri silk cocoon degumming Degumming method

Degumming procedure

Normal degumming

With sodium carbonate 2 g/L and sodium dodecyl sulphate 0.6 g/L at 100°C and a material to liquor ratio of 1 : 25 for 120 minutes. With sodium carbonate 2 g/L and sodium dodecyl sulphate 0.6 g/L at 120°C and a material to liquor ratio of 1 : 10 for 120 minutes.

Intensive degumming

no obvious crimp shape and a much lower crimp frequency than wool fibre. The wool and alpaca fibres have an elliptical cross-section and tend to bend about the longer axis of their cross-sections. Hence, as further illustrated in Fig. 6.17, the inner bend of the crimp is likely to be abraded during the bending-abrasion tests. Because of the high crimp, wool fibre will suffer a high magnitude of twisting during tests. This could be another reason for wool to have a shorter flexural fatigue life than alpaca.

6.3.8 Bending-abrasion fatigue of silk Silk fibres are very pliable, possess moderate tensile strength, and are very tough. Consequently, they can withstand a large number of repetitive bending cycles before they rupture. However, degumming conditions can dramatically affect the silk fibre fatigue life. Rajkhowa et al. (2008) degummed Eri cocoons using two degumming regimes as shown in Table 6.2, and then


117

6.18 Fibrillation of silk fibres due to bending abrasion.

performed bending-abrasion fatigue stress tests on single degummed fibre using the FibreStress tester. Their results show that the number of cycles required to break a fibre (average of 75 tests) were 2307 and 83 for normally degummed and intensively degummed Eri silk respectively. As the only differences between the two degumming methods were chemical concentration (liquor ratio) and operating temperature, it is clear that the conditions of chemical treatment on silk fibre significantly affect the silk fibre fatigue life. As can be observed in Fig. 6.18, under repeated cycles of bendingabrasion, significant fibrillation occurred in mulberry, muga and Eri filaments tested. These indicate that micro and nano fibrils within the silk fibre structure were split during the bending-abrasion process. Weakening the interfibrillar cohesiveness is the primary fatigue mechanism for silk fibre bending abrasion tests (Rajkhowa et al. 2008).

6.3.9 Factors affecting fibre fatigue life The evidence concerning the significance of various factors that influence the fatigue behaviour of natural fibres is sometimes confusing because the experimental data were obtained under widely different conditions and using different fatigue testing methods. Hence, fibre fatigue results may not be comparable. In addition, natural fibres are highly irregular, which results in great variations in fibre fatigue lives. Nevertheless, it is clear that fibre structure, physical and mechanical properties, treatments and service conditions (temperature, RH, bending/abrasion/twisting/buckling/tensioning, etc.) are the main factors affecting the fibre fatigue life. For example, when the fibre structure becomes more rigid or less able to absorb the energy, flex life will reduce. Hence, the flexural fatigue life of cotton fibre is inversely related to its brittleness (Kamal et al. 1984), and directly related to its cell wall thickness (Chauhan et al. 1980) and mercerization conditions. Compared to raw jute, mercerized jute fibres have a significantly longer fatigue life, and bleached jute fibres can withstand many more flex cycles before rupturing (Kundu 1987). In general, cellulosic fibres treated with NaOH solutions

118


are expected to have improved fatigue resistance, while the contrary is expected for keratin fibres. The ability of a fibre to withstand repeated distortion is the key to its fatigue life. High elongation, elastic recovery and work of rupture are considered to be more important factors for a good degree of fatigue resistance in a fibre than is a high strength. Therefore, despite its low strength, wool has the best bending fatigue life among cotton, natural silk, glass and many regenerated fibres (Thomson and Traill 1947). Although flax and other bast fibres have high tenacity, they cannot withstand repetitive bending because of their brittleness and low toughness. Upon bending and abrasion, they can easily divide into small single fibrils, then break.

6.3.10 Processing induced fatigue Fibres during processing and use are subjected to complex deformations; fatigue life of fibres is an important factor affecting end-product quality. The interrelationships exist between various physical and chemical stresses applied to fibres, and the impact of these stresses upon fibre properties. During fibre processing, fibre to fibre and fibre to machine parts interactions result in fibre fatigue and breakage. Figure 6.19 shows examples of such damage and breakage. The interactions cause the wool fibre tip splitting and length reduction. Adding lubricants can significantly reduce fibre damage and preserve fibre length.

6.3.11 Fabric pilling and fibre fatigue The primary uses of natural fibre are apparel and household items. Fabric wear-out and pilling are perhaps the major fatigue related concerns for natural fibres. Numerous distinct and independent mechanisms are involved in the wear of a polymer, such as abrasive wear, fatigue wear and adhesive wear. Fabric wear-out not only causes fibre loss, but also generates fabric pilling, which is one of the most serious quality problems, especially for knitwear. One consequence of wearing is the formation of surface fuzz and pills for apparel. When free fibres fatigue to the point where they fracture and are lost from the fabric surface, they easily get entangled with fibres that are firmly attached to the fabric surface and form pills. The pills bring about an unsightly appearance and degradation in fabric handle that may seriously compromise the fabric acceptability. Its development is undesirable and usually considered a nuisance by the consumer (Goswami et al. 1980). Fibre fatigue properties have an effect on fabric pilling and product performance. Single fibre abrasion and flex fatigue, and single fibre tensile properties affect the pilling propensity of their resultant fabrics. The fatigue


119

6.19 Wool fibre damage caused by (a) saw-tooth, and (b) round pin during processing.

of fibres in pills often results from a combination of slow and gradual bending and torsional deformation (Goswami et al. 1980). In most cases, fatigue damage within a pill appears to be either transverse cracking or kink-band cracking with a certain amount of skin-shedding in the crack zones (Cooke 1981). Goswami et al. (1980) have suggested that pilling might well be a result of the interlocking of the surface scale-like structure produced by slow and

120


gradual cyclic torsional deformation of the fibres. The majority of damaged fibres within the pills result from fatigue due to transverse cracking, kinkband cracking with a certain amount of skin-shedding in the crack zones or snarling together with apparent helical cracking and division of the fibre surface. A combination of bending and torsional deformation fatigue and the fatigue morphology characterizes the damaged fibres in use. The considerable fibre fatigue that occurs both around established pills and remote areas encourages fibre lashing or flailing, and as a consequence increases the probability of inter-fibre entanglements. Pilling can be considered to occur in three main stages: fuzz generation, pill formation and growth, and pill wear-off. The formation of fuzz on a fabric surface is the first stage of the pilling process. Fibre fatigue and fracture, to some extent, contribute to the fabric surface fuzz. After the initial formation of fuzz, a loose fibre entanglement develops within an area of high fuzz density, and a number of fibres eventually entangle and tighten to form a spherical mass (a pill), which attaches to the fabric with anchor fibres. Pill wear-off occurs when anchor fibres are broken and/or pulled out (Cooke 1985). Breakage occurs when the pill detachment force exceeds the combined tensile resistance of fatigued anchor fibres. With high twist yarns and/or tight fabric structures, anchor fibre fatigue and fracture dominate. The pills that form from fibres with high tensile strength are more likely to become immobile and remain on the fabric surface without wearing off. Fibre fatigue is important in the mechanism of fabric fuzz and pill growth, and it plays a dominant roll in the process of pill wear-off. Cooke (1984) has conducted wear trials and examined a wide range of knitted garments made from crossbred wool. It was found that fabric breakdown occurs largely as a result of yarn degradation, which may be caused by loss of fibres due to direct frictional wear and flexing and torsional fatigue that resulted in transverse cracking, breakdown and fall-out of short-fibre debris. Naylor and Williams (1988) studied wool knitwear that had been subjected to treatment with two different levels of potassium permanganate. Based on physical measurements, improvements in pilling performance when using the oxidant have been rationalized in terms of a significant reduction in fibre torsional fatigue. Changes in fibre tensile and frictional properties were insufficient to explain the large improvements in pilling performance they have observed. The work by Anderson et al. (1972) on fabric abrasion showed considerable wool-fibre fibrillation in torsionallyfatigued yarns. Fibre torsional fatigue is of major importance in fabric pilling and wear-off. Fibre properties such as breaking strength and flex life are considered to be the presiding factors that determine the mechanical resistance of a fibre to break, and govern the wearing-off of pills. For weaker fibres, as in the case of cotton and wool, the rate of wear-off is generally high. In most yarn


121

structures, the outer layer fibres are not fully secured. These surface fibres will develop fuzz and anchor on the surface. Provided that anchors can be replaced when they were either pulled out or fractured, pilling and pill growth will proceed throughout the life of a garment. All these indicate that modifying fibre fatigue life is a potential approach for pilling propensity control.

6.4

Methods of controlling fatigue in natural fibres

In most cases, long fatigue life is preferred to increase the service life of the fibre end products. In some situations, however, low fatigue life may be favourable, for example low fatigue life in fabric surface fibres will reduce fabric piling propensity. Hence, fatigue properties in natural fibres may be tailored to meet different application requirements. For instance, fibres in use will often undergo fatigue in the presence of moisture, high-temperature or liquid environment. Cotton fibres, as well as other natural cellulose fibres, are susceptible to damage by micro-organisms. Under proper conditions of temperature and humidity, the organisms become active and can cause drastic structural damage in cotton fibres. As an extreme case, only three weeks burial in moist biologically active soil at 200 ºC is sufficient to reduce cotton, linen or wool fabrics to such a tender state that they disintegrate under their own weight (Hearle et al. 1998). To eliminate biological attack, natural fibre materials should be stored under conditions of absence of water, low temperature ( 0 if Sa + Sm = ST. For smaller Sa the structure survives a hundred cycles Ns as a result of strain hardening and therefore the nearly horizontal asymptote occurs. Mathematically, this type of S-N curve (for the case of Sm = 0) is described by the so-called Basquin relation:

Effect of textile processing on fatigue

137

High cycle

Low cycle Finite life

Fatigue strength (Sa)

Infinite life

Nf

Sf

100

101

102

103

104

105

106

107

108

Number of stress cvcles (N)

7.3 Typical S-N curve.

Sa ≈ ST for N ≤ N S Sa = k1 ( N )

k2

7.1

for N s < N ≤ N f

Sa = Sf for N > N f For higher values of Sm > 0 the S-N curves are shifted down and Sf is reduced. The S-N curve can be then constructed replacing Sa by effective stress range Sef and is defined as ⎡ ⎛ S ⎞2 ⎤ Sef = R ⎢1 − ⎜ m ⎟ ⎥ ⎣ ⎝ ST ⎠ ⎦

−1

The effect of Sm on fatigue is not as large as the effect of Sa (fatigue is primarily a consequence of cyclic loading). The fatigue limit value Sf is often proportional to the material strength ST, i.e. Sf ≈ a ST

7.2

The proportionality factor for metals and alloys is in the range 0.35 < a < 0.5. It has been observed that the fatigue strength at 1000 cycles is approximately 0.9 ST. For the case when Ns = 1, k1 = ST. From these simplifications k2 = log(0.9)/log(1000) = −0.0153. In the case when external factors influencing fatigue are present modified fatigue limit is Stf expressed as

138

Fatigue failure of textile fibres m

log ( Stf ) = log ( Sf ) +∑ ki

7.3

i =1

where ki are correction factors (due to temperature, size, diameter, surface roughness, etc.). The Basquin relation defined by eqn. (7.1) was derived theoretically for fatigue of an ensemble of parallel elements having a two parameter Weibull distribution of tensile strength and for random orientation of these elements in plane [6]. For prediction of tensile fatigue life in polymeric systems (composites) the following assumptions are made [5]: •

the tensile strength Sn after n fatigue cycles undergoes a power law decreasing function of n. For n = 1, S1 = ST • the strength decay is linearly dependent on the stress amplitude 2Nt • the final failure takes place at n = N, when tensile strength is equal to critical value SN. After integration of the strength decay expression, the residual strength ST − Sn is expressed in the form ST − Sn = k3 SN (1 − R ) ( n k4 − 1)

7.3

where R is stress ratio, k3 and k4 are material constants. The number of cycles to failure can be then evaluated from expression ⎛ ST ⎞⎛ 1 ⎞ k − 1⎟ ⎜ ⎟ = k3 ( N 4 − 1) ⎜⎝ SN ⎠ ⎝ 1 − R ⎠

7.4

Equation (7.4) can be used for estimation of parameters k3 and k4 from experimental data couples (Ni, SNi). It is obviously postulated that the scatter in the fatigue is correlated with the scatter in strength data, i.e. the lower ST results in lower N. It was experimentally proven [5], that distribution of uniaxial tensile strength is nearly the same as distribution of fatigue strength F(S). For the strength F(S) the two parameter Weibull distribution ⎛ ⎛ S ⎞ CT ⎞ F ( S ) = 1 − exp ⎜ − ⎜ ⎟ ⎟ ⎝ ⎝ bT ⎠ ⎠

7.5

with scale parameter bT and shape parameter cT are often useful. The expected number of cycles to failure N* pertaining to a chosen probability of failure P(N*) is then expressed as N * = k4 1 +

1 ⎡ bT ⎤ 1c ln (1 − P ( N *)) T − 1⎥ k3 (1 − R ) ⎢⎣ SN ⎦

7.6


139

In fact the N* is quantile of lifetime distribution for probability equal to P(N*). The joint distribution of fatigue life N and stress amplitude Sa should have some limitations based on the physical and statistical constraints [8]. The suitable distribution function has the form [8] F ( N, Sa ) = k5 ( Sa − Sf ) 7 ( N − N min ) 6 k

k

7.7

where Nmin is minimum lifetime, k5 is combined scale factor for lifetime and stress amplitude (or level), and k6, k7 are shape parameters. The mean value of lifetime E(N) is expressed in the form ⎛ ⎞ 1 ⎜⎝ k ( S − S )k7 ⎟⎠ f E ( N ) = N min + 5 a 1 + 1 k6

1 k6

7.8

The simplified procedure for estimation of parameters of joint distribution function F(N, Sa) is described in the work [8]. Strain life curve In the case when failure occurs at low number of cycles the effect of plasticity should be added as well [4]. For this situation it is better to use the socalled strain life curve. In the strain life curve (see. Fig. 7.4) the stress amplitude Sa replaced by the total strain amplitude εt composed from elastic εe and plastic εp components (εt = εe + εp). The elastic strain amplitude for the case of Sm = 0 is computed from Basquin relation (eqn. (7.1)) which can be expressed as εe =

Sf* (2 N f )b E

7.9

Strain amplitude

Δε 2

1 0.1 0.01 0.001 10–4 10–5 100

101

102

103 104 Reversal, 2Nf

7.4 Typical strain life curve.

105

106

107

140


where S *f is the so-called fatigue strength coefficient (Sa intercept at 2Nf = 1), 2Nf is number of half cycles (reversals) to failure and b is fatigue strength exponent different from k2. The value of S *f is approximately equal to ST and b is usually in the range −0.05 < b < −0.12. In the case of means stress Sm > 0 eqn. (7.9) is modified and εe =

(S* − S ) (2 N ) f

m

b

7.10

f

E

The plastic strain amplitude is expressed by Manson-Coffin power law [12] c ε p = ε*f (2 N f )

7.11

where ε*f is fatigue ductility coefficient and c is fatigue ductility exponent. The ε*f is approximately equal to deformation at break εT and c is usually in the range −0.5 < c < −0.8. In Fig. 7.5, the separation of strain life curve into plastic and elastic components is shown For the extremely low cycle fatigue cases eqn. (7.11) does not give satisfactory results and an exponential type model is more useful. For this case fatigue life is expressed in the form [14] N=

exp ( w ) − 1 1 2 exp ( wΔε p ε T ) − 1

where w is material constant and Δεp is plastic strain amplitude. Ramberg and Osgood [13] proposed a technique for describing stresses and strains in material subjected to cyclic loading. The total strain amplitude is divided into two components, but described as a function of stress amplitude Sa

εf

(Δε2 )

0.1

Strain amplitude

1

0.01 0.001 10–4 10–5 100

Δε σf (2N )b c = f +εf(2Nf) E 2

c

b σf E 101

2Nt 102

103

104

Reversal (2Nf)

7.5 Decomposition of strain life curve.

105

106

107

Effect of textile processing on fatigue ε t = ε e + ε pe =

Sa ⎛ Sa ⎞ +⎜ ⎟ E ⎝ A⎠

141

h

where A and h are material parameters of plastic deformation. Conventional low cycle fatigue damage is a surface phenomenon where small micro cracks nucleate and grow on the surface of the material. Bulk stresses and strains are employed to describe fatigue damage because the micro crack growth is too complex. Many correlations between fatigue and tensile properties have been proposed. For a wide range of materials the prediction of fatigue based on the initial modulus, E, ultimate strength, ST, and true fracture strain, εf can be used. This prediction has the form [11] Δε ⎛S ⎞ = 0.623 ⎜ T ⎟ ⎝ E⎠ 2

0.832

( 2 N f )−0.09 + 0.0196 (ε f )0.155 ⎛⎜⎝

ST ⎞ ⎟ E⎠

−0.53

( 2 N f )−0.56

Thermo mechanical fatigue Thermo mechanical fatigue is caused by combined thermal and mechanical loading, where both the stresses and temperatures vary with time. This is typical for processing of textile materials where heat is a result of friction between textile materials alone or textile material and processing machinery parts. This thermal effect can very high, sufficient in some cases to partially (surface) melt thermoplastic materials or degrade thermosets (as natural fibres). This type of loading can be more damaging by more than an order of magnitude compared with isothermal fatigue at the maximum operating temperature. Material properties, mechanical strain range, strain rate, temperature, and the phasing between temperature and mechanical strain all play a role in the type of damage formed in the material. It is simple to assume that temperature dependence of fatigue life, N, is controlled by a thermally activated mechanism. By using eqn. (7.1) the fatigue life can be modified to be ⎛ E ⎞ N = N o exp ⎜ − (S )1 k2 ⎝ RT ⎟⎠ a where E [kJ/mol] is apparent activation energy of the thermally activated processes responsible for change of fatigue life, R = 8.31 10−3 kJ/mol is the universal gas constant and No is reference fatigue life. One of the major differences between isothermal and thermal mechanical fatigue is constraint. When heated, structures develop thermal gradients as they expand. Local expansion near contact points is often constrained by the surrounding cooler material. In this case thermal strain is converted into mechanical strain, which causes fatigue damage. Total constraint exists

142


when all of the thermal strain is converted into mechanical strain. Over constraint can occur in a stress concentration where the mechanical strain is greater than the thermal strain. One measure of the degree of constraint is the ratio of the thermal and mechanical strain rates. Damage here has three primary sources: fatigue (fatigue life Nf), oxidation (fatigue life No) and creep (fatigue life Nc) [9]. Damage from each process is summed to obtain an estimate of the total fatigue life, N N=

Nf No Nc No Nc + Nf Nc + Nf No

7.12

Frequently one of the damage mechanisms is dominant. From a modelling perspective, this suggests that the individual damage models and their associated material properties must be accurate only for those conditions where the life is dominated by that failure mechanism. Fatigue damage is most commonly described by a strain-life curve. Oxidation damage can occur in the form of an oxide surface degradation. An oxide layer can be formed on the surface. Oxidation damage is a function of the strain range, strain rate, and temperature. Creep damage is a function of the stresses, time and temperature. Some models for prediction of oxidation and creep damage, especially for metals, are presented in reference [10].

7.2.2 Fatigue accumulation during processing During processing some loading spectrum is applied on textile materials or structures. The result is shortening of structure fatigue life N. Let the percentage of original fatigue life remaining after loading spectrum application be equal to P [%]. Partial fatigue damage must then be present in material because its original life N has been reduced to P ⎞ ⎛ N r = N ⎜1 − ⎝ 100 ⎟⎠

7.13

The damage may still be invisible, but is present in the material. In the case of variable loading it is often possible to separate loading spectra into m segments (see. Fig. 7.6). In j th segment is constant stress amplitude Saj and number of cycles nj. Let the corresponding fatigue life time be Nj and mechanical work absorbed during nj similar cycles is proportional to nj. In this segment it is consumption of nj/Nj of the fatigue resistance. Let the work absorbed until failure have a constant value W. In the j th segment the relative portion of absorbed mechanical work is wj/W. As a consequence it implies that wj nj = = 1 − Pj 100 7.14 W Nj

143

Stress


Time

7.6 Division of loading spectrum into segments with constant loading.

The Pj [%] is percentage of original fatigue life remaining after j th segment loading. According to the so called Miner rule (linear cumulative damage) [1] the failure occurs when Σwj = W or if the fatigue resistance is fully consumed, i.e. m

nj

∑N j =1

=1

7.15

j

This very simple approach has some serious limitations. If some stress amplitudes Saj are below fatigue limit, the corresponding Nj is infinite and nj/Nj = 0. According to the Miner rule the material will not fail because eqn. (7.4) will be never valid. The Miner rule is therefore based on the physically incorrect assumption that segments with stress amplitudes Saj below fatigue limit are not damaging. In reality these segments are contributing to the extent of damage (in the dependence on actual load spectrum). For flat load spectrum the damage contribution of cycles with stress amplitudes Saj below fatigue limit is low in comparison with steep load spectrum. Results of c = Σnj/N values (assumption of constant N = Nj) at failure obtained from experiments vary from much smaller than 1 to significantly larger than 1. Small values are promoted by zero mean stress. High values are promoted by positive mean stress and step load spectra (relatively small number of severe load cycles). The residual stress contributes significantly to high c values. In a random load spectrum is often c ≈ 1 [2]. The replacing value 1 in eqn. (7.15) by value c ≤ 1 leads to the relative Miner rule. A very simple modification of Miner rule is use of extrapolation of S-N curve to lower stress amplitudes with Saj ≤ Sf in order to predict damage increments of cycles with an amplitude below the fatigue limit Sf . The corresponding Nj can be then calculated from relation ⎛ Saj ⎞ Nj = ⎜ ⎟ ⎝ k1 ⎠

1 k2

7.16

144


where k1 and k2 are parameters of S-N curve. In logarithmic coordinates log (k1) is equal to intercept and k2 is slope of S-N line. Hainbach [3] proposed to replace k2 value in eqn. (7.16) by value 2k2 − 1. Combining eqns (7.14) and (7.16) the percentage Pj is expressed as −1 k2 ⎡ ⎤ ⎛ Saj ⎞ Pj = 100 ⎢1 − n j ⎜ ⎟ ⎥ ⎝ k1 ⎠ ⎣ ⎦

7.17

For characterization of influence of processing or use of textile structures on fatigue life it is necessary to know loading spectrum segment amplitude, number of cycles in each segment and S-N curve parameters. This approach is useful for the case of failure due to high number of cycles. Another possibility for how to evaluate the damage accumulation is to use initial modulus of undamaged material ET, initial modulus just before fatigue failure Ef and initial modulus after some deformation history (loading spectrum) En. The accumulated fatigue damage is then expressed as damage index defined as D=

ET − En E T − Ef

7.18a

In the case when Ef = 0 the accumulated damage is simply expressed as D = 1−

En ET

7.18b

The cumulative damage can then simply be predicted from knowledge of initial modulus after realizing deformation history (loading spectrum). The damage index Dj for the j th segment with constant amplitude of cycles can be expressed as a function of number of cycles nj at constant stress amplitude Saj and fatigue life time Nj in the form [7] ⎛ Saj ⎞ ⎛ n j ⎞ Dj = ⎜ 1 − ⎟ ⎜ ⎟ ⎝ ST ⎠ ⎝ N ⎠

k2

7.19

The problem with use of damage index is nonlinear dependence on the number of loading cycles. For homogeneous materials D is practically zero till n/N ≈ 0.25 and then is progressively increased. For composite materials D rapidly grows for low and high values of n/N. One nonlinear model for describing dependence of D on n/N is published in [7]. For the case of tensile fatigue damage rate dD/dn can be described by relation [16] dD A ( Sa ST ) 5 = dn (1 − D )k6 k

where A, k5 and k6 are material constants.

7.20


145

A recent review of major fatigue models and lifetime prediction methodologies for fibre reinforced polymer composites is presented in [15]. Most of these models are modification of the basic models presented here. These models can be used for prediction of damage extent due to load history during processing and use of textile structures.

7.3

Fatigue of textile structures

Standard fatigue tests are realized under cyclic loading with constant amplitude. Results can be used directly for design purposes or comparison of various textiles for fabrics which will be dynamically (mainly cyclically) deformed during use. For investigation of damage and fatigue during manufacturing of yarn and fabrics the special tests simulating dynamic processes of structure creation are frequently used. The prediction of textile fatigue life during use needs to define wear and maintaining conditions. The cyclic mechanical loading is then often combined with surface abrasion and weathering (action of sunlight and moisture). In some cases cyclic loading is replaced by abrasion only. Textile products as woven and knitted fabrics are hierarchically complex, composed of fibres and yarns (see Fig. 7.7). All kinds of fabric mechanical exposures can be investigated at the microlevel, i.e. level of hierarchical components or macro-level, i.e. level of fabric. •

At micro-level fibre fatigue is caused by a change of their structure due to applied loading spectrum (variation of stress or strain during time). The fatigue process in fibres starts with dislocation movements, eventually forming persistent slip bands that nucleate short cracks. At the yarn level fatigue is caused by fibre fatigue and processes changing configuration

7.7 Typical structure of textile fabrics.

146

•


of fibres (as friction sliding). At the fabric level fatigue is caused by yarn fatigue and changes of yarns configuration. The overall fatigue is then a combination of fatigue phenomena in fibres, yarn and fabrics. At macro-level fabric is considered an anisotropic body and fatigue is characterized by the response of this body to the applied loading spectrum.

7.3.1 Effects of manufacturing processes on fatigue During manufacturing of textile structures textile materials (fibres, yarns) are progressively damaged due to following reasons: •

Action of dynamic load with variable amplitude and often high amplitude spikes. • Contacts between individual fibres or yarns and between fibres and machine parts. The result of these actions is surface abrasion and friction accompanied by production of local heat. • Environmental action of moisture, heat, oxidation and presence of special chemicals. In this section the processes of spinning and weaving from the point of view of process dynamic and fatigue damage are discussed. Spinning Spinning consists of the spinning preparation and the yarn formation. Spinning preparation represents processes in which fibres are transformed to produce a fibrous strand. The yarn forming converts this strand into a spun yarn. Spinning preparation consists of the opening and cleaning, carding, drafting, combing and roving. The spinning preparation may cause fibre damage due to a severe interaction between the working elements and the fibres. This damage may be represented by fibre fragmentation. The extent of fibre fragmentation is typically determined by testing short fibre content, by evaluation of fibre length distribution or by evaluation of bundle strength. The ‘short fibre content’ or SFC is arbitrarily defined as the percentage of fibres shorter than 12.7 mm (0.5 inch). In some situations, short fibre content in the output material is found to be slightly lower than that of the input material [30]. For this case, the simple expression based on mass balance principle for computation of the extent of fibre damage EFD [%] during spinning preparation was proposed by El Mogahzy [30] ⎡ SFCout − SFCin ⎤ ⎡ SFCw − SFCout ⎤ EFD = ⎢ +W ⎢ ⎥ ⎥⎦ SFCin SFCin ⎣ ⎦ ⎣

7.21


147

0.09 UHML: 27.9–28.4 mm

0.08 0.07 Density

0.06

a b c d

0.05 0.04 0.03 0.02 0.01 0

0

5

10

15

20 25 30 Length (mm)

35

40

45

50

7.8 Length distribution of cottons having UHM 27.9–28.4 mm [23].

where SFCout [%] is the short fibre content in the output material, SFCin [%] is the short fibre content in the input material, SFCw [%] is the short fibre content in the waste material, and W [%] is the percent of waste. Some examples of application of eqn. (7.21) are given in [30]. Krifa [23] investigated length distribution of the 67 cotton bales patterns by AFIS measurements. For all cottons the local peak in the range of very short fibres (3.2 mm) was identified (see Fig. 7.8). The origin of this fibre fragment accumulation is likely related to fibre breakage occurring during the mechanical processing of the fibre, or during the AFIS testing procedure itself. A similar peak was noticed by Schneider et al. [31] using a measurement method based on image analysis, and was attributed to fibre breakage during ginning. It was found that unimodal length distribution indicates extensive breakage of cotton fibres [23]. This pattern is characteristic of an immature-weak cotton, but could also be observed for a mature-strong cotton that has an aggressive processing history (i.e., was subjected to excessive damage). Figure 7.9 illustrates the change of the length distribution of cotton that had been processed through an aggressive opening action. The shape of the length distribution depends on the resistance to breakage inherent in the fibre; and on the aggressiveness of the mechanical processes to which the cotton is subjected. Evidence of bimodality observed in length distributions of bale (raw cotton) appeared to correlate with high resistance to breakage (higher strength and more mature fibres). Subjecting the raw fibre to mechanical stresses (spinning preparation) appeared, through a process of breakage, to gradually dissipate the bimodal structure of the distribution, which shifted towards shorter lengths. The unimodal structure can also result from excessive mechanical treatment of the fibre during harvesting, ginning, and lint cleaning.

148

Fatigue failure of textile fibres 0.09 0.08

Raw fibre

Extensively opened fibre

0.07 Density

0.06 0.05 0.04 0.03 0.02 0.01 0

0

5

10

15

20

25

30

35

40

45

50

Length (mm)

7.9 Influence of aggressive mechanical opening on the fibre length distribution [23].

The main types of spinning systems are ring spinning and rotor (or open end) technologies. Ring spinning is characterized by continuity of fibre flow from the input strand to yarn and tension-controlled spinning process [30]. The consolidation mechanism in ring spinning is twisting. Twist is inserted to the fibres by a traveller rotating around a ring flange. The amount of twist inserted in the yarn is controlled by the front roll (or delivery) speed and the traveller rotational speed. In rotor spinning, the drafting mechanism consists of [30]: • • •

mechanical opening using an opening roll air drafting using an air stream and transporting duct doubling mechanism.

The use of a sliver requires a large amount of draft to reduce its size down to that of the yarn size. One of the fundamental differences between rotor spinning and ring spinning is the lack of tension differential in rotor spinning. This low tension has a tendency to reduce yarn strength because of the low fibre migration and torque required to twist the fibres. In principle, rotor spun yarns need more fibres per cross-section than comparable ring-spun yarns due to the insufficient torque at the point of yarn formation. Spinning tension in ring spinning is defined as the tensile force applied on the yarn at the onset of twisting. Variation in spinning tension directly results in variation in yarn strength. Excessive tension or tension peaks may result in end breakage during spinning. In fact, more than 80% of end breakage during ring spinning is believed to result from tension peaks. The relationship between traveller speed and spinning tension FY has the form [30]

Effect of textile processing on fatigue Fy =

μr t Fc sin α

149 7.22

where μr/t is the coefficient of friction between ring and traveller, α is the angle between yarn from traveller to bobbin and a straight horizontal line from traveller to spindle axis. The centrifugal force Fc has the form Fc =

mt ⋅ Vt2 dr

7.23

where mt is traveller mass, Vt is traveller speed, and dr is ring diameter. It is an observed fact that almost all yarn breaks in the ring frame take place just after the delivery from the front nip in the spinning zone, i.e. between the front rollers’ nip and the thread guide. The yarn breakage phenomenon in ring spinning is slippage-dominated, i.e. there is no evidence of fibre breakage. The strength of yarn at the spinning zone is significantly less than the yarn strength obtained by a tensile tester. In general, the spinning tension is considerably greater than onethird of the single thread strength [40]. In fact, a very thin portion of yarn just after the delivery from the front nip causes a yarn breakage in ring spinning [41]. The variations of spinning tension in ring spinning are mainly caused by the irregularities in the rotation of the traveller around the ring. It is an established fact that in each rotation of the traveller there are five peak spinning tensions [40]. The increase of spindle rotation velocity is one possible way to raise the spinning frame efficiency. However, the consequence of this is an increase in yarn tension, which can lead to yarn breakage above an acceptable level. The number of yarn breakages can be limited by decreasing yarn tensions occurring during their processing with the spinning frame. In real conditions of a working spinning frame, it is very difficult to gain information regarding the influence of selected parameters and factors on the dynamic of the twisting-and-winding system. The effective method for obtaining data for the system’s dynamic is based on simulation based on the mathematical models. In reference [42] the simulation programme was used for estimation of the dependency between yarn material parameters and the system’s dynamic. The variation of winding tension during first layer winding (for the balloon heights Hmax) for various materials is in Table 7.1 [42]. A comparison of results for the balloon heights Hmax and Hmin leads to the conclusion that during package formation the average tension values for the yarns analyzed decrease, whereas the amplitudes rise [42]. On the basis of huge simulations it has been stated that the material of yarn and the yarn surface properties influence the yarn tension when spinning with a twisting-and-winding system working imperfectly, as well as with a correctly working system. However, the degree of influence is different depending on the yarn material. [42].

150


Table 7.1 Simulated tension for correctly working twisting-winding system (balloon heights Hmax) Parameter

Minimum tension

Maximum tension

Average tension

Tension amplitude

unit Cotton Wool Flax Polyester Viscose

[cN] 14.67 11.29 15.53 18.22 12.79

[cN] 16.06 12.70 16.07 19.07 14.20

[cN] 15.28 11.92 15.78 18.59 13.42

[cN] 0.463 0.486 0.164 0.291 0.475

Flax material influences most distinctly the course of yarn tension versus time and the balloon shape. The reason for this behaviour by the flax yarn is its high rigidity. The next raw material which reveals its characteristic features during spinning is cotton. The pulse character of yarn tensions (which signals a balloon collapse) can be observed during the final phase of cotton yarn package formation. For all other raw materials, the yarn tension courses are stable. [42]. To reduce yarn breaks during ring spinning, the following aspects should be taken into consideration [40]: •

Since yarn breakage in ring spinning is related to slippage of fibres at the spinning triangle as a result of peaks occurring in the spinning tension fibre, the grip at the front drafting rollers should be increased by having a higher top roller pressure. The use of softer cots also enhances the grip at the front rollers. If the total pressure on the rollers cannot be increased, the grip at the front rollers’ nip can be improved by reducing the width of the cots. A reduction in friction between ring and traveller could reduce the peak tension during the rotation of the traveller. Measures should be taken to reduce the mass irregularity of yarn straight after carding. The width of the drafted ribbon at the front roller nip should be reduced.

• • •

Spinning tension in rotor spinning is particularly sensitive to the term (Tr2 Rr ). The importance of yarn tension lies in its significant effect on yarn quality. Previous studies showed that there is a linear relationship between the variation in yarn tension and the yarn mass variation (Uster CV%). The increase in rotor speed or the increase in rotor diameter results in a consistent reduction in yarn breaking elongation [30]. The yarn tension outside the rotor Fout during rotor spinning can be expressed by the relation proposed by the Rieter group [30]. 2

Effect of textile processing on fatigue Ty ω 2r Rr2 ⎞ ⎛ exp (μ 2 ) Fout = ⎜ FP + ⎝ 2 ⎟⎠

151

7.24

where Fp is the yarn tension at the peeling off point (usually Fp ≈ 0), Ty is the yarn linear density [tex], Tr is the rotor rotational speed in radians, Rr is the rotor radius [mm], and μ is the coefficient of yarn-rotor surface friction. Grosberg and Mansour [49] derived practically the same equation for Fout [cN] Fout = 0.72 (Ty ω 2r Rr2 )

2000 1800 1600 1400 1200 1000 800 600 400 200 0

18 .7 8 21 .2 2 23 .6 6 26 .1 28 .5 5 30 .9 9 33 .4 3 35 .8 7 38 .3 1 40 .7 5 43 .2 45 .6 4 48 .0 8 50 .5 2 52 .9 6 55 .4 57 .8 4

Numbering of class

As in case of ring spinning, the tension on the yarn during rotor spinning also exhibits peaks. The increase in rotor speed may result in a rising in tension peaks even sharper than the average tension does. This effect increases the chance of more end breakage during spinning. In real spinning conditions the temporal tension values are more important than the mean values [48]. Investigation into the dynamic forces which act on rotor-spun yarn [49] confirmed the occurrence of high-frequency (short-wave) yarn tension oscillations of significant amplitudes, which even exceed the mean value of about 30%. Maximum yarn tensions directly influence the number of yarn breakages [49]. An increase in yarn tension during spinning causes a decrease in the elongation of yarn over stretching to break. This influence is very substantial, although the yarn tension values constitute only 10–20% of the yarn’s breaking force [50]. The experimental measurement of yarn tension during rotor spinning was realized by Lotka and Jackowski [47]. Cotton yarns with linear densities of 25, 35, 45, and 55 tex formed with the use of BD 200S rotor spinning machine were investigated. Figure 7.10 shows the histogram of yarn tension for the 25 tex yarn.

Yam tension, cN

7.10 Yarn tension histogram of the 25-tex yarn [47].

152


The highest coefficient of variation of yarn tension was obtained for yarn with a linear density of 25 tex, whereas the lowest was obtained for the 55 tex yarn. The phenomenon of dynamic yarn tension fluctuations is more intensive when yarns with low linear density are formed, as these yarns are characterized by a higher mass irregularity than coarse yarn. This phenomenon is very dangerous as it may cause yarn breakage. The increase in bundle strength through subsequent stages of processing may provide a useful index of processing performance. A process exhibiting a different trend (e.g., a reduction in bundle strength) may indicate a potential problem in the drawing or the combing stage [30]. The mechanical processes during spinning have negative effects on the fibre surface structure. The surface damage of cotton fibres after spinning – cracks, cuticle layer removal, stratification and the existence of wrinkles and surface folds were observed in [32, 34] by using scanning electron microscopic tests. The raw cotton surface was smooth, without damage and with visible fibrils. The raw cotton SEM images and images after passing through the finisher draw frame, open-end spinning system are shown in Fig. 7.11. Wrinkles and folding correspond to weak places in the secondary wall. The visible cracks develop parallel to the angle of the macrofibril spiral, which proves that weaknesses in the cotton take place along macrofibril twist. Hearle and Hasnain [33] observed the serious effects of folds and cracks in cotton fibres for cotton in fatigue tests. Incidence of mechanical damage of various fibres during OE-rotor spinning due to contact with opening rollers has received considerable attention during the various research works [36–38]. The other source of damage that will lead to mechanical weakening of fibres and catastrophic breakage is thermal damage. This kind of damage can occur through fibre processing in the opening roller, rotor and take off nozzle. Thermal damage of polyester fibres occurred during the ring spinning through carding, drafting in passages and yarn balloon controller and traveller in ring frame [39]. The intensity of the damage will be related to the amount of the generated heat,

(a)

(b)

(c)

7.11 SEM images: a) raw cotton, b) crack due to rotor spinning c) folds due to rotor spinning [34].


153

(a)

(b)

(c)

7.12 Thermal damage of PET fibres: a) contact with opening roller, b) a longitudinal crack on the fibre c) fibre flattening [35].

which could be mainly a function of the quality of spin finish, rotational speed of opening roller and rotor and type of rotor and opening roller. The thermal damage of PET fibres (fineness of 1.5 den) during the spinning on BDA 10N Elitex rotor spinning frame is shown in Fig. 7.12. The test was carried out when the rotational speed of opening roller and rotor were 7000 rpm and 70 000 rpm respectively. A self-pumping rotor with a diameter of 43 mm was used [35]. During manufacture and subsequent use of yarns, they are often transported from one cylindrical package (bobbins, cones). In the case of high transport speed, the balloon is formed by spinning yarn between package and guide-eye, causing nonlinearity in yarn tension, which can even cause breakage of yarn. A comprehensive investigation of the dynamics of overend unwinding has been reported by Padfield [43]. Kothari and Leaf [44]

cN

20 19 18 17 16 15 14 13 12 11 10 0

Fatigue failure of textile fibres 20 19 18 17 16 15 14 13 12 11 10 0

cN

154

20 40 60 80 100 120 140 160 180 200 54C (a)

20 40 60 80 100 120 140 160 180 200 54C (b)

7.13 Yarn tension during unwinding: a) without control b) with control [46].

reported extensive numerical calculation based on Padfield’s analyses. Goswami [45] continued investigations of nonlinear dynamics of over-end unwinding, but there is still no exact mathematical model of this process and even in the case of a simplified mathematical model, its parameters are quite dependent on many conditions. Tension in over-end unwinding process depends on a lot of parameters, such as transport speed, modulus of elasticity of yarn, package winding density. It could be possible to use PID controller with variable gains for this problem, but for better performance the fundamental idea of an active disturbance rejection controller (ADRC) has been used.[46] Figure 7.13a shows yarn tension without control (passive dancer). It can be seen that tension increases while diameter of package becomes smaller, which conforms to the investigation of Goswami [45]. Figure 7.13b shows yarn tension with ADRC control applied. It can be seen, that variance of tension is about two times smaller than in case of uncontrolled tension and there is no increasing tension during operation [46]. The variation of tension during unwinding is dependent on the speed of unwinding and the unevenness of the package (Fig. 7.14). The increase of unwinding rate leads to the higher yarn tension. The state of the input package has a significant effect on the variation of yarn during unwinding. It can be observed that though the tension minima remains the same, the tension maxima increases drastically. From variation of tension during spinning operations or winding it is possible to identify load spectrum and then accumulation of fatigue.

7.3.2 Weaving Weaving is a process of formation of fabric with interlacement of two or more sets of yarns using a loom. The weaving process in most looms can

Effect of textile processing on fatigue 2

30

60

Pmax = 1.73 cN

Tension, cN

Pmax = 0.65 cN

1

155

90

120 30 Time, sec

60

(a)

90

120

150

(b)

7.14 Yarn tension variation during unwinding: a) smooth surface, b) uneven surface.

Heddles Warp beam

Weft insertion

Shed Tensioning

Takeup

Reed

7.15 Major parts of weaving loom [18].

be divided into separate operations known as: warp beam winding (or warp creel set-up), warp let-off, warp tensioning, shedding, weft insertion, beating, and take-up of the fabric. The locations of these operations are shown in the Fig. 7.15 Warp beam winding: before weaving it is necessary to wind the warp yarns onto the warp beams of the loom. The number of warp beams can vary. The warp beams hold the warp yarns for weaving. Warp tensioning: the warp yarns pass through a series of tensioning devices. These devices consist basically of yarn guides that are loaded by weights hanging onto the guides. The purpose is to take-up any slack in the warp yarns to ensure they are taut for the next stages of weaving. As the yarns pass through the tensioning stage they may slide backwards and forwards against the guides many times as weaving progresses. Shedding: the taut warp yarns pass through the eyes of heddles that control the vertical displacement of the yarns during weaving. Each layer of yarn

156


is raised and lowered in turn to form a ‘shed’ through which the weft yarns are inserted. Weft insertion: the suitable device (shuttle, etc.) is used to pull weft yarns through the sheds formed between the warp yarns. Despite the relatively high speed and efficiencies in loom with shuttle picking, productivity of these machines will continue to be limited as long as their fundamental construction involves the use of shuttle propulsion. It could be shown that the power required for picking is proportional to the cube of the loom speed. Air jet loom has a maximum speed and maximum weft insertion rate. Because of the very high quality of yarn required, the yarn must be of very high standard, otherwise the loom stoppages due to warp breaks and weft breaks will be high. The cover of the fabric in air jet will not be as in good as projectile and rapier looms. Beating: by using a large ‘reed’, the weft yarns are tightly packed into the fabric. During beating, the weft overcomes the frictional resistance between the warp and the weft, and moves along the direction of the warp ahead depending on the reed. When the difference F between the warp tension FS and the fabric tension FF near the cloth-fell is bigger than the frictional resistance, the weft slips against the warp. In consequence, the density of fabric increases. On the contrary, when F is less than the strongest static friction, the weft moves forwards with the warp, therefore it will not affect the increase of fabric density. Fabric take-up: after beating, a take-up mechanism pulls the woven fabric along the loom so that the weft insertion and beating stages can be repeated. It is in the take-up stage that the woven fabric is complete. In these stages it is mainly the warp yarns that are damaged. It was discovered that the fibres are abraded against each other and the loom machinery during weaving, and the resulting abrasion damage and removal of sizing agent causes a reduction in yarn strength of between 30 and 50% depending on the type of yarn. Some fibres are also broken during weaving, and this causes a small reduction to the yarn stiffness and contributes to the large loss in yarn strength [18]. While there is clear evidence of fibre breakage during weaving, it does not occur to a large extent. A thorough examination of the yarns using scanning electron microscopy revealed that a large majority of the fibres were not broken [18]. Based on initial modulus values, it is estimated that approximately 5–7% of the fibres in the warp were broken during weaving [18]. Another damage mechanism that degrades the strength but not the stiffness of the yarns must occur during weaving to account for the large loss in strength. It is believed the strength is reduced due to abrasion caused by fibres dubbing against other fibres and the loom machinery during weaving, such as the warp beams, tensioning device, heddles, reed


157

125

120

120

115

115

110

Tension (cN)

Tension (cN)

and rapier. It is known that removing the sizing agent will cause random fibre slacks in the yarn, which can reduce the tensile strength. It is also proposed that the abrasion creates ultra-small scratches on the brittle fibres (as glass), which would reduce the strength while not affecting the initial modulus [18]. Warp tension is one of the parameters to be carefully controlled on a loom, so as to avoid excessive peak tension values, as well as too low values of it, which would lead to unsatisfactory shedding. It is known that the highest effect on the tension in the warp (about 59%) is exerted by shedding, while that of beating-up is assessed at about 39% and that of the beam and take up roll at about 5% [57]. Loom operation studies showed that 88% of all warp breakage occurred at a point between the back harness and the reed, probably due to fatigue of the yarn as it travels from loom beam to the woven fabric [61]. A computer simulation of warp yarn tension during weaving is therefore aimed to help in creating objective setting recommendations for weaving machines. Masjajtis [52] derived a simple model of warp tension variation based on the shedding geometry and assumption of elastic warp yarn behaviour. A more complex model is described in [58]. The viscoelastic yarn model (Eyring type) used for simulation of the dynamic behaviour of the system weft, warp and fabric during weaving cycle was published by Nosek [55]. Warp tension variations during weaving of twill on the projectile loom (speed 235 rpm) was measured by shell gauge tension meter [51]. Results are shown on Fig. 7.16a. The measured warp tension variations are predicted by a simulation model with sufficient precision (see Fig. 7.16b) [58].

110 105 100 95

100 95 90

90 85 0

105

0.5

1 1.5 Time (sec) (a)

2

85 0

0.5

1 1.5 Time (sec) (b)

7.16 Warp tension variations: a) measured by shell gauge tension meter b) simulated [51].

2

158


It is known that during the weaving process, warp tension varies continuously as a function of the crank angle. It reaches the minimum value corresponding to the warp yarn at the average line. When the warp yarn bangs the weft yarn, warp tension increases dramatically to the maximum value. This position corresponds to the position of warp yarn at lower shed with the crank angle ϕ = 70º. The other maximum corresponds to the position of warp yarn at upper shed with the crank angle of ϕ = 430º. The value of the maximum at lower shed is bigger that the one at the upper shed because the warp tension is proportional to the height of the shed. The model of mechanical behaviour of fabric and warp yarn during loom stops was developed in work [60]. The failure of warp yarns on a loom is often caused by repeated cyclical elongation at small stresses well below the breaking point applied under static load. Generally, there is no prior indication of impending failure due to fatigue. The fatigue behaviour of sized warp yarn under cyclical elongation accompanied by abrasion action as measured on a Sulzer-Ruti Webtester device was studied by Anandjiwala and Goswami [56]. Three fatigue criteria: fatigue life, damage rate, and visual appearance were used. Fatigue life (number of cycles to failure) exhibited a skewed probability density trace that deviated significantly from normal distribution. A threeparameter Weibull distribution was found as suitable for approximating experimental results. This distribution was also confirmed for lifetime and abrasion data of various textile materials [59]. The rate of fatigue damage expressed in terms of loss in tensile property indicators (usually breaking strength) was useful in assessing the fatigue-sustaining capacity of yarn. The deterioration of sized cotton/polyester warp yarn tenacity till 50% of fatigue life time is usually small. At fatigue cycles between 50 and 75% of fatigue life the loss of tenacity is huge (about twice tenacity at 50% of fatigue life). Beyond 75% of fatigue life the sudden loss of tenacity appears [56]. The fatigue lifetime of yarn is a decreasing power function of sample length [17]. For the tensile fatigue of a 50/50 viscose/polyester warp yarn the lifetime (at 2.8% strain level) is decreased with about the 1.2 power of specimen length [62]. In contrast, the Pierce theory of ‘weakest link’ effect predicts a decrease with the 0.2 power of specimen length. Barella [63] found the power 0.54 for the case of worsted yarn lifetime under constant cyclic strain. The ‘weakest link’ effect is therefore much higher for lifetime than for strength. The same power was found for standard deviation of fatigue lifetime. This fact implies that the coefficient of variation of fatigue lifetime is independent on specimen length. The fatigue lifetime of of a 50/50 viscose/polyester warp yarn was well fitted by the three-parameter Weibull distribution for wide range of strain levels εL. The logarithm of the estimated characteristic life (from Weibull distribution parameters) Ne was found to decrease with the 1.52 power of


159

εL. and scale parameter bT of Weibull distribution (see. eqn. (7.3)) decreases with the 1.8 power of εL log ( N e ) = 8.44 ε1L,52

b = 0.353 ε1.8 L

The minimum lifetime Nm (third parameter of Weibull distribution) remains near zero for strain levels above 2% (above yield point). For strain levels below 2% the Nm increases rapidly with decreasing εL. The loading pattern imposed to the sized yarns during weaving corresponds the strain level εL = 3% and frequency 80 cycles per minute and number of loads 1700 [62]. The Nm of yarns under loading conditions observed at the loom greatly exceeded the number of load cycles imposed on the warp during loading. The tensile fatigue is therefore often not an important cause of yarn failure at the loom. The abrasion in between neighbouring yarns and with the machine elements is probably responsible for yarn failure. The experimental investigation of warp yarn failure mechanisms was studied on the Webtester (Reutlingen) [64]. Due to repeat abrasion during weaving, the yarn structure is loosened, some protruding fibres come out from the fibre body and fuzz balls appear on the yarn surface. The inter-fibre slippage is facilitated as well. Unsized warp fails mainly due to slippage mechanisms. Sized yarns are more consolidated by the adhesion of size material with fibres and failure mainly occurs due to fibre breaking.

7.3.3 Yarn fatigue Yarn failure occurs due to a combination of fibre slippage and fibre breakage mechanisms. The typical patterns of yarn breakage due to slippage and fibre break are shown on the Fig. 7.17. In dependence on the type of yarn, one of these mechanisms predominates: slippage in the case of ring and rotor yarns and breakage in the case

7.17 Two main mechanisms of yarn failure (slippage – top, fibre break – bottom).

160


of compact yarns. The static and dynamic tensile strength of ring and compact yarns is lower in comparison with rotor yarn. Static breaking load and elongation, although correlated with some dynamic characteristics of yarn, do not allow for a good prediction of the yarn behaviour under dynamic loading, especially for prediction of the yarn behaviour under pulsatory loading. Yarn resistance to dynamic loading depends mainly on twist parameter which can determined in a way the compactness and ‘closure’ of the yarn structure. The results obtained show that yarn resistance to dynamic loading differs for different types of yarn. Openend and air jet yarns due to their structure, among others the presence of wrapper fibres, form more ‘compact and closed’ structure, and are generally more resistant to the dynamic loading than more ‘open’ ring-spun yarns. The fatigue behaviour of yarn strongly depends on the features of load applied during pulsatory loading – the number of pulses till breaks decreases exponentially with increasing load frequency and decreases, although slower, with increasing initial load [19]. Different type of the yarn structure effects different relationships between the static and fatigue strength of yarn. Open-end yarns of the same tensile strength can be characterized by much higher fatigue strength than ring spun yarns. When comparing the effect of fatigue and coupled effect of fatigue and abrasion, one can see that the one phase yarn – ring and compact – can be characterized by much higher resistance to the abrasion than the multi-phase rotor yarns [19]. The reason is the quite different surface structure of yarns (see Fig. 7.18). The yarn surface structure has great influence on the yarn-to-yarn (YY) and yarn-to-metal (YM) frictions. It was found that the behaviour of frictions for YY is different than that of YM. In case of YY friction, rotor yarn shows maximum friction followed by air-jet, and ring-spun yarns; however, a reverse order is noticed for YM friction [65]. The fatigue behaviour of cotton and cotton-polyester blended yarn produced by rotor spinning has been investigated by Jeddi [20]. The mechanism of staple yarn fatigue is based on three stages:

7.18 Surface structure of 29.5 tex cotton yarn: a) ring, b) experimental, c) rotor.

Effect of textile processing on fatigue • • •

161

yarn decrimping fibre slippage arising from inter-fibre friction fibre elongation.

The yarn decrimp mechanism occurs just at the initial stage of fatigue. The region of the sharp variations of the yarn viscoelastic properties up to nearly 1000 cycles of loading is attributed to the ability of the fibre to reorganize inside the yarn structure. Finally, the region of the steady variations of the viscoelastic properties of yarn seems to be due to fibre elongation. The fatigue results reveal that the polyester fibre component has an important role in improving the fatigue resistance of the yarn to tensile cyclic loading. It has also been demonstrated that greater variations of yarn viscoelasticity are obtained with higher strain percent, so that at 1% strain the 100% cotton yarn reveals immediate failure [20]. Barrela and Manich [22] investigated the resistance to abrasion and resistance to repeated extension (both expressed as number of cycles to break) for PES/cotton yarn prepared by ring, rotor and Dref III technology for two linear densities and three twist level. The yarn against yarn abrasion was performed using a Abrafil device (mechanical load 150cN and speed of abrasion 100 cycles/minute). Resistance to repeated extension was measured using a Comptis device (pretension 1.5 cN/tex, cycle amplitude 3.3 % frequency 200 cycles/minute). The friction spun yarns (Dref III) show a smaller resistance to abrasion and to repeated extension than the ring and rotor yarns. The ring-spun yarns show the best repeated extension resistance, followed by rotor-spun yarn. Dref III yarns are the weakest. The best abrasion resistance is for rotor yarns, followed by ring yarns. The Dref III yarns are again the weakest. The blend composition has serious influence on resistances. When the proportion of polyester is decreased form 70% to 50% the extension resistance decreased and abrasion resistance increased for rotor- and friction-spun yarns. This trend is less defined for ring-spun yarns. Both repeated tension and abrasion resistances increase with increase of twist level. The abrasion resistance is higher for coarser yarns for all yarn types. The dependence of abrasion resistance AT on repeated tension resistance RT is shown in Fig. 7.19. The straight lines were computed by linear least squares regression. The estimated regression lines have the form Ring: AT = 721.79 + 2.73 RT Rotor: AT = −311.47 + 23.51 RT Dreft III: AT = −192.7 + 15.70 RT The methodology for creation S-N fatigue curves for worsted wool yarns with linear density of 16 tex and 25 tex produced with the use of a FIOMAX 2000 ring spinning machine was describer in the work [66]. These curves were approximated by the semi-logarithmic model in the form

162

Fatigue failure of textile fibres 20000 Abrasion resistance AT [cycles]

Rotor ∗

15000

Dref III 10000

5000

∗

∗

Ring

0

–5000 0

200 400 600 800 1000 Repeated tension resistance RT [cycles]

1200

7.19 Dependence of abrasion resistance AT on repeated tension resistance RT.

yarn 16 tex: S/ST = 106.1 − 4.58 ln(N) yarn 16 tex: S/ST = 108.2 − 4.64 ln(N) The yarn linear density does not have significant influence on the course of the S-N curve. The influence of the fibre parameters (fineness Tf [tex], staple length Ls [mm], density ρf [mg/mm3]) and yarn parameters (fineness Ty [tex], twist Ky [turns/m], density ρy [mg/mm3]) on the fatigue life N under specific conditions is investigated in [67]. Four dimensionless complexes were obtained as a result of scaling theory and dimensionality analysis. After rearrangements the final prediction relation results N =c

Ty ρy K y Ls = c nf μ f K y Ls Tf ρf

where c is proportionality constant, nf is number of fibres in the yarn crosssection and μf is yarn packing density. Yarn fatigue is dependent on the modification of textile materials during processing and environmental conditions. Subramaniam et al. [53] have reported that slack mercerization treatment improves the fatigue life of ring- and rotor-spun yarn. Selvakumar [54] has reported that fatigue life of silk and treated silk yarn in water is lower than that in air.

7.3.4 Fabric fatigue The fabric fatigue phenomena were studied by Busse et al. [68]. A new device allowing testing dimensional behaviour of knitted fabric under a


163

large number of cyclic elongations was developed by Ben Abdessalem and co-workers [21]. This dynamic tester simulates knitted fabric deformation during garment wearing. This tester was associated with an imageprocessing device in order to measure fabric and loop deformation. The fatigue test applied to a plain knitted fabric made of cotton involved a variation of fabric dimensions and a permanent deformation that persists after relaxation. This permanent deformation depends largely on the number of repetitive cycles. The analysis of the loop geometry with the imageprocessing device permitted concluded that deformation involved by fatigue test corresponds to a displacement and a lengthening of the yarn composing the loop. Mechanism of fibre slippage in the yarn and fibre viscoelasticity were used in order to understand hysteresis phenomena of spun yarns and to predict plain knitted fabric dimensional behaviour from yarn and fibre properties [21].

7.4

Prediction of fatigue during wearing

It is a common experience that fabrics deteriorate mainly due to repeated deformation and abrasion in long-term wearing. This is considered to be a kind of fatigue caused mainly by the changes in fabric handle and mechanical properties during long-term use. The popular laboratory method of examining fatigue behaviour is periodic loading at some stress (or) strain level below in a tensile test. This is a logical form of experimentation since many textile items are subjected to this type of loading history during everyday use (e.g., tyre cords, S-belts and clothing). In the majority of the papers on fabric fatigue the fabric geometry, weight, breaking strength, elongation and abrasion resistance are investigated [69, 71] but few paid attention to wearability. This wearability is, however, very important in practical use of clothing, because it is directly connected with durability of textiles [70]. For evaluating fabric handle and precise measurement of fabric mechanical properties after simulated wearing the KES-F fabric-testing system developed by Kawabata can be used [72]. KES enables the measurement of mechanical properties under the deformation range which occurs in actual wear, and prediction of fabric handle by method based on fabric mechanical data [73]. The purpose of the study published by Hattori and co-workers [72] was to investigate the fatigue of clothing fabrics with respect to their mechanical properties and handle during wear by using the KES-F system and the simulation tester KES-FS. The 14 typical woven fabrics for men’s suits were subjected to repeat shear deformation (104, 105 and 106 cycles) superposed on a constant extension load to simulate the deformation in actual wear. Fabric specimens were measured for mechanical properties before, during, and after the above deformation and recorded as functions of deformation

164


cycles. Fabric hand at each stage was expressed by total hand value (THV) computed from measured mechanical data. The THV showed little change even after 106 cycles of wear simulation. The lower the initial fabric quality, the greater was the decrease in THV. In order to examine the effects of fabric tension on repeated shear fatigue, tests were carried out under three different levels of constant extensional load. Other conditions were kept at the standard condition. In tensile properties, the higher the constant extensional load, the more rapid the decrease in tensile energy (WT) and resiliency (RT) beyond 104 cycles [73]. In shear properties, effects of constant extensional load appear at the initial stage of the repeated loading. The increase in shear hysteresis (2HG) above 104 cycles becomes more pronounced under higher tensile load. As a simulation of clothing use under humid environments, shear fatigue tests were carried out under wet condition in which water content in fabric was 65 ± 5%. It was found that the fatigue behaviour of all samples (wool and polyester) under wet conditions was different from that under dry conditions [72]. In order to observe the recovery of fabrics from fatigue, the specimen was subjected to 105 cycles of shearing and mechanical properties were measured. Another measurement was made after keeping the specimen at rest for 18 hrs and this procedure was repeated 10 times. Recovery from fatigue expressed in the ratio of shear hysteresis 2HG/2HG0 was much better in higher quality worsted fabric than in low quality fabric. The higher quality fabric showed less change during the repeated testing, and the deformation and recovery is within a narrow range [72]. The fatigue behaviours of knitted fabrics were different from those of woven fabrics. Knitted fabrics fatigue is more severe than woven fabrics under the same testing conditions. The decrease of tensile resilience RT is pronounced at 103 cycles. In knitted fabrics, shear hysteresis 2HG increases linearly with increasing log (shearing cycles). [72]. The fatigue durability depends on fabric structure as well as on fibre and yarn properties. High-quality fabrics were superior to low-quality fabrics in the retention of fabric mechanical properties and hand, and also in the level of residual or permanent deformation after fatigue test. Results of repeated shear experiments are in good agreement with those of wear test on men’s slacks. Fatigue behaviours of clothing fabrics during actual long-term wear can be reproduced in a short time, using this fatigue testing system when a small amount of carborundum powder is sprayed on the surface of suitably wet specimens [72]. The influence of abrasive characteristics on wear mechanisms is still unknown. Some researchers [74] report that micro-abrasive particles create fine hairiness on the surface of the fabric, whereas macro grits induce longer pile. Wearing a structure with smaller abrasive particles seems to promote


165

the wear of emergent filaments that have larger degrees of freedom on their surfaces [75].

7.5

Conclusions

The fatigue of textile structures can be evaluated in the same ways as fatigue of other structures or by selection of specific characteristics as total hand value. The techniques based on the prediction of fatigue life use the threeor two- parameter Weibull distribution. Methods based on characterization of fatigue due to cyclic deformation are to some extent oriented to the investigation of cumulative damage. The evaluation of results is based on the experimental setup and purpose of investigation. The complex approach based on the identification of loading spectra is often replaced by simpler evaluation of breakage rate or number of breaks. The dynamic loading during spinning and weaving can be simulated by use of mechanistic or structural models. The fatigue of textile structures is often correlated with structural and geometrical parameters.

7.6

Acknowledgements

This work was supported by the research project No. 1M4674788501 (Textile Centre) of the Czech Ministry of Education.

7.7

References

1. Miner M.A.: Cumulative damage in fatigue, J. Appl. Mech. 12, A159 (1945). 2. Schutz W.: The prediction of fatigue life in the crack initiation and propagation stages, Eng. Fracture Mechanics, 11, 405–421 (1979). 3. Hainbach E.: Modified linear damage accumulation hypothesis accounting for decreasing fatigue strength during increasing fatigue damage (in German). LBF Darmstadt TM No 50 (1970). 4. Bishop N.W.N., Sherratt F.: Finite Element Based Fatigue Calculation, NAFEMS (2000). 5. Caprino G., Giorleo G.: Fatigue lifetime of glass fabric/epoxy composites, Composites A30, 299–304 (1999). 6. Atus E., Herszage A.: A two-dimensional micromechanical fatigue model, Mechanics of Materials 20, 209–223 (1995). 7. Mao H., Mahandevan S.: Fatigue damage modelling of composite materials, Composite Structures 58, 405–410 (2002). 8. Castillo E., Fernandez-Cantelli A., Hadi A. S.: On fitting a fatigue model to data, Int. J. of Fatigue 21, 97–106 (1999). 9. Sehitoglu H.: Advances in Fatigue Lifetime Predictive Techniques, ASTM 1122, 47–76, (1992). 10. Neu R.W., Sehitoglu H.: Metallurgical Transactions A20, 1769–1783 (1989). 11. Muralidharan U., Manson J.: Engineering Materials and Technology, 110, 55–58 (1988).

166


12. Manson S.S.: Fatigue a complex subject, Exp. Mechanics 5, 193–226 (1965). 13. Ramberg W., Osgood W.R.: Description of stress strain curves by three parameters, Technical Note No. 902, National Advisory Committee for Aeronautics, Washington DC, 1943. 14. Xue L.: A unified expression for cycle fatigue, Int. J. Fatigue 30, 1691–1698 (2008). 15. Degrieck J., Van Paepegem W.: Fatigue damage modelling of fibre reinforced composite materials, Applied Mechanics Rev. 54, 279–300 (2001). 16. Van Paepegem W., Degrieck J.: Modelling strategies for fatigue damage behaviour of fibre reinforced composites, European J. of Mechanical and Environmental Engineering, 46, 217–227 (2001). 17. Prevorsek D., Lyons L.L., Whitwell J.C.: Statistical treatment of data and extreme-value theory in relation to fatigue in textiles, Textile Res. J. 33(12), 963–973 (1963). 18. Rudov-Clarka S., Mouritza A.P., Leea L., Bannister M.K.: Fibre damage in the manufacture of advanced three-dimensional woven composites, Composites: Part A 34, 963–970 (2003). 19. Cybulska M.: Fatigue and tensile behaviour of staple yarns, The Fiber Society Spring Conference, St. Gallen, May 2005. 20. Jeddi A. A.A. et al.: A Comparative Study of the Tensile Fatigue Behaviour of Cotton–Polyester Blended Yarn by Cyclic Loading, Journal of Elastomers and Plastics, 39, 165–179 (2007). 21. Ben Abdessalem S., Elmarzougui S., Faouzi Sakli F.: Dynamic Fatigue of Plain Knitted Fabric, J. Textile and Apparel 5,No 2, 1–10 Summer 2006. 22. Barrela A., Manich A.M.: Friction spun yarn versus ring and rotor spun yarns, Text. Res. J., 59,767–769 (1989). 23. Krifa M.: Fiber Length Distribution in Cotton Processing: Dominant Features and Interaction Effects, Text. Res. J. 76, 426–435 (2006). 24. Anokhina Yu. E., Zinoveva V. A., Shablygin M. V.: Fibre damage in textile processing, Fibre Chemistry 33, 59–62 (2001). 25. Subramaniam V, et al.: Fatigue of Cotton and Viscose Stable Yarns by Biaxial Rotation over a Pin, Textile Res J, 60, 301–303(1990). 26. Miller B, Friedman H L, Turner R: Design and Rise of a Cyclic Abrader for Filaments and Yarns – A study of Polyester Filament Yarn, Text. Res. J.: 53, 733–740. (1983). 27. Dunlop J. I., Barker A.: Fatigue Testing of Fibre Under Compressive Flexing, Text. Res. J. 43, 739–382 (1973). 28. Goswami B C, Duckett K E and T L Vigo, Torsional Fatigue and the Initiation Mechanism of Pilling, Textile Res. J. 50, 481–485 (1980). 29. Cooke, W. D., Pilling Attrition and Fatigue, Text. Res. J. 55, 409 (1985). 30. El Moghazy Y.: Fiber-To-Yarn Engineering, book in preparation Auborn University, 2008. 31. Schneider, T., Rettig, D. and Mussig, J., Single Fiber based determination of short fibre content, Proc. ‘Beltwide Cotton Conferences’. San Diego, CA, USA January (1994), pp. 1511–1513. 32. Sarna E., Wlochowicz A.: The influence of the spinning systems on the surface damane of cotton fibres, Textile Res. J. 77, 810–811 (2007). 33. Hearle J W S., Hasnain N.: Fatigue of cotton fibres and yarns, Proc Annual Conference of the Textile Inst, New Delhi (1979), pp 163–165.


167

34. Sarna E., Wlochowicz A.: Damages of cotton surface during spinning process, Microsc. Microanal. 9 (Suppl. 3), 366–367 (2003). 35. Gharenghai A., A.: Thermal Damage of PET Fibres in OE Rotor Spinning, Proc. Int, Conf. ‘Textile Science’, Liberec, Czech Republic, May 2008, sec.4, pp 1–6. 36. Gharehaghaji A. A., Johnson, N.A.G.: Wool fibre microdamage caused by opening processes, Part 1: Sliver Opening, J. Tex. Inst., 84, No. 3, (1993). 37. Salhotra K. R., Chattopadahyay R.: Loss in fibre tenacity during separation in rotor spinning, Text. Res., J. 54, 194–197, (1984). 38. Salhotra K. R., Chattopadahyay R.: Incidence and mechanism of fibre breakage in rotor spinning, Text. Res. J., 52, 317–320, 1982. 39. Klein, W., Short staple spinning Series, Vol.6, Man-Made fibres and their processing, The Textile Institute, 1994. 40. DeBarr, A. E., Catling, H.: The Principles and Theory of Ring Spinning. p. 149, The Textile Institute, Butterworths, 1965. 41. Ghosh A., et al.: The mechanism of end breakage in ring spinning, AUTEX Research Journal, 4, 19–23, March 2004. 42. Król B., Przybył K.: Influence of yarn kind on the dynamic of the twisting and winding system of the ring spinning machine, AUTEX Research Journal, 2, 144–151, September 2002. 43. Padfield D. G.: The motion and tension of an unwinding thread, Proc. Roy. Soc. London. A245, 382–407 (1958). 44. Kothari V. K., Leaf G. A. V: The unwinding of yarns from packages. Parts I, II and III, J. Text. Inst. 5, 89–172 (1979). 45. Goswami B. C.: Nonlinear dynamics of high-speed transport for staple yarns, National Textile Center Annual Report, November, 2002. 46. Umirov U. R. et al.: Tension control for high quality take up of yarn transport machine, Res. Paper, Yeungnam University, Gyeongsan, Korea 2003. 47. Lotka M., Jackowski T.: Yarn tension in the process of rotor spinning, AUTEX Research Journal, 3, 23–27, March 2003. 48. Margulis W. E.: The form of balloon and yarn tension in centrifugal and pneumomechanical Spinning, Tekstilnaja.Promyszlenost., No 6, (1969). 49. Grosberg P., Monsour S.A.: High-speed open-end rotor spinning, J. Text. Inst., 66, 389–396 (1975). 50. Stadler H.: Der Einfluβ der Rotordrehzahl auf die Garn Herstellung, Textil Praxi International, 7, 175–184, (1975). 51. Mirjalili S.A., Bandura M.P.U.: Computer simulation of warp tension, Proc. Int, Conf. Textile Science, Liberec, Czech Republic, May 2008, sec.4, pp 1–9. 52. Masjajtis J.: Dynamic disturbances of warp tension in weaving and behaviour of the left off motion on the loom, Monograph, Gent 1998. 53. Subramaniam V, et al.: Fatigue of cotton and viscose staple yarns by biaxial rotation over a pin, Text. Res. J, 60, 301–303(1990). 54. Selvakumar N: A study on the effects the chemical and physical treatments on the properties of silk yarn, PhD Thesis, Anna University 1995. 55. Nosek S.: Theory of Weaving I–III., DT Pardubice 1988–1999 (book in Czech). 56. Anandjiwala R. D., Goswami B. C.: Tensile Fatigue Behaviour of Stable Yarn, Text. Res. J. 63, 392–403 (1993). 57. Bykadorov R.: Control of the quality of the woven fabric in the loom, Legkaja Industrija, Moscow 1984.

168


58. Mirjalili S,A,: Computer simulation of warp tension on a weaving machine, J. Text. Engn. 49, 7–13 (2003). 59. Barella A: The Application of the First and Third Asymptotic Distributions to Abrasion Fatigue and Repeated extension of yarns and to Bending of Fabrics, J Text. Inst., 36, T665–674. (1965). 60. Vangheluwe, L., and Kiekens, P., Modeling Mechanical Behavior of Fabric and Warp Yarn During Loom Stops, Text. Res. J. 66, 722–726 (1996). 61. Milovidov N.N.: Cyclic tensile strength of yarn and its connection with end breakage rate in weaving, Tecknol. Text. Prom. 2, 77–82 (1964). 62. Picciotto R., Hersh S. P.: The fatigue behavior of a warp yarn and its influence on weaving performance Text. Res. J. 42, 512–522 (1972). 63. Barella A.: Contribution to the study of the behavior of worsted yarns subjected to repeat stretching, Invest. Inform. Textile 4, 3–23 (1961). 64. Behera B.K., Mishra R.: Weavability of cold sized worsted warp yarns, J. Text. Engn. 52, 179–187 (2006). 65. Ghosh A. et al.: A Study on Dynamic Friction of Different Spun Yarns, J. Appl. Polym. Sci., 108, 3233–3238 (2008). 66. Drobina R.: Fatigue curves elaborated for selected worsted wool yarns, FIBRES & TEXTILES in Eastern Europe 15, 64–65 (2007). 67. Shustov Yu. S., Shustov E. Yu.: Prediction of the wear resistance of yarns made from chemical fibres, Fibre Chemistry, 34, 452–455 (2002). 68. Busse W. F., et al.: Fatigue of Fabrics, J. Appl. Physics 13, 715–724 (1942). 69. Lee J.S., Barbour H.F., Finker M.D.: Wearing comfort of selected New Mexico wool I, Text. Res. J. 31, 540–550 (1961). 70. Lee J.S., Finker M.D.: Wearing comfort of selected New Mexico wool II, Text. Res. J. 33, 855–863 (1963). 71. Dowlen R.P.: Durability of serge in trousers, Text. Res. J. 35, 1035–1041 (1965). 72. Hattori Y., Niwa M, Kawabata S.: Changes in mechanical properties of fabric after repeated deformation, J. Text. Mach. Soc. Japan, 30, 1–12 (1984). 73. Kawabata S.: The Standardization and Analysis of Fabric Hand, 2nd. edn, The Textile Machinery Society of Japan, Osaka 1982. 74. Seidel L. E.: Face Finishing for Knit Fabric, Textile Ind. 99–100 (1979). 75. Fontaine S., Bueno M., Renner M.: Tribologic phenomena during wear of fibrous structures, Text. Res. J. 73, 855–863 (2003).

8 Environmental aspects of fatigue K SLATER, University of Guelph, Canada

Abstract: This chapter deals with the environmental aspects of fatigue failure in textiles. It examines first how the environment in which a fibre is located affects its life, as a result of degradation, under the influence of moisture, radiation, heat, chemical or microbiological agents, then focuses on how fatigue fracture modifies the surroundings. It then deals with means of overcoming environmental effects and ends with a prediction of future trends and possibilities. Key words: environment, degradation, physical and chemical sources, protection and prevention, future needs, prospects and challenges.

8.1

Introduction: aspects of importance

In any discussion of the environmental effects of fatigue failure, we need to examine three possible interpretations of the term. The first of these is a consideration of how the environment influences failure potential, the second deals with how failure affects the environment, and the third is a consideration of how the effects of fatigue failure on the environment can be reduced or minimised.

8.1.1 Definitions In order to begin the examination, there are several definitions of terms that should be established at the outset. These include the environment, pollution, and degradation. In this chapter, I am assuming that the term ‘the environment’ is intended to be a comprehensive one that includes the entire planet, together with its mineral and biological inhabitants, and encompassing all its accompanying atmospheric shell to the location at which gravitational force no longer binds any substances to the earth’s neighbourhood. There is no point, for example, in assuming that the release of gases into the stratosphere is going to exert no effect on the planet, as it is precisely in this region that much of the origin of greenhouse effect responsible for global warming takes place. 169

170


In addition, the social environment, as well as merely the physical one, should be a part of the definition, since any effects that influence the mental state of human beings may well have some reflection in behaviour that can indirectly harm the planet or its inhabitants. The term ‘pollution’ is defined as including any material or intangible condition that is emitted from a process or event and that has any presence which is undesirable to human, animal or plant life in the vicinity. In this definition, the presence of noise or visual pollution is included, as also is the ‘pollution’ of offensive behaviour or language that can lead to reprisals or vengeful tendencies potentially harmful to the environment. This extension is not normally included in our commonly held impressions of what is meant by pollution, but these factors can become important in special circumstances where wilful damage results from their presence. Degradation is defined as any change in mechanical or structural properties that leaves the fibre with a lower resistance to destruction, by whatever applied sources, than it had before the change took place. The process may be cataclysmic or may take place over an extended period of time, but is usually irreversible, except possibly by a major recycling operation.

8.2

Effects of environment on fatigue fracture

The effects of the environment on fibre properties in general are normally classed as a specific case of degradation, and the entire subject of textile degradation has been discussed extensively elsewhere (Slater, 1991). Apart from this source, there is very little information in the literature relating to textile fibre assemblies as such, most of the work published relating to composites in which textile fibres are an important and integral part.

8.2.1 Degradation steps Fracture takes place for many different reasons, most of which are dealt with in other chapters of this book, but there are three distinct stages to the process of degradation responsible for fracture, depending on the energy applied to bring about molecular changes. At low levels of energy application, changes in end-group structure, in which portions of the molecule residing on a side-chain are modified, can take place. Fracture will only result after many such changes have occurred, if at all, because the integrity of the fibre molecule is reduced very slowly, if at all. An increase in the applied energy can bring about destruction of the side-chain, separating it from the long-chain main molecule giving the fibre its major source of integrity. This type of degradation will bring about fracture more rapidly, and will probably always result in destruction eventually. Finally, at high energy levels, main-chain scission can occur, with very rapid

Environmental aspects of fatigue

171

degradation causing fracture very rapidly. The degradative energy applied can take many forms, including mechanical, thermal, chemical, microbiological or radiative ones, all of which are affected significantly by environmental conditions existing when they attack the fibre. However, the immediate cause of fracture usually consists of destructive changes brought about by the application of mechanical force. The force may be applied in any one of a tensile, compressive, flexural, torsional, or rotational mode, or in a combination of two or more modes. Destruction may also be brought about by the action of frictional contact on fibre surfaces, wearing away the structure gradually until it is unable to support the load applied to it. All of these fatigue processes slowly change the internal molecular structure of the fibres, as described above, to the failure point. The rate at which the degradative actions to which the fibre is exposed produce fracture, is dependent on many factors, such as the actual molecular structure itself, the treatment received by the fibres during manufacture or use, and any treatments given to prolong the life of the structure. However, the basic criterion of durability in understanding the exposure of the fibres to environmental factors; this can be crucial in establishing whether or not fracture actually occurs and at what rate it is likely to bring about failure. The most obvious example of these factors is the all-pervasive presence of natural elements to which the fibres are exposed, such as the encompassing atmospheric conditions.

8.2.2 Degradation sources Air and moisture are the two most common elements present in the atmosphere. In the ideal case, these would be pure components, with no contaminants, though this situation virtually never happens. Impurities, in the form of such added sources of degradation as corrosive gases or chemicals, will be discussed shortly, but oxygen and moisture (in vapour or liquid form) can be considered independently of other constituents.

8.2.3 Effects of moisture The oxygen content of the air is not usually regarded as having much influence on degradation, though it can sometimes act as a catalyst in reactions where its presence is needed to bring about an oxidative process. Moisture, though, is a different matter. As is well known, standard atmospheres must be used in textile testing, because test results can vary enormously if moisture content changes. The moisture affecting the onset or rate of fatigue fracture may also be atmospheric, may be contained within the fibre molecule, or may be present as liquid water in which the fibre is immersed. Its

172


presence can increase hydrogen bonding to strengthen a fibre (and hence improve its resistance to fracture), provide a reduction in frictional contact by a lubrication effect, or bring about fibre swelling to force apart molecular chains to the point where their reduced ability to form cross-linking bonds can lead to end-group or side-chain modification to increase the risk of rupture. Suzuki (1982) examines the effect on frictional properties of various fabrics in contact with polypropylene, to the point of disintegration, as a means of predicting wear resistance. Wang and colleagues (1990) examine transient moisture effects in fibres and composites, noting that these are much less well-studied than steady-state ones. Their study particularly focuses on the creep of Kevlar (as fibres and in composites) in relative humidity conditions cycling between 5% and 95%, finding that results resemble those obtained with natural fibres, including accelerated creep that can lead to degradation. Guillemet (1995) deals with static fatigue of glass fibres in a wet environment, suggesting that hydrolysis produces crack growth during tensile stress. In work on composites, glass and sisal fibres are both studied in this context. Ellyin and Rohrbacher (2000; 2003) examine the effect of an aqueous environment and temperature on glass-epoxy structures immersed in distilled water, noting that immersion at higher temperatures leads to embrittlement of the glass and blistering of the matrix. Two groups of workers investigate sisal-reinforced composites. Jeon, Seo and Lim (2005) find that fracture toughness decreases as water absorption rate increases, while Kim and Seo (2006) note that mechanical properties (including fatigue) worsen if cycling from wet to dry and back is carried out frequently. Moisture can additionally affect other types of change, to be dealt with later in the chapter. Among the most familiar ones are the enhanced rate of microbiological growth and the increased likelihood of static electricity generation that can lead to a fire risk (that is, the extreme end-result of thermal degradation) under the right conditions.

8.2.4 Effects of radiation A common constituent of natural exposure is the presence of ultraviolet radiation from the sun, and this also has a notable effect on degradation. For the requisite changes in molecular structure to occur, the radiation must possess enough energy, usually imparted by exciting some region of the fibre molecule in a resonance reaction. The effects of ultraviolet radiation exposure have been familiar for some time, and many early references to the topic can be found. Solar energy coincides with many important bonddissociation energy levels, as discussed by Wall, Frank and Stevens (1971) in work specifically directed to nylon and polyester. Holt and Milligan


173

(1976) explain the photoyellowing of wool by mechanisms that can lead potentially to fatigue failure, while Shatkay and Wetherall (1977) interpret the kinetics of degradation by means of free radical reactions, which are also used by Hon (1977) to account for changes in cellulosic fibres.

8.2.5 Effects of combined degradative factors Weathering, a combination of air, moisture and ultraviolet radiation (possibly in conjunction with gaseous impurities), can be devastating in its destructive tendencies on many fibres (Barnett and Slater, 1991). It is only necessary to see the state reached by national flags flying outdoors in windy (i.e., mechanically stressed) conditions, for as short a time as one year and being reduced to tatters, to realise the powerful effect of this combination in causing fatigue fracture. Carlsson, Garton and Wiles (1977) explain fracture in polypropylene by postulating backbone fission followed by chain oxidation, possibly by a free radical route, and suggest that manufacturing conditions might have an important influence on the degree of weathering degradation. Brown et al. (1983) study surface abrasion damage during weaving of Kevlar and feel that subsequent photodegradation may be dependent on crystallinity. Other (artificial) light sources may also bring about degradation, but they are usually less harmful than solar exposure. However, exposure to other radiation wavelengths, such as X-ray or microwave wavelengths, may cause changes in structure that could eventually lead to loss of integrity over a long time period.

8.2.6 Effects of heat The influence of heat, a different type of radiation, should also not be ignored. Thermal degradation (Slater, 1976) can often be a major factor in the onset of fibre damage at a rapid rate. Virtually all fibres suffer relatively minor changes, such as discoloration, cockling or shrinkage, at fairly low temperatures. Many of them, notably the cellulosic ones, are affected even more significantly, igniting readily at temperatures that make their use dangerous in household conditions where open flames may be encountered. Thermal degradation itself, without ignition, is a cause of changes in mechanical behaviour. Nkiwane (2001) finds that nylon tyre cords fail more rapidly by creep at elevated temperatures. In composites, environmental effects on fibre degradation are studied for silicon carbide fibres used as reinforcement materials by Shimojo and Bowen (1998) and by Legrand et al. (2002) at temperatures of 450 ºC, indicating that air exposure has a significant effect, increasing crack potential, in comparison with vacuum conditions. Glass fibre reinforcement at a range 20 to 160 ºC is studied by Hale, Gibson and Speake (2002), who find that the failure envelopes for a phenolic

174


system are relatively unaffected by temperature, but weakening of an epoxy system occurs. Bittner-Rohrhofer, Humer and Weber (2002) investigate changes taking place at cryogenic temperatures in superconducting magnet coil glass-reinforced epoxy insulators and note that results indicate the strong influence of direction of winding and of radiation exposure. Kawai and colleagues (2001) use carbon-fibre reinforced epoxy composites, finding that fatigue strength is reduced significantly as temperature rises.

8.2.7 Effects of gas exposure Atmospheric gases, mainly the oxides of nitrogen, sulphur, and carbon, are often present in conjunction with humidity, ultraviolet radiation and solar heat, and their contribution to any change in the rate of degradation cannot be ignored. Little work has appeared on this topic, but it can be assumed that their specific effects will be similar to (but less rapid than) those produced when acids come into contact with the fibres to increase the rate at which the fatigue point is approached. In general, acids tend to degrade cellulosic fibres very rapidly, but have little or no effect on the structural integrity of synthetic ones (Bao and Slater, 1990).

8.2.8 Microbiological attack One further source of degradation that may be present under conditions of normal use is the microbiological one. Bacteria, moulds and similar agents are omnipresent and can drastically affect some fibres, notably again the natural ones. These will disintegrate fairly rapidly if left in contact with the agent (if they are laying on the ground or immersed in water, for instance), to the point where they are of no use and need to be discarded. Even the synthetic fibres, which may not strictly be degraded (that is, may not lose any significant structural integrity) are often discoloured and so become unfit for further use. In clothing, the presence of human bodily fluids (perspiration or urine, for example) can combine microbiological with chemical attack to weaken a fabric structure.

8.2.9 Chemical attack Contact with reagents in specialised uses can also occur. In household situations, salts, bleaching agents, detergents and organic chemicals, notably solvents, can bring about degradation when they come into contact with fibres during routine maintenance procedures. In industrial applications, the same substances, plus strong acids or bases, can have the same effect. Again, the exact result of such contact will depend entirely on the reagent/ fibre combination under consideration, but natural fibres are generally


175

more susceptible to inorganic chemical attack, while some synthetic fibres are degraded rapidly by solvents; any elementary textile chemistry text book can identify the specific results of contact. One interesting paper by Tyagi and colleagues (2004) considers the effect of mercerisation on cotton yarns. They find that the process improves packing by swelling and shrinking the fibres, while increasing tenacity, flexural rigidity and resistance to failure during weaving. Apart from this, recent work is sparse, the only notable paper being that of Song, Takeda and Kawamoto (2000), who study the degradation of silicon carbide fibres under hot loading in the presence of an oxidising agent and find that the presence of the latter produces major reduction in degradation resistance. Most of the evidence of interest dates from one or two decades ago. Carswell and Roberts (1980) compare the behaviour of glass-reinforced polyester resin under tensile fatigue loading in air and in acidic environments, noting significant differences in time-to-fracture, fracture surfaces, fibre pullout and delamination. Fujii et al. (1994) extend the work to show that no noticeable diffusion of acid takes place, postulating a relationship between fatigue damage and stress corrosion. Ashbee and Ashbee (1985) propose that fibre fracture in Kevlar/epoxy composites may be influenced by the presence of sodium sulphate left over from the production process. Yu, Ishii and Tohgo (1994) also note that sodium chloride may be influential in reducing fatigue strength of silicon carbide fibres used as reinforcement in an aluminium matrix.

8.2.10 Structural effects The structure of the fibre assembly is also capable of influencing its fatigue failure potential. Dooraki, Nemes and Bolduc (2006) discuss the factors influencing the strength of yarns, then derive a model to account for the changing failure rates under different conditions. Ghosh, Ishtiaque and Rengasamy (2005a) analyse spun yarn failure in two papers, the first relating failure to structure and testing parameters and the second (2005b) incorporating a translation of strength in fibre bundles to spun yarns. Ghosh (2006) subsequently carries out a theoretical investigation to explain yarn failure mechanism, based on fibre bundle tenacity. Cho and Jeong (2005) discuss the reasons underlying the magnitude of bundle strength, then develop a correlation between this measure and yarn strength from a statistical point of view. Pan, Hua and Qiu (2001) take the work a step further, establishing a correlation between yarn strength and the strength of individual fibres, taking into account such parameters as twist, fibre length, and packing density. Cybulska and Goswami (2001) and Zeidman and Sawhney (2002) add to this a consideration of the way in which fibre length distribution influences yarn strength.

176


8.2.11 Effects of manufacturing conditions The actual processing of fibres to produce a yarn may also have some influence on failure rates. Naik and Singh (2001) present an analytical means of predicting shear strength from twist angle and twist at different yarn radii. Haixia, Yuan and Shanyuan (2006) carry out a similar analysis to show that strength properties of rotor spun yarns are also influenced by twist. Saitta et al. (1999) point out that the break in a yarn containing a knot almost invariably takes place near the knot and investigate the effect of knots on strength. Behera and Joshi (2006) demonstrate that failure occurs at minimum diameter and is related to fibre slippage in Dref yarns. Roy, Basu and Majumder (2000) find that cellulosic yarn wrappings on wrap-spun jute yarns impart staggered fracture when warping is dense but catastrophic failure at low wrapping densities. Mukherjee and Majumder (2004) investigate the same type of yarn in comparison with flyer yams and note different modes of fracture (catastrophic and slip-stick) for the two cases. Lappage (2005) finds, in a study of yarn failure, that irregularity and linear density are crucial factors in breakage and that the process is affected by failure of splices, thin places and abrasion problems. Dubinskaile and Milasius (2005) use cyclic stress-to-failure tests to select optimum yarns for carpet manufacture. Fatigue properties of optical fibres are studied in two earlier papers. Glaesemann and Gulati (1991) note the difficulties of establishing a reliable prediction of failure, given that strength distributions of up to thousands of kilometres may be needed, while Gulati (1992) attributes failure to flaw development, finding that large flaws are critical in a fatigue environment. Several papers are devoted to the subject of failure in ropes, arguably the most critical application of all those where failure is important. Chaplin (1999) notes that torsional distortion in a wire rope is a major source of failure and develops a mechanism to account for the phenomenon. Rebel et al. (2000) note that wire ropes are very heavy and examine the feasibility of replacing them with ropes made from fibres. Hobbs et al. (2000) study failure in ropes as a result of buckling by axial compression, even though the rope may be in tension, and develop predictions of failure by mathematical modelling. Johnson, Petrina and Phoenix (2003) find that synthetic ropes made from aramid, PRO and HMPE fibres are superior to wire ones, because they are lighter, more resistant to corrosion, more flexible and less likely to fail. They design a tensioning system to improve even further the resistance to failure in these ropes. Sloan, Bull and Longerich (2005) use HMPE ropes in a study of failure, then use the results to change rope design and achieve a 100% improvement in precluding or delaying failure. McCorkle and co-workers (2003) support the belief that synthetic ropes are superior in properties, deriving information from a long-term study of


177

fatigue on high-performance synthetic ropes. Shin et al. (2006) measure cut-resistance of high-performance yarns, finding that this parameter depends on the slicing angle, the blade sharpness and yarn pre-tensioning, then explain their results by studying differences in failure modes.

8.2.12 Mechanical action Finally in this section, it is obvious that mechanical action can bring about degradation during manufacture or use of fibrous materials but, since this is the major thrust of the entire book, conditions under which this type of attack takes place are dealt with extensively elsewhere and need not be examined in any great detail at this point. A few examples will, though, serve to illustrate the type of work being carried out in this topic area. Fatigue endurance, of various kinds, is addressed by several workers. Sloan and Seymour (1990) present a simple technique, based on linear elastic fracture mechanics, for characterising crack growth in composites subjected to a hostile environment and discuss applications in fatigue testing. Huang, Ramakrishna and Tay (1999) use a micromechanical approach to estimate the tensile strength of a fabric-reinforced composite to allow the failure mode to be identified. Yuce, Varachi and Wei (1990) expose optical fibres to zero-stress ageing, showing that strength, fatigue resistance and ageing behaviour are highly dependent on coating material, chemical environmental exposure and ambient temperature. Glaesemann (1992) attempts to establish a safe stress level for optical fibres to allow prediction of fatigue resistance in a hostile environment to be carried out. Jeddi et al. (2007) compare fatigue behaviour of cotton and cotton-polyester yarns and note three stages in the process, yarn decrimping, fibre slippage under friction, and fibre elongation. The polyester component, as might be expected, is shown to have an important role in stabilising the structure to improve fatigue resistance. High strain rates can change failure modes, and two papers investigate this situation. Termonia (2004) presents a model of the change taking place under small projectile impacts in woven fabrics, by following the strain wave propagation along individual yarns from the impact point. Various projectile characteristics are considered, with slippage of yarns and the fracture process being analysed. Marques and colleagues (2004) subject viscose yarns to high rates of stress, then use SEM examination to show that granular fracture takes place but that the yarn surface becomes less granular as strain rate is increased. Some workers have applied degradation theory and practice to wire ropes, with results which might possibly be applicable to textile materials in rope form. Ridge, Chaplin and Zheng (2001) undertake tests to determine the effects of degradation and impaired quality on fatigue endurance,

178


and find that these factors are not highly significant. Giglio and Manes (2005) subject wire ropes to axial and bending loads under abnormal working conditions in an effort to develop reliable predictions of fatigue life.

8.3

Effects of fatigue fracture on the environment

When fracture takes place, some effect on the environment, small or large, is likely to be encountered, since every human activity leaves a trace of its occurrence in some way on the planet. There are, in general, five possible ways in which the ecological system can be affected. These may be classed as the effect on the air, the land, the water, the noise level and the visual impact.

8.3.1 Air pollution Air pollution is mainly present, almost exclusively in the form of dust, when particles of fibrous material are discarded as a result of fracture. Small fragments of the broken polymer are scattered into the surrounding air and, if they are small enough, remain there as microscopic particles. In addition, any reagents that have previously been applied to the fibres may also be present, either on the particles or as molecular clumps of the actual reagent itself. The presence of these materials can impose adverse conditions in any of three ways. First, the dust may be toxic or otherwise harmful to animals (including human beings) breathing the atmosphere. Asbestos and cotton dust, for example, are known to produce fatal diseases in the lungs, while some dyestuffs and finishes are also carcinogenic. Other illnesses may include bronchitis or pneumonia, if the dusty atmosphere is breathed for an extended period of time, there may be temporary discomfort, in the form of a cough, even after short exposures. Obviously, there would have to be major fatigue fractures taking place before high levels of dust problems of this kind occurred, but the potential for trouble is present if fracture takes place in an enclosed area and exposure time is long. The second effect is much less drastic but still a nuisance. If dust is produced in the atmosphere and machinery is operating nearby, then the lubrication of this machinery may be impaired and breakdown, early failure or malfunctioning can arise. At a minimum, maintenance operations will have to be carried out more frequently, adding significantly to the cost associated with the product manufactured by the machinery. Finally, the presence of dust can cause a different type of financial drawback. If textile goods are stored nearby, then these goods may be contaminated to the point where they are too dirty to be accepted by a customer.


179

This produces an unacceptable economic cost if they must be discarded as waste or, at least, cleaned before being sold.

8.3.2 Land pollution This matter of discarding goods earlier than could reasonably be anticipated from initial predictions of life expectancy is one cause of land pollution. If a textile material suffers fatigue fracture that renders it unsatisfactory for further use, then it is often thrown away. If the discarded textile is one which can easily be decomposed, then the net result is not serious. Cellulosic goods, for instance, decompose rapidly when exposed to weathering, as mentioned above, and the end-products of this process are harmless ones in terms of toxicity. If the fractured textile fabric is made from polyester, though, then decomposition is slower and some of the end-products of breakdown are likely to be toxic. Land pollution also takes place when any chemical compounds applied to the fabric are left in place to be removed over time. Products of decomposition, or the reagents themselves, may be picked up by plants during growth, and passed on into the food chain, or may be eaten directly by animals with the same end-result. Degradation products that can result may include substances such as furans, dieldrins, or other organochemicals capable of producing cancer, liver damage or birth defects in animals or humans. These dangerous compounds are not just left in place, but can be transferred into the water table by rainfall, and then become hazardous causes of water pollution. Leaching from garbage dumps is responsible for much of the pollution that can cause illness in humans, and can be taken up by plants using the water from the local aquifer as a source of nutrients. The net result is one of stunted growth in plants, or the ingestion by animals (including humans) of toxic substances from these plants, to bring about serious consequences if the toxins are concentrated in the liver. Again, the type and extent of harm will vary with fibre type, natural fibres in general being less damaging than synthetic ones. However, the dyes, finishes, etc., applied to fibres will also be leached into the ground water, and those applied to natural fibres (in the form of finishes as well as dyes) may be equally harmful.

8.3.3 Visual pollution An increase in visual pollution also results when premature rejection of textiles as a consequence of fracture occurs. The discarded fibrous material is unsightly, so spoiling the natural beauty of the location, and may also be distracting when, say, a car driver’s attention is caught by it. The cost of

180


removal, to eliminate or reduce these drawbacks, is considerable and adds significantly to the economic burden of society.

8.3.4 Safety concerns The question of safety should also be taken into account when considering the end consequences of fibre fracture. If the textile product is part of a protective arrangement, in the form of garments, ropes, belts, safety harness, etc., then failure at an earlier stage than expected can bring about disastrous consequences. Prediction of failure rate cannot always be reliable. Tests for life expectancy of nylon ropes, for instance, are normally carried out in conditions of tensile stress, but abrasion by contact with rock surfaces is a far more likely cause of failure. Work has shown (Barnett and Slater, 1991) that failure in abrasive mode can take place much earlier than in tensile mode after actinic degradation by outdoor solar exposure has occurred. Similar deviations from expected behaviour can occur when a fibre assembly is constrained to an unexpectedly small radius of curvature, working near its safety limit, or used in conditions for which it had not been designed (as when in contact with chemicals likely to cause weakening, etc.).

8.3.5 Ancillary concerns A further consequence of high fracture rate is the need to provide additional equipment for the manufacture or use of replacement goods. The machines for these applications will have to be maintained or replaced more frequently in such cases, with the resulting increase in undesirable side-effects, such as cotton dust, noise, lubricants, sizes, finishes, dyes, etc. (Slater, 2003). Thus, although the direct effects of fibre fracture may appear to have only small repercussions in terms of environmental consequences, their cumulative effects are widespread and may be considerable in many cases. For that reason it is advantageous to reduce or eliminate them wherever possible.

8.4

Overcoming environmental effects

8.4.1 Fibre factors In trying to minimise the harmful effects of fibre fracture on the environment, the obvious first step to take is an attempt to prevent the fracture from taking place at all. In this aim, the most successful approach begins with the selection of the optimum material that provides the best match of properties for the intended end-use. Thus, for instance, the choice of any


181

fibres that will fracture easily in the application should be avoided as far as possible. Typical examples include avoiding cotton to make ropes where outdoor exposure in contact with moisture (as in tentage or agricultural covers) will be needed, or nylon where abrasive contact under stress is accompanied by constant exposure to severe weathering conditions (as in flags or climbing ropes).

8.4.2 Treatment factors Treatments may also be applied to textile goods in an attempt to prolong service life and hence reduce the environmental loading of a too-frequent need to discard the product. Typical examples include sizing and desizing compounds to reduce frictional force (and hence yarn breakage or dust production) in weaving, lubricants applied during spinning to maintain yarn quality, reagents to provide ultraviolet protection by absorbing energy in the degradative segments of solar radiation, or softeners to reduce brittle contact between yarns of the end-product during use. The latter also enhance comfort, so allowing the textile material (especially if it is a garment) to provide satisfactory wear for a longer period of time, again reducing the possibility of it being discarded prematurely.

8.4.3 Compounded factors Although such treatments may be capable of increasing lifetime and hence reducing pollution when the textile is discarded, there is another factor that must be taken into account in arriving at a balanced decision. The actual chemical agents used to provide increased durability may themselves be environmentally harmful. Some knowledge of what happens when they are eventually discarded with the fibres must be available to assess the comparative risks of early discard of fibres only versus later discard of fibres plus reagents. Pollution by discarded cotton, for instance, is lower than that by discarded polyester, but the flame-resistant treatments that must often be given to cellulosic articles may be more damaging than the polyester fibres alone, especially if (as is often the case) the cotton has a much shorter life and is thus discarded more frequently. Water pollution is especially of concern in such cases, since the more rapid disintegration of cotton may allow faster dissipation of toxic substances into the water table than does the much slower degradation of the synthetic fibre.

8.4.4 Society standards Yet another factor of importance must be considered when any planned attempt to reduce environmental harm is put into practice. Avoiding

182


frequent discard is admirable in theory, but the practice may be less easy to establish. The problem is that society standards (especially in the developed regions of the world) are reluctant to permit lengthy use of textiles to continue. We will not be seen in public if we are dressed in a garment that contains unsightly worn areas, even if the visible damage is only slight and does not detract from the ability of the garment to provide protection or fulfil any modesty needs. We insist on changing soft furnishings in the home, even though they are still serviceable, if they show signs of wear or if they do not match a perceived need to incorporate a new colour scheme or style of decoration. We tolerate vandalism on public transport vehicles, only inflicting punishments that are so laughable as to be a joke even if the vandals are caught. In short, there is an urgent need to accept environmental importance in our entire societal structure. We ignore the destructive actions that are imposed on the planet, or pay mere lip service to the situation by carrying out token ‘solutions’ that are in fact completely ineffective, serving only to transfer any pollution to another location or disguise its presence by dilution or concealment.

8.5

Future trends

8.5.1 Initial approaches From the perspective of the industry as it presently exists, an obvious first approach is to devise some means of prolonging the life of a material so that it does not need to be discarded as quickly. In practical terms, there is an urgent need for increased reliability for every product made. This can only be achieved by extremely careful attention to such matters as product design, materials selection, construction techniques, coloration and finishing treatments. The aim must be to prolong life to provide a tremendously long service period, but there is no possible hope of this happening unless environmental constraints are always selected as the most crucial factor in designing goods.

8.5.2 Limitations Unfortunately, there is an inherent limit to this approach. Current states of knowledge do not allow us to manufacture a product that will never wear out, fade, or change dimensions. Fibres suffer degradation on exposure to light, heat, moisture, mechanical, biological or chemical stresses. Dyestuffs and finishes are slowly destroyed over a period of time. Shrinkage or stretching can take place during maintenance processes. Many decades of research into finding answers to these problems have already taken place, without


183

any complete or resounding success. Thus, one of the foremost needs is to focus attention on how or why degradation takes place, with the aim of reducing or eliminating it, a daunting task.

8.5.3 Industry challenges The task ahead for the textile industry, then, is to produce goods of high quality, with long life expectancy, that pose no threat to the environment. The first casualty of this aim will inevitably be a financial one. Environmental and economic health are incompatible. If the means of improving longevity is to produce a stronger final material, then it will cost more, whether because better quality raw materials (or more of them) must be used, or because a stronger starting fibre type is selected. Either option costs more (environmentally as well as financially) than a material which is designed to exceed the failure threshold by a lower margin. A further aim will be to find finishing treatments that are not harmful to the planet. This may well result in an enforced change in product properties that will have to be accepted by the consumer, with lower ability to satisfy the wishes of purchasers. A current factor militating against this is the highly competitive nature of the industry, when every little demand by the consumer is met in an effort to beat out the competition. Almost certainly, some kind of government intervention will be needed to enforce the reduction in standards caused by the move towards lower consumer satisfaction.

8.5.4 Short-term prospects Because of these over-riding difficulties, the immediate future is unlikely to bring about any real change in attitudes. We continually pay lip service to measures that are supposed to prevent or reduce environmental harm, but the approach is not comprehensive. The main problem is one of greed and selfishness. We insist on trying to preserve the economic strength of a nation or an industry, whenever environmental protection is suggested, but fail to realise that the two are incompatible. The environment is always sacrificed for the sake of the economy and this will continue to be our priority until we are forced to change our ways.

8.5.5 Hope for the future There is, however, currently considerable interest in the problem of global warming and this may be the first sign of a change in attitude. Real change is unlikely to happen while we live in an economy-driven society, but if climate change forces us to re-evaluate our priorities, then the textile industry should be ready to act in accordance with the new needs.

184


As described elsewhere (Slater, in preparation), it is unlikely that this real change will take place until a major catastrophe forces us to modify totally our lifestyle. The catastrophe will need to have at least three major components: 1. It will have to take place in an industrial nation (or simultaneously in many different parts of the industrialised world). 2. It will have to be the cause of death and major loss of property for millions of wealthy people. 3. It will have to be directly and obviously the result of human activity. Calamitous events already take place in the world, but they tend not to affect the rich nations, and arguments can be made that they are the result of natural causes. Floods, earthquakes and tidal waves, for instance, may be more prevalent as a result of global warming, but we cannot be sure either that the correlation is a direct one or that the increased adverse weather patterns are not brought about as part of some natural cycle. Until we experience a situation where (for example) an entire freeway is blocked with traffic during an inversion, causing thousands or millions to die from asphyxiation in the gridlock from carbon monoxide poisoning, we will not accept the need to reduce road traffic. As long as ‘free’ energy from solar, wind, tidal or other new sources is touted as a solution to our power shortage problems, we will continue to allow economies to expand, totally ignoring the environmental cost of the structures and distribution networks necessary for this ‘free’ energy to be made and transferred to its place of use, plus the environmental cost of the raw materials consumed in making the products (vehicles, houses, industrial goods, etc.) able to be created by its use.

8.5.6 Fundamental needs There is a vital need for education in the environmental field and, assuming that it happens, it is interesting to consider the likely result in the field of textiles and specifically in the area of fatigue failure. The first need is an increased margin of reliability for the product subjected to stress, so that it simply does not fracture for a tremendously long time after being put into service. This will require extremely careful product design, with selection of materials, finishes, etc., being carried out purely on environmental grounds, rather than on economic ones. Safety margins will have to be amortised over very long time periods in order to compensate for the inevitable slow degradation that will take place, so that stress resistance continues even after much of the initial property is lost. Next, we should investigate why (and whether) the end-product is actually needed. If the item likely to fracture is a luxury, rather than a necessity,


185

then its production should be discontinued. The main problem here is to convince society to redefine its concept of what counts as luxury and necessity, respectively. In general, ornamentation and safety may be said to lie at opposite ends of this scale, but ornamentation that protects from adverse weather conditions, or identifies an enemy in battle, and safety in the form of a seat belt for a gas-guzzling vehicle or a harness for luxury sport like bungee jumping or paragliding are obvious exceptions to this classification. Naturally, newer approaches to the production of textiles in general, such as effluent treatment, recycling, the replacement of wet treatments with more efficient (or dry) ones, lower-noise and lower-energy equipment and other such developments, should also be adopted in producing fractureresistant goods. Many of these techniques are genuinely helpful in the aim of preserving to some extent the planet’s health, though others merely substitute one harmful process by a different, but equally harmful, one. The important step to take is to distinguish between them so that genuine improvement in environmental protection can take place.

8.6

References

Ashbee, E. and Ashbee, K.H.G. (1985) J. Materials Science Letters, 4(3), 249–250. Bao Guoping and Slater, K. (1990) J. Text. Inst., 81, 59. Barnett, R.B. and Slater, K. (1991) J. Text. Inst., 82, 417. Behera, B.K. and Joshi, V.K. (2006) J. Text. Inst., 97(6), 503–512. Bittner-Rohrhofer, K., Humer, K. and Weber, H.W. (2002) Cryogenics, 42(5), 265–272. Brown, J.R. et al. (1983) Text. Res. J., 53, 214. Carlsson, D.J., Garton, A. and Wiles, D.M. (1977) J. Appl. Polym. Sci., 21, 2963. Carswell, W.S. and Roberts, R.C. (1980) Composites, 11(2), 95–99. Chaplin, C.R. (1999) Engineering Failure Analysis, 6(2), 67–82. Cho, K.H. and Jeong, S.H. (2005) J. Materials Science, 40(20), 5341–5347. Cybulska, M. and Goswami, B.C. (2001) Fibres and Textile in Eastern Europe, 9(4), 20–23. Dooraki, B.F., Nemes, J.A. and Bolduc, M. (2006) Journal de Physique IV, 134, 1183–1188. Dubinskaile, K. and Milasius R. (2005) Materials Science, 11(3), 288–291. Ellyin, F. and Rohrbacher, C. (2000) J. Reinforced Plastics and Composites, 19(17), 1405–1427. Ellyin, F. and Rohrbacher, C. (2003) J. Reinforced Plastics and Composites, 22(7), 615–636. Fujii, Y., Murakami, A., Kato, K., Yoshiki, T., Maekawa, Z. and Hamada, H. (1994) J. Materials Science, 29(16), 4279–4285. Ghosh, A. (2006) Fibers and Polymers, 7(3), 310–316. Ghosh, A., Ishtiaque, S.M. and Rengasamy, R.S. (2005a) Text. Res. J., 75(10), 731–740. Ghosh, A., Ishtiaque, S.M. and Rengasamy, R.S. (2005b) Text. Res. J., 75(10), 741–744.

186


Giglio, M. and Manes, A. (2005) Engineering Failure Analysis, 12(4), 549–568. Glaesemann, G.S.T. (1992) Proc. SPIE: International Society for Optical Engineering, 1580, 125–129. Glaesemann, G.S. and Gulati, S.T. (1991) Optical Engineering, 30(6), 709–715. Guillemet, C. (1995) Revue de Metallurgie, 92(2), 253–263. Gulati, S.T. (1992) Proc. 41st Intl. Wire and Cable Symposium, 612–621. Haixia, Z., Yuan, X. and Shanyuan, W. (2006) Cotton Textile Technology, 33(7), 16–19. Hale, J.M., Gibson, A.G. and Speake, S.D. (2002) Journal of Composite Materials, 36(3), 257–270. Hobbs, R.E., Overington, M.S., Hearle, J.W.S. and Banfield, S.J. (2000) J. Text. Inst., 91(3), 335–358. Holt, L.A. and Milligan, B. (1976) J. Text. Inst., 67, 269. Hon, N.S. (1977) J. Polym. Sci: Polym Chem., 15, 725. Huang, Z.M., Ramakrishna, S. and Tay, A.Q. (1999) J. Composite Materials, 33(19), 1758–1791. Jeddi, A.A., Nosraty, H., Taheriotaghsrara, M. and Karimi, M. (2007) J. Elastomers and Plastics, 39(2), 165–179. Jeon, Y.B., Seo, D.W. and Lim, J.K. (2005) Key Engineering Materials, 297–300, 213–216. Johnson, E., Petrina, P. and Phoenix, S.L. (2003) Oceans 2003: IEEE, 3, 1329–1334. Kawai, M., Yajima, S., Hachinohe, A. and Takano, Y. (2001) Journal of Composite Materials, 35(7), 545–576. Kim, H.J. and Seo, D.W. (2006) International Journal of Fatigue, 28(10), 1307–1314. Lappage, J. (2005) Text. Res. J., 75(6), 512–517. Legrand, N., Remy, L., Molliex, L. and Dambrine, B. (2002) International Journal of Fatigue, 24(2–4), 369–379. Marques, F.A., Silva, E.C., Silva, A.C. and Drean, J.Y. (2004) Text. Res. J., 74(8), 551–554. McCorkle, E., Chou, R., Stenvers, D., Smeets, P., Vlasblorn, M. and Grootendorst, E. (2003) Oceans 2003: IEEE, 2, 1058–1063. Mukherjee, S. and Majumder, P.K.A. (2004) Indian Journal of Fibre and Textile Research, 29(4), 436–439. Naik, N.K. and Singh, M.N. (2001) J. Text. Inst., 92(2), 164–183. Nkiwane, L.C. (2001) Kautschuk Gummi Kunststoffe, 54(12), 648. Pan, N., Hua, T. and Qiu, Y.P. (2001) Text. Res. J., 71(11), 960–964. Rebel, G., Chaplin, C.R., Groves-Kirkby, C. and Ridge, I.M.L. (2000) Insight, 42(6), 384–390. Ridge, I.M.J., Chaplin, C.R. and Zheng, J. (2001) Engineering Failure Analysis, 8(2), 173–187. Roy, A.N., Basu, G. and Majumder, A. (2000) Indian Journal of Fibre and Textile Research, 25(2), 92–96. Saitta, A.M., Soper, P.D., Wasermann, E. and Klein, M.L. (1999) Nature, 389(6731), 46–48. Shatkay, A. and Wetherall, I.L. (1977) J. Polym. Sci: Polym. Chem., 15, 1735. Shimojo, M. and Bowen, P. (1998) Fatigue and Fracture of Engineering Materials and Structures, 21(2), 171–182. Shin, H.S., Erlich, D.C., Simons, J.W. and Shockey, D.A. (2006) Text. Res. J., 76(6), 607–613.


187

Slater, K. (1976) Textile Progress, 8, No. 3. Slater, K. (1991) Textile Degradation, The Textile Institute, Manchester. Slater, K. (2003) Environmental impact of textiles, Woodhead, Cambridge, UK. Slater, K. (in preparation) Survival prospects for humanity. Sloan, F., Bull, S. and Longerich, R. (2005) Oceans 2005: IEEE, 1, 629–635. Sloan, F.E. and Seymour, R.J. (1990) J. Composite Materials, 24(7), 727–739. Song, D.Y., Takeda, N. and Kawamoto, M. (2000) Materials Science and Engineering A (Structural Materials – Properties, Microstructure and Processing), A278(1–2), 82–87. Suzuki, A. (1982) J. Text. Mach. Soc. Japan, 35(2), T23. Termonia, Y. (2004) Text. Res. J., 74(8), 723–728. Tyagi, G.K., Goyal, A., Parmar, S. and Bhatnagar, S. (2004) Textile Asia, 36(12), 29–33. Wall, M.J., Frank, G.C. and Stevens, J.R. (1971) Text. Res. J., 41, 38. Wang, J.Z., Dillard, D.A., Wolcott, M., Kamke, F.A. and Wilkes, G.L. (1990) J. Composite Materials, 24(9), 994–1009. Yu, S.-Y., Ishii, H. and Tohgo, K. (1994) Fatigue and Fracture of Engineering Materials and Structures, 17(5), 571–578. Yuce, H.H.,Varachi, J.P. and Wei, T. (1990) Proc. SPI: International Society for Optical Engineering, 1174, 279–288. Zeidman, M. and Sawhney, P.S. (2002) Text. Res. J., 72(3), 216–220.

9 Fatigue of polymer-matrix textile composite materials Y GOWAYED, Auburn University, USA

Abstract: In this chapter, some of the work on experimental evaluation and numerical modeling of the fatigue behavior of polymer matrix textile composites is presented, addressing the effect of test parameters such as frequency, max stress, and the nature of the cycles on the response of textile composites. The impact of the preform architecture varying from weaves to braids and knits, both 2D and 3D, is discussed and its impact on the time-dependent response is presented. Numerical models utilizing statistical analysis, stiffness degradation and residualstrength, either singly or combined, are also discussed. Key words: fatigue response, textile composites, numerical modeling, polymer matrix.

9.1

Introduction

The fatigue test is typically carried out by repeat loading and unloading in tension-tension, tension-compression or compression-compression depending on the service conditions the loading setup is trying to mimic. This experiment is defined by the R-ratio representing maximum to minimum loading stress (R = min stress/max stress) and the load can be applied in different repeat-forms, such as step or sinusoidal. The response of textile composites to fatigue loading is affected by different parameters depending on the nature of the load cycle. In tensiontension fatigue, the fibers are mostly in tension, and cracks may initiate at defect areas or cross-over points, where the stress has its maximum value due to yarn crimp, and then progresses through the rest of the cross-section. In compression-compression, impregnated yarns are prone to kinking, causing crack initiation at the yarn/matrix interface and subsequent propagation in zones of maximum stress or strain. The behavior of textile composites under fatigue loads is different to that of laminated composites due to the difference in the spatial relationship between their fiber architecture and the load direction. In laminated 188

Fatigue of polymer-matrix textile composite materials

189

composites, yarns lie in a group of planes (lamina) parallel to each other and the load has a defined angle relative to the fiber architecture within each lamina. Such regular yarn arrangement facilitates the understanding of the behavior and the prediction of the response. In textile composites, the story is more complicated by the architecture of the yarns that is not limited to few defined planes. The majority of reinforcing fabric-preforms are either woven or braided. In woven fabric composites, yarns are interlaced with each other in orthogonal planes allowing the yarns to have crimp within a certain plan that is perpendicular to planes of other yarns. Woven fabrics can be formed in two-dimensional, such as plain weaves and satins, or three-dimensional, such as orthogonal weaves, layer interlock and angle interlock. The complex yarn arrangement enhances certain properties, such as stiffness and strength, in the direction of the preform yarns. For example, in an angle interlock weave the warp yarn lies at an angle in the out-ofplane direction, as shown in Fig. 9.1, enhancing the out-of-plane shear, modulus and strength. Yarns are intertwined to form two-dimensional or three-dimensional braided architectures giving a unique feature to braided structures that is not limited to orthogonal planes. Two-dimensional braids such as biaxial and triaxial braids (both can be in a 1/1 or 2/2 form) are known for their high in-plane properties while three-dimensional braids, 2-step and 4-step braids, are known for their high out-of-planes shear moduli and strength. The reader is encouraged to read more on fabric preforms, for example (Chou and Ko, 1989), prior to reading this chapter to develop an understanding of the preforms and an appreciation of their possible role in fatigue response. It is a common belief, which is supported by few experimental observations, that toughness of textile composites is higher than laminated composites with comparable fiber volume fractions. This can be understood by looking at the crack path through the two types of composites. In laminated composites, if a crack starts at an interface between two laminas, or reaches that point along its path, it will rapidly progress along that un-reinforced plane without hindrance. On the other hand, in 3D textile composites, the

Weft yarns

Warp yarns

9.1 Image of angle interlock weaves with warp yarns (dark gray) and weft yarns (light gray) produced by pcGINA.

190


Crack Laminated composites

3D textile composites

9.2 Crack path in laminated and textile composites.

crack will start either at a void, which may be located at a resin-rich area under the cross-over point of two yarns, or at the cross-over point itself and progress along the interface between a fiber and matrix in some direction. This will not continue for long due to the interlacing or intertwining feature of woven and braided composites, respectively. The crack, trying to maintain a minimum energy path, will have to change direction prolonging its overall path and opening new surfaces, consuming further energy from its driving force. That tortuous crack path in textile composites, as compared to an almost straight path in laminated composites, will require more energy, explaining its added toughness (Fig. 9.2). In 2D textile composites where layers of 2D fabrics are stacked on top of one another, the story about crack path propagation may lie between the two extremes mentioned above. Although 2D layered structures can be considered as laminated systems, they are still textile composites with out-of-plane yarn crimping and movement. Each layer typically ‘nests’ within adjacent layers prolonging the crack path to a degree between laminated composites and fully integrated 3D textile composites. Few experimental and analytical activities have been conducted to evaluate and characterize the fatigue response of textile composite materials. In the next few sections a brief summary of these activities is presented.

9.2

Experimental evaluation of fatigue response

Some work is available in archived literature on experimental fatigue behavior of textile composites for different fabric preforms such as woven (Pandita et al., 2001, Kawai and Taniguchi, 2006, Toubal et al., 2006, Tsai et al., 2000), braided (Tate et al., 2006, Carlos, 1994, Portanova and Deaton, 1995, Fujihara et al., 2007) and knitted (Pandita and Verpoest, 2004, Chou et al., 1992) composites. For polymer composites, the fabric preform architecture controls the behavior of the composite both mechanically and thermally. It is important to view the fabric yarns as reinforcing elements within an encompassing matrix responsible for its integrity and protection. The orthogonal nature of woven fabrics, the out-of-plane inclined yarns of braided structures and the loops of knitted architectures control the composite behavior and dictate its response.


191

Tension-tension fatigue response has been studied by Pandita et al. (2001) for on and off-axis directions for a balanced (i.e., similar warp and weft yarn type and density) plain weave Eglass/epoxy for a fiber volume fraction of 50%, a stress ratio of 0.1 and a varying load frequency from 0.5 to 10 Hz. Tests were conducted in the weft direction, 45º and 60º angles to the weft direction. The static stress-strain curves related to these investigations showed almost linear response for the weft direction load, and a nonlinear response at 45º and 60º angles, along with a clear knee in the curve, with the degree of the nonlinearity at the 45º higher than that at 60º. The fatigue life in the weft direction was much longer than that in the bias directions at the same stress level. The weft direction loading failure sequence has been characterized by cracks at low number of cycles, followed by crack propagation in yarns transverse to the load direction. The transverse cracks grew either in a rich resin area or in the longitudinal tows within the same layer followed by fiber fracture and ultimate failure of the composite. For fatigue tests in the bias direction, the crack showed fiber/ matrix debonds followed by fiber reorientation, fiber failure and ultimate composite failure. The reorientation process included large strain values of up to 20%. In the bias direction loading, it was observed that the width of the hysteresis loops increased with the increase of the maximum stress value showing excessive damage. The temperature of the composite was monitored at high load frequency levels and an increase of approximately 10 ºC was observed at 10 Hz for on-axis fatigue load and did not have an impact on the results. A higher temperature rise in the bias, off-axis, direction was observed along with a major impact on the fatigue response. Similar experiments were conducted for a balanced plain weave T300/ epoxy composite at room temperature and 100 ºC in the 0º, 15º, 30º and 45º angles at loading frequencies of 2 and 10 Hz with a stress ratio of 0.1 (Kawai and Taniguchi, 2006). Nonlinear response was observed for all the off-axis direction loading with a knee in the curve and a reduction in modulus and strength with the increase in the off-axis angle. The 100 ºC static loading curve was qualitatively similar to the 0 ºC loading, but showed lower modulus and strength values. Fatigue response difference between on- and off-axis loading was similar to that observed by Pandita et al. (2001). The shape of the S-N curve at 100 ºC and the effect of orientation were similar to that at room temperature, with a minor decrease in the fatigue life at the higher temperature. An increase as high as 70 ºC in the surface temperature of the specimen was reported for 10 Hz in the off-axis direction showing a possible reason of fatigue response sensitivity to load frequency. The fatigue of 3D woven composites manufactured from three and five weft layer angle interlock carbon/epoxy composites as shown in Fig. 9.3 was studied by Tsai et al. (2000). The total fiber volume fraction for the threelayer material was 25% while the five-layer fabric had a 32.5% fiber volume

192

Fatigue failure of textile fibres Warp

Weft

(a) Warp Weft

(b)

9.3 Schematic illustration of three and five weft layer angle interlock weaves (Tsai et al., 2000).

fraction. Fatigue tests were carried out at a stress ratio of zero and a load frequency of 5 Hz in the direction of the weft yarns. The three-layer specimens had a fatigue life longer than that of the five-layer specimens under the same stress level. It was observed that cracks initiated in the tows transverse to the load direction (warp yarns) at low number of cycles. The authors believed that this event caused local bending due to weft yarn undulation producing matrix cracks starting at the specimen surface. At a higher number of cycles, the surface cracks grew further through the warp yarns. These cracks continued to increase in density until debonding between fibers and matrix occurred. Prior to failure, the cracks covered the entire specimen cross-section, breaking weft yarn fibers leading to fracture. A schematic drawing detailing failure steps as observed from micrographic images is shown in Fig. 9.4. The effect of braid angle on the fatigue performance of biaxially braided carbon/epoxy composites was investigated by Tate et al. (2006). The preforms were manufactured from AS4 carbon yarns at braid angles of 25º, 30º and 45º. Static tests showed a nonlinear stress-strain curve and that the increase in braid angle, measured as half the angle between the two braided yarns, caused a decrease in the longitudinal modulus and strength. Fatigue tests were conducted for a stress ratio of 0.1 at 10 Hz frequency. For an endurance limit set at 1 million cycles, the 25º and 30º braids reached the endurance limit at 40% of the ultimate tensile strength (UTS), while the 45º reached it at 50% of UTS. Failure was reported to occur in the last 10% of the load history and to be sudden and catastrophic including matrix cracking, ply delamination and fiber fracture.


193

(1)

(2)

(3)

(4)

(5)

9.4 Schematic of damage progression in three-layer angle interlock woven composites (Tsai et al., 2000).

The effect of yarn spatial orientation relative to the loading direction on the fatigue performance of textile composites is most evident in knitted composites where yarns are formed as loops with a varying angle to applied load. Additionally, knitted fabrics are highly deformable and can stretch to large strain values as compared to woven or braided composites. Hence, it is expected that the fiber/matrix interface will carry most of the load causing early cracks and delaminations. Tension-tension fatigue of warp and weftknitted Eglass/epoxy composites with a fiber volume fraction of 30% was carried out at a stress ratio of 0.1 and a load frequency of 0.5 to 10 Hz (Pandita and Verpoest, 2004) and compared to woven composites with a fiber volume fraction of 50%. It was reported that the fatigue life of the knitted composites was lower than that of woven composites load in the yarn direction but was comparable to that of woven composites loaded at

194


±45º. The failure in the knitted fabric composites was reported to be highly dependent on the orientation of the yarns and that crack propagation was faster where the yarns were perpendicular to the load direction. The effect of loading frequency on the rise in the surface temperature was stressed as having a major impact on the fatigue life and as being different for different composite performs. At 5 Hz the surface temperature of woven composites reached 70 ºC after a few thousand cycles and about 10 ºC for knitted fabric composites after a hundred thousand cycles.

9.3

Modeling of fatigue behavior

Different routes can be adopted to present a consistent scenario for parameters involved in the fatigue behavior of textile composites. These parameters can be external, such as strain or deformational values, or internal, such as crack propagation, energy release, yarn re-orientation, change in a material property such as stiffness or strength, or a combination of two or more of these parameters. Such parameters can be monitored and analyzed either statistically, by quantifying their performance and using a statistically significant number of experiments to generate empirical equations that can be used to predict behavior of other composite materials, or mechanistically by attempting to develop a mechanistic understanding for the material response that typically links constituent material behavior to composite response. Archived literature has recently been enriched with many approaches to model the fatigue behavior of textile composites such as statistical analysis, stiffness degradation (Yoshioka and Seferis, 2002, Wen and Yazdani, 2008, Huang and Ramakrishna, 2003), residual-strength and combined or cumulative damage approaches (Gowayed and Fan, 2001, Hochard et al., 2006). In residual strength approaches, fatigue failure is assumed to occur when residual strength becomes equal to the maximum applied-stress amplitude, while in the modulus degradation approach fatigue failure causes the value of the modulus to be reduced to an empirically predetermined value. A damage tolerance approach combines both residual strength degradation and modulus degradation by using the concept of strain energy release rate. Statistical approaches (Himmel and Bach, 2006, Gagel et al., 2006, Van Paepegem and Degrieck, 2002) have also been frequently used for the prediction of residual strength/stiffness and fatigue life of composite laminates in which, based on the statistical nature of composite fatigue response, a cumulative damage model can be developed to predict the residual strength and fatigue life of composite materials subjected to cyclic loading. Statistical analysis of fatigue response, for textile composites or any other material, tries to identify one of more of the basic characteristics of the


195

material under testing and their change with repeated cyclic load. As an example, cyclic strength degradation with load cycles was characterized by Himmel and Bach (2006) as follows: S ( N ) = A − B log ( N ) where, S is the cyclic strength, N is the number of cycles to failure, A and B are curve fitting parameters that are determined from the test data for 5%, 50% and 95% probability of survival using the least squared error method. Damage accumulation was accounted for using the model developed by Hahn and Kim (1975): Rn = R0 − [ R0 − Smax ]

k

where, Rn is the residual strength, n is the number of cycles, Smax is the maximum stress amplitude, R0 is the strength of the virgin material, and k is a curve fitting parameter. Graphs and tables were developed by Himmel and Bach (2006) for parameters A, B and k for various carbon fabric preforms infiltrated with toughened epoxy and polyvinylester resins using resin transfer molding (RTM) and vacuum assisted resin infusion (VARI) tested at different stress ratios of +0.1 and −1 at a maximum loading frequency of 5 Hz. A non-scalar approach was used by Gagel et al. (2006) to statistically describe the degradation and failure of textile composites utilizing information on the number of cycles until failure, stiffness degradation, and damage evolution. Stiffness was evaluated utilizing the slope of the stress-strain hysteresis curve for each cycle between maximum and minimum loading stress. Damage was characterized as the change in stiffness over a certain number of load cycles. Plotting damage as a function in stiffness degradation versus fraction of fatigue life defined three different stages of performance: i) rapid increase of damage in the first 10% of fractional fatigue life, ii) steady state damage progression until approximately 90% of fractional fatigue life and iii) rapid increase in damage in the final 10% of the life. Crack densities in ±45º and 90º were used to predict the stiffness degradation during fatigue. Load cycle dependent increase in crack density was described using the following empirical equation: ρ = mP ( σ max ) ⋅ log ( n N f ) + nP ( σ max ) where, ρ is the crack density in a certain direction, mP and nP are curve fitting parameters, n is the number of cycles, σmax is the maximum stress and Nf is the number of cycles to failure. Values were measured for mP and nP for glass-fiber non-crimp fabric reinforced epoxy composites tested at a stress ratio of −1 at a frequency of 6 Hz. In a stiffness degradation approach, the stiffness, or compliance, is first calculated using information on fiber, matrix and preform architecture

196


typically utilizing an established and experimentally verified method. A stepwise approach is then invoked to introduce a time-dependent or cycledependent decay to the function due to crack propagation and/or internal damage. Failure is assumed to occur when the stiffness and/or the strain reach a certain predefined criterion. In Yoshioka and Seferis (2002) the authors utilized a combination of the crimp model (Chou and Ko, 1989) and a modified shear lag model (Lee and Daniel, 1990) to evaluate the stiffness of the composite and the effect of transverse cracking on its deterioration. Warp yarn debonding was also taken into consideration by removing the contribution of the debonded warp yarn from the overall stiffness of the composite. Cumulative damage approaches typically depend on energy or strain considerations for crack initiation that would impact a property, or more, in the composite and allow it to degrade with the number of cycles. Similar to stiffness degradation, a predefined bound on degradation will allow for the prediction of the number of cycles to failure. In Hochard et al. (2006) and Payan and Hochard (2002) the authors utilized a damage approach using three different parameters, one for the warp, another for the weft, and a third for in-plane shear (i.e., ±45º). Cracks were assumed to run in any one of the three directions under the assumptions of plain stress conditions. The strain energy for each laminate was defined utilizing these damage parameters and split into tensile and compressive components to describe crack opening and closing as a function in the change of the damage parameter. It was assumed that the damage parameter for the in-plane shear was caused by the combinatory effect of the static load and the cyclic nature of the fatigue load. Applying this model for a balanced 4-harness satin carbon fabric composite under tension-tension fatigue loads on ±45º samples showed a rapid increase in the in-plane shear damage parameter in the first cycle, followed by a slow increase in damage progression until failure. A progressive damage model has been constructed by Gowayed and Fan (2001) based on modeling of textile composites using the Graphical Integrated Numerical Analysis (pcGINA) (Gowayed et al., 1996, Gowayed, 1997, Gowayed and Yi, 1997, and Gowayed and Fan, 2001), traditional and hybrid finite element analysis (FEA), and a functional relationship between life, and strength and modulus degradation of composite elements as shown schematically in Fig. 9.5. In this model, traditional FEA is used to evaluate the stiffness reduction and the energy release rate due to crack propagation in the composites. An ideal fabric geometrical representation is constructed by calculating the location of a set of spatial points ‘knots’ that can identify the yarn center-line path within the preform space based on the fabric forming process in a typical textile machine. This is followed by incorporating a B-spline function to approximate a smooth yarn center-line path

Fatigue of polymer-matrix textile composite materials Fiber & fabric info

197

Geometric model Apply fatigue loading (± ΔP) Hybrid FE analysis (GINA) Stiffness & stress-strain values

No

Stress > Strength?

Energy balance & Stiffness reduction Yes

Complete failure?

No

Crack initiation & propagation

Stop

9.5 Schematic flowchart for fatigue modeling of textile composite materials (Gowayed and Fan, 2001).

relative to the identified knots. A repeat unit cell of the modeled preform is identified from the geometric modeling and used to represent a complete yarn or tow pattern. A hybrid finite element approach is used to divide the unit cell into smaller subcells as hexahedral brick elements with fibers and matrix around each integration point. A virtual work technique is utilized to calculate the mechanical properties of the repeat unit cell. An iterative loading procedure is used to apply load and distribute stresses in the unit cell for each load cycle. Crack initiation and propagation is monitored, based on a maximum stress criterion, during each load cycle and strength and stiffness are accordingly degraded. Load can be concentrated, distributed, parabolic, etc. The displacement at each node in the unit cell continuum is calculated under the applied load. Strain and stresses at each node are calculated from the knowledge of the node displacement, the shape function and the stiffness around each node. The applied stress at each node is compared to the strength around that node. If the node fails to carry the applied stress, a crack is initiated at this node and mapped onto a global growing network of cracks within the unit cell. Energy balance and stiffness degradation are calculated for this crack and a functional relationship between the degradation of composite strength and the cycle number is used to compute the residual strength of the composite. Results obtained from this study were compared to data obtained in Tsai et al. (2000) and are shown in Figs 9.6 and 9.7.

Fatigue failure of textile fibres 1

Normalized max. stress

0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0

0

1

2

3

4 Log2N

5

6

7

9.6 Comparison between experimental work (circles) and model prediction (line) for angle interlock woven composites shown in Fig. 9.3a (Gowayed and Fan, 2001).

1 0.9 Normalized max. stress

198

0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0

0

1

2

3 4 Log2N

5

6

7

9.7 Comparison between experimental work (circles) and model prediction (line) for angle interlock woven composites shown in Fig. 9.3b (Gowayed and Fan, 2001).


9.4

199

Conclusions

In the past few years, archived literature witnessed an increased activity in the study of fatigue of polymer-matrix textile-composite materials. As textile composites performing and consolidation processes moved out of the labs into manufacturing companies, better control on material properties was achieved allowing reduction in material variability and opening venues for systematic analysis of these materials for 2D and 3D preforms. Such increase in activity, from an analytical perspective, can be attributed to a better understanding of fabric preform architectures and their spatial distribution facilitated by faster computing micro-processors and implementation of sophisticated mathematical models. These activities are expected to continue and increase as more textile composite products are introduced into the market to further our understanding of their fatigue response. This will not only be limited to polymer-matrix composite systems but will expand to ceramic-matrix textile-composite materials as well.

9.5

References

Carlos, K.G., ‘Fatigue behavior of three-dimensional triaxially braided composites’, M.Sc. Thesis, North Carolina A&T University, 1994. Chou, T.W., and Ko, F.K. (eds), Textile structural composites, Elsevier, Amsterdam, pp. 209–263, 1989. Chou, S., Chen, H.C., and Lai, C.C., ‘The fatigue properties of weft knit fabric reinforced epoxy resin composites’, Composite Science and Technology, Vol. 45, pp. 283–291, 1992. Fujihara, K., Yoshida, E., Nakai, A., Ramakrishna, S. and Hamada, H. ‘Influence of micro-structures on bending properties of braided laminated composites,’ Composites Science and Technology, Vol. 67, No. 10, pp. 2191–2198, 2007. Gagel, A., Fiedler, B., and Schulte, K., ‘On modelling the mechanical degradation of fatigue loaded glass-fibre non-crimp fabric reinforced epoxy laminates,’ Composite Science and Technology, Vol. 66, pp. 657–664, 2006. Gowayed, Y., ‘The effect of voids on the elastic properties of textile composites’, ASTM Journal of Composite Technology & Research, Vol. 18, No. 2, pp. 168–173, 1997. Gowayed, Y., and Fan, H., ‘Fatigue behavior of textile composite materials subject to tension-tension loads,’ Polymer Composites, Vol. 22, No. 6, pp. 762–769, 2001. Gowayed, Y., and Yi, L., ‘Mechanical behavior of textile composite materials using a hybrid finite element approach’, Polymer composites, Vol. 18, No. 3, pp. 313–319, 1997. Gowayed, Y., Pastore, C., and Howarth, C., ‘Modification and application of unit cell continuum model to predict the elastic properties of textile composites’, Composites Part A – Applied Science and Manufacturing, Vol. 27A, pp. 149–155, 1996. Hahn, T., and Kim, R.Y., ‘Proof testing of composite materials,’ Journal of Composite Materials, Vol. 9, pp. 297–311, 1975.

200


Himmel, N., and Bach, C., ‘Cyclic fatigue behavior of carbon fiber reinforced vinylester resin composites manufactured by RTM and VARI’, International Journal of Fatigue, Vol. 28, pp. 1263–1269, 2006. Hochard, Ch., Payan, J., and Bordreuil, C., ‘A progressive first ply failure model for woven ply CFRP laminates under static and fatigue loads,’ International Journal of Fatigue, Vol. 28, pp. 1270–1276, 2006. Huang, Z.M., and Ramakrishna, S., ‘Modeling inelastic and strength properties of textile laminates: a unified approach,’ Composites Science and Technology, Vol. 63, pp. 445–466, 2003. Kawai, M., and Taniguchi, T., ‘Off-axis fatigue behavior of plain weave carbon/epoxy fabric laminates at room and high temperatures and its mechanical modeling’, Composites: Part A 37, pp. 243–256, 2006. Lee, J.W., and Daniel, I.M., ‘Progressive transverse cracking of cross-ply composite laminates,’ Journal of Composite Materials, Vol. 24, No. 11, pp. 1225–1243, 1990. Pandita, S., and Verpoest, I., ‘Tension-tension fatigue behavior of knitted fabric composites,’ Composite Structures, Vol. 64, pp. 199–209, 2004. Pandita, S., Huysmans, G., Wevers, M., and Verpoest, I., ‘Tensile fatigue behavior of glass plain-weave fabric composites in on- and off-axis directions,’ Composites: Part A 32, pp. 1533–1539, 2001. Payan, J., and Hochard, Ch., ‘Damage modeling of carbon/epoxy laminated composites under static and fatigue loads,’ International Journal of Fatigue, Vol. 24, pp. 299–306, 2002. Portanova, M.A., and Deaton, J.W., ‘Impact and fatigue resistance of a [±30º/0º] three-dimensional braided carbon epoxy composites’, ASTM Special Technical Publication, p. 1230, 1995. Tate, J., Kelkar, A., and Whitcomb, J., ‘Effect of braid angle on fatigue performance of biaxial braided composites’, International Journal of Fatigue 28, pp. 1239–1247, 2006. Toubal, L., Karama, M., and Lorrain, B., ‘Damage evolution and infrared thermography in woven composite laminates under fatigue loading’, International Journal of Fatigue, Vol. 28, pp. 1867–1872, 2006. Tsai, K.H., Chiu, C.H., and Wu, T.H., ‘Fatigue behavior of 3D multi-layer angle interlock woven composite plates’, Composites Science and Technology, Vol. 60, No. 2, pp. 241–248, 2000. Van Paepegem, W., and Degrieck, J., ‘Coupled residual stiffness and strength model for fatigue of fibre-reinforced composite materials,’ Composites Science and Technology, Vol. 62, pp. 687–696, 2002. Wen, C., and Yazdani, S., ‘Anisotropic damage model for woven fabric composites during tension-tension fatigue,’ Composite Structures, 82, 127–131, 2008. Yoshioka, K., and Seferis, J., ‘Modeling of Tensile fatigue damage in resin transfer molded woven carbon fabric composites,’ Composites, Part A, Vol. 33, pp. 1593–1601, 2002.

10 Fatigue damage in structural textile composites: testing and modelling strategies W VAN PAEPEGEM, Ghent University, Belgium

Abstract: This chapter discusses the fatigue behaviour of structural textile composites, where the textile yarns are embedded in a plastic matrix to enhance their mechanical properties. Most chapters so far have dealt with the fatigue behaviour of dry, non-impregnated textile yarns, while here, the textile fibre architecture has been impregnated with a resin and has been cured to a solid state. The different testing and modelling strategies for fatigue of structural composites are presented and future trends and challenges are discussed. Key words: composites, fatigue, testing, numerical modelling, textile reinforcement.

10.1

Introduction

In the broadest sense, composite materials are a combination of two or more chemically different phases with a distinct interface between them. Thus a composite is heterogeneous and consists of two or more materials which together produce desirable properties that cannot be achieved with any of the constituents alone. Fibre-reinforced composite materials, in particular, consist of high strength and high modulus fibres in a matrix material. In these composites, fibres are the principal load-carrying members, and the matrix material keeps the fibres together, acts as a load-transfer medium between fibres, and protects fibres from being exposed to the environment (e.g., moisture, toxic agents, etc.) (Mallick, 1997; Herakovich, 1998; Reddy, 1997). The matrix material can vary from metals over ceramics to concrete, but thermosetting and thermoplastic resins are by far the most common matrix materials for structural composites (Stewart, 2002). The most important reinforcing fibres are glass, carbon and aramid. The fibre reinforcement can take many different geometrical arrangements: chopped fibres, unidirectional fibres, weaves, braids, knits, three-dimensional preforms, etc. 201

202


This chapter discusses the fatigue behaviour of structural textile composites, where the textile yarns are embedded in a plastic matrix to enhance its mechanical properties. Most chapters so far have dealt with the fatigue behaviour of dry, non-impregnated textile yarns, while here, the textile fibre architecture has been impregnated with a resin and has been cured to a solid state. Application of these structural textile composites can be found in many industrial areas, ranging from the automotive industry, wind turbine and aerospace industry, to the sports industry (Drechsler, 2000; SAMPE, 2004; Haberkern, 2006).

10.2

Materials

In textile composites, the mechanical properties of the matrix material, the fibre material and the fibre/matrix interface have a major impact on the fatigue behaviour, together with the geometrical arrangement of the fibre reinforcement. These four aspects are briefly discussed in the next subsections.

10.2.1 Matrix materials In general, there are two large classes of plastic matrices for structural composites: (i) thermosetting matrices, and (ii) thermoplastic matrices. Most common thermosetting matrices are epoxy, unsaturated polyester, vinylester and polyurethane. These matrices are characterised by a small failure strain, a rather good stiffness and quite a low fracture toughness. The low fracture toughness can cause early cracking during fatigue loading, which is why a lot of research effort is concentrated on toughening the thermosetting matrices with thermoplastic fillers (Zhang, 2003). Most common thermoplastic matrices are polypropylene (PP), polyamide (PA), saturated polyester (PET, PBT) and the high-end polyphenylene sulphide (PPS), polyetherimide (PEI) and polyetheretherketone (PEEK). Compared to thermosetting plastics, thermoplastic resins in general have a larger failure strain, a lower stiffness and a higher fracture toughness. Their visco-elastic nature can lead to self-heating under fatigue loading at higher frequencies. Table 10.1 lists the mechanical properties of the most commonly used thermosetting and thermoplastic resins for (textile) composites (Coronet, 2007; Samyn, 2007; Gibbs, 1998; Hancox, 1983).

10.2.2 Fibre materials If the annual production in tonnes is compared, glass is the most important reinforcing fibre for structural composites (Stewart, 2002). Carbon is much

Fatigue damage in structural textile composites

203

Table 10.1 Mechanical properties of most common resins for (textile) composites Stiffness Strength Max. strain [GPa] [MPa] [%]

Matrix

Type

Density [kg/m3]

Unsaturated polyester (UP) Epoxy Thermosetting polyimide (PI) Polyamide (PA) Polypropylene (PP) Thermoplastic polyimide Polyphenylene sulphide (PPS)

thermoset

1200–1500 2–2.4

40–90

2

thermoset thermoset

1100–1400 3–6 1300 3.2

35–100 55

1–6 1.5

thermoplast 1140 thermoplast 900 thermoplast 1330

1.4–2.8 1–1.4 2.7

60–75 25–38 90–120

40–80 >300 90

thermoplast 1430

3.6

70–90

4

Table 10.2 Mechanical properties of the most common reinforcing fibres for polymer composites Fibre

Density [kg/m3]

Stiffness [GPa]

Strength [GPa]

Max. strain [%]

E-glass Carbon HS Carbon IM Carbon HM Carbon UHM Aramid K49 PE Sp. 900

2580 1800 1750 1800 1900 1450 970

69–72 160–250 276–317 338–436 440–827 131 117

3.5–3.8 1.4–4.9 2.35–7.0 1.9–5.5 1.8–3.5 3.6–4.1 2.6

4.5–4.9 0.8–1.9 0.8–2.2 0.5–1.4 0.4–0.5 2.8 3.5

more expensive, and will mainly be used if high stiffness and low weight or specific thermal properties are required. Aramid fibre (often referred to as Kevlar, DuPont’s trademark) is a speciality product for anti-ballistic applications and the aerospace industry, but is less widespread than glass and carbon. Other reinforcing fibres include high density polyethylene, polypropylene, polyester and nylon. Table 10.2 lists the mechanical properties of the most common reinforcing fibres for polymer composites (Herakovich, 1998; Verpoest, 2006).

10.2.3 Fibre/matrix interface The quality of the fibre/matrix interface is a very critical property for fatigue performance (Galiotis and Koimtzoglou, 2003; Zhifei et al., 2005; Gamstedt

204


et al., 2002). In general, the adhesion between fibre and matrix is better for thermosetting matrices than for thermoplastic matrices, because the high viscosity of the thermoplastic resins makes the impregnation of the fibre reinforcement more difficult (Hancox, 1983). Surface finishes are nearly always applied to fibres both to allow handling with minimum damage and to promote fibre/matrix interfacial bond strength. With carbon and aramid fibres for use in composite applications, the surface finish or size applied usually performs both functions. The finish is applied to the fibre at the point of fibre manufacture and this finish remains on the fibre throughout the conversion process into fabric. Glass fibre yarns, however, when used for weaving, are treated in two stages. The first finish is applied at the point of fibre manufacture at quite a high level and is purely for protection of the fibre against damage during handling and the weaving process itself. This protective finish, which is often starch based, is cleaned off or ‘scoured’ after the weaving process either by heat or with chemicals. The scoured woven fabric is then separately treated with a different matrix-compatible finish specifically designed to optimise fibre to resin interfacial characteristics such as bond strength, water resistance and optical clarity (SP Systems Guide, 2007).

10.2.4 Geometrical arrangements of fibre reinforcement The geometrical arrangement of the fibre reinforcement has a major effect on the resulting stiffness and strength of the composite. If the plastic matrix is reinforced with chopped fibres, the in-plane behaviour is still quasiisotropic and the enhancement of the mechanical properties is only moderate. On the other hand, if unidirectional fibre reinforcement is used, the largest increase in stiffness and strength is achieved, but the resulting composite can be highly anisotropic. Textile reinforcement is situated in between these two extremes, and depending on the chosen textile architecture, the flexibility in terms of desired stiffness and strength is very high. Textile reinforcement can take many forms: woven fabrics (plain weaves, unidirectional fabrics), braided fabrics, knitted fabrics, stitched multi-axial multi-ply fabrics (non-crimp fabrics), 3-D preforms and sandwich preforms (Naik, 2003; Lomov, 2006). Composite laminates are then made by stacking the fabric layers at different angles. The lamination scheme and material properties of individual laminae provide an added flexibility; designers can therefore tailor the stiffness and strength of the laminate to match the structural requirements. Owing to the waviness of the textile fibre architecture, crack growth is arrested more easily by the undulation of the fibres. This phenomenon is called ‘crack containment’. Nishikawa et al. (2006) recently showed that


205

this crack containment in fatigue can be modified by strongly changing the aspect ratio of the individual fibre tows. The use of very wide and thin ‘spread tows’ resulted in a longer fatigue life for woven carbon fabric epoxy laminates.

10.3

Fatigue testing methods

For fibre-reinforced composites, a variety of fatigue tests can be done, because the number of parameters is very large: (i) the amplitude control (stress or strain), (ii) the testing frequency, (iii) the loading direction (axial, bending, bi-axial), (iv) the load ratio (tension/tension, tension/compression, compression/compression). However only one of these tests has been standardised in both ASTM and ISO standards: that is the tension-tension fatigue test with a constant-amplitude load (EN ISO 13003:2003: Fibrereinforced plastics – Determination of fatigue properties under cyclic loading conditions and ASTM D3479/D3479M-96(2002)e1 Standard Test Method for Tension-Tension Fatigue of Polymer Matrix Composite Materials). The latest fatigue standard published is EN ISO 13003:2003. This standard gives general principles for fatigue testing that can be applied, with care, to all modes of testing. The general aspects are generally applicable to all testing modes and types of composite materials. It is recommended that, when available, the equivalent static test methods should be used (Harris, 2003). The number of parameters that affect the result of fatigue experiments is very large. They can be classified in two categories: parameters that are inherent to the material used, and parameters that are related with the fatigue loading conditions. Parameters inherent to the composite specimen •

•

The matrix can be a thermosetting or a thermoplastic matrix. Owing to hysteretic heating, the load frequency effect is much more pronounced in the case of thermoplastic matrices. Temperature rises up to 100 ºC are not unusual for higher test frequencies (5–10 Hz) (Xiao, 1999). The thermal conductivity of the matrix is also important for temperature rises in the material. When the conductivity is low, the hysteresis losses accumulate locally and the material degrades faster due to the hysteretic heating. The frequency effect is also very pronounced for elastomer matrices, which are used in nylon fibre-reinforced elastomers for the carcass of bias aircraft tires (Lee and Liu, 1994). The stacking sequence of the laminate has a very important effect on the fatigue behaviour of the composite specimen (Adali, 1985). Kim (1980) showed that the onset of delaminations can be delayed by changing the stacking sequence of graphite/epoxy specimens. For the

206


[0º/±45º/90º]s layup, delaminations initiated at N = 200 cycles, while for the [0º/90º/±45º]s stacking sequence, delamination onset was only observed at N = 20 000 cycles. Further, a dependency of the maximum crack density state on the onset of delamination was noticed. • The mechanical properties of the constituents (fibres and matrix) and the fibre volume fraction, as well as the bond strength between fibres and matrix, affect the fatigue performance. Adali (1985) has shown that the fibre volume content is an important variable for design against fatigue. • Defects during fabrication can be a cause of deteriorated fatigue performance. For example, defects can be due to draping of the fibre reinforcement. McBride and Chen (1997) studied the evolution of microstructure in dry plain-weave fabric during large shear deformation. For instance, in various resin-transfer moulding techniques, dry fabric is shaped around complex mould shapes. The fabric can be in a state of large in-plane shear deformation which significantly alters the spatial orientation of the constituent yarns and fibres. These deformations can lead to shear wrinkling in the fabric and reductions of structural integrity, due to varying fibre volume content. Lomov et al. (2000) have shown that each textile geometry model should take this effect into account. • Discontinuities, like cuts or holes, or previously initiated damage (due to impact damage (Jones et al., 1987), for instance) act like stress concentrators and can alter the fatigue behaviour of the composite specimen significantly. • Residual stresses can significantly affect the time to first damage. Bassam et al. (1998) have performed quasi-static tensile tests on cross-ply Eglass/epoxy specimens. They observed a residual strain on unloading which was due to matrix cracking. In their opinion, matrix cracks give rise to the relief of thermal curing stresses, and any damage which locally reduces the balanced macroscopic curing stresses can lead to a change in laminate dimensions. The residual strain could be reasonably well predicted based on shear-lag or variational mechanics. Parameters of the fatigue setup •

•

In composite materials, the inhomogeneity resulting from the fibre distribution may be so great as to influence the fatigue response of a sample, the size of which is comparable with the scale of the inhomogeneity. Thus, in woven-roving laminates, the width of a test piece should be sufficiently large to include several repeats of the weaving pattern (Harris, 1981). The choice of an appropriate specimen shape is strongly influenced by the characteristic nature of the composite being tested and has given


• •

207

rise to some difficulties. Early attempts to use test specimens similar to metallic designs led to unrepresentative modes of failure in composites that were (i) highly anisotropic and (ii) had relatively low in-plane shear resistance, such as unidirectionally-reinforced carbon fibre plastics containing high-modulus untreated carbon fibres. Much early work was carried out on waisted samples of conventional kind, and designs were adapted for unidirectional carbon fibrereinforced composites by using long samples with very large radii and relatively little change in cross-section in the gauge length. An alternative approach that has been used to prevent the familiar shear splitting in composites of low shear resistance was to reduce the sample thickness rather than the width. For tensile fatigue testing, this has now been almost universally superceded by the technique of using parallel-sided strips with end-tabs, either of glass fibre-reinforced polymer or soft aluminium, bonded onto the samples for gripping. Carefully done, this eliminates the risk of grip-damage (and resultant premature failure) without introducing significant stress concentrations at the ends of the test length, and failures can usually be expected to occur in the test section (Harris, 1981). However, for tension-compression and compression-compression fatigue testing, various sorts of tabs are used (see, for example, Gathercole et al., 1994; Walsh and Pipes, 1982; Ramkumar, 1982). The nature of the fatigue stress (tension/compression/shear). σ min The stress ratio R = is a very important parameter in fatigue σ max testing. The stress levels σmin and σmax are evaluated with their algebraic sign, so a negative stress ratio (−∞ < R < 0) refers to tension-compression loading, while tension-tension loading comprises the range 0 < R < 1 and compression-compression loading comprises the range 1 < R < +∞ (see Fig. 10.1). The remaining cases are zero-tension loading (R = 0) and zero-compression loading (R = −∞). For most composite materials, the worst fatigue loading condition is fully reversed axial fatigue, or tension-compression loading (R = −1) (Curtis, 1989; Brocker and Woithe, 1991). Indeed, in compression, although the fibres remain the principal load-bearing elements, they must be prevented from becoming locally unstable and undergoing a micro-buckling type of failure. This is the task of the matrix and the fibre/matrix interface, the integrity of both being of far greater importance in compressive loading than in tensile loading. Under tensile fatigue loading, many of the laminate plies, without fibres in the test direction, develop intraply damage and this causes local layer delamination at relatively short lifetimes. In compression, this tensile-induced damage can lead to local layer instability and layer buckling, perhaps

208

Fatigue failure of textile fibres Ro Compression-tension

–1.0 Stress amplitude

–∞

Ro

Tension-compression –0.33

–3.0

0.0

∞

0.5Ro

Tensiontension

Compressioncompression

0.33

3.0

Rs = 1.0 –Ro

–0.5Ro

0.5Ro

Rs = 1.0 0 Mean stress

0.5Ro

Ro

10.1 Constant-life diagram showing lines of constant stress ratio R (designated in the graph as Rs) (Schaff and Davidson, 1997b).

• •

•

•

•

before resin and interfacial damage within the layers has initiated fibre micro-buckling. Thus fatigue lives in reversed axial loading are usually shorter than for zero-compression or zero-tension loading. The fatigue experiments can be load-controlled or strain-controlled. When the fatigue stress amplitude is low compared to the static strength, it is called ‘high-cycle fatigue’, because the specimen will sustain the load during a large number of cycles. While, on the other hand, when the fatigue stress amplitude is a large percentage of the static strength, it is called ‘low-cycle fatigue’. Damage types can be quite different, since static failure mechanisms are often involved in low-cycle fatigue, while typical fatigue damage mechanisms rather occur in high-cycle fatigue (Harik et al., 2000). The frequency is a very important parameter. Ellyin and Kujawski (1992) clearly showed the influence of frequency in fatigue tests on Eglass/epoxy specimens. At lower stress amplitudes they found that the cyclic creep increased and the fatigue life decreased with reduced cyclic frequency. In contrast, at higher stress amplitudes, the effect of loading frequency on both cyclic creep and fatigue life was opposite to that at lower stresses, in that the fatigue life decreased with higher frequency. Environmental conditions can play an important role. Bach (1996) has proved that moisture has a significant effect on the fatigue performance of glass fibre-reinforced plastics. The in-service fatigue loadings as they act on a real composite structure, are rarely reproduced in laboratory tests, mainly for two reasons: (i) most fatigue testing machines are not equipped for such complex


209

loading conditions, and (ii) the in-service fatigue loadings are often known insufficiently. Therefore, the in-service fatigue loadings are replaced by representative, standardised load spectra for fatigue testing purposes. Typical examples are: (i) the WISPER spectrum for wind loads on the rotor blades of wind turbines (WInd SPEctrum Reference) (ten Have, 1991; Brøndsted et al., 1997; Bond, 1999), and (ii) the FALSTAFF spectrum for the load on the wing root area of fighter aircraft (Fighter Aircraft Load STAndard For Fatigue evaluation) (Schultz, 1981; Schaff and Davidson, 1997a, 1997b). Another option is the use of block loading tests, where loading blocks with high load amplitude and low load amplitude are applied in a different order, to assess the effect of low-high and high-low load transitions on the fatigue life and the damage evolution. Van Paepegem and Degrieck (2002b) have shown that there is no consensus in international literature on which sequence is the most severe: a low-high load sequence or a high-low load sequence. It strongly depends on the type of material, stacking sequence and amplitude levels. Recent studies in block loading have been reported by Found and Quaresimin (2003), Bourchak et al. (2007) and Epaarachchi (2006). Naturally, the number of these variable amplitude tests should be limited due to their expensive and time-consuming nature. Therefore the fatigue life of the composite component under variable amplitude loads is often estimated, based on the fatigue testing results under constant amplitude loading. In the following paragraphs, the main fatigue testing methods are discussed: • • • • •

tension-tension fatigue tension-compression and compression-compression fatigue bending fatigue shear dominated fatigue multiaxial fatigue.

Attention will not only be given to the test procedure, but also to the instrumentation methods. The importance of online monitoring techniques cannot be stressed enough, because fatigue damage leads to measurable degradation of macroscopic (elastic) properties in almost all types of (textile) composites.

10.3.1 Tension-tension fatigue The uni-axial tension-tension fatigue test is the most widely used fatigue test. The coupon geometry is a parallel-sided specimen, instrumented with

210


tabs. The choice of the tabbing material differs among the testing laboratories. Some prefer steel or aluminium tabs, but most of them use glass/epoxy tabs, where the glass reinforcement has a [+45º/−45º]ns stacking sequence. In most cases, the tabs are straight-sided non-tapered tabs. A fatigue test is usually conducted with a servo-hydraulic testing machine, equipped with grips that clamp the specimen. The alignment of the specimen is very important. No bending loads must be induced in the specimen due to misalignment. The load is recorded by the load cell, while the extensometer records the axial strain. The grip displacement is recorded as well, but is not very useful. The transverse strain can be measured by a strain gauge or a bi-axial extensometer. A thermocouple can be read out to monitor the surface temperature of the composite. The test frequency is always chosen as high as possible to limit the duration of the test and minimise the cost, but the fatigue response of some composites strongly depends on the frequency (especially in the case of fibre-reinforced thermoplastics). In tension-tension fatigue tests, the stress ratio R is often chosen to be 0.1. Nevertheless the stress ratio (or the mean stress) also has a clear effect on the damage growth rate in tension-tension fatigue. Many authors have shown that if the maximum stress is kept constant, the damage growth is reduced for increasing mean strength. Wevers et al. (1987, 1990) studied the fatigue damage development in carbon/epoxy laminates. They reported that the number of matrix cracks in the cross-ply laminates was considerably smaller if the stress ratio for tension-tension fatigue was increased from R = 0.03 to R = 0.5. According to Wevers et al. (1987, 1990), this phenomenon could be due to crack closure. During the growth of 90º cracks, material debris is formed between the crack faces. When the crack is closing down, this excess material causes: (i) compressive forces in the 90º plies, for perpendicular cracks, or (ii) sliding forces, for inclined matrix cracks. For the fatigue tests with increased stress ratio, the cracks will stay open and no additional cracks can be formed at the low stress level. Several authors have reported similar results. Mallick (1997) presented representative fatigue data, in the form of a Goodman diagram, for several unidirectional composites at 107 cycles. If the stress ratio R is increased, the maximum stress can be larger for the same fatigue life. Beaumont (1987) studied the damage behaviour of quasi-isotropic carbon/epoxy laminates. He did not compare the damage growth rates for constant maximum stress σmax, but for constant stress amplitude Δσ. When recalculating his results it can be clearly seen that when the stress ratio R is increasing for constant maximum stress, the damage growth rate is decreasing, in agreement with the reported results by other researchers. In the international standards, the number of cycles to failure is considered as the main outcome of the tension-tension fatigue test. Yet it is worth


211

the effort to use online instrumentation methods. The most simple and effective online measurement is the axial stiffness evolution. The axial stiffness can be directly calculated from the axial stress (loadcell) and the axial strain (extensometer). The axial strain must never be calculated from the axial displacement and the gauge length, because the inevitable slip in the clamps can lead to serious errors in the strain calculation. Depending on the fibre and matrix type and the stacking sequence, the stiffness degradation can range from a few percent to several tens of percent (Hashin, 1985; Whitworth, 2000; Highsmith and Reifsnider, 1982; Yang et al., 1990; 1992; Kedward and Beaumont, 1992). If the transverse strain is measured, the Poisson’s ratio νxy can be followed as well. It has been recently showed by Van Paepegem et al. (2007) that the evolution of the Poisson’s ratio is a very sensitive parameter for fatigue damage. Figure 10.2 shows the evolution of the Poisson’s ratio for a unidirectional glass fabric/epoxy composite in tension-tension fatigue. The νxy − εxx curves in strain-controlled fatigue between 0.0006 (0.06%) and 0.006 (0.6%) show a highly nonlinear behaviour and are upper-bounded by the static degradation of the Poisson’s ratio. Another simple measurement is surface temperature evolution. Neubert et al. (1987) demonstrated a fairly good agreement between temperature change and stiffness degradation for carbon/epoxy laminates. Quaresimin

uxy versus exx for [0°/90°]2s fatigue test W_090_8 static [0°/90°]2s test IF4 static [0°/90°]2s test IF6 [0°/90°]2s fatigue test W_090_8: cycle 600 + 5 [0°/90°]2s fatigue test W_090_8: cycle 3600 + 5 [0°/90°]2s fatigue test W_090_8: cycle 37200 + 5

0.20 0.15 0.10

uxy [–]

0.05 –0.00 0.000

0.005

0.010

0.015

0.020

–0.05 –0.10 –0.15 –0.20

exx [–]

10.2 Evolution of Poisson’s ratio for a unidirectional glass fabric/epoxy composite in tension-tension fatigue (Van Paepegem et al., 2007).

212


(2002) showed that the high cycle fatigue strength of woven carbon/epoxy composites can be estimated, based on the infra-red thermal analysis of the specimen temperature. One of the popular online monitoring techniques is acoustic emission. The acoustic emission technique basically involves the detection of low intensity stress waves generated by failure events as they occur, such as delamination and fibre fracture. The stress waves are formed from the strain energy released during a failure event, and propagate to the surfaces where they are detected by acoustic sensors (Mouritz, 2003). The reliability of the acoustic emission technique to accurately monitor the progression of fatigue damage has been proven in numerous studies (for review see Wevers, 1987; Mouritz, 2003). Very recently, Bourchak et al. (2007) used acoustic emission energy as a fatigue damage parameter for unidirectional and woven fabric carbon composites. They suggested that the arbitrary choice of fatigue stress levels at a high percentage of the static ultimate tensile strength should be re-considered, given the significant acoustic emission energy at low stress levels in the first cycles of fatigue loading. Another online technique is the use of embedded optical fibre sensors with a Bragg grating. The Bragg grating is a periodical variation of the optical refractive index that is written in the core of the glass fibre and is typically a few millimetres in length. When broadband light is transmitted into the optical fibre, the Bragg grating acts as a wavelength selective mirror. For each grating only one wavelength, the Bragg wavelength, λB is reflected with a Full Width at Half Maximum of typically 100 pm, while all other wavelengths are transmitted. The Bragg wavelength is directly proportional with the period of the Bragg grating. If the optical fibre sensor is embedded in a composite laminate, the strain in the loaded laminate will cause the period of the Bragg grating to change, and hence the value of the reflected Bragg wavelength. The advantages are numerous: • •

the measurement is absolute and does not drift in time fibre optic sensors are rugged passive components resulting in a high lifetime (>20 years) and are insensitive to electromagnetic interference • the fibre Bragg grating forms an intrinsic part of the optical fibre and has very small dimensions which makes it very suitable for embedding in composite plates • many fibre Bragg gratings can be multiplexed employing only one optical line so more sensing points can be read out at the same time. Doyle et al. (1998) experimented on the use of fibre optic sensors for tracking the cure reaction of a fibre-reinforced epoxy, with success. They also successfully demonstrated the feasibility of these sensors for monitoring the stiffness reduction due to fatigue damage, for thermosetting matrix.


213

De Baere et al. (2007b) have shown that the optical fibre sensors also survive the production process for carbon fabric thermoplastics (both autoclave and compression moulding) and that the correspondence between the axial strain measurements from the extensometer and the optical fibre sensor were identical in tension-tension fatigue tests. That means that the adhesion of the embedded optical fibre sensor to the surrounding thermoplastic material is very good. Resistance measurement is a well-established damage detection technique for unidirectional carbon composites (Abry et al., 1999). For a long time, there has been disagreement between researchers whether the resistance should increase or decrease when local fibre fractures occur (Angelidis et al., 2004; Chung and Wang, 2006; Angelidis et al., 2006). In a recent series of articles, it has been clearly demonstrated that the resistance must increase with increasing damage, but a lot of researchers observe a decrease of resistance, due to bad contact of the electrodes. Recently, De Baere et al. (2007a) showed that resistance measurement also works very well for monitoring damage in carbon fabric reinforced thermoplastics under tension-tension fatigue loading. Figure 10.3 shows the evolution of relative resistance change ρ and axial fatigue stress σxx during fatigue cycles 4025 to 4030.

Relative resistance change ρ [–]

0.020

ρ

σxx

600

0.018

540

0.016

480

0.014

420

0.012

360

0.010

300

0.008

240

0.006

180

0.004

120

0.002

60

0.000 0.0

0.2

0.4

0.6

0.8

Stress σxx [MPa]

Relative resistance change ρ as a function of time plotted over the evolution of corresponding the stress

0 1.0

Time [s]

10.3 Evolution of relative resistance change ρ and axial fatigue stress σxx during fatigue cycles 4025 to 4030 in a 5-harness satin weave carbon/PPS laminate (De Baere et al., 2007a).

214


10.3.2 Tension-compression and compression-compression fatigue In general the damage growth rate in compression is smaller, when two restrictions are made: (i) there are no delaminations, (ii) the stress ratio R is not negative (Mallick, 1997). On the other hand, it is well known that the stress ratio R = −1 is very detrimental for stacking sequences which can develop delaminations. It has often been reported in literature that this sign reversal of the applied stress has a detrimental effect on fatigue life (Bartley-Cho et al., 1998; Gathercole et al., 1994; Adam et al., 1994; Badaliance et al., 1982; Gamstedt and Sjogren, 1999). This is because the plies, with and without fibres in the loading direction, develop intraply damage, and this causes local delamination at relatively short lifetimes. In tensile loading this is less serious, as the plies containing fibres aligned with the loading direction, continue to support the majority of the applied load. In compression, however, tensile-induced damage can lead to local instability and buckling, perhaps before resin and interfacial damage within the plies initiate fibre microbuckling. Thus fatigue lives in reversed tension-compression loading are usually shorter than for tension-tension loading. In tension-compression and compression-compression fatigue, the alignment of the specimen is very crucial. Bending induced by misalignment can cause premature failure of the specimen. ASTM E1012 is the only standard which specifically addresses the procedure for measurement of misalignment induced bending (ASTM E1012-05 Standard Practice for Verification of Test Frame and Specimen Alignment Under Tensile and Compressive Axial Force Application). A more extensive description of this procedure can be found in the ‘Code of practice for the measurement of misalignment induced bending in uniaxially loaded tension-compression test pieces’, published by the Institute for Advanced Materials of the European Commission (Bressers, 1995). Further, buckling of the specimen must be avoided. There are two options: (i) decrease the unsupported gauge length, or (ii) use anti-buckling guides for longer gauge lengths. Matondang and Schutz (1984) studied the influence of the anti-buckling guide design on the compression fatigue behaviour of carbon fibre-reinforced composites. Quaresimin (2002) showed that the static compressive strength of woven carbon composites increases as the unsupported length decreases and decided to carry out the tension-compression fatigue tests on short unsupported specimens. Gagel et al. (2006a) studied the tension-compression fatigue behaviour of E-glass multi-axial non-crimp fabric/epoxy laminates with the use of an anti-buckling guide and a PTFE coated paper to minimise friction.


215

10.3.3 Bending fatigue Uni-axial fatigue experiments in tension/compression are most often used in fatigue research (Fujii et al., 1993; Schulte et al., 1987; Hansen, 1997) and accepted as a standard fatigue test, while bending fatigue experiments are scarcely used to study the fatigue behaviour of composites (Ferry et al., 1997; Herrington and Doucet, 1992; Chen and Matthews, 1993b). Bending fatigue tests differ in several aspects: •

•

•

•

The bending moment is (piecewise) linear along the length of the specimen (3-point bending, 4-point bending, cantilever beam bending). Hence stresses, strains and damage distribution vary along the gauge length of the specimen. On the contrary, with tension/compression fatigue experiments, the stresses, strains and damage are assumed to be equal in each cross-section of the specimen. Owing to the continuous stress redistribution, the neutral fibre (as defined in the classic beam theory) is moving in the cross-section because of changing damage distributions. Once a small area inside the composite material has moved, for example from the compressive side to the tensile side, the damage behaviour of that area is altered considerably. The finite element implementation of related damage models gives rise to several complications, because each material point is loaded with a different stress, strain and possibly stress ratio, so that damage growth can be different for each material point. In tension/compression fatigue tests, the stress- or strain-amplitude is constant during fatigue life and differential equations describing decrease of stiffness or strength, can often be simply integrated over the considered number of loading cycles. Smaller forces and larger displacements in bending allow a more slender design of the fatigue testing facility.

Basically, three types of bending fatigue tests can be distinguished: (i) three-point bending (Sidoroff and Subagio, 1987; El Mahi et al., 2002), (ii) four-point bending (Caprino and D’Amore, 1998), and (iii) cantilever bending (Herrington and Doucet, 1992; Van Paepegem and Degrieck, 2001; 2002a; 2002b; 2002c; 2005). The success of these tests for fatigue of (textile) composites is quite limited, because the interpretation of the results is more difficult and in case of stiffness degradation, stress redistribution across the specimen height comes into play. Moreover, as long as the bending stiffness of the laminate is high enough (e.g., sandwich composites), the deflections are small and linear beam theory still applies, but once that the bending stiffness of the composite decreases (e.g., thin laminates), the deflections are large and geometric nonlinearities and friction at the roller supports affect the fatigue results. Van Paepegem et al. (2006c) have shown that in

216


three-point bending fatigue of woven carbon thermoplastics, the friction at the supporting rolls (due to large deflections) results in a ‘hysteresis-like’ behaviour of the stress-strain curve, which would normally be attributed to damage, but is entirely due to friction. The majority of flexural fatigue tests are performed on sandwich composites (El Mahi et al., 2004; Abbadi et al., 2006; Judawisastra et al., 1997; Farooq et al., 2002). Only a limited number of papers have been published on flexural fatigue testing of woven fabric composites. Miyano et al. studied the effect of loading rate and temperature on the flexural fatigue behaviour of a satin woven carbon/epoxy laminate (Miyano et al., 1994). A test set-up in three-point bending was used. They showed that the flexural fatigue strength of the laminate is affected by temperature, even at temperatures far below the glass transition temperature. Effect of sea water on the bending fatigue behaviour of glass/polyester composites with chopped strand mats and woven fabrics was investigated by Hasan et al. (1998). Caprino and Giorleo (1999) investigated the fatigue behaviour of plain weave glass/epoxy composites in four-point bending. A statistical fatigue model was presented based on the hypothesis of a two-parameter Weibull distribution of the static strength. Mahfuz et al. (2000, 2001) described the fatigue modulus degradation for a thick-section plain weave S2-glass fabric/ vinylester composite loaded in three-point bending fatigue. Van Paepegem and Degrieck (2001, 2002a, 2002b, 2005) investigated the fatigue behaviour of plain weave glass/epoxy composites in cantilever bending. A phenomenological damage model was developed for fatigue loading in warp/weft and bias direction.

10.3.4 Shear dominated fatigue Fatigue testing in pure shear is very difficult. Lessard et al. (1995) modified the static three-rail shear test (ASTM D 4255/D 4255M – 01) to do fatigue testing on carbon/epoxy plates. Much more common methods are the tension-tension fatigue tests on a [+45º/−45º]ns laminate, This test is based on the ASTM D3518/D3518M-94(2001) Standard Test Method for In-Plane Shear Response of Polymer Matrix Composite Materials by Tensile Test of a ±45º Laminate. This standard explains how the shear stress-strain curve can be derived from a static tensile test on a ±45º laminate, by measuring the longitudinal and transverse strain. The test is also called a bias tension test, because the bias (or cross-grain) direction is the 45º direction between warp and weft direction in case of fabric reinforced composites. In both pure shear and shear-dominated fatigue, the test frequency is a very important parameter. The shear stresses can lead to significant autogeneous heating and once the temperature exceeds the glass transition temperature, the deformations can be very large.


217

Shear stress-strain curve for cyclic [+45°/–45°]2s test IH2 60

Shear stress τ12 [MPa]

50 40 30 20 IH2 cyclic test IH4 static test

10 0 0.00

0.01

0.02

0.03 0.04 0.05 Shear strain γ12 [–]

0.06

0.07

10.4 Accumulation of permanent shear strain in cyclic loading of unidirectional glass fabric/epoxy composites (Van Paepegem et al., 2006a, 2006b).

A hardly studied phenomenon is the accumulation of permanent strain during shear-dominated fatigue loading. For composite materials with a thermoplastic matrix, creep effects seem to be dominant, while in case of thermosetting materials, permanent strain is simply neglected in most reported literature. Moreover, for both types of material, the phenomenon is not well understood. Van Paepegem et al. (2006a, 2006b) studied the accumulation of permanent shear strain in [+45º/−45º]2s glass/epoxy laminates under cyclic loading. They showed that the shear modulus significantly degrades, but that the accumulation of permanent shear strain is even more important. Figure 10.4 shows the accumulation of permanent shear strain in cyclic loading of unidirectional glass fabric/epoxy composites.

10.3.5 Multiaxial fatigue As opposed to metals, where there exists an extensive amount of research on biaxial/multiaxial fatigue, research in the same field on composite materials is far less complete. Literature reviews on multiaxial/biaxial fatigue of composites can be found in Shokrieh and Lessard, 2003; Chen and Matthews, 1993a; Quaresimin and Susmel, 2002; Philippidis and Vassilopoulos, 1999. The four main types of multiaxial fatigue set-ups described in literature are: (i) tension/torsion, (ii) internal pressure/tension, (iii) planar biaxial setups, and (iv) bending/torsion.

218


•

In case of tension/torsion set-ups (Lee and Hwang, 2001; Fujii et al., 1992; Inoue et al., 2000), composite tubes are clamped in a servo-hydraulic machine with two actuators. A combination of uniaxial load (tension/ compression) and torsional moment is applied to the composite tube and realises a multiaxial stress state in the tube. The main problem with these tests is the clamping of the tubes. The design of the tube ends must be such that the effect of the stress concentration at the grips is minimised and does not cause premature fatigue failure of the tube. • Internal pressure/tension (Ellyin and Martens, 2001; Perreux et al., 2005; Perreux and Joseph, 1997) set-ups realise a combination of axial stress and hoop stress in composite tubes. The ASTM D2992 Standard practice for obtaining hydrostatic or pressure design basis for fiberglass (GlassFiber-Reinforced Thermosetting-Resin) pipe and fittings can be used as a guideline. • For multiaxial fatigue testing on flat specimens, the planar biaxial set-ups are most commonly used. The specimen typically has a cruciform shape, but the precise geometry of the specimen is extremely important to generate the most critical stress/strain state in the centre of the specimen, and not in the loading arms (Smits et al., 2006). Chen and Matthews (1993b) clamped all edges of flat rectangular composite plates and applied fatigue load by a central indenter to obtain biaxial bending fatigue. • Bending/torsion set-ups can be used to apply three- or four-point bending and torsion simultaneously to flat rectangular specimens. More information can be found in the papers by Ferry et al. (1997, 1999). Only a few of the cited investigations deal with textile composites. Fujii et al. (1992) and Inoue et al. (2000) reported the results of tension/torsion biaxial cyclic loading on plain weave glass/polyester tubes. They found that the modulus decay in shear is affected by the loading path, while the modulus decay in tension is not. Also fibre misalignment and fibre reorientation during testing appeared to have a substantial influence on the results.

10.4

Typical fatigue damage in structural textile composites

10.4.1 Inspection techniques for visualisation of fatigue damage The easiest inspection technique is visual inspection. Depending on the difference in optical refraction index of the matrix and fibre materials, the transparency of the composite laminate can be very high. Gagel et al.


219

10.5 Microscopic image of a plain weave glass/epoxy composite loaded in bending fatigue (Van Paepegem, 2002).

(2006a,b) reported an extraordinary high transparency of E-glass multiaxial non-crimp fabric epoxy laminates. Matrix cracks, voids and inclusions could be detected easily by transmitted light. Optical or light microscopy provides a direct path from observations made with the naked eye, to what is visible at magnifications up to about 1000× (Hull, 1999). Fracture surfaces are embedded in resin and polished before observation. Figure 10.5 shows a microscopic image of the damage in a plain weave glass/epoxy composites loaded in bending fatigue (Van Paepegem, 2002). Scanning Electron Microscopy (SEM) is by far the most popular technique for fractographic studies. The fracture surface is scanned, on a rectangular or square raster, by a finely focused beam of high-energy electrons (typically 5–40 keV). The electrons penetrate the surface of the specimen and interact with the atoms of the material in a variety of elastic and inelastic scattering processes. By selectively collecting and measuring the scattered signals, information is obtained about the surface and near-surface properties and characteristics (Hull, 1999). Typically, for conventional modern instruments, the maximum resolution for SEM images is a few nanometres and the range of possible magnifications is very wide, from about 5× to 20 000×. The depth of field is much greater than the corresponding values for light microscopy (Hull, 1999). Edge replication of composite specimens has been used with considerable success in the study of ply cracking and edge delamination (Masters and Reifsnider, 1982; Wevers, 1987). The specimen edge is polished and a plastic replica of the surface is taken. This is achieved either by painting a film of plastic, such as cellulose acetate dissolved in acetone, on the surface, or pressing a thin sheet of a gel of the cellulose acetate onto the surface, so that it completely wets the surface and follows the intricate contours. When the plastic has fully dried, it is removed from the fracture surface by peeling and forms a negative replica of the original surface. This replica of the edge

220


can then be studied by a variety of microscopic techniques: light microscopy, SEM and TEM (Hull, 1999). A very common inspection technique for fatigue damage in (textile) composites is ultrasonics. Ultrasonics can be performed in various modes of operation, but the most common for fatigue damage detection is the through-transmission (C-scan) technique. Through-transmission ultrasonics basically consists of a transducer for emitting ultrasonic pulses that is placed at or near one surface and a receiver sensor that is located at the opposite surface. The technique applies to relatively low frequency sound beams, typically 0.5–15 MHz, having a small aperture. The transducer and receiver are coupled to the surfaces or they are immersed in water together with the composite. The ultrasound waves are attenuated by defects in the composite and the acoustic attenuation is monitored using the receiver (Mouritz, 2003). In classical C-scans, the size of the applied bounded beam and the applied wavelength are too big to characterise microscopic defects with high spatial resolution. Contrary to low frequency ultrasound, high frequency ultrasound, as used in acoustic microscopy, can detect such defects with very high in-depth resolution, if a high frequency focused beam with relatively low aperture is used (Declercq, 2005). Another well-known inspection technique is radiography. During passage of the radiation through the material, the rate of energy absorption can be changed by defects that have a different absorption coefficient to the parent material. Certain types of defects in composites are not easily detected using standard X-ray radiography because of poor contrast on the radiograph. In order to enhance the contrast between defects and the parent material, a variety of X-ray-sensitive penetrants can be used (Mouritz, 2003). High-resolution 3D X-ray micro-tomography or micro-CT is a relatively new technique which allows scientists to investigate the internal structure of their samples without actually opening or cutting them (Cnudde et al., 2006). Without any form of sample preparation, 3D computer models of the sample and its internal features can be produced with this technique. In order to perform tomography, digital radiographs of the sample are made from different orientations by rotating the sample along the scan axis from 0 to 360 degrees. After collecting all the projection data, the reconstruction process is producing 2D horizontal cross-sections of the scanned sample. These 2D images can then be rendered into 3D models, which enable a virtual view of the object. Figure 10.6 shows the micro-tomography images of plain woven glass/ epoxy composite (left) and damaged 5-harness satin weave carbon/PPS (right). Another technique is thermography, which registers the infra-red radiation and estimates the surface temperature of the specimen. Hansen (1999)


221

10.6 Micro-tomography images of plain woven glass/epoxy composite (left) and damaged 5-harness satin weave carbon/PPS (right) (De Baere et al., 2007b).

successfully applied this method to the tension-tension fatigue of woven glass fabric/epoxy composites. Toubal et al. (2006) used infra-red thermography for their study of [±45º]2s woven carbon fabric/epoxy composites in tension-tension fatigue loading. They observed temperature increases up to 80 ºC, which can be attributed to the very high testing frequency of 10 Hz. Gagel et al. (2006b) observed similar temperature rises for E-glass multi-axial non-crimp fabric/epoxy composites in tension-tension fatigue at 6 Hz. Other inspection techniques that have not been mentioned so far, are the leaky Lamb wave technique (Chimenti and Nayfeh, 1985; Seale et al., 1993), acoustic polar scans (Declercq, 2005; Maes, 1998), acoustography, vibrothermography or SPATE (Stress Patterns Analysis by the measurement of Thermal Emissions) and Moiré interferometry (Mouritz, 2003). These techniques are rarely used to study fatigue damage in composites.

10.4.2 Typical fatigue damage mechanisms Fibre-reinforced polymer composites have a quite good rating as regards life time in fatigue. The same does not apply to the number of cycles to initial damage. Although the fatigue behaviour of fibre-reinforced polymer composites has been studied for many years, it is so diverse and complex that present knowledge is far from complete. There are a number of important differences between the fatigue behaviour of metals and of continuous fibre-reinforced polymer composites. In metals, the stage of gradual and invisible deterioration takes a relatively large part of the total life. No significant reduction of stiffness is observed in metals during the fatigue process. The final stage of the process starts with the formation of small cracks, which are the only form of macroscopically observable damage. Gradual growth and coalescence of these cracks quickly produce a large crack and final failure of the structural component.

222


Continuous fibre-reinforced polymers are made of long, reinforcing fibres embedded in a polymer matrix. This makes them heterogeneous and anisotropic. The first stage of deterioration by fatigue is observable by the formation of ‘damage zones’, which contain a multitude of microscopic cracks and other forms of damage, such as fibre/matrix interface debonding and pull-out of fibres from the matrix. It is important to observe that damage starts very early, after only a few or a few hundred loading cycles. This early damage is followed by a second stage of very gradual degradation of the material, characterised by a gradual reduction of the stiffness. More serious types of damage appear in the third stage, such as fibre breakage and unstable delamination growth, leading to an accelerated decline and finally catastrophic failure. This three-stage stiffness reduction was first reported by Schulte et al. (1985, 1987) and Schulte (1984) but has since then been observed for many different types of composite materials, also for textile composites. Fujii et al. (1993) have performed fatigue tests on plain woven glass fibrereinforced composites with a polyester matrix. The tests were performed in uni-axial zero-tension loading along the warp direction. They reported a modulus degradation which was very similar to the one described by Schulte (1984). According to the experiments, the weft tows first debond from the matrix (see Fig. 10.7). Simultaneously matrix cracks occur in the resin-rich regions between two woven sheets. In a second stage, the debonds propa-

Debonding

Stage 1 Meta delamination

Stage 2

Stage 3

10.7 Microscopic fatigue processes in a plain woven glass/epoxy laminate (Fujii et al., 1993).


223

gate in the weft along the fibre and connect with other debonds that are present near the adjacent cross-over points. Small delaminations (so-called meta-delaminations) occur between warp and weft fibre bundles near the fabric cross-over points. Finally fibre fracture occurs. These observations have been confirmed by Pandita et al. (2001). Tensiontension fatigue tests were performed on plain woven glass/epoxy specimens and damage development was monitored by means of acoustic emission and SEM inspection (Scanning Electron Microscopy). Later on, Pandita and Verpoest (2004) conducted similar experiments on E-glass knitted fabric epoxy laminates. There the matrix cracks and yarn-matrix debonds initiated from the part of the knitting loop that was perpendicular to the loading direction. The fatigue damage then propagated following the knitting structure. Gagel et al. (2006b) described the damage evolution in E-glass multi-axial non-crimp fabric epoxy laminates under tension-tension fatigue loading. The gradual development of matrix cracks in the 45º and 90º directions was observed with an optical scanner in transmitted light mode. The authors did not report whether or not the polyester stitching yarn affected the fatigue damage onset and propagation. Quaresimin (2002) reported large-scale inter-layer delaminations in tension-compression fatigue of twill carbon fabric epoxy laminates, where a considerable amount of 45º layers were present in the stacking sequence. Later on, Found and Quaresimin (2003) reported similar observations for 5-harness satin weave carbon epoxy laminates in tension-tension fatigue. If textile composites with a central hole are tested in fatigue, the hole acts as a stress concentrator and damage concentrates around the hole. Recent investigations for woven carbon fabric laminates are discussed by Toubal et al. (2006), Hochard et al. (2006) and Bourchak et al. (2007).

10.5

Modelling strategies for fatigue damage in textile composites

A good deal of the work that has been done on the investigation of the fatigue of composites reflects the much more extensive body of knowledge relating to the fatigue of metallic materials. And this is not unreasonable since the established methods of accumulating and analysing metallic fatigue data provided a reliable means of describing fatigue phenomena and designing for fatigue. The danger was, and is, in making the assumption that the underlying mechanisms of material behaviour that give rise to the stress/life (σ/log Nf) curve are the same in metals and composites. Indeed the study of fatigue of composites is complicated by the fact that most fatigue phenomena which are well-known for metals, cannot be simply transposed to their ‘composite’ equivalent:

224


•

Most fibre-reinforced composites are strongly orthotropic or transversely isotropic, as opposed to the isotropic metals. • The fracture mechanism is not defined by one single crack, but by various damage mechanisms that can interact with each other. • The finite element modelling of multi-layered composites or composites reinforced with textile structures is not straightforward. Calculating the correct stress states often requires advanced finite element analysis. • As opposed to metals, fibre-reinforced composites show a typical degradation of stiffness and strength during fatigue life.

Before the development of fracture mechanics and its use in treating metallic fatigue as a crack-growth problem, the only available design information on fatigue behaviour was the stress/life, or S/N curve. It represented directly the perceived nature of fatigue in terms of experimental results, but gave no indication of the mechanisms of fatigue damage, of the presence or behaviour of cracks, or of changes in the characteristics of the material as a consequence of the fatigue process. Another presentation is the constant-life diagram, where the expected life is shown for a given combination of the alternating component of the stress σ − σ min σ + σmin and the mean stress σm = max . This automatically σ alt = max 2 2 σ introduces the concept of the stress ratio R = min , and the question of the σmax relative importance of any compression component of stress. In metallic fatigue it was frequently assumed, however, that compression stresses were of no significance because they acted only to close the fatigue cracks, unlike tensile forces. Master diagrams of this kind are presented in a variety of forms, all more or less equivalent, but the most familiar is that which is usually referred as the Goodman diagram. The abscissa contains the normalised mean stress m, while the ordinate axis shows the normalised alternating stress a. According to Fong (1982), there are two technical reasons why fatigue damage modelling in general is so difficult and expensive. The first reason is that there are several scales where damage mechanisms are present: from micro-scale (fibres and matrix), through the meso-scale (individual lamina), to the component and structural levels. The second reason is the impossibility of producing ‘identical’ specimens with well-characterised microstructural features. Fong also draws attention to some pitfalls of fatigue damage modelling: • •

Confusion over scale: information from measurements on different scale levels is combined improperly and leads to erroneous results. False generalisation: for example, stiffness reduction can often be divided in three regimes (sharp initial reduction – more gradual decrease – final


225

failure (Schulte et al., 1985; Daniel and Charewicz, 1986)), but the related models are not always valid in the three stages. • Oversimplification: curve fitting of experimental data is done by using oversimplified expressions. This last statement was confirmed by Barnard et al. (1985). He presented evidence that much of the scatter of the S-N curve drawn from his experimental data was caused by a change in failure mode, generating a discontinuity in the S-N curve. Indeed a student’s t-distribution indicated that his test data were falling apart in two distinct and statistically significant populations. The remaining scatter was a consequence of static strength variations. Next, many models have been established for laminates with a particular stacking sequence and particular boundary conditions, under uni-axial cyclic loading with constant amplitude, at one particular frequency. Their application to real structures with a stacking sequence varying from point to point, and more complex variations of the loads, is very complicated. Indeed some serious difficulties have to be overcome when fatigue life prediction of composite materials under general loading conditions is pursued: • The governing damage mechanism is not the same for all stress level states (Daniel and Charewicz, 1986; Barnard et al., 1985). Failure patterns vary with cyclic stress level and even with number of cycles. • The load history is important. When block loading sequences are applied in low-high order or in high-low order, there can be a considerable difference in damage growth (Hwang and Han, 1986). • Most experiments are performed in uni-axial stress conditions (e.g., uniaxial tension/compression), although more complex stress states do exist in real structures. • The residual strength and fatigue life of composite laminates have been observed to decrease more rapidly when the loading sequence is repeatedly changed after only a few loading cycles (Farrow, 1989). This socalled ‘cycle mix effect’ was described in detail by Farrow (1989) and means that the residual strength and the fatigue life of composite laminates are decreasing more rapidly when the (block) loading sequence shows frequent transitions from high-to-low or low-to-high stress levels after only a few loading cycles. • The frequency can have a major impact on the fatigue life. Ellyin and Kujawski (1992) investigated the frequency effect on the tensile fatigue performance of glass fibre-reinforced [±45º]5S laminates and concluded that there was a considerable influence of test loading frequency. Especially for matrix dominated laminates and loading conditions, frequency becomes important because of the general sensitivity of the matrix to the loading rate and because of the internal heat generation and associated temperature rise.

226


Clearly a lot of research has still to be done in this domain. However several attempts have been made to extend models for uni-axial constant amplitude loading to more general loading conditions, such as block-type and spectrum loading and to take into account the effect of cycling frequency and multi-axial loads. A rigorous classification of the most important fatigue models and life time prediction methodologies for fatigue testing of fibre-reinforced polymers is difficult, but a workable classification can be based on the classification of fatigue criteria by Sendeckyj (1990). According to Sendeckyj, fatigue criteria can be classified in four major categories: the macroscopic strength fatigue criteria, criteria based on residual strength and those based on residual stiffness, and finally the criteria based on the actual damage mechanisms. A similar classification has been used by Degrieck and Van Paepegem (2001) to classify the large number of existing fatigue models for composite laminates and consists of three major categories: (i) fatigue life models, which do not take into account the actual degradation mechanisms but use S-N curves or Goodman-type diagrams and introduce some sort of fatigue failure criterion, (ii) phenomenological models for residual stiffness/ strength, and (iii) progressive damage models which use one or more damage variables related to measurable manifestations of damage (transverse matrix cracks, delamination size). The next paragraphs briefly justify the classification. Although the fatigue behaviour of fibre-reinforced composites is fundamentally different from the behaviour exposed by metals, many models have been established which are based on the well-known S-N curves. These models make up the first class of so-called ‘fatigue life models’. This approach requires large amounts of experiments for each material, layup and loading condition (Schaff and Davidson, 1997a), and does not take into account the actual damage mechanisms, such as matrix cracks and fibre fracture. The second class comprises the phenomenological models for residual stiffness and strength. These models propose an evolution law which describes the (gradual) deterioration of the stiffness or strength of the composite specimen in terms of macroscopically observable properties, as opposed to the third class of progressive damage models, where the evolution law is proposed in direct relation with specific damage. Residual stiffness models account for the degradation of the elastic properties during fatigue. Stiffness can be measured frequently or even continuously during fatigue experiments, and can be measured without further degrading the material (Highsmith and Reifsnider, 1982). The residual stiffness model may be deterministic, in which a single-valued stiffness property is predicted, or statistical, in which predictions are for stiffness distributions. The other approach is based on the composite’s strength. In many applications of composite materials it is important to know the residual strength of the


227

composite structure, and as a consequence the remaining life time during which the structure can bear the external load. Therefore the so-called ‘residual strength’ models have been developed, which describe the deterioration of the initial static strength during fatigue life. From their early use, strength-based models have generally been statistical in nature. Most commonly, two-parameter Weibull functions are used to describe the residual strength and probability of failure for a set of laminates after an arbitrary number of cycles. Since the damage mechanisms which govern the fatigue behaviour of fibre-reinforced composites, have been studied intensively during the last decades, a last class of models have been proposed which describe the deterioration of the composite material in direct relation with specific damage (e.g., transverse matrix cracks, delamination size). These models correlate one or more properly chosen damage variables to some measure of the damage extent, quantitatively accounting for the progression of the actual damage mechanisms. These models are often designated as ‘mechanistic’ models. Summarized, fatigue models can be generally classified in three categories: (i) the fatigue life models, (ii) the phenomenological models for residual stiffness/strength, and (iii) the progressive damage models. One of the important outcomes of all established fatigue models is the life time prediction. Each of the three categories uses its own criterion for determining final failure and as a consequence for the fatigue life of the composite component. The fatigue life models use the information from SN curves or Goodman-type diagrams and introduce a fatigue failure criterion which determines the fatigue life of the composite specimen. Regarding the characterisation of the S-N behaviour of composite materials, Sendeckyj (1981) advises taking into account three assumptions: 1. S-N behaviour can be described by a deterministic equation, 2. Static strengths are uniquely related to the fatigue lives and residual strengths at runout (termination of cyclic testing). An example of such a relationship is the commonly used ‘strength-life equal rank assumption’ which states that for a given specimen its rank in static strength is equal to its rank in fatigue life (Hahn and Kim, 1975; Chou and Croman, 1978). 3. Static strength data can be described by a two-parameter Weibull distribution. Residual strength models have in fact an inherent ‘natural failure criterion’: failure occurs when the applied stress equals the residual strength (Harris, 1985; Schaff and Davidson, 1997a). Residual stiffness models are dealing with different definitions of ‘failure’ and already in the early 1970s, Salkind (1972) suggested drawing a family of S-N curves, being contours of a specified percentage of stiffness loss, to present fatigue data. In another

228


approach, fatigue failure is assumed to occur when the modulus has degraded to a critical level which has been defined by many investigators. Hahn and Kim (1976) and O’Brien and Reifsnider (1981) state that fatigue failure occurs when the fatigue secant modulus degrades to the secant modulus at the moment of failure in a static test. According to Hwang and Han (1986), fatigue failure occurs when the fatigue resultant strain reaches the static ultimate strain. Progressive damage models and life time prediction methodologies are very often inherently related, since the fatigue life can be predicted by establishing a fatigue failure criterion which is imposed to the progressive damage model. For specific damage types, the failure value of the damage variable(s) can be determined experimentally. The mechanistic or progressive damage models which quantitatively account for the progression of damage in composite laminates, offer the long-term promise to be applicable to a wide variety of materials, layups and loadings with a minimal amount of experimentally obtained input. At present, however, most of these models have only been applied to simple fatigue loadings or to a very specific class of materials (Schaff and Davidson, 1997a). Moreover, the vast majority of these models have been developed theoretically but their implementation and validation for practical applications is very limited up till now. Nevertheless, fatigue damage models should be based on a firm knowledge and sound modelling of the actual micro-scale damage mechanisms, yet resulting on a meso-scale in phenomenological residual stiffness models which characterise the composite’s fatigue behaviour in terms of macroscopically observable properties. Moreover, with regard to the implementation in numerical software, the fatigue damage models, which correlate damage with the degradation of macroscopic stiffness properties, can be used for real structures as well. When full-scale structural components are subjected to in-service fatigue loadings, stiffness is a very adequate parameter as it can be measured nondestructively and the residual stiffness exhibits much less statistical scatter than residual strength (Hashin, 1985; Whitworth, 2000; Highsmith and Reifsnider, 1982; Yang et al., 1990; 1992; Kedward and Beaumont, 1992).

10.6

Future trends and challenges

At least three major trends and challenges can be observed: (i) the trend towards multi-scale modelling of fatigue damage in composites, (ii) the trend towards better testing and instrumentation methods, and (iii) the challenge for faster assessment of the fatigue performance of new composite material combinations.


229

10.6.1 Towards multi-scale modelling Currently, fatigue simulation tools exist to predict the crack density in one layer of unidirectional fibres (= one UD-ply) (Joffe and Varna, 1999; Allen et al., 1987; Ogin et al., 1985), or to predict the degradation of the homogenised elastic properties of a UD-based laminate (Whitworth, 1998; Highsmith and Reifsnider, 1982; Yang et al., 1990), but none of these simulation tools bridges the gap between the micromechanical damage phenomena and the structural response of the damaged laminate. Moreover, very little work has been carried out on the fatigue behaviour of textile based composites (Fujii et al., 1992; Schulte et al., 1987; Caprino and Giorleo, 1999). The promising approach to these problems is multi-level modelling, which allows the inclusion of mesoscopic behaviour features in macroscopic descriptions, without the need for an a priori postulated macroscopic constitutive law (Zako et al., 2003; Carvelli and Poggi, 2001; Tang and Whitcomb, 2003; Edgren et al., 2004). Macroscopic constitutive relations (material properties on the laminate level) are obtained from scaling up material modelling at lower (meso- and micro-)scales, where the detailed material structure with its specific material behaviour is represented. The multi-scale approach couples the advantages of a pure micro- or mesomechanical approach to those of a pure macroscopic modelling. The complex material behaviour is properly captured by modelling at lower scales, while largescale analyses at the macroscale remain numerically feasible.

10.6.2 Better testing and instrumentation methods For a long time, fatigue testing of composites was only focused on providing the S-N fatigue life data. No efforts were made to gather additional data from the same test by using more advanced instrumentation methods. The development of methods such as digital image correlation (strain mapping) and optical fibre sensing allows for much better instrumentation, combined with traditional equipment such as extensometers, thermocouples and resistance measurement. Validation with finite element simulations of the realistic boundary conditions and loading conditions in the experimental set-up must maximise the generated data from one single fatigue test. Figure 10.8 shows an example of the comparison between experimentally measured and numerically simulated strain fields in a cruciform specimen for biaxial fatigue loading (Lamkanfi et al., 2006). Currently a lot of research effort is spent on the integration of smart sensors into composite structures. Such embedded sensors must monitor the deformation and local strains in the composite component under

230


0 0,3 0,6 0,9 1,2 1,5 1,8 2,1 2,4 2,7[%]

10.8 Comparison of experimentally measured (Courtesy of VUB) and numerically simulated strain fields in a cruciform specimen for biaxial fatigue loading (Lamkanfi et al., 2006).

in-service loading conditions and must be capable of detecting damage at an early stage. In that way, structural health monitoring would lead to (i) smaller safety factors at design (because the loads are better known), (ii) early repair of fatigue damage, and (iii) extended fatigue lives (because the load history is well-known).

10.6.3 Faster assessment of fatigue performance of new composite materials The constituent materials for textile composites are developing very rapidly: new resins are developed, toughening fillers are added, new geometrical arrangements of fibre reinforcement are created, and new fibre types are explored (silk, banana and bamboo). This contrasts sharply with the very time-consuming and expensive characterisation of the fatigue performance of only one fibre/matrix combination. Therefore faster methods for assessment of the fatigue performance of new composite materials should be developed.

10.7

Sources of further information and advice

The textbook Fatigue in composites. Science and technology of the fatigue response of fibre-reinforced plastics edited by Bryan Harris (2003) is one of the best reference books for further information about fatigue of composite materials. The three-yearly ‘International Conference on Fatigue of Composites’ provides a forum to all researchers active in this field. The first conference was held in 1997 (Paris, France), followed by the conferences in 2000 (Williamsburg, USA) and 2004 (Kyoto, Japan) and 2007 (Kaiserslautern, Germany).


231

Relevant testing standards for fatigue of structural composites are: • •

• • •

ASTM D3479/D3479M-96(2002)e1 Standard Test Method for TensionTension Fatigue of Polymer Matrix Composite Materials ASTM D6115–97(2004) Standard Test Method for Mode I Fatigue Delamination Growth Onset of Unidirectional Fiber-Reinforced Polymer Matrix Composites EN ISO 13003:2003 Fibre-reinforced plastics – Determination of fatigue properties under cyclic loading conditions EN ISO 14692 Petroleum and natural gas industries – Glass-reinforced plastics (GRP) piping EN 12245 Transportable gas cylinders – Fully wrapped composite cylinders

For a broader study of textile composites, the textbook Design and manufacture of textile composites can be recommended (Long, 2006).

10.8

References

Abbadi, A., Azari, Z., Belouettar, S., Gilgert, J. and Pluvinage, G. (2006). Static and fatigue characterization of sandwich panels. Materials Science Forum, 518, 555–562. Abry, J.C., Bochard, S., Chateauminois, A., Salvia, M. and Giraud, G. (1999). In situ detection of damage in CFRP laminates by electrical resistance measurements. Composites Science and Technology, 59, 925–935. Adali, S. (1985). Optimization of fibre reinforced composite laminates subject to fatigue loading. In: Marshall, I.H. (ed.). Composite Structures 3. Proceedings of the third International Conference on Composite Structures (ICCS-3), 9–11 September 1985, Scotland, Elsevier, pp. 43–55. Adam, T., Gathercole, N., Reiter, H. and Harris, B. (1994). Life prediction for fatigue of T800/5245 carbon-fibre composites: II. Variable-amplitude loading. International Journal of Fatigue, 16(8), 533–547. Allen, D.H., Harris, C.E. and Groves, S.E. (1987). A thermomechanical constitutive theory for elastic composites with distributed damage – II. Application to matrix cracking in laminated composites. International Journal of Solids & Structures, 23(9), 1319–1338. Angelidis, N., Wei, C.Y. and Irving, P.E. (2004). The electrical resistance response of continuous carbon fibre composite laminates to mechanical strain. Composites Part A, 35, 1135–1147. Angelidis, N., Wei, C.Y. and Irving, P.E. (2006). Short communication – Response to discussion of paper: The electrical resistance response of continuous carbon fibre composite laminates to mechanical strain. Composites Part A, 37, 1495–1499. Bach, P.W. (1996). The effect of moisture on the fatigue performance of glass fibre reinforced polyester coupons and bolted joints. Energieonderzoek Centrum Nederland (ECN), http://www.ecn.nl/library/ecn_rap/abstr96/c_015.html, 34 p. Badaliance, R., Dill, H.D. and Potter, J.M. (1982). Effects of spectrum variations on fatigue life of composites. In: Daniel, I.M. (ed.) Composite materials: Testing and

232


design (Sixth Conference). ASTM STP 787. Philadelphia, American Society for Testing and Materials, pp. 274–286. Barnard, P.M., Butler, R.J. and Curtis, P.T. (1985). Fatigue scatter of UD glass epoxy, a fact or fiction? In: Marshall, I.H. (ed.) Composite Structures 3. Proceedings of the third International Conference on Composite Structures (ICCS-3), 9–11 September 1985, Scotland, Elsevier, pp. 69–82. Bartley-Cho, J., Lim, S.G., Hahn, H.T. and Shyprykevich, P. (1998). Damage accumulation in quasi-isotropic graphite/epoxy laminates under constantamplitude fatigue and block loading. Composites Science and Technology, 58, 1535–1547. Bassam, F., Boniface, L., Jones, K. and Ogin, S.L. (1998). On the behaviour of the residual strain produced by matrix cracking in cross-ply laminates. Composites Part A, 29(11), 1425–1432. Beaumont, P.W.R. (1987). The fatigue damage mechanics of composite laminates. In: Wang, A.S.D. and Haritos, G.K. (eds) Damage mechanics in composites. Presented at the Winter Annual Meeting of the ASME, 13–18 December 1987, Boston, Massachusetts, ASME, pp. 53–63. Bond, I.P. (1999). Fatigue life prediction for GRP subjected to variable amplitude loading. Composites Part A, 30(8), 961–970. Bourchak, M., Farrow, I.R., Bond, I.P., Rowland, C.W. and Menan, F. (2007). Acoustic emission energy as a fatigue damage parameter for CFRP composites. International Journal of Fatigue, 29(2007), 457–470. Bressers, J. (ed.) (1995). A code of practice for the measurement of misalignment induced bending in uniaxially loaded tension-compression test pieces. European Commission, Institute of Advanced Materials, ISBN 92-826-9681-2, 52 pp. Brocker, W. and Woithe, K. (1991). Fatigue life behaviour of composite structures. Report for Industrieanlagen-Betriebsgesellschaft MBH Ottobrunn (Germany). Brøndsted, P., Andersen, S.I. and Lilholt, H. (1997). Fatigue damage accumulation and lifetime prediction of GFRP materials under block loading and stochastic loading. In: Andersen, S.I., Brøndsted, P., Lilholt, H., Lystrup, Aa., Rheinländer, J.T., Sørensen, B.F. and Toftegaard, H. (eds) Polymeric Composites – Expanding the Limits. Proceedings of the 18th Risø International Symposium on Materials Science, 1–5 September 1997, Roskilde, Denmark, Risø International Laboratory, pp. 269–278. Caprino, G. and D’Amore, A. (1998). Flexural fatigue behaviour of random continuous-fibre-reinforced thermoplastic composites. Composites Science and Technology, 58, 957–965. Caprino, G. and Giorleo, G. (1999). Fatigue lifetime of glass fabric/epoxy composites. Composites Part A, 30(3), 299–304. Carvelli, V. and Poggi, C. (2001). A homogenization procedure for the numerical analysis of woven fabric composites. Composites Part A, 32(10): 1425–1432. Chen, A.S. and Matthews, F.L. (1993a). A review of multiaxial/biaxial loading tests for composite materials. Composites, 24(5), 395–406. Chen, A.S. and Matthews, F.L. (1993b). Biaxial flexural fatigue of composite plates. In: Miravete, A. (ed.) ICCM/9 Composites: properties and applications. Volume VI. Proceedings of the Ninth International Conference on Composite Materials, 12–16 July 1993, Madrid, Spain, Woodhead Publishing Limited, pp. 899–906. Chimenti, D.E. and Nayfeh, A.H. (1985). Leaky lamb waves in fibrous composite laminates. Journal of Applied Physics, 58(12), 4531–4538.


233

Chou, P.C. and Croman, R. (1978). Residual Strength in Fatigue Based on the Strength-Life Equal Rank Assumption. Journal of Composite Materials, 12, 177–194. Chung, D.D.L. and Wang, S. (2006). Discussion on paper ‘The electrical resistance response of continuous carbon fibre composite laminates to mechanical strain’ by Angelidis N, Wei CY, Irving PE, Composites: Part A, 2004; 35: 1135–1147. Composites Part A, 37, 1490–1494. Cnudde, V., Masschaele, B., Dierick, M., Vlassenbroeck, J., Van Hoorebeke, L. and Jacobs, P. (2006). Recent progress in X-ray CT as a geosciences tool. Applied Geochemistry, 21(5), 826–832. Coronet – Thermoplastic Composites Infrastructure Cooperation Network (2007). http://www.coronet.eu.com/uploads/Matrix%20Material%20Properties.doc. Curtis, P.T. (1989). The fatigue of organic matrix composite materials. In: Partridge, I.K. (ed.) Advanced composites. Londen, Elsevier, pp. 331–367. Daniel, I.M. and Charewicz, A. (1986). Fatigue damage mechanisms and residual properties of graphite/epoxy laminates. Engineering Fracture Mechanics, 25(5/6), 793–808. De Baere, I., Van Paepegem, W. and Degrieck, J. (2007a). The use of rivets for electrical resistance measurement on carbon fibre-reinforced thermoplastics. Smart Materials & Structures, 16(5), 1821–1828. De Baere, I., Voet, E., Van Paepegem, W., Vlekken, J., Cnudde, V., Masschaele, B. and Degrieck, J. (2007b). Strain monitoring in thermoplastic composites with optical fibre sensors: embedding process, visualization with micro-tomography and fatigue results. Journal of Thermoplastic Composite Materials, 20(5), 453–472. Declercq, N. (2005). The interaction of complex harmonic elastic waves with periodically corrugated surfaces and with anisotropic viscoelastic or piezoelectric layered media. Chapter XIII. Acoustic microscopy. PhD thesis, Ghent University, Ghent, 2005, pp. 595–620. Degrieck, J. and Van Paepegem, W. (2001). Fatigue Damage Modelling of FibreReinforced Composite Materials: Review. Applied Mechanics Reviews, 54(4), 279–300. Doyle, C., Martin, A., Liu, T., Wu, M., Hayes, S., Crosby, P.A., Powell, G.R., Brooks, D. and Fernando, G.F. (1998). In-situ process and condition monitoring of advanced fibre-reinforced composite materials using optical fibre sensors. Smart Materials & Structures, 7(2), 145–158 APR 1998. Drechsler, K. (2000). Advanced textile structural composites – Needs and current developments. Proceedings of the Fifth International Conference on Textile Composites, Leuven, Belgium, 18–20 September 2000. Edgren, F., Mattsson, D., Asp, L.E. and Varna, J. (2004). Formation of damage and its effects on non-crimp fabric reinforced composites loaded in tension. Composites Science and Technology, 64, 675–692. El Mahi, A., Berthelot, J.-M. and Bezazi, A. (2002). The fatigue behaviour and damage development in cross-ply laminates in flexural tests. Proceedings of the Tenth European Conference on Composite Materials (ECCM-10), Brugge, Belgium, 3–7 June 2002. El Mahi, A., Farooq, M.K., Sahraoui, S. and Bezazi, A. (2004). Modelling the flexural behaviour of sandwich composite materials under cyclic fatigue. Materials & Design, 25(3), 199–208.

234


Ellyin, F. and Kujawski, D. (1992). Fatigue testing and life prediction of fibreglassreinforced composites. In: Neale, K.W. and Labossière, P. (eds) First International Conference on Advanced Composite Materials in Bridges and Structures (ACMBSI), 1992, Sherbrooke, Québec, Canada, Canadian Society for Civil Engineering, pp. 111–118. Ellyin, F. and Martens, M. (2001). Biaxial fatigue behaviour of a multidirectional filament-wound glass-fiber/epoxy pipe. Composites Science and Technology, 61, 491–502. Epaarachchi, J.A. (2006). A study on estimation of damage accumulation of glass fibre reinforced plastic (GFRP) composites under a block loading situation. Composite Structures, 75(1–4), 88–92. Farooq, M.H., Sahraoui, S. and El Mahi, A. (2002). Evaluation of fatigue behaviour in sandwich composite materials. Proceedings of the Tenth European Conference on Composite Materials (ECCM-10), Brugge, Belgium, 3–7 June 2002. Farrow, I.R. (1989). Damage accumulation and degradation of composite laminates under aircraft service loading: assessment and prediction. Volumes I and II. Cranfield Institute of Technology, Ph.D. Thesis. Ferry, L., Gabory, D., Sicot, N., Berard, J.Y., Perreux, D. and Varchon, D. (1997). Experimental study of glass-epoxy composite bars loaded in combined bending and torsion loads. Fatigue and characterisation of the damage growth. In: Degallaix, S., Bathias, C. and Fougères, R. (eds) International Conference on fatigue of composites. Proceedings, 3–5 June 1997, Paris, France, La Société Française de Métallurgie et de Matériaux, pp. 266–273. Ferry, L., Perreux, D., Varchon, D. and Sicot, N. (1999). Fatigue behaviour of composite bars subjected to bending and torsion. Composites Science and Technology, 59, 575–582. Fong, J.T. (1982). What is fatigue damage? In: Reifsnider, K.L. (ed.) Damage in composite materials. ASTM STP 775. American Society for Testing and Materials, pp. 243–266. Found, M.S. and Quaresimin, M. (2003). Two-stage fatigue loading of woven carbon fibre reinforced laminates. Fatigue & Fracture of Engineering Materials & Structures, 26(1), 17–26. Fujii, T., Amijima, S., Lin, F. and Sagami, T. (1992). Study on strength and nonlinear stress-strain response of plain woven glass fiber laminates under biaxial loading. Journal of Composite Materials, 26(17), 2493–2510. Fujii, T., Amijima, S. and Okubo, K. (1993). Microscopic fatigue processes in a plain-weave glass-fibre composite. Composites Science and Technology, 49, 327–333. Gagel, A., Fiedler, B. and Schulte, K. (2006a). On modelling the mechanical degradation of fatigue loaded glass-fibre non-crimp fabric reinforced epoxy laminates. Composites Science and Technology, 66(5), 657–664. Gagel, A., Lange, D. and Schulte, K. (2006b). On the relation between crack densities, stiffness degradation, and surface temperature distribution of tensile fatigue loaded glass-fibre non-crimp-fabric reinforced epoxy. Composites Part A – Applied science and manufacturing, 37(2), 222–228. Galiotis, C. and Koimtzoglou, C. (2003). The effect of the interface on the fatigue performance of fibre composites. In: Harris, B. (ed.) Fatigue in Composites. Cambridge, Woodhead Publishing and CRC Press, 2003, pp. 147–172.


235

Gamstedt, E.K. and Sjogren, B.A. (1999). Micromechanisms in tension-compression fatigue of composite laminates containing transverse plies. Composites Science and Technology, 59(2), 167–178. Gamstedt, E.K., Skrifvars, M., Jacobsen, T.K. and Pyrz, R. (2002). Synthesis of unsaturated polyesters for improved interfacial strength in carbon fibre composites. Composites Part A – Applied science and manufacturing, 33(9), 1239–1252. Gathercole, N., Reiter, H., Adam, T. and Harris, B. (1994). Life prediction for fatigue of T800/5245 carbon-fibre composites: I. Constant-amplitude loading. International Journal of Fatigue, 16(8), 523–532. Gibbs, H.H. (1998). High temperature resins. In: Peters, S.T. (ed.) Handbook of Composites, Chapman & Hall, London, 1998, ISBN 0 412 54020 7. Haberkern, H. (2006). Multiaxials – Tailor-made reinforcements. Reinforced Plastics, April 2006, pp. 28–33. Hahn, H.T. and Kim, R.Y. (1975). Proof testing of composite materials. Journal of Composite Materials, 9, 297–311. Hahn, H.T. and Kim, R.Y. (1976). Fatigue behaviour of composite laminates. Journal of Composite Materials, 10, 156–180. Hancox, N.L. (1983). High performance composites with resin matrices. In: Kelly, A. and Mileiko, S.T. (eds) Fabrication of Composites. Handbook of Composites vol. 4. Amsterdam, North Holland, 1983, pp. 1–44. Hansen, U. (1997). Damage development in woven fabric composites during tensiontension fatigue. In: Andersen, S.I., Brøndsted, P., Lilholt, H., Lystrup, Aa., Rheinländer, J.T., Sørensen, B.F. and Toftegaard, H. (eds) Polymeric Composites – Expanding the Limits. Proceedings of the 18th Risø International Symposium on Materials Science, 1–5 September 1997, Roskilde, Denmark, Risø International Laboratory, pp. 345–351. Hansen, U. (1999). Damage development in woven fabric composites during tensiontension fatigue. Journal of Composite Materials, 33(7), 614–639. Harik, V.M., Klinger, J.R. and Bogetti, T.A. (2000). Low cycle fatigue of unidirectional laminates: stress ratio effects. Journal of Engineering Materials and Technology, 122, 415–419. Harris, B. (1981). Fatigue testing of fibre composites. In: Advisory Group for Aerospace Research & Development (AGARD). Fatigue test methodology. AGARD Lecture Series No. 118 (LS-118), pp. 6.1–6.18. Harris, B. (1985). Fatigue behaviour of polymer-based composites and life prediction methods. AIB-Vinçotte Leerstoel, 2 maart 1995, Belgium, Nationaal Fonds voor Wetenschappelijk Onderzoek, 28 p. Harris, B. (ed.) (2003). Fatigue in composites. Science and technology of the fatigue response of fibre-reinforced plastics. Cambridge, Woodhead Publishing Ltd., 742 pp. Hasan, E.H., Abdel Hakeem, H.M. and El-Sayed, A.A.M. (1998). Environmental rotating bending fatigue behavior of fiber glass/polyester composites. In: Crivelli Visconti, I. (ed.) ECCM-8: Science, technologies and applications. Volume 3. Proceedings of the Eight European Conference on Composite Materials, 3–6 June 1998, Naples, Italy, Woodhead Publishing Limited, pp. 191–198. Hashin, Z. (1985). Cumulative damage theory for composite materials: residual life and residual strength methods. Composites Science and Technology, 23, 1–19.

236


Herakovich, C.T. (1998). Mechanics of fibrous composites. New York, John Wiley & Sons, Inc. Herrington, P.D. and Doucet, A.B. (1992). Progression of bending fatigue damage around a discontinuity in glass/epoxy composites. Journal of Composite Materials, 26(14), 2045–2059. Highsmith, A.L. and Reifsnider, K.L. (1982). Stiffness-reduction mechanisms in composite laminates. In: Reifsnider, K.L. (ed.). Damage in composite materials. ASTM STP 775. American Society for Testing and Materials, pp. 103–117. Hochard, C., Payan, J. and Bordreuil, C. (2006). A progressive first ply failure model for woven ply CFRP laminates under static and fatigue loads. International Journal of Fatigue, 28(10), 1270–1276. Hull, D. (1999). Fractography: observing, measuring and interpreting fracture surface topography. Cambridge, Cambridge University Press, 366 pp. Hwang, W. and Han, K.S. (1986). Cumulative damage models and multi-stress fatigue life prediction. Journal of Composite Materials, 20, 125–153. Inoue, A., Fujii, T. and Kawakami, H. (2000). Effect of loading path on mechanical response of a glass fabric composite at low cyclic fatigue under tension/torsion biaxial loading. Journal of Reinforced Plastics and Composites, 19(2), 111–123. Joffe, R. and Varna, J. (1999). Analytical modeling of stiffness reduction in symmetric and balanced laminates due to cracks in 90º layers. Composites Science and Technology, 59, 1641–1652. Jones, R., Williams, J.F. and Tay, T.E. (1987). Is fatigue testing of impact damaged laminates necessary? Composite Structures, 8, 1–12. Judawisastra, H., Ivens, J. and Verpoest, I. (1997). Bending fatigue behaviour and damage development of PUR-epoxy 3D woven sandwich composites. In: Degallaix, S., Bathias, C. and Fougères, R. (eds) International Conference on fatigue of composites. Proceedings, 3–5 June 1997, Paris, France, La Société Française de Métallurgie et de Matériaux, pp. 197–204. Kedward, K.T. and Beaumont, P.W.R. (1992). The treatment of fatigue and damage accumulation in composite design. International Journal of Fatigue, 14(5), 283–294. Kim, R.Y. (1980). Experimental assessment of static and fatigue damage of graphite epoxy laminates. ICCM 3. Volume 2, pp. 1015–1028. Lamkanfi, E., Van Paepegem, W., Smits, A., Van Hemelrijck, D. and Sol, H. (2006). Comparison of DIC strain measurements and FEM simulations in cruciform specimens for biaxial fatigue loading. Internal report for the FWO project ‘Experimental and theoretical study of the damage behaviour of fibre-reinforced composite systems under biaxial fatigue loading’. Lee, B.L. and Liu, D.S. (1994). Cumulative damage of fiber-reinforced elastomer composites under fatigue loading. Journal of Composite Materials, 28(13), 1261–1286. Lee, C.-S. and Hwang, W.-B. (2001). Lifetime prediction of fiber-reinforced polymermatrix composite materials under tension/torsion biaxial loading. In: Hui, D. (ed.) Eighth International Conference on Composites Engineering (ICCE/8). Proceedings. Tenerife, Spain, 5–11 August 2001, pp. 535–536. Lessard, L.B., Eilers, O.P. and Shokrieh, M.M. (1995). Testing of in-plane shear properties under fatigue loading. Journal of Reinforced Plastics and Composites, 14, 965–987. Lomov, S.L. (2006). Manufacturing and internal geometry of textiles. In: Long, A.C. (ed.) Design and manufacture of textile composites. Taylor and Francis, ISBN 0849325935.


237

Lomov, S.V., Huysmans, G., Luo, Y., Prodromou, A. and Verpoest, I. (2000). Textile composites: modelling strategies. Proceedings of the Fifth International Conference on Textile Composites, Leuven, Belgium, 18–20 September 2000. Long, A.C. (ed.) (2006). Design and manufacture of textile composites. Taylor and Francis, ISBN 0849325935. Maes, K. (1998). Non-destructive evaluation of degradation in a fibre-reinforced plastic. Master thesis (in Dutch). Ghent University, Ghent, 105 pp. Mahfuz, H., Zaman, K., Haque, A., Foy, C., Mohamed, H. and Jeelani, S. (2000). Fatigue life prediction of thick-section S2-glass/vinylester composites under flexural loading. Journal of Engineering Materials and Technology – Transactions of the ASME, 122, 402–408. Mahfuz, H., Saha, M., Zaman, K. and Jeelani, S. (2001). Life prediction of woven composites under flexural fatigue. In: Hui, D. (ed.) Eighth International Conference on Composites Engineering (ICCE/8). Proceedings. Tenerife, Spain, 5–11 August 2001, pp. 611–612. Mallick, P.K. (1997). Composites Engineering Handbook. New York, Marcel Dekker Inc. Masters, J.E. and Reifsnider, K.L. (1982). An investigation of cumulative damage development in quasi-isotropic graphite/epoxy laminates. In: Reifsnider, K.L. (ed.) Damage in composite materials. ASTM STP 775. American Society for Testing and Materials, pp. 40–62. Matondang, T.H. and Schutz, D. (1984). The influence of anti-buckling guides on the compression-fatigue behaviour of carbon fibre-reinforced plastic laminates. Composites, 15(3), 217–221. McBride, T.M. and Chen, J. (1997). Unit-cell geometry in plain-weave fabrics during shear deformations. Composites Science and Technology, 57, 345–351. Miyano, Y., McMurray, M.K., Enyama, J. and Nakada, M. (1994). Loading rate and temperature dependence on flexural fatigue behavior of a satin woven CFRP laminate. Journal of Composite Materials, 28(13), 1250–1260. Mouritz, A.P. (2003). Non-destructive evaluation of damage accumulation. In: Harris, B. (ed.) Fatigue in Composites. Cambridge, Woodhead Publishing and CRC Press, 2003, pp. 242–266. Naik, N.K. (2003). Woven-fibre thermoset composites. In: Harris, B. (ed.) Fatigue in Composites. Cambridge, Woodhead Publishing and CRC Press, 2003, pp. 296–313. Neubert, H., Harig, H. and Schulte, K. (1987). Monitoring of fatigue induced damage processes in CFRP by means of thermometric methods. In: Matthews, F.L., Buskell, N.C.R., Hodgkinson, J.M. and Morton, J. (eds) Sixth International Conference on Composite Materials (ICCM-VI) & Second European Conference on Composite Materials (ECCM-II): Volume 1. Proceedings, 20–24 July 1987, London, UK, Elsevier, pp. 1.359–1.368. Nishikawa, Y., Okubo, K., Fujii, T. and Kawabe, K. (2006). Fatigue crack constraint in plain-woven CFRP using newly-developed spread tows. International Journal of Fatigue, 28(10), 1248–1253. O’Brien, T.K. and Reifsnider, K.L. (1981). Fatigue damage evaluation through stiffness measurements in boron-epoxy laminates. Journal of Composite Materials, 15, 55–70. Ogin, S.L., Smith, P.A. and Beaumont, P.W.R. (1985). Matrix cracking and stiffness reduction during the fatigue of a (0/90)s GFRP laminate. Composites Science and Technology, 22(1), 23–31.

238


Pandita, S.D. and Verpoest, I. (2004). Tension-tension fatigue behaviour of knitted fabric composites. Composite Structures, 64(2), 199–209. Pandita, S.D., Huysmans, G., Wevers, M. and Verpoest, I. (2001). Tensile fatigue behaviour of glass-plain weave fabric composites in the on and off-axis directions. Composites Part A: Applied Science and Manufacturing, 32(10), 1533–1539 Sp. Iss. SI, 2001. Perreux, D. and Joseph, E. (1997). The effect of frequency on the fatigue performance of filament-wound pipes under biaxial loading: experimental results and damage model. Composites Science and Technology, 57, 353–364. Perreux, D., Thiebaud, F., Farines, L. and Uskokovic, P.S. (2005). Investigation on the static and fatigue failure of bi-directional composite pipes. Materials Science Forum, 494, 435–442. Philippidis, T.P. and Vassilopoulos, A.P. (1999). Fatigue strength prediction under multiaxial stress. Journal of Composite Materials, 33(17), 1578–1599. Quaresimin, M. (2002). Fatigue of woven composite laminates under tensile and compressive loading. Proceedings of the Tenth European Conference on Composite Materials (ECCM-10), Brugge, Belgium, 3–7 June 2002. Quaresimin, M. and Susmel, L. (2002). Multiaxial fatigue behaviour of composite laminates. Key Engineering Materials, 221(2), 71–79. Ramkumar, R.L. (1982). Compression fatigue behavior of composites in the presence of delaminations. In: Reifsnider, K.L. (ed.) Damage in composite materials. ASTM STP 775, pp. 184–210. Reddy, J.N. (1997). Mechanics of Laminated Composite Plates. Theory and Analysis. Florida, CRC Press. Salkind, M.J. (1972). Fatigue of composites. In: Corten, H.T. (ed.) Composite Materials Testing and Design (Second Conference). ASTM STP 497. Baltimore, American Society for Testing and Materials, pp. 143–169. SAMPE Journal. Special Issue ‘Prepregs and Fabrics’. March/April 2004. Official Journal of the Society for the Advancement of Material and Process Engineering. Samyn, P. (2007). Tribophysical Interpretation of Scaling Effects in Friction and Wear for Polymers. Chapter 3. Test materials. PhD thesis, Ghent University, Ghent. Schaff, J.R. and Davidson, B.D. (1997a). Life prediction methodology for composite structures. Part I – Constant amplitude and two-stress level fatigue. Journal of Composite Materials, 31(2), 128–157. Schaff, J.R. and Davidson, B.D. (1997b). Life prediction methodology for composite structures. Part II – Spectrum fatigue. Journal of Composite Materials, 31(2), 158–181. Schulte, K. (1984). Stiffness reduction and development of longitudinal cracks during fatigue loading of composite laminates. In: Cardon, A.H. and Verchery, G. (eds) Mechanical characterisation of load bearing fibre composite laminates. Proceedings of the European Mechanics Colloquium 182, 29–31 August 1984, Brussels, Belgium, Elsevier, pp. 36–54. Schulte, K., Baron, Ch., Neubert, H., Bader, M.G., Boniface, L., Wevers, M., Verpoest, I. and de Charentenay, F.X. (1985). Damage development in carbon fibre epoxy laminates: cyclic loading. In: Proceedings of the MRS-symposium ‘Advanced Materials for Transport’, November 1985, Strassbourg, 8 p. Schulte, K., Reese, E. and Chou, T.-W. (1987). Fatigue behaviour and damage development in woven fabric and hybrid fabric composites. In: Matthews, F.L.,


239

Buskell, N.C.R., Hodgkinson, J.M. and Morton, J. (eds) Sixth International Conference on Composite Materials (ICCM-VI) & Second European Conference on Composite Materials (ECCM-II): Volume 4. Proceedings, 20–24 July 1987, London, UK, Elsevier, pp. 4.89–4.99. Schutz, D. (1981). Variable amplitude fatigue testing. In: AGARD Lecture Series No. 118. Fatigue test methodology, pp. 4.1–4.31 Seale, M.D., Smith, B.T., Prosser, W.H. and Masters, J.E. (1993). Lamb wave response of fatigued composite samples. Technical Report conf-rpqnde-93-p1261, NASA Langley Technical Report Server. Sendeckyj, G.P. (1981). Fitting models to composite materials fatigue data. In: Chamis, C.C. (ed.) Test methods and design allowables for fibrous composites. ASTM STP 734. American Society for Testing and Materials, pp. 245–260. Sendeckyj, G.P. (1990). Life prediction for resin-matrix composite materials. In: Reifsnider, K.L. (ed.) Fatigue of composite materials. Composite Material Series 4. Elsevier, pp. 431–483. Shokrieh, M.M. and Lessard, L.B. (2003). Fatigue under multiaxial stress systems. In: Harris, B. (ed.) Fatigue in Composites. Cambridge, Woodhead Publishing and CRC Press, 2003, pp. 63–113. Sidoroff, F. and Subagio, B. (1987). Fatigue damage modelling of composite materials from bending tests. In: Matthews, F.L., Buskell, N.C.R., Hodgkinson, J.M. and Morton, J. (eds) Sixth International Conference on Composite Materials (ICCM-VI) & Second European Conference on Composite Materials (ECCM-II): Volume 4. Proceedings, 20–24 July 1987, London, UK, Elsevier, pp. 4.32–4.39. Smits, A., Van Hemelrijck, D., Philippidis, T.P. and Cardon, A. (2006). Design of a cruciform specimen for biaxial testing of fibre reinforced composite laminates. Composites Science and Technology, 66, 964–975. SP Systems Guide to Composites (2007). http://www.gurit.com/. Stewart, R. (2002). US composites industry expands despite economy. Reinforced Plastics, November 2002, 20–26. Tang, X. and Whitcomb, J.D. (2003). Progressive failure behaviour of 2D woven composites. Journal of Composite Materials, 37(14), 1239–1259. ten Have, A.A. (1991). Wisper and WisperX final definition of two standardised fatigue loading sequences for wind turbine blades. NLR report NLR TP91476. Toubal, L., Karama, M. and Lorrain, B. (2006). Damage evolution and infrared thermography in woven composite laminates under fatigue loading. International Journal of Fatigue, 28(12), 1867–1872. Van Paepegem, W. (2002). Development and finite element implementation of a damage model for fatigue of fibre-reinforced polymers. PhD thesis. Gent, Belgium, Ghent University Architectural and Engineering Press (ISBN 90-76714-13-4), 403 p. Van Paepegem, W. and Degrieck, J. (2001). Fatigue Degradation Modelling of Plain Woven Glass/epoxy Composites. Composites Part A, 32(10), 1433–1441. Van Paepegem, W. and Degrieck, J. (2002a). A New Coupled Approach of Residual Stiffness and Strength for Fatigue of Fibre-reinforced Composites. International Journal of Fatigue, 24(7), 747–762. Van Paepegem, W. and Degrieck, J. (2002b). Effects of Load Sequence and Block Loading on the Fatigue Response of Fibre-reinforced Composites. Mechanics of Advanced Materials and Structures, 9(1), 19–35.

240


Van Paepegem, W. and Degrieck, J. (2002c). Modelling damage and permanent strain in fibre-reinforced composites under in-plane fatigue loading. Composites Science and Technology, 63(5), 677–694. Van Paepegem, W. and Degrieck, J. (2005). Simulating Damage and Permanent Strain in Composites under In-plane Fatigue Loading. Computers & Structures, 83(23–24), 1930–1942. Van Paepegem, W., De Baere, I. and Degrieck, J. (2006a). Modelling the nonlinear shear stress-strain response of glass fibre-reinforced composites. Part I: Experimental results. Composites Science and Technology, 66(10), 1455–1464. Van Paepegem, W., De Baere, I. and Degrieck, J. (2006b). Modelling the nonlinear shear stress-strain response of glass fibre-reinforced composites. Part II: Model development and finite element simulations. Composites Science and Technology, 66(10), 1465–1478. Van Paepegem, W., De Geyter, K., Vanhooymissen, P. and Degrieck, J. (2006c). Effect of friction on the hysteresis loops from three-point bending fatigue tests of fibrereinforced composites. Composite Structures, 72(2), 212–217. Van Paepegem, W., De Baere, I., Lamkanfi, E. and Degrieck, J. (2007). Poisson’s ratio as a sensitive indicator of (fatigue) damage in fibre-reinforced plastics. Fatigue and Fracture of Engineering Materials & Structures, 30, 269–276. Verpoest, I. (2006). Module 1: Composite materials – Fibre-reinforced plastics. IVPV course on Composite Materials, Ghent University, Ghent, 2006. Walsh, R.M.Jr. and Pipes, R.B. (1982). Compression fatigue behaviour of notched composite laminates. Journal of Materials Science, 17, 2567–2576. Wevers, M. (1987). Identification of fatigue failure modes in carbon fibre reinforced composites. Part I: Text. Master thesis, Leuven, Belgium, Faculteit Toegepaste Wetenschappen, 210 p. Wevers, M., Verpoest, I., Aernoudt, E. and De Meester, P. (1987). Fatigue damage development in carbon fibre reinforced epoxy composites: correlation between the stiffness degradation and the growth of different damage types. In: Matthews, F.L., Buskell, N.C.R., Hodgkinson, J.M. and Morton, J. (eds) Sixth International Conference on Composite Materials (ICCM-VI) & Second European Conference on Composite Materials (ECCM-II): Volume 4. Proceedings, 20–24 July 1987, London, UK, Elsevier, pp. 4.114–4.128. Wevers, M., Verpoest, I. and De Meester, P. (1990). Is crack closure due to fatigue loading causing more damage in carbon fibre reinforced epoxy composites? In: Füller, J., Grüninger, G., Schulte, K., Bunsell, A.R. and Massiah, A. (eds) Developments in the science and technology of composite materials. Proceedings of the Fourth European Conference on Composite Materials (ECCM/4), 25–28 September 1990, Stuttgart, Elsevier Applied Science, pp. 181–188. Whitworth, H.A. (1998). A stiffness degradation model for composite laminates under fatigue loading. Composite Structures, 40(2), 95–101. Whitworth, H.A. (2000). Evaluation of the residual strength degradation in composite laminates under fatigue loading. Composite Structures, 48(4), 261–264. Xiao, X.R. (1999). Modeling of load frequency effect on fatigue life of thermoplastic composites. Journal of Composite Materials, 33(12), 1141–1158. Yang, J.N., Jones, D.L., Yang, S.H. and Meskini, A. (1990). A stiffness degradation model for graphite/epoxy laminates. Journal of Composite Materials, 24, 753–769.


241

Yang, J.N., Lee, L.J. and Sheu, D.Y. (1992). Modulus reduction and fatigue damage of matrix dominated composite laminates. Composite Structures, 21, 91–100. Zako, M., Uetsuji, Y. and Kurashiki, T. (2003). Finite element analysis of damaged woven fabric composite materials. Composites Science and Technology, 63, 507–516. Zhang, M.R. (2003). A Review of the Epoxy Resin Toughening. Department of Chemical Engineering and Materials Science, Syracuse University, Syracuse. http://writing.syr.edu/~cslipson/419619/LiteratureReviewrevised.pdf. Zhifei, S., Yanhua, C. and Limin, Z. (2005). Micromechanical damage modeling of fibre/matrix interface under cyclic loading. Composites Science and Technology, 65, 1203–1210.

Index

Abrafil device, 161 abrasion fatigue, 109–11 resistance, 110 acoustic emission technique, 212 acoustic polar scans, 221 acoustography, 221 acrylic fibre, 60, 62 active disturbance rejection controller, 154 AFIS measurements, 147 air jet loom, 156 AKZO, 68 alpaca fibres bending-abrasion fatigue, 111, 113–16 SEM images, 97 aramid fibre, 27, 29 as reinforcing fibre for structural composites, 203, 204 torsional fatigue, 69, 70 asbestos, 178 Atlas random tumble pilling test, 122 B-spline function, 196–7 backbone fission, 173 Basquin relation, 136–8, 139–40 BD 200S rotor spinning machine, 151 BDA 10N Elitex rotor spinning frame, 153 bending-abrasion fatigue silk, 116–17 wool and alpaca fibres, 111, 113–16 bending fatigue, 215–16 bias tension test, 216 biaxial rotation fatigue factors affecting fibre fatigue, 75–81 axial splitting development, 75 neutral plane movement, 79 single end driven rotation fibre fracture morphologies, 79–81 torque in testing, 76, 78–9 fibre rotating over a pin, 76 neutral plane movement effect on fatigue characteristics, 80 new developments and future trends, 90 nylon 6.6 effect of strain amplitude and weight, 85 pH on rotational fatigue life, load 2 gm, 86

242

pH on rotational fatigue life, load 12.1 gm, 86 strain amplitude and weight on fatigue lifetimes, 85 water temperature on rotational fatigue life, 87 polyester fibre showing extensive splitting characteristics, 89 tested using biaxial rotation over a pin, 77 principles, 73 rotation over a pin (single end drive), 82–9 apparatus, 82–3 broken Kevlar-29 fibre, 88 comparison of fibres, 84–5 disadvantages and advantages, 82 effect of water, 83–4 failure of fibres in use, 89 fibres tested, drive from one end, 81 high modulus fibres, 88–9 high tenacity nylon, polyester and polypropylene fibre tests, 84 medium tenacity nylon and polyester in air and water tests, 83 pH effect, 85–7 preconditioning fibres effect, 87 strain amplitude and supported weight effect, 85 temperature effect, 87 shear stresses, 75 strong covalent bonds of polymer molecules, 76 suitable types of fibres for testing, 74 techniques for fatigue testing, 74 ten-station fibre over a pin testing apparatus, 83 testing methods, 74–5 textile fibres, 73–91 torque values, 79 ways of reduction, 90–1 block loading tests, 209 Bodine motor, 54 Bragg grating, 212 break-twist-angle test, 107–9 BTA, see break-twist-angle test

Index carbon, as reinforcing fibre for structural composites, 202–3 classic beam theory, 215 composite materials, 201–2 application, 202 fatigue behaviour, 202 matrix, 201 reinforcing fibres, 201 Comptis device, 161 conditioner, for hair care, 128 cotton average flexural fatigue cycles to break single fibre, 107 flexural bending fatigue life, 102 mean flex life and BTA, 109 schematic structure and longitudinal section image, 99 spinning, 146–54 structure and morphology, 98–9 torsional fatigue, 53, 59, 60, 61 ‘crack containment,’ 204–5 crack density, 195 ‘cycle mix effect,’ 225 degradation definition, 170 environmental effects on fatigue fracture sources, 171 steps, 170–1 thermal, 172, 173–4 degumming condition, 116, 117, 125 15-den/dtex nylon 66, 56 digital image correlation (strain mapping), 229 Dref III technology, 161 DSM, 68 Dupont, 68 Dyneema SK65, 68, 69, 70 edge replication, for visualisation of fatigue damage, 219–20 environment aspects of importance, 169–70 definition, 169–70 effects on fatigue fracture, 170–8 chemical attack, 174–5 combined degradative factors, 173 degradation sources, 171 degradation steps, 170–1 gas exposure effects, 174 heat effects, 173–4 manufacturing conditions effects, 176–7 mechanical action, 177–8 microbiological attack, 174 moisture effects, 171–2 radiation effects, 172–3 structural effects, 175 fatigue, 169–85 fatigue failure effects, 178–80 air pollution, 178–9 ancillary concerns, 180 land pollution, 179 safety concerns, 180 visual pollution, 179–80

243

future trends, 182–5 fundamental needs, 184–5 hope for the future, 183–4 industry challenges, 183 initial approaches, 182 limitations, 182–3 short-term prospects, 183 overcoming effects, 180–2 compounded factors, 181 fibre factors, 180–1 society standards, 181–2 treatment factors, 181 epoxy, 202 Eri silk, 117 methods for cocoon degumming, 116 extensometer, 210, 211, 213, 229 FALSTAFF spectrum, 209 fatigue, 6–7 accumulation during processing, 142–5 basic principles, 3–10 future trends, 7–8 S-N and survival diagrams, 7 straight tensile failure vs tensile fatigue failure, 5 stress–strain curve for fibres, 6 damage in structural textile composites, 201–30 data representation, 6–7 definition, 99 effect of textile processing, 133–65 aggressive mechanical opening effect on fibre length distribution, 148 cotton yarn structure, 135 cyclic variation of stress, 136 decomposition of strain life curve, 140 dependence of abrasion resistance, 162 length distribution of cottons, 147 loading spectrum division into segments, 143 major parts of weaving loom, 155 S-N curve, 137 SEM image of raw cotton, 152 strain life curve, 139 surface structure of 29.5 tex cotton yarn, 160 textile fabric structure, 145 thermal damage of PET fibres, 153 warp tension variations, 157 winding tension variation during first layer winding, 150 yarn failure mechanisms, 159 yarn tension during unwinding, 154 yarn tension histogram of the 25-tex yarn, 151 yarn tension variation during unwinding, 155 see also specific kind of fatigue environmental aspects, 169–85 effects of environment on fatigue failure, 170–8 effects of fatigue fracture of environment, 178–80 future trends, 182–5 overcoming environmental effects, 180–2

244

Index

fabric, 162–3 factors, 134–5 failure, 4–5 of fibre, 4–6 factors affecting data, 6–7 general features, 134 life, 133–4 life models, 226 material, 136–42 S-N curve, 136–9 strain life curve, 139–41 thermo mechanical fatigue, 141–2 of materials and structures, 134–45 models fatigue life models, 226 phenomenological models, 226–7, 227–8 progressive damage models, 226, 227, 228 natural fibres, 99–121 phenomena, 135, 146, 162, 223 of polymer-matrix textile composite materials, 188–99 prediction during wearing, 163–5 of textile structures, 145–63 fabric fatigue, 162–3 importance, 133–4 manufacturing processes effects, 146–54 spinning, 146–54 weaving, 154–9 yarn fatigue, 159–62 fatigue strength coefficient, 140 fibre bundle tenacity, 175 fibre creep phenomenon, 126 fibre crimp, 97 fibre fatigue, 4–6 basic principles, 3–10 failure, 4 macro-level, 146 micro-level, 145–6 Fibrestress tester, 113, 117 finite element analysis hybrid, 197 traditional, 196 FIOMAX 2000 ring spinning machine, 161 flax fibres, 35 flax material, 150 flex fatigue, 51 central crack early stage development in nylon 6 fibre, 49 failure obtained under non-ideal conditions, 47 humidity and temperature effects, 44–7, 51 nylon 6 fibres, 45 nylon fibres, 45 polyester, 45 kink bands, 38, 40 formation, 39 methods of achieving flex failure, 36 methods of flexing fibres, 35–8 multi-station tester, 38 natural fibres, 106–7 nylon 6 where a central crack is dominant, 46 polyester where multiple parallel splits are dominant, 46

surface abrasion and crossing kink band marks, 41 temperature effect, 40–4 nylon 6 at various temperatures, 42 nylon 6.6 at various temperatures, 43 polyester, nylon 6 and nylon 6.6, 42 polyester at various temperatures, 44 tensile failure of partially flexed polyester, 48 tensile failure of unflexed and partially flexed nylon 6, 48 of textile fibres, 34–51 diagram, 51 theoretical aspects, 49–50 glass, as reinforcing fibre for structural composites, 202 Goodman diagram, 210, 224 Gossypium barbadense cotton, 108 Goswami’s instrument, 35 Graphical Integrated Numerical Analysis, 196 helix angle, 55 ‘high-cycle fatigue,’ 208 high modulus polyethylene fibre, 31–2 split and fused break of, 32 high-performance polymer fibres torsional fatigue, 67–70 comparison of torsional fatigue lifetimes, 69–70 fracture morphologies, 70 pre-tension and torsion angle effects, 68–9 properties, 68 high-resolution 3D X-ray micro-tomography for visualisation of fatigue damage, 220 Hooke’s Law, 12, 49, 50 hygral fatigue, 128 hysteresis phenomena, 163 ICI pillbox testing, 122 KES-F fabric-testing system, 163 KES-FS, 163 Kevlar 29, 37 broken by rotation over a pin technique, 88 torsional fatigue, 70 Kevlar 129, 68 Kevlar fibre, 27, 29, 30, 31 broken in tension, 29 torsional fatigue, 53, 60, 68 leaky Lamb wave technique, 221 liquid crystal technology, 27, 31 loadcell, 211 ‘low-cycle fatigue,’ 208 Manson–Coffin power law, 140 mass balance principle, 146 mercerisation, 99, 117 cotton, 106 slack, 107, 162 Merino wool yarn, 110 meta-delaminations, 223

Index micro-deformation, 35, 38, 40 Micronaire, 106 Miner rule, 143 Moiré interferometry, 221 muga, 117 mulberry, 117 multiaxial fatigue, 217–18 bending/torsion set-ups, 218 internal pressure/tension, 218 planar biaxial set-ups, 218 tension/torsion set-ups, 218 natural fibres broken fibre ends from bending-abrasion test, 114 cotton flexural bending fatigue life, 102 flexural fatigue cycles to break single fibre, 107 schematic structure and longitudinal section image, 99 cracks due to bending, 115 diameter and bending-abrasion fatigue cycles, 113 effect of lubricant addition on topmaking performance, 123 effect of structure–property relationships on fatigue failure, 95–129 fatigue, 99–121 abrasion, 109–11 bending-abrasion, 111, 113–17 buckling, 109 and fabric pilling, 118–21 factors affecting fibre fatigue life, 117–18 flex, 106–7 and irregularity, 100–1 processing induced, 118 tensile, in some protein fibres, 102–6 twisting and flex fatigue, 107–9 fatigue controlling methods, 121–9 attrition, 124–5 fibre blends, 125 hair care, 125–9 lubrication, 123–4 pilling control through fabric surface treatments, 121–2 fatigue tests, 96 fatigued and non-fatigued fibre materials tensile properties, 104 length and fineness after relaxing fatigued samples, 104 mean flex life and BTA of wool and cotton fibres, 109 pills from fabric tested with Martindale pilling test method, 112 protein fibres recovery behaviour, 105 silk effect of fibre pre-treatments on milling efficiency, 125 fibres fibrillation due to bending abrasion, 117 methods for Eri silk cocoon degumming, 116

245

structure and morphology, 97–9 cotton, 98–9 wool, 97–8 wool and alpaca fibres, 97 carpet yarn fibres buckling fatigue, 110 curvature effect on abrasion, 116 effect of relative humidity on flexural fatigue life, 108 fibre damage caused by saw-tooth and round pin during processing, 119 fibre diameter variation and 3D profiles, 101 fibre tensile recovery behaviour, 103 knitted fabrics after ICI pillbox testing, 122 nylon effect of water, 83–4 flex fatigue, 34–51 as reinforcing fibre for structural composites, 203 nylon 6 flex fatigue, 34–51 nylon 6.6 complementary initiation points of fatigue break, 21 complex truncated fatigue breaks, 25 complimentary ends broken in fatigue, 14 complimentary ends broken in tension, 14 crack initiation inside the fibre, 20 flex fatigue, 34–51 high temperature fatigue breaks, 24 macro and nano-structure, 19 microtomed sections of initiation region of fatigue crack, 23 particle at the point of initiation of conical fatigue crack, 21 tensile fatigue, 10–33 OE-rotor spinning, 152 optical fibre sensing, 229 organic fibre properties and structures, 28 tensile properties, 27 PA 66, see nylon 6.6 PBO chemistry, 27 PEN, see polyethylene naphthalate PET, see polyester; polyethylene terephthalate fibre PID controller, 154 Pierce theory of ‘weakest link,’ 158 pilling, 135 of fibre and fibre fatigue, 118–21 control through fabric surface treatment, 121–2 stages, 120 Poisson’s ratio, 211 pollution definition, 170 fatigue failure effects on environment air, 178–9 land, 179 visual, 179–80

246

Index

polyamide as thermoplastic matrix, 202 polyester as reinforcing fibre for structural composites, 203 saturated, as thermoplastic matrix, 202 unsaturated, as thermosetting matrix, 202 polyester fibre classical tensile or creep fracture morphology, 21 complex truncated fatigue breaks, 25 effect of water, 83–4 energy dissipation during cyclic loading, 18 failure strains and stresses as function of temperature, 26 fatigue lifetimes as function of temperature and stress level, 26 final failure by fatigue occurs behind the fatigue crack tip, 16 occurs from the fibres surface, 16 flex fatigue, 34–51 high temperature fatigue breaks, 24 optical micrograph of truncated fatigue fracture, 24 particle at the crack initiation point, 22 survival graphs, 18 tensile fatigue, 10–33 tongue end of fibre broken at room temperature at 50 Hz, 15 torsional fatigue, 53, 56, 57, 59, 60 deformed at low cyclic torsional amplitude, 61 polyetheretherketone, as thermoplastic matrix, 202 polyetherimide, as thermoplastic matrix, 202 polyethylene, high density, as reinforcing fibre for structural composites, 203 polyethylene naphthalate fibre, 13, 15 advancement of fatigue crack, 17 polyethylene terephthalate fibre tensile fatigue, 10–33 torsional fatigue, 62–3, 65–6 fibrillation or cracks along the axis of fibre, 67 peeled, 64 properties, 63 strength vs. torsional fatigue cycles, 64 tenacity and breaking strain, 71 transverse cracking, 65 polymer-matrix textile composite angle interlock weaves with warp and weft yarns, 189 crack path in laminated and textile composites, 190 detailed failure steps from micrographic images, 193 fatigue behaviour, 188–99 fatigue modelling, 197 fatigue response evaluation, 190–4 five-layer angle interlock weaves comparison of experimental work and model prediction, 198 schematic illustration, 192

modelling of fatigue behaviour, 194–7 crimp model, 196 cumulative damage model, 194, 196 modified shear lag model, 196 progressive damage model, 196 three-layer angle interlock weaves comparison of experimental work and model prediction, 198 schematic illustration, 192 polymeric fibres, fatigue properties and fracture morphology, 56–7 poly(p-phenylene terephthalamide), 27 polyphenylene sulphide, high-end, as thermoplastic matrix, 202 polypropylene as reinforcing fibre for structural composites, 203 as thermoplastic matrix, 202 polyurethane, as thermosetting matrix, 202 protein fibres, tensile fatigue, 102–6 radiography, for visualisation of fatigue damage, 220 Raman Spectroscopy, 19 resin transfer moulding, 195 resistance measurement, 213, 229 rotation over a pin, 82–9 S-N curve, 136–9 scanning electron microscopy torsional fatigue, 56, 70, 71 for visualisation of fatigue damage, 219, 223 servo hydraulic testing machine, 210 shear dominated fatigue, 216–17 shear fatigue, 164 shell gauge tension meter, 157 ‘short fibre content,’ 146 silicon carbide fibres, 173, 175 silk, 100 effect of fibre pre-treatments on milling efficiency, 125 fibres, 35 fibres fibrillation due to bending abrasion, 117 single end drive, 82–9 Single Fibre Analyser, 113 spinning, 146–54 melt, thermoplastic textile fibres produced by, 13–19 tension, 148–9, 150 standard bending theory, 49 static three-rail shear test, 216 strain amplitude, 80 strain life curve, 139–41 Stress Patterns Analysis by the measurement of Thermal Emissions, 221 Sulzer-Ruti Webtester device, 158, 159 survival diagrams, 7 tensile fatigue, 10–33 different ways of conducting tests, 12 fibre failure at temperature and in structures, 23–7

Index high modulus polyethylene, 31–2 liquid crystal, 29–31 mechanisms involved in fatigue, 19–22 Kevlar fibre broken in tension, 29 organic fibre properties and structures, 28 tensile properties, 27 PA 66 fibre complementary initiation points of fatigue break, 21 complex truncated fatigue breaks, 25 complimentary ends broken in fatigue, 14 complimentary ends broken in tension, 14 crack initiation inside the fibre, 20 high temperature fatigue breaks, 24 macro and nano-structure, 19 microtomed sections of initiation region of fatigue crack, 23 particle at the point of initiation of conical fatigue crack, 21 PEN fibre advancement of fatigue crack, 17 PET fibre classical tensile or creep fracture morphology, 21 complex truncated fatigue breaks, 25 energy dissipation during cyclic loading, 18 failure strains and stresses as function of temperature, 26 fatigue lifetimes as function of temperature and stress level, 26 final failure by fatigue occurs behind crack tip, 16 final failure by fatigue occurs from the fibres surface, 16 high temperature fatigue breaks, 24 optical micrograph of truncated fatigue fracture, 24 particle at the crack initiation point, 22 survival graphs, 18 tongue end of fibre broken at room temperature at 50 Hz, 15 principles, 11–13 some protein fibres, 102–6 split and fused break of high modulus polyethylene fibre, 32 of textile fibres, 10–33 thermoplastic textile fibres, 13–19 fracture morphologies, 13, 15–17 loading conditions, 17–19 Zylon fibre regularly spaced compression bands, 31 splitting fibre, 31 textile composite constant-life diagram of constant stress ratio, 208 fatigue damage, 201–30 fatigue of polymer-matrix materials, 188–99 fatigue testing methods, 205–18 bending fatigue, 215–16 multiaxial fatigue, 217–18 parameters inherent to composite specimen, 205–6

247

parameters of fatigue set-up, 206–9 shear dominated fatigue, 216–17 tension-compression and compressioncompression fatigue, 214 tension-tension fatigue, 209–13 future trends and challenges, 228–30 better testing and instrumentation methods, 229–30 faster assessment of new materials fatigue performance, 230 towards multi-scale modelling, 229 materials, 202–5 fibre materials, 202–3 fibre/matrix interface, 203–4 geometrical arrangements of fibre reinforcement, 204–5 matrix materials, 202 mechanical properties reinforcing fibres for polymer composites, 203 resins for (textile) composites, 203 microscopic fatigue processes in plain woven glass/epoxy laminate, 222 modelling strategies for fatigue damage, 223–8 plain weave glass/epoxy composite and damaged 5-harness satin weave carbon/PPS, 221 loaded in bending fatigue, 219 relative resistance evolution during fatigue cycles, 213 strain fields in cruciform specimen for biaxial fatigue loading, 230 typical fatigue damage, 218–23 inspection techniques, 218–21 mechanisms, 221–3 unidirectional glass fabric/epoxy composite permanent shear strain accumulation in cyclic loading, 217 Poisson’s ratio evolution, 211 textile fibre biaxial rotation fatigue, 73–91 flex fatigue, 34–51 kink bands, 38, 40 methods of flexing, 35–8 tensile fatigue, 10–33 thermoplastic, see thermoplastic textile fibre torsional fatigue failure, 53–71 textile processing effect on fatigue, 133–65 textiles effect of processing on fatigue, 133–65 fatigue, 145–63 forms of reinforcement, 204 thermo mechanical fatigue, 141–2 sources of damage, 142 vs isothermal fatigue, 141–2 thermocouples, 229 thermography infra-red thermography, 221 vibrothermography, 221 for visualisation of fatigue damage, 220–1

248

Index

thermoplastic matrices, 202 thermoplastic textile fibre produced by melt spinning, 13–19 fracture morphologies, 13, 15–17 loading conditions, 17–19 thermosetting matrices, 202 through-transmission (C-scan) technique, 220 torsion angle, 55 effect on torsional fatigue life, 68–9 torsional fatigue acrylic fibre, 62 apparatus for single fibres, 55 cotton fibre, 61 cycling mechanism close up view, 55 deformed medium-tenacity polyester fibre cyclic torsional amplitude of 3º, 60 torsional strain amplitude of 10º, 59 effect of strain amplitude on strength reduction, 66 factors affecting fibre fatigue, 56–71 failure in fibres, 53–71 high-performance polymer fibres, 67–70 comparison of torsional fatigue lifetimes, 69–70 fracture morphologies, 70 pre-tension and torsion angle effects, 68–9 properties, 68 influence of fine structure, 62–3, 65–6 effect of draw ratio and strain amplitude, 63, 65 involved mechanisms, 65–6 mechanical degradation, 63 variation of fibre structure with draw ratio, 63 Kevlar fibre, 62 microstructure investigations, 70–1 multi-stage instrumentation with cycling mechanism, 55 PET fibres fibrillation or cracks along the axis of fibre, 67 peeled, 64 properties, 63 strength vs. torsional fatigue cycles, 64 tenacity and breaking strain, 71 transverse cracking, 65 polyester fibre deformed at low cyclic torsional amplitude, 61 polymeric fibres fatigue properties and fracture morphology, 56–7, 59–60 principles, 53 ruptured 15-den/dtex nylon 66 monofilament torsional amplitude ± 47º at 5.2 Hz, time to break 13h (× 165), 57 torsional amplitude ± 47º at 5.2 Hz, time to break 13h (× 370), 58 single-station apparatus, 54 testing methods, 53–6

types of fibres affected, 53 ways of reducing, 71 total hand value, 164 transmission electron microscope, 71 Twaron, 27, 29, 31, 68 Twaron 2000, 68 ultrahigh molecular weight polyethylene, 68, 69, 70 ultrasonic, for visualisation of fatigue damage, 220 V notch, 47 vacuum assisted resin infusion, 195 Van der Waals’ bond, 29, 30, 63 vibrothermography, 221 vinylester, as thermosetting matrix, 202 viscoelastic yarn model, 157 warp tension, 157 weak link theory, 100, 113 weathering, 173 weaving, 154–9 operations beating, 156 fabric take-up, 156 shedding, 155–6 warp beam winding, 155 warp tensioning, 155 weft insertion, 156 Weibull distribution, 100–1 three-parameter, 158, 159 two-parameter, 138, 165, 216, 227 WISPER spectrum, 209 Wöhler curve, 136 wool bending-abrasion fatigue, 111, 113–16 carpet yarn fibres buckling fatigue, 110 curvature effect on abrasion, 116 effect of relative humidity on flexural fatigue life, 108 fibre damage caused by saw-tooth and round pin during processing, 119 fibre diameter variation and 3D profiles, 101 knitted fabrics after ICI pillbox testing, 122 mean flex life and BTA, 109 SEM images, 97 structure and morphology, 97–8 tensile recovery behaviour, 103 yarn breakage phenomenon, 149 fatigue, 159–62 staple fatigue mechanism, 160–1 Young’s Modulus, 27, 29, 30, 50 Zylon fibre, 27, 30, 31 regularly spaced compression bands, 31 splitting fibre, 30

Fatigue Failure of Textile Fibres (Woodhead Publishing Series in Textiles)

Handbook of Textile Fibres (Woodhead Publishing Series in Textiles)

Identification of Textile Fibers (Woodhead Publishing Series in Textiles)

Handbook of Sustainable Textile Production (Woodhead Publishing Series in Textiles)

Smart Fibres, Fabrics and Clothing (Woodhead Publishing Series in Textiles)

Handbook of textile design (Woodhead Publishing Series in Textiles)

Soft Computing in Textile Engineering (Woodhead Publishing Series in Textiles)

Advances in Textile Biotechnology (Woodhead Publishing Series in Textiles)

Biodegradable and sustainable fibres (Woodhead Publishing Series in Textiles)

High performance fibres (Woodhead Publishing Series in Textiles)

Smart Textile Coatings and Laminates (Woodhead Publishing Series in Textiles)

Handbook of Tensile Properties of Textile and Technical Fibres (Woodhead Publishing Series in Textiles)

Recycling Textile and Plastic Waste (Woodhead Publishing Series in Textiles)

Recycling in Textiles (Woodhead Publishing in Textiles)

Physical Testing of Textiles (Woodhead Publishing Series in Textiles)

Knitting Technology (Woodhead Publishing Series in Textiles)

Handbook of Medical Textiles (Woodhead Publishing Series in Textiles)

Chemical Finishing of Textiles (Woodhead Publishing Series in Textiles)

Handbook of Technical Textiles (Woodhead Publishing Series in Textiles)

Surface Modification of Textiles (Woodhead Publishing Series in Textiles)

Polyolefin Fibres: Industrial and Medical Applications (Woodhead Publishing in Textiles)

Friction in Textile Materials (Woodhead Publishing in Textiles)

Textiles in Automotive Engineering (Woodhead Publishing Series in Textiles)

Handbook of Textile Fibre Structure, Volume 1: Fundamentals and Manufactured Polymer Fibres (Woodhead Publishing in Textiles)

Modelling and Predicting Textile Behaviour (Woodhead Publishing in Textiles)

Coated and Laminated Textiles (Woodhead Publishing Series in Textiles)

Technical Textiles Yarns (Woodhead Publishing Series in Textiles)

Interior Textiles: Design and Developments (Woodhead Publishing Series in Textiles)

Digital Printing of Textiles (Woodhead Publishing in Textiles)

Digital Printing of Textiles (Woodhead Publishing in Textiles)

Biomechanical Engineering of Textiles and Clothing (Woodhead Publishing in Textiles)

Fatigue Failure of Textile Fibres (Woodhead Publishing Series in Textiles)

Handbook of Textile Fibres (Woodhead Publishing Series in Textiles)

Identification of Textile Fibers (Woodhead Publishing Series in Textiles)

Handbook of Sustainable Textile Production (Woodhead Publishing Series in Textiles)

Smart Fibres, Fabrics and Clothing (Woodhead Publishing Series in Textiles)

Handbook of textile design (Woodhead Publishing Series in Textiles)

Soft Computing in Textile Engineering (Woodhead Publishing Series in Textiles)

Advances in Textile Biotechnology (Woodhead Publishing Series in Textiles)

Biodegradable and sustainable fibres (Woodhead Publishing Series in Textiles)

High performance fibres (Woodhead Publishing Series in Textiles)

Smart Textile Coatings and Laminates (Woodhead Publishing Series in Textiles)

Handbook of Tensile Properties of Textile and Technical Fibres (Woodhead Publishing Series in Textiles)

Recycling Textile and Plastic Waste (Woodhead Publishing Series in Textiles)

Recycling in Textiles (Woodhead Publishing in Textiles)

Physical Testing of Textiles (Woodhead Publishing Series in Textiles)

Knitting Technology (Woodhead Publishing Series in Textiles)

Handbook of Medical Textiles (Woodhead Publishing Series in Textiles)

Chemical Finishing of Textiles (Woodhead Publishing Series in Textiles)

Handbook of Technical Textiles (Woodhead Publishing Series in Textiles)

Surface Modification of Textiles (Woodhead Publishing Series in Textiles)

Polyolefin Fibres: Industrial and Medical Applications (Woodhead Publishing in Textiles)

Friction in Textile Materials (Woodhead Publishing in Textiles)

Textiles in Automotive Engineering (Woodhead Publishing Series in Textiles)

Handbook of Textile Fibre Structure, Volume 1: Fundamentals and Manufactured Polymer Fibres (Woodhead Publishing in Textiles)

Modelling and Predicting Textile Behaviour (Woodhead Publishing in Textiles)

Coated and Laminated Textiles (Woodhead Publishing Series in Textiles)

Technical Textiles Yarns (Woodhead Publishing Series in Textiles)

Interior Textiles: Design and Developments (Woodhead Publishing Series in Textiles)

Digital Printing of Textiles (Woodhead Publishing in Textiles)

Digital Printing of Textiles (Woodhead Publishing in Textiles)

Biomechanical Engineering of Textiles and Clothing (Woodhead Publishing in Textiles)

Recommend Documents