Zhang, Xiangwu - Fundamentals of Fiber Science-DeStech Publications (2014)

of
Xiangwu Zhang, Ph.D.

Professor
North Carolina State University
Fiber and Polymer Science
Textile Engineering, Chemistry and Science
F u n d a m en t a l s o f F i b er S c i en ce
DEStech Publications, Inc.

4 3 9 N o r t h Du k e S t r e e t
Lancaster, Pen nsyl van ia 1 7602 U.S.A.
Copyrig ht © 2014 by DEStech Publications, Inc.

A l l r i g h t s r es e r v e d
No part of this pub licati on may be reprodu ced, sto red in a

retrieval system, o r transmitt ed, in any fo rm or by an y means,
electro nic, mechan ical, photo cop ying, recording, or otherwi se,
without the prior writ ten permissio n of the pub lisher.
Prin ted in the United States of Ameri ca

10 9 8 7 6 5 4 3 2 1
M ai n en t r y u n d e r t i t l e :
Fund amentals of Fib er Scien ce
A DE S t e c h P u b l i c a t i o n s b o o k
Bibliog rap hy: p .
In c lu d e s in d e x p . 4 0 9
Library of Congress Contro l Nu mber: 20139 56174

ISBN No. 9 78-1-60 595-119 -5
HOW TO ORDER THIS BOOK

BY PHONE: 877-500-4337 or 717-290-1660, 9AM–5PM Eastern Time
BY FAX: 717-509-6100
Order Department
BY MAIL:
DEStech Publications, Inc.

439 North Duke Street
Lancaster, PA 17602, U.S.A.
BY CREDIT CARD: American Express, VISA, MasterCard, Discover
BY W W W SIT E: http://www.destechpub.com
Contents
Preface xiii
Chapter 1. Introduction........................................................................... 1
Problems.......................................................................................... 6
Part I FIBER STRUCTURE
Chapter 2. Chemical Structure of Synthetic Polymer Fibers.................. 9

2.1 Repeating Units................................................................ 9
2.1.1 Concept of Repeating Unit................................ 9
2.1.2 Repeating Units of Synthetic Fibers................ 10
2.2 Molecular Weight........................................................... 12
2.2.1 Average Molecular Weights............................ 12
2.2.2 Effect of Molecular Weight.............................. 15
2.3 Configurations................................................................. 16
2.3.1 Head-to-Head and Head-to-Tail..................... 16
2.3.2 Tacticity........................................................... 17
2.3.3 Skeletal Structure............................................ 20
2.3.4 Copolymers..................................................... 21
2.4 Conformations................................................................ 22
2.5 Bonding........................................................................... 25
2.5.1 Preliminary Bonding....................................... 25
2.5.2 Secondary Bonding......................................... 28
References..................................................................................... 31
Problems........................................................................................ 33
iii
iv Contents
Chapter 3. Physical Structure of Synthetic Polymer Fibers.................. 35

3.1 Solid Phases.................................................................... 35
3.2 Unit Cells........................................................................ 37
3.2.1 Concept of Unit Cell....................................... 37
3.2.2 Unit Cells of Synthetic Polymers.................... 39
3.3 Crystalline Models.......................................................... 42
3.3.1 Fringed Micelle Model.................................... 43
3.3.2 Folded-Chain Model and Switchboard Model... 43
3.3.3 Modified Fringed Micelle Model and
Fringed Fibril Model...................................... 44
3.4 Morphology of Crystallites............................................. 46
3.4.1 Spherulitic Morphology.................................. 46
3.4.2 Fibrillar Morphology...................................... 46
3.4.3 Other Morphologies........................................ 48
3.5 Morphology of Synthetic Fibers..................................... 48
3.5.1 Cross-Sectional Morphology........................... 48
3.5.2 Longitudinal Morphology............................... 50
References..................................................................................... 51
Problems........................................................................................ 52
Chapter 4. Chemical Structure of Natural Polymer Fibers................... 53

4.1 Natural Cellulose Fibers................................................. 53
4.1.1 Cellulose.......................................................... 53
4.1.2 Cotton and Other Natural Cellulose Fibers... 55
4.2 Manufactured Cellulose Fibers....................................... 56
4.3 Natural Protein Fibers..................................................... 59
4.3.1 Protein............................................................. 59
4.3.2 Wool and Silk................................................... 60
4.4 Manufactured Protein Fibers.......................................... 62
References..................................................................................... 62
Problems........................................................................................ 63
Chapter 5. Physical Structure of Natural Polymer Fibers..................... 65

5.1 Natural Cellulose Fibers................................................. 65
5.1.1 Cotton.............................................................. 65
5.1.2 Other Natural Cellulose Fibers...................... 68
5.2 Manufactured Cellulose Fibers....................................... 71
5.3 Natural Protein Fibers..................................................... 74
5.3.1 Wool................................................................. 74
5.3.2 Silk................................................................... 76
Contents v
5.4 Manufactured Protein Fibers.......................................... 78

References..................................................................................... 78
Problems........................................................................................ 82
Chapter 6. Structure of Non-Polymer Fibers........................................ 83

6.1 Carbon Fibers.................................................................. 83
6.1.1 Atomic Order................................................... 83
6.1.2 Morphology..................................................... 85
6.2 Glass Fibers.................................................................... 86
6.2.1 Atomic Order................................................... 87
6.2.2 Composition.................................................... 88
6.2.3 Morphology..................................................... 90
References..................................................................................... 91
Problems........................................................................................ 93
Chapter 7. Structure of Nanofibers....................................................... 95

7.1 A Brief Description of Electrospinning Process............. 95
7.2 Material Type.................................................................. 96
7.2.1 Polymers.......................................................... 96
7.2.2 Carbons........................................................... 98
7.2.3 Ceramics......................................................... 98
7.2.4 Metals.............................................................. 99
7.2.5 Composites...................................................... 99
7.3 Fiber Diameter.............................................................. 101
7.4 Fiber Morphology......................................................... 101
7.4.1 Porous Nanofibers......................................... 101
7.4.2 Core-Sheath Nanofibers................................ 103
7.4.3 Hollow Nanofibers........................................ 103
7.4.4 Multichannel Nanofibers............................... 104
7.4.5 Nanofibers with Other Morphologies........... 104
7.5 Fiber Assemblies........................................................... 105
7.5.1 Nanofiber Nonwovens................................... 106
7.5.2 Nanofiber Filaments and Aligned Nanofiber
Sheets............................................................ 107
7.5.3 Nanofiber Arrays........................................... 108
7.5.4 Nanofiber Yarns............................................. 109
References................................................................................... 110
Problems...................................................................................... 113
vi Contents
Part II FIBER FORMATION
Chapter 8. Flow Behavior of Polymers...............................................117

8.1 Basic Definitions........................................................... 117
8.2 Newtonian and Non-Newtonian Fluids........................ 119
8.2.1 Newtonian Fluid............................................ 119
8.2.2 Non-Newtonian Fluids.................................. 119
8.3 Time-Dependent Fluids................................................ 123
8.4 Flow Behavior of Polymers.......................................... 125
8.4.1 Flow Curves of Polymers.............................. 125
8.4.2 Carreau Equation.......................................... 125
8.5 Capillary Flow of Polymers.......................................... 127
8.5.1 Shear Stress at Wall....................................... 127
8.5.2 Shear Rate at Wall......................................... 129
8.5.3 Calculation of Viscosity................................. 133
8.5.4 Corrections of Flow Curves.......................... 133
8.6 External Factors Affecting the Viscosity of Polymers.... 134
8.6.1 Effect of Temperature.................................... 134
8.6.2 Effect of Shear Rate....................................... 136
8.6.3 Effect of Pressure.......................................... 138
8.7 Structural Factors Affecting the Viscosity of Polymers.....139
8.7.1 Effect of Chain Rigidity and Intermolecular
Interaction..................................................... 139
8.7.2 Effect of Molecular Weight............................ 139
8.7.3 Effect of Molecular Weight Distribution....... 141
8.8 Elastic Effects .............................................................. 142
8.8.1 Die Swelling.................................................. 142
8.8.2 Melt Fracture................................................ 144
8.9 Elongational Viscosity.................................................. 144
References................................................................................... 147
Problems...................................................................................... 148
Chapter 9. Formation of Synthetic Polymer Fibers............................ 151

9.1 Melt Spinning............................................................... 151
9.1.1 Fiber Formation Process.............................. 151
9.1.2 Force Balance............................................... 153
9.1.3 Energy Balance............................................. 158
9.1.4 Molecular Orientation.................................. 162
9.1.5 Crystallization............................................... 164
9.1.6 Cross-Sectional Effects................................. 168
9.1.7 Multifilament Effects..................................... 170
Contents vii
9.2 Solution Spinning......................................................... 171

9.2.1 Dry Spinning................................................. 171
9.2.2 Wet Spinning.................................................. 172
9.3 Gel Spinning................................................................. 174
9.4 Liquid Crystal Spinning................................................ 175
9.5 Dispersion Spinning...................................................... 177
9.6 Reaction Spinning......................................................... 177
9.7 Drawing—An Important “Post-Spinning” Process...... 178
9.7.1 Modes of Drawing......................................... 178
9.7.2 Drawing of Initially Amorphous Filaments.... 179
9.7.3 Drawing of Crystallized Filaments............... 181
9.8 Heat Treatment after Drawing...................................... 182
References................................................................................... 183
Problems...................................................................................... 185
Chapter 10. Formation of Natural Polymer Fibers............................... 187

10.1 Natural Cellulose Fibers............................................... 187
10.1.1 Types of Cotton.............................................. 187
10.1.2 Formation of Cotton...................................... 188
10.2 Manufactured Cellulose Fibers..................................... 191
10.2.1 Derivative Methods....................................... 191
10.2.2 Direct Methods.............................................. 194
10.3 Natural Protein Fibers................................................... 195
10.3.1 Formation of Wool........................................ 195
10.3.2 Formation of Silk.......................................... 197
10.4 Manufactured Protein Fibers........................................ 198
10.4.1 First and Secondary Generations................. 198
10.4.2 Third Generation........................................... 199
References................................................................................... 200
Problems...................................................................................... 203
Chapter 11. Formation of Non-Polymer Fibers.................................... 205

11.1 Carbon Fibers................................................................ 205
11.1.1 PAN Process.................................................. 205
11.1.2 Mesophase Pitch Process.............................. 208
11.1.3 Cellulose Process.......................................... 209
11.1.4 Hydrocarbon Gas Process............................ 210
11.2 Glass Fibers.................................................................. 210
11.2.1 Wool Process................................................. 210
11.2.2 Continuous Filament Process....................... 211
viii Contents
11.2.3 Marble Process............................................. 212

References................................................................................... 214
Problems...................................................................................... 215
Chapter 12. Formation of Nanofibers by Electrospinning.................... 217

12.1 Electrospinning Systems............................................... 217
12.2 Fiber Formation Process............................................... 219
12.3 Solution Properties........................................................ 221
12.3.1 Rheological Behavior.................................... 221
12.3.2 Conductivity.................................................. 222
12.3.3 Surface Tension............................................. 223
12.3.4 Nanoparticle Filler....................................... 223
12.3.5 Salt Additive.................................................. 224
12.3.6 Mixed Solvent................................................ 225
12.4 Operational Conditions................................................. 226
12.4.1 Voltage........................................................... 226
12.4.2 Nozzle Diameter............................................ 226
12.4.3 Nozzle-Collector Distance............................ 226
12.4.4 Solution Flow Rate........................................ 227
12.4.5 Other Operational Conditions...................... 227
References................................................................................... 227
Problems...................................................................................... 230
Chapter 13. Formation of Nanofibers by Other Methods..................... 233

13.1 Centrifugal Spinning..................................................... 233
13.1.1 Fiber Formation Process.............................. 233
13.1.2 Centrifugal Spinning System......................... 235
13.1.3 Fluid Properties............................................ 235
13.1.4 Operational Conditions................................. 236
13.2 Melt Blowing................................................................ 239
13.3 Bicomponent Fiber Spinning ....................................... 240
13.4 Phase Separation........................................................... 241
13.5 Template Synthesis....................................................... 242
13.6 Self-Assembly............................................................... 243
References................................................................................... 245
Problems...................................................................................... 247
Contents ix
Part III FIBER PROPERTIES
Chapter 14. Primary and Secondary Properties.................................... 251

14.1 Primary Properties........................................................ 251
14.1.1 Aspect Ratio.................................................. 251
14.1.2 Strength......................................................... 252
14.1.3 Flexibility...................................................... 254
14.1.4 Cohesiveness................................................. 254
14.1.5 Uniformity..................................................... 255
14.2 Secondary Properties.................................................... 255
14.2.1 Physical Shape.............................................. 255
14.2.2 Density.......................................................... 256
14.2.3 Modulus, Elongation, Elastic Recovery,
and Resilience............................................... 256
14.2.4 Thermal Properties....................................... 259
14.2.5 Electrical Properties..................................... 259
14.2.6 Color and Optical Properties........................ 260
14.2.7 Environmental Properties............................. 260
14.2.8 Other Properties............................................ 261
References................................................................................... 263
Problems...................................................................................... 264
Chapter 15. Mechanical Properties of Fibers........................................ 265

15.1 Basic Definitions........................................................... 265
15.1.1 Hooke’s Law.................................................. 265
15.1.2 Tensile Stress, Strain and Modulus............... 267
15.1.3 True Stress and True Strain........................... 268
15.1.4 Specific Stress................................................ 269
15.1.5 Poisson’s Ratio.............................................. 269
15.1.6 Bulk Modulus................................................ 270
15.1.7 Shear Modulus.............................................. 272
15.2 Tensile Properties.......................................................... 273
15.2.1 Ideal Stress-Strain Behavior......................... 273
15.2.2 A Simple Model............................................. 274
15.2.3 Deviations from Ideal Stress-Strain Behavior...276
15.2.4 Response of Polymer Chains......................... 278
15.2.5 Stress-Strain Behavior of Commercial Fibers..282
15.2.6 External Factors Affecting the Stress-Strain
Behavior........................................................ 284
15.2.7 Structural Factors Affecting the Stress-Strain
Behavior........................................................ 286
15.2.8 Elastic Recovery............................................ 289
x Contents
15.3 Compressive Properties................................................ 292

15.3.1 Longitudinal Compressive Properties........... 292
15.3.2 Transverse Compressive Properties.............. 295
15.4 Torsional Properties...................................................... 296
15.4.1 Torsion of Fibers........................................... 296
15.4.2 Shear Modulus of Fibers............................... 299
15.5 Bending Properties........................................................ 301
References................................................................................... 304
Problems...................................................................................... 305
Chapter 16. Viscoelastic Properties of Fibers....................................... 309

16.1 Molecular Mechanisms of Viscoelastic Behavior........ 309
16.2 Phenomenological Aspects of Viscoelastic Behavior... 311
16.2.1 Creep............................................................. 312
16.2.2 Stress Relaxation........................................... 315
16.2.3 Dynamic Mechanical Properties................... 318
16.3 Time-Temperature Equivalence.................................... 323
16.3.1 Time-Temperature Equivalence..................... 323
16.3.2 Time-Temperature Superposition Principle.. 325
16.4 Models of Viscoelastic Behavior.................................. 327
16.4.1 Ideal Spring and Ideal Dashpot.................... 327
16.4.2 Maxwell Model.............................................. 328
16.4.3 Kelvin-Voigt Model....................................... 330
16.4.4 Four-Element Model..................................... 332
16.4.5 Multi-Relaxation Time Model....................... 333
References................................................................................... 336
Problems...................................................................................... 337
Chapter 17. Thermal Properties of Fibers............................................. 339

17.1 Heat Capacity and Specific Heat.................................. 339
17.1.1 Heat Capacity and Specific Heat.................. 339
17.1.2 Effects of Temperature and Humidity............ 340
17.2 Thermal Conductivity................................................... 342
17.2.1 Longitudinal and Transverse Thermal
Conductivities................................................ 342
17.2.2 Thermal Conductivities of Fiber-Based
Products........................................................ 344
17.3 Thermal Expansion and Contraction............................ 344
17.3.1 Coefficient of Thermal Expansion................. 344
17.3.2 Dimensional Stability and Heat Setting........ 346
Contents xi
17.4 Glass Transition............................................................ 347

17.4.1 Glass Transition Temperature....................... 347
17.4.2 Free Volume Theory...................................... 349
17.4.3 Factors Affecting the Glass Transition
Temperature................................................... 353
17.4.4 Other Transitions.......................................... 356
17.5 Melting.......................................................................... 356
17.5.1 Melting Temperature..................................... 357
17.5.2 Factors Affecting the Melting Temperature.. 358
17.6 Degradation and Decomposition.................................. 361
17.6.1 Degradation vs. Decomposition.................... 361
17.6.2 Decomposition of Polymer Fibers................ 362
17.6.3 Thermally Stable Polymer Fibers................. 363
References................................................................................... 364
Problems...................................................................................... 366
Chapter 18. Electrical Properties of Fibers........................................... 367

18.1 Electrical Conductivity................................................. 367
18.1.1 Basic Concepts.............................................. 367
18.1.2 Electrical Conductivities of Polymer Fibers.....369
18.1.3 Electrically Conductive Polymer Fibers....... 371
18.2 State Electricity............................................................. 372
18.2.1 Basic Concepts.............................................. 372
18.2.2 Mechanisms of Static Charge Generation.... 373
18.2.3 Antistatic Treatments..................................... 376
18.2.4 Utilization of Static Charging....................... 377
References................................................................................... 377
Problems...................................................................................... 379
Chapter 19. Frictional Properties of Fibers........................................... 381

19.1 Basic Concepts.............................................................. 381
19.1.1 Amontons’ Law.............................................. 381
19.1.2 Static Friction vs. Kinetic Friction............... 382
19.2 Nature of Friction......................................................... 384
19.3 Fiber-on-Fiber Friction................................................. 385
19.3.1 Friction Coefficient....................................... 385
19.3.2 Factors Affecting the Friction Coefficient..... 386
19.4 Fiber-on-Other-Material Friction.................................. 387
19.4.1 Friction Coefficient....................................... 387
19.4.2 Factors Affecting the Friction Coefficient..... 390
19.5 Lubrication.................................................................... 390
xii Contents
References................................................................................... 392
Problems...................................................................................... 393
Chapter 20. Optical Properties of Fibers.............................................. 395

20.1 Polarization and Light................................................... 395
20.2 Refractive Index and Birefringence.............................. 397
20.2.1 Refractive Index............................................ 397
20.2.2 Birefringence................................................. 398
20.3 Reflection and Luster.................................................... 402
20.4 Absorption and Dichroism............................................ 404
References................................................................................... 405
Problems...................................................................................... 407
Index 409
About the Author 415

Preface
Fiber science is the study of the formation, structure, and properties of fibers
on various scales, ranging from the atomic to microscopic to macroscopic (large
enough to be visible). It provides the fundamental knowledge for the industrial
application of fibers in a wide range of areas, including apparel, home furnishing,
nonowovens, composites, biomedical materials, energy storage and conversion,
etc. Fiber science is, therefore, an essential part of the education of fiber scientists,
engineers, and technologists in both academic and industry settings.
This book deals with the fundamental aspects of the formation, structure, and
properties of fibers. It starts with the chemical and physical structures of polymer
fibers, non-polymer fibers, and nanofibers, followed by their formation mecha-
nisms. Properties of fibers are then discussed and correlated with their structure
and formation mechanisms to establish the formation-structure-property relation-
ships, enabling readers to advance to more complex engineering and design for
numerous applications and to grasp the underlying concepts of fiber manufactur-
ing techniques. Although the book is arranged to give a sense of direction to the
readers to start from the beginning and proceed to the end, each chapter is self-
contained and can be read independently.
This book covers both polymer and non-polymer fibers, as well as novel nano-
fibers. The book is not intended to provide a comprehensive review on all aspects
of fibers. Instead, it is to provide the background of knowledge and understanding
of fiber science, and to establish the foundation necessary to understand and con-
tribute to the development of the subject in the 21st century. Hence, the book will
be useful to all scientists and engineers involved in academic and industrial re-
search related to different aspects of fiber science and to undergraduate and gradu-
ate students in the fields of fiber science, textiles, composites, polymer science
and engineering, materials science and engineering, and chemical engineering.
Xiangwu Zhang
College of Textiles
North Carolina State University
xiii
CHAPTER 1
Introduction
Fibers are used for making materials for a wide range of applications, including
apparel, home furnishing, nonwovens, composites, biomedical materials, energy
storage and conversion, etc. The study of these materials begins with an under-
standing of the fibers from which they are made. Fiber science is the study of the
formation, structure and properties of fibers on various scales, ranging from the
atomic to microscopic to macroscopic (large enough to be visible). These three
aspects are not independent from each other. Figure 1.1 shows the relationships
among the formation, structure, and properties of fibers. Establishing quantitative
and predictive relationships among the way fibers are formed, their structures and
properties is fundamental to the study of fibers.
There are many different types of fibers. Most fibers have diameters greater
than 1 micrometer, and they can be divided into polymer fibers and non-polymer
fibers. Polymer fibers include synthetic polymer fibers and natural polymer fi-
bers. Synthetic polymer fibers are made from polymers synthesized from raw
Figure 1.1. The triangle shows the interdependence of formation, structure, and properties
of fibers.
1
2 Introduction
materials, such as petroleum-based chemicals or petrochemicals. In general, syn-

thetic polymer fibers are created by forcing, usually through extrusion, polymers
through small holes (called spinnerets) into air or other mediums to form fila-
ments. Natural polymer fibers include those produced by plants and animals. They
are typically biodegradable and can be classified as natural cellulose fibers and
natural protein fibers. Celluloses and proteins also can be modified and extruded
into fibers using methods similar to those used in making synthetic polymer fi-
bers. The resultant fibers are typically called manufactured cellulose and protein
fibers. Non-polymer fibers are those that are not made from polymers, and include
carbon, glass, ceramic, metal, and composite fibers, etc. In addition to traditional
classifications, microscale polymer and non-polymer fibers, nanoscale fibers (i.e.,
nanofibers) have been developed using methods, such as electrospinning, cen-
trifugal spinning, melt blowing, bicomponent fiber separation, phase separation,
template synthesis, and self-assembly. Nanofibers also can be made of different
materials and are typically intended for special applications. Figure 1.2 shows the
fiber classification that is used in this book.
Part 1 of this book focuses on the chemical and physical structures of fibers,
and it starts with synthetic polymer fibers. The fundamental knowledge on the
chemical structure of synthetic polymer fibers is essential for understanding their
physical structure and properties. The chemical structure, such as chain structure
and configuration, of synthetic polymer fibers is determined when the polymer is
synthesized. In general, the process for making the polymer into fibers does not
change the chemical structure. The chemical structure of synthetic polymer fibers
Figure 1.2. Classification of fibers. This is different from the commonly used classification
of textile fibers. In the textile fiber classification, fibers are divided into natural fibers and
man-made fibers.
Introduction 3
does not depend on their shape or morphology, either. In addition to chemical

structure, the physical structure of synthetic polymer fibers also is important in
determining the fiber properties. Unlike chemical structure, the physical struc-
ture does depend on the processing and the final shape of the fibers. Chapters 2
and 3 deal with the chemical and physical structures of synthetic polymer fibers,
respectively.
The chemical and physical structures of natural polymer fibers are more com-
plex than those of synthetic polymer fibers. Two most important building units
for natural polymer fibers are cellulose and protein. Natural cellulose fibers come
from the “stringy” portions of plants ranging from the fine seed fibers of the cot-
ton plant to the coarse pineapple leaf fibers. Natural protein fibers are hairs of ani-
mals, like the sheep and the delicate filaments spun by silkworms and insects. In
addition to these natural fibers, manufactured cellulose and protein fibers also are
based on natural biopolymers. Although these fibers are processed like synthetic
polymer fibers, they are discussed in this book together with natural cellulose
and protein fibers since they have similar chemical structures. The chemical and
physical structures of natural polymer fibers are discussed in Chapters 4 and 5,
respectively.
Chapter 6 addresses the structure of non-polymer fibers. A wide range of
non-polymer fibers, such as carbon, glass, silicon carbide, boron, asbestos, and
metal fibers, now is available commercially. Compared with polymer fibers, non-
polymer fibers often are stronger, stiffer, more heat resistant, and nonflammable.
However, except for metal fibers, non-polymer fibers also are characterized by
their brittleness. These property characteristics are directly related to the atomic
arrangement and the defect structure of non-polymer fibers. Chapter 6 discusses
the structure of two most used non-polymer fibers: carbon and glass fibers.
Nanofibers are an important class of material that is useful in a variety of ap-
plications, including filtration, tissue engineering, protective clothing, compos-
ites, battery separators, energy storage, etc. Nanoscience is the study of atoms,
molecules, and objects with sizes ranging from 1 to 100 nm. However, the term
“nanofibers” has been traditionally used for fibers with diameters less than 1000
nm. Chapter 7 addresses the main structural characteristics of nanofibers.
In Part II, the formation of different fibers is discussed. Synthetic polymer
fibers and manufactured natural polymer fibers can be produced by melt spinning,
solution spinning, gel spinning, liquid crystal spinning, dispersion spinning, etc.
To convert the polymer components into the desired fiber structures, it is impor-
tant to understand the flow behavior of polymers. Chapter 8 describes the basic
knowledge related to the flow behavior of polymers. The use of such knowledge
in the formation of synthetic polymer fibers and manufactured cellulose and pro-
tein fibers is covered in Chapters 9 and 10, respectively. Chapter 10 also addresses
the formation processes of natural cellulose and protein fibers, such as cotton,
wool and silk, which are controlled by the genetic codes and are complex.
4 Introduction
Non-polymer fibers can be produced by many different methods, depending

on the type of materials used. For example, carbon fibers often are made by high
temperature treatment of carbon precursors in an inert atmosphere. Glass fibers
and some ceramic fibers can be directly spun from their melts. Ceramic fibers also
can be obtained by the calcination of ceramic precursor fibers or by the chemical
vapor deposition of precursor gas on a carbon fiber substrate. Chapter 11 focuses
on the formation of carbon and glass fibers.
Electrospinning is the most reported method for producing nanofibers.
Electrospinning uses an electrical charge to draw fine jets from a solution or melt.
It has been combined with other methods to produce nanofibers of polymers, car-
bons, ceramics, metals, and composites. In addition to electrospinning, there are
other methods that can be used to produce nanofibers. For example, centrifugal
spinning utilizes the centrifugal force generated by a rotating spinneret to produce
nanofibers. Melt blowing is typically used to produce fibers with diameters great-
er than 1 μm by using high-velocity air; however, it can produce nanofibers by
using carefully selected materials and processing parameters. Bicomponent fiber
separation, phase separation, template synthesis and self-assembly also are useful
methods for fabricating nanofibers from different materials. Chapter 12 discusses
the fiber formation process and processing-structure relationships of electrospin-
ning. Other nanofiber formation methods are covered in Chapter 13.
Part III of the book discusses different properties of fibers. Fiber properties can
be classified into primary and secondary properties. Primary properties are those
that fibers must possess so they can be converted into useful products. Examples
of primary properties are aspect ratio, strength, flexibility, cohesiveness, and uni-
formity. Secondary properties are those that are desirable and can improve con-
sumer satisfaction with the end-products made from the fibers. Secondary proper-
ties include, but are not limited to physical shape, density, modulus, elongation,
elastic recovery, resilience, thermal properties, electrical properties, color and op-
tical properties, moisture regain, resistance to chemical and environmental con-
ditions, resistance to biological organisms, and resistance to insects. Chapter 14
provides an introduction to these primary and secondary properties.
Among various properties, mechanical properties probably are the most impor-
tant properties of fibers. There are many different types of mechanical properties,
including tensile, torsional, bending, and compressional properties. Among them,
tensile properties are the most intensively studied for fibers, probably because of
their unique shape. However, other types of mechanical properties also are impor-
tant. Chapter 15 first describes the basic definitions of Hooke’s law, stress, strain,
and tensile, bulk and shear moduli, and then gives more detailed discussion on the
tensile, torsional, bending, and compressional properties of fibers.
Fibers often exhibit both viscous and elastic characteristics when undergoing
deformation. As a matter of fact, all materials can exhibit elastic and viscous char-
acteristics simultaneously if the time scale of observation is comparable to the
relaxation times needed for large-scale atomic rearrangements in these materials.
Introduction 5
However, the relaxation times of most non-polymer fibers are significantly greater
than the time scale of normal observations, and hence it is hard to observe their
viscoelastic behavior at room temperature. On the other hand, polymer fibers have
relaxation times that are comparable to the time scale of observation, and they
easily display viscoelastic behavior. Chapter 16, therefore, focuses on the visco-
elastic properties of polymer fibers.
Thermal properties of fiber-based products are important in many applications.
For example, the main function of textile fabrics is to protect the wearer from
cold or heat, and to ensure appropriate heat transfer between the human body and
the environment in order to maintain the physiological thermal balance of the
wearer. Composites also need appropriate thermal properties so they can be useful
in aerospace and space industries. The thermal properties of fibers are the starting
point for understanding the final properties of these products, although many oth-
er factors also play important roles. Chapter 17 deals with the most basic thermal
properties of fibers, including heat capacity, specific heat, thermal conductivity,
thermal expansion and contraction, glass transition, melting, and degradation and
decomposition.
The electrical behavior of non-polymer fibers is varied from excellent electri-
cal conductors (e.g., carbon fibers) to good insulators (e.g., glass fibers). However,
with only a few exceptions, pure polymer fibers are insulators with electrical con-
ductivities in the order of 10–16 S cm–1. Static charges can be easily generated and
accumulated on the surface of polymer fibers. This could lead to serious conse-
quences, such as handling problems during fiber processing, breaking down of
sensitive electronic devices, ignition of flammable vapors and dusts in certain
environments, and clinging tendency and annoying electrical shocks during con-
sumer use. Chapter 18 focuses on the electrical conductivity and static charging
of polymer fibers.
Frictional properties of fibers affect the processing, structure and properties of
all fiber-based products. For example, friction is the force that holds the fibers to-
gether in yarns and fabrics. Without sufficient friction, the strength and structural
integrity of yarns and fabrics will be lost. However, if the friction is too high, it
could cause equipment failure, fiber surface damage, and even fiber breakage.
Chapter 19 addresses the basic principles associated with the frictional properties
of polymer fibers.
The transmission, reflection and absorption of light determine the visual ap-
pearance of an individual fiber. The appearance of fiber assemblies is then the
result of the combined effects of individual fibers, although it also is affected by
the arrangements of fibers. Optical properties of fibers also provide a convenient
measure of many structural characteristics, especially the molecular orientation.
Chapter 20 discusses the practical aspects of the optical properties of fibers.
6 Introduction
PROBLEMS
(1) Read any scientific or engineering paper in fiber science written in the last
five years, and discuss how it advances the field. Provide exact reference:
author, title, journal, volume, page, and year.
(2) Pick at least one type of fiber from the list: polyester, nylon, acrylic, olefin,
cotton, wool, silk, and glass fibers. Then, carry out a literature search to find
out the production and consumption of that fiber in the United States and
in the world.
PART I
Fiber Structure
CHAPTER 2
Chemical Structure of Synthetic Polymer Fibers
The fundamental knowledge on the chemical structure of synthetic polymer fibers

(or simply synthetic fibers) is essential for understanding their physical structure
and properties. The chemical structure, such as chain structure and configuration,
of synthetic fibers is determined when the polymer is synthesized. The process of
making the polymer into fibers typically does not change the chemical structure.
The chemical structure of synthetic fibers does not depend on their shape or mor-
phology, either. In addition to chemical structure, the physical structure of syn-
thetic fibers also is important in determining the fiber properties. Unlike chemical
structure, the physical structure does depend on the processing and the final shape
of the fibers. The physical structure of synthetic fibers also is influenced by their
chemical structure. This chapter focuses on the chemical structure of synthetic
fibers, and the next chapter deals with the physical structure.
2.1 REPEATING UNITS

2.1.1 Concept of Repeating Unit
One essential concept to define the chemical structure of synthetic fibers is the re-
peating unit. The term repeating unit denotes an elementary unit whose repetition
would produce the entire polymer chain, except for the chain ends. Basically, a
polymer chain is constructed by connecting the repeat units together successively
along the main chain, like the beads of a necklace (Figure 2.1). A repeating unit
is not to be confused with the term monomer, which refers to the small molecule,
from which a polymer is synthesized. Another important term for describing the
chemical structure is monomeric unit, which is a structural unit, defined by the
monomer, constituting part of the repeating unit.
9
10 Chemical Structure of Synthetic Polymer Fibers
Figure 2.1. Repeating units connect together successively forming a linear polymer chain.
Repeating unit, monomer and monomeric unit are three different concepts.
They can be coincident, but in many cases, they are different from each other.
Table 2.1 shows three different cases. In some polymers, such as polyethylene and
polypropylene, the monomers do not lose any atoms during the synthesis process,
and the repeating and monomeric units keep the same composition and structure
with the monomers (Case 1). In some other polymers, the monomers lose atoms
during the synthesis process, and the repeating and monomer units differ from the
monomers by a limited number of atoms in the chemical formula (Case 2). One
such example is polyamide 6 (nylon 6), which can be synthesized from monomer
ε-aminocaproic acid. During the synthesis of nylon 6, water molecules are lost
and the resultant repeating and monomeric units have fewer atoms than the cor-
responding monomer structure. There are also polymers, in which the repeating
unit, monomer and monomeric unit are all different from each other (Case 3). For
example, polyamide 6,6 (nylon 6,6) can be synthesized by alternate combination
of two different monomers hexamethylenediamine and adipic acid. As a result,
the repeating unit of nylon 6,6 is formed by two monomeric units, which also are
different from the monomer structures due to the loss of water molecules.
2.1.2 Repeating Units of Synthetic Fibers

The physical structure and properties of synthetic fibers are largely determined by the
repeating units. Polymers with all kinds of repeating units have been developed for
different applications. However, only a few of these polymers have been made into
synthetic fibers. This is because to make useful fibers, the polymers have to meet cer-
tain structure requirements, which will be discussed later. Table 2.2 shows the repeat-
ing units for some important polymers that can be made into synthetic fibers.
Table 2.1. Comparison of monomers, monomeric units, and repeating units.
Monomeric Polymer
Polymer Monomer Unit Repeating Unit Structure
Case 1: Monomer = Monomeric Unit = Repeating Unit
Polyethylene
(PE)
Polypropylene
(PP)
Case 2: Monomer ≠ Monomeric Unit = Repeating Unit
Polyamide 6
(Nylon 6)
Case 3: Monomer ≠ Monomeric Unit ≠ Repeating Unit
Polyamide 6,6
(Nylon 6,6)
Repeating Units 11
Table 2.2. Repeating units of some important polymers.

Type Polymer Name Repeating Unit
Polyethylene (PE)
Polyolefins
Polypropylene (PP)
Polyvinyl chloride (PVC)
Polyvinylidene chloride (PVDC)
Polyvinylidene fluoride (PVDF)
Vinyl Polytetrafluoroethylene (PTFE)

Polymers
Polyvinyl alcohol (PVA)
Polystyrene (PS)
Polyacrylonitrile (PAN)
Polyvinylidene dinitrile
Polyethylene terephthalate (PET)
Polytrimethylene terephthalate (PTT)
Polyesters Polybutylene terephthalate (PBT)
Polyethylene naphthalate (PEN)

n
Poly-1,4-cyclohexylene
-dimethylene terephthalate (PCDT)
Nylon 6
Nylon 6,6
Polyamides
Nylon 11
Nylon 6,10
Polyparaphenylene terephthalamide
(Kevlar®)
Aramids
Polymetaphenylene isophthalamide
(Normax®)
Type Polymer Name Repeating Unit
Polyphenylene sulfide (PPS)
Polycarbonate (PC)
Polybenzimidazole (PBI)
Others
Polyparaphenylene benzobisoxazole (PBO)
Polyparaphenylene benzobisthiazole (PBT)
Polyhydroquinone-diimidazopyridine (M-5)
2.2 MOLECULAR WEIGHT

The size of polymer chains can be characterized by measuring the molecular
weight (M). Typically, the size of polymer chains increases with increase in mo-
lecular weight. The molecular weight of a polymer is determined by the degree of
polymerization (N), i.e., the number of repeating units in the polymer chain, by
the following equation:
M = NM 0 (2.1)
where M0 is the molecular weight of the repeating unit.
2.2.1 Average Molecular Weights

With very few exceptions, the molecular weight of a polymer does not have a
single value because different polymer chains can have different lengths. Hence,
there is a distribution of molecular weights. The molecular weight distribution is
typically shown graphically by plotting the number of polymer chains against the
molecular weight, as exemplified in Figure 2.2.
The distribution curve of molecular weights is essential for some applications.
However, in most practical uses of polymers, it is more convenient to character-
ize the distribution in terms of average molecular weights. Since the molecular
weight of a polymer changes in intervals of M0, the distribution of molecular
weights is actually discontinuous. As a result, the average molecular weights are
defined by assuming the polymer chains exist in discrete fractions i containing Ni
molecules of molecular weight Mi.
Molecular Weight 13
Figure 2.2. A distribution of molecular weight along with various average molecular

weights.
Average molecular weights can be defined several different ways. Among

them, the two most important ones are number average molecular weight ( M n )
and weight average molecular weight ( M w ). The number average molecular
weight is defined as the total weight of the polymer divided by the total number of
the molecules, i.e.,
∑ NM
∞
Total Weight i =1 i i
Mn = = (2.2)
∑ N
∞
Total Number i
i =1
∞
Here, the term N i / ∑N i is physically the number fraction (xi) of polymer chains
i =0
with molecular weight Mi. Therefore, Equation 2.2 can be rewritten as:
∞
M n = ∑xi M i (2.3)
i =1
Many polymer properties are not just dependent on the number of polymer
chains, but on the size or weight of each polymer chain. In addition, many mo-
lecular weight measurements, such as light scatting and ultracentrifuge methods,
are based on the contributions of polymer chains according to their sizes. For
these cases, we need to use the weight average molecular weight, which can be
described as:
∑
∞
N i M i2
Mw = i =1
(2.4)
∑
∞
i =1
Ni M i
∞
In Equation 2.4, N i M i / ∑N i M i is the weight fraction (wi) of polymer chains
i =1
with molecular weight Mi. Hence, an alternative form for the weight average mo-
lecular weight is:
∞
M w = ∑wi M i (2.5)
i =1
Comparing Equations 2.3 and 2.5, it is clear that M n is the average Mi weighed
according to the number fraction and M w is the average Mi weighed according to
the weight fraction.
In addition to M n and M w , higher molecular weight averages are also used.
For example, M w can be derived from M n simply by replacing Ni with NiMi in
Equation 2.2. This process can be generalized to replace Ni by N i M ik to obtain an
average molecular weight denoted as M k :
∑
∞
N i M ik +1
Mk = i =1
(2.6)
∑
∞
N M ik
i =1 i
From Equation 2.6, it is seen that M n = M 0 and M w = M 1 . When k is greater

than 1, z average molecular weight M z and z+1 average molecular weight M z +1
can be defined as:
∑
∞
N i M i3
Mz = M2 = i =1
(2.7)
∑
∞
i =1
N i M i2
and
∑
∞
N i M i4
M z +1 = M 3 = i =1
(2.8)
∑
∞
N M i3
i =1 i
which often are used in ultracentrifugation experiments.

One average molecular weight that often is used, but does not fit into the form
of M k is the viscosity average molecular weight M v :
Molecular Weight 15
1
 ∑ ∞ N M 1+ α  α
i i
M v =  i =∞1  (2.9)
 ∑ N i M i 
i =1
where α is a constant that is determined by the polymer/solvent system used dur-

ing the viscosity measurements.
The relationship between various average molecular weights is shown in
Figure 2.2. It is seen that, the average molecular weights rank in the order:
M n ≤ M v ≤ M w ≤ M z ≤ M z +1 (2.10)
The equalities hold only for monodisperse polymers, i.e., only when all polymer
chains have the same molecular weight. As discussed previously, most polymers
are polydisperse since different polymer chains can have different lengths. Hence,
the different average molecular weights are all different from each other, and they
rank in the order shown in Equation 2.10. Therefore, the difference between the
average molecular weights is a good indication of the distribution of molecular
weight. For example, the ratio of the weight average molecular weight to the num-
ber average molecular weight is known as the polydispersity index (PDI).
Mw
PDI =
Mn (2.11)
The polydispersity index is always greater than or equal to one. It is equal to one
only if the polymer is monodisperse. For polydisperse polymers, the polydisper-
sity index is greater than one. A larger polydispersity index indicates a broader
distribution of molecular weights.
2.2.2 Effect of Molecular Weight

Properties of synthetic fibers are closely related to their molecular weight. To
obtain useful fibers, the molecular weight of a polymer must exceed a critical
value. Below the critical molecular weight, the polymer does not have the physi-
cal properties needed for fiber application. Beyond the critical molecular weight,
the mechanical strength of the resultant fibers increases rapidly with increases in
molecular weight. However, the increase of mechanical strength becomes lower
at high molecular weights, and the strength eventually becomes relatively con-
stant. In addition to mechanical strength, many other physical properties, such as
modulus and glass transition temperature, have the same trend. Figure 2.3 shows
the effect of molecular weight on fiber properties.
Although higher molecular weight leads to better fiber properties, synthetic

fibers are not always made from polymers that have very high molecular weights.
With increase in molecular weight, the polymer chains tend to entangle more,
which significantly increases the viscosity of polymer melt or solution. Eventually,
the viscosity of the polymer melt or solution becomes so high and the formation
of fibers is no longer possible or economic. Therefore, there is a certain molecular
weight range that is suitable for fiber application, as shown in Figure 2.3.
2.3 CONFIGURATIONS
Although there are many different polymers, not all of them can be made into
useful fibers. To be made into useful fibers, the polymer chains must be linear
and the molecular weight needs to be high enough for mechanical stability, but
low enough for dissolving or melting for forming fibers. The polymer chains also
must have the appropriate configuration. Configuration refers to the permanent
geometry resulted from the spatial arrangement of bonds in a polymer chain. Only
polymer chains with certain configuration can form fibers with desirable physical
structure and properties.
2.3.1 Head-to-Head and Head-to-Tail

One simple but important configuration type is the head-to-head or head-to-
tail placement of repeating units along the polymer chain. Figure 2.4 show the
Figure 2.3. Effects of molecular weight on the property of synthetic fibers and the viscosity
of polymer melt or solution.
Configurations 17
head-to-head and head-to-tail configurations of polypropylene, in which the α

carbon connected to the methyl group is considered as the head and β carbon as
the tail.
The head-to-tail configuration is the thermodynamically and spatially preferred
structure for most polymers that contain side groups. However, it is possible to
deliberately design the synthesis process to obtain polymer chains with head-to-
head configuration. Polymers with head-to-tail and head-to-head configurations
have different properties. For example, poly-isobutylene with the head-to-tail
configuration has a glass transition temperature of –61°C and a melting tempera-
ture of 5°C; it is an excellent rubber material. On the other hand, poly-isobutylene
with the head-to-head configuration has a glass transition temperature of 87°C
and a melting temperature of 187°C; it cannot be used as a rubber. In fiber ap-
plication, synthetic fibers are typically made from polymers with the head-to-tail
configuration.
2.3.2 Tacticity
Tacticity is the relative stereochemistry of adjacent chiral centers within a poly-
mer chain. The term chiral here is used to describe a structure that is not super-
imposable on its mirror image. Figure 2.5 shows an example of a chiral center
and shows that the chemical structures on both sides of the mirror have identical
chemical composition, but they represent different spatial configurations since
they are nonsuperimposable. The cause of the two different configurations is the
asymmetric carbon in the center of the structure, known as the chiral center.
For a carbon atom to become a chiral center, it must be linked to four different
atoms or groups. Figure 2.6 shows a typical chiral center in a polymer chain. If
the C* atom is not in the middle of the polymer chain, the two chain segments on
each side are of unequal length and can be considered to be different groups. As
Figure 2.4. Head-to-tail and head-to-head configurations of polypropylene.

Figure 2.5. Illustration of a chiral center.
a result, the C* atom becomes a chiral center if the R group is different from the
H atom. For polymers with chiral centers, the different spatial arrangements lead
to their tacticity.
Many polymers, such as polypropylene, polyvinyl chloride, and polyacryloni-
trile, have R groups on every other carbon atom. If the chiral centers connected
to all these R groups have the same configuration, the polymer is called isotactic.
If the chiral centers alternate in configuration, the polymer is called syndiotactic.
The polymer becomes atactic if the chiral centers have random order in configura-
tion (Figure 2.7).
Figure 2.7 illustrates the three-dimensional structure of isotactic, syndiotactic
and atactic configurations. The differences between these configurations also can
be presented in two dimensions (Figure 2.8). The two-dimensional illustration is
made by using Fisher projections, in which the R groups are placed either up or
down. All up or all down represent the isotactic configuration. Alternating up and
down represents syndiotactic configuration. The atactic configuration is shown as
random up and down.
The tacticity of polymers has important effects on their physical structure and
properties. Atactic polymers usually are amorphous unless the side group is ex-
tremely polar and allows some crystallinity. On the other hand, both isotactic and
syndiotactic structures can crystalize due to their regularity along the polymer
Configurations 19
Figure 2.6. A chiral center (C*) in a polymer chain.
Figure 2.7. Three-dimensional illustration of isotactic, syndiotactic and atactic configura-

tions. The triangular and dotted lines indicate bonds to atoms or groups above or below the
plane of the carbon-carbon backbone, respectively.
chains. However, their unit cells and melting temperatures are different. As a re-
sult, to make synthetic fibers, it is important to select polymers with appropriate
tacticity. For example, polypropylene fibers typically are made from the isotactic
structure, which is highly crystalline and has a melting point of at least 160°C due
to the closely packed regular chains. Syndiotactic or atactic polypropylene is not
Figure 2.8. Two-dimensional illustration of isotactic, syndiotactic and atactic configurations.
being made into synthetic fibers since they either have low degree of crystallinity
or are completely amorphous.
2.3.3 Skeletal Structure

There are many different skeletal structures. Figure 2.9 shows three most impor-
tant skeletal structures: linear, branched, and network. In linear polymers, repeat-
ing units are arranged in a chainlike fashion with no branches or bridges between
the chains. On the other hand, branched polymers have side chains, i.e., branches,
of significant length that are bonded to the main chains. Network polymers con-
tain branches that connect different polymer chains into a “network”.
Synthetic fibers typically are made from linear polymers since they can be
aligned and packed closely during fiber formation to achieve the desired physical
structure and properties. Branched polymer chains are hard to pack and hence
they are seldom made into synthetic fibers. Network polymers do not melt or
Configurations 21
Linear Branched Network

Figure 2.9. Linear, branched and network skeletal structures.
dissolve, and they cannot be spun into fibers. However, after linear polymers are
made into fibers, they can be crosslinked to form a network structure to alter the
fiber properties by using chemical crosslinkers or radiation.
2.3.4 Copolymers
The polymers discussed so far in this chapter are all homopolymers that are com-
posed of a single type of repeating unit. However, a large and growing number
of commercial polymers actually are composed of different types of repeating
units. These polymers are called copolymers. It is an important strategy for fiber
engineers to manipulate the formation and properties of synthetic fibers by using
copolymers. This is because it is time consuming and expensive to develop a new
homopolymer for a specific application. However, it is much more time and cost
effective to obtain a copolymer with desired properties by simply introducing a
second repeating unit to an existing homopolymer.
There are different types of copolymers, depending on the particular arrange-
ment of the repeating units along the polymer chain. Figure 2.10 shows four dif-
ferent types of copolymers:
• Random copolymers, in which the distribution of repeating units is random;

• Alternating copolymers with two types of repeating units arranged alter-
nately along the polymer chain;
• Block copolymers, in which the repeating units exist only in long sequenc-
es, i.e., blocks, of the same type;
• Graft copolymers, a special type of branched polymers in which the
branches are composed of repeating units that are different from those in the
backbone.
Random and alternating copolymers typically exhibit properties that are be-
tween those of the corresponding homopolymers. Therefore, they often are de-
veloped to combine the properties of homopolymers into a single fiber. This is
difficult to achieve by simply blending the homopolymers because most polymers
are immiscible with each other. For example, fibers of polyacrylonitrile homopol-
ymer have been used in hot gas filtration systems, outdoor fabrics, and precursors
for carbon fibers. However, many commercial acrylic fibers actually are made
Figure 2.10. Random, alternating, block and graft copolymers.
of copolymers that contain at least 85% acrylonitrile repeating units. During the
synthesis of these copolymers, either neutral and/or ionic co-monomers are added
to change the fiber morphology, provide dye sites, or improve water sorption.
Block and graft copolymers often show the characteristic of each of the ho-
mopolymers, and they also may have some unique new properties caused by the
chemical bonds between the homopolymer segments. Block and graft copoly-
mers often are used to make synthetic fibers for special applications. For example,
polypropylene fibers are mechanically strong and are a good reinforcement for
polymer composites. However, polypropylene is inert and does not form strong
intermolecular interaction with the matrix, which could lead to the failure of the
composite. One approach to address this problem is to add different monomers
during the melt spinning of polypropylene fibers to form a graft copolymer, which
can have good intermolecular interaction with the matrix. Composites using these
polypropylene-based copolymer fibers are significantly stronger and more dura-
ble than those made of polypropylene homopolymer fibers.
2.4 CONFORMATIONS
Polymer chains shown from Figure 2.7 to 2.10 are all static representations of real
chains. Static representations do not show an important aspect of real polymer
chains, i.e., their oscillation and movement caused by the thermal vibrations of
the polymer chain structure. With an increase in temperature, the motions of poly-
mer chains increase in both frequency and amplitude. There are many different
Conformations 23
motions of the polymer chain structure, and among them the most important one
probably is the rotation of single bonds. Double and triple bonds are rigid and can-
not rotate without being broken. Only single bonds can rotate, which lead to the
movement of adjacent atoms. Conformation can, therefore, be defined as the order
that arises from the rotation of molecules around the single bonds.
Conformation and configuration are two diffrent concepts. The configuration
of a polymer cannot be altered unless chemical bonds are broken and reformed.
When a polymer changes its configuration, it turns into a different polymer.
However, the conformation of a polymer can be changed by rotating around the
single bonds, during which no bonds are broken or formed. Hence, the polymer
remains the same when its conformation changes.
During the rotation of single bonds, there are preferred low-energy conforma-
tions. Using a simple molecule of n-butane as an example, Figure 2.11 shows
the potential energies related to different conformations. The n-butane molecule
consists of four carbon atoms linked together in a linear chain with the hydrogen
atoms along the periphery. When the adjacent hydrogen atoms are aligned, there
will be considerable interference among them, leading to the high-energy eclipsed
conformation, which is not desirable. The interference can be significantly lowered
when they rotate around their axis to a staggered conformation (Figure 2.11A). By
rotating the single bond connecting the two central carbon atoms, it is seen that
there are two positions where the outer carbon atoms can be located to lower the
interference. One position where the outer carbons are in opposite direction is
called the trans conformation, and the other where they are adjacent is called the
gauche conformation (Figure 2.11B). Depending on the rotational angle, there is
one trans conformation for a single bond in the center of n-butane molecule, but
two possible gauche conformations, named as gauche (-) and gauche (+). From
Figure 2.11C, it is clear from the n-butane model that the trans conformation has
lower energy than both gauche (-) and gauche (+) conformations.
As compared with small molecules such as n-butane, polymer chains have
many more single bonds on the backbone. Each single bond can have one trans
and two gauche conformations, and hence a polymer chain with N bonds on the
backbone can have 3N different conformations. For example, a polyethylene chain
with 10,000 bonds can have 310,000 (= 104,800) different conformations. When a
polymer chain changes the conformation, its size and shape also change. As a
result, the ability of polymers to change their conformations leads to an internal
degree of freedom that is not available to small molecules.
Synthetic fibers typically are semicrystalline. During the fiber formation pro-
cess, the polymer chains change their conformations to form the crystalline phase.
After the fiber formation, the polymer chains in the crystalline phase are largely
in trans conformations, which remain unchanged until the crystalline structure
is damaged or altered. However, in the amorphous phase, the polymer chains
still can change their conformations after the fiber formation. For example, the
(A)
(B)
(C)
Figure 2.11. (A) Eclipsed and staggered conformations, (B) trans and gauche conforma-
tions, and (C) potential energy of n-butane molecule.
drawing of a synthetic fiber can change some single bonds from gauche to trans
conformation, leading to a better aligned structure.
In the amorphous phase of synthetic fibers, the ratio (Ngauche/Ntrans) of gauche
and trans conformations can be estimated by
N gauche  ∆E 
= 2exp  − (2.12)
N trans  kT 
Bonding 25
where ∆E is the energy difference between the two conformations, k the Boltzmann
constant, and T the absolute temperature. It is seen that with increase in tempera-
ture, the Ngauche/Ntrans ratio increases. The result is the polymer chains tend to con-
tract in size at elevated temperatures and change to a less ordered structure.
Many properties of synthetic fibers are related to how many possible confor-
mations the polymer chains can have. However, not all polymer chains have the
same number of possible conformations even if they have the same chain length.
Figure 2.12 compares a rigid rod-like polymer chain and a flexible random coil
polymer chain. Rigid rod-like polymer chains typically assume more extended
conformations than flexible polymer chains, and as a result, fibers made from
rigid polymers are generally stiffer and have a higher glass transition temperature.
One example of a rigid rod-like polymer fiber is Kevlar®.
2.5 BONDING
Repeating units are the smallest building blocks discussed so far in this chap-
ter. Repeating units are formed by atoms, which are held together by bonds. In
polymer fibers, there are two categories of bonds, i.e., primary and secondary
(Figure 2.13). Primary bonds link the atoms together to form repeating units and
eventually the polymer chains. Secondary bonds, also called intermolecular inter-
actions or intermolecular bonds, are significantly weaker than preliminary bonds,
but they are critical for binding the polymer chains into fibers. In polymer fi-
bers, all primary bonds are covalent. However, there are many different types of
secondary bonds, including ionic secondary bond, hydrogen bond, dipole-dipole
force, aromatic ring association, and van der Waals force.
2.5.1 Preliminary Bonding

Primary bonds are the attractive forces that hold atoms together to form polymer
chains. The breaking of primary bonds leads to the degradation or even decompo-
sition of the polymer. In synthetic fibers, the primary bonds typically are covalent
bonds, which are formed by the sharing of pairs of electrons between atoms. The
electrons being shared are called valence electrons and are in the highest energy
(A) (B)
Figure 2.12. (A) Rigid rod-like polymer chain, and (B) flexible random coil polymer chain.
Figure 2.13. Primary and secondary bonds.
level of the atom, i.e., outer shell. The number of valence electrons of an atom
is the same as the group number in the periodic table. Table 2.3 shows the group
numbers and valence electrons of the atoms that often are seen in synthetic fi-
bers, including hydrogen, carbon, nitrogen, oxygen, sulfur, fluorine, and chlorine.
During the formation of covalent bonds, these atoms share the valence electrons
in a way that enables them to acquire a stable electronic configuration, i.e., full
valence shell. This means that these atoms tend to have eight electrons at the
outer shell (except H, which tends to have two electrons) by sharing electrons.
To acquire the stable electronic configuration, two atoms can share two electrons
to form a single bond. It also is possible for two atoms to form double bonds by
sharing four electros, or triple bonds by sharing six electrons.
The number of covalent bonds that an atom can form is determined by how
many valence electrons it has. For example, hydrogen can form one bond, carbon
can form four, nitrogen three, oxygen two, etc. Figure 2.14 shows the formation of
covalent bonds in a small molecule: methane. It is seen that each carbon atom pro-
vides four valence electrons to be combined with the four electrons from four dif-
ferent hydrogen atoms. In the resultant methane molecule, the carbon atom shares
eight electrons on its outer shell while each hydrogen shares two electrons. Thus,
all atoms in methane achieve their stable electronic configuration. The ethane
structure is the result of the formation of seven covalent bonds by sharing fourteen
valence electrons among two carbon and six hydrogen atoms, while polyethylene
Table 2.3. Group numbers and valence electrons of

atoms that often are seen in synthetic fibers.
Atom Group Number Valence Electron

Hydrogen (H) Group I 1 valence electrons: 1s1
Carbon (C) Group IV 4 valence electrons: 2s2, 2p2
Nitrogen (N) Group V 5 valence electrons: 2s2, 2p3
Oxygen (O) Group VI 6 valence electrons: 2s2, 2p4
Sulfur (S) Group VI 6 valence electrons: 3s2, 3p4
Fluorine (F) Group VII 7 valence electrons: 2s2, 2p5
Chlorine (Cl) Group VII 7 valence electrons: 3s2, 3p5
Bonding 27
has 3n + 1 covalent bonds formed by the sharing of 6n + 2 electrons among n

carbon and 2n + 2 hydrogen atoms (Figure 2.15).
Two important characteristics of covalent bonds are length and strength. Bond
length is defined as the distance between the centers of two bonded atoms, while
bond strength is the amount of energy required to break a bond. The length and
strength of covalent bonds depend on many factors, e.g., electron affinities, sizes
and electronegativity of atoms. However, the length and strength of the same
bond in various molecules are consistent. Table 2.4 compares the lengths and
strengths of the covalent bonds that are commonly seen in synthetic fibers.
The strength of covalent bonds is directional. Hence, the bonds in a molecule
have the tendency to maximize their separation in space. Figure 2.16 shows the
possible geometry of the covalent bonding of methane and polyethylene. In meth-
ane, the four covalent bonds maximize their separation by forming a tetrahedral
geometry. The bond angles are 109.5°. In the case of polyethylene, the maximiza-
tion of bond separation theoretically gives the polymer chain a three-dimensional
zigzag geometry. However, since the polymer chain can change its conformation
by rotating the single bonds, it has many other geometries, but the bond separation
is always maximized.
Figure 2.14. Formation of covalent bonds in methane.
Figure 2.15. Covalent bonds in ethane and polyethylene.

Table 2.4. Bond lengths and bond strengths of covalent bonds in synthetic fibers.
Bond Length (pm) Strength (kJ/mol)
C-C 154 348
C-H 109 413
C-N 147 308
C-O 143 360
C-F 135 488
C-Cl 177 330
C=O 122 745
C≡N 115 890
N-H 101 391
O-H 96 366
2.5.2 Secondary Bonding

Secondary bonds are the attractive forces that hold polymer chains together to
form fibers. Secondary bonds also are called intermolecular interactions or inter-
molecular bonds. In fibers, the polymer chains can change their conformation by
breaking and reforming secondary bonds. The melting of synthetic fibers is in part
determined by the total strength of secondary bonds. There are different types of
secondary bonds, and the most important ones are ionic secondary bond, hydro-
gen bond, dipole-dipole force, aromatic ring association, and van der Waals force.
The bond lengths of secondary bonds are longer than those of primary bonds, and
they are in the order of 300–500 pm. The bond strengths of secondary bonds are
Figure 2.16. Geometry of covalent bonding in methane and polyethylene with all-trans

conformation.
Bonding 29
Table 2.5. Bond strengths of primary and secondary bonds.

Bond Strength (kJ/mol)
Covalent Bond 300–400
Ionic Secondary Bond 100–350
Hydrogen Bond 10–60
Dipole-Dipole Force 5–30
Aromatic Ring Association 5–30
Van der Waals force <5
weaker than those of primary bonds. Table 2.5 compares the bond strengths of
primary and secondary bonds.
Among all secondary bonds, ionic bond is the strongest, with the bond strength
ranging from 100 to 350 kJ/mol. An ionic secondary bond is a type of chemical
bond formed through an electrostatic attraction between two oppositely charged
atoms on two neighboring polymer chains (Figure 2.17). Unlike in a covalent
bond where two atoms acquire the stable electronic configuration by sharing elec-
trons at the outer shells, the atoms involved in an ionic secondary bond have very
different electronegativities and one of the atoms lose its electron(s) to the other
atom. Therefore, the formation of ionic secondary bonds results from the transfer
of electrons between atoms on adjacent polymer chains. Ionic secondary bonds
are just slightly weaker than covalent primary bonds. Ionic secondary bonds can
be found in natural polymer fibers, e.g., wool. However, ionic secondary bonds
are not common in synthetic polymer fibers.
The next strongest secondary bond is hydrogen bond. Hydrogen bond is the at-
tractive intermolecular interaction between a hydrogen atom and an electronega-
tive atom, such as nitrogen, oxygen, or fluorine. Not all hydrogen atoms can form
hydrogen bonds. For a hydrogen bond to be formed, the hydrogen atom also must
be linked to an electronegative atom of oxygen, nitrogen or fluorine, which is
called the hydrogen-bond donor. This hydrogen-bond donor attracts the electrons
shared with the hydrogen atom and leaves the hydrogen a positive partial charge.
The positive partial charge then attracts a lone pair of electrons on the hydrogen-
bond acceptor to form a hydrogen bond. In synthetic fibers, hydrogen bonds are
mainly formed between OH groups (e.g., polyvinyl alcohol) and between >C=O
Figure 2.17. Example of ionic secondary bonding between polymer chains.

and NH groups (e.g., nylon). Figure 2.18 shows the formation of hydrogen bonds
in polyvinyl alcohol and nylon. The bond strengths of hydrogen bonds are be-
tween 10–60 kJ/mol. The formation of hydrogen bonds affects the crystallization
process and mechanical properties of synthetic fibers. In addition, synthetic fibers
that can form hydrogen bonds tend to have better moisture absorption properties.
A third type of relatively strong secondary bond is dipole-dipole force, which
is the attractive interaction between the positive end of a polar group and the
negative end of another polar group. Dipoles occur due to the unequal sharing
of electrons between atoms in polar groups. In a polar group, the atom that is
more electronegative than others pull the electrons closer to itself to form a
dipole, in which one side of the group possesses a partially negative charge
and the other side a partially positive charge. For example, in polyacrylonitrile,
the -C≡N group is a polar group, in which the carbon atom is partially positive
and the nitrogen is partially negative. When two -C≡N get close enough, the
dipole-dipole force is formed by the attraction of the partially positive carbon
and partially negative nitrogen in two groups (Figure 2.19). The strengths of
dipole-dipole forces are weaker than those of hydrogen bonds, and typically are
in the range of 5–30 kJ/mol.
Another type of secondary bond that has similar strength with dipole-dipole
force is aromatic ring association. The aromatic ring association, also called π-π
stacking, is the attraction interaction between two aromatic rings. The delocal-
ized π electrons on aromatic rings are somewhat mobile. The shifting of these
delocalized π electrons leads to the formation of attraction interaction between
two neighboring aromatic rings. The aromatic ring association can be found in the
Figure 2.18. Hydrogen bonds in polyvinyl alcohol (left) and nylon (right).
Figure 2.19. Dipole-dipole force in polyacrylonitrile.

References 31
crystallization region of some synthetic fibers, such as polyethylene terephthalate

(PET), where the aromatic rings stack regularly on each other. Figure 2.20 shows
the aromatic ring association between two aromatic rings in two adjacent PET
polymer chains.
The weakest secondary bond is the van der Waals force. The formation of
van der Waals force lies in the instantaneously induced dipoles generated by the
momentary shifting of electrons surrounding the nucleus of electrically neutral
atoms. Unlike other types of secondary bonds, which occur between certain atoms
or groups, the var der Waals force always exists as long as two atoms or groups are
close enough to each other (Figure 2.21). Therefore, although the van der Waals
force is the weakest secondary bond, the sum of van der Waals forces between
two neighboring polymer chains is significant. Both the melting and dissolving
processes are closely related to the van der Waals force between polymer chains.
In addition to the five secondary bonds mentioned above, there are other types
of interactions existing between polymer chains. Examples include ion-dipole
force, ion-aromatic ring interaction, and ion-induced dipole force. These sec-
ondary bonds are formed between an ionic group with a dipole, aromatic ring,
and induced dipole, respectively. Ion-dipole force and ion-aromatic have similar
strength, and they are weaker than ionic secondary bonding, but may be slightly
stronger than dipole-dipole force or aromatic ring association.
REFERENCES
[1] Bower, D.I., An Introduction to Polymer Physics, Gambridge University
Press, 2002.
Figure 2.20. Aromatic ring association in PET.

Figure 2.21. Van der Waals force in polyethylene.
[2] Callister, W.D., Materials Science and Engineering: An Introduction,

Seventh Edition, John Wiley & Sons, 2007.
[3] Carey, F.A., and Sundberg, R.J., Advanced Organic Chemistry, Part A:
Structure and Mechanisms, Fifth Edition, Springer, 2007.
[4] Cook, J.G., Handbook of Textile Fibers, Vol. II. Man-Made Fibers, Fifth
Edition, Woodhead Publishing Limited, 1984.
[5] Flory, P.J., Principles of Polymer Chemistry, Cornell University Press,
1953.
[6] Grosberg, A.Y., and Khokhlov, A.R., Giant Molecules: Here, There, and
Everywhere, Academic Press, 1997.
[7] Hearle, J.W.S., “Fibre Structure: Its Formation and Relation to Performance”,
in Handbook of Textile Fibre Structure. Volume I: Fundamentals and
Manufactured Polymer Fibers, editors Eichhorn, S.J., Hearle, J.W.S., Jaffe,
M., and Kikutani, T., Woodhead Publishing Limited, 2009.
[8] Israelachvili, J.N., Intermolecular and Surface Forces, Third Edition,
Academic Press, 2011.
[9] Mark, J.E., Eisenberg, A., Graessley, W.W., Mandelkern, L. Samulski, E.T.,
Koenig, J.L. and Wignall, G.D., Physical Properties of Polymers, Second
Edition, American Chemical Society, 1993.
[10] Painter, P.C., and Coleman, M.M., Fundamentals of Polymer Science: An
Introductory Text, Second Edition, Technomic, 1997.
[11] Peebles, L.H., Molecular Weight Distributions in Polymers, Wiley, 1971.
Problems 33
[12] Rubinstein, M., and Colby, R.H., Polymer Physics, Oxford University
Press, 2003.
[13] Rudin, A., and Choi, P., The Elements of Polymer Science & Engineering,
Third Edition, Academic Press, 1982.
[14] Sperling, L.H., Introduction to Physical Polymer Science, 4th Edition, John
Wiley & Sons, Inc., 2006
[15] Tonelli, A.E., Polymers From the Inside Out—An Introduction to
Macromolecules, John Wiley & Sons, Inc., 2001.
[16] Warner, S.B., Fiber Science, Prentice Hall, 1995.
PROBLEMS
(1) Both polyethylene fiber and wax are essentially –CH2–. How does a poly-
ethylene fiber differ from wax?
(2) Natural fibers can have very long chains but polymers used for synthetic
fibers typically have shorter chain lengths. Why?
(3) Consider a nylon 6,6 fiber having the following distribution of polymer
chains:
20 chains of degree of polymerization 20
(i) Calculate the number average molecular weight.
(ii) Calculate the weight average molecular weight.
(iii) Calculate the polydispersity.
(4) Write the chemical structures of polyethylene, polypropylene, nylon 6, ny-

lon 6,6, polyester, polyacrylonitrile, and Kevlar® fibers.
(5) Different polypropylenes may be described as isotactic, syndiotactic, or
atactic. What does this mean? Which polypropylene is used to make com-
mercial textile fibers?
(6) What is the difference between a block and a graft copolymer? What are
the possible applications for fibers made from block and graft copolymers?
(7) A polyester fiber is described as having a polydispersity of 1.07. What does

this mean?
(8) Name the most prevalent primary bond found in polymer fibers. Describe
its electronic structure or mention what causes it to be formed between two
atoms.
(9) What are the main characteristics of covalent bonds?
(10) What kind of secondary bonds can be found in polypropylene, nylon, poly-
ester, and polyacrylonitrile fibers, respectively?
(11) Describe how ionic secondary bond, hydrogen bond, dipole-dipole force
and aromatic ring association are formed.
(12) All fibers have van der Waals force. Describe what causes the van der Waals
force to form.
CHAPTER 3
Physical Structure of Synthetic Polymer Fibers
The chemical structure of synthetic fibers deals with how the atoms and repeating
units are arranged in polymer chains, while the physical structure addresses how
the polymer chains are assembled to form fibers. In general, the chemical struc-
ture determines the maximum potential a fiber can achieve. However, the physical
structure determines how well the potential actually is achieved in practice. The
chemical structure affects the physical structure of synthetic fibers. This chapter
focuses on the physical structure of synthetic fibers.
3.1 SOLID PHASES

States of matter are the distinct forms, in which different phases of matter exist.
Three most commonly seen states of matter are solid, liquid, and gas. Figure 3.1
shows these three states of small molecule compounds. In the solid state, the
molecules are packed closely together. The secondary bonds between the mol-
ecules are strong enough so the molecules cannot move freely; they can only vi-
brate. There are two types of solids: crystalline and amorphous. The molecules in
crystalline solids are packed in a three-dimensionally ordered, repeating pattern,
while those in amorphous solids are distributed randomly. When the temperature
increases, the molecules become more excited and active, and the material can
change from the solid state to liquid. In the liquid state, the secondary bonds
are weakened due to the increased intermolecular distance. The molecules have
enough energy to have translational movement and the structure is mobile. As a
result, the shape of the liquid is not definite but is determined by its container.
When the temperature continues to increase, the molecules eventually can have
enough energy to completely break all secondary bonds, leading to the transi-
tion from the liquid state to gas. In the gas state, the distance between adjacent
35
36 Physical Structure of Synthetic Polymer Fibers
Figure 3.1. Three states of small molecule compounds.
molecules is much greater than the molecular size, and the material has no definite
shape or volume. The molecules in the gas state keep their translational move-
ment until they hit each other or the wall of the container.
Figure 3.2 shows the states of polymers. Polymers do not exhibit a gas state
because of the presence of secondary bonds. To change polymers to a gas state, all
secondary bonds must be completely broken to separate polymer chains from each
other. Although an individual secondary bond is weaker than a primary bond, the
sum of secondary bonds is significant due to the extremely long length of polymer
chains. Therefore, it is impossible to break all secondary bonds without damag-
ing the primary bonds, i.e., the decomposition of polymers. Therefore, polymers
only have liquid and solid states. The liquid state of polymers has significantly
higher viscosity than that of small molecular compounds. The behavior of poly-
mer liquids is essential for the formation of synthetic fibers and will be discussed
in Chapter 8. This chapter focusses on the solid state.
As shown in Figure 3.2, polymers also have two types of solid states: crys-
talline and amorphous. In reality, solid-state synthetic fibers are typically semi-
crystalline, which means both the amorphous and crystalline phases co-exist in
synthetic fibers. In the amorphous phase, the arrangement of polymer chains is dis-
ordered. However, in the crystalline phase, the polymer chains are assembled in a
three-dimensionally ordered manner. The crystalline phase has higher mechanical
strength than the amorphous one. During the fiber formation, an important task is
to manipulate the spinning parameters to achieve a high degree of crystallization.
Due to the extremely long length of polymer chains, it is almost impossible to
Unit Cells 37
Figure 3.2. States of polymers.
assemble them in a completely ordered structure in most practical fiber formation

processes. Therefore, synthetic fibers are typically semi-crystalline.
3.2 UNIT CELLS

3.2.1 Concept of Unit Cell
The crystalline phase is composed of smaller units, called unit cells. A unit cell is
basically a small set of atoms, ions, or molecules that are arranged in a particular
way, which is periodically repeated in three dimensions on a lattice. Unit cells
are the building block of the crystallites. The entire crystallite structure can be
constructed by pure translational repetition of unit cells. The overall properties of
a crystallite structure also can be calculated, at least theoretically, by considering
an individual unit cell.
There are totally 14 different types of unit cells, which meet the following
criteria:
• The unit cell is the simplest repeating unit in the crystal.

• Opposite faces of a unit cell are parallel.
• The edge of the unit cell connects equivalent points.
These fourteen unit cells belong to seven categories, which differ in the three
edge lengths (a, b, and c) and three internal angles (a, β, and γ). Figure 3.3 shows
the fourteen different unit cells, and Table 3.1 shows their edge lengths and inter-
nal angles.
In a real crystallite, each unit cell contains a few atoms, ions or molecules,
depending on the type of the material. For example, cesium chloride (CsCl) has a
&XELF 6LPSOH %RG\&HQWHUHG )DFH&HQWHUHG
ȕ F
Į
D
E E E
+H[DJRQDO

7HWUDJRQDO 6LPSOH %RG\&HQWHUHG

5KRPERKHGUDO
7ULJRQDO

2UWKRUKRPELF 6LPSOH %DVH&HQWHUHG %RG\&HQWHUHG )DFH&HQWHUHG

0RQRFOLQLF 6LPSOH %DVH&HQWHUHG

7ULFOLQLF

Figure 3.3. Fourteen types of unit cells.
body-centered cubic unit cell structure (Figure 3.4). In a CsCl unit cell, the Cl– is
located in the body center and the Cs+ ion in the corner. In Figure 3.4, it looks like
the unit cell has eight Cs+ ions since it has eight corners. However, on each corner,
the Cs+ is shared by eight neighboring unit cells, and hence each unit cell only
contains 8 × 1/8 = 1 Cs+ ion. As a result, each unit cell of CsCl contains two ions:
one Cl– ion and one Cs+ ion.
Unit Cells 39
Table 3.1. Edge lengths and internal angles of the seven categories of unit cells.
Category Edge lengths Internal angles

Cubic a=b=c α = β = γ = 90°
Hexagonal a=b≠c α = 120°, β = γ = 90°
Tetragonal a=b≠c α = β = γ = 90°
Rhombohedral (Trigonal) a=b=c α = β = γ ≠ 90°
Orthorhombic a≠b≠c α = β = γ = 90°
Monoclinic a≠b≠c α ≠ 90°, β = γ = 90°
Triclinic a≠b≠c α ≠ β ≠ γ ≠ 90°
Figure 3.4. Unit cell structure of CsCl.
3.2.2 Unit Cells of Synthetic Polymers

Synthetic fibers are composed of long polymer chains. Research shows the edge
lengths of unit cells of polymers are in the same order of magnitude as the co-
valent bond lengths. Hence, it is impossible for an entire polymer chain to be
included in a single unit cell. Instead, each polymer chain passes through many
unit cells. The long length and entanglement of polymer chains impede the chain
motion and leave regions that are amorphous. This is the main reason why most
synthetic fibers are not 100% crystalline. The following discusses the unit struc-
tures of several important fiber-forming synthetic polymers.
Polyethylene is a polymer with the simplest repeating unit structure. The crys-
talline phase of polyethylene typically is composed of orthorhombic unit cells,
with edge lengths of a = 0.740 nm, b = 0.493 nm, and c = 0.254 nm. In the crystal-
line phase, the chains of polyethylene are in all-trans conformation and have a lin-
ear zigzag conformation. Figure 3.5 shows the unit cell structure of polyethylene.
It is seen that each unit cell contains two repeating units.
Vinyl polymers have side groups on their backbones, and only isotactic or syn-
diotactic vinyl polymers can crystallize. Due to the substantial steric hindrance
from the side groups, vinyl polymers adopt a helical conformation during crystal-
lization, where the extra side groups are accommodated on the outside of the helix
Figure 3.5. Unit cell structure of polyethylene.
by twisting of the entire polymer chain. For example, isotactic polypropylene

crystallized at normal conditions can have either hexagonal or monoclinic unit
cell. In either case, the polypropylene forms a helix that makes one complete turn
for every three repeating units. Figure 3.6 compares the zigzag conformation of
polyethylene and the helix conformation of polypropylene. Of course, both left-
and right-handed helices are possible. Figure 3.7 shows that left- and right-handed
helices of polypropylene face each other, leading to a closer packing of polymer
chains in the crystalline phase. Other vinyl polymers also form helical conforma-
tion in the crystalline phase, but they may require different numbers of repeating
units for making a complete turn.
The polymers discussed above are nonpolar and van der Waals force is the
only type of secondary bond that holds the polymer chains together in the crys-
talline phase. When polymers have other types of stronger secondary bonds, the
crystallite structure tends to be affected by these bonds. Figure 3.8 shows the
effect of hydrogen bonding on the chain arrangement in the crystalline phase of
nylon 6,6. The unit cell structure of nylon 6,6 is triclinic, with the edge lengths of
a = 0.49 nm, b = 0.54 nm, and c = 1.73 nm, and internal angles of α = 48°, β = 77°,
and γ = 63°. To maximize the number of hydrogen bonds, the nylon 6,6 chains
tend to form a fully extended, planar zigzag conformation.
The crystalline phase of polyethylene terephthalate (PET) also has a unit cell
structure of triclinic. The edge lengths are: a = 0.456 nm, b = 0.594 nm, and
c = 1.075 nm, and internal angles are: α = 98.5°, β = 118°, and γ = 112°. Figure 3.9
Unit Cells 41
Figure 3.6. Zigzag structure of polyethylene and helical structure of polypropylene.
Figure 3.7. Arrangement of helical structures of isotactic polypropylene.

Figure 3.8. Planar zigzag structure of nylon 6,6.
shows the effect of aromatic ring association on the arrangement of polymer

chains in PET crystalline phase. The aromatic rings tend to be stacked together to
maximize the secondary bonding (aromatic ring association), leading to the for-
mation a planar zigzag structure (Figure 3.9). Both nylon 6,6 and PET fibers have
relatively high melting temperatures (around 260°C) due to the hydrogen bonding
in the former case and the aromatic ring association in the latter case. It must be
mentioned that in the case of PET, the polymer chain is relatively stiff, which also
contributes to its planar zigzag conformation and high melting temperature.
3.3 CRYSTALLINE MODELS

Synthetic fibers are semicrystalline. The unit cell structure deals with how the
polymer chains are organized inside the crystalline phase. However, it does not
tell how the crystallites look like in semicrystalline synthetic fibers. To address
this, several crystalline models have been developed.
Crystalline Models 43
Figure 3.9. Planar zigzag structure of PET.
3.3.1 Fringed Micelle Model

It has been discussed that the polymer chains pass through many unit cells in the
crystalline phase. Due to the large molecular weights, the polymer chains are cal-
culated to be even longer than the crystallites. Hence, it is believed the polymer
chains also pass in and out of many crystallites. This leads to the fringed micelle
model (Figure 3.10). According to the fringed micelle model, the polymer chains
wander from the amorphous phase through a crystallite, and then back out into
the amorphous phase. Because of the polymer chains’ length, they pass through
several crystallites to bind them together. This model is successful in explaining
the good tensile strength and flexible nature of semicrystalline synthetic fibers.
3.3.2 Folded-Chain Model and Switchboard Model

One important feature of the fringed micelle model is that all the polymer chains
run continuously between the crystalline and amorphous phases without folding
back. This view changed when single crystals of polyethylene were prepared by
Keller in 1957. These single crystals were made by precipitation from dilute so-
lutions of hot xylene, and they typically have diamond shape, with thicknesses
ranging from 10 to 20 nm. To explain the structure of polyethylene single crys-
tals, a folded-chain model was developed (Figure 3.11). The folded-chain model
explains the chain-folding phenomenon, which is not captured in the fringed mi-
celle model. However, the folded-chain model can only be used to explain spe-
cial crystalline structure, such as single crystals. This is because the folded-chain
model requires large-scale, long-range reorganization of all polymer chains. Due
to the long length and entanglement of polymer chains, this is difficult during the
fiber formation process. Therefore, a switchboard model also has been developed.
Figure 3.10. Fringed micelle model with the crystalline phase shadowed.
Figure 3.11. Folded-Chain model.
Figure 3.12 compares the chain re-entry in folded-chain and switchboard models.

In the folded-chain model, the polymer chains reenter the crystalline phase right
next to where they left. On the other hand, in the switchboard model, the polymer
chains do not have to follow adjacent re-entry. Instead, they can re-enter the crys-
talline phase randomly. The switchboard model requires minimum movement and
re-organization of polymer chains.
3.3.3 Modified Fringed Micelle Model and Fringed Fibril Model

Considering the fact that polymer chains can fold back and re-enter the crystal-
lites, the fringed micelle model has been modified. Figure 3.13 shows the modi-
fied fringed micelle model. At the ends of the crystallites in a modified fringed
micelle model, there is a mixture of chain folding and fringing into tie-molecules,
leading to other crystallites. The modified fringed micelle model is being widely
used to explain the crystalline structure of polymers.
Crystalline Models 45
Figure 3.12. Comparison of polymer chain re-entry in folded-chain and switchboard

models.
Figure 3.13. Modified fringed micelle model with the crystalline phase shadowed.
In synthetic fibers, the micelles are different from those in many other polymer
products. The micelles in synthetic fibers often have the fibril shape, with diam-
eters ranging from several nm to 100 nm. The lengths of fibrils depend on the
polymer type and the processing conditions. This leads to a fringed fibril model.
Figure 3.14 shows a basic fringed fibril model. The bonding between fibrils is
relatively weak, and hence fibrils actually can be observed in fracture studies of
some synthetic fibers.
The actual crystalline structure of synthetic fibers is largely affected by the
polymer type and the fiber formation conditions. The models discussed above
cannot explain all the crystalline structures that have been found in synthetic fi-
bers. Hence, other models also have been developed, and many of them are modi-
fied versions of above-mentioned models so they can better explain the structure
and properties of different synthetic fibers.
Figure 3.14. Fringed fibril model with the crystalline phase shadowed.
3.4 MORPHOLOGY OF CRYSTALLITES

3.4.1 Spherulitic Morphology
Under quiescent conditions, the crystallization of polymers begins at numerous
nucleation points and the crystalline phase grows outward (spherically) from
these centers to form shperulites. Spherulites can be observed in polarized opti-
cal microscope, which produces a Maltese Cross appearance for the individual
spherulites. Figure 3.15 shows both a polarized light microscopy image of spheru-
lites and a sketch of the spherulite structure. Spherulites contain both crystalline
and amorphous phases. In spherulites, polymer chains are oriented normal to the
radial direction. The crystals in the spherulites are called lamellar crystals and of-
ten are described using the switchboard model. The diameters of spherulites grow
linearly with time until they impinge on each other. Eventually, the spherulites can
pervade the entire space of the material.
3.4.2 Fibrillar Morphology

Spherulites are formed under quiescent conditions (no flow and no large-scale
molecular orientation). However, synthetic fibers are formed by using different
spinning techniques. Polymer chains and crystallites can orient themselves during
the fiber spinning processes. To enhance the mechanical properties, spun fibers
also are drawn or stretched to further increase the orientation. As a result, the
spherulitic structure gives away to the fibrillar morphology. For many synthetic
fibers, fibrils are formed directly by oriented polymer chains during spinning
processes and the fibrillar morphology is enhanced by drawing. However, some
polymers have high crystallization rate and may form some spherulites in as-spun
fibers under certain spinning conditions. In this case, the shperulites are stretched
during drawing and some folded polymer chains slip, twist or even unfold to form
fibrils. As a result, spherulites are not common in commercial synthetic fibers.
Morphology of Crystallites 47
Figure 3.16 shows an example of the fibrillar morphology. It is seen the orien-

tation direction of fibrils is the same as that of the polymer chains in the crystalline
phase. In synthetic fibers, the diameters of fibrils can range from tens of nano-
meters to several micrometers. Each fibril also may contain both crystalline and
amorphous phases. The fibrillar morphology of synthetic fibers can be described
using the fringed fibril model (Figure 3.14) or its modified version with higher
degrees of crystallinity and orientation.
Figure 3.15. Appearance and structure of spherulites. (Lotz, B., et. al., Polymer, 46, 577–
610, 2005.)
Figure 3.16. AFM images of (A) fibrils and (B) polymer chains in the crystalline phase for
stretched polyethylene. (Snetivy, D., et. al., Journal of Materials Chemistry, 2, 891–892,
1992.)
3.4.3 Other Morphologies

Compared with the spherulitic structure, the fibrillar morphology is more desir-
able for synthetic fibers. During the formation of oriented fibrils, the polymer
chains are arranged in a way that allows alignment of more primary bonds along
the fiber axis direction. This leads to increased mechanical properties for synthetic
fibers. However, as shown in the fringed fibril model (Figure 3.14), a significant
amount of polymer chains still are folded rather than extended. As a result, the ac-
tual strength and modulus of most synthetic fibers are only a small fraction of the
theoretical values that perfectly aligned polymer chains can achieve. The key to
maximize the mechanical properties of synthetic fibers is to fully extend the poly-
mer chains and align them along the fiber axis. Such structure could be achieved
by high-speed spinning of fibers, spinning fibers directly from liquid crystalline
polymers, and gel spinning. Figure 3.17 shows a model for polyester fibers made
by high-speed spinning. Figure 3.18 shows a model for Kevlar® fibers spun from
the liquid crystalline state and Spectra® polyethylene fibers made by gel spinning.
The structures discussed above are just examples that can be found in many
synthetic fibers. Fiber scientists now can manipulate the fiber formation processes
to obtain a wide range of structures for different applications. Therefore, other
crystalline morphologies and models exist for synthetic fibers.
3.5 MORPHOLOGY OF SYNTHETIC FIBERS

3.5.1 Cross-Sectional Morphology
The cross-sectional shape of synthetic fibers is largely determined by the spin-
neret profiles used during fiber spinning. Circular cross-sectional shape is the
simplest and it can be produced with spinnerets with circular holes. However,
the cross-sectional shapes also are affected by the spinning method used. For
example, when fibers are spun from polymer solutions, the fiber skin is the first
region to be solidified during the removal of solvent. With the continuous removal
of solvent after the solidification of the skin, the fiber diameter decreases and the
skin has to be deformed to accommodate the diameter change. As a result, fibers
spun from solutions are not round. Instead, they may present crenulated, bilobal,
trilobal, or other cross-sectional shapes (Figure 3.19).
Non-circular synthetic fibers also can be obtained by using different spinneret
profiles. Figure 3.20 shows several different spinneret hole shapes. While produc-
ing synthetic fibers with different cross-sectional shapes, it is important to con-
sider the unique properties of polymer melts or solutions. For example, polymers
swell when they exit the spinneret holes, and the swelling is uneven from non-
circular spinneret holes. Typically, the swelling at corners is less significant than
at other locations. Therefore, to produce square or triangle fibers, the spinneret
holes need have stretched corners (Figures 3.20A and B).
Morphology of Synthetic Fibers 49
Figure 3.17. Model for polyester fibers made by high-speed spinning.
Figure 3.18. Model for Kevlar® fibers spun from the liquid crystalline state and Spectra®
polyethylene fibers made by gel spinning.
Figure 3.19. Crenulated, bilobal and trilobal cross-sectional shapes of synthetic fibers.
Figure 3.20. Examples of spinneret holes for making fibers with different cross-sections of
(a) square, (b) triangle, (c) hollow, (d) T, and (e) star shapes.
Some synthetic fibers consist of two polymers arranged in separated phases on

the cross-section. These bicomponent fibers typically are produced by spinning
two polymers from the same spinneret with both polymers contained within the
same filament. Side-by-side, segmented pie, core-sheath and islands-in-the-sea are
examples of the cross-sectional structures for bicomponent fibers (Figure 3.21).
Bicomponent fibers can have a number of advantages over single-component fi-
bers. For example, in some bicomponent fibers, the second component can serve
as a reinforcing agent to enhance the mechanical properties of the fibers. In some
other bicomponent fibers, the second component can improve the dyeing capabil-
ity of the fibers. Some bicomponent fibers also are used as self-crimping fibers.
In addition, some bicomponent fibers are designed to make special fibers that are
not easy to produce using regular spinning method. For example, islands-in-sea
bicomponent fibers often are used to make micro/nanofibers or hollow fibers by
dissolving one of the two components.
3.5.2 Longitudinal Morphology

As-spun synthetic fibers are typically straight. However, some synthetic fibers
possess a wavy, undulating structure, which is called crimp. Crimped fibers can be
produced by different methods. The most straightforward method is to introduce
crimp by thermomechanical means. In this method, straight fibers are mechani-
cally deformed into a crimped structure, followed by heat setting.
An alternative approach is to produce self-crimped fibers by using the bicom-
ponent fiber structures discussed in section 3.5.1. The two components in the fi-
bers can be different polymers or the same polymer but with different average
molecular weights, different molecular weight distributions, different additives,
or other structural differences. The key to introduce self-crimp is to design struc-
tural asymmetry across the fiber cross-section.
Figure 3.21. Side-by-side, segmented pie, core-sheath and islands-in-the-sea structures of

bicomponent fibers.
References 51
REFERENCES
[1] Callister, W.D., Materials Science and Engineering: An Introduction,
Seventh Edition, John Wiley & Sons, 2007.
[3] Geil, P.H., Polymer Single Crystals, Interscience, 1963.
[4] Grosberg, A.Y., and Khokhlov, A.R., Giant Molecules: Here, There, and
Everywhere, Academic Press, 1997.
[5] Hearle, J.W.S., “Fibre Structure: Its Formation and Relation to Performance”,
in Handbook of Textile Fibre Structure, Volume I: Fundamentals and
Manufactured Polymer Fibers, editors Eichhorn, S.J., Hearle, J.W.S., Jaffe,
M., and Kikutani, T., Woodhead Publishing Limited, 2009.
[6] Kikutani, I., Radhakrishnan, J., Arikawa, S., Takaku, A., Okui, N., Jin,
N., Niwa, F., and Kudo, Y., “High-Speed Melt Spinning of Bicomponent
Fibers: Mechanism of Fiber Structure Development in Poly(ethylene tere-
phthalate)/Propylene System”, Journal of Applied Polymer Science, 62,
1913–1924, 1996.
[7] Lotz, B., and Cheng, Z.D., “A Critical Assessment of Unbalanced Surface
Stresses as the Mechanical Origin of Twisting and Scrolling of Polymer
Crystals”, Polymer, 46, 577–610, 2005.
[8] Mark, J.E., Eisenberg, A., Graessley, W.W., Mandelkern, L. Samulski, E.T.,
Koenig, J.L. and Wignall, G.D., Physical Properties of Polymers, Second
Edition, American Chemical Society, 1993.
[9] Mather, R.R., “The Structure of Polyolefin Fibers”, in Handbook of Textile
Fibre Structure, Volume I: Fundamentals and Manufactured Polymer
Fibers, editors Eichhorn, S.J., Hearle, J.W.S., Jaffe, M., and Kikutani, T.,
Woodhead Publishing Limited, 2009.
Introductory Text, Second Edition, Technomic, 1997.
[11] Rudin, A., and Choi, P., The Elements of Polymer Science & Engineering,
Third Edition, Academic Press, 1982.
[12] Salem, D.R., Structure Formation in Polymer Fibers, Hanser Publishers,
2001.
[13] Snetivy, D., Yang, H., and Julius Vancso, G., “Imaging of Different Crystal
Planes in Oriented Polyethylene by Atomic Force Microscopy”, Journal of
Materials Chemistry, 2, 891–892, 1992.
Wiley & Sons, Inc., 2006.
[15] Tonelli, A.E. Polymers From the Inside Out—An Introduction to
Macromolecules, John Wiley & Sons, Inc., 2001.
PROBLEMS
(1) Explain why polymer does not have a gas state.
(2) Why are some polymers transparent? Others are not?
(3) The lengths of polymer chains are significantly greater than the edge lengths
of unit cells. How can polymer chains crystallize in fibers by forming unit
cells?
(4) What are the common conformations found in the crystalline phase of poly-
mer fibers?
(5) How does the hydrogen bonding affect the crystalline structure of nylon
fibers?
(6) Spherulites are commonly seen in many polymer products. However, spher-
ulites should be avoided during the formation of synthetic fibers. Why?
(7) The crystallites in many polymer fibers have fibrillar morphology. How
does the drawing process affect the fibrillar morphology?
(8) What are the possible routes for making high-performance synthetic fibers?
(9) What determines the cross-sectional shape of melt-spun fibers?
(10) What are the typical cross-sectional shapes of solution-spun fibers?
CHAPTER 4
Chemical Structure of Natural Polymer Fibers
The chemical structure of natural polymer fibers is more complex than that of
synthetic fibers. Two most important building units for natural polymer fibers are
cellulose and protein. Natural cellulose fibers come from the “stringy” portions
of plants, ranging from the fine seed fibers from the cotton plant to the coarse
pineapple leaf fibers. Natural protein fibers are hairs of animals like the sheep and
the delicate filaments spun by silkworms and insects. In addition to these natural
fibers, manufactured cellulose and protein fibers also are important fibers that are
based on natural biopolymers, but are processed like synthetic polymer fibers.
4.1 NATURAL CELLULOSE FIBERS

Natural cellulose fibers also are called plant fibers or vegetable fibers. Natural
cellulose fibers include cotton together with flax, jute, jute, ramie and other fibers
produced by plants. Natural cellulose fibers can be classified to seed, bast, leaf,
and fruit fibers. Figure 4.1 shows the classification of natural cellulose fibers.
4.1.1 Cellulose
The basic building unit of natural cellulose fibers is cellulose, the most abundant
biopolymer in nature. Figure 4.2 shows the chemical structure of cellulose, which
basically is a polysaccharide consisting of a linear chain of several hundred to
more than ten thousand β-linked glucose units. The glucose unit has a ring struc-
ture of one oxygen -O- atom and five -CH- groups. Three hydroxyl (–OH) side
groups also are connected to the ring structure. Among them, two -OH groups
are linked directly to the -CH- group on the ring while the other -OH group is
linked to -CH- through a -CH2- group. The remaining two -CH- groups on the ring
53
54 Chemical Structure of Natural Polymer Fibers
Figure 4.1. Classification of natural cellulose fibers.
Figure 4.2. Chemical structure of cellulose.
structure are connected by -O- atoms to neighboring glucose units on the linear
polymer chain.
Two important features of the chemical structure of cellulose are the glucose
ring structure and –OH side groups. The glucose ring structure provides stiffness
to the polymer chain and the –OH groups can form hydrogen bonds. These fea-
tures define many important properties of natural cellulose fibers.
Starch is another important biopolymer that is found in plants and has a simi-
lar structure to cellulose. Like cellulose, starch also consists of glucose units.
However, the glucose units in cellulose and starch are connected with different
types of linkages, and the resultant polymer chains have different configurations.
Figure 4.3 compares the polymer chains of starch and cellulose. The glucose units
Natural Cellulose Fibers 55
Figure 4.3. Comparison of starch and cellulose.
in starch are connected with α linkages and the polymer chain has a α configura-
tion, i.e., all –CH2-OH groups on the same side of the polymer chain. However,
the glucose units in cellulose are connected by β linkages, and the polymer chain
has a β configuration. In the β configuration, each successive glucose unit is ro-
tated around the axis of the polymer chain, and -CH2-OH groups on neighboring
rings are on the alternative sides of the polymer chain. This small difference in
the chemical structures of starch and cellulose has significant impacts on their
properties. For example, starch is soluble in water when heated and is used by
plants to store energy in their carbon-hydrogen bonds. However, cellulose does
not dissolve in water and is strong enough to form usable fibers.
4.1.2 Cotton and Other Natural Cellulose Fibers

Cotton fibers are the most important natural cellulose fibers. Cotton fibers are one
of the purest sources of cellulose. The cellulose content in cotton fibers is in the
range of 88.0% – 96.0%. The noncellulosic components of cotton fibers include
proteins, pectic substances, sugars, wax, and organic acids. Table 4.1 shows the
contents of different components in dried cotton fibers. The actual contents of dif-
ferent components may vary, depending on the variety and maturity of cotton and
environmental conditions such as climate, soil, water source, farming method,
etc. Treatments can remove most of the noncellulosic components, and the cel-
lulose content of the treated cotton fibers could be more than 99%.
In addition to cotton fibers, they are many other types of natural cellulose fi-
bers. One major difference between cotton and other natural cellulose fibers is the
cellulose content. Cotton fibers are the purest and other natural cellulose fibers
have significantly lower content of cellulose. For example, both cotton and kapok
are seed fibers, but cotton has the highest cellulose content and kapok has the
lowest (13 wt%). The cellulose contents of other natural cellulose fibers are in
the range of 40%–90%. Table 4.2 shows the cellulose contents of several natural
cellulose fibers.
Table 4.1. Chemical composition of dried cotton fibers.

Component Content (wt %)
Cellulose 88.0–96.0
Protein 1.1–1.9
Ash 0.7–1.6
Pectic substances 0.7–1.2
Organic acids 0.5–1.0
Wax 0.4–1.0
Sugars 0.1–1.0
Others —
Source: Wakelyn, P.J., et al., Cotton Fiber Chemistry and Technology, CRC press, 2007.
Table 4.2. Cellulose contents of different natural cellulose fibers.
Cellulose Fiber Plant Origin Content (wt %)

Cotton Seed 88–96
Kapok Seed 13
Hemp Bast 70–92
Ramie Bast 68–76
Flax Bast 60–81
Jute Bast 51–84
Kenaf Bast 44–57
Bagasse Bast 32–48
Bamboo Bast 26–43
Phormium Leaf 67
Pineapple Leaf 70–82
Abaca Leaf 61–64
Banana Leaf 60–65
Sisal Leaf 43–78
Coir Fruit 43–46
Source: Mwaikambo, L.Y., African Journal of Science and Technology, 7, 120–133, 2006.
4.2 MANUFACTURED CELLULOSE FIBERS

Manufactured cellulose fibers are man-made fibers. To make manufactured cel-
lulose fibers, the cellulose component of the plant is dissolved in a solvent, which
then is spun into fibers in the same ways that synthetic polymer fibers are made.
Examples of manufactured cellulose fibers are rayon, acetate, triacetate, Lyocell,
and Cupro.
Manufactured Cellulose Fibers 57
Manufactured cellulose fibers can be made by both derivative and direct meth-
ods. In derivative methods, cellulose is chemically modified to form a cellulose
derivative, which is spun into a coagulation bath to produce manufactured cel-
lulose fibers. Rayon is the first manufactured cellulose fiber produced by the de-
rivative approach, and the process used is called viscose process. In this process,
cellulose in the form of wood pump is treated with sodium hydroxide and carbon
disulfide to form cellulose xanthate. The cellulose xanthate is extruded into an
acid bath, where the xanthate groups are removed from the polymer chains to
regenerate the cellulose in the fiber form, called rayon (Figure 4.4). However, the
cellulose derivative is not always regenerated back into cellulose. For example,
acetate and triacetate fibers are produced by acetylation with acetic acid and acetic
anhydride with the presence of sulfuric acid. The acetylation converts hydroxyl
groups in cellulose to acetate groups (Figure 4.5), which render the cellulose
polymer much more soluble in organic solvents. After the fiber formation, the
cellulose derivative is not converted back into cellulose. The difference between
acetate and triacetate fibers lies in the degree of acetylation. In acetate fibers, less
than 92% but at least 74% of the –OH groups are acetylated. In triacetate fibers,
at least 92% of the hydroxyl groups are acetylated.
In the direct methods, cellulose is dissolved and spun into fibers by using
special solvent systems without chemical modification of the polymer struc-
ture. For example, Lyocell fibers are produced by using an organic solvent,
N-methylmorpholine N-oxide, to dissolve cellulose. Cupro fibers are manufac-
tured by dissolving cellulose in a mixed solution of copper salts and ammonia.
(Some researchers consider Cupro fibers to be produced by a derivative method
because interaction is formed between cellulose and the cuprammonium ions in
the solution, as discussed in Chapter 10.) More recently, researchers also are using
ionic liquids as the solvent to fabricate manufactured cellulose fibers.
The fabrication methods and chemical structures of manufactured cellulose
fibers are summarized in Table 4.3.
Figure 4.4. Chemical modifications involved in the viscose process of Rayon.
Figure 4.5. Acetate group.

Table 4.3. Fabrication methods, main characteristics, and

chemical structures of manufactured cellulose fibers.
Fiber Fabrication Main
Fiber Structure
Name Method Characteristics
Rayon Derivative Produced by the Cellulose:
method viscose process
Acetate Derivative Less than 92% but Cellulose acetate:

method at least 74% of the
hydroxyl groups are
acetylated
Triacetate Derivative At least 92% of the Cellulose triacetate:

method hydroxyl groups are
acetylated
Lyocell Direct method Produced by using Cellulose:

an organic solvent,
N-methylmorpholine
N-oxide
Cupro Direct Processed by using Cellulose:

method* a mixed solution
of copper salts and
ammonia
*Some researchers consider Cupro fibers to be produced by a derivative method because interac-
tion is formed between cellulose and the cuprammonium ions in the solution.
Natural Protein Fibers 59
Figure 4.6. Structure of a polypeptide chain.
Figure 4.7. Reaction of amino acids to form a polypeptide.
4.3 NATURAL PROTEIN FIBERS

Natural protein fibers can be classified into two basics groups, animal hair fibers
and extruded protein fibers. Animal hair fibers include but are not limited to wool
(from sheep), cashmere, mohair, yak, and alpaca. All animal hair fibers have simi-
lar chemical structure. Extruded protein fibers are spun by silkworm, spider, etc.,
and they also have similar chemical structure. In this section, the discussion on
natural protein fibers will focus on wool and silk (by silkworm).
4.3.1 Protein
Among all fibers, protein fibers have the most complex chemical structure.
Proteins consist of one or more polypeptides folded together. A polypeptide is
a single linear polymer chain of amino acids bonded together by peptide link-
ages between the carboxyl and amino groups of adjacent amino acid residues.
Figure 4.6 shows a linear polypeptide chain, and Figure 4.7 shows the reaction of
amino acids to form a polypeptide.
There are twenty different types of amino acids, which are distinguished by
their different side (-R’) groups (Figure 4.8). The sequences of different amino
acid residues in proteins are defined by the genetic codes, and the different se-
quences represent the different biological functions of proteins. Twenty different
amino acids lead to an enormous list of possible sequences that make proteins the
most versatile biopolymers. In the case of wool and other hair fibers, the genetic
codes trigger the production of different proteins with different compositions and
sequences of amino acid residues when the fibers grow out of hair follicles. The
Figure 4.8. Chemical structures of twenty amino acids.
formation process of silk filaments also is affected by the protein properties, which
are determined by the compositions and sequences of the amino acid residues.
4.3.2 Wool and Silk

Table 4.4 shows typical contents of different amino acid residues in the proteins
of wool and silk fibers. The contents of amino acid residues vary with breed, diet
and other factors. For example, wool fibers from different sources could have dif-
ferences of up to 20% in the contents of some amino acid residues.
Table 4.4. Amino acid residue contents of wool and silk fibers.
Amino Acid Residue

Type Side Group Amino Acid Content (wt %)
Wool Silk
Inert -H Glycine 6.5 43.8
-CH3 Alanine 4.1 26.4
-CH(CH3)2 Valine 5.5 3.2
-CH2CH(CH3)2 Leucine 9.7 0.8
-CH(CH3)CH2CH3 Isoleucine — 1.37
-CH2C6H5 Phenylalanine 1.6 1.5
Acidic -CH2COOH Aspartic acid 7.27 3.0
-CH2CH2COOH Glutamic acid 16.0 2.03
Basic -(CH2)4NH2 Lysine 2.5 0.88
-(CH2)3NHC(NH)NH2 Arginine 8.6 1.05
Histidine 0.7 0.47
Hydroxyl -CH2OH Serine 9.5 12.6

-CH(OH)CH3 Threonine 6.6 1.5
-CH2C6H4OH Tyrosine 6.1 10.6
Ring Proline 7.2 1.5
Double -CH2-S-S-CH2- Cystine 11.8 —

Other -CH2CH2-S-CH3 Methionine 0.35 —
Tryptophan 0.7 —
Source: Harris, M., Handbook of Textile Fibers, Harris Research Laboratories, 1954.
Due to the large amount of functional side groups and the >C=O and >N-H
groups on the polymer main chains, many different types of intermolecular bonds
can be formed in wool and silk fibers:
• Hydrogen bonds can be formed between >C=O and >N-H groups on the
main chains and between the hydroxyl and amide side groups.
• Ionic secondary bonds occur between acidic and basic side groups.
• Aromatic ring association can be formed between the aromatic rings in
some of the side groups.
• Covalent amide bonds can be formed between polymer chains during to the
presence of –COOH and –NH2 groups, which is not common in synthetic
polymer fibers.
• Covalent disulfide bonds also can occur between polymer chains due to the
presence of cysteine (-CH2-SH). During protein synthesis, cysteine is oxi-
dized to cystine, -CH2-S-S-CH2-, which provides the disulfide bonds.
In wool and silk fibers, intermolecular bonds are so extensive that the polymers
cannot melt. When heated, the primary bonds on the main chains break before
all intermolecular bonds can be damaged. As a result, both wool and silk behave
like thermosetting polymers. However, the ability for wool and silk fibers to form
intermolecular bonds is different. As shown in Table 4.4, the chemical structure
of silk is relatively simple and contains mainly residues of four types of amino
acids: glycine, alanine, serine, and tyrosine. Wool has a more complex chemical
structure and consists of many different types of amino acid residues. As a result,
more types of intermolecular bonds can be found in wool fibers.
4.4 MANUFACTURED PROTEIN FIBERS

Manufactured protein fibers, often called azlons, are man-made fibers produced
from animal or plant proteins. Examples of protein sources are milk, chicken
feathers, soy beans, peanuts, corns, etc. Traditionally, most manufactured pro-
tein fibers were made directly from proteins dissolved in solvents. Recent trends
in the research and development of manufactured protein fibers include the use
of biochemistry to modify the source proteins and the introduction of synthetic
polymers such as polyvinyl alcohol and polyacrylonitrile to improve the fiber
mechanical properties. Antibacterial agents are often being added during the fiber
formation process to provide health benefits to the manufactured protein fibers.
As a result, the chemical structure of manufactured protein fibers is becoming
more complex.
Currently, the market share of manufactured protein fibers is smaller than most
other types of fibers. However, advances in manufacturing process and structure
modification in the latest manufactured protein fibers indicate they may have a
chance to increase the public acceptance.
REFERENCES
[1] Brooks, M.M., “Regenerated Protein Fibres: a Preliminary Review”, in
Handbook of Textile Fibre Structure, Volume II: Natural, Regenerated,
Inorganic and Specialist Fibres, editors Eichhorn, S.J., Hearle, J.W.S.,
Jaffe, M., and Kikutani, T., Woodhead Publishing Limited, 2009.
[2] Choudhury, A.K.P., Textile Preparation and Dying, Science Publishers, 2006.
Problems 63
[3] Cook, J.G., Handbook of Textile Fibres, Volume I—Natural Fibres,

[4] Ciechanska, D., Wesolowska, E., Wawro, D., “An Introduction to Cellulosic
Fibres”, in Handbook of Textile Fibre Structure, Volume II: Natural,
Regenerated, Inorganic and Specialist Fibres, editors Eichhorn, S.J.,
Hearle, J.W.S., Jaffe, M., and Kikutani, T., Woodhead Publishing Limited,
2009.
[5] Dyer J., Daul, G., “Rayon Fibers”, in Handbook of Fibre Science and
Technology, Volume IV. Fibre Chemistry, editors Lewin, M., Pearce, E.M.,
Marcel Dekker, 1985.
[6] Ganster, J., Pink, H.P., “The Structure of Man-Made Cellulosic Fibres”,
in Handbook of Textile Fibre Structure, Volume II: Natural, Regenerated,
[7] Harris, M., Handbook of Textile Fibers, Harris Research Laboratories, 1954.
[8] Hearle, J.W.S., “An Introduction to Protein Fibres”, in Handbook of Textile
Fibre Structure, Volume II: Natural, Regenerated, Inorganic and Specialist
Fibres, editors Eichhorn, S.J., Hearle, J.W.S., Jaffe, M., and Kikutani, T.,
[9] Hsieh, Y.L., “Chemical Structure and Properties of Cotton”, in Cotton:
Science and Technology, editors Gordon S., and Hsieh, Y.L., Woodhead
Publishing Limited, 2007.
[10] Kosan, B., Michels, C., and Meister, F., “Dissolution and Forming of
Cellulose with Ionic Liquid”, Cellulose, 15, 59–66, 2008.
[11] Mwaikambo, L.Y., “Review of the History, Properties and Application of
Plant Fibres”, African Journal of Science and Technology, 7, 120–133,
2006.
[12] Wakelyn, P.J., Edwards, J.V., Bertoniere, N.R., Triplett, B.A., Hunter, L.,
French, A.D., Rousselle, M.A., McAlister, D.D., Thibodeaux, D.P., Goynes,
W.R. Jr., and Gamble, G.R., Cotton Fiber Chemistry and Technology, Taylor
& Francis Group, 2007.
PROBLEMS
(1) What are the four main types of natural cellulose fibers?
(2) Describe the repeating unit of cellulose.
(3) How do the hydroxyl groups affect the properties of natural cellulose fibers?
(4) What is the major difference between the chemical compositions of cotton
and other natural cellulose fibers?
(5) What are the two methods for making manufactured cellulose fibers?
Describe their major difference.
(6) How does triacetate differ from acetate?
(7) Describe the chemical structure of natural protein fibers.
(8) Describe the main differences (chain structure, side groups, and morphol-
ogy, whatever are applicable) between the structures of wool and silk.
(9) What are the secondary bonds that can be found in wool and silk fibers?
(10) Why do wool and silk give different smells in the burning test?
CHAPTER 5
Physical Structure of Natural Polymer Fibers
The physical structure of natural cellulose and protein fibers is determined by the
genetic codes. Human can only have limited influence on the physical structure
of natural cellulose and protein fibers by selecting the species being raised and
the environmental conditions such as soil, water source, farming method, etc.
However, the physical structure of manufactured cellulose and protein fibers can
be manipulated by controlling the chemical structure and processing conditions.

5.1.1 Cotton
As discussed in Chapter 4, cellulose is by far the major component in cotton
fibers, but its precise content depends on the growth environment, geographical
location and maturity of the plant. The glucose rings are not flat but bent in a chair
configuration and each successive ring is rotated by 180° with respect to the poly-
mer main chain. The hydroxyl groups attached to the ring units can form hydro-
gen bonds with each other and link cellulose chains into thin sheets (Figure 5.1).
These sheets are hold together by van der Waals forces to form crystallites.
Figure 5.2 shows a unit cell structure of the cellulose in cotton, with 1.030 nm for
two glucose units along the polymer chain axis, 0.835 nm for spacing between
adjacent chains in the sheets, and 0.790 nm between sheets. The crystal lattice in
cotton is known as Cellulose I, which is different from the Cellulose II structure in
manufactured cellulose fibers. When cellulose is dissolved and regenerated to the
fiber form, the parallel arrangement of cellulose chains is altered and the resultant
manufactured cellulose fibers will have a Cellulose II structure with anti-parallel
chains.
65
66 Physical Structure of Natural Polymer Fibers
Figure 5.1. Possible assembly of cellulose chains in a sheet with hydrogen bonds.
Figure 5.2. Unit cell structure of cellulose in cotton.

Cellulose in cotton is synthesized by the condensation of glucose monomers at

enzyme complexes. Each enzyme complex produces 30 cellulose chains, which
lie in the same direction to crystallize into long microfibrils. The microfibrils have
an average diameter of several nanometers. From this aspect, the cellulose in cot-
ton is considered to be 100% crystalline since all cellulose chains contribute to
the formation of microfibrils. However, the density of cotton is lower than the
crystal density, and experimental measurements using diffraction, spectroscopic,
thermal and other techniques show cotton is about 60% crystalline. This can be
explained by the imperfect packing when the microfibrils are formed. Microfibrils
consisting of aligned cellulose chains are the main structural components of cot-
ton fibers.
Figure 5.3 shows the cellular structure of a cotton fiber. A cotton fiber basically
is a large single cell. During the fiber formation, the cell first grows to the full
length of the fiber by stacking microfibrils to form a primary wall. The microfi-
brils in the primary wall are randomly oriented. After the formation of the primary
wall, secondary walls are deposited in daily growth rings inside the primary wall,
leaving a small lumen at the center. Secondary walls include parallel microfibrils
aligned with helix angles ranging from 20°–30°. In addition, there is also a cuticle
layer outside the primary wall. The cuticle is a waxy protective barrier that pro-
vides water resistance to the fibers as they grow. During wet processing of cotton
fibers, the cuticle is typically removed.
The lumen is a hollow tube and it is filled with fluid when the cotton fibers are
growing. Fresh cotton fibers have a circular cross-sectional structure. However,
when cotton fibers are dried, the lumen losses its fluid and collapses (Figure 5.4).
The collapse of the lumen leads to the kidney-bean cross-sectional shape and flat,
twisted longitudinal structure of dried cotton fibers (Figure 5.5).
Figure 5.3. Cellular structure of a cotton fiber.

Figure 5.4. The collapse of lumen in a cotton fiber.
Figure 5.5. SEM images of cotton fibers with (A) cross-sectional and (B) longitudinal
views.
5.1.2 Other Natural Cellulose Fibers

Like in cotton, cellulose chains in other natural cellulose fibers also crystallize
by forming the same unit cell structure as shown in Figure 5.2. Cellulose chains
stack together to form microfibrils. The diameters of microfibrils in most natural
cellulose fibers are in the same range, but the internal structure of the microfibrils
show different degrees of crystallinity. In addition, the helix angles of microfi-
brils are different in different natural cellulose fibers (Table 5.1). Except coir, all
other natural cellulose fibers have lower helix angles than cotton when microfi-
brils stack together to form the walls. Smaller spiral angles result in higher tensile
moduli for these fibers.
Another major difference among different natural cellulose fibers is their mor-
phology. Kapok fibers are obtained from the seed pods of the tropical kapok tree.
Like cotton, kapok fibers are unicellular fibers. However, kapok fibers do not col-
lapse and twist after dried (Figure 5.6). Dried kapok fibers have circular, hollow
(lumen) cross-sectional structure with total wall thicknesses of around 2 μm and
fiber diameters ranging from 15 to 35 μm. As a result, kapok fibers have lower
densities (0.31–0.38 g/cm3) than most other natural cellulose fibers.
Many other natural cellulose fibers form bundles. For example, short flax fi-
bers (27–36 mm) overlap each other and are held together by a mixture of non-
cellulosic polymers, including hemicellulose, lignin, and pectins (Figure 5.7).
The resultant fiber bundles have irregular shapes and contain multiple lumens.
Individual flax fibers have an average diameter of around 20 μm, but the bundle
Table 5.1. Helix angles of natural cellulose fibers.
Cellulose Fiber Plant Origin Helix Angle (°)

Cotton Seed 20–30
Kapok Seed —
Hemp Bast 6.2
Ramie Bast —
Flax Bast 5
Jute Bast 8.1
Kenaf Bast —
Bagasse Bast —
Bamboo Bast —
Phormium Leaf —
Pineapple Leaf 6–14
Abaca Leaf —
Banana Leaf 11–12
Sisal Leaf 10–22
Coir Fruit 39–49
Figure 5.6. SEM image of a kapok fiber. (Abdullah, M.A., et. al., Journal of Hazardous
Materials, 177, 683–691, 2010.)
size varies considerably. Figure 5.8 through Figure 5.10 show SEM images of

bundle structures formed by kenaf, hemp, and sisal fibers, respectively.
The sizes and densities of natural cellulose fibers vary considerably. Table 5.2
shows the typically lengths, diameters, and densities of natural cellulose fibers.
The actual sizes and densities are affected by the farming technique, growth envi-
ronment, geographical location and maturity of the plant.
Figure 5.7. Cross-section schematic (A) and SEM image (B) of flax fibers. (Baley, C., et.
al., Materials Letters, 60, 2984–2987, 2006.)
Figure 5.8. SEM image of kenaf fibers. (Shibata, S., et. al., Polymer Testing, 24, 1005–
1011, 2005.)
Figure 5.9. SEM image (longitudinal) of hemp fibers. (Troëdec, M.L., et. al., Journal of
the European Ceramic Society, 29, 1861–1868, 2009.)
Figure 5.10. SEM image of sisal fibers. (Thomason, J.L., et. al., Composites Science and
Technology, 71, 1008–1015, 2011.)

Manufactured cellulose fibers have a very different physical structure from natu-
ral cellulose fibers. The crystal lattice formed by native cellulose chains in natural
cellulose fibers is called Cellulose I (Figure 5.2). When cellulose is dissolved and
spun into fibers, the crystal lattice changes to Cellulose II. Table 5.3 compares
the unit cell parameters and crystal densities of Cellulose I and Cellulose II. The
Table 5.2. Lengths, diameters and densities of natural cellulose fibers.
Fiber Length Fiber Diameter Bulk Density (g/

Cellulose Fiber Plant Origin
(mm) (μm) cm3)
Cotton Seed 20–64 11.5–17 1.55
Kapok Seed 8–32 15–35 0.31–0.38
Hemp Bast 8.3–14 17–23 1.40–1.50
Ramie Bast 60–250 28.1–35 1.55
Flax Bast 27–36 17.8–21.6 1.40–1.50
Jute Bast 1.9–3.2 15.9–20.7 1.30–1.50
Kenaf Bast — 17.7–21.9 1.22–1.40
Bagasse Bast 1.7 20 0.550–1.35
Bamboo Bast 2.7 10–40 1.50
Phormium Leaf 5.0–5.7 15.4–16.4
Pineapple Leaf — 20–80 1.44–1.56
Abaca Leaf 4.6–5.2 17–21.4 1.50
Banana Leaf 2–3.8 — 1.30–1.35
Sisal Leaf 1.8–3.1 18.3–23.7 1.30–1.55
Coir Fruit 0.9–1.2 16.2–19.5 1.15–1.25
Table 5.3. Edge lengths, internal angles and crystal

densities of Cellulose I and Cellulose II.
Edge Lengths (nm) Internal Crystal
Crystal Type Angle β Density (g/
a b c (o) cm3)
Cellulose I 0.835 1.030 0.790 84 1.625
Cellulose II 0.802 1.030 0.903 63 1.620
unit cell size of Cellulose II is slightly larger than that of Cellulose I, leading to a
lower crystal density.
Another important difference between Cellulose I and Cellulose II is the ori-
entation of cellulose chains. In Cellulose I, all cellulose chains are aligned in the
same parallel direction, while those in Cellulose II have anti-parallel directions. In
some manufactured cellulose fibers, the chemical structure of cellulose chains is
altered by replacing the hydroxyl groups. Acetate and triacetate fibers are the most
important examples of such fibers. In acetate and triacetate fibers, the polymer
chains also have anti-parallel arrangement when they crystallize. However, both
acetate and triacetate fibers do not have sufficient hydroxyl groups to form a huge
amount of hydrogen bonds between adjacent chains, and hence their degrees of
crystallinity and bulk densities are relatively low.
The crystallite morphology of manufactured cellulose fibers is similar to that
of synthetic polymer fibers since both are man-made spun fibers. Two basic mod-
els often are used for manufactured cellulose fibers: fringed micelle model and
fringed fibril model. In both cases, polymer chains are more or less aligned along
the fiber axis. The fringed fibril model with fringe chains connecting the crys-
tallites is the more appropriate model for most manufactured cellulose fibers.
Figure 5.11 shows a fringed fibril model proposed by Fink and his coworker for
describing the crystalline structure of manufactured cellulose fibers. Although
thfringed fibril model is more desirable, the fringed micelle model still is suitable
for a few manufactured cellulose fibers such as rayon.
Manufactured cellulose fibers are spun from solutions. During fiber forma-
tion, the fiber skin is solidified first and the sequence solidification of the fiber
core reduces the fiber diameter and deforms the fiber skin. As a result, manufac-
tured cellulose fibers typically have an irregular, crenulated cross-sectional shape.
Figure 5.12 shows a transmission electron microscopy (TEM) image of a rayon
fiber. Due to the different solidification rates of the fiber skin and core, this par-
ticular rayon fiber has a skin-core morphology, in which the core has voids in
the order of 20 –150 nm and the skin is a densified layer of about 1.5–2.5 μm
thick. The crystalline structures in the skin and core also are different. Figure 5.13
shows schematically the crystallite orientations in skin and core regions. The
cross-sectional collapse occurred during fiber formation leads to a radial orienta-
tion of crystallites in the skin, but the core has more disordered orientation.
Figure 5.11. Fringed fibril model for manufactured cellulose fibers. (Fink, H.P., et. al.,
Journal of Applied Polymer Science, 30, 3779–3790, 1985.)
Figure 5.12. TEM image of a rayon fiber. (Fink, H.P., at. al., Progress in Polymer Science,
26, 1473–1524, 2001.)
Figure 5.13. Crystallite orientation in skin and core of a rayon fiber.

Manufactured cellulose fibers with other morphologies can be obtained by us-

ing non-circular spinneret holes to achieve different cross-sectional shapes, or by
injecting a gas into the spinning solution to produce inflated fibers. In addition,
the “all-skin” morphology also is obtained in fibers used for tire yarns by adding
dimethylamine or dimethylformamide.

5.3.1 Wool
The physical structure of wool is complex. The cross-section of wool is circular,
or slightly elliptical in shape. The diameter of wool fibers is in the range of 20–40
μm. Figure 5.14 shows a schematic of the cross-section of a wool fiber, which
consists of three major components, cuticle, cortex, and medulla.
Mechanically, the most important component of wool fibers is the cortex.
Figure 5.15 shows the structure of the cortex in a wool fiber. The protein in wool
fibers is called keratin. Three keratin chains in wool form helices with helix angles
ranging from 30° to 35°. Three helix chains twist compactly together to form a
protofibril with a diameter of around 2 nm. Eleven protofibrils assemble to form
a microfibril with a diameter of 7–8 nm. Microfibrils also are called intermediate
filaments and they helically wind together into a macrofibril with a diameter of
Figure 5.14. Cross-section structure of a wool fiber.
Figure 5.15. Structure of the cortex in a wool fiber.

about 0.4–0.5 μm. Macrofibrils are aligned to form a cortical cell. Cortical cells
assemble to form the cortex component of the wool fiber.
Wool fibers contain three types of cortical cells, called ortho-, meso- and para-
cortical cells. Figure 5.16 shows the distribution of these three types of cells in a
wool fiber. Cortical cells are separated from each other by a cell membrane com-
plex (CMC) with a thickness of around 25 nm. Most wool fibers mainly contain
ortho- and para-cortical cells, which are arranged bilaterally. In ortho-cortical
cells, the macrofibrils are discrete and twisted along their long axis, with micro-
fibrils forming a whorl pattern due to the twisting of the peripheral microfibrils
around the cell core. In para-cortical cells, the macrofibrls are fused together
around the cell periphery and the microfibrils are all parallel to the fiber axis. In
addition, para-cortical cells contain more sulfur-containing cysteine groups than
ortho-cortical cells. Due to the structural differences between ortho- and para-
cortical cells, wool fibers typically have a natural helical crimp with the para-
cortical cells on the inside.
The cuticle also is an important component of wool fibers. Figure 5.17 shows
a SEM image of a wool fiber. It is seen that the cuticle is an overlapping scale
structure covering the cortex of wool fibers. The thicknesses of the cuticle and the
scale are both 1–2 μm, indicating the scales are just overlapped slightly to provide
a covering for the fiber with the thickness of one or two layers of scales. The scale
structure leads to a directional friction effect, with the fibers being smoother in
the growth direction. Unlike the cortex, in which the keratin chains crystallize by
forming helices, the cuticle cells are largely amorphous.
The third component is the medulla, which may be empty or may contain
a group of open cell walls. The cells in medulla do not contain sulfur in their
Figure 5.16. TEM image of the cross-section of a wool fiber. Ortho-cortical (O), meso-
cortical (M) and para-cortical (P) cells and the cuticle (Cu) are indicated. (Marshall, R.C.,
et. al., Electron Microscopy Reviews, 4, 47–83, 1991.)
Figure 5.17. SEM image of a wool fiber.
proteins. Mechanically, the medulla represents empty space. The medulla can be
easily observed in coarse wool fibers, but it may not be present in some finer
fibers.
In addition to wool fibers obtained from sheep, there are many other animal
hair fibers. These animal hair fibers differ from each other mainly in fiber diame-
ter, length, number of scales in the cuticle, and natural color. For example, Mohair
comes from Angora goat, and it possesses little crimp and has flat scales. Camel
hair is the hair fiber of the two-humped camels of Mongolia, Tibet and other ar-
eas of Asia. Camel hair has poorly defined scales and it does not felt like wool.
Cashmere is an extremely fine and soft fiber from Asian Cashmere goat. Alpaca
fiber is a hair fiber from the alpaca of South America, and it is durable, silky, and
lustrous. Vicuna, from the vicuna of South America, is lightweight and has fine
scales that result in a smooth hand and high luster.
5.3.2 Silk
Silk has a simpler structure than wool. The protein in silk is called fibroin
and it mainly has four types of simple side groups: -H, -CH3, -CH2OH, and –
CH2C6H5OH. Sulfur-containing cystine groups and other side groups are in small
quantities. The intermolecular bonds between polymer chains are hence mostly
the hydrogen bonds between >C=O and >N-H groups on the main chains, and a
few hydrogen bonds and ionic secondary bonds between side groups.
The fibroin chains are held together by hydrogen bonds and are aligned along
the fiber axis to form β-pleated sheets, leading to the formation of partially ex-
tended chain crystallites (Figure 5.18). However, the bulky side groups are un-
desirable for forming crystallites. Hence, the sections of the fibroin chains that
contain bulky side groups tend to be distributed in the amorphous phase.
In addition to fibroin protein, the as-spun silk produced by silkworm has an-
other component, called sericin, which also is a protein. The un-degummed silk
has around 75% fibroin and 25% sericin. Figure 5.19 shows a schematic of un-
degummed silk. It is seen that two fibroin fibers are held together with sericin
coating. Fibroin is insoluble in water, but sericin can be removed by boiling with
soap water, i.e., the degumming process. Figure 5.20 shows a degummed silk
fiber, which has a triangle cross-section.
Figure 5.18. Possible alignment of fibroin chains in silk.
Figure 5.19. Schematic of un-degummed silk.

Figure 5.20. SEM image of degummed silk. (Zhang, K., et. al., Acta Biomaterialia, 6,
2165–2171, 2010.)

Manufactured protein fibers typically are made from proteins of milk, peanut,
corn, soybean, etc. These proteins often are coiled into a compact ball-shape
structure, with coils linked together by a large number of intermolecular bonds.
Treatments have to be carried out to destroy these intermolecular bonds and align
protein chains to form fibers.
Manufactured protein fibers developed between late nineteenth and mid-twen-
tieth centuries often had poor mechanical properties, and they often were blended
with other natural or synthetic polymer fibers. Some manufactured protein fibers,
such as milk-based protein fibers, had an unpleasant odor when wet. As a re-
sult, due to technical and economic challenges, these manufactured protein fibers
failed to become mainstream fibers.
Since late twentieth century, manufactured protein fibers with different struc-
tures and properties are being developed with new technologies, such as the use
of chemical and biochemical treatments and the spinning of protein/synthetic
polymer bicomponent fibers. Figure 5.21 shows a SEM image of a core-sheath
soybean protein/polyvinyl alcohol bicomponent fiber.
REFERENCES
[1] Abdullah, M.A., Rahmah, A.U., and Man, Z., “Physicochemical and
Sorption Characteristics of Malaysian Ceiba Pentandra (L.) Gaertn. as a
Natural Oil Sorbent”, Journal of Hazardous Materials, 177, 683–691, 2010.
[2] Ansell, M.P., and Mwaikambo, L.Y., “The Structure of Cotton and Other
Plant Fibres”, in Handbook of Textile Fibre Structure, Volume II: Natural,
Regenerated, Inorganic and Specialist Fibres, editors Eichhorn, S.J., Hearle,
J.W.S., Jaffe, M., and Kikutani, T., Woodhead Publishing Limited, 2009.
References 79
Figure 5.21. SEM image of a core-sheath soybean protein/polyvinyl alcohol fiber. (Zhang,

Y., et. al., Journal of Applied Polymer Science, 71, 11–19, 1999.)
[3] Baley, C., Perrot, Y., Busnel, F., Guezenoc, H., and Davies, P., “Transverse
Tensile Behaviour of Unidirectional Plies Reinforced with Flax Fibres”,
Materials Letters, 60, 2984–2987, 2006.
[4] Brooks, M.M., “Regenerated Protein Fibres: a Preliminary Review”, in
[6] Ciechanska, D., Wesolowska, E., and Wawro, D., “An Introduction to
Cellulosic Fibres”, in Handbook of Textile Fibre Structure, Volume II:
Natural, Regenerated, Inorganic and Specialist Fibres, editors Eichhorn,
S.J., Hearle, J.W.S., Jaffe, M., and Kikutani, T., Woodhead Publishing
Limited, 2009.
[7] Dyer J., and Daul, G., “Rayon Fibers”, in Handbook of Fibre Science and
Technology, Volume IV. Fibre Chemistry, editors Lewin, M., and Pearce,
E.M., Marcel Dekker, 1985.
[8] Feughelman, M., “Natural Protein Fibers”, Journal of Applied Polymer
Science, 83, 489–507, 2002.
[9] Fink, H.P., Philipp, B., “Models of Cellulose Physical Structure form the
Viewpoint of the Cellulose I to Cellulose II Transition, Journal of Applied
Polymer Science, 30, 3779–3790, 1985.
[10] Fink, H.P., Weigel, P., Purz, H.J., and Ganster, J., “Structure Formation
of Regenerated Cellulose Materials from nmMO-Solutions”, Progress in
Polymer Science, 26, 1473–1524, 2001.
[11] Ganster, J., Pink, H.P., “The Structure of Man-Made Cellulosic Fibres”,
[12] Harris, M., Handbook of Textile Fibers, Harris Research Laboratories, 1954.
[13] He, J.H., Ren, Z.F., Fan, J., and Xu, L., “Hierarchy of Wool Fibers and its
Interpretation Using E-Infinity Theory”, Chaos, Solitons and Fractals, 41,
1893–1841, 2009.
[14] Hearle, J.W.S., “An Introduction to Protein Fibres”, in Handbook of Textile
Fibre Structure, Volume II: Natural, Regenerated, Inorganic and Specialist
Fibres, editors Eichhorn, S.J., Hearle, J.W.S., Jaffe, M., and Kikutani, T.,
[15] Hearle, J.W.L., “Physical Structure and Properties of Cotton”, in Cotton:
[16] Hearle, J.W.S., “Physical Structure and Fibre Properties”, in Regenerated
Cellulose Fibres, editor Woodings, C., Wooding Publishing Limited, 2001.
[18] Khalifa, I.B., Ladhari, N., Touay, M., “Application of Sericin to Modify
Textile Supports”, Journal of the Textile Institute, 103, 370–377, 2012.
[19] Kosan, B., Michels, C., and Meister, F., “Dissolution and Forming of
Cellulose with Ionic Liquid”, Cellulose, 15, 59–66, 2008.
[20] Lim, T.T., and Huang, X., “Evaluation of Kapok (Ceiba Pentandra (L.)
Gaertn.) as a Natural Hollow Hydrophobic—Oleophilic Fibrous Sorbent
for Oil Spill Cleanup”, Chemosphere, 66, 955–963, 2007.
[21] Marshall, R.C., Orwin, D.F.G., and Gillespie, J.M., “Structure and
Biochemistry of Mammalian Hard Keratin”, Electron Microscopy Reviews,
4, 47–83, 1991.
[22] McKittrick, J., Chen, P.Y., Bodde, S.G., Yang, W., Novitskaya, E.E., and
Meyers, M.A., “The Structure, Functions, and Mechanical Properties of
Keratin”, The Journal of The Minerals, Metals & Materials Society, 64,
449–468, 2012.
References 81
[23] Meyer, K.H., and Misch, L., “Cellulose: Crystal Structure”, Helv Chim
Acta, 20, 232–244, 1937.
2006.
[25] O’Sullivan, A.C., “Cellulose: the Structure Slowly Unravels”, Cellulose, 4,
173–207, (1997).
[26] Qi, H., Cai, J., Zhang, L., Nishiyama, Y., and Rattaz, A., “Influence of
Finishing Oil on Structure and Properties of Multi-Filament Fibers from
Cellulose Dope in NaOH/Urea Aqueous Solution”, Cellulose, 15, 81–89,
2008.
[27] Shibata, S., Cao, Y., and Fukumoto, I., “Press Forming of Short Natural
Fiber-Reinforced Biodegradable Resin: Effects of Fiber Volume and Length
on Flexural Properties”, Polymer Testing, 24, 1005–1011, 2005.
[28] Thomason, J.L., Carruthers, J., Kelly, J., and Johnson, G., “Fibre Cross-
section Determination and Variability in Sisal and Flax and its Effects on
Fibre Performance Characterisation”, Composites Science and Technology,
71, 1008–1015, 2011.
[29] Tortora, P.G., Understanding Textiles, Fourth Edition, Macmillian
Publishing Company, 1992.
[30] Troëdec, M.L., Peyratout, C.S., Smith, A., Chotard, T., “Influence of Various
Chemical Treatments on the Interactions between Hempfibres and a Lime
Matrix”, Journal of the European Ceramic Society, 29, 1861–1868, 2009.
[31] Vollrath, Porter, D., Dicko, C., “The Structure of Silk”, in Handbook of
Textile Fibre Structure, Volume II: Natural, Regenerated, Inorganic and
Specialist Fibres, editors Eichhorn, S.J., Hearle, J.W.S., Jaffe, M., and
Kikutani, T., Woodhead Publishing Limited, 2009.
[34] Woodings, C., “Applications Development”, in Regenerated Cellulose
Fibres, editor Woodings, C., Woodhead Publishing, 2001.
[35] Mortmann, “The Structure and Properties of Wool and Hair Fibres”, in
[36] Zhang, K., Si, F.W., Duan, H.L., and Wang, J., “Microstructures and
Mechanical Properties of Silks of Silkworm and Honeybee”, Acta
Biomaterialia, 6, 2165–2171, 2010.
[37] Zhang, Y., Ghasemzadeh, S., Kotliar, A.M., Kumar, S., Presnell, S. Williams,
L.D., “Fibers from Soybean Protein and Poly(vinyl alcohol)”, Journal of
Applied Polymer Science, 71, 11–19, 1999.
PROBLEMS
(1) Natural cellulose and manufactured cellulose fibers have different crystal-
line structures. Describe the difference.
(2) Describe the cellular structure of cotton fibers.
(3) What is the helix angle of the microfibrils when they are aligned in parallel
in the secondary walls of cotton fibers?
(4) How do the cross-sectional and longitudinal shapes change when fresh cot-
ton fibers are dried?
(5) What are the major differences between the physical structures of cotton
and other natural cellulose fibers?
(6) Describe the main physical structure features of manufactured cellulose
fibers.
(7) Describe the physical structure of wool.
(8) Most wool fibers have a natural helical crimp. Why?
(9) Describe the physical structure of silk.
(10) What are the major differences between the physical structures of wool and
silk?
CHAPTER 6
Structure of Non-Polymer Fibers
A wide range of non-polymer fibers, such as carbon, glass, silicon carbide, boron,
asbestos, and metal fibers, now is available commercially. Among them, carbon
fibers often are made by heat-treating precursor fibers of polymer materials, such
as polyacrylonitrile and rayon. Glass fibers, silicon carbide fibers and metal fibers
are made directly from non-polymer raw materials. Compared with polymer fi-
bers, non-polymer fibers often are stronger, stiffer, more heat resistant, and less
flammable. However, except for metal fibers, non-polymer fibers also are charac-
terized by their brittleness. These property characteristics are directly related to
the atomic arrangement and the defect structure of non-polymer fibers. There are
many different types of non-polymer fibers, but this chapter only discusses the
structure of two most used non-polymer fibers: carbon and glass fibers.
6.1 CARBON FIBERS

Carbon fibers have diameters ranging from 5 to 10 μm, and they are mainly com-
posed of carbon atoms. Carbon fibers have excellent properties, such as high
modulus, high strength, low weight, high chemical resistance, high temperature
tolerance and low thermal expansion, and often are combined with other materials
to form high-strength, high-modulus, and low-weight composites.
6.1.1 Atomic Order

The ideal graphite crystal is composed of infinite sheets of carbon atoms arranged
in a hexagonal pattern, i.e., graphene sheets. These graphene sheets are stacked in
a regular ABAB sequence with atoms in alternate planes aligning with each oth-
er (Figure 6.1). An important characteristic of the graphite crystal is the different
83
84 Structure of Non-Polymer Fibers
Figure 6.1. Arrangement of carbon atoms in graphite.
bonding types within and between the graphene sheets. Within the graphene sheets,
the carbon atoms are held together by strong covalent bonds and the bond length is
0.142 nm. Van der Waals forces hold these graphene sheets together and the distance
between adjacent sheets is 0.335 nm. This unique characteristic makes the graphite
crystal one of the most anisotropic structures, i.e., exhibiting different properties
when measured along different directions. For example, the van der Waals forces
can break easily under external stress, and as a result the graphene sheets slide over
one another. This sliding makes graphite an excellent lubricant.
Unlike a graphite crystal that is formed by perfect graphene sheets, the basic
unit of carbon fibers is the imperfect graphene sheet. Imperfect graphene sheets
contain a high number of defects, such as vacancies and impurity atoms, and are
strongly distorted in-plane (Figure 6.2). Carbon fibers are based on the stack-
ing of these imperfect graphene sheets. Due to the high-density in-plane defects
and distorted sheet shape, the stacking of graphene sheets in carbon fibers is less
ordered than that in graphite crystal, leading to a so-called “turbostratic” struc-
ture (Figure 6.3). The arrangement of the sheet planes determines the mechanical
properties of carbon fibers. To achieve high fiber modulus and strength, the good
alignment of the graphene sheets parallel to the fiber axis is desired. Although
the alignment of graphene sheets in the “turbostratic” structure of carbon fibers
is not perfect, it can be improved by selectively adjusting the processing pa-
rameters used during the fiber formation. For example, when the heat-treatment
Carbon Fibers 85
temperature exceeds 1700°C, carbon fibers are converted to graphite fibers, in

which the alignment of graphene sheets is similar to that in graphite crystals.
6.1.2 Morphology
Carbon fibers typically are straight, without crimp (Figure 6.4). The cross-section-
al shape of carbon fibers typically is circular (Figure 6.5). However, carbon fibers
with other cross-sectional shapes also exist.
The arrangement of crystallites in the cross-section of carbon fibers is com-
plex and is affected by the precursor type and structure, fiber spinning conditions,
Figure 6.2. Imperfect graphene sheet in carbon fibers.
Figure 6.3. Turbostratic structure of carbon fibers.

Figure 6.4. Longitudinal SEM image of carbon fibers. (Sharma, S.P., et. al., Surface and
Coatings Technology, 203, 1329–1335, 2009.)
Figure 6.5. Cross-sectional SEM image of carbon fibers. (Petrović, Ž., et. al., Thin Solid
Films, 513, 193–200, 2006.)
and carbonization temperature. In carbon fibers, graphene sheets stack together to

form crystallites and the stacking thickness varies from a few nanometers to tens
of nanometers. These crystallite sheets are preferentially oriented along the fiber
axis. Figure 6.6 shows the arrangement of crystallite sheets in a carbon fiber. In this
particular carbon fiber, the crystallite sheets twist, fold and join each other to form a
radial texture at the fiber cross-section. In addition to the radial structure, other types
of cross-sectional morphology can be found in carbon fibers. Figure 6.7 shows a
few examples of possible cross-sectional morphology of carbon fibers.
6.2 GLASS FIBERS

Glass fibers are based on silica (SiO2), with the addition of different oxides. Glass
fibers also have good mechanical properties and often are used as an insulating
material or reinforcement for making composites. Although they are not as strong
as carbon fibers, glass fibers are much cheaper and less brittle.
Glass Fibers 87
Figure 6.6. Arrangement of crystallite sheets in a carbon fiber. (Naito, K., et. al., Carbon,
46, 189–195, 2008.)
Figure 6.7. Possible cross-sectional morphology of carbon fibers.
6.2.1 Atomic Order

In its pure form, silica exist as a macromolecule, (SiO2)n, in which SiO4 groups
are configured as a tetrahedron with the silicon atom in the center and four oxygen
atoms at the corners (Figure 6.8). When SiO2 crystallizes, the SiO4 tetrahedrons
form crystals by sharing the oxygen atoms at the corners. Figure 6.9 shows the
two-dimensional arrangement of silicon and oxygen atoms in a SiO2 crystal.
Unlike pure SiO2 crystals, the atoms in glass fibers are arranged in a less or-
dered pattern (Figure 6.10). This is related to the fiber formation process. During
fiber formation, the raw materials are melted, extruded, and allowed to cool
quickly. The viscosity of the fluid is so high that the atoms are unable to move
to the low potential energy positions in the crystal unit cells. The disordered pat-
tern of the liquid is “frozen” into the solid fiber. Therefore, glass fibers usually
are amorphous and regarded as a supercooled fluid. Glass fibers are transparent
because the length scale over which the disorder occurs is smaller than the wave-
length of the visible light and is unable to scatter the light.
Figure 6.8. Tetrahedral structure of SiO2.
Figure 6.9. Two-dimensional arrangement of silicon and oxygen atoms in a SiO2 crystal.
In pure silica, each silicon atom is linked to four oxygen atoms by covalent
bonding, and hence a high processing temperature is required for fiber formation.
Therefore, it is common to introduce impurity atoms (such as Ca, Na, K, etc.) that
have lower valence to break up the network by forming ionic bonds with oxygen
(Figure 6.11). The introduction of impurities lowers the processing temperature
and the cost of the glass fibers. However, the strength and modulus of glass fibers
also are reduced.
6.2.2 Composition
Commercial glass fibers contain significant amounts of impurity atoms that lower
the processing temperature and the product cost. Atoms listed in the first column
(e.g., Na and K), second column (e.g., Mg, Ca, and Ba) and third column (e.g., B
and Al) of the periodic table make up a significant percentage of the composition
of glass fibers. But other atoms (e.g., Fe and Ti) also are introduced in many glass
Glass Fibers 89
Figure 6.10. Two-dimensional arrangement of silicon and oxygen atoms in glass fibers.
Figure 6.11. Atomic arrangement in glass fibers after the introduction of impurities.
fibers. These atoms exist as cations and disrupt the corner bonding of the tetrahe-
dral SiO4 units, resulting in significant changes in the properties of glass fibers.
Table 6.1 shows the composition of several important glass fibers.
Glass fibers are designated as “A” glasses when they are made of soda-lime-
silica glasses that are similar to these used in glass bottles and flat glass. Among
the glass fibers shown in Table 6.1, A-glass fibers have the highest SiO2 content.
On the other hand, E-glass fibers (E for electrical) has the lowest SiO2 content.
They draw well and have good mechanical, electrical and weathering proper-
ties. S-glass (S for strength) and C-glass (C for corrosion) fibers have the same
SiO2 content, but the contents of other components are different. S-glass fibers
have higher strength, modulus, and thermal resistance than other types of glass
Table 6.1. Composition of different glass fibers.
A Glass E-glass S-Glass C-Glass Wool Glass

SiO2 71.8 52.4 64.4 64.4 63.5
Na2O + K2O 14.2 0.8 0.3 9.6 16.2
CaO 8.8 17.2 — 13.4 9.0
MgO 3.8 4.6 10.3 3.3 5.2
B2O3 — 10.6 — 4.7 2.9
BaO — - — 0.9 —
Al2O3 + Fe2O3 1.5 14.4 25.0 4.1 3.0
fibers, but C-glass fibers have better resistance to correction. Wool glass fibers,
also called fiberglasses, are used as thermal insulators for buildings.
6.2.3 Morphology
Glass fibers are amorphous and have simple morphology. Figures 6.12 and 6.13
show longitudinal and cross-sectional SEM images of glass fibers, respectively.
Most glass fibers are straight and have circular cross-section. However, glass fi-
bers with other cross-sectional structures also exist. Figure 6.14 shows a SEM
image of hollow glass fibers. These hollow glass fibers have higher rigidity than
regular fibers. When used as the reinforcement in composites, hollow glass fibers
cannot only provide the desired structural improvement, but also introduce a res-
ervoir suitable for the containment of a healing agent, which makes the compos-
ites self-healing.
Figure 6.12. Longitudinal SEM image of glass fibers. (Etcheverry, M., et. al., Composites
Part A: Applied Science and Manufacturing, 39, 1915–1923, 2008.)
References 91
Figure 6.13. Cross-sectional SEM image of glass fibers. (Lassila, L.V.J., et. al.,
Biomaterials, 23, 2221–2229, 2002.)
Figure 6.14. SEM image of hollow glass fibers. (Murphy, E.B., et. al., Progress in Polymer
Science, 35, 223–251, 2010.)
REFERENCES
[1] Cato, A.D., and Edie, D.D., “Flow Behavior of Mesophase Pitch”, Carbon,
41, 1411–1417, 2003.
[2] Chand, S., “Review Carbon Fibers for Composites”, Journal of Materials
Science, 35, 1303–1313 (2000).
[3] Donnet, J.B., and Qin, R.Y., “Study of Carbon Fiber Surfaces by Scanning
Tunnelling Microscopy, Part I. Carbon Fibers from Different Precursors and
After Various Heat Treatment Temperatures”, Carbon, 30, 787–796, 1992.
[4] Etcheverry, M., Ferreiraa, M.L., Capiati, N.J., Pegoretti, A., and Barbosa,
S.E., “Strengthening of Polypropylene-Glass Fiber Interface by Direct
Metallocenic Polymerization of Propylene onto the Fibers”, Composites
Part A: Applied Science and Manufacturing, 39, 1915–1923, 2008.
[5] Frank, E., Hermanutz, F., and Buchmeiser, M.R., “Carbon Fibers:
Precursors, Manufacturing, and Properties”, Macromolecular Materials
and Engineering, 297, 493–501, 2012.
[6] Hong, S.H., Korai, Y., and Mochida, I., “Development of Mesoscopic
Textures in Transverse Cross-section of Mesophase Pitch-Based Carbon
Fibers”, Carbon, 37, 917–930, 1999.
[7] Hull, D., and Clyne, T.W., An Introduction to Composite Materials, Second
Edition, Cambridge University Press, 1996.
[8] Johnson, D.J., Tomizuka, I., and Watanabe, O., “The Fine Structure of
Pitch-Based Carbon Fibres”, Carbon, 13, 529–534, 1975.
[9] Jones, F.R., and Huff, N.T., “The Structure and Properties of Glass Fibres”,
Handbook of Textile Fibre Structure. Volume 2: Natural Regenerated,
Inorganic and Specialist Fibres, Editors Eichhorn, S.J., Hearle, J.W.S.,
[10] Lassila, L.V.J., Nohrström, T., and Vallittu, P.K., “The Influence of Short-
Term Water Storage on the Flexural Properties of Unidirectional Glass
Fiber-Reinforced Composites”, Biomaterials, 23, 2221–2229, 2002.
[11] Matsumoto, T. “Mesophase Pitch and its Carbon Fibers”, Pure and Applied
Chemistry, 57, 1553–1562, 1985.
[12] Murphy, E.B., and Fred Wudl, F., “The World of Smart Healable Materials”,
Progress in Polymer Science, 35, 223–251, 2010.
[13] Naito, K., Tanaka, Y., Yang, J.M., and Kagawa, Y., “Tensile Properties of
Ultrahigh Strength PAN-Based, Ultrahigh Modulus Pitch-Based and High
Ductility Pitch-Based Carbon Fibers”, Carbon, 46, 189–195, 2008.
[14] Paris, O., and Peterlik, H., “The Structure of Carbon fibers”, Handbook
of Textile Fibre Structure. Volume 2: Natural Regenerated, Inorganic and
Specialist Fibres, Editors Eichhorn, S.J., Hearle, J.W.S., Jaffe, M., and
[15] Petrović, Ž., Metikoš-Huković, M., Z. Grubač, Z., and Omanović, S.,
“The Nucleation of Ni on Carbon Microelectrodes and Its Electrocatalytic
Activity in Hydrogen Evolution”, Thin Solid Films, 513, 193–200, 2006.
[16] Sharma, S.P., and Lakkad, S.C., “Morphology Study of Carbon Nanospecies
Grown on Carbon Fibers by Thermal CVD Technique”, Surface and
Coatings Technology, 203, 1329–1335, 2009.
[17] Sheshin, E.P., “Properties of Carbon Materials, Especially Fibers, For Field
Emitter Applications”, Applied Surface Science, 215, 191–200, 2002.
Problems 93
[18] Toyoda, M., and Inagaki, M., “Exfoliation of Carbon Fibers”, Journal of
Physics and Chemistry of Solids, 65, 109–117, 2004.
PROBLEMS
(1) Compared with polymer fibers, what are the unique features of non-poly-
mer fibers?
(2) What are the excellent properties of carbon fibers?
(3) Describe the turbostratic structure of carbon fibers.
(4) How are the crystallite sheets arranged in carbon fibers?
(5) What are the main differences between carbon and graphite fibers?
(6) What is the most important component in glass fibers? What are the other
possible components?
(7) Describe the atomic order of glass fibers.
(8) What are wool glass fibers? What is the major application for wool glass
fibers?
(9) Are common glass fibers amorphous or crystalline? Why?
(10) What are the typical longitudinal and cross-sectional shapes of carbon and
glass fibers?
CHAPTER 7
Structure of Nanofibers
Nanoscience is the study of atoms, molecules, and objects with sizes ranging from
1 to 100 nm. However, the term “nanofibers” traditionally has been used for fibers
with diameters less than 1000 nm. Nanofibers are an important class of material
that is useful in a variety of applications, including filtration, tissue engineering,
protective clothing, composites, battery separators, energy storage, etc. Methods
used to produce nanofibers include, but are not limited to, electrospinning, centrif-
ugal spinning, melting blowing, bicomponent fiber separation, phase separation,
template synthesis, and self-assembly. This chapter focuses on the main structural
characteristics of electrospun nanofibers. Chapter 12 will discuss the fiber forma-
tion process of electrospinning and the diameter control of electrospun nanofi-
bers. Chapter 13 will discuss the other nanofiber formation methods.
7.1 A BRIEF DESCRIPTION OF ELECTROSPINNING

PROCESS
Electrospinning is a simple, non-mechanical technique for fabricating nanofibers.
Figure 7.1 shows a nozzle-based electrospinning setup. High voltage is applied
between a spinning fluid contained in a syringe and a metallic collector. When the
voltage reaches a critical value, electrostatic forces overcome the surface tension
of the fluid and eject a liquid jet from the nozzle of the spinneret (metallic needle).
The electrically charged jet then undergoes a stretching-and-whipping process,
during which the jet diameter decreases from hundreds of micrometers to as small
as tens of nanometers. The as-spun nanofibers are accumulated on the surface of
the grounded collector. The electrospinning process is relatively slow, and a single
syringe shown in Figure 7.1 can only produce 1–100 mg nanofibers per hour. The
production rate can be improved significantly by connecting a large number of
95
96 Structure of Nanofibers
Figure 7.1. Schematic of a basic electrospinning setup.
individually charged syringes in parallel or by directly spinning from a liquid bath

using a nozzle-less process.
7.2 MATERIAL TYPE

Historically, electrospinning was used to produce nanofibers from polymer mate-
rials. However, electrospun nanofibers of carbon, ceramic, metal and composite
materials also have been recently developed.
7.2.1 Polymers
Nanofibers can be electrospun from both synthetic and natural polymers. Figures 7.2
and 7.3 show SEM images of electrospun polyacrylonitrile and silk fibroin nanofi-
bers, respectively.
During the electrospinning of polymer nanofibers, a high degree of molecular
alignment could be achieved by the large effective draw ratio exerted by the elec-
trostatic force. However, the crystallization process is retarded due to the rapid
solidification of the stretched polymer chains during electrospinning. Basically,
the stretched polymer chains do not have sufficient time to form crystallites.
Therefore, electrospun nanofibers often have lower degree of crystallinity and
lower crystallization temperature than the corresponding bulk material, but the
polymer chains in the amorphous phase may show a certain degree of orientation.
The crystallite structure of electrospun nanofibers can be described using a modi-
fied fringed micelle model (Figure 7.4). However, other crystallite structures also
can be formed by selectively adjusting the electrospinning parameters.
Material Type 97
Figure 7.2. SEM image of polyacrylonitrile nanofibers.
Figure 7.3. SEM image of silk fibroin nanofibers.
Figure 7.4. Modified fringed micelle model with the crystalline phase shadowed.
7.2.2 Carbons
Carbon nanofibers often are prepared by the electrospinning of precursor poly-
mers, such as polyacrylonitrile, polyimide, and pitch, etc., followed by the car-
bonization of the resultant precursor fibers. Figure 7.5 shows a SEM image of
carbon nanofibers, which were prepared by the carbonization of electrospun poly-
acrylonitrile nanofibers at 700°C in argon gas. The crystalline structure of carbon
nanofibers is mainly determined by the precursor type and carbonization con-
ditions, such as carbonization temperature and heating rate. Electrospun carbon
nanofibers shown in Figure 7.5 have a disordered carbon structure due to the rela-
tively low carbonization temperature. However, with increase in carbonization
temperature, carbon nanofibers present more ordered structure. When the carbon-
ization temperature is greater than 1800°C, carbon nanofibers could be converted
to graphite nanofibers, which have highly ordered graphite structure.
7.2.3 Ceramics
Ceramic nanofibers also can be synthesized by the calcination of electrospun
precursor nanofibers. Electrospun ceramic nanofibers were first reported by Dai
and his coworker in 2002. Since then, more than twenty ceramic systems (e.g.,
SiO2, Al2O3, and ZrO2) have been fabricated as nanofibers. Ceramic fibers can be
either amorphous or polycrystalline, depending on the material type and process-
ing conditions. Figure 7.6 shows ZrO2 nanofibers, which were synthesized from
zirconium oxychloride (ZrOCl2) precursor. During electrospinning, polyvinyl al-
cohol (PVA) was added to achieve desirable solution viscosity and improve the
electrospinnability. The electrospun ZrOCl2/PVA nanofibers were heat-treated to
form ZrO2 nanofibers, during which ZrOCl2 was converted to ZrOCl2 while PVA
was pyrolyzed.
Figure 7.5. SEM image of carbon nanofibers.

Material Type 99
Figure 7.6. SEM image of ZrO2 nanofibers.
7.2.4 Metals
Metal nanofibers also have been produced using electrospinning. Bognitzki et al.
fabricated Cu nanofibers by the electrospinning of a copper nitrate-polyvinylbu-
tyral (PVB) solution, followed by heat treatments in air and hydrogen. The heat
treatment in air converts copper nitrate to copper oxide and decomposes PVB,
while the heat treatment in hydrogen changes copper oxide to metallic copper. In
addition to copper nanofibers, iron, cobalt and nickel nanofibers also have been
prepared by the similar approach.
7.2.5 Composites
Composite nanofibers have been made by electrospinning solutions that contain
nanoparticles, such as silica, titania, carbon black, silver, and iron oxides. The
nanofiber matrices can be polymer, carbon, or ceramic. For carbon and ceramic
matrices, post-electrospinning treatments are needed to convert their precursors
into carbon and ceramic materials.
Figure 7.7 shows a TEM image of Si/PAN composite nanofibers. These nano-
fibers were prepared by direct electrospinning of a PAN solution containing Si
nanoparticles. One common problem of using this method is the aggregation of
nanoparticles inside the fiber matrix. The formation of aggregates can be mini-
mized by adding surfactants to the electrospinning solution, modifying the sur-
face of the nanoparticles, or using a soluble precursor of the nanoparticles. For ex-
ample, Ag/PAN composite nanofibers have been produced by the electrospinning
of AgNO3/PAN nanofibers, following by the reduction of AgNO3 to Ag nanoparti-
cles using plasma treatment. Since the Ag nanoparticles are reduced directly from
the soluble AgNO3 salt, they are homogeneously dispersed in the PAN nanofiber
matrix (Figure 7.8).
Figure 7.7. SEM image of Si/PAN composite nanofibers. (Ji, L., et al., Carbon, 47, 3219–
3226, 2009.)
Figure 7.8. TEM image of an Ag/PAN composite nanofiber. (Shi, Q., et. al., European
Polymer Journal, 47, 1402–1409, 2011.)
Composite nanofibers also have been made by introducing nanoparticles

into or onto nanofibers after electrospinning. Figure 7.9 shows a TEM image of
Pt-Ru/carbon composite nanofibers, which were fabricated by electrodeposit-
ing Pt nanoparticles directly onto the surface of electrospun carbon nanofibers.
Electrodeposition of metal particles can be carried out by applying a constant
voltage or successive voltammetric cycles. The nanoparticle size can be con-
trolled by selectively adjusting the electrodeposition conditions, such as voltage,
time, cycle rate, and cycle number. In addition to electrodeposition, chemical de-
position also can be used.
Fiber Morphology 101
Figure 7.9. TEM image of Pt-Ru/carbon composite nanofibers. (Lin, Z., et. al., Journal of
Materials Research, 25, 1329–1335, 2010.)
7.3 FIBER DIAMETER

Fiber diameter is the most important structure parameter for electrospun nano-
fibers. The smallest fiber diameter achieved so far is 1.6 nm by Huang and his
coworker for electrospun Nylon 4,6 nanofibers. However, most electrospun nano-
fibers have diameters ranging from 50 nm to 1 μm. The diameter of electrospun
nanofibers is dependent on a number of processing parameters: (i) intrinsic prop-
erties of the spinning solution such as material type, molecular weight, solution
concentration, rheological behavior, conductivity, and surface tension; (ii) opera-
tional conditions such as voltage, flow rate, nozzle diameter, spinneret configura-
tion, and spinneret-collector distance. Detailed discussion on the diameter control
for electrospun nanofibers is presented in Chapter 12.
7.4 FIBER MORPHOLOGY

Electrospun nanofibers usually exhibit a solid interior. However, electrospun
nanofibers with other morphologies also exist.
7.4.1 Porous Nanofibers

Porous nanofibers have large surface areas and are useful for applications such
as filtration, catalysis, fuel cells, supercapacitors, and tissue engineering. The po-
rous structure typically is obtained by adding pore generators during the electro-
spinning process. Examples of pore generators include, but are not limited to,
polymers, nanoparticles, and salts. These generators are added directly into the
electrospinning solutions. After electrospinning, they are removed to form small
pores. Figure 7.10 shows a SEM image of a porous carbon nanofiber, prepared
by using poly(L-lactide) (PLLA) as the pore generator. During the preparation of
these nanofibers, PAN and PLLA were first blended and electrospun. The as-spun
precursor nanofibers then were heat-treated in argon to decompose PLLA while
converting PAN to carbon. The decomposition of PLLA led to the formation of
a porous structure. Figures 7.11 and 7.12 show SEM images of porous carbon
and MnO2/carbon nanofibers, in which the pores were generated by using SiO2
nanoparticles and Mn acetate salt, respectively.
Figure 7.10. SEM image of porous carbon nanofibers, in which the pores were generated
by PLLA polymer. (Ji, L., et. al., Journal of Polymer Science: Part B: Polymer Physics,
47, 493–03, 2009.)
Figure 7.11. SEM image of porous carbon nanofibers, in which the pores were generated
by SiO2 nanoparticles. (Ji, L., et. al., Carbon, 47, 3346–3354, 2009.)
Fiber Morphology 103
7.4.2 Core-Sheath Nanofibers

Core-sheath nanofibers have been produced by co-electrospinning two different
spinning solutions using a spinneret comprising two coaxial needles. Figure 7.13
illustrates a schematic of the coaxial spinneret system. During electrospinning,
two different solutions are used as inner and outer fluids, respectively. Figure 7.14
shows a TEM image of a polybutadiene rubber (BR)/polyvinylpyrrolidone (PVP)
nanofiber prepared by the coaxial spinneret system. In this nanofiber, BR is the
core and PVP is the sheath.
7.4.3 Hollow Nanofibers

Hollow nanofibers also can be prepared using the same coaxial spinneret system,
as shown in Figure 7.13. To obtain the hollow structure, the material used for
Figure 7.12. SEM image of porous MnO2/carbon nanofibers, in which the pores were gener-
ated by Mn acetate salt. (Ji, L., et. al., Journal of Materials Chemistry, 19, 5593–5601, 2009.)
Figure 7.13. Schematic of a coaxial spinneret system.

Figure 7.14. TEM image of a core-sheath BR/PVP nanofiber. (Haiyan Wu, et. al., Materials
Letters, 84, 5–8, 2012.)
the fiber core must be degradable or extractable without damaging the sheath.
Figure 7.15 shows a SEM image of hollow TiO2 nanofibers. During electrospin-
ning, mineral oil was used as the core fluid and a titanium isopropoxide [Ti(OiPr)4]/
PVP solution as the outer fluid. The electrospun nanofibers were then heat-treated
to remove mineral oil and PVP while converting Ti(OiPr)4 to TiO2.
7.4.4 Multichannel Nanofibers

Multichannel nanofibers can be prepared by using a multichannel spinner-
et (Figure 7.16). Examples of multi-channel TiO2 nanofibers are shown in
Figure 7.17. The cross-sectional structures of the multichannel spinnerets used
for preparing these nanofibers also are shown. During electrospinning, paraffin oil
was used as the inner fluid and a titanium isopropoxide/PVP solution as the outer
fluid. The electrospun nanofibers were then heat-treated to decompose PVP and
covert titanium isopropoxide into TiO2 for forming the sheath. The paraffin oil
also was removed during heat treatment for creating the hollow channels.
7.4.5 Nanofibers with Other Morphologies

In addition to the different nanofibers discussed previously, nanofibers with other
morphologies, such as porous core-sheath, porous hollow, ribbon-like, branched,
helical structures, also have been developed.
Fiber Assemblies 105
Figure 7.15. SEM image of hollow TiO2 nanofibers. (Li, D., et. al., Advanced Materials,
16, 1151–1170, 2004.)
Figure 7.16. Schematic of a multichannel spinneret system.
7.5 FIBER ASSEMBLIES

In addition to the structures of individual nanofibers, the assembly structures play
a significant role in determining the properties and applications of electrospun
nanofibers. Electrospun nanofibers often are deposited as nonwoven porous mats,
and the pore size and porosity of these mats need be controlled. Other assembly
structures, such as continuous nanofiber filaments, aligned nanofiber sheets, nano-
fiber arrays, and nanofiber yarns, also have been produced by using special nano-
fiber collecting systems that can adjust the fiber jet movement through controlling
the distribution of electric field.
Figure 7.17. SEM images of multichannel TiO2 nanofibers with different channel numbers.
The inset in each figure shows the cross-sectional structure of the spinneret system used
for fabricating the nanofiber. (Zhao, Y., et. al., Journal of American Chemical Society, 129,
764–765, 2007.)
7.5.1 Nanofiber Nonwovens

The most important assembly for electrospun nanofibers is the randomly-oriented,
porous nonwoven mat. Electrospun nanofiber nonwovens typically are collected
by depositing dry fibers on the surface of a flat collecting plate, and the bending
instability associated with the spinning jet helps achieve the randomly oriented
nanofiber assembly. Through careful control of processing parameters, electros-
pun nanofiber nonwovens can be produced over a wide range of porosity values,
from nearly nonporous films to very porous and delicate fibrous structures. The
pore sizes of electrospun nanofiber nonwovens can vary from hundreds of nano-
meters to tens of micrometers.
According to the statistical geometry of electrospun nanofiber nonwovens,
the pore size is dominantly determined by the fiber diameter, but other fiber
variables, such as porosity, areal density, and fiber density, also have influence.
During electrospinning, the pore size can be reduced by decreasing the distance
between the spinneret and the collector, increasing the time of nanofiber collec-
tion, increasing polymer concentration, or decreasing applied voltage. Figure 7.18
shows that the pore size of PAN nanofiber mats decreases significantly when the
spinneret-collector distance decreases from 15 to 3 cm.
In most electrospinning setups, the nanofiber collecting plate is made of a con-
ductive material, such as aluminum foil, which is electrically grounded. When a
nonconductive material is used as the collector or placed on the collector surface,
charges on the electrospinning jet can quickly accumulate on the nonconductive
collecting surface, which hinders the further deposition of nanofibers. In addition,
due to the repulsive forces of the accumulated charges, nanofibers that are col-
lected on the nonconductive surface usually have a lower packing density than
those collected on a conducting surface. In some cases, the repulsive forces of the
accumulated charges on the deposited nanofibers may even cause the formation
of large-pore structures, such as dimples.
7.5.2 Nanofiber Filaments and Aligned Nanofiber Sheets

Continuous nanofiber filaments can be obtained by collecting fibers on a cylin-
drical collector with a high rotating speed, i.e., a rotating drum (Figure 7.19A).
By increasing the duration of electrospinning, flat sheets or tubes consisting of
aligned nanofibers also can be obtained by using this method. The orientation of
nanofibers in these assemblies is an important structure parameter, which can be
controlled by changing the rotating speed of the drum. Matthews et al. studied the
electrospinning of collagen nanofibers and found that, at low speeds (< 500 rpm),
a random mix of nanofibers was collected. However, when the rotating speed was
increased to 4,500 rpm, the nanofibers showed significant alignment along the
axis of rotation.
Figure 7.18. SEM images of PAN nanofiber nonwovens with different pore sizes.
Spinneret-collector distance: (A) 15 cm, and (B) 3 cm.
Figure 7.19. Schematics of nanofiber collecting systems for forming different nanofiber

assembly structures: (A) rotating drum for continuous nanofiber filament and aligned nano-
fiber sheet; (B) parallel electrodes for aligned nanofiber array; (C) counter-electrodes for
cross-bar nanofiber array; (D) blades for nanofiber yarn; (E) parallel rings for twisted nano-
fiber yarn; and (F) liquid bath for continuous nanofiber yarn.
One advantage of using a simple rotating drum is that a large area of aligned
nanofiber sheets can be easily fabricated. However, the disadvantage is that nano-
fibers are easy to break at high rotating speeds and highly aligned fiber sheets are
difficult to produce. Modifications have been made to lower the rotating speed,
avoid the nanofiber breakage, and/or improve the nanofiber alignment. For ex-
ample, a sharp pin structure can be introduced into the rotating drum to create an
electric field that starts from the tip of the needle and converges at the tip of the
sharp pin. Using this modified drum, Sundaray et al. fabricated highly aligned
polymethyl methacrylate (PMMA) nanofiber sheets at a rotating speed of 2,000
rpm. Figure 7.20 shows a SEM image of a thin sheet of aligned PMMA nanofi-
bers. Another simple approach is to use a rotating wire drum, with which highly
aligned nanofiber sheets can be produced at a rotating speed as low as 1 rpm.
Highly aligned nanofiber sheets also can be fabricated by placing knife-edge elec-
trodes below the rotating drum.
7.5.3 Nanofiber Arrays

The orientation of nanofibers in assemblies also can be controlled by using a col-
lector consisting of two parallel conducting electrodes separated by an insulating
gap (Figure 7.19B). The introduction of the insulating gap to the collector changes
the configuration of the electric field, which stretches the fibers to be aligned
Figure 7.20. SEM image of a thin sheet of aligned PMMA nanofibers. (Sundaray, B., et.
al., Applied Physics Letters, 84, 1222–1224, 2004.)
perpendicular to the electrodes to form a nanofiber array. The electrostatic repul-

sions between the charged nanofibers can further enhance the alignment in the
array. One major difference of these nanofiber arrays from the aligned nanofiber
sheets described in section 7.5.2 is that the size of nanofiber arrays is limited by
the width of the insulating gap, which typically is no more than several centi-
meters. When the gap width is too large, there will be a greater tendency for the
nanofibers to be deposited on the electrode rather than across the gap, and the
degree of alignment will be significantly decreased.
However, one advantage of this parallel-electrode collector over the rotating
drum is the resultant nanofiber arrays can be easily transferred onto other sub-
strates. Single nanofibers also can be collected across the gap and removed for
the construction of single-fiber devices. In addition, more complex nanofiber ar-
rays can be obtained by modifying the parallel-electrode collector system. For
example, cross-bar arrays of electrospun nanofibers can be obtained by using four
counter-electrodes, as shown in Figure 7.19C. A SEM image of cross-bar array of
PVP nanofibers is shown in Figure 7.21. Other complex arrangements of nano-
fiber arrays also can be obtained by adding more counter electrodes and/or re-
arranging the electrodes.
7.5.4 Nanofiber Yarns

One simple means to obtain nanofiber yarns is to directly twist electrospun nano-
fiber mats. Figure 7.22 shows a poly(vinylidene fluoride-co-hexafluoropropylene)
(PVDF-HFP) nanofiber yarn prepared by twisting a narrow fibrous strip cut di-
rectly from an electrospun nanofiber mat. Higher-quality nanofiber yarns can be
obtained by using a collector consisting of two conducting blades placed in line
(Figure 7.19D) or using a collector with two rings placed in parallel (Figure 17E).
Figure 7.21. SEM image of cross-bar array of aligned PVP nanofibers. (Li, D., et. al.,
Advanced Materials, 16, 361–366, 2004.)
Figure 7.22. SEM image of a PVDF-HFP nanofiber yarn. (Zhou, Y., et. al., Journal of
Materials Research, 27, 537–544, 2012.)
Nanofiber yarns described above have limited length. Continuous nanofiber

yarns can be obtained by depositing nanofibers into a water bath, followed by
drawing these nanofibers from the water in the form of a yarn using a rotating
drum (Figure 17F).
REFERENCES
[1] Bognitzki, M., Frese, T., Steinhart, M., Greiner, A., Wendorff, J.H., Schaper,
A., and Hellwig, M., “Preparation of Fibers with Nanoscaled Morphologies:
Electrospinning of Polymer Blends”, Polymer Engineering and Science,
41, 982–989, 2001.
References 111
[2] Dai, H., Gong, J., Kim, H., and Lee, D., “A Novel Method for Preparing
Ultra-Fine Alumina-Borate Oxide Fibres via an Electrospinning Technique”,
Nanotechnology, 13, 674–677, 2002.
[3] Garg, K., Sell, S.A., Bowlin, G.L., “Electrospinning and Its Influence on the
Structure of Polymeric Nanofibers”, in Handbook of Textile Fibre Structure.
Volume 1: Fundamentals and Manufactured Polymer Fibres, Editors
Eichhoorn, S.J., Hearle, J.W.S., Jaffe, M., and Kikutani, T., Woodhead
[4] Haiyan Wu, H., Hu, Q., Zhang, L., Fong, H., and Tian, M., “Electrospun
Composite Nanofibers of Polybutadiene Rubber Containing Uniformly
Distributed Ag Nanoparticles”, Materials Letters, 84, 5–8, 2012.
[5] Huang, C., Chen, S., Lai, C., Reneker, D.H., Qiu, H., Ye, Y., and Hou, H.,
“Electrospun Polymer Nanofibers with Small Diameters”, Nanotechnology,
17, 1558–1563, 2006.
[6] Ji, L., Lin, Z., Medford, A.J., and Zhang, X., “Porous Carbon Nanofibers
from Electrospun Polyacrylonitrile/SiO2 Composites as an Energy Storage
Material”, Carbon, 47, 3346–3354, 2009.
[7] Ji, L., Medford, A.J., and Zhang, X., “Fabrication of Carbon Fibers with
Nanoporous Morphologies from Electrospun Polyacrylonitrile/Poly(L-
lactide) Blends”, Journal of Polymer Science: Part B: Polymer Physics, 47,
493–503, 2009.
[8] Ji, L., Medford, A.L., and Zhang, X., “Porous Carbon Nanofibers
Loaded with Manganese Oxide Particles: Formation Mechanism And
Electrochemical Performance as Energy-Storage Materials”, Journal of
Materials Chemistry, 19, 5593–5601, 2009.
[9] Ji, L., Saquing, C., Khan, S.A., and Zhang, X., “Preparation and
Characterization of Silica Nanoparticulate-Polyacrylonitrile Composite and
Porous Nanofibers”, Nanotechnology, 19, 085605, 2008.
[10] Ji, L., and Zhang, X., “Electrospun Carbon Nanofibers Containing Silicon
Particles as an Energy-Storage Medium”, Carbon, 47, 3219–3226, 2009.
[11] Katta, P., Alessandro, M.R., Ramsier, R.D., and Chase, G.G., “Continuous
Electrospinning of Aligned Polymer Nanofibers onto Wire Drum Collector”,
Nano Letters, 4, 2215–2218, 2004.
[12] Li, D., Wang, Y., and Xia, Y., “Electrospinning Nanofibers as Uniaxially
Aligned Arrays and Layer-by-Layer Stacked Films”, Advanced Materials,
16, 361–366, 2004.
[13] Li, D., Wang, Y., and Xia, Y., “Electrospinning of Polymeric and Ceramic
Nanofibers as Uniaxially Aligned Arrays”, Nano Letters, 3, 1167–1171,
2003.
[14] Li, D., and Xia, Y., “Electrospinning of Nanofibers: Reinventing the Wheel”,
Advanced Materials, 16, 1151–1170, 2004.
[15] Lin, Z., Ji, L., Toprakci, O., Krause, W., and Zhang, X., “Electrospun Carbon
Nanofiber-Supported Pt–Pd Alloy Composites for Oxygen Reduction”,
Journal of Materials Research, 25, 1329–1335, 2010.
[16] Matthews, J.A., Wnek, G.E., Simpson, D.G., and Bowlin, G.L.,
“Electrospinning of Collagen Nanofibers”, Biomacromolecules, 3, 232–
238, 2002.
[17] McCann, J.T., Li, D., and Xia, Y., “Electrospinning of Nanofibers with
Core-Sheath, Hollow, or Porous Structures”, Journal Materials Chemistry,
15, 735–738, 2005.
[18] Norris, I.D., Shaker, M.M., Ko, F.K., and MacDiarmid, A.G., “Electrostatic
Fabrication of Ultrafine Conducting Fibers: Polyaniline/Polyethylene Oxide
Blends”, Synthetic Metals, 114, 109–114, 2004.
[19] Ramakrishna, S., Fujihara, K., Teo, T.C., and Ma, Z., An Introduction to
Electrospinning and Nanofibers, World Scientific Publishing, 2005.
[20] Shi, Q., Vitchuli, N., Nowak, J., Caldwell, J.M., Breidt, F., Bourham, M.,
Zhang, and McCord, M., “Durable Antibacterial Ag/Polyacrylonitrile (Ag/
PAN) Hybrid Nanofibers Prepared by Atmospheric Plasma Treatment and
Electrospinning”, European Polymer Journal, 47, 1402–1409, 2011.
[21] Sundaray, B., Subramanian, V., Natarajan, T.S., Xiang, R.Z., Chang, C.C.,
and Farm, W.S., “Electrospinning of Continuous Aligned Polymer Fibers”,
Applied Physics Letters, 84, 1222–1224, 2004.
[22] Teo, W.E., Kotaki, M., Mo, X.M., and Ramakrishna, S., “Porous
Tubular Structures with Controlled Fibre Orientation Using a Modified
Electrospinning Method”, Nanotechnology, 16, 918–924, 2005.
[23] Teo, W.E., and Ramakrishna, S., “A Review on Electrospinning Design and
Nanofibre Assemblies”, Nanotechnology, 17, R89–R106, 2006.
[24] Teo, W.E., and Ramakrishna, S., “Electrospun Fibre Bundle Made of
Aligned Nanofibres over Two Fixed Points”, Nanotechnology, 16, 1878–
1884, 2005.
[25] Toprakci, O., Toprakci, H.A.K., Ji, L., Lin, Z., Xu, G., and Zhang, X.,
“Carbon Nanotube-Loaded Electrospun LiFePO4/Carbon Composite
Nanofibers as Stable and Binder-Free Cathodes for Rechargeable Lithium-
Ion Batteries”, ACS Applied Materials & Interfaces, 4, 1273–1280, 2012.
[26] Wendorff, J.H., Agarwal, S., and Greiner, A., Electrospinning: Materials,
Processing, and Applications, Wiley-VCH, 2012.
Problems 113
[27] Zhao, Y., Cao, X., and Jiang, L., “Bio-mimic Multichannel Microtubes by
a Facile Method”, Journal of American Chemical Society, 129, 764–765,
2007.
[28] Zhang, S., Lu, Y., Xu, G., Li, Y., and Zhang, X., “LiF/Fe/C Nanofibers as
High-Capacity Cathode Material for Li-Ion Batteries”, Journal of Physics
D: Applied Physics, 45, 395301, 2012.
[29] Zhang, X., “Processing-Structure relationships of Electrospun Nanofibers”,
in Nanofibers: Fabrication, Performance, and Applications, Editor Chang,
W.N., Nova Science Publishers, 2009.
[30] Zhou, Y., Fang, J., Wang, X., and Lin, T. “Strip Twisted Electrospun
Nanofiber Yarns: Structural Effects on Tensile Properties”, Journal of
Materials Research, 27, 537–544, 2012.
PROBLEMS
(1) What are the main components of a basic electrospinning set-up?
(2) Describe the electrospinning process.
(3) How carbon and ceramic nanofibers can be made by electrospinning?
(4) What are the main approaches for making electrospun composite nanofibers?
(5) How can hollow and core-sheath nanofibers be made by electrospinning?
(6) What are the possible assemblies of electrospun nanofibers?
PART II
Fiber Formation
CHAPTER 8
Flow Behavior of Polymers
Synthetic polymer fibers and manufactured natural polymer fibers typically are
produced by melt, solution, dispersion, or gel spinning. To convert the polymer
components into the desired fiber structure, it is important to understand their flow
behavior. Knowledge in polymer flow behavior can lead to: (i) the intelligent se-
lection of suitable polymers or polymer compounds that can be made into fibers;
(ii) the establishment of qualitative, and to some extent quantitative, relationships
between processing conditions, fiber structures, and fiber properties; and (iii) the
minimization or removal of the processing faults and defects that are of rheologi-
cal origin. This chapter addresses the basic knowledge related to the flow behav-
ior of polymers. The use of such knowledge in the formation of synthetic polymer
fibers and manufactured natural polymer fibers will be discussed in Chapters 9
and 10, respectively.
8.1 BASIC DEFINITIONS

Figure 8.1 shows the flow profile of a fluid (e.g., polymer melt or solution) be-
tween two large parallel plates. The bottom plate is stationary and the top plate is
movable. An external force is applied to the top plate and moves it at a constant
velocity. The polymer fluid in contact with each plate has the same velocity as that
plate. The velocities of intermediate layers of the fluid increase uniformly from
one plate to another, as shown by the arrows. As a result, the fluid layers slide
smoothly over one another and the flow is laminar.
In the laminar flow as shown in Figure 8.1, the shear stress (τ) is the applied
force (F) divided by the area (A) of each plate:
F
τ= (8.1)
A
117
118 Flow Behavior of Polymers
Figure 8.1. Simple shear flow of a polymer fluid.
Shear rate ( γ ), i.e., the rate of change of shear strain, is given by:
dγ
γ = (8.2)
dt
where γ is the shear strain, i.e., dx/dy.
Assuming the displacement of the top plate in time dt is dx, the velocity (u) of
the top plate is:
dx
u= (8.3)
dt
The velocity gradient in the y direction, i.e., from the bottom plate to the top plate
is:
du u
velocity � gradient = = (8.4)
dy l
Equation 8.2 then can be rearranged to:
d d  dx  d dx du u
γ = ( γ ) =   =   = = (8.5)
dt dt  dy  dy  dt  dy l
Therefore, Equations 8.2 and 8.4 are equivalent expressions and velocity gradient
is an alternative definition of shear rate.
The viscosity (η) of the fluid can then be defined as:
Shear Stress F / A τ
η≡ = = (8.6)
Shear Rate u/l γ
Newtonian and Non-Newtonian Fluids 119
Viscosity is the fluid’s resistance to flow and is a measure of the frictional forces
between the molecules in the fluid. Hence, viscosity is an important parameter for
studying the flow behavior of polymers. According to Equation 8.6, the unit of
viscosity is that of force times distance, divided by area times speed. The SI unit
is 1 N·m/[m2·(m/s)] = 1 N·s/m2 = 1 Pa·s. The cgs unit for viscosity is dyn·s/cm2.
Another commonly used unit is poise, in honor of the French scientist Jean Louise
Marie Poiseuille: 1 poise = 1 dyn·s/cm2 = 10–1 N·s/m2.
8.2 NEWTONIAN AND NON-NEWTONIAN FLUIDS

8.2.1 Newtonian Fluid
The ideal Newtonian fluid is a fluid in which the shear stress-shear rate curve is
linear and passes through the origin. Figure 8.2A shows the shear stress-shear rate
curve of a Newtonian fluid. The slope of the curve is the viscosity of the fluid and
it keeps constant when the shear rate changes (Figure 8.2B). Newtonian fluids
typically are formed by small molecules, where the structure and orientation do
not change with the intensity of shearing. Examples of Newtonian fluids include
water, gasoline, motor oils, and many organic solvents.
8.2.2 Non-Newtonian Fluids

Many fluids, including polymer melts and solutions, are non-Newtonian. The
three most important non-Newtonian fluids are Bingham, shear-thickening, and
shear-thinning fluids. Figure 8.3 compares the shear stress-shear rate curves of
these three non-Newtonian fluids with that of an ideal Newtonian fluid.
The Bingham fluid has a linear shear stress-shear strain relationship, but a criti-
cal shear stress is required to initiate the flow. The shear stress-shear rate curve
of a Bingham fluid does not pass through the origin, and it can be expressed as:
Figure 8.2. (A) Shear stress-shear rate curve, and (B) viscosity-shear rate curve of a
Newtonian fluid.
Figure 8.3. Shear stress-shear rate curves of Newtonian and non-Newtonian fluids.
τ = τ critical + ηγ (8.7)
where τcritical is the critical stress required to initiate the flow. The Bingham fluid
typically has an internal structure that collapses above the critical shear stress.
Examples of Bingham fluids include suspensions, slurries, and pulps.
The shear-thickening fluid, also called dilatant fluid, shows an increase in its
resistance to flow with increasing shear rate. The most common shear-thickening
fluids are concentrated suspensions. In a concentrated suspension, as long as there
are balanced inter-particle interactions, such as van der Waals forces, the par-
ticles remain suspended. However, when the shear rate increases, particles enter
a state of flocculation and can be no longer held in suspension. The suspension
has the tendency to behave like a solid, and this causes the viscosity to increase.
However, when the shear stress is removed, the particles spread apart and once
again form a stable suspension. Some polymer pastes also have shear-thickening
behavior at certain shear rate ranges. In addition, polymers that crystallize under
stress may exhibit shear-thickening behavior.
Most polymer melts and solutions are shear-thinning fluids, which also are
called pseudoplastic fluids. The resistance of a shear-thinning fluid decreases with
increasing shear rate. The shear-thinning behavior typically is related to the “dis-
entanglement” of polymer chains with increasing shear rate (Figure 8.4). In poly-
mer fluids, the long and one-dimensional polymer chains often are extensively
entangled. Under shear, polymer chains start to disentangle and this reduces the
fluid’s resistance to flow, causing the shear-thinning behavior. In addition to the
disentanglement effect, the orientation of polymer chains under shear also plays a
role in the shear-thinning behavior.
Newtonian and Non-Newtonian Fluids 121
Theoretically, in shear-thickening and shear-thinning flows, there is not a char-

acteristic “viscosity” since η = τ/ γ is not a constant. Therefore, the term “apparent
viscosity” (ηa) is introduced and it is defined as the shear stress divided by shear
rate:
τ
ηa ≡ (8.8)
γ
This apparent viscosity increases with shear rate for shear-thickening fluids, is
independent of shear rate for Newtonian fluids, and decreases with shear rate for
shear-thinning fluids (Figure 8.5). Therefore, while reporting the apparent vis-
cosities of polymer melts or solutions, it is important to mention the shear rates
or shear stresses used during the measurements. In this book, for the purpose of
simplicity, “apparent viscosity” will be called as “viscosity” and be given the
symbol η.
To mathematically describe the flow behaviors of shear-thickening and shear-
thinning fluids, a power law equation can be used:
Figure 8.4. “Disentanglement” of polymer chains under stress.
Figure 8.5. Apparent viscosity-shear rate curves of Newtonian, shear-thickening, and

shear-thinning fluids.
τ = K ( γ ) (8.9)
n

where K is the fluid consistency index, and n the power law index. In logarithmic
form, Equation 8.9 can be written as:
log τ = log K + n log γ (8.10)
This means that the logarithmic shear stress-shear strain curve produced by
Equation 8.10 is a straight line. Figure 8.6 compares the log τ – log γ curves of
Newtonian, shear-thickening, and shear-thinning fluids. The curves of all three
fluids are straight lines, but with different slopes. The curve slope is determined
by the n value of the fluid. The fluid exhibits: Newtonian behavior when n = 1,
shear-thickening behavior when n > 1, and shear-thinning behavior when n < 1.
Based on the definition of viscosity, the power law equation can be rearranged
to obtain:
η = K ( γ )
n −1
(8.11)
Equation 8.11 also can be rewritten in the logarithmic form:
log η = log K + ( n − 1) log γ (8.12)
According to Equation 8.12, the log η – log γ curves of Newtonian, shear-thick-

ening, and shear-thinning fluids also are straight lines and the slopes are deter-
mined by the n-1 values.
Figure 8.6. log τ – log γ curves of Newtonian, shear-thickening, and shear-thinning fluids.

Time-Dependent Fluids 123
8.3 TIME-DEPENDENT FLUIDS

The Newtonian and non-Newtonian fluids discussed in section 8.2 are time inde-
pendent; that is, the viscosity remains constant as long as the shear rate does not
change. However, some fluids exhibit time-dependent flow behavior and their
viscosities change with the time of shearing. Two most important time-dependent
fluids are thixotropic and rheopectic. At a fixed shear rate, the viscosity of a thixo-
tropic fluid decreases with time, while the viscosity of a rheopectic fluid increases
with time (Figure 8.7).
The decrease of viscosity of the thixotropic fluid indicates a progressive break-
down of its structure. One classic example of thixotropic fluid is the ketchup,
which splashes after a period of vigorous tapping. Another example is the syno-
vial fluid found in joints between some bones. Thixotropic fluid often is com-
pared with the shear-thinning fluid. Figure 8.8 shows the shear stress-shear rate
curves of thixotropic and shear-thinning fluids under continuously increasing and
then decreasing shear rate. For the thixotropic fluid, in the time interval between
making the measurement at decreasing shear rate and making measurement at in-
creasing shear rate, the viscosity already has decreased since it is time dependent,
and this leads to the formation of the characteristic hysteresis curve, as shown
in Figure 8.8A. However, for the shear-thinning fluid, although the viscosity
changes with shear rate, it does not change with time. As a result, the curves
with increasing shear rate and with decreasing shear rate overlap with each other
(Figure 8.8B).
The rheopectic fluid also is called an anti-thixotropic fluid. In a rheopectic
fluid, the increase in viscosity with time often is caused by the formation of
Figure 8.7. Evolution of viscosity with time for time-independent, thixotropic, and

rheopectic fluids.
intermolecular forces. Wet concrete, in which water molecules react with the ce-
ment and bond the other components together, is a good example of rheopectic
fluid. The comparison between rheopectic and shear-thickening fluids is similar to
that between thixotropic and shear-thinning fluids. Figure 8.9 compares the shear
stress-shear rate curves of thixotropic and shear-thickening fluids. The rheopectic
fluid shows the characteristic hysteresis, with the shear rate decreasing curve lo-
cated above the shear rate increasing curve. However, the shear-thickening fluid
is time-independent and the two curves overlap (Figure 8.9B).
Polymer melts and solutions may present thixotropic or rheopectic behavior
if they degrade or crosslink during shear, especially at elevated temperatures.
Polymers that do not experience reactions during shear traditionally are con-
sidered as time-independent fluids. However, recent research shows the disen-
tanglement of polymer chains during shear may cause time-dependent effects.
Compared with most traditional thixotropic and rheopectic fluids, the time-de-
pendent effects caused by polymer chain disentanglement are relatively small.
Hence, in this book, all polymer melts and solutions are treated as time-indepen-
dent shear-thinning fluids.
Figure 8.8. Shear stress-shear rate curves for (A) a thixotropic fluid, and (B) a shear-
thinning fluid under continuously increasing and then decreasing shear rate.
Figure 8.9. Shear stress-shear rate curves for (A) a rheopectic fluid, and (B) a shear-thick-
ening fluid under continuously increasing and then decreasing shear rate.
Flow Behavior of Polymers 125
8.4 FLOW BEHAVIOR OF POLYMERS

8.4.1 Flow Curves of Polymers
As discussed above, polymer melts and solutions can be treated as shear-thinning
fluids. However, this is true only over a certain shear rate range. Figure 8.10 shows
the typical logarithmic flow curves of polymer melts and solutions over a very
wide range of shearing. Three different regions can be found in the flow curves:
1. A “lower Newtonian” region exists at low shear rates or stresses, and this
leads to the definition of zero-shear viscosity (η0), i.e., value of η as γ  0.
2. The shear-thinning region is observed over several decades of intermediate
shear rates.
3. At very high shear rates, an “upper Newtonian” region is attained and the
viscosity is defined as η∞.
The presence of three regions is mainly due to the disentanglement effect of

polymer chains. At low shear rates, the polymer chains are extensively entangled
and have the greatest resistance to flow. The shear stress at this region is not suf-
ficient to disentangle the polymer chains. As a result, the viscosity is independent
on the shear rate or stress and the polymer fluid is at its “lower Newtonian” re-
gion. At that region, the polymer fluid has its highest viscosity, i.e., η0, which can
be used as an important characteristic parameter for describing the polymer fluid.
As the shear rate increases to a high enough value, the shear stress starts to dis-
entangle and orient the polymer chains, and higher shear stress or strain leads to
greater disentanglement effect. As a result, the polymer fluid enters the shear-thin-
ning region. However, at very high shear rates, the polymer chains are completely
disentangled and oriented, and the viscosity becomes constant again, indicating
the polymer fluid enters an “upper Newtonian” region.
For the formation of synthetic polymer fibers and manufactured natural poly-
mer fibers, it is important to disentangle and orient polymer chains to achieve
good mechanical properties. Therefore, the formation of these polymer fibers is
not carried out at the “lower Newtonian” region. On the other hand, the “up-
per Newtonian” region also is not desired. The intense shearing at the “upper
Newtonian” region could lead to extensive breakage of polymer chains, leading
to mechanical degradation. Therefore, polymer fibers are processed in the shear-
thinning region. In the remaining sections of this chapter, only the shear-thinning
region will be discussed.
8.4.2 Carreau Equation

Many equations have been proposed to quantitatively represent the flow curves of
polymer melts and solutions. Among them, the power law equation discussed in
section 8.2.2 is probably the simplest and is adequate for many engineering pur-
poses. However, the power law can reasonably approximate only portions of the
actual flow curves over a relatively narrow shear rate range. To represent the flow
curves over a wider shear rate range, the Carreau equation often is used:
n −1
η − η∞ 
= 1 + ( λc γ ) 
2
2
(8.13)
η0 − η∞  
where λc is the time constant, and n the power law index. η0 and η∞ are the viscos-
ity at zero shear rate and the viscosity at infinite shear rate, respectively, as shown
in Figure 8.10. The Carreau equation has four parameters (λc, n, η0, and η∞) and
can quantitatively represent the flow curves of polymer over a wide range of shear
rate.
In practical application, to represent the polymer flow curves in a relatively
narrow shear rate range (e.g., one or two decades of shear rate), the power law
equation has sufficient accuracy. To represent the polymer flow curves in a wider
shear rate range (e.g., three or four decades), the Carreau equation can be used.
In addition to these two equations, other models also have been developed. The
interested reader is referred to the works by Tanner (2000), Dealy and Larson
(2006), and Brazel and Rosen (2012).
Figure 8.10. Typical log τ – log γ and log η – log γ curves of polymer melts and solutions
over a very wide range of shearing rates.
Capillary Flow of Polymers 127
8.5 CAPILLARY FLOW OF POLYMERS

Synthetic polymer fibers and manufactured natural polymer fibers typically are
produced by melt, solution, dispersion, or gel spinning. In these processes, poly-
mer fluids are extruded through dies to form one-dimensional fiber structures
(Figure 8.11). The dies typically have small diameters so thin fibers can be ob-
tained. As a result, in order to form fibers with desired structure and properties, it
is critically important to understand the capillary flow of polymers, i.e., the flow
behavior in a small-diameter pipe.
To quantitatively describe the flow of polymer fluids through a small-diameter
pipe, the following assumptions are necessary:
1. No-slip at the wall, i.e., the velocity of the polymer fluid at the wall of the
pipe is zero.
2. The flow pattern is constant along the pipe.
3. The polymer fluid is time-independent.
4. The polymer fluid is incompressible.
5. The polymer flow is isothermal; and
6. The force of gravity is negligible.
8.5.1 Shear Stress at Wall

The shear stress at the wall of the pipe can be obtained by considering the balance
of forces. Figure 8.12 shows the balance of forces on an element of polymer fluid
Figure 8.11. Polymer flow through a die.

Figure 8.12. Balance of forces on a fluid element in a pipe with length L and radius R.
in a pipe with length L and radius R. The length and radius of the fluid element are
dz and r, respectively. The total pressure drop along the pipe is ΔP and the pres-
sure drop along the fluid element is dP.
If the polymer flow is steady, the sum of all the forces acting on the fluid ele-
ment is zero:
∑F = F + F1 2 + F3 = 0 (8.14)
where F1 and F2 are hydrostatic forces, and F3 is the drag on the surface of the
element. F1, F2 and F3 can be expressed by:
F1 = Ppr2
F2 = –[P + dP]pr2
F3 = (surface area) × (drag/unit surface)
= surface area × shear stress
= 2prdzt
As a result, Equation 8.14 can be rewritten as:
∑F = P πr − [ P + dP ] πr
2 2
+ 2 πrdz τ = 0 (8.15)
The shear stress on the fluid element then can be obtained:
r  dP 
τ=   (8.16)
2  dz 
Since the flow pattern is constant along the pipe length, dP/dz is constant and
Equation 8.16 can be rewritten as:
r ∆P
τ= (8.17)
2L
From Equation 8.17, it can be concluded the shear stress τ has a linear relationship
with the element radius r, and it increases from zero at the pipe center linearly
with r to a maximum (τw) at the wall. The shear stress at the wall is then given by:
R ∆P
τw = � (8.18)
2L
Therefore, when a polymer fluid flows through a pipe with known dimensions
(i.e., L and R), the shear stress at the pipe wall can be easily obtained from the
pressure drop between the ends of the pipe. In addition, the shear stress at an ar-
bitrary point in the pipe can be obtained by:
r
τ= τw (8.19)
R
8.5.2 Shear Rate at Wall

The shear rate at the wall of the pipe is related to the volumetric flow rate Q,
which can be measured experimentally. The volumetric flow rate also can be ob-
tained by integrating the velocity profile over the cross-sectional area of the pipe.
Figure 8.13 shows the velocity profile of a polymer fluid in a pipe with radius R.
The velocity of the fluid at distance r from the pipe center is uz. The velocity at the
wall (i.e., r = R) is zero since there is no slip at the wall. Shear rate is the same as
the velocity gradient, and hence the shear rate at any point in the pipe is defined
as -dur/dr.
As shown in Figure 8.13, the volumetric flow rate can be calculated from the
volume under the curve:
R R
Q = ∫ 2 πru z dr = 2 π ∫ru z dr (8.20)
0 0
where 2πru z dr is the volume of a fluid shell with thickness dr, height uz, and a
radius of r.
Equation 8.20 can be integrated by parts by letting g = uz and dh = rdr in the
equation of by parts:
Figure 8.13. Velocity profile of a polymer fluid in a pipe with radius R.
∫g � dh = hg − ∫� h dg (8.21)
Hence, dg = duz and h = r2/2. Equation 8.20 can then be rewritten as:
  r 
R R R
2

Q = 2 π ∫u z rdr = 2 π   u z  − ∫ ( r 2 / 2 ) du z  (8.22)
2 0 0
0   
Since uz = 0 when r = R, the first term in Equation 8.22 is zero. As a result,
R R
du z
Q = − π ∫r 2 du z = − π ∫r 2 dr (8.23)
0 0
dr
As discussed in the section 8.5.1, the shear stress is also related to r, which is the
distance from the pipe center. From Equation 8.19, the following equation can be
obtained:
R
r= τ (8.24)
τw
Substituting Equation 8.24 and γ = -dur/dr into Equation 8.23 leads to:
τ
πR 3 w 2
τ 3w ∫0
Q=  dτ
γτ (8.25)
which can be rearranged to:

τ
τ 3w Q w 2
πR 3 ∫0
 dτ
= γτ (8.26)
By differentiating both sides with respect to τw, the following equation can be
obtained:
1  3 dQ 
 τw + 3τ 2w Q  = γ w τ w 2 (8.27)
πR 3  d τ w 
Substituting τw = RΔP/2L into Equation 8.27 gives the Rabinowitsch equation:

1  dQ  (8.28)
γ w =
πR 3 3Q + ∆P d ∆P 
The Rabinowitsch equation allows the calculation of the shear rate at the pipe
wall from three measurable quantities: R, Q, and ΔP. Since the flow properties of
a fluid are independent of the pipe structure, the τw- γ w curve of the fluid is the
same as the τ- γ curve. Therefore, by using Equation 8.18 and the Rabinowitsch
equation, the flow curve, i.e., τ- γ or log τ-log γ , can be obtained. However, the
use of the Rabinowitsch equation sometimes is tedious, and hence it often is sim-
plified by using the following mathematical rearrangements.
Equation 8.27 can first be rearranged to:
3  4Q  τ w d ( 4Q / πR )
3
γ w =   + (8.29)
4  πR 3  4 d τw
d ( log x ) 1 dx d log x
Since = , y becomes x . As a result, Equation 8.29 can be
dx x dy d log y
rearranged to:
3  4Q  1  4Q  d log ( 4Q / πR )
3
γ w =  +   (8.30)
4  πR 3  4  πR 3  d log τ w
To simplify Equation 8.30, the derivative can be defined as 1/ n ' , i.e.,
d log τ w d log ( R ∆P / 2 L )
n' = = (8.31)
(
d log 4Q / πR 3
) (
d log 4Q / πR 3 )
Equation 8.30 can then be rewritten as:
 3n ' + 1  4Q
γ w =  '  3 (8.32)
 4n  πR
This simple equation is known as the Metzner form of the Rabinowitsch equation.
Now, the flow curve (τ- γ or log τ-log γ ) of the fluid can be obtained by using
Equations 8.18 and 8.32.
From Equation 8.31, the following relationship also can be obtained:
 R ∆P   4R 
log  = n ' log  + log K ' (8.33)
 2 L   πR 3 
According to Equation 8.33, the shear stress at the wall can be calculated by:
n'
R ∆P  4Q 
τw = = K'  (8.34)
2L  πR 3 
If the polymer fluid follows the power law relationship, i.e., τ w = K ( γ w ) ,
n
Equation 8.32 can be used to obtain:
n n
 3n ' + 1   4Q 
τw = K  '   3 
(8.35)
 4n   πR 
Comparing Equations 8.34 and 8.35, it is seen that K ' and n ' can be obtained
from the fluid consistency index K and the power law index n:
n
 3n + 1 
K' = K   (8.36)
 4n 
and
n' = n (8.37)
The Metzner form of the Rabinowitsch equation (Equation 8.32) can then be re-
written to:
 3n + 1 4Q (8.38)
γ w = 
 4n  πR 3
For Newtonian flow (n = 1), the term 4Q/πR3 is in fact the true shear rate at all.
However, for non-Newtonian flow, the term 4Q/πR3 is known as the apparent
shear rate at wall ( γ w, a ), i.e.,
4Q
γ w, a = (8.39)
πR 3
In many applications, the flow curves (τ- γ or log τ-log γ ) of the fluids are ob-
tained by using Equations 8.18 and 8.39, and Equation 8.38 is used as a correction
for non-Newtonian flows.
8.5.3 Calculation of Viscosity

The viscosity of the polymer fluid can be calculated from Equations 8.32 and
8.34:
−1 n' −1
τw  3n ' + 1   4Q 
η= = K'  '   3 
(8.40)
γ w  4n   πR 
For polymer fluids that follow the power law relationship, Equation 8.40 can be
rearranged to:
n −1 n −1 n −1
 3n + 1   4Q   3n + 1 
( γ )
n −1
= K ( γ w )
n −1
η= K   3 
= K  w, a (8.41)
 4n   πR   4n 
8.5.4 Corrections of Flow Curves

As discussed above, the flow curves of polymer fluids can be obtained by
Equations 8.18 and 8.38 (or 8.39), and the viscosities of the fluids can be calcu-
lated by Equation 8.41. While deriving these equations, one of the assumptions is
that the flow pattern is constant along the pipe. However, in a real capillary flow,
the polymer fluid exhibits different flow patterns in the entrance and exit regions
of the pipe. For example, the pressure drops at the die entrance and exit regions
are different from ΔP/L. Therefore, corrections, e.g., Bagley correction, are need-
ed to address the entrance and exit effects. Another assumption is that there is no
slip at the wall. However, in a real flow, polymer fluid may slip at the wall and this
reduces the shear rate near the wall. The Mooney analysis can be used to address
the effect of the wall slip. In addition, the velocity profile shown in Figure 8.13
is a parabolic flow. However, the true flow in the die orifice is not necessarily a
simple parabolic flow, and hence Weissenberg-Rabinowitsch correction often is
used to correct the shear rate at the wall for the non-parabolic velocity profile.
Detailed mathematical treatments for these corrections are not discussed here,
but they can be found in the references listed at the end of this chapter.
8.6 EXTERNAL FACTORS AFFECTING THE VISCOSITY OF

POLYMERS
The flow of a polymer occurs when the polymer chains slide against each other.
The viscosity of a polymer fluid is its resistance to flow and is a measure of the
frictional forces between the molecules in the fluid. The viscosity of polymer
fluids is affected by a number of external factors, such as temperature, shear rate,
and pressure.
8.6.1 Effect of Temperature

It is well-known that an increase in temperature can reduce the polymer viscos-
ity. The viscosity-temperature relationship of polymers often is described by the
Arrhenius equation:
η = Ae ∆E / RT (8.42)
where A is a structure constant, ΔE the activation energy, R the gas constant, and T
the absolute temperature of the polymer. The increase in temperature leads to in-
creased intermolecular distance and reduced intermolecular interaction, which—
in turn—results in improved polymer chain mobility. As a result, the viscosity of
the polymer fluid decreases with increase in temperature.
The Arrhenius equation can be rearranged to:
∆E 1
ln η = ln A + (8.43)
R T
This indicates that the ln η-1/T plot is a straight line and the slope is ΔE/R
(Figure 8.14). Since the value of the gas constant is 8.314 J·mol–1 K–1, the activa-
tion energy can be calculated directly from the slope of the ln η-1/T plot. When
a polymer has a higher activation energy, its viscosity is more sensitive to the
change in temperature.
The activation energy of polymers is closely related to their chemical struc-
tures. Table 8.1 shows the activation energies of some polymer melts. In general,
flexible polymer chains with weak intermolecular interactions (e.g., polyethyl-
ene and polypropylene) have low activation energies, while polymers with rigid
chains or strong intermolecular interactions (e.g., acetate and polyacrylonitrile)
have high activation energies. Figure 8.15 shows the effect of polymer chemical
structure on the viscosity-temperature relationship. It is seen that the viscosity
of polymers with rigid chains or strong intermolecular interactions is sensitive
to the change in temperature. However, the viscosity of flexible polymer chains
with weak intermolecular interactions is less sensitive to temperature. This pro-
vides important guidance for the control of polymer spinning processes. To spin
External Factors Affecting the Viscosity of Polymers 135
Figure 8.14. A typical ln η-1/T plot.
fibers from polymers with rigid chains or strong intermolecular interactions (e.g.,
acetate and polyacrylonitrile), it is effective to reduce their viscosities by increas-
ing the processing temperature. On the other hand, to spin fibers from polymers
with flexible chains and weak intermolecular interactions (e.g., polyethylene and
Table 8.1. Activation energies of polymer melts.
Polymer Activation Energy

(kcal/mole)
Polyethylene, linear 6.3–8.2
Polypropylene 9.0–10.1
Nylon 6 17.2–18.6
Nylon 6,6 16.2
Polyester 13.1–20.1
Polycarbonate 21.9
Polystyrene 25.5
Polyacrylonitrile 36.5–41.0
Acetate 34.2
Sources: Porter, R.S., et. al., Journal of Polymer Science, Part C. Polymer Symposia, 15,
365–371, 1967.; Saini, D.R., et al., Journal of Macromolecular Science, Part B. Physics, 22,
437–449, 1983.
Figure 8.15. Effect of chemical structure on the viscosity-temperature relationship of

polymers.
polypropylene), increasing temperature may not be sufficient to reduce their vis-

cosities and other approaches are needed.
For fiber spinning processes, another commonly used equation for the viscosi-
ty-temperature relationship is:
η = ae − bT (8.44)
where both a and b are constants. The constant b indicates the temperature sensi-
tivity of the polymer and is typically in the range of 0.01–0.1°C–1.
The validity of Equations 8.43 and 44 is related to the glass transition tempera-
ture (Tg) of polymers. For many polymers, these two equations only hold at T > Tg
+ 100°C. For temperatures ranging from Tg to Tg + 100°C, the Williams-Landel-
Ferry (WLF) equation can be used:
η −17.4 (T − Tg )
log = (8.45)
ηTg 51.6 + T − Tg
where ηTg is the viscosity at Tg.
8.6.2 Effect of Shear Rate

As discussed in previous sections, most polymer fluids exhibit shear-thinning be-
havior, i.e., the viscosity decreases with increase in shear rate. The shear-thinning
External Factors Affecting the Viscosity of Polymers 137
behavior of polymer fluids is related to the “disentanglement” phenomenon. With

increase in shear rate, more polymer chains are being disentangled and oriented,
leading to lower resistance to the flow, i.e., lower viscosity. The viscosity-shear
rate relationship can be described by using the power law equation or the Carreau
equation.
The disentanglement process is affected by the chemical structure of polymers.
Figure 8.16 shows the effect of polymer chemical structure on the viscosity-shear
rate relationship. For polymers with flexible chains and weak intermolecular in-
teractions (e.g., polyethylene and polypropylene), their conformations can be
changed easily and increasing shear rate is effective for disentangling the polymer
chains. As a result, their viscosities decrease rapidly with an increase in shear rate.
However, for polymers with rigid chains or strong intermolecular interactions
(e.g., acetate and polyacrylonitrile), it is more difficult to change their conforma-
tions by shear and their viscosities are less sensitive to the shear rate.
Based on the discussion in sections 8.6.1 and 8.6.2, the following guidelines
can be used if the polymer viscosity needs to be reduced during fiber formation:
1. For polymers with flexible chains and weak intermolecular interactions, it is

more effective to increase the shear rate.
2. For polymers with rigid chains or strong intermolecular interactions, it is
more effective to increase the temperature.
In real applications, many other factors can affect the actual viscosity of a
polymer. As a result, these guidelines may not work for some polymers. In addi-
tion, there are limitations while increasing the temperature or shear rate. For ex-
ample, the processing temperature cannot be too close to the polymer degradation
Figure 8.16. Effect of chemical structure on the viscosity-shear rate relationship of

polymers.
temperature since polymer chains tend to break at elevated temperatures. The

shear rate also is limited by the capacity of the equipment used for the fiber forma-
tion. Moreover, extremely high shear rate could cause instable flow and make the
fiber spinning process impossible.
8.6.3 Effect of Pressure

During fiber formation, polymer fluids are extruded through die orifices under
pressure. Figure 8.17 shows the effect of pressure on the viscosity of polymers. In
general, the increase in pressure leads to reduced intermolecular distance and in-
creased intermolecular interaction. As a result, the viscosity increases. The effect
of increasing pressure is similar to that of decreasing temperature. For example,
for many polymer melts, applying a pressure of 10 MPa is equivalent to decreas-
ing the temperature by around 5°C.
The relationship between viscosity and pressure (P) can be described by:
η = αe βP (8.46)
where α and β are constants. The constant β typically is in the order 2 × 10–8 Pa–1.
This indicates a pressure increase of 10 MPa could cause the viscosity to increase
by 22%.
Equations 8.44 and 46 can be combined to give the following equation that
includes both the temperature and pressure dependences of the viscosity:
η = De − bT + βP (8.47)
Figure 8.17. Effect of pressure on the viscosity of polymers.

Structural Factors Affecting the Viscosity of Polymers 139
where D is a constant. Equation 8.47 can be used in conjunction with the power

law equation or the Carreau equation to describe the viscosity as a function of T,
γ , and P.
8.7 STRUCTURAL FACTORS AFFECTING THE VISCOSITY

OF POLYMERS
Polymers with different structures show different flow behaviors. For fiber forma-
tion, the most important structural factors are chain rigidity, intermolecular inter-
action, molecular weight, and molecular weight distribution. Other factors, such
as crosslinking degree, side chain length and flexibility, also affect the viscosity
of polymers, but they are less important since fiber spinning processes require
polymers that are linear and are not crosslinked.
8.7.1 Effect of Chain Rigidity and Intermolecular Interaction

As discussed in the previous section, the polymer chain rigidity and intermolecu-
lar interaction affect the viscosity-temperature and viscosity-shear rate relation-
ships. In general, polymers with rigid chains have higher resistance to flow and
their viscosities often are greater than those of flexible chains. Similarly, polymers
with strong intermolecular interactions have higher viscosities than those with
weak interactions. In addition, the viscosities of polymers with rigid chains or
strong intermolecular interactions are sensitive to the change in temperature, but
are less sensitive to shear rate.
8.7.2 Effect of Molecular Weight

With increase in molecular weight, the viscosities of polymers increase since lon-
ger polymer chains have greater resistance to the flow. The actual viscosity-mo-
lecular weight relationship of polymers is complex. For most polymers, the vis-
cosity at zero shear rate (η0) has the following relationship with the weight average
molecular weight ( M w ):
η0 = K L M w1.0 , at M w < M c (8.48)
η0 = K H M w3.4 , at M w < M c (8.49)
where KL and KH are constants. M c is the critical molecular weight, at which the
chain entanglements begin to dominate the rate of slippage of the polymer chains.
When the weight average molecular weight is less than M c , the average polymer
chain length is smaller than the average distance between entanglements, and as a
result, the viscosity is determined mainly by the average size of the polymer
chains and is proportional to the molecular weight. However, when the weight
average molecular weight is greater than M c , there is a significant amount of
entanglements and the resistance to flow increases quickly. As a result, the viscos-
ity is proportional to M w3.4 . Figure 8.18 shows the effect of weight average mo-
lecular weight on the zero-shear viscosity of polymers.
For most polymers, the M c value ranges from 1,000 to 30,000 g/mol. The
weight average molecular weights of most polymer fibers are greater than their
corresponding M c values so they can be spun into the fiber form and have suffi-
cient mechanical properties. However, compared with most other polymer products,
polymer fibers typically have lower molecular weights. According to Equation 8.49,
lower molecular weight leads to significantly smaller viscosity, which allows the
polymer fluids to pass through the small die orifices to form fibers.
Equations 8.48 and 49 show the effect of molecular weight on the zero-shear
viscosity. The formation of polymer fibers is carried out at high shear rates.
Figure 8.19 shows the viscosity-molecular weight relationships at different shear
rates. It is seen that at M w < M c , the shear rate does not affect the viscosity-mo-
lecular weight relationship since there is no significant entanglements in the sys-
tem. However, at M w > M c , the slope of the η- M w curve decreases with increase
in shear rate since more entanglements are disentangled at high shear rates. At
very high shear rates, the viscosity could be proportional to the molecular weight.
Figure 8.18. Effect of weight average molecular weight on the zero-shear viscosity of

polymers.
Structural Factors Affecting the Viscosity of Polymers 141
8.7.3 Effect of Molecular Weight Distribution

At a fixed molecular weight, the viscosity of polymers changes with change in
molecular weight distribution. Figure 8.20 shows the effect of molecular weight
distribution on the η- γ curve of polymers. Typically, when the molecular weight
distribution is narrower, the polymer fluid is more Newtonian and the power law
index n is closer to unit. When the molecular weight distribution is broader, the
polymer fluid becomes more non-Newtonian and n decreases. This is because
polymers with broad molecule weight distributions have a larger number of ex-
tremely long chains. These long chains can form entanglements easily, leading to
Figure 8.19. Effect of shear rate on the viscosity-molecular weight relationship of polymers.
Figure 8.20. Effect of molecular weight distribution on the viscosity of polymers.

higher viscosities at low shear rates. However, at high shear rates, the entangle-
ments formed by the long chains can be disentangled and the viscosity decreases
rapidly with an increase in shear rate.
8.8 ELASTIC EFFECTS

As discussed above, polymer chains tend to be disentangled and oriented under
shear. One important result of the disentanglement and molecular orientation un-
der shear is the decrease of viscosity with increasing shear rate. However, the
flow of polymers is not pure viscous flow, and it has elastic component since the
change of chain conformations is not completely irreversible. Upon the release of
the shear, the polymer chains tend to recoil and be pulled back by the restraining
force. Such elastic response has significant effect on the fiber formation processes.
8.8.1 Die Swelling

One important elastic effect is the die swell behavior, i.e., the extrudate diameter
is greater than the die diameter. This is because that while passing through the die,
the polymer chains are being disentangled and oriented, producing normal stress.
Due to the limited die length, disentanglement and orientation of polymer chains
are not complete in the die. When exiting the die, the normal stress disappears and
the polymer chains coil. As result, the extrudate expands perpendicular to the flow
direction (Figure 8.21). The die swell ratio (B) is defined as:
Dmax
B= � (8.50)
D0
Figure 8.21. Die swell behavior of polymers.

Elastic Effects 143
where D0 is the die diameter, and Dmax the maximum diameter of the extrudate
(e.g., as-spun fiber before drawing). Under certain conditions, the die swell ratio
could reach 400% or even higher. The die swell ratio is closely related to the mo-
lecular weight of polymers. Typically, with an increase in molecular weight, the
amount of die swell increases.
The die swell ratio is affected by the design of the die. For example, when the
die diameter is fixed, the die swell ratio decreases with an increase in die length
(Figure 8.22). The die swell ratio also is affected by temperature and shear rate.
In general, the die swell ratio decreases with increase in temperature. However,
the die swell-shear rate relationship is more complex. Figure 8.23 shows a typical
die swell-shear rate curve of polymers. The die swell ratio increases with shear up
to a limit, beyond which the die swell decreases. It is important to notice that the
maximum die swell often occurs at a shear rate, which is just slightly lower than
the onset of melt fracture.
Figure 8.22. Effect of die length on the die swell.
Figure 8.23. Effect of shear rate on the die swell ratio of polymers.

8.8.2 Melt Fracture

As shown in Figure 8.23, when the shear rate exceeds a critical value, melt frac-
ture occurs and polymers no longer extrude smoothly. Melt fracture starts with
barely visible surface defects, but at higher shear rates the as-spun fibers can have
different morphologies. In some cases, as-spun fibers present “sharkskin”, in oth-
ers the fibers look like screwed threads. The as-spun fibers also can have rod-like
cross-sections, twisted into the form of spirals. These are just a few examples, and
many other types of morphologies can be observed when melt fracture occurs. At
very high shear rates, the melt fracture could lead to the loss of cohesion.
It is commonly agreed the melt fracture behavior is related to the elasticity of
polymer fluids. The formation of different morphologies may be due to some
secondary effects. For the formation of polymer fibers, melt fracture is not de-
sired, and hence the fiber spinning processes need to be carried out below the criti-
cal shear rate ( γ c ), i.e., the onset of melt fracture. Therefore, it is important to
determine the critical shear rate for polymers. In general, with increase in molecu-
lar weight, the critical shear rate decreases (Figure 8.24). This means the melt
fracture occurs more easily for longer polymer chains since they are more elastic.
In practical applications, the critical shear rate can be increased by increasing the
processing temperature (Figure 8.25).
8.9 ELONGATIONAL VISCOSITY

So far, the discussion in this chapter focuses on shear flow. However, the elon-
gational flow of polymers also is important for fiber formation. In a pure shear
flow, the velocity gradient is normal to the flow direction. On the other hand, the
Figure 8.24. Effect of molecular weight on the critical shear rate of polymers.

Elongational Viscosity 145
Figure 8.25. Effect of temperature on the critical shear rate of polymers.
velocity gradient in an elongational flow is in parallel with the flow direction.

When polymer fluid flows through the die, the velocity gradient is normal to the
flow direction and hence the flow is a shear flow. However, when the polymer
fluid is converging into the small-diameter die, velocity gradient also is devel-
oped in parallel with the flow direction, and hence the flow in the entry region is
partially an elongational flow. When the polymer fluid exits the die orifice, the
velocity gradient becomes in parallel with the flow direction again due to the ten-
sion force, and the shear flow changes to elongational flow until the polymer is
solidified. Therefore, it is important to study the elongational flow of polymers.
The elongational viscosity (ηe) is the resistance of the fluid to elongational
flow. The elongational viscosity can be defined as:
du
ηe = σ / (8.51)
dz
where σ is the tensile stress, and du/dz is velocity gradient in the flow direction.
Trouton studied the relationship between the elongational viscosity and shear vis-
cosity (η), and found the elongation to shear viscosity ratio is equal to 3:
ηe
= 3 (8.52)
η
This relationship, known as the Trouton ratio, is valid at low stresses or rates.
Figure 8.26 compares the elongational viscosity-stretch rate ( ε ) relationship to
the shear viscosity-shear rate relationship over a wide range of strain rates. At low
rates, the elongational viscosity is three times of the shear viscosity. Shear
Figure 8.26. Elongational viscosity-stretch rate and shear viscosity-shear rate relationships

of polymers.
viscosity decreases with increase in shear rate, i.e., shear-thinning behavior.

However, the elongational viscosity-stretch rate curve has a strain hardening re-
gion, in which the elongational viscosity increases with stretch rate. After the
strain hardening region, the elongational viscosity decreases again with increase
in stretch rate.
When the elongational viscosity starts to decrease after the strain hardening
region, the elongational viscosity-stretch rate relationship can be described by:
ηe , 0
ηe = (8.53)
1 + ( a ηe , 0 ε )
b
where ηe,0 is the zero stretch rate viscosity, and a and b are empirical constants.
The elongational viscosity also is a function of temperature. At T > Tg + 100°C,
the elongational viscosity-temperature relationship follows the Arrhenius equa-
tion. When the temperature is between Tg and Tg + 100°C, the temperature depen-
dence of elongational viscosity can be described by the WLF equation.
When the polymer fluid exits the die orifice during fiber formation, the crystal-
line structure starts to be developed. The elongational viscosity increases rapidly
as the crystallization process starts. When the degree of crystallinity still is low,
the effect of crystallinity on the elongational viscosity is given by:
ηe = Hexp ( cθ d ) (8.54)
where θ is the relative crystallinity, and H, c and d are empirical constants.

When the degree of crystallinity becomes higher, the small crystals act
as “physical crosslinks”. Ziabicki suggests when the number of such physical
References 147
crosslinks exceeds a critical value, the following equation can be used to describe
the temperature and crystallinity dependences of the elongational viscosity:
α
 θ 
ηe = f (T ) / 1 −  (8.55)
 θ cr 
where f(T) is a function of temperature, θcr the crystallinity when the number of
physical crosslinks exceed the critical value, and α a constant. Experimental data
show the θcr value could be very low, 2–3%. Therefore, even a relatively low
degree of crystallinity can drastically change the elongational viscosity and then
significantly affects the fiber formation process.
REFERENCES
[1] Baird, D.G., and Collias, D.I., Polymer Processing: Principles and Design,
John Wiley & Sons, 1998.
[2] Barnes, H.A., Hutton, J.F., and Walters, K., An Introduction to Rheology,
Elsevier, 1989.
[3] Brazel, C.S., and Rosen, S.L., Fundamental Principles of Polymeric
Materials, Third edition, John Wiley & Sons, 2012.
[4] Brydson, J.A., Flow Properties of Polymer Melts, Van Nostrand Reinhold
Company, 1981.
[5] Dealy, J.M., and Larson, R.G, Structure and Rheology of Molten Polymers,
Hanser Gardner Publications, 2006.
[6] Han, C.D., Rheology and Processing of Polymeric Materials: Volume 1:
Polymer Rheology, Oxford University Press, 2007.
[7] Han, C.D., Rheology and Processing of Polymeric Materials: Volume 1I:
Polymer Processing, Oxford University Press, 2007.
[8] Kontopoulou, M., Applied Polymer Rheology: Polymeric Fluids with
Industrial Applications, John Wiley & Sons, 2012.
[9] Park, H.M., Hong, S.M., and Lim, J.Y., “Estimation of Rheological
Parameters Using Velocity Measurements”, Chemical Engineering Science,
62, 6806–6815, 2007.
[10] Park, H.M., Shin, K.S. and Choi, Y.J., “Rheometry Using Velocity
Measurements”, Rheologica Acta, 48, 433–445, 2009.
[11] Porter, R.S., and Johnson, J.F., “Temperature Dependence of Polymer

Viscosity. The Influence of Shear Rate and Stress”, Journal of Polymer
Science, Part C. Polymer Symposia, 15, 365–371, 1967.
Introductory Text, Second Edition, CRC Press, 1997.
[13] Saini, D.R., and Shenoy, A.V., “A New Method for the Determination of
Flow Activation Energy of Polymer Melts”, Journal of Macromolecular
Science, Part B. Physics, 22, 437–449, 1983.
[14] Shaw, M.T., Introduction to Polymer Rheology, John Wiley & Sons, 2012.
[15] Tanner, R.I., Engineering Rheology, Second Edition, Oxford University
Press, 2000.
[16] Zhang, X., Pan, Y., Cheng, Jianfeng, and Yi, X., “The Influence of Low-
Melting-Point Alloy on the Rheological Properties of a Polystyrene Melt”,
J. Mater. Sci., 35, 4573–4581, 2000.
[17] Ziabicki, A., “The Mechanisms of ‘Neck-Like’ Deformation in High-Speed
Melt Spinning. 1. Rheological and Dynamic Factors”, Journal of Non-
Newtonian Fluid Mechanics, 30, 141–155, 1988.
[18] Ziabicki, A., “The Mechanisms of Neck-Like Deformation in High-Speed
Melt Spinning .2. Effects of Polymer Crystallization”, Journal of Non-
PROBLEMS
(1) What are Newtonian, Bingham, shear-thinning, and shear-thickening flows?
(2) Why do polymers often exhibit shear-thinning behavior?
(3) What are rheopectic and thixotropic fluids?
(4) A typical shear stress-shear rate curve for a polymer can be divided into
lower Newtonian, shear-thinning, and upper Newtonian regions if the shear
rate range is very wide. Explain how the polymer chains behave under shear
in these three regions.
(5) What are the shear stress and shear rate at wall when a polymer flows
through a small-diameter pipe.
(6) Discuss and explain how you would expect the viscosity of a polypropylene
melt to vary with:
(i) Molecular weight
(ii) Chain rigidity
Problems 149
(iii) Intermolecular interaction

(iv) Temperature
(v) Strain rate
(7) Two major approaches can be used to reduce the polymer viscosity: (i) in-
creasing the temperature, and (ii) increasing the shear rate. To reduce the
viscosity of a polymer composed of flexible chains, which approach is more
effective? How about a polymer with strong intermolecular interaction?
(8) The viscosity of a polymer can be reduced by two simple approaches: (i)
increasing the temperature, and (ii) increasing the shear rate. What are the
limitations for these two approaches?
(9) What is die swell? Why does the die swell behavior occur when a polymer
is extruded?
(10) What is the cause of melt fracture?
(11) What is the relationship between elongational viscosity and shear viscosity?
CHAPTER 9
Formation of Synthetic Polymer Fibers
Despite the variety of synthetic polymers, only a few of them can be made into
useful fibers. Fiber-forming polymers typically are linear and are capable of be-
ing oriented. If side groups are present, they must be simple and/or polar. The
molecular weights of fiber-forming polymers must be high enough for achiev-
ing good mechanical and thermal stability, and at the same time, low enough for
dissolving or melting for extruding into fibers. These linear, moderate-molecular
weight polymers can be made into fibers by melt spinning, solution spinning, gel
spinning, liquid crystal spinning, dispersion spinning, etc. This chapter focuses on
the fundamental aspects of the melt spinning process. Other spinning processes
will be briefly discussed.
9.1 MELT SPINNING

Melt spinning typically is used to make fibers from polyolefins, nylons, and poly-
esters. One major advantage of melt spinning is that it does not require a puri-
fication step since no solvent is used. Another advantage is that melt spinning
has very high production speed, which ranges from several hundred meters per
minute to several thousand meters per minute.
9.1.1 Fiber Formation Process

The melt spinning process involves melting and extrusion of the fiber-forming
polymer through a spinneret, i.e., a multihole capillary die, followed by cooling
and solidification to form filaments. Figure 9.1 shows a schematic of the melt
spinning process. The fiber-forming polymer used typically is in the form of dried
granules or pellets. The dry polymer is fed into an extruder through a hopper.
151
152 Formation of Synthetic Polymer Fibers
Figure 9.1. Schematic of the melt spinning process.
The extruder melts and conveys the polymer into a metering pump. The metering
pump delivers the polymer melt at an accurately controlled rate into the spin pack,
where the polymer is filtered and pushed through the holes of the spinneret. The
extruded filaments pass through a cooling chamber to be solidified. The rates of
cooling and solidification in the cooling chamber typically are controlled by cross
air blow. Heated cells and liquid baths also may be used in some cases to reduce
or increase the cooling and solidification rates. During the solidification process,
a take-up device typically is used to create a tensional force to draw down the
filaments to smaller diameters. The resultant filaments often are passed directly to
the drawing process to improve the physical structure and mechanical properties.
Many processing parameters affect the structure of melt-spun fibers. The most
important processing parameters are:
1. Spinneret orifice shape, dimension, and number.

2. Extrusion temperature.
3. Mass flow rate through each spinneret hole.
4. Take-up velocity.
5. Length of the spinline, and
6. Cooling condition.
During melt spinning, these parameters must be carefully controlled so the

desired structure and properties can be achieved for melt-spun fibers. To under-
stand how to control these processing parameters, an engineering analysis of the
Melt Spinning 153
melt spinning process is required. The basic engineering analysis involves the
flow of polymer melts, the balances of forces and energies, and the development
of molecular orientation and crystalline structure. The flow behavior of polymers
has been discussed in Chapter 8. The following sections address the force balance,
energy balance, and molecular orientation and crystallization in the spinline.
9.1.2 Force Balance

Figure 9.2 shows forces acting on a single filament in a spinline. The balance of
forces on the filament can be described as:
Frheo , z = Frheo , 0 + Fsurf + Finert + Fdrag − Fgrav (9.1)
where Frheo,z is the rheological force at a distance z from the spinneret, Frheo,0 the
rheological force at z = 0 (i.e., spinneret exit), Fsurf the surface tension at the in-
terface between the filament and the cooling medium, Finert the inertial force pro-
duced by the acceleration of the filament, Fdrag the drag force caused by the fiber
moving through the cooling medium, and Fgrav the gravitational force.
For melt spinning, the surface tension of polymer melt typically is small com-
pared to other forces, and it can be assumed that:
Fsurf = 0 (9.2)
The inertial force can be expressed by:
Finert = w ( u z − u0 ) (9.3)
Figure 9.2. Forces acting on a spinning filament.

where w is the mass throughput rate per spinneret hole, uz the velocity of the
filament at a given point, and u0 the velocity of the filament at the exit of the
spinneret.
The drag force can be written as:
z
Fdrag = ∫πd z τ z dz (9.4)
0
where dz is the diameter of the filament at a given point, τz the shear stress at the
fiber surface caused by the friction with the cooling medium. If the cooling me-
dium is air, the shear stress at the fiber surface is given by:
1
τz = ρa urel 2 Cdrag (9.5)
2
where ρa is the density of air, urel the relative axial velocity of the filament and the
cooling air, and Cdrag the drag coefficient. If the cooling air does not move parallel
to the filament, the relative axial velocity urel is the same as the filament velocity
uz. Equation 9.4 then can be rewritten as:
z
1
2 ∫0
Fdrag = ρa Cdrag u z 2 πd z dz (9.6)
There are many different methods for calculating the drag coefficient. One of
the most commonly used methods is to relate the drag coefficient to the Reynolds
number (Re) by:
Cdrag = KRe − n (9.7)
where the exponent n typically ranges from 0.6 to 0.8, and K ranges from 0.23 to
>1. The Reynolds number Re can be expressed by:
ρQ
Re = (9.8)
πD0 η
where ρ is the density of the polymer, Q the volumetric throughput rate per spin-
neret hole, D0 the diameter of the spinneret hole, and η the viscosity of the poly-
mer melt. When cooling mediums other than air are used, Equation 9.6 still may
be valid, but appropriate medium densities and drag coefficient values must be
used.
Melt Spinning 155
The gravitational force is given by:
z 2
d 
Fgrav = ∫π  z  ρgdz (9.9)
0 
2 
where ρ is the density of the filament, and g the acceleration of gravity. According
to the material balance, the mass throughput rate can be written as:
2
d 
w = ρAz u z = ρπ  z  u z (9.10)
 2 
where Az is the cross-sectional area of the filament at a given point. The gravita-
tional force can then be rewritten as:
z
w
Fgrav = ∫ gdz (9.11)
u
0 z
Based on Equations 9.2, 9.3, 9.6, and 9.11, the force balance can be rewritten
as:
z z
1 w
Frheo , z = Frheo , 0 + w ( u z − u0 ) +
2 ∫0
ρa Cdrag u z 2 πd z dz − ∫ gdz (9.12)
u
0 z
Compared with Frheo,0, it is more convenient to measure the rheological force Frheo,L
at the take-up device, i.e., z = L, where L is the distance between the take-up
device and the spinneret. Therefore, the rheological force at a given point can be
given by:
L L
1 w
Frheo , z = Frheo , L − w ( uL − u z ) −
2 ∫z
ρa Cdrag u z 2 πd z dz + ∫ gdz (9.13)
u
z z
where uL is the velocity of the filament at the take-up device, i.e., the take-up ve-
locity. At the take-up device, the Frheo,L is equal to the measurable take-up tension
force:
Frheo , L = Ftake −up (9.14)

By using Equation 9.13, the rheological force at a given point in the spinning fila-
ment can be calculated from measurable quantities. The gradient of axial tension
along the spinning filament can be obtained from:
dFrheo , z du z 1 w
=w + ρa Cdrag u z 2 πd z − g (9.15)
dz dz 2 uz
In melt spinning, the tensile stress developed in the filament is a very important
parameter and it largely determines the final structure and properties of the melt-
spun filaments. The tensile stress (σz) at a given point is related to the rheological
stress by:
Frheo , z
σz = 2
(9.16)
d 
π z 
 2 
Therefore, the tensile stress along the filament also can be obtained from measur-
able quantities.
Based on the force balance analysis described above, the contributions of dif-
ferent forces to the rheological force and tensile stress can be obtained. Figure 9.3
shows the rheological force, inertial force, drag force and gravitational force
along a thin spinning filament. The actual curve shapes vary from polymer to
polymer and are affected by various processing parameters. Curve shapes that are
completely different from those shown in Figure 9.3 are possible. For the example
Figure 9.3. Evolutions of rheological, inertial, drag and

gravitational forces along the spinline.
Melt Spinning 157
shown in Figure 9.3, the gravitational force is the dominant force in the vicinity of
the spinneret. However, the gravitational force is small compared to other forces
when the distance from spinneret is significant. Except for the vicinity of the
spinneret, the inertial and drag forces are the major components for the rheologi-
cal force. The inertial force develops earlier than the drag force, and it levels off
as the filament is solidified and the diameter becomes relatively constant. Since
the structure of the polymer filament is mainly developed before the filament is
completely solidified, the inertial force is the dominant force that determines the
physical structure of the filament. The drag force develops later, and it continues
to increase even after the filament is solidified. As a result, it eventually could
become greater than the inertial force if the length of the spinline is long enough.
The velocity, velocity gradient and filament diameter also can be obtained
from the force balance analysis. Figure 9.4 shows the filament velocity as a func-
tion of distance from the spinneret. In the vicinity of the spinneret, the velocity
of filament may reduce slightly due to the die swell phenomenon. After that, the
velocity of the filament keeps on increasing until the filament is solidified, after
which the velocity levels off. At higher spinning speed, the velocity developed
along the spinline also is higher.
The velocity gradient (duz/dz) along the spinline is shown in Figure 9.5. The
entire spinline can be divided into three regions:
1. Die swell region (0 ≤ z < ls): negative velocity gradient (duz/dz < 0).
2. Elongation region (ls ≤ z < l∞): positive velocity gradient (duz/dz > 0).
3. Rigid motion region (l∞ ≤ z < L): zero velocity gradient (duz/dz = 0).
Figure 9.4. Evolution of filament velocity along the spinline.

Figure 9.5. Evolution of velocity gradient along the spinline.
The length (ls) of the die swell region is short, ranging from a fraction of a mil-
limeter to several millimeters. In the die swell region, the molecular orientation
developed in the spinneret channel is relaxed and polymer chains coil, leading
to a negative velocity gradient. The elongation region plays the most important
role in determining the diameter and physical structure of the melt-spun filament.
The velocity gradient is positive and molecular orientation occurs because of the
elongational flow. The length (l∞ - ls) of the elongation region ranges from 0.3 to
1.5 meters, depending on the cooling condition, filament thickness, take-up ve-
locity, etc. The solidification and crystallization are completed in the rigid region.
The velocity gradient is zero in the rigid region. However, the crystallization pro-
cess in the rigid region is promoted by the molecular orientation, and hence the
rigid region still affects the structure and properties of the resultant fiber.
The evolution of filament diameter is shown in Figure 9.6. In the vicinity of
the spinneret, the filament diameter experiences a small increase due to the die
swell phenomenon. At high spinning speed, the filament diameter reduces rapidly
after the die swell region and then it levels off after the filament is solidified. At
low spinning speed, the die swell in the vicinity of the spinneret does not have ap-
parent change; however, the reduction rate of filament diameter in the elongation
region becomes lower.
9.1.3 Energy Balance

The temperature of the spinning filament is a function of distance from the spin-
neret and is determined by the energy balance along the spinline. Heat transfer
Melt Spinning 159
from the filament involves three mechanisms: conduction, radiation, and convec-
tion. The conduction of heat occurs: (1) from the hotter section of the filament to
the cooler section of the filament, and (2) from the filament to cooler objects in
contact with the filament, such as the take-up device. Compared to other mecha-
nisms, the conduction of heat in the melt spinning process is negligible.
The heat transferred by radiation is strongly dependent on the temperature of
the spinning filament. The apparent heat transfer coefficient for radiation can be
expressed by:
T 4 − Ta 4
α*rad = K (9.17)
T − Ta
where K is a constant, T the absolute temperature of the radiating body, and Ta the
absolute temperature of air (or other cooling medium). For the spinning of inor-
ganic fibers, such as glass fibers, the temperature of the spinning filament is high
(600–1100°C) and the radiation is significant. However, polymers typically are
extruded at temperatures below 300°C, and the temperature of the spinning fila-
ment decreases quickly after exiting the spinneret. Hence, during the melt spin-
ning of polymer fibers, the effect of radiation is appreciable only in the vicinity of
the spinneret. Experimental work shows that even in the vicinity of the spinneret,
the radiation heat transfer does not exceed 10–15% of convection heat transfer.
Therefore, the convection of heat is the main mechanism for the heat transfer
along the spinline. In most theoretical analyses, radiation is either neglected or
is incorporated into the convective contribution by selecting an appropriate heat
transfer coefficient.
Figure 9.6. Evolution of filament diameter along the spinline.

Many theories have been developed to predict the heat transfer coefficients of
spinning filaments. One of the commonly used relationships is given by Kase and
Matsuo:
1
1
k  ρa u f d    u c 6
2
3 
α* = 0.42 a   1 +  8   (9.18)
d  ηa    u f  
 
where ka is the thermal conductivity of air, d the diameter of the spinning filament,
ρa the density of air, ηa the viscosity of air, uf the velocity of the filament, and uc
the transverse velocity of air (i.e., the component of air velocity perpendicular to
the filament).
In addition to heat transfer, the crystallization of the polymer also contrib-
utes to the change of temperature along the spinning filament. When a polymer
crystallizes, it releases a certain amount of heat to form a low-energy, ordered
structure. The latent heat of crystallization slows the temperature reduction along
the spinline. The effect of crystallization on the filament temperature may not be
apparent when polyester (especially, polyethylene terephthalate, PET) is melt-
spun at low or moderate spinning speeds. This is because polyester has slow crys-
tallization kinetics and it may not crystallize in the spinline. However, for the
melt spinning of polyolefins and nylons, the crystallization effect has to be taken
into consideration since these polymers crystallize rapidly along the spinline. In
addition, for the high-speed melt spinning of polyester fibers, the crystallization
effect cannot be neglected since the crystallization process is promoted by the
high spinning speed.
Neglecting the radial temperature variations within the filament, the energy
balance of the spinning filament can be written as:
dT πd α* (T − Ta ) ∆H c d θ
=− + (9.19)
dz wC p C p dz
where w is the mass throughput rate per spinneret hole, Cp the heat capacity of the
polymer, ΔHc the heat of crystallization, and θ the degree of crystallinity. On the
right side of Equation 9.19, the first term is the contribution of heat transfer, and
the second term is the contribution of crystallization. Based on the energy bal-
ance, the temperature distribution along the filament can be obtained.
Figure 9.7 shows the filament temperature as a function of distance from the
spinneret. When polyester fibers are melt-spun, the crystallization effect can be
neglected, the filament temperature decreases with increasing distance from the
spinneret due to the heat transfer. When polyolefin and nylon fibers are melt-spun,
a slight shoulder appears on the temperature-distance curve due to the crystalliza-
tion effect. Figure 9.8 compares the temperature and crystallinity of a polyolefin
Melt Spinning 161
Figure 9.7. Evolution of filament temperature along the spinline.
or nylon filament along the spinline. The shoulder at the temperature-distance

curve occurs at the same point when the crystallization starts. In some cases, the
crystallization effect is so significant a slight increase in filament temperature can
be observed.
Figure 9.8. Evolution of filament temperature and crystallinity along the spinline.

9.1.4 Molecular Orientation

Molecular orientation is critically important for achieving good mechanical be-
havior for polymer fibers. The high degree of molecular orientation is produced
mainly in the drawing step, which will be discussed in section 9.7. However, some
orientation also is developed during the melt spinning process. As discussed in
Chapter 8, the capillary flow of the polymer in the spinneret channels facilitates
the orientation of polymer chains along the flow direction. However, compared
with the molecular orientation developed along the spinline after the polymer
exits from the spinneret, the degree of molecular orientation produced in the spin-
neret channels is insignificant due to the following two reasons:
1. The time for the polymer to pass through the spinneret channel is rath-
er short and is not sufficient for the molecular orientation to be fully
developed.
2. The polymer jet exiting from the spinneret is subject to relaxation, which is
indicated by the die swell phenomenon. As a result, the molecular orienta-
tion developed in the spinneret is largely relaxed.
After the polymer is extruded from the spinneret, the elongational flow of the
spinning filament leads to molecular orientation in both amorphous and crystal-
line phases. For the melt spinning of polyester fibers, the polymer crystallizes
slowly and the molecular orientation developed along the spinline is contributed
mainly by the amorphous phase. However, for polyolefin and nylon fibers, the
rapid crystallization process determines that the molecular orientation involves
both the amorphous and crystalline phases.
For uniaxial molecular orientation, an orientation factor can be used to de-
scribe the orientation of polymer chains relative to the filament axis. In general,
the orientation factor (f ) of a spinning filament is expressed by:
α z − αr 3cos 2 φ − 1
f = = (9.20)
α1 − α2 2
where αz – αr is the mean difference between the components of the polarizabil-

ity in the filament axial and radial directions, α1 – α2 the difference between the
polarizability parallel to and perpendicular to the polymer chain segment, and
f the angle between the polymer chain axis and the filament axis. According to
Equation 9.20, f = 1 and f = 0 when all polymer chains are oriented parallel to
the filament axis, and f = 0 and f = 54°44’ when all polymer chains are randomly
dispersed without molecular orientation.
One of the most convenient methods to study the molecular orientation in a
filament is the use of birefringence. The birefringence (Δn) of a filament is basi-
cally the refractive index difference between the axial and radial directions:
Melt Spinning 163
∆n = nz − nr (9.21)
where nz and nr are the reflective indices on the filament axial and radial direc-
tions, respectively. For fibers made from the same type of polymer, a higher bire-
fringence value indicates a higher degree of molecular orientation.
By using the birefringence, the orientation factor of the spinning filament can
be rewritten as:
∆n
f = (9.22)
∆nintri
where Δnintri is the intrinsic birefringence of the polymer, i.e., the maximum pos-
sible birefringence when all polymer chains are perfectly oriented along the fila-
ment axis. Since the orientation factor is proportional to the birefringence, many
literature reports use the measurable birefringence to directly describe the mo-
lecular orientation of polymer fibers.
In the spinline, the birefringence of a completely amorphous filament is pro-
portional to the applied tensile stress (σz):
∆n = Copt σ z (9.23)
where Copt is the stress optical coefficient. The applied tensile stress at the take-
up device is measurable, and the applied tensile stress at a given point along
the spinline can be calculated based on the force balance discussed in sec-
tion 9.1.2. Therefore, the birefringence of a spinning filament can be obtained by
Equation 9.23 as long as the stress optical coefficient is known. According to the
rubber elasticity theory, the stress optical coefficient can be given by:
(n + 2)
2 2
Copt =k ( α1 − α2 ) (9.24)
nT
where k is a constant, T the absolute temperature of the spinning filament, n is the
average reflective index of the polymer, and α1 - α2 the difference between the po-
larizability parallel to and perpendicular to the polymer chain segment. Although
the stress optical coefficient is inversely proportional to the absolute temperature,
many literature reports treat it as a constant while analyzing the birefringence of
spinning filaments.
When the molecular orientation of the crystallize phase is taken into consider-
ation, the birefringence of the spinning filament can be expressed by:
∆n = (1 − θ ) f a ∆nintri , a + θf c ∆nintri , c + ∆n form (9.25)

where θ is the degree of crystallinity, fa and fc the orientation factors of the amor-
phous and crystalline phases, Δnintri,a and Δnintri,c the intrinsic birefringences of the
amorphous and crystalline phases, and Δform a form birefringence due to the inter-
action of the two phases with the light.
Figure 9.9 shows the evolution of birefringence or orientation factor along the
spinline at different spinning speeds. The spinning speed affects the development
of molecular orientation. With increase in spinning speed, the molecular orienta-
tion develops at an earlier time and the maximum orientation that can be devel-
oped in the spinline also is higher.
9.1.5 Crystallization
Similar to molecular orientation, the development of crystalline structure plays
an important role in melt spinning. In general, the crystallization of polymers oc-
curs between the glass transition temperature (Tg) and melting temperature (Tm).
Literature reports show the crystallization could occur above the melting tem-
perature, especially under high molecular orientation. However, the amount of
such crystallization is small and has no apparent effect on the final structure and
properties of most melt-spun fibers, and hence this chapter will only address the
crystallization below Tm.
The crystallization of a polymer involves nucleation and growth processes. In
an isothermal crystallization process, the crystallization occurs immediately when
the temperature is lower than Tm. However, if the filament temperature is only
slightly lower than Tm, the driving force for the nucleation process is low and the
polymer crystallizes slowly. On the other hand, if the temperature is too low and is
Figure 9.9. Evolution of filament birefringence or orientation factor along the spinline.

Melt Spinning 165
Table 9.1. Tm, Tg, Tmax, and Kmax of some polymers.
Tm Tg Tmax Kmax
Polymer
(°C) (°C) (°C) (sec–1)
Polypropylene 180 –20 65 0.55
Polyester, polyethylene terephthalate 267 67 190 1.6×10–2
Nylon 6 228 45 146 0.14
Nylon 6,6 264 45 150 1.64

Sources: Ziabicki, A., Fundamentals of Fiber Formation: the Science of Fibre Spinning and
Drawing, John Wiley & Sons, 1976.
close to the glass transition temperature, the growth of crystals is slow due to the
lack of molecular mobility. As a result, the crystallization rate reaches its maxi-
mum at an intermediate temperature (Tmax). Table 9.1 shows the values of Tm, Tg,
Tmax, and maximum crystallization rate (Kmax) at Tmax of polymers that commonly
are used in melt spinning.
In melt spinning, the crystallization process in the spinning filament involves
very high cooling rates (up to 103–104°C/s) and hence it is absolutely non-isother-
mal. Increasing the cooling rate reduces the time for the polymer to crystallize at
a given temperature. Therefore, the polymer crystallization process mainly occurs
at temperatures much lower than the melting temperature. If the cooling rate is
sufficiently high, the polymer may not crystallize even at the temperature of maxi-
mum crystallization rate.
Figure 9.10 shows the so-called “continuous cooling transformation diagram”,
i.e., the relationship between cooling condition and crystallization kinetics, of a
polymer under quiescent condition. In the diagram, the cooling curves are plotted
Figure 9.10. Continuous cooling transformation diagram under quiescent condition. A

higher number on the cooling curve indicates a higher cooling rate.
for a polymer on temperature versus log time. The cool curves are numbers 1–6,
with a higher number indicating a higher cooling rate. The “C curve” on the right
bottom corner shows the start of crystallization under quiescent condition. As
shown in Figure 9.10, with increase in cooling rate, the crystallization starts at a
lower temperature, i.e., greater supercooling is required. If the cooling rates are
very high and miss the “nose” of the c-curve, the crystallization will not happen
and the resultant product will be completely amorphous (see curves 4, 5 and 6).
In melt spinning, the rate of crystallization is increased due to the development
of molecular orientation under stress. As a result, the c-curve of the spinning fila-
ment is moved to shorter times. Figure 9.11 shows the continuous cooling trans-
formation diagram of a spinning filament, which is under the tension created by
the take-up device. For comparison, the c-curve of the polymer under quiescent
condition also is shown in the figure using a dash line. It is seen the c-curve of the
filament is shifted to the left, and as a result, the crystallization occurs at higher
temperatures. For the cooling curves 4 and 5, the crystallization process does
not occur under quiescent condition, but it can be initiated when the polymer is
oriented in the spinning filament. The crystallization process in the spinning fila-
ment is called stress-induced crystallization or orientated crystallization since it is
developed under tensile stress and is highly influenced by molecular orientation.
In the classical treatment, the crystallization kinetics, i.e., the crystallinity (θ)-
time (t) relationship, of the isothermal crystallization process is described by the
Avrami equation:
θ = 1 − exp ( −kt n ) (9.26)
where k is the crystallization rate constant, and n the Avrami index. Both k and n can
be obtained from isothermal crystallization experiments. For the non-isothermal
Figure 9.11. Continuous cooling transformation diagram of a spinning filament. A higher

number on the cooling curve indicates a higher cooling rate.
Melt Spinning 167
crystallization process in melt spinning, the Avrami equation is modified by incor-

porating an “isokinetic approximation”:
  t  
n
θ = 1 − exp −  ∫K (T ( t ') ) dt '  (9.27)

  0  
where n is the same Avrami index derived from isothermal crystallization experi-
ments, T(t’) the temperature at time t’, and K(T) the non-isothermal crystallization
rate at temperature T. For practical treatments, Equation 9.27 often is rewritten to
the following differential form:
n −1
dθ   1  n
= nK (T ) (1 − θ) ln   (9.28)
dt   1 − θ 
Here, the non-isothermal crystallization rate, K(T), is related to the isothermal

crystallization rate, k(T), by:
1
K (T ) =  k (T )  n (9.29)
Equation 9.28 can be used to describe the kinetics of non-isothermal crystal-

lization process under quiescent condition. However, the crystallization process
in the spinning filament is non-quiescent, and the molecular orientation developed
under the tensile stress affects the crystallization rate. Therefore, the traditional
non-isothermal crystallization rate, K(T), must be replaced with the non-isother-
mal, stress-induced crystallization rate, K(T,f), where f is the orientation factor.
K(T,f) also is called the total crystallization rate. With the total crystallization rate,
Equation 9.28 can be rewritten to give:
n −1
dθ   1  n
= nK (T , f ) (1 − θ) ln   (9.30)
dt   1 − θ 
The relationship between the total crystallization rate and the orientation factor
is given by:
K (T , f ) = K (T , 0 ) exp  A (T ) f 2  (9.31)
where A(T) is an empirical parameter and is always positive. For example, the
A(T) value for polyester is 210 at 95°C and 940 at 115°C. Therefore, the stress-
induced crystallization rate increases monotonically with increase in molecular
orientation. Ziabicki suggested the value of K(T,0) at quiescent condition can be

approximated by:
 2 (T − Tmax ) 
2
K (T , 0 ) = K max exp  −4 ln  (9.32)

 D2 
where Kmax is the rate constant at the temperature (Tmax) of the maximum crystal-
lization rate, and D the half-width of the K-T curve. Substituting Equation 9.32
into Equation 9.31 gives:
 2 (T − Tmax ) 
2
K (T , f ) = K max exp  −4 ln 2
+ A (T ) f 2  (9.33)
 D 
In addition to Equation 9.33, the total crystallization rate K(T,f) also can be ob-
tained by the combination of rubber elasticity theory and quiescent crystallization
theory. For example, Katayama et al. and Patel et al. suggested the total crystal-
lization rate also can obtained by:
 U*   C0 
K (T , f ) = K 0 exp  −  exp  − 2 2
(9.34)
 R (T − T∞ )   T ∆T + C1T f 
where K0 and C0 are constants obtained from quiescent crystallization kinetics

data, C1 a constant determined by back calculation from experimental data ob-
tained from the melt spinline, U* the activation energy for segmental jumping, R
the gas constant, T∞ = Tg – 30, and ΔT = Tm – T.
With the knowledge of the total crystallization rate K(T,f), the kinetics of the
non-isothermal, stress-induced crystallization process in melt spinning can be de-
scribed by using Equation 9.30. Figure 9.12 shows the development of crystallini-
ty along the spinning filament. The crystallinity of the spinning filament increases
with increase in spinning speed. Higher spinning speed leads to greater molecular
orientation, which in turn promotes the crystallization process. However, the in-
creased spinning speed also leads to higher cooling rate, resulting in lower crys-
tallization temperature. In this case, the crystallization could decrease with in-
crease in spinning speed.
9.1.6 Cross-Sectional Effects

The analyses discussed in previous sections focus on the structural develop-
ment along the filament axis with the assumption the cross-sectional structure
is uniform. However, several factors could lead to a non-uniform cross-sectional
Melt Spinning 169
Figure 9.12. Evolution of filament crystallinity along the spinline.
structure for melt-spun filaments. One of the most important factors is the radial
temperature gradient. The heat transfer in melt spinning starts from the surface of
the spinning filament, and hence the center of the filament typically is hotter than
the surface. Theoretical modeling results show the radial temperature gradient of
the spinning filament could range from 103 to 105°C/cm. Such large temperature
gradients cause variations of viscosity and stress in the fiber cross-section, which
in turn lead to non-uniform distributions of molecular orientation and crystal-
linity. Figure 9.13 shows the variation of molecular orientation across the radial
direction of a spinning filament. The skin of the filament has higher molecular
orientation than the core. The radial distribution of the crystallinity is similar to
that of the molecular orientation. In addition, the size and structure of the crystals
also may vary along the radial direction in the cross-section.
Another important factor for causing the non-uniform cross-sectional structure
is the cooling method. One commonly used cooling method is the cross air blow
process (Figure 9.14). The effect of cross air blow on the molecular orientation
distribution across the fiber cross-section is shown in Figure 9.15. Compared with
the leeward direction, the cooling rate on the windward direction is greater, and
hence the molecular orientation often is higher. Similarly, the distribution of crys-
tallinity in the cross-section also is affected by the non-uniform cooling caused
by cross air blow.
The cooling process also can be affected by the cross-sectional shape of the
spinning filament. When a non-circular filament is being spun, the cooling pro-
cess is non-uniform in the cross-section and different parts of the cross-section
can exhibit different degrees of molecular orientation and crystallinity.
Figure 9.13. Radial distribution of molecular orientation in a filament cross-section.
9.1.7 Multifilament Effects

In practical melt spinning, multi filaments are extruded simultaneously from the
spinneret. During that process, the conditions applied to each filament may vary.
For example, the hot filaments can heat the cooling air as it passes through the
multifilament bundle, and as a result the filaments on the leeward side of the
bundle experience slower cooling rates than those on the windward side. The
high-speed filaments also may impart an axial velocity to the incoming air and
Figure 9.14. Schematic of cross air blow process.

Solution Spinning 171
Figure 9.15. Effect of cross air blow on the cross-sectional distribution of molecular ori-
entation in a filament.
reduce the transverse velocity in the cross-blow direction, which in turn lead to
variations in drag forces created by the cooling air. As a result, the structure and
properties vary from one filament to another in the multifilament bundle.
Although multifilament effects are important, there are few literature reports
on this topic mainly because of the complexities associated with accurately de-
scribing the heat transfer and dynamics during multifilament spinning. The inter-
ested reader is referred to the original works by Yasuda et al. (1978), Ishihara et
al. (1989), and Dutta (1987).
9.2 SOLUTION SPINNING

Solution spinning can be used to produce fibers from polymers that do not form
thermally stable melts, but can be dissolved in solvents. Two important solution
spinning methods are dry spinning and wet spinning. Dry spinning typically is
used for polymers that are soluble in volatile solvents. Wet spinning often is used
for polymers that are dissolved in non-volatile or thermally unstable solvents.
9.2.1 Dry Spinning

Figure 9.16 shows a schematic of the dry spinning process. The spinning solu-
tion is delivered into the multi-hole spinneret by a metering pump at an accu-
rately controlled rate. The spinneret often is placed in an enclosed drying tower
for solvent recovery. Upon exiting the spinneret, the polymer solution comes into
contact with a stream of hot inert gas (typically, air), and the solvent evaporates.
The hot gas carries most, if not all, of the solvent evaporated from the polymer
Figure 9.16. Schematic of the dry spinning process.
solution and leaves the drying tower. During this process, the polymer concentra-
tion in the filaments increases and the filaments solidify. Dry spinning (100–800
meters per minute) is faster than wet spinning, but slower than melt spinning.
Examples of dry-spun fibers are polyacrylonitrile, polyvinyl chloride, cellu-
lose acetate fibers, among others. Solvents used in dry spinning typically have low
boiling point and low vaporization heat. Solvents also should be thermally stable,
easily recoverable, nontoxic, and non-explosive. Examples of solvents used in dry
spinning include but are not limited to alcohols, acetone, ether, and tetrahydrofu-
ran. For effective solvent evaporation and recovery, the drying gas often is heated
and the inlet gas temperature varies from 100 to 250°C, depending on the nature
of polymer and solvent. To reduce the amount of solvent used, concentrated poly-
mer solutions are preferred, if possible. Typically, polymer concentrations range
from 15 to 40% in practical dry spinning.
The engineering analysis of dry spinning is complex. Compared with melt
spinning, dry spinning involves an additional process — mass transfer. The mech-
anisms of mass transfer in dry spinning include: 1) flash vaporization that takes
place at the spinneret exit as a result of decompression of the heated polymer solu-
tion, 2) diffusion of solvent within the spinline, and 3) convection of solvent from
the spinline surface to the dry gas medium. The mass transfer affects the force
balance and energy balance. In the force balance, both the inertial force and drag
force are different from those in melt spinning. In the energy balance, the latent
heat of solvent vaporization and the convective heat transfer have to be included.
For detailed discussion, the interested reader is referred to the works published
by Ziabicki (1976), Ohzawa and Nagano (1970), and Gou and McHugh (2004).
9.2.2 Wet Spinning

Figure 9.17 shows the most basic wet spinning process. The spinneret is placed in
the spinning bath, which contains a liquid that is miscible with the solvent in the
Solution Spinning 173
Figure 9.17. Schematic of a basic wet spinning process.
polymer solution, but cannot dissolve the polymer. During spinning, the polymer
solution is extruded from the spinneret into the spinning bath and the polymer
is precipitated to form solid filaments. The solvent in the polymer solution is re-
moved in the spinning bath by the counter-diffusion mechanism.
There are many different designs of the wet spinning systems. The spinning
bath can be situated horizontally or vertically. In vertical spinning bath, the fila-
ments can move upward or downward. In addition to the spinning bath shown
in Figure 9.17, more coagulating baths may be needed to remove the solvent,
and sometimes further manufacturing steps (e.g., drawing, washing, drying, heat-
setting) are incorporated into the wet spinning process.
Wet spinning often is used to process polyacrylonitrile, polyvinyl chloride,
polyvinyl alcohol, and cellulose fibers. Polymer concentrations used in wet spin-
ning vary from 5 to 30%, which typically are lower than those in dry spinning.
Compared with melt spinning and dry spinning, the wet spinning process has
much lower spinning speeds, ranging from 50 to 300 meters per minutes. This is
mainly because the filaments undergo much greater drag forces in the liquid baths.
The engineering analysis of the wet spinning process also is complex. During
wet spinning, the skin of the filament is solidified first, and as a result, the mass
transfer mainly involves the counter-diffusion of solvent and nonsolvent between
the solidified skin and the solution core, and between the filament skin and the
liquid bath. The solidification process and the diffusion rates are affected by the
energy balance. For the force balance, some researchers believe the drag force is
the main factor for determining the dynamics of wet spinning; however, others
argue the rheological force also plays an important role. Another challenge in
the engineering analysis is many physical constants that are needed to describe
the wet spinning process are difficult to determine under the complex conditions
in the liquid bath. The interested reader is referred to works by Han and Segal
(1970), and Ziabicki (1976).
9.3 GEL SPINNING

Gel spinning, also known as dry-wet spinning, can be used to produce high-perfor-
mance polymer fibers. Different forms of gel spinning are available. Figure 9.18
shows the schematic of a basic gel spinning process. The polymer solution is
pumped into the spinneret, which is placed above the spinning bath to create an
air gap of short distance. The extruded filaments first pass through the air gap,
which is similar to the dry spinning process, and then into the spinning bath, like
in the wet spinning. The process allows the formation of gel filaments by thermo-
reversible gelation. These gel filaments have sufficient mechanical stability and
are transferred into an oven to be drawn to high-performance fibers.
For melt and solution spinning processes, the drawing step also is necessary
to achieve high molecular orientation and good mechanical properties, and the
effects of drawing will be discussed in section 9.7. However, the drawing for gel
spinning is unique and is highlighted here. In general, the as-spun gel filaments
still contain a large amount of solvent and form swollen networks joined by small
crystalline regions (Figure 9.19). The drawing of such gel filaments typically is
carried out at an elevated temperature in an oven. Although drawing is possible
after complete removal of the solvent prior to the drawing step, the presence of
solvent during drawing can help facilitate the movement of polymer chains and
lead to extremely high molecular orientation in gel-spun fibers after drawing
(Figure 9.20).
Gel spinning enables the production of high-performance fibers from polymers
with flexible polymer chains, such as ultra-high molecular weight polyethylene
(UHMW-PE). To form gel filaments, the concentration of polymer solution is
typically low, e.g., 1–2% for UHMW-PE. The maximum attainable draw ratio
(λmax) is related to the initial polymer concentration (c) by:
λmax ∝ c −0.5 (9.35)
Therefore, lower polymer concentration leads to higher draw ratio, which is ben-
eficial for achieving ultra-high molecular orientation.
Figure 9.18. Schematic of a basic gel spinning process.

Liquid Crystal Spinning 175
Figure 9.19. Schematic of the swollen network structure in a gel filament.
9.4 LIQUID CRYSTAL SPINNING

High-performance fibers also can be obtained by the spinning of liquid crystalline
polymers. Liquid crystals are highly structured liquids, with orientational (nematic
and cholesteric) or positional (smectic) order of molecules (Figure 9.21). Liquid
crystallinity in polymers may be realized either by dissolving a polymer in a sol-
vent (i.e., lyotropic liquid crystalline polymers) or by heating a polymer above its
glass or melting transition point (i.e., thermotropic liquid crystalline polymers).
Rod-shaped liquid crystalline polymer chains can make high-performance fibers
because it is easy to orient these polymer chains in the fiber longitudinal direction,
without chain folding.
Figure 9.22 shows the unusual concentration dependence of the solution vis-
cosity of lyotropic liquid crystalline polymers. With increase in polymer con-
centration, the viscosity first increases. However, the viscosity decreases rapidly
after the concentration reaches a critical value. This is because at low concentra-
tions, the liquid crystalline phase has not been formed and the polymer solution
is isotropic, without ordered distribution of polymer chains. The viscosity of the
isotropic solution increases with increasing polymer concentration. When the
Figure 9.20. Schematic of the high molecular orientation in a gel-spun fiber.

Figure 9.21. Nematic, cholesteric and smectic phases of liquid crystals.
concentration reaches the critical value, the liquid crystalline phase is formed and
the orientational order of polymer chains reduces the resistance to flow, thereby
leading to lower viscosity. For thermotropic liquid crystalline polymers, the vis-
cosity of the polymer melt also decreases rapidly at the temperature when the
liquid crystalline phase appears.
Lyotropic liquid crystalline polymers can be spun into fibers by using a wet
or gel spinning process, while thermotropic liquid crystalline polymers can be
made into fibers by melt spinning. Although both lyotropic and thermotropic liq-
uid crystalline polymers can be made into fibers, most successful liquid crys-
tal fibers are based on lyotropic polymers. One such example is polyphenylene
terephthalamide (PPTA, Kevlar®), which was first produced by DuPont in 1972.
Figure 9.22. Viscosity-concentration relationship of lyotropic liquid crystal polymers.

Reaction Spinning 177
Kevlar® fibers can be produced by a method that is similar to the gel spinning
process (Figure 9.23). However, unlike the gel spinning of UHMW-PE fibers, in
which low-concentration solutions are used, Kevlar® fibers are spun from high-
concentration solutions, e.g., 20% PPTA in 80% sulfuric acid. In addition, the
high molecular orientation of Kevlar® fibers is achieved directly from the liquid
crystalline phase.
9.5 DISPERSION SPINNING

Dispersion spinning is used to produce fibers from polymers that are not soluble
and can melt only at very high temperatures. For example, polytetrafluoroethylene
(Tm = 327°C) cannot be spun into fibers by regular melt or solution spinning. To
form fibers, fine polytetrafluoroethylene particles can be dispersed in an aqueous
solution of destructible, fiber-forming polymer (e.g., polyvinyl alcohol) to make
a homogenous dispersion. The dispersion then can be spun into fibers using dry
or wet spinning. After the spinning process, a thermal-treatment step is applied to
decompose the fiber-forming polymer and sinter or fuse polytetrafluoroethylene
particles into fibers.
9.6 REACTION SPINNING

Reaction spinning of monomers or prepolymers can be used to form fibers from re-
active starting materials. One good example for reaction spinning is the formation
Figure 9.23. Spinning of Kevlar® fibers.

of spandex fibers. Some spandex fibers are produced by using two types of pre-
polymers: one is a flexible macroglycol, while the other is a stiff diisocyanate.
The macroglycol is a long chain polymer that has hydroxyl groups (-OH) on both
ends. This macroglycol part of the spandex fiber is flexible and is responsible
for its stretching characteristic. The diisocyanate is a shorter chain polymer, with
isocyanate groups (-NCO) on both ends. The diisocyanate part of spandex is rigid
and provides strength. During the fiber formation process, these two prepolymers
are mixed in a solution with some additives and are forced through the spinneret.
After exiting the spinneret, the filaments are heated in the presence of nitrogen
and solvent and the two prepolymers react to form solidified spandex fibers.
9.7 DRAWING—AN IMPORTANT “POST-SPINNING”

PROCESS
The spinning methods discussed in previous sections can convert polymer melts
or solutions into fibers. In many cases, the mechanical properties of as-spun fi-
bers are not sufficient for practical applications. The drawing process is, there-
fore, introduced to improve the properties of fibers by providing a large, irrevers-
ible elongation to as-spun fibers in the solid state to 20–8000% of their original
lengths. This elongation enables the extension and parallelization of polymer
chains and crystallites along the fiber axial direction, and the molecular orienta-
tion developed in drawing often is accompanied by changes in phase structure,
such as crystallization or partial destruction of crystallites, and in other structural
characteristics.
9.7.1 Modes of Drawing

In laboratory research, the drawing of fibers can be conducted by using tensile
testing apparatus with a furnace surrounding the fibers. The drawing typically is
carried at constant extension or strain rates. However, in most commercial fiber
production processes, drawing is carried out continuously either directly after the
spinning or else in a separate step with undrawn filaments previously collected on
drums, bobbins, etc. During such drawing, the undrawn filaments are supplied at a
constant velocity via a rotating feed roll, and taken up on another roll (or rolls) at
higher velocities, thereby resulting in the elongation of the filaments in proportion
to the draw ratio, i.e., the ratio of the take-up and feed velocities. Depending on
the polymer type, the drawing can be carried out in air at room temperature (e.g.,
nylon), in a hot air chamber (e.g., polyvinyl alcohol), on contact with a hot plate
or drums (e.g., polyester), or in liquid baths that contain plasticizers and/or other
agents (e.g., acrylic).
The limitation of the laboratory tensile experiments at constant extension or
strain rates is the maximum strain rate of tensile testing instruments is one order
Drawing—An Important “Post-Spinning” Process 179
of magnitude lower than that of the continuous force drawing process used in
commercial production. As a result, the differences in the magnitude and kinet-
ics of the structural development in laboratory tensile experiments and commer-
cial drawing process are significant. However, results from laboratory tensile
experiments have provided important fundamental understanding on the struc-
tural development of filaments during drawing, and they have helped interpret the
structure changes of filaments under the more complex drawing in commercial
production.
9.7.2 Drawing of Initially Amorphous Filaments

Some polymers (e.g., polyester) have slow crystallization kinetics, and they may
not have sufficient time to crystallize during spinning. As a result, as-spun fila-
ments made from such polymers are amorphous before drawing. However, during
drawing, the development of molecular orientation reduces the energy barrier be-
tween the amorphous and crystalline states, permitting crystallization at tempera-
tures where none would occur in the undrawn filaments.
Figure 9.24 shows the development of molecular orientation and crystallinity
of initially amorphous fibers. At low draw ratios, the polymer does not crystallize,
Figure 9.24. Development of molecular orientation and crystallinity of initially amorphous

filaments.
but molecular orientation is being developed. At a critical draw ratio (λ1), the
strain-induced crystallization occurs, and both crystallinity and molecular orien-
tation develop rapidly. There is a second critical draw ratio (λ2), after which the
development of molecular orientation and crystallinity slows down. The critical
draw ratio λ1 is determined by the onset of the crystallization process. The critical
ratio λ2 indicates the degree of crystallinity has reached a point, where some of
tie molecules between crystallites are fully extended (Figure 9.25). The slowing
of orientation and crystallization at draw ratios greater than λ2 is the result of the
formation of taut inter-crystalline tie molecules, which hinder the uncoiling and
disentangling of neighboring tie-chains that are not fully extended. The formation
of taut tie molecules between crystallites can be demonstrated by rapidly increas-
ing stress required to maintain the constant strain rate at draw ratios greater than
λ2 in laboratory tensile experiments (Figure 9.26).
Figure 9.25. Schematic of the formation of fully extended tie molecules between crystal-
lites at high draw ratios.
Figure 9.26. Stress-draw ratio curve of initially amor-

phous filaments in laboratory tensile experiments.
Drawing—An Important “Post-Spinning” Process 181
9.7.3 Drawing of Crystallized Filaments

Many as-spun filaments already are crystallized prior to drawing. For example,
undrawn polyethylene and polypropylene fibers often exhibit high level of crys-
tallinity, ranging from 50 to 80%. The degree of crystallinity in undrawn nylon
filaments is lower, but it still can reach 35%. However, the crystallites in these
filaments are either un-oriented or only have a low orientation degree, and as a
result, the mechanical properties of crystallized filaments prior to drawing still are
unsatisfactory. By applying the drawing process, the molecular orientation and
crystallinity of crystallized filaments can be further improved, leading to higher
mechanical properties.
Figure 9.27 shows the development of amorphous and crystalline phases in
crystallized filaments. The drawing process increases the molecular orientation
for both amorphous and crystalize phases. In general, at low draw ratios, the
amorphous phase in the filaments is deformed first and the molecular orienta-
tion of the amorphous phase increases. The deformation of the crystalline phase
becomes significant when the tie molecules between the adjacent crystallites are
highly extended. The deformation of the crystalline phase starts with the increase
of the crystallite thickness, followed by the tilting of the crystalline phase. With
increase in draw ratio, large crystallites may break up under stress to form smaller
fibril-like crystallites. These newly-formed fibrils still are attached to each other
by tie molecules, and they slide against each other to align along the drawing
direction.
The increased molecular orientation also induces higher degree of crystallinity,
especially when the drawing is conducted above the glass transition temperature.
Figure 9.28 shows the contents of amorphous and crystalline phases in crystallized
Figure 9.27. Development of molecular orientation in both amor-

phous and crystalline phases of crystallized filaments.
Figure 9.28. Change of phase contents during drawing of crystallized filaments.
filaments. With increase in draw ratio, more amorphous phase is converted into
crystalline phase. Studies also show that, in certain crystalline polymers, drawing
could cause the transformation of one crystal morphology to another.
9.8 HEAT TREATMENT AFTER DRAWING

The drawing process introduces some instabilities to the fiber structure and prop-
erties. For example, the formation of high molecular orientation and the destruc-
tion of original crystallites lead to a thermodynamically unstable system, which,
given enough time, tends toward some equilibrium by changing structural char-
acteristics, including fiber dimensions, crystallinity, molecular orientation, etc.
At room temperature, the changes usually are slow unless stimulated by swelling
agents. At elevated temperatures, the changes could occur quickly.
One effective approach to stabilize the fiber structure and properties is to apply
a heat treatment process after drawing. This post-drawing process often is called
“heat setting”. The purpose of the heat treatment is to:
1. stabilize the fiber dimensions,

2. equilibrate the crystalline structure,
3. improve the mechanical properties, and/or
4. modify the chemical structure, e.g., thermal crosslinking of polyvinyl alco-
hol fibers.
Depending on the polymer structure and fiber type, the heat treatment process
can be carried out by using dry air, water vapor, solid heating elements, or heated
liquid baths. Swelling agents (e.g., water) can be used to promote solvent-induced
References 183
molecular motions, which accelerate the structural stabilization. The heat treat-
ment temperature should be higher than the use temperature of the fibers to ensure
their stability under application conditions. Depending on the desired mechanical
properties, the heat treatment can be conducted in a free state or under tension.
In general, the heat treatment in a free state is more effective in minimizing the
residual shrinkage, but the treatment under tension is effective in reducing the
strain-at-break.
REFERENCES
[1] Baird, D.G., and Collias, D.I., Polymer Processing: Principles and Design,
[2] Barham, P.J., and Keller, A., “Review: High-Strength Polyethylene Fibres
from Solution and Gel Spinning”, Journal of Materials Science, 20, 2281–
2302, 1985.
[3] Cheng, S.Z.D., Li, F., Li, C.Y., McCreight, K.W., Yoon, Y., and Harris,
F.W., “Fibers from Liquid Crystalline Polymers”, in Structure Formation
in Polymeric Fibers, Editor Salem, D.R., Hanser Gardner Publications Inc,
2000.
[4] Cuculo, J.A., Hotter, J.F., and Zhou, Q., “Advances in the Control of Spinline
Dynamics for Enhanced Properties”, in Structure Formation in Polymeric
Fibers, Editor Salem, D.R., Hanser Gardner Publications Inc., 2000.
[5] Dutta, A., “Role of Quench Air Profiles in Multifilament Melt Spinning of
PET Fibers”, Textile Research Journal, 57, 13–19, 1987.
[6] George, H.H., “Spinline Crystallization of Polyethylene Terephthalate”,
in High-Speed Fiber Spinning: Science and Engineering Aspects, Editors
Ziabicki A., and Kawai, H., Krieger Publishing Company, 1991.
[7] Gou, Z, and McHugh, A.J., “Dry Spinning of Polymer Fibers in Ternary
Systems—Part I: Model Development Predictions”, International Polymer
Processing, 19, 244–253, 2004.
[8] Gou, Z, and McHugh, A.J., “Dry Spinning of Polymer Fibers in Ternary
Systems—Part II: Data Correlation and Predictions”, International Polymer
Processing, 19, 254–261, 2004.
[9] Gou, Z., McHugh, A.J., “Two-Dimensional Modeling of Dry Spinning of
Polymer Fibers”, Journal of Non-Newtonian Fluid Mechanics, 118, 121–
136, 2004.
[10] Han, C.D., Rheology and Processing of Polymeric Materials: Volume 1I:
Polymer Processing, Oxford University Press, 2007.
[11] Han, C.D., and Segai, L., “A Study of Fiber Extrusion in Wet Spinning. I.
Experimental Determination of Elongational Viscosity”, Journal of Applied
Polymer Science, 14, 2973–2998, 1970.
[12] Han, C.D., and Segai, L., “A Study of Fiber Extrusion in Wet Spinning. II.
Effects of Spinning Conditions on Fiber Formation”, Journal of Applied
Polymer Science, 14, 2999–3019, 1970.
[13] Ishihara, H., Hayashi, S., and Ikeuchi, H., “Computer-Simulation of Multi
Filament Air Jet Melt Spinning”, International Polymer Processing, 4, 91–
95, 1989.
[14] Kase, S., “Mathematical Simulation of Melt Spinning Dynamics: Steady-
State Conditions and Transient Behavior”, in High-Speed Fiber Spinning:
Science and Engineering Aspects, Editors Ziabicki A., and Kawai, H.,
Krieger Publishing Company, 1991.
[15] Kase, S., and Matsuo, T., “Studies on Melt Spinning. I. Fundamental
Equations on the Dynamics of Melt Spinning”, Journal of Polymer Science:
Part A, 3, 2541–2554, 1965.
[16] Kase, S., and Matsuo, T., “Studies on Melt Spinning. II. Steady-State and
Transient Solutions of Fundamental Equations Compared with Experimental
Results”, Journal of Applied Polymer Science, 11, 251–287, 1967.
[17] Katayama, K., and Yoon, M.G., “Polymer Crystallization in Melt Spinning:
Mathematical Simulation”, in High-Speed Fiber Spinning: Science and
Engineering Aspects, Editors Ziabicki A., and Kawai, H., Krieger Publishing
Company, 1991.
[18] Kubo, S., “Air Boundary Layer on A Filament in High-Speed Spinning”,
in High-Speed Fiber Spinning: Science and Engineering Aspects, Editors
Ziabicki A., and Kawai, H., Krieger Publishing Company, 1991.
[19] Lemstra, P.J., Bastiaansen, C.W.M., and Rastogi, S., “Basic Aspects of
Solution(Gel)-Spinning and Ultra-Drawing of Ultra-High Molecular
Weight Polyethylene”, in Structure Formation in Polymeric Fibers, Editor
Salem, D.R., Hanser Gardner Publications Inc, 2000.
[20] Matsui, M., “Fiber Formation Process in High-Speed Spinning of
Polyethylene Terephthalate”, in High-Speed Fiber Spinning: Science
and Engineering Aspects, Editors Ziabicki A., and Kawai, H., Krieger
[21] Nakamura, K., Watanabe, H., Katayama, K., and Amano, T., “Some
Aspects of Nonisothermal Crystallization of Polymers. I. Relationship be-
tween Crystallization Temperature, Crystallinity, and Cooling Conditions”,
Journal of Applied Polymer Science, 16, 1077–1091, 1972.
Problems 185
[22] Ohzawa, Y., and Nagano, Y., “Studies on Dry Spinning. II. Numerical
Solutions for Some Polymer–Solvent Systems Based on the Assumption
That Drying is Controlled by Boundary-Layer Mass Transfer”, Journal of
Applied Polymer Science, 14, 11179–1899 (1970).
[23] Patel, R.M., and Spruiell, J.E., “Crystallization Kinetics during Polymer
Processing - Analysis of Available Approaches for Process Modeling”,
Polymer Engineering and Science, 31, 730–738, 1991.
[24] Salem, D.R., “Structure Formation during Drawing of Flexible Chain
Polymers”, in Structure Formation in Polymeric Fibers, Editor Salem,
D.R., Hanser Gardner Publications Inc, 2000.
[25] Smook, J., and Pennings, A.J., “Preparation of Ultra-High Strength
Polyethylene Fibres by Gel-Spinning/Hot Drawing at High Spinning
Rates”, Polymer Bulletin, 9, 75–85, 1983.
[26] Spruiell, J.E., “Structure Formation during Melt Spinning”, in Structure
Formation in Polymeric Fibers, Editor Salem, D.R., Hanser Gardner
Publications Inc, 2000.
[27] Yasuda, H., Ishihara, H., and Yanagawa, H., “Computer Simulation of Melt
Spinning and Its Application to the Actual Process”, Sen-I-Gakkaishi, 34,
20–27, 1978.
[28] Ziabicki, A., Fundamentals of Fiber Formation: the Science of Fibre
Spinning and Drawing, John Wiley & Sons, 1976.
[29] Ziabicki, A., “The Mechanisms of ‘Neck-Like’ Deformation in High-Speed
Melt Spinning. 1. Rheological and Dynamic Factors”, Journal of Non-
[30] Ziabicki, A., “The Mechanisms of Neck-Like Deformation in High-Speed
Melt Spinning .2. Effects of Polymer Crystallization”, Journal of Non-
[31] Ziabicki, A., and Jarecki, L., “The Theory of Molecular Orientation and
Oriented Crystallization in High Speed Spinning”, in High-Speed Fiber
Spinning: Science and Engineering Aspects, Editors Ziabicki A., and
Kawai, H., Krieger Publishing Company, 1991.
PROBLEMS
1. What are the main processing parameters that affect the structure of melt-
spun fibers?
2. What are the forces that a spinning filament experiences during melt spin-
ning? Which one is the dominating force?
3. Describe the evolution of filament velocity along the spinline during melt
spinning.
4. What are the three heat transfer mechanisms? Which one is the dominating
mechanism for melt spinning?
5. Describe the evolution of filament temperature along the spinline during
melt spinning.
6. What is birefringence and what is its relationship with molecular
orientation?
7. Describe the evolution of molecular orientation along the spinline during
melt spinning.
8. Describe the evolution of crystallinity along the spinline during melt spinning.
9. Compare the structural evaluation of nylon, polyolefin and polyester fila-
ments during melt spinning.
10. How does the physical structure vary along the radial direction of the cross-
section of a melt-spun fiber?
11. Describe dry spinning and wet spinning.
12. Compare the pros and cons of melt spinning, dry spinning, and wet spinning.
13. What is gel spinning? Describe the unique features of gel spinning.
14. What is liquid crystal spinning? Describe the unique features of liquid
crystal spinning.
15. Describe the development of molecular orientation and crystallinity while
drawing an initially amorphous filament.
16. Describe the development of molecular orientation in both amorphous and
crystalline phases while drawing a crystallized filament.
17. What are the possible effects of the “heat setting” process?
18. Several methods can be used to obtain high-performance synthetic fibers.
Give three examples.
CHAPTER 10
Formation of Natural Polymer Fibers
The fiber formation processes of natural cellulose and protein fibers are controlled
by the genetic codes and are complex. Humans can only have limited influence
on the fiber formation processes by selectively adjusting the environmental condi-
tions, such as climate, soil, water source, farming method, etc. On the other hand,
manufactured cellulose and protein fibers are formed by using methods similar to
those for making synthetic fibers. The processing conditions of these fibers can be
adjusted to manipulate the fiber structure and properties.

Cotton is the most important natural cellulose fiber in the world. For the last
twenty years, about 40% of the fiber consumed worldwide is cotton. This section
focuses on the formation of cotton fibers.
10.1.1 Types of Cotton

Cotton fibers are seed hairs from plants belonging to the order Malvales, family
Malvaceae, tribe Gossypieae, and genus Gossypium. There are thirty-three recog-
nized species, but only four are grown on a commercial scale in the world. These
four species of cotton are: hirsutum, barbadense, aboreum, and herbaceum.
Gossypium hirsutum was developed in the United States from cotton that is
native to Central America, Caribbean and Mexico. Currently, about 90% of cot-
ton produced worldwide is G. hirsutum. The length of G. hirsutum lint fibers
ranges from 25 to 36 mm and is suitable for apparel, home furnishing, and indus-
try applications.
187
188 Formation of Natural Polymer Fibers
Gossypium barbadense is of South American origin and has the longest lengths,
ranging from 33 to 36 mm. Egypt and Sudan are the largest producers of G. bar-
badense fibers, but the United States and South America also produce some. G.
barbadense fibers provide 8% of current world cotton production and are mainly
used for high-quality apparel, luxury fabrics, and specialty yarns.
Gossypium arboretum and Gossypium herbaceum are Old World cottons that
are native to India and Pakistan, and Southern Africa and the Arabian Peninsula,
respectively. Both cottons are short (9.5–19 mm) and coarse. They mainly are
produced in South Asia and Southeast Asia.
Some cotton species (e.g., G. hirsutum) produce two types of fibers: lint fibers
and linters (or fuzz fibers). The lint fibers are long (e.g., 25 mm or longer) and thin
and can be removed from the seed surfaces easily. On the other hand, linters are
short (15 mm or shorter) and coarse, and they are tightly adhered to the seed. Due
to the differences in fiber lengths and diameters, lint fibers are suitable for mak-
ing textile fabrics, but fuzz fibers are not suitable for textile processing. However,
fuzz fibers still can be used for batting and padding in bedding, automobile ap-
plications, paper making, chemical feed stock for making plastics and rayons, etc.
Traditionally, varietal development of cottons was confined to the breeding
methods that depend on crossing parents within species. However, modern bio-
technology has enabled the production of biotech or transgenic cottons. The ini-
tial biotech cottons were developed for insect resistance and herbicide tolerance.
Currently, biotech cottons have been or are being developed to have properties,
such as good stress tolerance, high yield potential, improved agronomic perfor-
mance, etc.
Although there are many different types of cottons, the formation process of
cotton fibers is basically the same, as discussed below.
10.1.2 Formation of Cotton

Cotton fibers are the dried cell walls of formerly living cells, which are believed to
be the largest single cells in nature. These living cells are the seed coat hairs that
outgrow from individual epidermal cells on the outer integument of the ovules
in the cotton fruit. Typically, the cotton flower only blooms for one day and be-
comes senescent afterward. On the day of full bloom, or anthesis, the petals of
cotton flower are creamy white (Figure 10.1A), but they quickly change to yel-
low, then pink, and eventually, dark red. After two or three days, the flower petals
wither and fall off from the developing green pod, i.e., the so-called “cotton boll”
(Figure 10.1B). Inside the boll, moist cotton fibers are formed and push out from
the ovule surfaces of the newly formed seeds. As the fibers continue to grow, the
cotton boll turns to brown. Eventually, the boll opens and the fluffy cotton fibers
burst forth (Figure 10.1C).
The formation and growth of cotton fibers can be divided into four overlapping
but distinct stages, as described below:
Stage 1—Initiation: In general, each cotton boll contains 20–30 ovules,

spread in three to five segmented compartments (locules). The ovules are
attached to the plant by a connection called the funiculus. On, or slightly
before the day of anthesis, the initiation of lint fibers begins from the epider-
mal cells on the ovule surfaces. Typically, lint fiber development starts at
the more rounded end of the seed and proceeds around the seed surface to
the micropyle. Approximately, lint fiber cells outgrow from one out of every
four epidermal cells. At the initiation stage, the lint fiber cells exhibit a bul-
bous appearance and are visible above the formerly smooth ovule epidermal
surface (Figure 10.2). For some cotton species (e.g., G. hirsutum), a second
type of fiber cells (linters) begin growing after about six or seven days. The
distribution of the lint fibers and linters over the seed surface is neither uni-
form nor random. In general, the base of the seed produces more lint fibers,
but the cells near the tip of the seed mostly produce linters. For other cotton
species (e.g., G. barbadense), linters are not formed.
Figure 10.1. Photographs of (A) cotton flower, (B) cotton boll, and (C) open cotton boll.
(Provided by Jeffery C. Silvertooth, The University of Arizona.)
Figure 10.2. SEM image of cotton seed surface with the initiation of lint fiber cells on the
day of anthesis. (Hsieh, Y.L., in Cotton: Science and Technology, Editors Gordon, S., et al.,
Woodhead Publishing Limited, 2007.)
Stage 2—Elongation: The elongation of fiber cells begins on the day of

anthesis by spherical expansion above the ovular surface and this process
continues for 15 to 20 days (Figure 10.3). Initially, the cell elongation ori-
ents against the micropyle end of the ovule, but it becomes spiral after two
or three days. The primary cell walls continue to elongate, reaching the final
fiber lengths in 21–35 days post anthesis. The formation rate of cellulose
chains in primary cell walls is around 2 ng/min in this stage. The cellulose
microfibrils formed in the primary wall are randomly oriented. The primary
cell walls are very thin (0.2 to 0.4 μm) and are covered by a waxy layer or
cuticle which provides water resistance to the fiber cells as they grow. The
maximum potential length of fiber cells is controlled genetically by different
cotton genotypes, but environmental conditions during cell elongation can
change the ability of fiber cells to reach their full potential length.
State 3—Secondary Wall Thickening and Maturation: The formation
of secondary walls starts around 15 to 20 days post anthesis, and continues
for 30 to 45 days until the cell walls are 2–6 μm thick. By the time of fiber
mature, more than 90% of the dry weight of the fiber is cellulose, and a lu-
men also is formed at the center of the fiber. The cellulose formation rate for
secondary walls is around 130 nm/min, which is significantly higher than
that of primary walls. In addition, the molecular weight of secondary wall
cellulose is higher than that of primary wall cellulose. In secondary walls,
cellulose microfibrils are subsequently deposited in a helical conformation.
Periodically, the orientation gyre of the helical microfibrils changes and cre-
ates weak points where fiber breakage is likely under tension. The orienta-
tion of microfibrils in secondary walls is an important factor in determining
the fiber strength, and it is regulated by the structural orientation in the
cytoplasm collectively called the cytoskeleton. At the time of fiber mature,
the newly formed cylindrical cotton fibers are fully hydrated and exhibit
high intrinsic mobility and porosity in their structure.
State 4—Desiccation: At the last stage, the cotton boll opens and the fibers
are desiccated. The desiccation of cotton fibers involves the removal of
Figure 10.3. SEM images of cotton seed surface at the beginning of elongation of fiber
cells in one and two days post anthesis. (Hsieh, Y.L., in Cotton: Science and Technology,
editors Gordon, S., et al., Woodhead Publishing Limited, 2007.)
fluids from the lumens and the loss of intermolecular water from the cellu-
lose. The removal of fluids causes the cylindrical fibers to collapse, thereby
leading to the kidney-bean cross-sectional shape and flat, twisted longi-
tudinal structure of dried cotton fibers. The loss of intermolecular water
reduces the intermolecular distance between cellulose chains and promotes
the formation of intermolecular hydrogen bonds. The collapse of cylindrical
fibers and the formation of hydrogen bonds result in irreversible morphol-
ogy changes, such as increased structural heterogeneity, decreased porosity,
and reduced sorption capacity. In addition, the structural changes also lead
to an increase in the molecular strains and this ultimately reduces the chain
mobility. As a result, these irreversible changes determine the final mechani-
cal properties and surface chemistry of cotton fibers.
The desiccation process completes the formation of cotton fibers, which then
can be harvested. Cotton fibers then are separated from the seeds by using a pro-
cess, called ginning. The ginning process, which can be considered as a part of the
harvesting, normally includes conditioning, seed-fiber separation, cleaning, and
packaging. After ginning, the cotton fibers then can be turned into different final
products by using different processing methods.

Manufactured cellulose fibers are the first man-made fibers ever produced start-
ing with Rayon fibers and the viscose process invented in late 1800s. Currently,
manufactured cellulose fibers are produced by either derivative methods or direct
methods. In derivative methods, cellulose polymer chains are chemically modi-
fied to form cellulose derivatives, which then are dissolved and spun into fibers. In
the direct methods, special solvent systems are used without the chemical modifi-
cation of the cellulose chains.
10.2.1 Derivative Methods

Due to the considerable hydrogen bonding, high molecular weight, and complex
crystalline and amorphous structures, cellulose does not melt nor does it dissolve
in common organic solvents. Therefore, the production of manufactured cellu-
lose fibers often starts by derivatizing the cellulose polymer chains by chemical
approaches.
Among various derivative methods, the viscose process, invented by Cross,
Bevan and Beadle in 1892, still is dominating in commercial cellulose fiber pro-
duction with an annual production of more than 2 million tons. Figure 10.4 shows
the flow chart of the viscose process. The viscose process begins with the alkaliza-
tion step, where the cellulose pulp is treated with an 18–20 % solution of NaOH
Figure 10.4. Flow chart of the viscose process.
at room temperature to form alkali cellulose. The following is the main reaction
in the alkalization step:

&HOO 2+1D2+ &HOO 2 1D +2
(10.1)
The resultant alkali cellulose is pressed to remove most of the excess NaOH so-
lution and is shredded to increase the surface area and make the material easier to
process. The shredded alkali cellulose is pre-aged with the presence of oxygen in
the ambient air to reduce and control the cellulose molecular weight. After reach-
ing the desired molecular weight, carbon disulfide (CS2) is added to the shredded
alkali cellulose in a specially designed vessel called a Barette at 20–30°C for 1–3
hours to carry out the xanthation step, which leads to the formation of sodium cel-
lulose xanthate. The following is the main reaction in the xanthation step:
6
&HOO
2 1 D
&6 &HOO 2 & 61D (10.2)
The xanthation process is complex because in addition to Reaction 10.2, many

side reactions occur simultaneously upon the addition of CS2. The following are
some important side reactions:
6
&6+2 +6&2+ +6&26 (10.3)
6
&26+2 +2&2+ +6&2 (10.4)
&21D2+ (10.5)
1D&2+2
6 6
1D2+
+6&6 +6&6+ 1D6&61D+2 (10.6)
6
1D6&61D+62 1D62&6+6 (10.7)
After the xanthation step, the sodium cellulose xanthate crumb is dissolved
into a 5–8 % NaOH solution below 10°C to form a yellow cellulose xanthate
solution, i.e., the so-called viscose. The viscose is ripened (aged) over a period of
several hours to several days to achieve a uniform distribution of xanthate groups
in cellulose. After the ripening, the viscose is extruded through a spinneret into a
spinning bath (coagulation bath), containing H2SO4, Na2SO4, and ZnSO4. In the
coagulation bath, the extrudate is first coagulated and subsequently regenerated
into cellulose fibers. CS2 is recovered and reused. The following is the main reac-
tion during regeneration:
&HOO 2 & 61D+62 &HOO 2+&61D+62 (10.8)
The structure and properties of viscose fibers can be controlled by selectively

modifying the viscos process conditions. Fibers can be made with a symmetrical
cross-section or a variety of cross-sectional shapes. By controlling the spinning
conditions, the fibers can be spun in a form of skin-core structure. By using addi-
tives either in the viscose solution and/or in the coagulation bath, high-wet modu-
lus fibers can be produced.
In addition to the viscose process, other derivative methods exist. For example,
the cuprammonium process (or Bembergy process) uses a so-called cuprammo-
nium solution, which is a mixture of copper hydroxide and aqueous ammonia
solution, to dissolve cellulose. The cellulose chain interacts with the cupram-
monium ions (Cu(NH3)42+) to form a complex (Figure 10.5). The interaction be-
tween the cellulose chain and the cuprammonium ions is not a primary bond, and
hence some researchers consider the cuprammonium process as a direct method.
Cuprammonium decomposes when exposed to light, and hence precaution must
be taken to limit the exposure of the cuprammonium cellulose solution to light.
Typically, the cuprammonium cellulose solution is spun into a coagulation/re-

generating bath using a funnel as a processing aid to protect the delicate as-spun
filaments. The alkaline coagulant liquid travels down the funnel and simultane-
ously coagulate and stretch the extrudate. The resultant fibers are washed with 5%
H2SO4 solution to remove any remnant copper or other chemical deposits.
Another important derivate method is the carbamate process, which uses urea
instead of CS2 or cuprammonium as the solvent. The product of the reaction be-
tween cellulose pulp and urea is cellulose carbamate, which easily can be dis-
solved in aqueous NaOH solution and spun into a coagulation bath to be regener-
ated into cellulose fibers.
One common feature of the viscose, cuprammonium, and carbamate processes
is the chemically modified cellulose is regenerated into cellulose after the extru-
sion. Fibers made from these processes often are called rayon fibers. However,
useful fibers can be produced by derivative methods without the regeneration of
cellulose. Two important examples are cellulose acetate and cellulose triacetate
fibers. Cellulose acetate can be obtained by acetylation of cellulose with acetic
acid and acetic anhydride with sulfuric acid as a catalyst. Cellulose triacetate,
which is partly saponified to get the desired degree of substitution, is produced
by a similar process. Both cellulose acetate and cellulose triacetate keep their de-
rivative structure in final fibers. The major difference between these two fibers is
that in cellulose acetate fibers, less than 92% but at least 74% of the –OH groups
are acetylated, but in triacetate fibers, at least 92% of the hydroxyl groups are
acetylated.
10.2.2 Direct Methods

Although cellulose is not soluble in common organic solvents, a series of special
solvents have been found to be able to dissolve cellulose. However, only some of
them can be used to spin cellulose fibers via direct methods. So far, all these viable
direct solvents consist of two components, e.g., N-methylmorpholine-N-oxide/
water (NMMO/H2O), lithium chloride/dimethylacetamide, trifluoroacetic acid/di-
chloroethane, calcium thiocyanate/water, ammonia/ammonium thiocyanate, zinc
chloride/water, and sodium hydroxide/water.
Figure 10.5. Interaction between cellulose and cuprammonium.

The most important direct method is the NMMO or Lyocell process, in which
the NMMO/H2O mixture is used at elevated temperatures as the solvent. In the
NMMP process, cellulose pulp is first added to the NMMP mixture, which then
is stirred and heated to obtain a homogeneous cellulose solution. N-propyl gallate
also is added as an anti-oxidant to stabilize the molecular weight of the cellulose.
The cellulose solution is extruded through the spinneret into water to obtain cel-
lulose filaments. NMMP is recovered from the bath and reused. Cellulose fibers
made by the NMMO process typically have high crystallinity, high molecular
orientation, and high tenacity. However, the NMMP process has not replaced the
viscose process in the industry partially because of the fibrillation problem of
fibers made by this direct method.

Natural protein fibers include animal hair fibers and extruded protein fibers. This
section discusses the formation of two most important natural protein fibers: wool
(by sheep) and silk (by silkworm).
10.3.1 Formation of Wool

Wool fibers are sheep hairs that grow from cavities (follicles) extended from the
surface of the sheep skin through the epidermis into the dermis (Figure 10.6).
One of the most important functions of the hair follicle is to produce a hair shaft,
i.e., the fiber. Follicle formation occurs only once in a sheep’s lifetime. At birth,
the entire follicle population is initiated and normally does not change thereafter.
Although the total number of follicles does not change during the lifetime of
an individual sheep, the hair follicle actually is a regenerating system, manifested
by the so-called follicle cycle. A complete follicle cycle consists of four phases:
growth (anagen), regression (catagen), rest (telogen), and shedding (exogen).
After the shedding phase, the next anagen starts. Hair follicle cycling is a devel-
oping process, which occurs over the entire lifetime of a sheep. However, due to
the forcefully selection of long fiber length, wool-producing sheep does not ac-
tively cycle through all the four phases, but experiences an extremely long anagen
period so that high-quality wool fibers can be obtained for practical applications.
Figure 10.6 shows a schematic of a follicle with the major morphological com-
ponents of the developing fiber. A pear-shaped dermal papilla is located at the bot-
tom of the hair follicle. The dermal papilla plays an important role in inducing the
formation of the hair follicle and it also retains this instructive ability throughout
the life of the hair follicle. During anagen, the outer cells of the dermal papilla
deliver a signal to the surrounding epithelial cells and stimulates them to divide
rapidly, thereby leading to the formation of the hair matrix, also known as the
follicle bulb. The follicle bulb cells continue to divide for the life of the follicle
Figure 10.6. Schematic of a follicle showing the major morphological components of the

developing fiber.
and provide daughter cells that differentiate to form inner root sheath cells and
fiber cells. The inner root sheath consists of three concentric layers: Henle’s layer,
Huxley’s layer, and the inner root sheath cuticle. The Henle’s layer is adjacent to
the outer root sheath, which separates the entire structure from the dermis. The
inner root sheath cuticle provides the support to the maturing fiber until it reaches
the level of sebaceous gland, where the inner root sheath cells die and shed from
the surface of the epidermis.
Surrounded by the inner root sheath, the fiber cells multiply and the constant
stream of cell production pushes the fiber cells upward toward the skin surface. As
the fiber cells move up the hair follicle, they begin to differentiate into particular
cell types: cuticle and cortex. In addition to cuticle and cortex cells, the follicle
also produces a central strand of cells that are loosely organized, forming the
medulla in the center of the hair fiber. Eventually, the tip of the fiber penetrates
the superficial layers of the epidermis and the mature fiber emerges from the skin
surface. In the mature fiber, cells are keratinized and hardened.
Although the keratinized cells in a mature fiber already are dead, the fiber
length continues to increase due to the formation of new fiber cells in the follicle.
Mature fibers can be sheared after reaching certain length. In most parts of the
world, mature fibers can be sheared once a year, in spring or summer. Sheared
fibers then can be sorted and graded for use.
10.3.2 Formation of Silk

Silk is a protein fiber spun by silkworm, which is the caterpillar or larva of the
domesticated silkmoth, Bombyx mori. In addition to silkworm, many other insects
or non-insects (e.g., spider) can spin silk fibers. However, the silk of Bombyx mori
is the mainstay of commercial silk production and consumption. This section will
only discuss the formation of silk fiber by silkworm.
Each silkworm has two tubular spinning glands. Each silkworm can use these
two glands to extrude two single silk fibers, which are joined together to form one
single thread by muscular reaction. Figure 10.7 shows the schematic of two spin-
ning glands, which join together at one end. The main components of the spinning
glands are: secretory portion, reservoir, duct, and spinneret. Each gland has its
own secretory portion and reservoir, but two glands share one duct and spinneret.
The secretory portion is by far the longest and is believed to be responsible for
the secretion of the silk protein, fibroin. The reservoir is about one-fourth of the
length of the secretory portion, but it has a significantly larger diameter and is as-
sociated with the storage of the spinning solution. It also is believed the second
silk protein, sericin, is synthesized in the reservoir. The short and narrow duct
opens to the outside through the spinneret, which is located on the silkworm’s lip.
During the formation of silk, fibroin is synthesized by epithelial cells in the
secretory portion of the gland and then stored in a 20–30% aqueous solution in
the reservoir. At the same time, the reservoir also synthesizes sericin as a sepa-
rate layer around the fibroin. The two proteins in the aqueous solution proceed
together without mixing into the duct, where the shear rate increases from 2 to
400 s–1. The significant increase in shear rate is made possible by the formation of
a liquid crystalline intermediate phase of the silk proteins in the spinning solution.
At the spinneret, the two fibroin cores are extruded together, surrounded by a layer
of sericin, forming one thread. Upon exiting the spinneret, the silk proteins are
Figure 10.7. Schematic of two silk spinning glands that join just before the duct. (Willcox,
P.J., et. al., Macromolecules, 29, 5106–5110, 1996.)
converted from a liquid crystalline intermediate phase into semicrystalline fibers,

with a high degree of molecular orientation.
The silkworm spins the silk in a figure-8 pattern around itself to form a com-
plete cocoon. The cocoon then can be immersed in hot water to loosen the fibers,
which are pulled from the water for use.

Proteins are more complex than cellulose and it is more difficult to produce manu-
factured protein fibers. So far, there are totally three generations of manufactured
protein fibers. The first generation was developed in the late nineteenth and early
twentieth centuries, the second in the mid-twentieth century, and the third in the
late twentieth and early twenty-first centuries. The production methods for the
first and second generations of manufactured protein fibers were similar; how-
ever, the methods used for the third generation changed significantly. Therefore,
this section discusses the first and second generation protein fibers together, but
the third generation separately.
10.4.1 First and Secondary Generations

The first generation of manufactured protein fibers was developed in the late
nineteenth and early twentieth centuries. These manufactured protein fibers were
intended as substitutes for silk, but most of them were never commercially pro-
duced. The second generation of manufactured protein fibers was developed and
produced in the mid-twentieth century. Many of these manufactured fibers ini-
tially were developed as substitutes for wool or silk.
A wide range of animal and vegetable protein sources, such as milk casein,
egg white, chicken feathers, fish albumen, collagen, gelatin, hair, hooves, horn,
soy beans, corn zein, peanuts, sunflower seeds, hemp seeds, castor oil seeds, were
explored as the protein sources for manufactured protein fibers. Many production
methods also were developed for converting these proteins into fibers, and most
of them at least involved four steps: separation of protein, solubilization of pro-
tein, fiber spinning, and insolubilization of protein.
Step 1—Protein Separation: In general, proteins are not in a readily acces-

sible form. For example, many protein sources, such as milk, peanuts and
seeds, contain a large amount of oils and fats. Therefore, proteins first were
separated from the source material into a suitable form for fiber formation.
In some cases, the proteins also were dried and grounded into granules.
Step 2—Solubilization: An appropriate solubilization process is critically
important for obtaining a useful spinning solution. For some proteins, such
as gelatin, they were dissolved directly in either hot water or glacial acetic
Manufactured Protein Fibers 199
acid after being obtained from slaughterhouse waste. However, if the pro-
tein was dried and grounded into granules, a stronger solvent was used. The
obtained spinning solutions often were aged to obtain the desired viscos-
ity. During that process, the oxidation and bacterial activity were carefully
controlled.
Step 3—Spinning: The spinning solution was extruded through a spinneret
into a coagulation bath to form filaments. The coagulation bath typically
contained salts and acids, such as sodium sulfate, aluminum sulfate, magne-
sium sulfate, and sulfuric acid. With the presence of salts and acids, the fila-
ment surface coagulated first and this minimized the tendency of filaments
to clump together.
Step 4—Insolubilization: The as-spun protein filaments were often soluble.
Hence, an insolubilization step was introduced to form networks between
protein chains. The degree of cross-linking was important. When the cross-
linking degree was too low, the filaments would not have sufficient wet
strength to be made into useful products. When the cross-linking degree was
too high, the filaments would be brittle. The insolubilization of protein often
was carried out in a liquid bath so the filaments could be drawn at the same
time to improve the molecular orientation.
Many manufactured protein fibers were commercialized in mid-twentieth cen-

tury. However, due to technical and economic problems, manufactured protein
fibers were not able to compete with either natural fibers or the newly develop-
ment synthetic fibers at that time. All manufactured protein fibers developed in
mid-twentieth century eventually were abandoned by manufacturers.
10.4.2 Third Generation

In the late twentieth and early twenty-first centuries, manufactured protein fibers
attracted attention again due to the demand for fibers with a reduced ecological
footprint. Many manufactured protein fibers are being developed with a primary
emphasis on imitating the structure and properties of silk.
For this third generation of manufactured protein fibers, milk and soy beans are
the two most important protein sources. Chicken feathers also are being widely
used in the development. In addition to the common methods used in the first and
second generations, many technical innovations have been applied to the produc-
tion of these manufactured protein fibers. Examples include the use of biochemis-
try to modify the protein structure, the incorporation of synthetic polymers to im-
prove the fiber strength and modulus, the formation of protein-based copolymers
by chemical grafting, etc.
REFERENCES
[1] Adelson, D.L., Cam, G.R., DeSilva, U., and Franklin, I.R., “Gene Expression
in Sheep Skin and wool (Hair)”, Genomics, 83, 95–105, 2004.
[2] Anthony, W.S., “The Harvesting and Ginning of Cotton”, in Cotton: Science
and Technology, editors Gordon S., and Hsieh, Y.L., Woodhead Publishing
Limitted, 2007.
[3] Asakura, Yao, J., Yang, M., Zhu, Z., and Hirose, H., “Structure of the
Spinning Apparatus of a Wild Silkworm Samia Cynthia Ricini and
Molecular Dynamics Calculation on the Structural Change of the Silk
Fibroin”, Polymer, 48, 2064–2070, 2007.
[4] Brooks, M.M., “Regenerated Protein Fibres: A Preliminary Review”, in
[5] Bond, J.J., Wynn, P.C., Brown, G.N., and Moore, G.P.M., “Growth of Wool
Follicles in Culture”, In Vitro Cellular & Developmental Biology, 30A,
90–98, 1994.
[6] Cappello, J., and McGrath, K.P., “Spinning of protein Polymer Fibers”, in
Silk Polymers, editors Kaplan, D., Adams, W.W., Farmer, B., and Viney, C.,
ACS Symposium Series, American Chemical Society, 1993.
[7] Ciechanska, D., Wesolowska, E., and Wawro, D., “An Introduction to
Cellulosic Fibres”, in Handbook of Textile Fibre Structure, Volume II:
Natural, Regenerated, Inorganic and Specialist Fibres, editors Eichhorn,
S.J., Hearle, J.W.S., Jaffe, M., and Kikutani, T., Woodhead Publishing
Limited, 2009.
[8] Cuculo, J.A., Aminuddin, N., and Frey, M.W., “Solvent Spun Cellulose
Fibers”, in Structure Formation in Polymeric Fibers, editor Salem, D.R.,
Hanser Gardner Publications, 2000.
[9] Ganster, J., and Fink, H.P., “The Structure of Man-Made Cellulosic Fibres”,
[10] Ho, C.P., Shen, S.M., Tang, P.S., and Yu, S.H., “Physiology of the Silkworm.
II. Mechanism of Silk Formation as revealed by X-Ray Analysis of the
Contents of the Silk Gland in Bombyx Mori”, Physiological Zoology, 17,
78–82, 1944.
References 201

Publishing Limitted, 2007.
[12] Kerkam, K., Yiney, C., Kaplan, D., and Lombardi, S., “Liquid Crystallinity
of Natural Silk Secretions”, Nature, 349, 596–598, 1991.
[13] Kudlicka, K., Brown, R.M., Li, L., Lee, J.H., Shin, H., and Kuga, S.,
“β-Glucan Synthesis in the Cotton Fiber”, Plant Physiology, 107, 111–123,
1995.
[14] Lawrence, C., “The Opening, Blending, Cleaning and Carding of Cotton”,
in Cotton: Science and Technology, editors Gordon, S., and Hsieh, Y.L.,
Woodhead Publishing Limitted, 2007.
[15] Li, G., and Yu, T., “Investigation of the Liquid-Crystal State in Silk Fibroin”,
Makromolekulare Chemie. Rapid communications, 10, 387–389, 1989.
[16] Li, G., Zhou, P., Shao, Z., Xie, X., Chen, X., Wang, H., Chunyu, L., and
Yu, T., “The Natural Silk Spinning Process – A Nucleation-Dependent
Aggregation Mechanism”, European Journal of Biochemistry, 268, 6600–
6606, 2001.
[17] Magoshi, J., Magoshi, Y., and Nakamura, S., “Mechanism of Fiber Formation
of Silkworm”, in Silk Polymers, editors Kaplan, D., Adams, W.W., Farmer,
B., and Viney, C., ACS Symposium Series, American Chemical Society,
1993.
[18] Meinert, M.C., and Delmer, D.P., “Changes in Biochemical Composition of
the Cell Wall of the Cotton Fiber during Development”, Plant Physiology,
59, 1088–1097, 1977.
[19] Mercer, E.H., “Formation of Silk Fibre by the Silkworm”, Nature, 168,
792–793, 1951.
[20] Moriya, M., Ohgo, K., Masubuchi, Y., and Asakura, “Flow Analysis
of Aqueous Solution of Silk Fibroin in the Spinneret of Bombyx Mori
Silkworm by Combination of Viscosity Measurement and Finite Element
Method Calculation”, Polymer, 49, 952–956, 2008.
[21] Naithani, S.C., Rao, N.R., and Singh, Y.D., “Physiological and Biochemical
Changes Associated with Cotton Fibre Development. I. Growth Kinetics
and Auxin Content”, Physiologia Plantarum, 54, 225–229, 1982.
[22] Orford, S., Delaney, S., and Timmis, J., “The Genetic Modification of
Cotton”, in Cotton: Science and Technology, editors Gordon S., and Hsieh,
Y.L., Woodhead Publishing Limitted, 2007.
[23] Riffer, R., and Broido, A., “Asymmetric Distribution in the Biosynthesis of
Cotton Cellulose-[U-14C]”, Journal of Experimental Botany, 25, 216–218,
1974.
[24] Silva-Zacarin, E.C.M., Silva De Moraes, R.L.M., and Taboga, S.R., “Silk
Formation Mechanisms in the Iarval Salivary Glands of Apis Mellifera
(Hymenoptera: Apidae)”, Journal of Biosciences, 28, 753–764, 2003.
[25] Stenn, K.S., and Paus, R., “Controls of Hair Follicle Cycling”, Physiological
Reviews, 81, 449–488, 2001.
[26] Tsuchida, K., Jouni, Z.E., Gardetto, J., Kobayashi, Y., Tabunoki, H., Azuma,
M., Sugiyama, H., Takada, N., Maekawa, H., Banno, Y., Fujii, H., Iwano,
H., and Wells, M.A., “Characterization of the Carotenoid-Binding Protein
of the Y-gene Dominant Mutants of Bombyx Mori”, Journal of Insect
Physiology, 50, 363–372, 2004.
[27] Vollrath, F., Porter, D., and Dicko, C., “The Structure of Silk”, in Handbook
of Textile Fibre Structure, Volume II: Natural, Regenerated, Inorganic and
Specialist Fibres, editors Eichhorn, S.J., Hearle, J.W.S., Jaffe, M., and
[28] Wakelyn, P.J., and Chaudhry, M.R., “Organic Cotton”, in Cotton: Science
and Technology, editors Gordon S., and Hsieh, Y.L., Woodhead Publishing
Limitted, 2007.
[30] Willcox, P.J., Gido, S.P., Muller, W., and Kaplan, D.L., “Evidence of a
Cholesteric Liquid Crystalline Phase in natural Silk Spinning Processes”,
Macromolecules, 29, 5106–5110, 1996.
[31] Worktmann., F.J., “the Structure and Properties of Wool and Hair Fibres”,
Problems 203
PROBLEMS
(1) What are the four formation and growth stages of cotton fibers?
(2) Describe derivative and direct methods for making manufactured cellulose
fibers.
(3) Describe the formation process of wool fibers.
(4) Describe the formation process of silk fibers.
(5) What are the major steps that are needed for making manufactured protein
fibers?
CHAPTER 11
Formation of Non-Polymer Fibers
Many different methods can be used to produce non-polymer fibers. For example,
carbon fibers are made by high temperature treatment of carbon precursors in an
inert atmosphere. Glass fibers and some ceramic fibers can be directly spun from
their melts. Ceramic fibers also can be obtained by the calcination of ceramic pre-
cursor fibers or by the chemical vapor deposition of precursor gas on a carbon fi-
ber substrate. This chapter discusses the formation of two important non-polymer
fibers: carbon and glass fibers.
11.1 CARBON FIBERS

Carbon fibers typically are produced by the high temperature treatment of struc-
turally stable precursor fibers that have been thermally, mechanically, and/or
chemically pre-treated. Based on the precursor type, carbon fibers can be divided
into four major classes, namely (i) polyacrylonitrile (PAN)-based, (ii) mesophase
pitch-based, (iii) cellulose-based, and (iv) vapor grown carbon fibers.
11.1.1 PAN Process

PAN currently is the most used precursor for carbon fibers. Figure 11.1 shows the
main processes for producing carbon fibers by using the PAN route. PAN can be
synthesized from acrylonitrile through a radical polymerization process. During
the synthesis of PAN, co-monomers, such as methyl acrylate (up to 5%) can be
added to improve the processability of PAN-based precursor fibers. The addition
of co-monomers also can improve the mechanical properties of the final carbon
fibers by increasing the molecular orientation.
205
206 Formation of Non-Polymer Fibers
Figure 11.1. Processes to produce carbon fibers by using the PAN route.
PAN precursor fibers often are obtained by solution spinning methods, such as
wet spinning and dry spinning. Melt spinning of PAN fibers is not commonly used
in commercial production. This is because PAN-based copolymers contain strong
polar nitrile groups and experience thermally induced cyclization reaction before
melting, which make the melt spinning of PAN impossible unless a large amount
of solvent additives are added to lower the melting temperature. In addition to
traditional wet spinning and dry spinning, the so-called dry-wet spinning with an
air gap between the spinneret and the spinning bath also can be used to produce
PAN precursor fibers with high molecular orientation.
The high temperature treatment of PAN precursor fibers starts with a stabi-
lization process between 200 and 260°C. Stabilization can be carried out in the
absence of oxygen, but oxygen (from air) typically is used to promote inter- and
intramolecular reactions that are important for improving the carbon fiber quality.
The main purpose of the stabilization process is to convert the PAN chains into a
heteroaromatic structure that is non-meltable and flame-resistant. The stabiliza-
tion process is complex, in terms of chemical reactions. Many different structures
are formed during the stabilization process. For example, a cyclized and dehy-
drated structure is formed by cyclization and dehydrogenation reactions, and it is
widely accepted as the product of low temperature stabilization (Figure 11.2A).
An azomethine cross-link structure also may be formed at the same time dur-
ing stabilization (Figure 11.2B). With the presence of oxygen from air, oxidized
structures can be formed (Figures 11.2C, D, and E). The abovementioned struc-
tures can be combined in oxidized chains (Figures 11.2F, G, and H). In structures
Carbon Fibers 207
shown in Figures 11.2G and H, non-reacted nitrile (-CN) groups still are present
due to the random initiation sites on the PAN chains.
The stabilization process introduces significant volume reduction to the pre-
cursor fibers. The longitudinal shrinkage is resisted mechanically, but the diam-
eter of the fibers is allowed to reduce. During that process, the orientation of PAN
chains is largely preserved by the strong intermolecular interactions and by the
newly-formed rigid ladder chain structure.
Figure 11.2. Possible chemical reactions and the resultant structures for the stabilization
of PAN precursor fibers.
After the stabilization process, the precursor fibers are structurally stable and
non-meltable, and the atmosphere for high temperature treatment then is changed
from air to an inert medium, such as nitrogen or argon, to carry out the carbon-
ization process. During carbonization, the cyclized and dehydrated precursor
crosslinks in the low temperature range, and then condensation reactions follow
up to 1700°C. These condensation reactions enable the removal of non-carbon
elements from the fibers, leading to the formation a “turbostratic” carbon phase
structure. This carbon phase is oriented along the fiber axis, but still has many tet-
rahedral carbon-type crosslinks between the graphite-type carbon layers. Carbon
fibers with this type of structure have high tensile strength. The turbostratic car-
bon structure has been discussed in Chapter 6. Figure 11.3 shows two examples of
the condensation reactions occurred in the carbonization process.
The treatment temperature can be increased to above 1700°C, up to 3000°C.
The process above 1700°C is called graphitization. While the carbonization pro-
cess can be carried out in nitrogen, graphitization must be conducted in argon
because carbon reacts with nitrogen at such high temperatures to form cyanogen.
The graphitization process heals some of the “defects” in the turbostratic carbon
structure and leads to the formation of graphitic structures. The resultant fibers
are called graphite fibers, which have higher modulus than regular carbon fibers
prepared at temperatures lower than 1700°C.
11.1.2 Mesophase Pitch Process

Pitches can be obtained as a byproduct from petroleum or coal industry.
Figure 11.4 shows a typical pitch structure. Fibers made from untreated pitches
are isotropic and exhibit relatively poor mechanical properties. To make useful
carbon precursor fibers, pitches typically are pre-treated between 350–500°C in
an inert atmosphere to develop a continuous anisotropic phase (similar to a me-
sophasic liquid crystal) or a mixture of anisotropic and isotropic phases that can
Figure 11.3. Examples of chemical reactions during the carbonization process.

Carbon Fibers 209
Figure 11.4. Typical pitch structure.
turn into an oriented structure during spinning. The resultant mesophase pitches
then are converted to carbon precursor fibers by melt spinning. During spinning,
the shear stress developed in the spinneret facilitates the orientation of the disc-
shaped molecules, which then is improved by post-spinning drawing.
The stabilization process of mesophase pitch fibers is carried out between 250–
350°C in air. During stabilization, pitch molecules form crosslinks and the precur-
sor fibers become non-meltable, ready for carbonization. The stabilized precursor
fibers then are carefully pre-carbonized between 700–900°C to remove volatiles
without the formation of voids. The pre-carbonized fibers are then carbonized at
1500–2000°C to form carbon fibers. To improve the fiber modulus, a graphitiza-
tion process up to 3000°C can be applied to obtain graphite fibers.
11.1.3 Cellulose Process

Many cellulose fibers, including both natural and manufactured cellulose fibers,
can be made into carbon fibers by high temperature treatment. However, carbon
fibers made from natural cellulose fibers, such as cotton and ramie, often have
high defect contents, which limit the strength of the fibers. Currently, rayon, a
manufactured cellulose fiber, is the most used cellulose-based carbon precursor.
The formation of rayon fibers has been discussed in Chapter 10. Rayon fi-
bers have a large number of hydroxyl groups and absorb water molecules easily.
Therefore, before the carbonization process, rayon fibers are first heat-treated be-
tween 25 and 150°C to carry out the physical desorption of water. The conversion
of dried rayon fibers into carbon fibers typically involves a stabilization process
between 300 and 400°C in air and a carbonization process up to 1500°C in inert
atmosphere. Graphite fibers can be obtained by increasing the temperature up to
3000°C in inert atmosphere.
11.1.4 Hydrocarbon Gas Process

Carbon fibers also can be produced by the pyrolytic deposition of hydrocarbon
gases. Many hydrocarbon gases, such as methane, naphthalene, and benzene,
have been used to produce carbon fibers with deposition temperatures of 1000–
1200°C. During the pyrolysis process, thin tubes of carbon are first formed on
ultra-fine particles. The tubes then grow by a surface diffusion mechanism, and
the subsequent high temperature treatment with a temperature up to 2500°C re-
sults in the formation of carbon fibers with diameters ranging from 10 nm to more
than 100 μm. Carbon fibers produced from hydrocarbon gases often have central
hollow cores.
11.2 GLASS FIBERS

Raw materials used to make glass fibers include but are not limited to sand, lime-
stone, soda ash, and cullet. These raw materials are melted in a furnace to obtain
molten glass. The molten glass then can be made into glass fibers, mainly by three
methods: wool process, continuous filament process, and marble process.
11.2.1 Wool Process

The wool process mainly is used to produce discontinued (short) glass fibers for
thermal insulation and filtration applications. These fibers typically are used in the
form of batts, blankets or boards, and they often are called fiberglass or glass wool.
Several methods can be used to produce glass wool. Currently, the most popu-
lar method is centrifugal spinning, also called rotary spinning. Figure 11.5 shows
the schematic of a typical centrifugal spinning process. A stream of molten glass
is fed into a spinner, which basically is a metal container with several thousand
holes uniformly distributed around on the sidewall. To keep the glass at its liquid
state, the temperature in the spinner typically is maintained at 900–1100°C. The
spinner rotates at a high speed, ranging from 200 rpm to more than 300 rpm. The
centrifugal force created at the high rotating speed causes molten glass to flow
through the small holes in the spinner sidewall to form fine streams of glass. The
glass streams coming out of the side wall are attenuated by high velocity air (or
other gas or steam) and are broken into fine (< 10 μm) glass fibers, which are
several centimeters in length. A binder is sprayed on the glass fibers when they
are formed. Broken glass fibers move through a “forming hood” placed under the
spinner, and they subsequently are collected on a conveyor belt in the form of a
mat. The function of the forming hood is to distribute the fibers evenly to form
a random alignment across the width and length of the collected glass fiber mat.
The conveyor belt delivers the glass fiber mat through a curing oven which dries
and cures the binder. The glass fiber mat then can be cut in desired lengths and
widths for different uses.
Glass Fibers 211
Figure 11.5. Schematic of a typical glass wool process. Forming hood is not shown.
11.2.2 Continuous Filament Process

Glass fibers can be produced by the continuous filament process with fiber lengths
up to 10 kilometers. These continuous glass fibers typically are used for reinforce-
ment purposes. Glass fibers produced with the continuous filament process have
nearly uniform diameters, ranging from 5–30 μm, which typically are larger than
those (< 10 μm) of discontinued glass fibers produced by the wool process.
Figure 11.6 shows the schematic of a typical continuous glass filament process.
Continuous glass filaments are produced by using a direct-melt bushing, i.e., spin-
neret, which contains a few hundred to several thousand small holes in a flat plate
located in the bottom. The bushing typically is made from a platinum-rhodium
alloy that can sustain a very high temperature. The temperature in the bushing
typically is in the range of 1150–1300°C, depending on the composition of the
glass. The molten glass flows out of the bushing under gravity to form filaments
with an initial diameter of around 1 mm at a velocity of a few millimeters per
second. The final diameter of the glass filaments is a function of the tension ap-
plied to the filaments as they are drawn. The linear velocity of the filaments at the
winder ranges from a few meters per second to more than 30 meters per second.
The huge velocity increase creates a tension that reduces the filament diameter
from around 1 mm to 5–30 μm.
On the spinline, a mixture of air and water (i.e., a fine mist) often is sprayed to
the newly formed filaments just below the bushing to help remove the heat from
the vicinity of the bushing and cool the individual filaments. This often leads to
a quick thermal quench rate, ranging from 500,000 to 1,000,000 K s–1. Such a
high quench rate makes the resultant glass fibers to exhibit relative low densities.
In addition to the mixture of air and water, a sizing applicator also introduces a
mixture of film-forming polymer, lubricant, etc., to provide filament cohesion and
protect the glass surface from abrasion.
In industry operations, multiple bushings are placed in parallel immediately
under a forehearth, which is connected to a large furnace (Figure 11.7). The fur-
nace and forehearth continuously melt, refine, and homogenize the raw materials
and deliver the resultant molten glass into the bushings for fiber formation. This
large-scale production is considered as an energy effective process, as compared
with the marble process discussed below.
11.2.3 Marble Process

Continuous glass fibers also can be produced by re-melting and spinning glass
marbles. Figure 11.8 shows the schematic of the marble making process. Raw
Figure 11.6. Schematic of a typical continuous glass filament process.
Figure 11.7. Schematic of the continuous glass filament process in a typical industrial operation.
Glass Fibers 213
materials are fed into a furnace to be melted to form molten glass, which flows
to the forehearth. The still-molten glass is released from the forehearth and the
resultant glass stream is cut into individual “gobs” by an automatic cutting de-
vice. The gobs travel down a metal ramp to grooved rollers that roll the gobs into
sphere-shaped marbles.
The marbles are cooled and shipped to the fiber manufacturing facility, where
they are re-melted in specially designed bushings and drawn into continuous fila-
ments. The fiber formation equipment is similar to that used in the continuous
filament process described in the previous section. The major difference is the de-
sign of bushing. In the continuous filament process, the bushing receives molten
glass from the forehearth and ejects glass filaments directly. However, the bush-
ing in the marble process receives glass marbles that are in solid state and need
to be re-melted. Therefore, the marble busing has more functions, e.g., melting
the marbles and conditioning the glass to the correct fiber-formation temperature.
Figure 11.9 shows the schematic of a typical marble bushing. A perforated plate is
located within the bushing to retain the glass marbles until they are melted.
In the marble process, the glass first must be melted for making marbles and
then is re-melted in the bushing for fiber formation. Due to the re-melting process,
Figure 11.8. Schematic of the marble making process.
Figure 11.9. Schematic of a typical bushing used in the marble process.

the energy used in the marble process is high, and hence most commercial contin-
uous glass fibers are manufactured by using the more energy-effective continuous
filament process. Currently, the marble process is only used for making specialty
glass fibers at smaller scales. One advantage of re-melting glass marbles in a
smaller scale is the filaments exiting the bushing can be quenched rapidly without
the need for mist spray.
REFERENCES
[1] Bahl, O.P., Shen, Z., Lavin, J.G., and Ross, R.A., “Manufacture of Carbon
Fibers”, in Carbon Fibers, Third Edition, Editors Donnet, J.B., Wang, T.K.,
Rebouillat, S., and Peng, J.C.M., Marcel Dekker, 1998.
[2] Cato, A.D., and Edie, D.D., “Flow Behavior of Mesophase Pitch”, Carbon,
41, 1411–1417, 2003.
[3] Chand, S., “Review Carbon Fibers for Composites”, Journal of Materials
Science, 35, 1303–1313 (2000).
[4] Donnet, J.B., and Bansal, R.C., Carbon Fibers, Marcel Dekker, 1984.
[5] Frank, E., Hermanutz, F., and Buchmeiser, M.R., “Carbon Fibers:
Precursors, Manufacturing, and Properties”, Macromolecular Materials
and Engineering, 297, 493–501, 2012.
[7] Johnson, D.J., “Carbon Fibers”, in Structure Formation in Polymeric
Fibers, Editor Salem, D.R., Hanser Gardner, 2000.
[9] Lafdi, K., and Wright, M.A., “Carbon Fibers”, in Handbook of Composites,
Editor Peters, S.T., Chapman & Hall, 1998.
[10] Loewenstein, K.L., “The Manufacture of Continuous Glass Fibres—Present
Trends in the Use of Platinum Alloys”, Platinum Metals Review, 19, 82–87,
1975.
[11] Lowenstein, K.L., The Manufacturing Technology of Continuous Glass
Fibres, Elsevier, 1973.
[12] Matsumoto, T. “Mesophase Pitch and its Carbon Fibers”, Pure and Applied
Chemistry, 57, 1553–1562, 1985.
Problems 215
[13] Morgan, P., Carbon Fibers and Their Composites, CRC Press, 2005.
[15] Sharma, S.P., and Lakkad, S.C., “Morphology Study of Carbon Nanospecies
Grown on Carbon Fibers by Thermal CVD Technique”, Surface and
Coatings Technology, 203, 1329–1335, 2009.
PROBLEMS
(1) What are the four most important precursors that can be used for making
carbon fibers?
(2) What the main steps for making carbon fibers from polyacrylonitrile? What
are the main reactions in each step?
(3) What are the three major processes for making glass fibers?
(4) Describe the wool process of glass fibers.
(5) Describe the continuous filament process of glass fibers.
CHAPTER 12
Formation of Nanofibers by Electrospinning
Electrospinning is the most reported method for producing nanofibers. A litera-

ture search using the Web of ScienceTM database shows that Year 2012 alone had
publication of nearly 10,000 journal articles in the electrospinning of nanofibers.
Among these publications, more than 50% focused on the investigation of the
electrospinning process and the characterization of the resultant nanofibers, and
the others mainly address the innovative use of electrospun nanofibers in vari-
ous applications. Chapter 7 has discussed the main structural characteristics of
electrospun nanofibers. This chapter focuses on the fiber formation process and
processing-structure relationships of electrospinning.
12.1 ELECTROSPINNING SYSTEMS

The simplest electrospinning system is based on a single nozzle design
(Figure 12.1). A typical single-nozzle electrospinning system at least contains
a syringe, a metal nozzle, a power supply, and a collector. During electrospin-
ning, the electrically charged fluid forms a conical droplet (i.e., Taylor cone),
from which a liquid jet is generated and is elongated. Nanofibers are deposited on
the collector or a substrate placed on the collector. The single-nozzle electrospin-
ning system is easy to build and is widely used in academic research. However,
the production rate of single-nozzle system is slow, and a single nozzle shown in
Figure 12.1 can only produce 1–100 mg nanofibers per hour. The production rate
can be improved by connecting a large number of nozzles in parallel. Companies,
such as MECC (Japan), Fuence (Japan), Toptec (Korea), Yflow (Spain), have de-
veloped and commercialized large-scale electrospinning machines based on vari-
ous multi-nozzle designs.
217
218 Formation of Nanofibers by Electrospinning
Figure 12.1. Schematic of a single-nozzle electrospinning system.
Nozzle-less electrospinning systems also have been developed for mass-

production of nanofibers. Figure 12.2 shows a basic nozzle-less electrospinning
system. The spinning fluid is placed in an open bath. To form liquid jets, a cylin-
drical electrode is rotated in the bath to be wetted by the spinning fluid. When a
high voltage is applied to the fluid, multiple jets are ejected from the electrode
surface in an “up-spinning” fashion. The resultant nanofibers are deposited onto
the grounded collector or a substrate that is moved continuously at a fixed speed
passing under the collector. The production rate of the nozzle-less electrospin-
ning system is relatively high since multiple jets are generated simultaneously
on the electrode surface. In theory, the production rate can be improved by sim-
ply increasing the length and the number of the spinning electrodes. The most
well-known nozzle-less electrospinning system is the NanospiderTM developed by
Elmarco (Czech Republic).
In the nozzle-less electrospinning system, the structure of the spinning elec-
trode is important. To facilitate the generation of liquid jets, different types of
spinning electrodes have been developed. Figure 12.3 compares two electrode
designs: smooth cylinder type and patterned wire type. The smooth-cylinder elec-
trode allows the generation of multiple jets from a thin layer of liquid fluid coated
on the smooth surface. The patterned-wire electrode generates jets from the fluid
drips carried by the specially-designed pattern on the thin wires. The selection of
electrodes during nozzle-less electrospinning is mainly dependent on the proper-
ties of the spinning fluid.
Although the design of the nozzle-less electrospinning system is different from
that of nozzle-based system, both utilize the electrostatic force to generate liquid
jets, which elongate to reduce the jet diameter and eventually form nanofibers.
The discussion in the following sections will be based on the nozzle-based elec-
trospinning system. However, the fundamental knowledge covered in these sec-
tions applies to the nozzle-less system.
Fiber Formation Process 219
Figure 12.2. Schematic of a nozzle-less electrospinning system.
Figure 12.3. Schematics of two different spinning electrodes: smooth cylinder type and
patterned wire type.
12.2 FIBER FORMATION PROCESS

In a nozzle-based electrospinning system, the high voltage is applied to the fluid
through the metal nozzle. The charged liquid droplet on the tip of the nozzle ex-
periences two types of electrostatic forces: mutual electrostatic repulsion between
the surface charges, and the Coulombic force resulted from the external electric
field. Under these electrostatic forces, the liquid droplet elongates into a Talyor
cone. When the intensity of the electric field exceeds a certain critical value, the
electrostatic forces overcome the surface tension of the fluid and eject the liquid
jet from the tip of the Taylor cone.
Figure 12.4 shows the path of the liquid jet after it is ejected from the Taylor
cone. The jet path begins with a straight zone. The length of the straight zone
increases in response to the electrostatic repulsion between the surface charges.
At the end of the straight zone, the bending perturbations begin and grow rap-
idly to form a three-dimensional coil (i.e., the first bending instability) under the
influence of the charges carried by the jet. Figure 12.5 shows the influence of
the charges on the formation of bending perturbations. The charged jet is forced
(Fdown-out) downward and outward by the charges above, and at the same time, it
also is forced (Fup-out) upward and outward by the charges below. The sum of the
two forces (Fsum = Fdown-out + Fup-out) is in the radial direction and it causes the for-
mation of the three-dimensional coil. As shown in Figure 12.4, the coil is carried
downstream and the diameter of the trajectory increases as both the elongation
Figure 12.4. Schematic of the path of the liquid jet formed during electrospinning.
(Reneker, D.H., et. al., Polymer, 49, 2387–2425, 2008.)
Figure 12.5. Forces acting on a segment of a charged jet.

Solution Properties 221
and the bending continue. The increase in the trajectory diameter is mainly caused
by the electrostatic repulsion between the surface charges. When the diameter of
the trajectory reaches a certain critical value, the liquid jet transforms to a smaller
three-dimensional coil (i.e., the second bending instability), with an axis follow-
ing the curved path of the first coil. The second coil eventually transforms to
an even smaller three-dimensional coil (i.e., the third bending instability). More
bending instabilities could occur, and the formation of new and smaller coils could
continue until the jet is solidified and reaches the collector. Due to these electrical
bending instabilities, the diameter of the liquid jet is reduced significantly and
leads to the formation of nanofibers.
In addition to the electrical bending instabilities, other types of instabilities are
related to the branching of the jet and the formation of beads. However, the elec-
trical bending instabilities are the most important since they directly determine
the diameter of the electrospun nanofibers. The bending instabilities are caused
by the electrostatic forces and are influenced by fluid properties and operational
conditions. If the electrospinning fluid is a polymer solution, the diameter of the
electrospun polymer nanofibers is dependent on: (i) intrinsic properties of the
polymer solution, such as rheological behavior, conductivity, surface tension,
polymer molecular weight, polymer concentration, and additive type and concen-
tration; and (ii) operational conditions, such as voltage, solution flow rate, nozzle
diameter, nozzle-collector distance, and motion of the collector. However, not all
these parameters are fundamental nor are all independent. For example, solution
rheology is a function of both polymer molecular weight and solution concentra-
tion. In addition, voltage, solution flow rate, and nozzle diameter are interrelated,
as are nozzle-collector distance and motion of the collector.
The following two sections discuss the effects of different solution properties
and operational conditions on the structure (especially, the diameter) of electros-
pun polymer nanofibers. Many carbon, ceramic, metal and composite nanofibers
are prepared from solutions that contain polymers, and hence the fundamental
knowledge covered in the following sections also gives insight to the structure
control of these nanofibers.
12.3 SOLUTION PROPERTIES

12.3.1 Rheological Behavior
Solution rheology has a significant influence on the diameter of electrospun nano-
fibers; for example, McKee at al. studied the fiber formation process of electros-
pun linear and branched polyester and found that fiber diameter (D) increases with
the zero shear rate viscosity (η0) of the solution by:
D = 0.05η0 0.8 (12.1)

The viscosity of the spinning solutions can be controlled by adjusting the poly-
mer molecular weight and polymer concentration. Typically, the solution viscos-
ity increases with increase in either molecular weight or concentration. Hence,
nanofibers with smaller diameters can be obtained by reducing polymer molecu-
lar weight and/or solution concentration. However, it also is important to ensure
both molecular weight and solution concentration exceed certain critical values
to allow the formation of sufficient entanglement of polymer chains during elec-
trospinning. The entanglement of the polymer chains prevents the electrically
driven jet from breaking up when the jet is elongated after leaving the nozzle tip.
Figure 12.6 shows SEM images of electrospun polyethylene oxide (PEO) nano-
fibers prepared using two different polymer concentrations. The chain entangle-
ment in the 2 wt% PEO solution is not sufficient and a large amount beads are
formed. When the PEO concentration increases to 5 wt%, straight PEO nanofibers
are formed and few beads are present.
In addition to polymer molecular weight and polymer concentration, many
other approaches can be used to adjust the viscosity of electrospinning solutions.
For example, adding nanoparticle fillers and salt additives are two commonly
used methods for adjusting the solution viscosity. These two approaches are dis-
cussed in sections 12.3.4 and 12.3.5, respectively.
Besides viscosity, it also is found that viscoelasticity particularly relaxation
time, plays a significant role in determining the electrospinnability of polymer
solutions and the diameter of the resultant nanofibers.
12.3.2 Conductivity
The structure of electrospun nanofibers also is influenced by the net charge den-
sity carried by the moving jet, which is determined from the electric current and
the mass of nanofibers collected on the surface of the grounded collector. The
electrostatic repulsion from the excess charge in the liquid jet always tries to in-
crease the surface area, which opposes the formation of spheres and results in thin
fibers. Net volume charge density can be described by:
Figure 12.6. SEM images of PEO nanofibers. Polymer concentrations: (A) 2 wt%, and
(B) 5 wt%. (Shi, Q., et. al., Journal of Polymer Science: Part B: Polymer Physics, 49,
115–122, 2011.)
t
Net volume charge density = c ρ∫Idt / m (12.2)
0
where I is the jet current, t the collecting time, c the solution concentration, ρ the
solution density, and m the mass of electrospun nanofibers. Generally, the net
volume charge density is proportional to the conductivity of the spinning solu-
tion. Therefore, the solution conductivity can be used as one of the determining
parameters to control the diameter of electrospun fibers since it is easy to measure
and adjust the solution conductivity. The simplest approach to adjust the solution
conductivity is to introduce conductive nanoparticle fillers or soluble salt addi-
tives into the solution, as discussed in sections 12.3.4 and 12.3.5.
12.3.3 Surface Tension

In addition to rheological behavior and conductivity, the solution surface tension
also plays an important role in the nanofiber formation, and the final structure of
electrospun nanofibers strongly depends on the interplay between surface tension,
electrostatic repulsion (conductivity), and viscoelastic force (rheological behav-
ior). Surface tension can be defined as the force with which the surface molecules
attract each other. Typically, when the surface tension is too high, beads tend to be
formed during electrospinning. However, the formation of beads can be avoided
by achieving a balance among surface tension, electrostatic repulsion, and vis-
coelastic force. The surface tension of polymer solutions typically is controlled
by adding salts or mixing solvents, as discussed in sections 12.3.5 and 12.3.6,
respectively.
12.3.4 Nanoparticle Filler

Adding nanoparticle fillers into the spinning solutions can change the solution
properties, and hence the electrospinnability and the diameter of resultant nano-
fibers. Generally, adding nanoparticle fillers increases the viscosity of spinning
solutions although the extent differs from system to system. The viscosity of
nanoparticle-added solutions sometimes can be estimated using the following
equation:
η
= 1 + 2.5θ (12.3)
ηs
where η is the viscosity of the nanoparticle-added solution, ηs the viscosity of the

solution without nanoparticles, and θ the nanoparticle content (or volume frac-
tion). Therefore, adding nanoparticle fillers can increase the fiber diameter by
increasing the solution viscosity. However, if the nanoparticle fillers are made
of conducting materials, they also can affect the fiber structure by increasing the
solution conductivity. Figure 12.7 shows SEM images of Au nanoparticle-added
polyvinylpyrrolidone (PVP) nanofibers. The addition of conductive Au nanopar-
ticles actually reduces the fiber diameter due to increased solution conductivity.
In addition to solution viscosity and conductivity, some nanoparticle fillers
also can alter the surface tension of the spinning solutions by forming strong
interactions with polymer chains and/or solvent molecules. These nanoparticle
fillers can affect the structure of electrospun nanofibers by changing the solution
surface tension.
12.3.5 Salt Additive

In addition to nanoparticle fillers, adding salt additives is another means to adjust
the solution properties and eventually the nanofiber structure. Figure 12.8 shows
SEM images of nanofibers electrospun from ZnCl2-added PAN solutions in di-
methylformamide (DMF). The average diameter of the electrospun nanofibers
gradually decreases from 290 to 250, 180, and 130 nm when the ZnCl2 salt con-
centration increases from 0 to 5, 10, and 15 wt%. The decreased fiber diameter is
caused by the changes in solution viscosity, conductivity and surface tension. The
addition of ZnCl2 salt can significantly increase the solution conductivity since
more free ions are available. At the same time, ZnCl2 salt also slightly increases
the solution viscosity and surface tension due to the formation of salt-PAN and
salt-DMF interactions (Figure 12.9). These changes work together to form a new
balance among the viscosity, conductivity and surface tension, which in turn re-
sults in the reduction in fiber diameter with the addition of ZnCl2 salt.
Figure 12.7. SEM images of Au/PVP nanofibers with different Au nanoparticle contents.

(A) 0 (pure PVP), (B) 2.6, (C) 3.7, and (D) 7.2 wt%. (Wang, Y., et. al., Journal of Applied
Polymer Science, 105, 3618–3622, 2007.)
Figure 12.8. SEM images of ZnCl2/PAN nanofibers. ZnCl2 concentration: (A) 0 (pure

PAN), (B) 5, (C) 10, and (D) 15 wt%.
Figure 12.9. Salt-DMF and salt-PAN interactions. Dash lines represent intermolecular

interactions.
The effect of salt additives on the structure of electrospun nanofibers is de-

pendent on the actual spinning solution used. As discussed above, the addition of
ZnCl2 salt in PAN solution results in smaller diameter for the electrospun nano-
fibers. However, the addition of iron acetylacetonate (IAA) salt to the PAN solu-
tion in DMF does not lead to apparent change in the diameter of electrospun PAN
nanofibers. The IAA salt also affects the solution viscosity, surface tension and
conductivity by forming intermolecular interactions with PAN chains and DMF
molecules. However, the changes in these solution properties are not so signifi-
cant and the resultant nanofibers do not exhibit apparent diameter change.
12.3.6 Mixed Solvent

Mixing solvents with different surface tensions is a simple way to adjust the so-
lution properties, especially surface tension. For example, Liu et. al. studied the
electrospinning of cellulose acetate. They found neither acetone (surface tension:
23.7 dyne/cm) nor dimethylacetamide (surface tension: 32.4 dyne/cm) allows
formation of continuous fibers, but a 2:1 acetone:dimethylacetamide solvent mix-

ture, with a surface tension around 26.6 dyne/cm, enables cellulose acetate to be
continuously electrospun into non-woven mats with fiber diameters ranging from
100 nm to 1 μm. Fong et. al., added ethanol to a polyethylene oxide (PEO)/water
solution and produced smooth fibers without the formation of beads.
12.4 OPERATIONAL CONDITIONS

In addition to solution properties, the structure (especially, the fiber diameter) of
electrospun nanofibers also can be controlled by selectively adjusting the electros-
pinning operational conditions, such as voltage, nozzle diameter, nozzle-collector
distance, solution flow rate, and motion of the collector.
12.4.1 Voltage
One of the crucial operational parameters for electrospinning is the voltage between
the nozzle and the grounded collector. A higher voltage indicates larger electric field
strength. In general, a voltage of greater than 5 kV is able to cause the spinning solu-
tion drop at the tip of the nozzle to elongate into the shape of a Taylor Cone during
the initiation of nanofiber electrospinning. A higher voltage often leads to greater
stretching of the solution due to the greater Columbic forces in the jet as well as the
stronger electric field. As a result, the fiber diameter becomes smaller. In addition,
when a solution of lower viscosity is used, a higher voltage favors the branching of
the liquid jet, which in turn leads to fibers with smaller diameters.
However, when the voltage is too high, the flight time for the polymer jet
will be reduced significantly, which may cause larger fiber diameters due to the
reduced time for the fibers to experience the bending instabilities. In addition,
higher voltage also leads to larger mass throughput of the jet, and this may further
increase the fiber diameter.
12.4.2 Nozzle Diameter

Nozzle diameter also is an important operational parameter in controlling the fiber
diameter. A decrease in nozzle diameter typically leads to smaller fiber diameters
and fewer beads. However, if the nozzle diameter is too small, it may not be pos-
sible to extrude the spinning solution through the nozzle to form nanofibers.
12.4.3 Nozzle-Collector Distance

The distance between the nozzle and collector has a direct influence on both
the flight time and electric field strength. During the electrospinning of polymer
References 227
nanofibers, the liquid jet must have enough time for most of the solvents to be
evaporated. In general, a smaller nozzle-collector distance provides higher elec-
tric field strength, leading to greater jet stretching and elongation, and eventually
smaller fiber diameter. On the other hand, with a smaller nozzle-collector dis-
tance, the jet has a shorter distance to travel before reaching the collector, which
may increase the fiber diameter. Therefore, the actual effect of nozzle-collector
distance on the diameter of electrospun nanofibers varies from system to system.
In addition, when adjusting nozzle-collector distance, it is very important to en-
sure the liquid jet has enough flight time for the solvent to evaporate.
12.4.4 Solution Flow Rate

Changing the solution flow rate is another means to control the nanofiber diam-
eter. When the flow rate is increased, there is a corresponding increase in the fiber
diameter simply because a greater volume of spinning solution is drawn away
from the nozzle. In addition, when the flow rate is too high, a significant amount
of beads may form during electrospinning. The formation of beads at high flow
rates may be caused by the reduced charge densities in the liquid jet. To maintain
the high charge densities and reduce the formation of beads, higher voltages are
needed when high flow rates are used.
12.4.5 Other Operational Conditions

In addition to the operational conditions described above, many other processing
parameters, such as temperature, humidity, pressure, and atmosphere type, can be
used to control the diameter of electrospun nanofibers. For example, increasing
temperature can reduce the viscosity of the spinning solution, and hence leads to
smaller fiber diameter. Increasing humidity can cause water condensation on the
fiber surface and influence the surface morphology by improving non-uniformity
or even creating surface pore structures. Reducing pressure can increase the ten-
dency of the spinning solution to flow out of the needle, but may cause unstable
jet initiation. A very low pressure also may cause direct discharge of the electrical
charges and make the electrospinning impossible. Changing air to other gases with
higher breakdown voltages, such as Freon-2, also can increase the fiber diameter.
REFERENCES
[1] Bergshoef, M.M., and Vancso, G.J., “Transparent Nanocomposites
with Ultrathin Electrospun Nylon-4,6 Fiber Reinforcement”, Advanced
Materials, 11, 1362–1365, 1999.
[2] Casper, C.L., Stephens, J.S., N.G., T., Chase, D.B., and Rabolt, J.F.,
“Controlling Surface Morphology of Electrospun Polystyrene Fibers:
Effect of Humidity and Molecular Weight in the Electrospinning Process”,
Macromolecules, 37, 573–578, 2004.
[3] Dai, H., Gong, J., Kim, H., and Lee, D., “A Novel Method for Preparing
Ultra-Fine Alumina-Borate Oxide Fibres via an Electrospinning Technique”,
Nanotechnology, 13, 674–677, 2002.
[4] Deitzel, J.M., Kleinmeyer, J.D., Hirvonen, J.K., and Tan, N.C.B.,
“Controlled Deposition of Electrospun Poly(Ethylene Oxide) Fibers”,
Polymer, 42, 8163–8170, 2001.
[5] Du, J., and Zhang, X., “Role of Polymer-Salt-Solvent Interactions in the
electrospinning of Polyacrylonitrile/Iron Acetylacetonate”, Journal of
Applied Polymer Science, 109, 2935–2941, 2008.
[6] Fong, H., Chun, I., and Reneker, D.H., “Beaded Nanofibers Formed During
Electrospinning”, Polymer, 40, 4585–4592, 1999.
[7] Huang, C., Chen, S., Lai, C., Reneker, D.H., Qiu, H., Ye, Y., and Hou, H.,
“Electrospun Polymer Nanofibers with Small Diameters”, Nanotechnology,
17, 1558–1563, 2006.
[8] Ji, L., Lin, Z., Guo, B., Medford, A.J., and Zhang, “Assembly of Carbon-
SnO2 Core-Sheath Composite Nanofibers for Superior Lithium Storage”,
Chemistry—A European Journal, 16, 11543–11548, 2010.
[9] Ji, L., Lin, Z., Medford, A.J., Zhang, X., “Porous Carbon Nanofibers from
Electrospun Polyacrylonitrile/SiO2 Composites as an Energy Storage
Material”, Carbon, 47, 3346–3354, 2009.
[10] Ji, L., Medford, A.J., and Zhang, X., “Fabrication of Carbon Fibers with
Nanoporous Morphologies from Electrospun Polyacrylonitrile/Poly(L-
lactide) Blends”, Journal of Polymer Science: Part B: Polymer Physics, 47,
493–503, 2009.
[11] Ji, J., Saquing, C., Khan, S.A., and Zhang, X., “Preparation and
Characterization of Silica Nanoparticulate-Polyacrylonitrile Composite and
Porous Nanofibers”, Nanotechnology, 19, 085605, 2008.
[12] Khil, M.S., Kim, H.Y., Kim, M.S., Park, S.Y., and Lee, D.R. “Nanofibrous
Mats of Poly(Trimethylene Terephthalate) via Electrospinning”, Polymer,
45, 295–301, 2004.
References 229
[13] Kim, C., Kim, Y.J., and Kim, Y.A., “Fabrication and Structural
Characterization of Electro-Spun Polybenzimidazol-Derived Carbon
Nanofiber by Graphitization”, Solid State Communications, 132, 567–571,
2004.
[14] Li, D., and Xia, Y., “Electrospinning of Nanofibers: Reinventing the Wheel”,
Advanced Materials, 16, 1151, 2004.
[15] Li, Y., Lin, Z., Xu, G., Yao, Y., Zhang, S., Toprakci, O., Alcoutlabi, M., and
Zhang, X., “Electrochemical Performance of Carbon Nanofibers Containing
an Enhanced Dispersion of Silicon Nanoparticles for Lithium-Ion Batteries
by Employing Surfactants”, ECS Electrochemistry Letters, 1, A31–A33,
2012.
[16] Liang, Y., Ji, L., Guo, B., Lin, Z., Yao, Y., Li, Y., Alcoutlabi, M., Qiu,
Y., and Zhang, X., “Preparation and Electrochemical Characterization of
Ionic-Conducting Lithium Lanthanum Titanate Oxide/Polyacrylonitrile
Submicron Composite Fiber-Based Lithium-Ion Battery Separators”,
Journal of Power Sources, 196, 436–441, 2011.
[17] Lin, Z., Ji, L., Toprakci, O., Krause, W., and Zhang, X., “Electrospun Carbon
Nanofiber-Supported Pt-Pd Alloy Composites for Oxygen Reduction”,
Journal of Materials Research, 25, 1329–1335, 2010.
[18] Lin, Z., Ji, L., Woodroof, M.D., Medford, A.J., Shi, Q., Krause, W., and
Zhang, X., “Electrocatalytic Interaction of Nano-Engineered Palladium
on Carbon Nanofibers with Hydrogen Peroxide and β-NADH”, Journal of
Solid State Electrochemistry, 15, 1287–1294, 2011.
[19] Liu, H., and Hsieh, Y.L., “Ultrafine Fibrous Cellulose Membranes from
Electrospinning of Cellulose Acetate”, Journal of Polymer Science Part B:
Polymer Physics, 40, 2119–2129, 2002.
[20] Matthews, J.A., Wnek, G.E., Simpson, D.G., and Bowlin, G.L.,
“Electrospinning of Collagen Nanofibers”, Biomacromolecules, 3, 232–
238, 2002.
[21] McKee, M.G., Wilkes, G.L., Colby, R.H., and Long, T.E., “Correlations
of Solution Rheology with Electrospun Fiber Formation of Linear and
Branched Polyesters”, Macromolecules, 37, 1760–1767, 2004.
[22] Ramakrishna, S., Fujihara, K., Teo, T.C., and Ma, Z., An Introduction to
Electrospinning and Nanofibers, World Scientific Publishing, 2005.
[23] Reneker, D.H., and Yarin, A.L., “Electrospinning Jets and Polymer
Nanofibers”, Polymer, 49, 2387–2425, 2008.
[24] Shi, Q., Vitchuli, N., Nowak, J., Lin, Z., Guo, B., McCord, M., Bourham,
M., and Zhang, X., “Atmospheric Plasma Treatment of Pre-Electrospinning
Polymer Solution: A Feasible Method to Improve Electrospinnability”,
Journal of Polymer Science: Part B: Polymer Physics, 49, 115–122, 2011.
[25] Stephens, J.S., Chase, D.B., and Rabolt, J.F., “Effect of the Electrospinning
Process on Polymer Crystallization Chain Conformation in Nylon-6 and
Nylon-12”, Macromolecules, 37, 877–881, 2004.
[26] Taylor, G., “Disintegration of Water Drops in an Electric Field”, Proceedings
of the Royal Society A: Mathematical, Physical and Engineering Sciences,
280, 918–924, 1964.
[27] Teo, W.E., Ramakrishna, S., “A Review on Electrospinning Design and
Nanofibre Assemblies”, Nanotechnology, 17, R89–R106, 2006.
[28] Toprakci, O., Toprakci, H.A.K., Ji, L., Lin, Z., Gu, R., and Zhang, X.,
“LiFePO4 Nanoparticles Encapsulated in Graphene-Containing Carbon
Nanofibers for Use as Energy Storage Materials”, Journal of Renewable
and Sustainable Energy, 4, 013121, 2012.
[29] Toprakci, O., Toprakci, H.A.K., Ji, L., Lin, Z., Xu, G., and Zhang, X.,
“Carbon Nanotube-Loaded Electrospun LiFePO4/Carbon Composite
Nanofibers as Stable and Binder-Free Cathodes for Rechargeable Lithium-
Ion Batteries”, ACS Applied Materials & Interfaces, 4, 1273–1280, 2012.
[30] Wang, Y., Li, Y., Sun, G., Zhang, G., Liu, H., Du, J., Yang, S., Bai, J., and Yang,
Q., “Fabrication of Au/PVP Nanofiber Composites by Electrospinning”,
Journal of Applied Polymer Science, 105, 3618–3622, 2007.
[31] Wendorff, J.H., Agarwal, S., and Greiner, A., Electrospinning: Materials,
Processing, and Applications, Wiley-VCH, 2012.
[32] Zhang, S., Lu, Y., Xu, G., Li, Y., and Zhang, X. “LiF/Fe/C Nanofibers as
High-Capacity Cathode Material for Li-Ion Batteries”, Journal of Physics
D: Applied Physics, 45, 395301, 2012.
[33] Zhang, X., “Processing-Structure relationships of Electrospun Nanofibers”,
in Nanofibers: Fabrication, Performance, and Applications, editor Chang,
W.N., Nova Science Publishers, 2009.
PROBLEMS
(1) Describe the nozzle-based electrospinning process.
(2) How does the nozzle-less electrospinning process work?
(3) How does the bending instability occur during electrospinning?
Problems 231
(4) Discuss and explain how you would expect the structure of electrospun fi-
bers to vary with the following solution properties:
(i) Viscosity
(ii) Conductivity
(iii) Surface tension
(iv) Nanoparticle filler
(v) Salt additive
(5) Discuss and explain how you would expect the structure of electrospun fi-
bers to vary with the following operational conditions:
(i) Voltage
(ii) Nozzle diameter
(iii) Nozzle-collector distance
(iv) Solution flow rate
CHAPTER 13
Formation of Nanofibers by Other Methods
Electrospinning is the most well-known method for fabricating nanofibers.

However, there are other methods that can be used to produce nanofibers. For
example, centrifugal spinning utilizes the centrifugal force generated by a rotat-
ing spinneret to produce nanofibers. Melt blowing typically is used to produce
fibers with diameters greater than 1 μm by using high-velocity air; however, it
can produce nanofibers by using carefully selected materials and processing pa-
rameters. Bicomponent fiber separation, phase separation, template synthesis and
self-assembly also are useful methods for fabricating nanofibers from different
materials.
13.1 CENTRIFUGAL SPINNING

Centrifugal spinning, also called rotational jet spinning or forcespinning, is a
useful method for producing nanofibers at high speed and low cost. In general,
centrifugal spinning is used to produce polymer nanofibers. However, carbon,
ceramic and metal nanofibers can be fabricated by the centrifugal spinning of
their precursors, followed by thermal treatments. Composite nanofibers can be
produced either by adding nanoparticle fillers directly into the centrifugal spin-
ning fluid or by a post-process, such as electrochemical deposition and chemical
deposition, after the centrifugal spinning.
13.1.1 Fiber Formation Process

Figure 13.1 shows a basic centrifugal spinning set-up. During fiber formation, a
spinning fluid is placed in a rotating spinneret, which contains multiple nozzles
on the sidewall. When the rotating speed reaches a critical value, the centrifugal
233
234 Formation of Nanofibers by Other Methods
force will overcome the surface tension of the spinning fluid to eject liquid jets
from the nozzle tips of the spinneret. Figure 13.2 shows the path of a liquid jet
after it is ejected from the nozzle tip. The centrifugal force is accompanied by the
air frictional force to elongate the jets, leading to the formation of nanofibers. In
addition to the centrifugal force and air friction force, other forces that may affect
the nanofiber formation process include rheological force, surface tension, and
gravitational force. The as-spun nanofibers are deposited on the surface of the
collector. Figure 13.3 shows a SEM image of polyacrylonitrile (PAN) nanofibers
centrifugally-spun from a 14 wt% PAN solution. The centrifugal spinning process
is simple and scalable, and it avoids the use of high electric voltage and can be
used to produce many different materials for various applications.
Figure 13.1. Schematic of a basic centrifugal spinning setup.
Figure 13.2. Schematic of the path of a liquid jet ejected from the nozzle tip during cen-
trifugal spinning. Rheological force, surface tension and gravitational force may affect the
nanofiber formation process, but they are not shown here for the purpose of simplicity.
Centrifugal Spinning 235
13.1.2 Centrifugal Spinning System

Two key components for practical centrifugal spinning systems are carefully de-
signed spinneret and collector. Figure 13.4 shows two cylinder-shape spinneret
designs. In Figure 13.4A, the spinneret has two nozzles, which are imbedded on
the opposite sides of the cylinder wall. The nozzle diameter and length are impor-
tant for controlling the fiber diameter. This cylinder-shape spinneret can be used
to spin nanofibers directly from a polymer solution. To spin nanofibers directly
from a polymer melt, inductive heating coils need to be used to melt the poly-
mer chips. To increase the production rate, multiple nozzles can be introduced
(Figure 13.4B). In addition to cylinder spinnerets, other types of spinnerets, with
spheroid, oblate spheroid, or trapezoidal shapes, also exist.
Different types of collectors have also been designed to collect nanofibers in
different forms. Figure 13.5A shows a circular metal or plastic collector, which
can collect nanofibers on the surface of the inner wall. The diameter of the cir-
cular collector determines the nozzle-collector distance, which is an important
parameter for determining the nanofiber structure. The circular collector is suit-
able for batch production of nanofibers. A nanofiber sheet collector is shown in
Figure 13.5B. In this case, nanofibers are collected onto a porous substrate (such
as textile fabric, paper, or other porous membranes) placed on the top of the col-
lector. A suction force is used to assist the formation of a uniform nanofiber sheet
(or nonwoven) on the surface of the porous substrate. The sheet collector is suit-
able for continuous production of nanofibers. Figure 13.5C shows a yarn collec-
tor. Nanofibers are collected in a water bath and the use of a rotating roller allows
the collection of continuous yarns of nanofibers.
13.1.3 Fluid Properties

The structure of centrifugally spun nanofibers is affected by the properties of the
spinning fluid. Two most important fluid properties that affect the nanofiber struc-
ture are viscosity and surface tension. Typically, increasing fluid viscosity can
decrease the number of so-called bead defects and increase the diameter of the
fibers. On the other hand, surface tension often is the driving force for the forma-
tion of beads and beaded fibers because it tends to convert the liquid jet into spher-
ical droplets by making the surface area smaller. During centrifugal spinning, the
effect of surface tension is suppressed by the centrifugal force and air frictional
force, which tend to elongate the spinning jets and increase the surface area.
The viscosity and surface tension of the spinning fluid can be controlled by
selectively adjusting other fluid properties and/or operational conditions. For ex-
ample, when the spinning fluid is a polymer solution, the viscosity can be changed
by adjusting the polymer molecular weight, polymer concentration, solvent type,
and additive type and content. The surface tension can be changed by adjust-
ing the solvent type and additive type and content. When the spinning fluid is a
polymer melt, polymer molecular weight, additive, and temperature are the main
parameters that can be used to control the fluid viscosity and surface tension.
13.1.4 Operational Conditions

Operational conditions, such as rotating speed, spinneret diameter, nozzle diameter,
nozzle-collector distance, etc., also affect the structure of centrifugally spun nanofibers.
Rotating Speed—The rotating speed of the spinneret is one of the most

important operational conditions since it directly affects the centrifugal
force and air frictional force. During centrifugal spinning, the centrifugal
force combines with the air frictional force to elongate the liquid jet into
nanofibers. When the spinning fluid is contained in a rotating nozzle tip, the
centrifugal force (Fcentri) formed on the liquid can be described by:
Figure 13.3. SEM image of PAN nanofibers produced by centrifugal spinning.
(A)
(B)
Figure 13.4. Cylinder-shape spinnerets with (A) two nozzles, and (B) multiple nozzles.
Centrifugal Spinning 237
Figure 13.5. Schematics of three different nanofiber collectors: (A) circular collector,

(B) sheet collector, and (C) yarn collector.
Fcentri = mω2 D / 2 (13.1)
where m is the mass of the fluid, ω the rotating speed of the spinneret, and
D the diameter of the spinneret. To eject the liquid jet from the nozzle tip,
the rotating speed must exceed a critical value to generate sufficient centrif-
ugal force to overcome the surface tension of the spinning fluid. Therefore,
to produce nanofibers, it is important to determine the critical rotating speed
of the spinning fluid.
After the liquid jet is ejected from the nozzle tip, the frictional force applied
onto the jet still can be calculated by using Equation 13.1, but ω becomes
the rotating speed of the jet and D becomes the diameter of the path of the
jet. At the same time, the air frictional force (Ffri) can be given as:
1
Ffri = − πC ρAω2 D 2 (13.2)
2
where C is a numerical drag coefficient, ρ the air density, A the cross-sec-

tional area of the jet, ω the rotating speed of the jet, and D the diameter of
the path of the jet.
While the liquid jet travels toward the collector, the rotating speed of the
jet decreases gradually. However, a higher rotating speed of the spinneret
always means a higher rotating speed of the jet after it travels a certain dis-
tance. Therefore, with increase in rotating speed of spinneret, the centrifugal
force and air frictional force applied on the liquid jet also increase, which
leads to greater elongation of the liquid jet and favors the reduction of fiber
diameter.
However, when the rotating speed is too high, the flight time for the liquid
jet to reach the collector becomes shorter, which may cause larger fiber
diameter due to the reduced time for the jet to stretch and elongate. In ad-
dition, higher rotating speed also leads to larger mass throughput of the jet,
which may further increase the fiber diameter. Therefore, it is important
to obtain the optimal rotating speed that can result in the smallest fiber
diameter.
Spinneret Diameter—The diameter of the spinneret is another im-
portant parameter that affects the structure of nanofibers. According to
Equation 13.1, if the rotating speed is fixed, the centrifugal force increases
with increase in spinneret diameter. As a result, it is easier to eject the liquid
jet by using larger spinneret. In addition, a larger spinneret diameter also
may lead to greater stretching and elongation of the liquid jet, which is
beneficial for forming thinner fibers.
Nozzle Diameter—Changing the nozzle diameter is another means to
control the nanofiber structure. A decrease in nozzle diameter typically leads
to smaller mass throughput of the liquid jet, which in turn reduces the fiber
diameter. However, if the nozzle diameter is too small, it may not be pos-
sible to eject the liquid jet and hence nanofibers cannot be formed.
Nozzle-Collector Distance—The distance between the nozzle and collector
has a direct influence on the flight time of the liquid jet. When a solution is
used as the spinning fluid, a minimum nozzle-collector distance is required
so the liquid jet can have sufficient time for most of the solvent to be evapo-
rated before reaching the collector. In addition, when the nozzle-collector
distance increases, the liquid jet will have a longer distance to travel and
this favors the fiber diameter reduction (This applied to both solution and
melt). However, when the nozzle-collector distance is too large, the liquid
jet may not be able to reach the collector due to the insufficient centrifugal
force.
Melt Blowing 239
Temperature—The temperature, at which centrifugal spinning is carried

out also plays an important role in controlling the fiber structure. Typically,
with increase in temperature, the fluid viscosity decreases. This favors the
reduction of fiber diameter, but also may lead to the formation of beads.
Air Flow—During centrifugal spinning, a suction force often is used to
collect nanofibers onto a porous substrate (Figure 13.5B). Sometimes, air
also is blown onto the liquid jet to assist the spinning process. In both cases,
the air flow may enhance the elongation of the liquid jet and lead to smaller
fiber diameter.
13.2 MELT BLOWING

Melt blowing was developed in the 1950s at the Naval Research Laboratory in
Washington, D.C., and it currently is one of the leading fiber and non-woven
manufacturing processes. Fibers produced by melt blowing typically have diam-
eters greater than 1 μm. However, by the careful control of operation conditions,
it is possible to produce nanofibers from certain materials.
Figure 13.6 shows a schematic of the melt-blowing process. After the polymer
is melt, jets are extruded simultaneously through multiple nozzles, and then at-
tenuated by heated, high-velocity air streams, in which the jet diameters decrease
significantly to form fine fibers. The fibers finally are deposited on a collector in
random orientations to form a non-woven web. Figure 13.7 shows a SEM image
of polybutylene terephthalate (PBT) nanofibers produced by melt blowing.
The structure of melt-blown fibers can be controlled by varying flow rates of
polymer melt and air, temperatures of polymer melt and air, nozzle geometry,
and intrinsic properties of polymer melt. For example, higher air flow rate can
Figure 13.6. Schematic of the melt blowing process. For simplicity, only a single nozzle is
shown. In actual production, multiple nozzles are used simultaneously.
enhance the attenuation of jets and favor the formation of smaller-diameter fibers.
Increasing the temperature of polymer melt also allows longer attenuation time
and leads to smaller fiber diameters.
13.3 BICOMPONENT FIBER SPINNING

Bicomponent fibers that have two immiscible polymers arranged in the cross-
section can be used to produce nanofibers by: (i) splitting the different compo-
nents, or (ii) removing one of the components. To make nanofibers, at least one of
the components has a cross-section with nanosized diameter or thickness. Cross-
sectional shapes of bicomponent fibers that can produce nanofibers include, but
are not limited to islands-in-the-sea, segmented pie, and hollow segmented pie
(Figure 13.8). These bicomponent fibers can be prepared by using two extruders
to spin different polymers through specially designed spinneret systems. Some
of the bicomponent fibers also can be prepared by spinning fibers directly from
polymer blends.
The splitting of bicomponent fibers involves mechanically working with the fi-
bers by drawing, needle-punching, beating, twisting, carding or hydroentangling.
When the splitting method is used, a relatively weak bonding between the fiber
Figure 13.7. SEM image of PBT nanofibers produced by melt blowing. (Ellison, C., et. al.,
Polymer, 48, 3306–3316, 2007.)
Figure 13.8. Typical cross-sectional shapes of bicomponent fibers used for producing

nanofibers.
Phase Separation 241
components is required to facilitate the splitting process. However, the bonding

still should be sufficient to allow the formation of bicomponent fibers during fiber
spinning.
The removal of one component from bicomponent fibers typically involves the
dissolving of the component using hot water, alkali solution, or other solvents. To
use the component removal method, the bicomponent fibers must be composed
of one soluble polymer and one un-soluble polymer. The most-commonly used
cross-sectional structure for this technology is islands-in-the-sea. Figure 13.9
shows a SEM image of carbon nanotube (CNT)-filled polyacrylonitrile (PAN)
composite nanofibers prepared by dissolving the sea component from an islands-
in-the-sea bicomponent fiber, in which CNT/PAN is the island component and
polymethyl methacrylate (PMMA) is the sea.
13.4 PHASE SEPARATION

Nanofibers can be produced using the phase separation phenomenon of polymer
solutions. A typical phase separation process for making nanofibers involves
multi-steps: polymer dissolution, gelation, phase separation, solvent removal,
and drying. Figure 13.10 shows the nanofiber formation process using the phase
Figure 13.9. SEM images of CNT/PAN composite nanofibers prepared from an islands-in-

the-sea bicomponent fiber (island: CNT/PAN, sea: PMMA). (A) Low-magnification image
showing the separation of CNT/PAN nanofibers from the islands-in-the-sea fiber after the
removal of the sea component by using nitromethane solvent, and (B) high-magnification
image showing separated CNT/PAN nanofibers. (Chae, H.G., et. al., Composites Science
and Technology, 69, 406–413, 2009.)
Figure 13.10. Schematic of the phase separation process for producing nanofibers.

separation method. The polymer is first dissolved in a solution at room tempera-

ture or an elevated temperature. The homogeneous solution then is kept at the
gelation temperature, which is dependent on the polymer type, solvent type, and
polymer concentration. At the gelation temperature, the polymer solution forms a
gel and phase separate to form a nanofibrous matrix. Nanofibers are obtained after
removing the solvent and drying the matrix. Figure 13.11 shows a SEM image
of poly(L-lactic acid) (PLLA) prepared by the phase separation method using a
tetrahydrofuran solvent.
13.5 TEMPLATE SYNTHESIS

In the template synthesis method, nanofibers are produced in hollow channels of
porous ceramic or polymer templates. Figure 13.12 shows the nanofiber forma-
tion process using the template synthesis method. The porous template first is
filled with monomers. Polymer nanofibers are synthesized from the monomers
chemically or electrochemically in the hollow channels of the porous template.
Separated nanofibers are obtained after the removal of the template by dissolving
or etching. Figure 13.13 shows a SEM image of polypyrrole (PPy) nanofibers
prepared by the template synthesis method. Nanofibers produced by template
synthesis often have a hollow structure because the synthesized polymer tends to
precipitate onto the inner surface of the hollow channels.
In addition to the use of monomers, it is possible to form nanofibers directly
from a polymer solution. In this case, the polymer solution is filled into the hol-
low channels and solidified into nanofibers by removing the solvent. Nanofibers
produced from the polymer solution typically have larger diameters than those
synthesized from monomers. This is because of the high viscosity of polymer
solutions, which does not allow the use of hollow channels with very small di-
ameters. Figure 13.14 shows a SEM image of PAN nanofibers prepared by filling
the hollow channels of a porous aluminum oxide template with a PAN solution,
followed by polymer solidification and template removal.
Figure 13.11. SEM image of PLLA nanofibers produced by phase separation. (Ma, P.X.,
et. al., Journal of Biomedical Materials Research, 46, 60–72, 1999.)
Self-Assembly 243
Figure 13.12. Schematic of the template synthesis process for producing nanofibers.
Figure 13.13. SEM image of PPy nanofibers prepared by template synthesis. (Martin,

C.R., Chemistry of Materials, 8, 1739–1746, 1996.)
13.6 SELF-ASSEMBLY
The self-assembly method produces nanofibers by holding small molecules to-
gether via intermolecular interactions. Depending on the chemical structure of
the small molecules, there are a few possible mechanisms that can be used to as-
semble nanofibers. One of the commonly used mechanisms is based on the forma-
tion of hydrogels, which contain two interpenetrated phases, i.e., the solid phase
and the liquid phase. The liquid phase typically is water, and the solid phase is
a network of nanofibers formed by the self-assembly of hydrogelator molecules.
Figure 13.14. SEM image of PAN nanofibers prepared by template synthesis. (Feng, L., et.
al., Angewandte Chemie International Edition, 41, 1221–1223, 2002.)
Dried nanofibers can be obtained by removing the liquid phase from the hydro-
gels. Figure 13.15 shows the formation of nanofibers by the self-assembly of hy-
drogelator molecules that have both hydrophilic and hydrophobic ends. These
hydrogelator molecules form nanofibers by weak intermolecular hydrogen bond-
ing and hydrophobic interactions. Figure 13.16 shows an example of such hydro-
gelator and the nanofibers formed by the hydrogelator molecules.
Figure 13.15. Assembly of nanofibers by hydrogelator molecules.

Figure 13.16. (A) Chemical structure of a hydrogelator molecule, and (B) SEM image of
nanofibers formed by the self-assembly of the hydrogelator molecules. (Yang, Z.M., et. al.,
Chemical Communications, 2424–2425, 2004.)
References 245
REFERENCES
[1] Anantharamaiah, N., Verenich, S., and Pourdeyhimi, B., “Durable Nonwoven
Fabrics via Fracturing Bicomponent Islands-in-the-Sea Filaments”, Journal
of Engineered Fibers and Fabrics, 3, 1–9, 2008.
[2] Badrossamay, M.R., McIlwee, H.A., Goss, J.A., and Parker, A.K.,
“Nanofiber Assembly by Rotary Jet-Spinning”, Nano Letters, 10, 2257–
2261, 2010.
[3] Begenir, A., Michielsen, S., and Pourdeyhimi, B., “Melt-Blowing
Thermoplastic Polyurethane and Polyether-Block-Amide Elastomers:
Effect of Processing Conditions and Crystallization on Web Properties”,
Polymer Engineering and Science, 49, 1340–1349, 2009.
[4] Chae, H.G., Choi, Y.H., Minus, M.L., and Kumar, S., “Carbon Nanotube
Reinforced Small Diameter Polyacrylonitrile Based Carbon Fiber”,
Composites Science and Technology, 69, 406–413, 2009.
[5] Ellison, C., Phatak, A., Giles, D.W., Macosko, C.W., and Bates, F.S.,
“Melt Blown Nanofibers: Fiber Diameter Distributions and Onset Of Fiber
Breakup”, Polymer, 48, 3306–3316, 2007.
[6] Feng, L., Li, S., Li, H., Zhai, J., Song, Y., Jiang, L., and Zhu, D., “Super-
Hydrophobic Surface of Aligned Polyacrylonitrile Nanofibers”, Angewandte
Chemie International Edition, 41, 1221–1223, 2002.
[7] Ikegame, M., Tajima, K., and Aida, T., “Template Synthesis of Polypyrrole
Nanofibers Insulated within One-Dimensional Silicate Channels: Hexagonal
versus Lamellar for Recombination of Polarons into Bipolarons”,
Angewandte Chemie International Edition, 42, 2154–2157, 2003.
[8] Lee, B., Ko, J., and Han, S., “Characteristics of PP/PET Bicomponent Melt
Blown Nonwovens as Sound Absorbing Material”, Advanced Materials
Research, 123, 935–938, 2010.
[9] Josephs, H., and Huston, R., Dynamics of Mechanical Systems, CRC Press,
2002.
[10] Ma, P.X., and Zhang, R.Y., “Synthetic Nano-Scale Fibrous Extracellular
Matrix”, Journal of Biomedical Materials Research, 46, 60–72, 1999.
[11] Martin, C.R., “Membrane-Based Synthesis of Nanomaterials”, Chemistry
of Materials, 8, 1739–1746, 1996.
[12] Martin, C.R., “Nanomaterials—A Membrane-Based Synthetic Approach”,
Science, 266, 1961–1966, 1994.
[13] McCulloch, J.G., “The History of the Development of Melt Blowing

Technology”, International Nonwovens Journal, 8, 109–121, 1999.
[14] Mellado, P., Mcllwee, H.A., Badrossamay, M.R., Goss, J.A., Mahadevan,
L., and Parker, K.K., “A Simple Model for Nanofiber Formation by Rotary
Jet-Spinning”, Applied Physics Letters, 99, 2031107, 2011.
[15] Michielsen, S., Pourdeyhimi, B., and Desai, P., “Review of Thermally
Point-Bonded Nonwovens: Materials, Processes, and Properties”, Journal
of Applied Polymer Science, 99, 2489–2496, 2006.
[16] Ohkoshi, Y., “Melt Spinning and Other techniques for the Production of
Nanofibers and Microfibres”, in Handbook of textile Fibre Structure. Volume
1: Fundamentals and manufactured Polymer Fibres, Editors Eichhorn, S.J.,
Hearle, J.W.S., Jaffe, M., and Kikutani, T., Woodhead Publishing Limited,
2009.
[17] Pourdeyhimi, B., Fedorova, N., Dondero, W., Gorga, R.E., Michielsen,
S., Ghosh, T., Chhaparwal, S., Barrera, C., Rinaldi, C., Satcher, M., and
Hinestroza, J.P., “Textile Nanotechnology”, in Handbook of Nanoscience,
Engineering, and Technology, Second Edition, Editors Goddard, W.A.III.,
Brenner, D.W., Lyshevski, S.E., and Lafrate, G.J., CRC Press, 2007.
[18] Sarkar, K., Gomez, C., Zambrano, S., Ramirez, M., Hoyos, E., Vasquez, H.,
and Lozano, K., “Electrospinning to Forcespinning”, Materials Today, 13,
12–14, 2010.
[19] Wang, L., Shi, J., Liu, L., Secret, E., and Chen, Y., “Fabrication of Polymer
Fiber Scaffolds by Centrifugal Spinning for Cell Culture Studies”,
Microelectronic Engineering, 88, 1718–1721, 2011.
[20] Wang, X., Yao, J., and Pan, X., “Fiber Splitting of Bicomponent Meltblown
Nonwovens by Ultrasonic Wave”, International Journal of Chemistry, 1,
26–33, 2009.
[21] Weitz, R.T., Harnau, L., Rauschenbach, S., Burghard, M., and Kern, K.,
“Polymer Nanofibers via Nozzle-Free Centrifugal Spinning”, Nano Letters,
8, 1187–1191, 2008.
[22] Wu, C.G., and Bein, T., “Conducting Polyaniline Filaments in a Mesoporous
Channel Host”, Science, 264, 1757–1759, 1994.
[23] Yang, Z.M., Gu, H.W., Fu, D.G., Gao, P., Lam, J.K., and Xu, B., “Enzymatic
Formation of Supramolecular Hydrogels”, Advanced Materials, 16, 1440–
1444, 2004.
Problems 247
[24] Yang, Z.M., and Xu, B., “A Simple Visual Assay Based on Small
Molecule Hydrogels for Detecting Inhibitors of Enzymes”, Chemical
Communications, 2424–2425, 2004.
[25] Yang, Z.M., and Xu, B., “Supramolecular Hydrogels Based on Biofunctional
Nanofibers of Self-Assembled Small Molecules”, Journal of Materials
Chemistry, 17, 2385–2393, 2007.
[26] Zhang, Y., Gu, H.W., Yang, Z.M., and Xu, B., “Supramolecular Hydrogels
Respond to Ligand-Receptor Interaction”, Journal of the American
Chemical Society, 125, 13680–13681, 2003.
[27] Zhou, F. and Gong, R., “Manufacturing Technologies of Polymeric
Nanofibres and Nanofibre Yarns”, Polymer International, 57, 837–845,
2008.
PROBLEMS
(1) Describe the fiber formation process of centrifugal spinning.
(2) Describe the fiber formation process of melt blowing.
(3) How can nanofibers be formed by bicomponent fiber spinning?
(4) Describe the phase separation process for making nanofibers.
(5) Describe the template synthesis process for making nanofibers.
(6) Describe the self-assembly process for making nanofibers.
PART III
Fiber Properties
CHAPTER 14
Primary and Secondary Properties
Fiber properties can be classified into primary and secondary properties. Primary
properties are those that fibers must possess so they can be converted into useful
products. Examples of primary properties are aspect ratio, strength, flexibility,
cohesiveness, and uniformity. Secondary properties are those that are desirable
and can improve consumer satisfaction with the end-products made from the fi-
bers. Secondary properties include, but are not limited to physical shape, density,
modulus, elongation, elastic recovery, resilience, thermal properties, electrical
properties, color and optical properties, moisture regain, resistance to chemical
and environmental conditions, resistance to biological organisms, and resistance
to insects. This chapter provides a brief introduction on these primary and sec-
ondary properties. Chapters 15–20 give more detailed discussion on some of the
important properties.
14.1 PRIMARY PROPERTIES

Primary properties are essential for converting the fibers into useful products. The
requirement for primary properties varies from application to application. The
following is a brief discussion on five primary properties that often are required
in many fiber-based products.
14.1.1 Aspect Ratio

One important structural characteristic of fibers is their lengths are considerable
greater than diameters. Fibers with relatively short lengths, measured in terms of
centimeters or inches, are called staple fibers. Long fibers, measured in kilome-
ters or miles, are called filament fibers. The diameters of most fibers range from
251
252 Primary and Secondary Properties
several micrometers to tens of micrometers. However, nanofibers, i.e., fibers with

diameters less than 1 micrometer, also are being produced for certain applications.
Due to the long lengths and small diameters, fibers typically have large aspect
ratios, i.e., length-to-diameter ratios. All natural polymer fibers except silk are
staple fibers, and the aspect ratios range from 1000 to 5000. Synthetic polymer
fibers are produced as filaments, but they can be cut into staple fibers with desired
aspect ratios. Inorganic fibers and nanofibers often are produced in continuous fil-
ament form. But they also can be made into staple fibers for certain applications.
Aspect ratio has an important effect on the structure and properties of the
products made from fibers. For example, in the textile industry, staple fibers are
twisted or otherwise assembled together to make continuous yarns. Yarns made
from staple fibers often have dull appearance since there are many fiber ends on
the yarn surface. On the other hand, filaments can be made into yarns with little
or no twisting, and the resultant yarns look smooth and lustrous. In the composite
application, filaments can be assembled into preforms, and the introduction of
resins leads to the formation of composites with excellent mechanical properties.
Staple fibers also can be used to make composites by directly mixing with resins.
Composites made from very short staple fibers are often weaker than those made
from filaments. However, when the aspect ratio is beyond 100, staple fiber com-
posites could have comparable properties as filament composites.
14.1.2 Strength
Fibers must have adequate strength to survive the fabrication process used for
making the end-products. Strength is the ability of a material to withstand an ap-
plied stress without failure. In the textile field, fiber strength is defined as tenacity.
The tenacity of a fiber basically is the tensile strength expressed as force per unit
linear density. A commonly used unit for tenacity is gram per denier, in which
denier is a unit of fiber or yarn measurement equal to the weight in grams of 9,000
meters of fiber or yarn. In addition to gram per denier, the unit gram per tex is
also often used, in which tex is a unit of fiber or yarn measurement equal to the
weight in grams of 1,000 meters of fiber or yarn. In fields other than textiles, the
International System (S.I.) unit, i.e., pascal (Pa) or newton per square meter (N/
m²), generally is used.
Table 14.1 shows the strengths of some synthetic polymer, natural polymer,
and inorganic fibers. The strength or tenacity of fibers is determined by factors,
such as chemical structure, degree of crystallinity, molecular orientation, and uni-
formity. For example, most polymer fibers have strengths in the range of 2.5–10
grams per denier. However, Kevlar and Spectra fibers can have tenacities greater
than 10 grams per denier due to their unique chemical and physical structures.
Inorganic fibers also can have very high strengths.
The strength value of fibers provides an importance indication on the poten-
tial strength of the resultant fiber-based products. However, the actual strength
Primary Properties 253
Table 14.1. Strengths of some synthetic polymer, natural polymer, and

inorganic fibers. Data obtained at 20°C and 65 % relative humidity.
Fiber Strength (grams per denier)

Synthetic Polymer Fibers
Polyester, regular 2.5–6.0
Polyester, high-strength 6.0–10.5
Acrylic 2.0–3.6
Nylon 6,6, regular 4.3–6.0
Nylon 6,6, high-strength 6.0–9.0
Nylon 6, regular 5.7–7.7
Nylon 6, high-strength 7.7–9.5
Kevlar 20–25
Spectra 25–43
Natural Polymer Fibers (including regenerated fibers)
Cotton 3.0–5.0
Flax 2.6–7.7
Hemp 5.8–6.8
Jute 3.0–5.8
Ramie 5.3–7.4
Wool 1.0–1.7
Silk 2.4–6.1
Acetate 1.2–1.5
Triacetate 1.1–1.4
Rayon, regular 1.8–3.0
Rayon, high-strength 3.0–5.0
Inorganic Fibers
Carbon 8.6–40
Glass 6.3–20
Sources: Hull, D., et. al., An Introduction to Composite Materials, Second Edition, Cambridge
University Press, 1996.; Morton, W.E., et. al., Physical Properties of Textile Fibres, Fourth
Edition, Woodhead Publishing Limited, 2008.; Mwaikambo, L.Y., African Journal of Science and
Technology, 7, 120–133, 2006.; Tortora, P.G., Understanding Textiles, Fourth Edition, Macmillian
of the end-products also will be affected by many other factors. For example,
the twisting of staple fibers plays an important role in determining the actual
strength of yarns made from staple fibers. Weak staple fibers could be made into
relative strong yarns by selecting appropriate processing parameters. In addition,
high-strength fibers are desired in making composites. However, the strength of
fiber-reinforced composites also is determined by the arrangement of fibers and

the interface between the fibers and the matrix.
14.1.3 Flexibility
The flexibility of fibers refers to their ability to be bent or folded. In practical ap-
plications, fibers often need to be bent or folded to make the end-products. During
the use of products, the flexible feature of fibers may also be required. However,
different processes and applications have different requirements for the fiber flex-
ibility. For example, wood is composed of cellulose fibers. These fibers lack flex-
ibility and cannot be used as textile fibers, but they can be used for making paper.
The flexibility of fibers is affected by both the chemical and physical struc-
tures. For example, most polymer fibers are flexible since they are made of flex-
ible chains. However, the flexibility is reduced with increase in the orientation of
polymer chains. The flexibility of fibers can be increased by introducing the crimp
structure along the longitudinal direction or changing the circular cross-section
to an oval one. Another important and effective method to increase the fiber flex-
ibility is by reducing the fiber diameter. Nanofibers are very flexible due to their
nanoscale diameters.
14.1.4 Cohesiveness
Fibers tend to hold together during processing due to the longitudinal contour or
the cross-sectional shape that allows the fibers to stick together. The ability of
fibers to stick together during processing is called cohesiveness. The cohesiveness
of fibers is required in some applications, but it can be undesirable in others. For
example, during yarn manufacturing, the cohesiveness is essential for fibers to be
held together and be twisted into yarns, which can be then processed into other
products. On the other hand, for composite application, the cohesiveness of fibers
could introduce weakness into the structure since it is difficult for resin to wet the
entire fiber surface if fibers stick together to form bundles.
Natural polymer fibers have unique structures that allow them to stick together.
For example, the crimp nature of cotton and the surface scales of wool provide
these fibers with good cohesiveness. However, as-spun synthetic polymer fibers
often have smooth surface and simple physical structure. The cohesiveness of syn-
thetic polymer fibers can be improved by increasing the aspect ratio, reducing the
fiber diameter, changing the circular cross-sectional shape into a more complex
structure, and introducing the crimp shape to the longitudinal direction. Similarly,
the cohesiveness of inorganic fibers and nanofibers also can be controlled.
Secondary Properties 255
14.1.5 Uniformity
The uniformity of fibers affects the properties and quality of end-products. The
strength of an individual fiber is determined by its weakest point. For synthetic
polymer fibers, man-made inorganic fibers and nanofibers, the uniformity can be
controlled during the fiber formation process to minimize the structural irregulari-
ties. However, it is difficult to control the uniformity of natural fibers since the
structure of these fibers is affected by many environmental factors. Although the
uniformity of an individual natural fiber is uncontrollable, it is possible to im-
prove the overall uniformity of fibers in end-products by blending natural fibers
from many different batches.
14.2 SECONDARY PROPERTIES

Secondary properties are not essential for the conversion of fibers into the end-
products. However, many of them are needed for the products to be accepted by
the customers.
14.2.1 Physical Shape

The physical shape of fibers affects their appearance, luster, hand and feel, cohe-
siveness, and many other properties. The shape of fibers can be examined both
in longitudinal and cross-sectional directions. The simplest longitudinal shape of
fibers is straight. But many natural polymer fibers, such as cotton and wool, have
crimp longitudinal shape. Synthetic nanofiber fibers can be made to have straight,
crimp, coiled, or spiral shape.
The cross-sectional shape of fibers is more complex. Many natural polymer
fibers have unique cross-sectional shapes. For example, the cross-section of dry
cotton is kidney-shaped, while that of degummed silk is nearly a triangle. These
cross-sectional shapes are controlled by genetic codes, and human has limited
influence. However, the cross-sectional shapes of synthetic polymer fibers, inor-
ganic fibers, and nanofibers can be manipulated by controlling the fiber formation
processes. The cross-sectional shapes of these fibers range from circular to oval,
triangular, dog bone, trilobal or multilobal, hollow, etc.
Compared with fibers with other types of cross-sections, circular fibers have
the smallest surface area, and can pack closely together into yarns, composites,
or other structures. Circular fibers often are lustrous and have soft and smooth
handle. On the other hand, dog bone-shaped and flat cross-section fibers have a
harsher, less smooth feel. Trilobal and multilobal fibers exhibit an increased luster
and have excellent ability to obscure the object placed beneath them because they
can reflect the light not only from the surface but also from one lobe to another.
Hollow fibers have low packing density and provide greater bulk with less weight.
In addition, hollow fibers have high absorbency due to their large surface area.
In addition to the longitudinal and cross-sectional shapes, the surface structure

of fibers varies. Many fibers have smooth and even contour. However, some fibers
are rough, uneven, or have unique surface features. For example, wool fibers are
covered with many scales that make them cling closely together.
14.2.2 Density
The mass density or density of a material is its mass per unit volume:
m
ρ= (14.1)
V
where ρ is the density, m the mass, and V the volume. In some cases, the density is
expressed by a dimensionless quantity, specific gravity or relative density, which
is obtained by dividing the density of a fiber by that of a standard material, such
as water. A specific gravity or relative density of less than one means that the fiber
floats on water. Density is an intensive property, which means that increasing the
amount of a fiber does not increase its density.
The densities of fibers mainly are determined by the types and packing of the
atoms used for forming the fibers. Table 14.2 shows the densities of some synthet-
ic polymer, natural polymer, and inorganic fibers. Polymer fibers have densities
ranging from 0.90 to 1.60 g/cm3 since they are composed of light atoms such as
carbon, hydrogen, nitrogen, and oxygen. Polyolefin fibers have the lowest densi-
ties. For example, the density of polypropylene fibers is between 0.90–0.91 g/cm3,
which is lower than that of water. Inorganic fibers have higher densities either due
to the heavy atoms (e.g., Si in glass fibers) or because of the compactly packed
structures (e.g., glass and carbon fibers).
14.2.3 Modulus, Elongation, Elastic Recovery, and Resilience

As discussed in Section 14.1.2, strength probably is the most essential mechani-
cal property of fibers. Fibers must have adequate strength to survive the fabrica-
tion process and be made into end-products. Other mechanical properties, such as
modulus, elongation, elastic recovery, resilience, etc., may be less essential, but
they are needed for the products to be accepted by the customers. Most applica-
tions have basic requirements for these secondary mechanical properties. In ad-
dition, although secondary mechanical properties are less essential than the fiber
strength, they still may affect the processing parameters that are important for the
production of the end-products.
The modulus of a fiber is a measure of its stiffness. In tensile tests, the fiber
modulus is defined as the ratio of the tensile stress over the tensile strain in the
range of stress in which Hooke’s law holds. Experimentally, the fiber modulus is
determined by the slope of a stress-strain curve. The fiber modulus is an intrinsic
Table 14.2. Densities of some synthetic polymer, natural polymer, and inorganic fibers.
Fiber Density (g/cm3)

Polyester 1.22–1.39
Acrylic 1.14–1.19
Nylon 6,6 1.14
Kevlar 1.43–1.47
Spectra 0.97
Cotton 1.52–1.56
Flax 1.40–1.50
Hemp 1.40–1.48
Jute 1.30–1.50
Ramie 1.51–1.55
Wool 1.30–1.32
Silk 1.25–1.36
Acetate 1.30–1.32
Rayon 1.49–1.53
Inorganic Fibers
Carbon 1.75–2.00
Glass 2.48–2.60
University Press, 1996.; Mwaikambo, L.Y., African Journal of Science and Technology, 7,
120–133, 2006.; Tortora, P.G., Understanding Textiles, Fourth Edition, Macmillian Publishing
Company, 1992.
property. Most polymer fibers have significantly greater moduli than the corre-
sponding bulk polymers since polymer fibers have higher degrees of crystallinity
and molecular orientation. However, the moduli of polymer fibers typically are
lower than those of inorganic fibers, such as carbon fibers.
Breaking elongation and elastic recovery are closely related to each other.
Elongation measures the percentage change in length when a fiber is stretched.
Breaking elongation, also called strain-at-break, is the elongation that occurs to
the point when the fiber breaks. Elastic recovery is the ability of a fiber to return
to its original length after the removal of stretching force and is measured as the
percentage of return from elongation towards the original length. In many ap-
plications, breaking elongation and elastic recovery must be considered together.
Table 14.3 shows the breaking elongation and elastic recovery of some synthetic
Table 14.3. Breaking elongation and elastic recovery of some

synthetic polymer, natural polymer, and inorganic fibers.
Elastic Recovery at %
Fiber Breaking Elongation (%)
Elongation (%)
Polypropylene 15–50 100 @ 2%
Polyester 10–60 85–99 @ 2%
Acrylic 25–55 80–99 @2%
Nylon 15–40 100 @ 8%
Kevlar 2–5 —
Spectra 3–4 —
Cotton 5–10 75 @ 2%
Flax 1.3–3.3 65 @ 2%
Jute 1.3–3.0 74 @ 1%
Ramie 3.0–7.0 52 @ 2@
Wool 20–45 99 @ 2%
Silk 10–25 92 @ 2%
Acetate 25–35 94 @ 2%
Triacetate 25–35 90–92 @ 2%
Rayon 15–28 82–98 @ 2%
Inorganic Fibers
Carbon 0.7–2.5 100 @ 1%
Glass 2–5.4 100 @ 2%
University Press, 1996; Morton, W.E., et. al., Physical Properties of Textile Fibres, Fourth
polymer, natural polymer and inorganic fibers. In general, polymer fibers exhibit
high breaking elongation values. Inorganic fibers have small breaking elongation,
but higher elastic recovery.
Resiliency is the ability of a fiber to return to its original shape after bending,
compression, or similar deformation. Resiliency is different from elastic recovery,
which is the fiber’s ability to return to the original length after stretching. Good
resiliency is important for some applications such as carpets. Polyester, wool and
nylon have good resilience and they often are used for making carpets. However,
when soft fabrics are needed, good resilience is not desirable. In this case, low-
resilience fibers such as cotton and rayon are the better choice.
14.2.4 Thermal Properties

Common thermal properties include glass transition temperature, melting tem-
perature, decomposition temperature, heat capacity, thermal conductivity, dimen-
sional stability, and flammability.
Glass transition temperature is the temperature at which the polymer chain
segments in a fiber start to gain mobility. The movement of chain segments only
occurs in the amorphous phase. Hence, above the glass transition temperature,
polymer fibers still are solid if the temperature still is lower than the melting tem-
perature and decomposition temperature. The melting temperature is the template
at which the crystalline phase of the fiber melts. Above melt temperature, the sec-
ondary bonds between the polymer chains cannot hold the chains together in the
ordered crystalline phase, and polymer fibers can no longer keep their fiber form.
The decomposition temperature is the temperature at which the primary bonds
break. Above the decomposition temperature, the breaking of primary bonds leads
to the formation of nonpolymeric products or a new structure completely different
from the original polymer.
Heat capacity is the amount of energy required to raise the temperature of a
fiber by one degree, while the thermal conductivity is the property of a fiber’s
ability to conduct heat. Dimensional stability is the ability of a fiber to keep its di-
mensions while raising the temperature. Flammability is about how easily a fiber
will burn or ignite, causing fire or combustion.
All these properties are important, and different applications have different
requirements for them.
14.2.5 Electrical Properties

Electrical conductivity is the ability of a fiber to carry or transfer electrical charg-
es. In general, polymer fibers have low conductivities. This is because all elec-
trons are either held to the atom nuclei or shared in the covalent bonds. Polymer
fibers with low conductivities can easily build up electrical charges and they cling
or even produce electrical shocks. The conductivity of inorganic fibers varies sig-
nificantly. Glass fibers have low conductivity when they are dry, but carbon fibers
are highly conductive.
Although polymer fibers generally are insulators, their conductivity could in-
crease significantly when they absorb moisture. Therefore, in some production
processes, the relative humidity is controlled to reduce the buildup of static charg-
es. However, not all polymer fibers absorb moisture. Fibers that do not absorb
moisture can be surface-treated to reduce the static charge accumulation by in-
creasing their ability to absorb and hold moisture. Conductive additives also can
be added to synthetic polymer fibers to directly increase the electrical conductivity.
14.2.6 Color and Optical Properties

Synthetic polymer fibers typically are white or off-white when they are manufac-
tured. Some special synthetic polymer fibers have other types of color. For example,
as-spun Kevlar fibers typically are yellow. In addition, synthetic polymer fibers with
different colors also can be produced by adding pigments during fiber formation or
by dyeing after the fibers are formed. Natural cellulose and protein fibers exhibit
great color difference by nature, and their color also can be modified by dyeing.
In addition to color, luster also is an important secondary property. Luster re-
fers to the gloss, sheen, or shine that a fiber has and is based on amount of light
reflected by the fiber. Lust may be necessary in some applications, but undesirable
in others. Synthetic polymer fibers may have bright luster if they have circular
cross-section. The luster of synthetic polymer fibers decreases if they are made
with complicated cross-sectional or surface features. Synthetic polymer fibers
also can have reduced luster by adding pigments including TiO2. The luster of
natural cellulose and protein fibers varies. For example, silk has a very high luster,
but cotton has low luster prior to treatments.
14.2.7 Environmental Properties

Fibers are exposed to a variety of chemical and environmental conditions during
processing (dyeing and other surface treatments), care (laundering and cleaning),
and use (air pollution and UV radiation). Among the different chemical and envi-
ronmental conditions, chemicals such as acids, alkalis and organic solvents have
the most rapid effect since they can directly swell, dissolve and degrade some
fibers. Table 14.4 shows the comparative resistance of different fibers to acids,
alkalis, and organic solvents. Polyester, acrylic and carbon fibers are resistant to
most chemicals. However, other fibers can be harmed or weakened by at least
some of the chemicals. In addition to chemical conditions, the ultraviolet (UV)
radiation can degrade many fibers by breaking their primary bonds. Inorganic
fibers typically have good resistance to UV radiation. Among various polymer
fibers, polyester and acrylic fibers have relatively good resistance to UV radiation,
but wool and silk fibers degrade under UV radiation.
Some fibers such as cotton support the growth of biological organisms, such
as molds or mildew. Biological organisms grow fast, especially in warm, dark
and damp environments, and they could deteriorate the fibers. However, there are
fibers (e.g., synthetic polymer and inorganic fibers) that do not support the growth
of biological organisms. Special treatments also can be given to inhibit or kill
biological organisms on fibers.
Natural cellulose and protein fibers are food sources for some insects. Carpet
beetles, clothes moths, and silverfish are examples of common insect pests that at-
tack these fibers. Insect grubs do not surface on articles that are mobile, frequently
washed, or regularly vacuumed and cleaned, and hence one important approach to
Table 14.4. Resistance of fibers to acids, alkalis, and organic solvents.
Resistance to Resistance to Organic

Fiber Resistance to Acids
Alkalis Solvents
Polypropylene Resistant Resistant Harmed by chlorinated
hydrocarbons
Polyester Resistant Resistant Resistant
Acrylic* Resistant to most Resistant to weak Resistant
Nylon Harmed Resistant Harmed by phenol and
formic acid
Cotton Harmed Resistant Resistant
Flax Harmed Resistant Resistant
Wool Resistant Harmed Resistant
Silk Harmed by mineral Harmed Resistant
acids
Acetate Weakened Little effect Harmed by acetone, phe-
nol, and chloroform
Rayon Harmed Resistant to weak Resistant
Inorganic Fibers
Carbon Resistant to most Resistant Resistant
Glass Resistant Harmed Resistant
*Acrylic is a fiber made from a copolymer composed of at least 85% by acrylonitrile units, -CH2-
CH(CN)-.
Sources: Morton, W.E., et. al., Physical Properties of Textile Fibres, Fourth Edition, Woodhead
Publishing Limited, 2008.; Tortora, P.G., Understanding Textiles, Fourth Edition, Macmillian
prevent damage from insects is to provide appropriate care. Chemical treatments

also can be used to provide the fibers with the ability to resist insects.
14.2.8 Other Properties

In addition to the abovementioned properties, there are other secondary proper-
ties that are important for practical applications of fibers. Among them, moisture
regain is an important secondary property. Moisture regain is the ability of a dry
fiber to absorb moisture, and can be calculated by:
wet weight − dry weight

moisture regain = ×100 (14.2)
dry weight
Table 14.5. Moisture regains of some synthetic polymer, natural polymer,

and inorganic fibers. Data obtained at 20°C and 65% relative humidity.
Fiber Moisture Regain (%)

Polypropylene 0
Polyester 0.4
Acrylic 1.3–2.5
Nylon 3.5–5.0
Kevlar 2.7–7
Cotton 7.0–10.0
Flax 7.0–12.0
Jute 12.0–13.7
Ramie 6.0–8.5
Wool 13.6–18.0
Silk 10.0–11.0
Acetate 6.0–7.0
Rayon 10.7–16.0
Inorganic Fibers
Carbon 0
Glass 0
Publishing Limited, 2008.; Mwaikambo, L.Y., African Journal of Science and Technology, 7,
120–133, 2006.; Tortora, P.G., Understanding Textiles, Fourth Edition, Macmillian Publishing
Company, 1992.
where the dry weight is the weight of a fiber dried thoroughly to its bone-dry state
in a drying oven, and the wet weight is the weight of the fiber at standard condi-
tions of 20°C and 65% relative humidity.
Table 14.5 shows moisture regains of some synthetic polymer, natural poly-
mer, and inorganic fibers. The ability of fibers to form hydrogen bonds has a sig-
nificant effect on the moisture regain. Wool, silk, cotton and other natural polymer
fibers can form a large amount of hydrogen bonds with water molecules and they
have high moisture regains. Polypropylene and polyester fibers do not form hy-
drogen bonds with water molecules and they have no or low moisture regain. The
packing of polymer chains in fibers also affects the moisture regain. In general,
with increase in crystallinity and molecular orientation, the moisture regain de-
creases since fewer water molecules can enter the fiber. The creation of a porous
structure is an effective approach to increase the moisture regain. For example,
References 263
regular solid carbon fibers have almost no moisture regain at standard conditions,
but activated carbon fibers with micropores and large surface areas can have high
moisture regains of greater than 30%.
Another important secondary property is fiber friction. On one hand, without
friction, fibers cannot be processed into useful products. On the other hands, fiber
friction also could cause equipment failure and/or damage to the fiber surface if it
is not controlled. During manufacturing, lubrication often is applied to reduce the
fiber friction to protect equipment and fiber surface.
REFERENCES
[1] Bueno, M.A., Aneja, A.P., and Renner, M., “Influence of the Shape of Fiber
Cross Section on Fabric Surface Characteristics”, Journal of Materials
Science, 39, 557–564, 2004.
[4] Gohl, E.P.H., and Vilensky, L.D., Textile Science: An Explanation of Fibre
Properties, Longman Group Ltd., 1981.
[7] Matsudaira, M., Tan, Y., and Kondo, Y., “The Effect of Fibre Crosssectional
Shape on Fabric Mechanical Properties and Handle”, Journal of the Textile
Institute, 84, 376–386, 1993.
[8] Morton, W.E., and Hearle, J.W.S., Physical Properties of Textile Fibres,
Fourth Edition, Woodhead Publishing Limited, 2008.
2006.
[11] Pizzuto, J., Fabric Science, Tenth Edition, Fairchild Publications Inc, 2010.
PROBLEMS
(1) Describe the definitions of primary and secondary properties of fibers.
(2) Why is the aspect ratio of a fiber considered as a primary property?
(3) What is the difference between strength and tenacity?
(4) You are given a filament whose length is 75 cm and mass is 9 × 10–2 mg.
Calculate its denier.
(5) Out of tenacity, luster, color, moisture regain, uniformity, density, aspect
ratio, elongation, and flexibility, which are secondary properties and why
are they considered secondary properties?
(6) Describe the physical shapes of cotton, silk, and wool fibers.
(7) The density of polyolefin fibers is lower than that of water? Is this good or
bad?
(8) What is resiliency?
(9) Name five thermal properties.
(10) Describe the benefits and penalties when a fiber has a very high moisture
regain.
CHAPTER 15
Mechanical Properties of Fibers
Mechanical properties probably are the most important properties of fibers. Fibers
must possess sufficient mechanical properties so they can be converted into use-
ful products, such as textiles, composites, etc. The mechanical properties of fibers
also limit the performance potential that can be achieved by these products. There
are many different types of mechanical properties, including tensile, torsional,
bending, and compressional properties. Among them, tensile properties are the
most widely studied for fibers, probably because of their unique shape. However,
other types of mechanical properties also are important. This chapter begins with
the basic definitions of Hooke’s law, stress, strain, and tensile, bulk and shear
moduli, followed by a detailed discussion on tensile, torsional, bending, and com-
pressional properties of fibers.
15.1 BASIC DEFINITIONS

15.1.1 Hooke’s Law
Fibers are subject to forces or loads when they are processed or in use. It is im-
portant to understand how fibers deform (elongate, compress, twist) or break as a
function of applied force. When the force is relatively small and does not exceed
the elastic limit, fibers obey the Hooke’s law, i.e., the law of elasticity. Hooke’s
law was discovered by the English physicist Robert Hooke in 1660, and it states
the deformation of an object is directly proportional to the applied force. In addi-
tion, the object returns to its original shape and size upon the removal of the force.
One common form of Hooke’s law is the spring equation (Figure 15.1). When
a force is applied to a spring, the end of the spring is displaced from its equilibri-
um position, and at the same time a restoring force is generated and tends to re-
store the spring back to its equilibrium. The spring equation relates the restoring
265
266 Mechanical Properties of Fibers
Figure 15.1. Hooke’s law describes the relationship between the restoring force of a spring
and the displacement of the spring’s end from its equilibrium position.
ur
u 
force ( Fr ) of a spring to the displacement ( x ) of the spring’s end from its equi-
librium position by a spring constant, k.
ur
u r
Fr = −kx (15.1)
The negative sign indicates the restoring force of the spring is in the opposite di-
rection
uur to the displacement. Similarly, the relationship between the applied force
( Fa ) and the displacement can be expressed as:
uur r
Fa = kx (15.2)
There is no negative sign in Equation 15.2 because the applied force is in the same
direction with the displacement.
Hooke’s law is useful in describing how a particular object deforms under a
certain force. However, it does not take into consideration the geometrical shape
and size of the object and is not convenient for describing the mechanical prop-
erties of fibers. For example, under a given force, a long piece of fiber always
deforms more than a shorter piece of the same fiber. Similarly, a thicker fiber can
sustain a larger force before breaking than a thinner fiber made of the same mate-
rial. Therefore, to compare the mechanical properties of fibers, stresses and strains
are defined to take out the effect of shape and size.
There are three basic types of stresses. If a fiber is stretched or compressed, it
is subjected to a tensile stress. If a force is applied over an entire surface of a fiber,
changing its volume, the fiber is said to be experiencing a bulk stress. Finally, if
Basic Definitions 267
the force is acting tangentially to the surface, causing it to twist, the fiber is subject
to a shear stress.
15.1.2 Tensile Stress, Strain and Modulus

Figure 15.2 shows a fiber with cross-sectional area A0 being subject to equal and
opposite tensile forces (F) at the ends. The tensile stress (σ) can be defined as the
ratio of the force to the cross-sectional area.
F
σ= (15.3)
A0
The SI unit of stress is newton per square meter (N/m2), which also is called the
pascal (Pa). However, a tensile stress of 1 pascal is extremely small, and hence the
unit of megapascal (MPa) often is used for convenience.
The tensile stress causes the elongation of the fiber. The fractional amount
a fiber elongates when it is subject to a tensile stress is called the tensile strain.
Mathematically, the tensile strain (ε) can be written as:
l − l0
ε= (15.4)
l0
where l0 is the original length of the unstressed fiber, and l is the length of the
stressed fiber. Strain is a dimensionless quantity since it is a ratio of length to
length, and it often is reported in percentage.
Figure 15.2. Deformation of a fiber under tensile force F.

According to the Hooke’s law, the amount of tensile strain experienced by a

fiber is directly proportional to the tensile stress. The tensile modulus (E) can then
be defined as:
σ
E= (15.5)
ε
Tensile modulus is also called Young’s modulus or elastic modulus. The unit
of tensile modulus also is pascal, or more often gigapascal (GPa) for the pur-
pose of convenience. Fibers with high tensile modulus are stiff and deform very
little under stress, whereas fibers with low tensile modulus are soft and deform
significantly.
15.1.3 True Stress and True Strain

The definitions of tensile stress and tensile strain are based on fixed reference
quantities, i.e., the original cross-sectional area and original length, respectively.
In many engineering applications, these definitions can be considered to be accu-
rate if the cross-sectional area and length of the fiber do not change substantially
while a tensile force is applied. As a result, tensile stress and tensile strain often
are called engineering stress and engineering strain. However, in some other situ-
ations, the cross-sectional area and the length of the fiber can change substantially.
In such cases, the stress and strain calculated based on original cross-sectional
area and original length cease to be accurate measures. To address this problem,
true stress and true strain are defined.
True stress (σT) is the ratio of the tensile force to the instantaneous cross-
sectional area:
F
σT = (15.6)
A
True stress can be related back to the engineering stress by assuming there is no
volume change in the fiber. Under this assumption:
A × l = A0 × l0 (15.7)
which leads to:
F F l F l0 + l − l0
σT = = × = × = σ (1 + ε) (15.8)
A A0 l0 A0 l0
True strain is the sum of all the instantaneous engineering strains:

l
dl l
εT = ∫d ε = ∫ = ln (15.9)
l0
l l0
True strain also can be related to the engineering strain by:
l l + l − l0
εT = ln = ln 0 = ln (1 + ε ) (15.10)
l0 l0
Under a small strain, true stress and strain are comparable to engineering stress
and strain. However, under large strains, the true stress and strain could be signifi-
cantly different from engineering stress and strain.
15.1.4 Specific Stress

The concept of tensile stress is simple and straightforward. However, the cross-
sectional areas of many fibers are not well-defined due to their unique cross-sec-
tional shapes. One useful parameter to describe the fiber cross-sectional fineness
is the linear density (μ), i.e., mass per unit length. For fibers with comparable
bulk densities, their cross-sectional area is proportional to the fiber linear den-
sity. Hence, the term specific stress (σs) often is used, especially in the textile
community:
F
σs = (15.11)
µ
Common units for specific stress are gram per denier or gram per tex. Denier is
the unit of fiber or yarn measurement equal to the weight in grams of 9,000 meters
of fiber or yarn, while tex is the unit of fiber or yarn measurement equal to the
weight in grams of 1,000 meters of fiber or yarn.
When specific stress is used in calculating the tensile modulus, the modulus
unit becomes gram per denier or gram per tex.
15.1.5 Poisson’s Ratio

When a fiber is subject to a tensile stress, it does not just deform in the direction
of the applied force, but also contracts in directions perpendicular to the force, as
shown in Figure 15.2. This phenomenon is called the Poisson effect. Poisson’s ra-
tio (ν), also called Poisson ratio or Poisson coefficient, is a measure of the Poisson
effect. Poisson’s ratio is the negative ratio of the transverse contraction strain to
longitudinal tensile strain.
εT
ν=− (15.12)
εL
where εT is the transverse contract strain and εL the longitudinal tensile strain.
Table 15.1 shows the Poisson’s ratios of some materials. An incompressible
material that does not change volume when elongates would have a Poisson’s
ratio of exactly 0.5. For example, rubber has a Poisson’s ratio of nearly 0.5.
Cork’s Poisson ratio is close to 0, indicating very little transverse contract when
stretched. Auxetic materials, such as polymer foams, have negative Poisson’s ra-
tios, and they become thicker in perpendicular directions when stretched. Most
other materials, including metals, ceramics, and polymers, have Poisson’s ratios
ranging from 0 to 5.
Poisson’s ratios shown in Table 15.1 are for isotropic materials that have iden-
tical structure and properties in all directions. Fibers, especially polymer fibers,
are anisotropic and their Poisson’s ratios deviate from the values reported in
Table 15.1.
15.1.6 Bulk Modulus

The second basic type of stress is bulk stress, i.e., hydrostatic pressure. Fibers
diminish in volume when exposed to a uniform hydrostatic pressure (P), as shown
in Figure 15.3. The isotropic compression due to the pressure can be described by
the engineering compression ratio (κ).
V − V0
κ= (15.13)
V0
where V0 is the original volume of the fiber before applying the pressure, and V the
volume of the fiber under pressure.
The bulk modulus (B) is a measure of resistance to the bulk stress or hydro-
static press, and it can be defined as:
P
B=− (15.14)
κ
Since the engineering compression ratio is dimensionless, the unit of bulk modu-
lus is the same as the hydrostatic pressure, i.e., pascal.
For isotropic materials, the bulk modulus has a simple relationship with tensile
modulus and Passion’s ratio:
E
B= (15.15)
3 (1 − 2υ )
Table 15.1. Poisson’s ratios of materials.
Material Poisson’s Ratio
Gold 0.42
Copper 0.33
Magnesium 0.35
Titanium 0.34
Stainless Steel 0.30–0.31
Glass 0.24
Concrete 0.20
Clay 0.30–0.45
Sand 0.20–0.45
Polystyrene 0.33
Polymethyl Methacrylate 0.33
Nylon 6,6 0.33
Rubber 0.50
Cork ca. 0.00
Auxetics negative
Sources: Boresi, A.P., et. al., Advanced Mechanics of Materials, Fifth Edition, John Wiley &
Sons, Ltd., 1993.; Ward, I.M., et. al., Mechanical Properties of Solid Polymers, Third Edition,
John Wiley & Sons, Ltd., 2013.
Figure 15.3. Deformation of a fiber under hydrostatic pressure P.

However, fibers are typically anisotropic, and do not follow Equation 15.15.

Due to the molecular orientation, the ratio of tensile modulus to bulk modulus for
most fibers is greater than what is predicted by Equation 15.15.
15.1.7 Shear Modulus

In addition to tensile stress and bulk stress (hydrostatic pressure), shear stress is
another basic type of stress. There are different ways to apply shear stress to a fi-
ber, which will be discussed later in this chapter. Here, for the purpose of simplic-
ity, the deformation of a cubic object under shear stress is discussed. Figure 15.4
shows the deformation that takes place when a shear force is applied parallel to
one face of a cubic object while the opposite face is held fixed by another equal
force. The face that experiences the shear force is named arbitrarily as the x, y face
in a Cartesian system. The shear stress (τxy) then can be defined as the force (F)
applied parallel to the face divided by the area (Axy) of this face:
F
τ xy = (15.16)
Axy
The shear strain (γxy) is defined as the deformation (δ) parallel to the direc-
tion of the shear force, divided by the original length (l0) perpendicular to this
direction:
δ
γ xy = (15.17)
l0
Shear modulus is the ratio of shear stress to shear strain:
τ xy
G= (15.18)
γ xy
Figure 15.4. Deformation of a cubic object under shear force F.

Tensile Properties 273
Similar to bulk modulus, the shear modulus of isotropic materials also has a sim-
ple relationship with tensile modulus and Passion’s ratio:
E
G= (15.19)
2 (1 + υ )
The relationship between shear modulus, tensile modulus and Poisson’s ratio of
anisotropic fibers deviates from Equation 15.19. In general, the ratio of tensile
modulus to shear modulus for most anisotropic fibers is greater than what is pre-
dicted by Equation 15.19.
15.2 TENSILE PROPERTIES

In many practical applications, tensile properties are the most important mechani-
cal properties of fibers since they typically are under tension or complex stress
states that include tension. This section focuses on the typical stress-strain behav-
ior and factors that affect the stress-strain behavior of polymer fibers. The elastic
recovery of polymer fibers also is discussed.
15.2.1 Ideal Stress-Strain Behavior

The relationship between the tensile stress and tensile strain that a fiber displays
is known as that fiber’s stress-strain behavior. If a fiber obeys the Hooke’s law, its
tensile stress is directly proportional to the tensile strain, up to the point of failure,
where the fiber breaks without yielding or plastic deformation (Figure 15.5). In
this case, the fiber exhibits ideal stress-strain behavior.
The ideal stress-strain curve shown in Figure 15.5 provides important infor-
mation: (i) the stress to break, which also is called the strength; (ii) the tensile
modulus, which is a measure of stiffness and is given by the slope of the curve,
i.e., stress divided by strain; and (iii) the energy to break, which is proportional
to the area under the curve and is a measure of toughness. Strength is an extrinsic
property, which means it depends on the weakest point of the fiber. Defects or
impurities can significantly reduce the strength of a fiber. However, modulus is
an intrinsic property and is less sensitive to defects and impurities. Toughness is a
measure of the amount of work required to break a fiber, and tough fibers are the
opposite of brittle fibers.
Figure 15.6 shows the ideal stress-strain curves of inorganic and polymer fi-
bers. Inorganic fibers typically have higher strength and modulus, but they break
at lower strain and are more brittle.
15.2.2 A Simple Model

A simple model can be used to describe the ideal stress-strain behavior. Figure 15.7
shows a one-dimensional array of atoms held together by chemical bonds. This
array can be used to represent part of an ideal material that has identical rows of
atoms placed adjacent to one another. When a tensile force is applied to this mate-
rial, the atoms are displaced from their equilibrium positions, creating a restor-
ing force that is equal, but opposite to the applied tensile force. The mechanical
response of the material is governed by the internal energy, i.e., the sum of the
Figure 15.5. Ideal stress-strain behavior.
Figure 15.6. Ideal stress-strain curves of inorganic and polymer fibers.

Figure 15.7. An array of atoms under tensile force.
potential energy and the kinetic energy. The kinetic energy is associated with the
vibrations of the atoms around their equilibrium positions and can be neglected
while considering the tensile properties of the material. The potential energy is
associated with the chemical bonds that hold the atoms together in the array, and
it determines the mechanical response of the material under the applied tensile
force.
Figure 15.8A shows the potential energy for a pair of atoms. There is an
equilibrium position, at which the potential energy is the lowest. The two atoms
repel strongly when their distance is smaller than the equilibrium distance, but
they attract to each other when their distance becomes larger than the equilib-
rium distance. When three or more atoms are involved, the curve for the potential
energy changes. However, the concept of the equilibrium position still is valid
(Figure 15.8B). Under the applied tensile stress, the potential energy (U) caused
by the displacement (x) of atoms from their equilibrium positions is:
1 2
U= kx (15.20)
2
where k is the force constant. When the displacement is relatively large, there are
some deviations, and Equation 15.20 can be rewritten by using a power series:
1 2 1 1
U= kx + k ' x 3 + k " x 4 +… (15.21)
2 3 4
As a result, the restoring force can be calculated by:
dU
force = = kx + k ' x 2 + k " x 3 +… (15.22)
dx
When the displacement is small, Equation 15.22 becomes:
force = kx (15.23)
Therefore, this simple model indicates the mechanical response of the material
follows the Hooke’s law and exhibits the ideal stress-strain behavior if the tensile
strain is small. However, deviations occur under larger strains.
Figure 15.8. Potential energy (A) for a pair of atoms, and (B) of the middle atom in an
array of three atoms.
15.2.3 Deviations from Ideal Stress-Strain Behavior

According to the simple model described in the previous section, the mechanical
response of the material deviates from the ideal stress-strain behavior at large
strains since the potential energy needs to be expressed using a power series
(Equation 15.21). In fibers, the material structure is far from perfect and the ar-
ray of atoms shown in Figure 15.7 is not an approximate description of the fiber
structure. For example, most polymer fibers have both amorphous and crystalline
phases. In each phase, the molecular orientation is not perfect. As a result, at both
longitudinal and transverse directions of a polymer fiber, there is a combination of
primary and secondary bonds and the potential energy cannot be simply described
using Equation 15.21.
Figure 15.9 shows a typical stress-strain curve of a polymer fiber, in which
the molecular orientation has not been fully developed. The stress-strain curve is
composed of elastic and plastic regions. In the elastic region, the strain is small
and the fiber deforms more-or-less elastically. The deformation in the elastic re-
gion is largely reversible. The tensile modulus of the fiber is obtained from the
initial slope of the stress-strain curve in the elastic region. When the tensile stress
reaches the yield strength, the fiber yields and the stress-strain curve enters the
Figure 15.9. Typical stress-strain curve of an un-oriented polymer fiber.
plastic region. At the yield point, a “neck” is formed in the fiber and the deforma-
tion becomes irreversible. There is an actual decrease in the tensile stress, i.e., the
so-called strain softening, in the initial stage of the plastic region. The deforma-
tion then continues with a slow increase in the tensile stress. At the end of the
plastic region, the fiber appears to gain resistance to the stress more rapidly and
ultimately breaks at a higher stress. The tensile strength basically is the stress at
the point of break when the fiber experiences its maximum strain. The maximum
strain also is called strain-at-break or breaking elongation.
The stress-strain curve of a polymer fiber depends on its structure. Figure 15.10
compares the stress-strain curves of un-oriented amorphous, semicrystalline, and
elastomer polymer fibers. Most polymer fibers are semicrystalline. However,
amorphous polymer fibers could be obtained prior to drawing for polymers with
slow crystallization kinetics. At temperatures well below the glass-transition tem-
perature (Tg), both amorphous and semicrystalline polymer fibers do not display
yielding or plastic deformation. They have high modulus and high strength, but
they are brittle. Both amorphous and semicrystalline polymer fibers yield and
exhibit plastic deformation at high strains when the temperature is close to or
above Tg. Although the modulus and strength decrease, the fibers become tougher
because of the yielding and plastic deformation. Elastomer fibers, such as some
silicone and polyurethane fibers, have the smallest modulus and strength. But they
have the largest strain-at-break. This is not caused by yielding or plastic deforma-
tion, but is an inherent non-linear elastic property of elastomeric materials.
The stress-strain curves shown in Figures 15.9 and 10 are for as-spun polymer
fibers, in which molecular orientation has not been fully developed. However,
Figure 15.10. Typical stress-strain curves of un-oriented amorphous, semicrystalline, and

elastomeric polymer fibers.
commercial polymer fibers have high molecular orientation, and hence more
polymer chains are aligned along the fiber longitudinal direction to bear the ap-
plied load. These fibers have higher modulus and strength than un-oriented poly-
mer fibers. But their strain-at-break is lower. Figure 15.11 shows a typical stress-
strain curve of a polymer fiber with high molecular orientation. The fiber does not
present a yield point. However, the stress-strain curve still deviates from the ideal
behavior, especially at relatively high strains.
15.2.4 Response of Polymer Chains

To understand the stress-strain behavior of polymer fibers, it is important to know
how the polymer chains respond to the tensile strain. Figure 15.12 divides the
stress-strain curve of an un-oriented amorphous fiber into three stages. Stage I
is the same as the elastic region. The plastic region of the curve is divided into
Stages II and III. The following is how the polymer chains respond to the tensile
strain in these three stages.
Stage I: The strain mainly is caused by the changes in the chain length and
angle. The deformation is reversible and approximately follows the Hooke’s
law.
Stage II: The polymer chain segments start to move and orientation starts to
occur. The characteristic of this stage is strain softening, i.e., the stress de-
creases with increase in strain. (The stain softening phenomenon disappears
if true stress and true strain are used in the stress-strain curve. However, the
Figure 15.11. Typical stress-strain curve of a molecularly-oriented polymer fiber.
Figure 15.12. Three-stage stress-strain curve of an un-oriented amorphous polymer fiber.
movement and orientation of polymer chains remain the same.) The defor-
mation is largely irreversible since Stage II is in the plastic region.
Stage III: Polymer chain segments continue to move and cold flow occurs.
The deformation is irreversible. Due to the molecular orientation and the
formation of new interactions such as entanglements, the stress starts to

increase again until the fiber breaks.
The stress-strain curve of an un-oriented semicrystalline fiber also can be di-

vided into three stages (Figure 15.13). The following is how the polymer chains
respond to the tensile strain in these three stages.
Stage I: The amorphous phase is deformed first and this is realized by

changing the chain length and angle (Figures 15.14A and B). The deforma-
tion of the crystalline phase occurs when the tie molecules between the
adjacent crystallites are highly extended. The deformation of the crystalline
phase starts with the increase of the crystallite thickness (Figure 15.14C),
followed by the tilting of the crystalline phase (Figure 15.14D). The fiber
deformation in this stage is largely reversible and approximately follows the
Hooke’s law.
Stage II: The polymer chain segments in the amorphous phase start
to move and large-scale molecular orientation occurs. Large crystal-
lites may be unraveled under stress to form smaller fibril-like crystallites
(Figure 15.14E). These newly formed fibrils still are attached to each other
by tie molecules, and they slide against each other to align along the tensile
direction (Figure 15.14F). The deformation in Stage II is irreversible.
Figure 15.13. Three-stage stress-strain curve of an un-oriented semicrystalline polymer fiber.

Stage III: The sliding and orientation still occur, but are largely limited by
tie molecules or entanglements. The stress increases until the fiber breaks.
The deformation in this stage is irreversible.
These are just some typical responses that may be observed in an un-oriented
semicrystalline fiber while a tensile stress is applied. In reality, the chemical and
crystalline structures affect how the polymer chains behave under stress. As a
result, for some un-oriented semicrystalline polymer fibers, the responses of poly-
mer chains under tensile stress may be different from what is discussed above.
There may not be significant yielding or plastic deformation in molecularly
oriented fibers. However, the stress-strain curve of molecularly oriented fibers
still is nonlinear and can be divided into two stages (Figure 15.15):
Figure 15.14. Structural evolution of an un-oriented semicrystalline fiber under tensile

stress. (A) Before deformation, (B) deformation of amorphous phase, (C) deformation of
crystalline phase, (D) tilting of crystalline phase, (E) unraveling and sliding of crystalline
phase, and (F) final plastic deformation. For simplicity, tie molecules are not shown.
Stage I: The stress-strain curve in this stage is a straight line and is revers-
ible. The deformation follows the Hooke’s law and is caused mainly by the
changes in chain length and angle.
Stage II: The stage is not apparent if the degree of molecular orientation is
very high. However, Stage II can be clearly observed in polymer fibers that
have low or moderate molecular orientation. In this stage, polymer chain
segments could move under stress and the sliding between the fibril-like
crystallites could occur. However, the tensile stress continues to increase
until the fiber breaks.
15.2.5 Stress-Strain Behavior of Commercial Fibers

Although un-oriented amorphous and semicrystalline polymer fibers are impor-
tant and they often are the intermediate products of molecularly oriented fibers,
most commercial polymer fibers are molecularly oriented, except for a few elas-
tomeric polymer fibers. The molecular orientation in natural polymer fibers oc-
curs by itself under the control of genetic codes. For synthetic polymer fibers,
molecular orientation is introduced purposely by high-speed spinning or drawing
to enhance the mechanical properties of these fibers so they can be converted into
useful final products.
Based on the stress-strain behavior, molecularly oriented polymer fibers can
be divided into three groups: high modulus-high tenacity (HM-HT) fibers, tough
Figure 15.15. Two-stage stress-strain curve of a molecularly-oriented polymer fiber.

fibers, and regular fibers. Figure 15.16 shows the stress-strain curves of these
three groups of polymer fibers. For comparison, the stress-strain curves of in-
organic and elastomeric polymer fibers also are shown. Inorganic fibers have
high modulus and strength, but they are brittle. HM-HT polymer fibers also have
high modulus and strength. In Figure 15.16, the modulus of inorganic fibers is
higher than that of HM-HT polymer fibers. But that relationship is relative since
some commercial HM-HT polymer fibers can have higher modulus than typical
inorganic fibers. Tough fibers typically are synthetic and have high strength, but
relatively low strain-at-break. Regular fibers, including natural and regenerated
polymer fibers and some synthetic polymer fibers, have moderate strength and
strain-at-break. Elastomeric polymer fibers have low strength, but extremely high
strain-at-break.
Figure 15.17 shows typical stress-strain curves of some commercial synthetic
polymer, natural polymer, and inorganic fibers. Carbon and Kevlar have the high-
est strengths and moduli, but their strains-at-break are the lowest. Polyester and
nylon 6,6 fibers with high degrees of molecular orientation and crystallinity are
considered as high-strength fibers in certain applications, but their strengths and
moduli are lower than those of carbon and Kevlar fibers. Regular polyester and
nylon 6,6 fibers have much lower strengths and moduli, but their strains-at-break
are much higher. Acrylic fibers often have lower strength than regular polyester
and nylon 6,6 fibers, but not necessarily lower modulus or stain-at-break. Among
the three natural polymer fibers shown in Figure 15.17, silk and cotton have
higher strengths and moduli than wool fibers, but wool fibers have the largest
Figure 15.16. Stress-strain curves of commercial inorganic fibers, HM-HT polymer fibers,

tough polymer fibers, regular polymer fibers, elastomeric polymer fibers.
Figure 15.17. Typical stress-strain curves of some synthetic polymer, natural polymer, and
inorganic fibers. Data obtained at 20°C and 65% relative humidity. Acrylic fibers are made
from copolymers composed of at least 85% by acrylonitrile units, -CH2-CH(CN)-.
strain-at-break. Polyurethane fibers are elastomeric with a strain-at-break up to

500%.
The stress-strain curves shown in Figure 15.17 just present the general prop-
erty trends. In reality, the curve shapes may vary from fiber to fiber. For example,
the stress-strain curves of natural polymer fibers are affected by the farming con-
ditions of the plants or animals. Depending on the applications, the processing
conditions of synthetic polymer fibers can be selectively adjusted to produce dif-
ferent physical structures and mechanical properties. Table 15.2 shows typical
tensile strength, modulus and strain-at-break values of some commercial fibers.
15.2.6 External Factors Affecting the Stress-Strain Behavior

External factors that affect the stress-strain behavior of fibers (especially polymer
fibers) include, but are not limited to: moisture, temperature, and strain rate. The
discussion in this section focuses on the general trends of regular, molecularly-
oriented polymer fibers. For simplicity, the typical stress-strain behavior shown in
Figure 15.11 is used in the discussion.
Figure 15.18A shows the effect of moisture, i.e., relative humidity (RH%),
on the stress-strain curve of a typical polymer fiber. Water molecules can dif-
fuse into the fiber structure and increases the intermolecular distance between the
Table 15.2. Tensile strengths, moduli, and strains-at-break of synthetic polymer, natural

polymer, and inorganic fibers. Data obtained at 20°C and 65% relative humidity.
Strength Modulus
Strain-at-Break
Fiber (grams per (grams per
(%)
denier) denier)
Polypropylene 2.0–7.8 50–90 15–50
Acrylic 2.0–3.6 40–75 25–55
Nylon 6,6, regular 4.3–6.0 15–40 15–40
Nylon 6,6, high-strength 6.0–9.0 50–60 13–16
Polyester, regular 2.5–6.0 105–120 10–60
Polyester, high-strength 6.0–10.5 150–160 7–8
Kevlar (HM-HT fiber) 20–25 400–1500 2–5
Spectra (HM-HT fiber) 25–43 850–1800 3–4
Polyurethane (elastomeric fiber) 0.4 0.1 500
Cotton 3.0–5.0 45–85 5–10
Flax 2.6–7.7 200–460 1.3–3.3
Jute 3.0–5.8 110–350 1.3–3.0
Ramie 5.3–7.4 175–190 3.0–7.0
Wool 1.0–1.7 25–35 20–45
Silk 2.4–6.1 80–90 10–25
Acetate 1.2–1.5 45 25–35
Triacetate 1.1–1.4 37–40 25–35
Rayon, regular 1.8–3.0 50–70 15–28
Rayon, high-strength 3.0–5.0 100–150 7–10
Inorganic Fibers
Carbon 8.6–40 1500–2600 0.7–2.5
Glass 6.3–20 350–450 2–5.4
University Press, 1996.; Morton, W.E., et. al., Physical Properties of Textile Fibres, Fourth
polymer chains. This softens the fiber and causes a reduction in fiber modulus.
At the same time, the fiber strength decreases and the strain-at-break increases.
However, this softening effect only occurs for nylon, wool, silk, and other poly-
mer fibers that are capable of forming hydrogen bonds with water molecules.
Fibers made of polyethylene, polypropylene, or other polymers that do not form
Figure 15.18. Effect of relative humidity on the stress-strain behavior of polymer fibers

that are capable of forming hydrogen bonding. (A) Typical polymer fibers, and (B) cotton.
hydrogen bonds are not affected by moisture. Another special case is cotton. As
shown in Figure 15.18B, the strength of cotton fibers actually increases when they
absorb moisture. One explanation for this is that significant residual stress exists
in cotton and the diffusion of moisture helps release the residual stress, leading to
increased fiber strength. However, work still is needed to provide strong experi-
mental support for this argument.
The effect of temperature is similar to that of moisture. Figure 15.19 shows the
typical stress-strain curves of a polymer fiber at low, medium, and high tempera-
tures. With increase in temperature, the fiber strength and modulus decrease, but
the strain-at-break increases. The actual changes in strength, modulus and strain-
at-break vary from fiber to fiber. In addition, exposure to high temperatures for
prolonged time could lead to permanent degradation of fibers.
The stress-strain behavior of fibers generally is tested by increasing the strain
from zero at a fixed rate. The rate of the strain is another factor influencing the
stress-strain behavior of fibers. In most cases, increased strain rate leads to in-
creased fiber strength and modulus, but reduced strain-at-break (Figure 15.20).
This is because, at high rates, it becomes more difficult to initiate the different
types of motions of the polymers, which will be discussed in Chapter 16.
15.2.7 Structural Factors Affecting the Stress-Strain Behavior

Many structural factors affect or even determine the stress-strain behavior of fi-
bers. This section focuses on discussing how the different structural factors affect
the stress-strain behavior of regular, molecularly oriented polymer fibers.
Polymer chemical structure limits the maximum potential properties that poly-
mer fibers can theoretically achieve. In general, fibers made of rigid polymer
chains have higher strength and larger modulus than those of flexible polymer
chains. For example, Kevlar is an aramid fiber formed by polyparaphenylene
terephthalamide, which contains aromatic rings on the polymer backbone. The
presence of aromatic rings makes the polymer chains very rigid, and as a result,
Figure 15.19. Effect of temperature on the stress-strain behavior of polymer fibers.
Figure 15.20. Effect of strain rate on the stress-strain behavior of polymer fibers.
Kevlar has excellent tensile properties, i.e., high strength and modulus. Similar to
rigid polymer chains, polymers that have strong secondary bonding also can form
fibers with high strength and modulus. Although rigid polymer chains and strong
secondary bonding can lead to high tensile properties, polymer fibers with high
strength and modulus could be obtained by other means. For example, polyethyl-
ene chains are flexible and can only form weak secondary bonding (van der Waal
force). However, Spectra fibers made from gel-spun ultra-high molecular weight
polyethylene have extremely high strength and modulus (Table 15.2), mainly due
to the extra-long polymer chains, good molecular orientation, and high degree of
crystallinity. Therefore, molecular weight, molecular orientation and crystallinity
also are important structural parameters that affect the final properties of polymer
fibers.
Figure 15.21 shows the effect of molecular weight on the tensile strength of
polymer fibers. At low molecular weights, the tensile strength increases rapidly
with increase in molecular weight. However, at high molecular weights, the ten-
sile strength only increases slightly when molecular weight increases. Although
it is feasible to improve the fiber tensile strength by increasing the polymer mo-
lecular weight, this approach is limited in practice by the polymer viscosity. With
increase in molecular weight, the polymer viscosity increases rapidly, especially
at high molecular weights, and this makes it difficult to produce fibers from high
molecular weight polymers.
Polymer chemical structure limits the maximum potential, but the physical
structure of polymers determines the actual tensile properties that can be achieved
by fibers in practice. One of the most important physical structural factors for
polymer fibers is the molecular orientation along the fiber longitudinal direction.
Figure 15.22 shows the effect of molecular orientation on the stress-strain be-
havior of polymer fibers. When the degree of molecular orientation is low, the
fiber exhibits yielding and plastic deformation, and the stress-strain curve can
be divided into three stages, as discussed in section 15.2.4. When the molecular
orientation increases, the strength and modulus increase, but the strain-at-break
decreases. The plastic region reduce quickly, but still may present when the fiber
has medium orientation. When the molecular orientation is high, the plastic region
disappears, and both the strength and modulus increase significantly. One simple
way to understand the effect of molecular orientation is that: while increasing the
Figure 15.21. Effect of molecular weight on the tensile strength of polymer fibers.

Figure 15.22. Effect of molecular orientation on the stress-strain behavior of polymer fibers.
molecular orientation, more primary bonds are distributed along the fiber longi-
tudinal direction to bear the applied force. This is the main cause of increased
strength and modulus since primary bonds along the polymer main chains are
significantly stronger and stiffer than the secondary bonds between the polymer
chains. Therefore, to improve the tensile properties of polymer fibers, it is es-
sential to achieve high molecular orientation in the fibers. For synthetic polymer
fibers, high molecular orientation is achieved by using a take-up speed that is
greater than the extrusion speed, and by introducing a post-spinning drawing step.
Natural polymer fibers typically have high molecular orientation by themselves.
Although high molecular orientation is desired in most cases, fibers with low
molecular orientation also are useful in some applications, especially, when large
strain-at-break is needed.
Similar to molecular orientation, the degree of crystallinity also is critically im-
portant. The crystalline phase has orderly arranged polymer chains and has better
mechanical properties than the amorphous phase. As a result, with increase in crys-
tallinity, the strength and modulus of polymer fibers increase, but the strain-at-break
decreases. All useful natural polymer fibers have relatively high degrees of crystal-
linity. The degree of crystallinity of synthetic polymer fibers can be controlled by
selectively adjusting the processing conditions used during the fiber formation.
15.2.8 Elastic Recovery

Elastic recovery is the ability of a fiber to recover from its deformation after the re-
moval of the applied stress. As discussed in the previous sections, the deformation
of a fiber includes both elastic and plastic components. The elastic deformation
mainly is caused by the changing of chain length and angle under applied stress,
and it is recoverable. The plastic deformation is associated with the molecular
orientation, unraveling and sliding of crystalline phase, etc., and it is unrecover-

able. As a result, the ealstic recovery of fibers is typically less than 100%.
Figure 15.23 shows both the tensile stretching curve and the recovery curve
of a typical fiber. The stretching curve shows the typical stress-strain behavior,
which includes both elastic and plastic deformations. The recovering curve shows
only the elastic deformation (or strain) is recovered after the removal of the ap-
plied tensile stress. The elastic recovery, or strain recovery, can then be defined as:
Recovered Strain
Elastic � Recovery = (15.24)
Total � Strain
Table 15.3 shows the elastic recovery of some fibers under different strains
and in different humidities. Nylon, polyester and wool fibers have higher elastic
recovery values than most other fibers. Both strain and humidity affect the fiber
elastic recovery. In general, the elastic recovery of fibers decreases with increase
in strain. However, the effect of humidity is complicated. Polyethylene fibers do
not absorb moisture and their elastic recovery is not affected by the humidity.
However, for fibers that can absorb moisture, their ability to recover from strain is
affected by moisture. In some cases, the elastic recovery increases with increase
in humidity. In other cases, the elastic recovery decreases with humidity.
Fibers sometimes are subject to repeated cycles of stretch and recovery. The
elastic recovery can change during repeated cycles. Figure 15.24 shows the
stretch and recovery curves of a fiber under multi-cycle stretch and recovery tests.
Figure 15.23. Stretch and recovery curves of a typical fiber.

Table 15.3. Elastic recovery of fibers at different strains and humidities.
Recovery (%) from
Fiber 1% Strain 5% Strain 10% Strain
60 90 60 90 60 90
RH% RH% RH% RH% RH% RH%
Polyethylene 80 80 80 83 78 80
Acrylic 92 90 50 48 43 39
Nylon 6,6 90 92 89 90 86 —
Polyester 98 92 65 60 51 47
Cotton 91 83 52 59 — —
Wool 99 94 69 82 51 56
Silk 84 78 52 58 34 45
Acetate 96 75 46 37 24 22
Rayon 67 60 32 28 23 27
Source: Beste, L.F., et. al., Textile Research Journal, 20, 441–453, 1950.
For this particular fiber, the elastic recovery at the first cycle is relatively low.
However, after a few cycles, the fiber becomes conditioned and the stretch and re-
covery curves tend to fall on a loop, i.e., 100% elastic recovery. This phenomoum
is important for some applications, such as tire cords. However, not all fibers can
have 100% elastic recovery in repeated cycles.
In addition to dimensional recovery, work recovery often is studied for ten-
sile deformation. While stretching a fiber, the total work done is either stored in
chemical bonds or is lost, typically in the form of heat. The work stored in the
chemical bonds is recoverable, but the work lost is not. In Figure 15.23, the area
under the stretch curve is the total work per unit volume done during stretching.
The area under the recovery curve is the work per unit volume returned during
recovery. Work recovery can then be defined as:
Work � Returned
� during � Recovery
Work � Recovery = (15.25)
TotalWork
� � done � during� Stretching
Table 15.4 shows the work recovery of some fibers under different strains and
in different humidities. Inorganic fibers, such as glass, have very high work recov-
ery at a low strain of 1%. However, they break at moderate and high strains, and
no work recovery can be measured. The work recovery of polymer fibers is lower
than their elastic (strain) recovery due to the non-linear stress-strain behavior. In
addition, the work recovery also is affected by both strain and humidity.
Figure 15.24. Stretch and recovery curves of a fiber under multi-cycle tests.
15.3 COMPRESSIVE PROPERTIES

Compressive properties of fibers also are important. Fibers often experience com-
pressive forces while being made into the final products. The final products also
may be under compression when they are in use. Therefore, it is important to
understand the compressional properties of fibers.
15.3.1 Longitudinal Compressive Properties

A longitudinal compressive stress basically is a negative tensile stress, while
a longitudinal compressive strain is the negative tensile strain. Figure 15.25
shows an ideal longitudinal stress-strain curve of a fiber. The longitudinal
compressive modulus is the same as (or comparable to) the tensile modulus.
However, the compressive strength typically is lower than the tensile strength.
Table 15.5 compares the longitudinal compressive strength and tensile strength
of some fibers.
The buckling of fibers under a longitudinal compressive force is a major is-
sue in fiber applications. Figure 15.26 shows the possible response of an uncon-
strained fiber under longitudinal compression. Under the longitudinal compres-
sive force, the fiber buckles easily even at low strains if there is no lateral support.
The critical compressive force (Fcritical), above which the fiber buckling occurs, is
given by the Euler’s equation:
Compressive Properties 293
Table 15.4. Work recovery of fibers at different strains and humidities.
Recovery (%) from
Fiber 1% Strain 5% Strain 10% Strain
60 RH% 90 RH% 60 RH% 90 RH% 60 RH% 90 RH%

Polyethylene 55 63 50 55 47 51
Acrylic 64 66 23 21 21 16
Nylon 6,6 66 68 55 64 52 —
Polyester 82 76 35 30 24 21
Cotton 50 42 31 33 - -
Wool 83 67 37 45 21 25
Silk 57 47 23 24 17 23
Acetate 80 70 18 14 9 8
Rayon 50 27 11 10 9 11
Inorganic Fiber
Glass 97 — — — — —
Source: Beste, L.F., et. al., Textile Research Journal, 20, 441–453, 1950.
Figure 15.25. Ideal stress-strain curve of a fiber.

Table 15.5. Longitudinal compressive and tensile strengths of

fibers. Data obtained at 20°C and 65% relative humidity.
Tensile Strength Compressive Strength

Fiber
(GPa) (GPa)
Kevlar 149 3.4–3.5 0.32–0.46
Spectra 1000 3.0 0.17
Carbon, polyacrylonitrile-based 1.4–3.9 0.48–1.15
Carbon, pitch-based 1.8–3.8 1.06–2.55
Glass 3.5–4.5 > 0.81
Source: Kumar, S., Indian Journal of Fiber and Textile Research, 16, 52–64, 1991.
π 2 EI
Fcritical = (15.26)
( KL )
2
where E is the modulus of the fiber, I the area moment of inertia of the fiber cross-
section, K the effective length factor, and L the fiber length. The effective length
factor is determined by the conditions of fiber ends: K = 1.0 when both ends are
pinned (hinged, free to rotate); K = 0.5 when both ends are fixed; K = 0.699 when
one end is fixed and the other end is pinned; and K = 2.0 when one end is fixed and
the other end is free to move laterally. The KL is the effective length of the fiber.
Figure 15.26. Buckling of an unconstrained fiber under longitudinal compressive force.

Compressive Properties 295
The Euler’s equation then can be rearranged to obtain the critical compres-
sional stress (σcritical) that is needed for a fiber to buckle:
Fcritical π2 E
σ critical = = (15.27)
( KL / R )
2
A
Therefore, the fiber’s resistance to buckling can be improved by increasing the

fiber modulus, increasing the fiber radius, and reducing the fiber length. Providing
lateral support is another effective approach to enhance the fiber’s resistance to
buckling. In composites, the buckling of fibers is limited by the matrix when the
compressive force is along the fiber axis (Figure 15.27A).
15.3.2 Transverse Compressive Properties

In addition to longitudinal compressive properties, transverse compressive prop-
erties also are important. Fibers in composites need have sufficient transverse
compressive modulus and strength since the external force sometimes is applied
in the transverse direction to the fiber axis (Figure 15.27B). In both woven and
nonwoven fabrics, fibers also may experience transverse compressive forces
while in use.
Table 15.6 shows compressive moduli of some fibers under the longitudinal
and transverse directions. In general, the transverse compressive moduli of poly-
mer fibers are lower than their longitudinal compressive moduli. This is mainly
because of high molecular orientation in these polymer fibers. Carbon fibers also
have lower transverse compressive modulus than longitudinal compressive mod-
ulus since the graphene sheets in these fibers are largely oriented. However, glass
fibers are isotropic without any orientation, and hence have comparable trans-
verse and longitudinal compressive moduli. Similar to the moduli, the transverse
Figure 15.27. A composite under compressive force: (A) along fiber axis direction, and (B)
along fiber transverse direction.
Table 15.6. Longitudinal and transverse compressive moduli of fibers.

Data obtained at 20°C and 65% relative humidity shows:
Longitudinal Modulus Transverse Modulus

Fiber
(GPa) (GPa)
Polyethylene 89.3 1.21
Kevlar 63.4–179.1 2.31–2.59
Wool 3.55–3.73 0.97–1.01
Carbon, polyacrylonitrile-based 234.6–343.2 6.03–10.08
Carbon, pitch-based 126.2–379.0 3.08–9.95
Glass 77.4 67.87
Source: Muraki, C., et. al., Journal of the Textile Institute, 81, 12–21, 1994.; Kawabata, S., Journal
of the Textile Institute, 81, 432–447, 1990.
compressive strengths of most fibers also are lower than their longitudinal com-
pressive strengths.
15.4 TORSIONAL PROPERTIES

Fibers often are twisted while being converted to useful products. For example,
fibers are twisted to form yarns, which then are made into textile fabrics or ropes.
The final products made of fibers also may be twisted during use. The torsional
properties determine how fibers respond while being twisted.
15.4.1 Torsion of Fibers

Torsion basically is shear with a different geometry. Figure 15.28 shows the de-
formation of a fiber under a torsional force (F). The fiber has a length of L and
a radius of R. After the fiber is twisted through an angle θ, line OA on the cross-
section of the fiber is displaced to a new position OB. Point O is the center of the
fiber cross-section. When θ is small, the displacement (δ) of point A along the
torsional force direction is:
δ = r cos θ = r θ (15.28)
where r is the length of line OA. The shear strain (γ) and shear stress (τ), respec-
tively, can be obtained by:
r
γ=θ (15.29)
L
and
Torsional Properties 297
Figure 15.28. Deformation of a fiber under a torsional force F.
r
τ = Gθ (15.30)
L
where G is the shear modulus of the fiber. Therefore, the shear stress and shear
strain in a fiber are proportional to the radial distance from the fiber center. The
maximum shear stress and strain are obtained at the fiber surface. Figure 15.29
shows the shear stress-shear strain curve at the surface of a fiber. Shear stress
increases with increase in shear strain. Shear strength is defined as the shear stress
when the fiber breaks. Table 15.7 compares the tensile and shear strengths of
fibers. It is seen the fiber shear strength is smaller than the fiber tensile strength.
The torque of the fiber often is studied by twisting a fiber. Torque basically is
the measure of how much a torsional force acting on a fiber causes that fiber to
twist. In mechanics, when the force is applied on a lever arm, the torque can be
defined by:
Torque = LF sin ϕ (15.31)
where L is the length of the lever arm, F the magnitude of the force, and φ the
angle between the force and the lever arm (Figure 15.30).
While twisting a fiber shown in Figure 15.28, the torque produced on the fiber
by the torsional force can be calculated by:
πR 4 G θ
Torque = (15.32)
2L
Figure 15.29. Typical shear stress-shear strain curve of a fiber.
Table 15.7. Tensile and shear strengths of fibers. Data

obtained at 20°C and 65% relative humidity.
Tensile Strength Shear Strength

Fiber
(grams per denier) (grams per denier)
Nylon 4.4 1.3
Cotton 2.7 1.0
Flax 2.9 0.91
Silk 3.5 1.3
Rayon 2.0 0.72
Acetate 1.3 0.66
Source: Warner, S.B., Fiber Science, Prentice Hall, 1995.
Figure 15.30. Definition of torque.

Torsional Properties 299
In most measurements, the torque is plotted against the twist, i.e., the number of
turns per unit fiber length (turns/cm or turns/mm). Figure 15.31 shows a typical
torque-twist curve of a fiber, which is similar to the shear stress-shear strain curve.
According to Equation 15.32, under a torsional force, the fiber resists the tor-
sional displacement (θ/L) by a factor of πR4G/2. Therefore, the torsional rigidity
can be defined as:
πR 4 G
Torsional Rigidity = (15.33)
2
The torsional rigidity of a fiber contains a material term (G) and a geometric
term (πR4/2). The geometric term contains the fiber radius to the fourth power,
and hence the radius or diameter of a fiber has a significant effect on its torsional
rigidity. Equation 15.33 can only be used for calculating the torsional rigidities of
circular fibers, i.e., those with circular cross-sections. When a non-circular fiber
is studied, the shape factor (λT), i.e., the ratio of the perimeter of the non-circular
fiber to that of a circular fiber of the same cross-sectional area, should be used:
πR 4 G
Torsional Rigidity = λT (15.34)
2
15.4.2 Shear Modulus of Fibers

As discussed in section 15.4.1, the shear modulus of fibers is important, and it de-
termines the fiber torsional rigidity. The shear modulus of a fiber can be measured
by using the torsion pendulum technique. Figure 15.32 shows the schematic of a
Figure 15.31. Typical torque-twist curve of a fiber.

Figure 15.32. Schematic of a basic torsion pendulum. More sophisticated designs are

available.
basic torsion pendulum. The fiber is mounted between a fixed end and an inertial
mass. The inertial mass should be sufficiently small so the tensile deformation of
the fiber can be neglected. At the same time, the inertial mass also should be suffi-
ciently large so it can produce stable oscillation for accurate measurement. During
the measurement, the fiber is subject to a shear strain caused by the rotation of the
inertial mass. The frequency of oscillation is measured. The shear modulus of the
fiber is given by:
2 LM ω2
G= (15.35)
πR 4
where L is the length of the fiber between the fixed end and the inertial mass, M
the moment of inertia of the inertial mass, ω the frequency of the oscillation, and
R the radius of the fiber. When the inertial mass (m) is a cylindrical disc, its mo-
ment of inertia is:
2
mRdisc
M = (15.36)
2
where Rdisc is the radius of the inertial mass.
Equation 15.35 only gives accurate shear modulus values of circular fibers,
i.e., those with circular cross-sections. When a non-circular fiber is studied, the
shape factor should be used. Table 15.8 shows typical shear moduli and shape
factors of some fibers.
For fibers, the relationship between shear modulus and tensile modulus is in-
teresting. Earlier in this Chapter, it has been mentioned the shear modulus of
isotropic materials has a simple relationship with tensile modulus and Passion’s
ratio (ν). The ratio of tensile modulus to shear modulus is:
E
= 2 (1 + ν ) (15.37)
G
Bending Properties 301
Table 15.8. Shear moduli and shape factors of fibers. Data

obtained at 20°C and 65% relative humidity.
Shear Modulus
Fiber Shape Factor
(GPa)
Polypropylene 0.75 1.00
Polyester 0.8 1.00
Acrylic, bilobal 1.3 0.57
Nylon 6,6 0.4 1.00
Kevlar 49 1.4 1.00
Kevlar 149 1.2 1.00
Cotton 2.2 0.71
Flax 1.4 0.94
Wool 1.3 0.99
Silk 2.4 0.84
Rayon 1.0 0.94
Carbon 4.7 1.00
Source: Warner, S.B., Fiber Science, Prentice Hall, 1995.
The Passion’s ratios of most isotropic polymers are between 0.3 and 0.4, and
hence the E/G ratios of isotropic polymers range from 2.5 to 3.0. However, poly-
mer fibers are anisotropic due to the high molecular orientation. With increase in
molecular orientation, the tensile modulus increases, but the shear modulus de-
creases. This is because the molecular orientation is beneficial for increasing the
fiber tensile modulus; however, with increased molecular orientation, more fibrils
are formed in the fiber and the fibrils are easily separated by twisting although
they have excellent tensile properties. Table 15.9 shows the tensile moduli, shear
moduli, and their ratios for some fibers. Some fibers have very large E/G ratios
due to the high molecular orientation.
15.5 BENDING PROPERTIES

To bend a fabric, composite, or any fiber-containing products, the bending of fi-
bers is required. In each case, the arrangement of fibers is important in determin-
ing the bending stiffness of the final product. However, to understand the bending
of the final products, the bending properties of fibers must be studied first.
The deformation of a fiber under bending is shown in Figure 15.33. Outer
layers of the fiber are stretched and the inner layers are compressed. While be-
ing bent, the length of the tensile side of the fiber increases, but the length of
the compressive side decreases. However, the neural plane between the tensile
and compressive sides remains unchanged in length. In addition to tensile and
Table 15.9. Tensile moduli, shear moduli, and their ratios of

fibers. Data obtained at 20°C and 65% relative humidity.
Tensile Modulus Shear Modulus Tensile Modulus

Fiber
(GPa) (GPa) Shear Modulus
Polyester, high-speed spun 1.98 0.65 3.05
Polyester, high-speed spun and drawn 8.81 0.85 10.4
Nylon 6 3.41 0.49 6.96
Polypropylene 2.09 0.57 3.67
Polyethylene, gel-spun 93.7 0.84 112
Kevlar 49 94.2 1.60 58.9
Source: Zeronian, S.H., et. al., Journal of the Textile Institute, 85, 293–300, 1994.
compressive stresses, shear stress also may be developed, depending on how the
bending is carried out.
Figure 15.34 compares the typical tensile and bending stress-strain curves of
polymer fibers. In most cases, the bending stress-strain curve lay below the tensile
curve. This is because the compressive side of the fiber yields more easily than
the tensile side. The bending rigidity (or flexural rigidity) of a circular fiber is
given by:
πR 4 E
Bending Rigidity = (15.38)
4
Similar to the torsional rigidity, the bending rigidity also has a material term (E,
tensile modulus) and a geometric term (πR4/4). The geometric term contains the
fiber radius to the fourth power, and hence the fiber radius or diameter has a sig-
nificant influence on the bending rigidity. For non-circular fibers, the fiber shape
factor should be used while calculating the bending rigidity:
πR 4 E
Bending Rigidity = λ B (15.39)
4
The value of the shape factor (λB) for bending rigidity may be different from
that (λT) for the torsional rigidity. However, they have the same tendency to in-
crease or decrease when the cross-sectional shape of fibers changes. According to
Equations 15.34 and 39, with increase in shape factors, both bending and torsional
rigidities increase (Figure 15.35).
Table 15.10 compares the bending and torsional rigidities of some fibers. The
bending rigidities of fibers generally are greater than their torsional rigidities.
Bending Properties 303
Figure 15.33. Bending of a fiber.
Figure 15.34. Typical tensile and bending stress-strain curves of a fiber.
Figure 15.35. Changes in shape factors, and torsional and bending rigidities with cross-
sectional shape. The bending direction is defined by the dash line.
Table 15.10. Bending and torsional rigidities of fibers.

Data obtained at 20°C and 65% relative humidity.
Bending Rigidity Torsional Rigidity

Fiber
(Pa mm4) (Pa mm4)
Polypropylene 143 41.3
Polyester 118–357 25.2–80.0
Acrylic 193–377 67.0–194
Nylon 6 181 61
Nylon 6,6 58.0–209 13.8–81.0
Cotton 194–218 61.0
Wool 600 319
Silk 39.0 10.6
Acetate 125 31.4
Triacetate 220 80.0
Rayon 89.0–800 15.6–191
Source: Owen, J.D., Journal of the Textile Institute Transactions, 56, T329–T339, 1965.
REFERENCES
[1] Beste, L.F., and Hoffman, R.M., “A Quantitative Study of Resilience”,
Textile Research Journal, 20, 441–453, 1950.
[2] Boresi, A.P., Schmidt, R.J., and Sidebottom, O.M., Advanced Mechanics of
Materials, Fifth Edition, John Wiley & Sons, Ltd., 1993.
[5] Gohl, E.P.H., and Vilensky, L.D., Textile Science: An Explanation of Fibre
Properties, Longman Group Ltd., 1981.
[6] Hearle, J.W.S., and Miraftab, M., “The Flex Fatigue of Polyamide and
Polyester Fibres”, Journal of Materials Science, 26, 2861–2867, 1991.
[7] Hearle, J.W.S., and Wong, B.S., “Flexural Fatigue and Surface Abrasion of
Kevlar-29 and Other High-Modulus Fibres”, Journal of Materials Science,
12, 2447–2455, 1977.
Problems 305
[9] Kawabata, S., “Measurements of Transverse Mechanical Properties of

High-Performance Fibers”, Journal of the Textile Institute, 81, 432–447,
1990.
[10] Kumar, S., “Advances in High Performance Fibers”, Indian Journal of
Fiber and Textile Research, 16, 52–64, 1991.
[12] Muraki, C., Niwa, M., Amino, N., and Kawabata, S., “Changes in
Anisotropic Elastic Moduli of Wool Fibres during Worsted Spinning,
Weaving and Finishing”, Journal of the Textile Institute, 81, 12–21, 1994.
2006.
[14] Owen, J.D., “The Application of Searle’s Single and Double Pendulum
Methods to Single Fibre Rigidity Measurements”, Journal of the Textile
Institute Transactions, 56, T329–T339, 1965.
[16] Pizzuto, J., Fabric Science, Tenth Edition, Fairchild Publications Inc, 2010.
[17] Schultz, J., Polymer Materials Science, Prentice-Hall Inc., 1974.
[19] Ward, I.M., and Sweeney, J., Mechanical Properties of Solid Polymers,
Third Edition, John Wiley & Sons, Ltd., 2013.
[21] Zeronian, S.H., Buschler-Diller, G., Holmes, S., and Inglesby, M.K.,
“Relationships between the Mechanical Properties of Synthetic Fibers”,
Journal of the Textile Institute, 85, 293–300, 1994.
PROBLEMS
(1) Describe Hooke’s law, tensile stress, tensile strain, tensile modulus,
Poisson’s ratio, bulk modulus, and shear modulus.
(2) What is the difference between true stress and engineering stress? How do
you calculate the true stress from engineering stress?
(3) What is specific stress?

(4) In general, polymer fibers do not exhibit ideal stress-strain behavior. Why?
(5) Estimate the modulus, tenacity, true tenacity, and strain-to-break for an un-
oriented polymer fiber shown below.
(6) An as-spun, unoriented, amorphous polymer fiber is stretched until it breaks.

Discuss all the possible places the energy might have gone.
(7) An as-spun, unoriented, semicrystalline polymer fiber is stretched until it
breaks. Discuss all the possible places the energy might have gone.
(8) A molecularly oriented polymer fiber is stretched until it breaks. Discuss all
the possible places the energy might have gone.
(9) Discuss and explain how you would expect the tensile properties of a poly-
mer fiber to vary with:
(i) Molecular weight
(ii) Molecular orientation
(iii) Chain rigidity
(iv) Intermolecular interaction
(v) Moisture
(vi) Temperature
(vii) Strain rate
(10) Compare elastic recovery with work recovery.

(11) Which has larger value for polymer fibers, shear tenacity or tensile tenac-
ity? Why?
Problems 307
(12) What kind of deformation is involved when you are bending a fiber?
(13) For commercial polymer fibers, the ratio of tensile modulus (E) to shear
modulus (G) is greater than the value predicted by:
E
= 2 (1 + ν )
G
where v is the Passion’s ratio. Explain why.

(14) Compare axial compressive modulus with axial tensile modulus. Which is
larger? Why?
(15) Compare axial compressive strength with axial tensile strength. Which is
larger? Why?
(16) Compare transverse compressive modulus with axial compressive modulus.
Which is larger? Why?
CHAPTER 16
Viscoelastic Properties of Fibers
Viscoelasticity is the property of materials that exhibit both viscous and elas-
tic responses under applied stress. Viscoelastic properties haven been observed
throughout recorded history. For example, archers in ancient time knew they
should never leave their bows strung when not in use because the tension in the
bowstring would decrease over time. The first scientific study on the viscoelas-
tic properties of fibers is probably by Weber, who noted in 1835 that silk fibers
under tension presented an immediate deformation and a delayed extension that
increased with time. As a matter of fact, all materials can exhibit elastic and vis-
cous characteristics simultaneously if the time scale of observation is comparable
to the relaxation times needed for large-scale atomic rearrangements in these ma-
terials. However, the relaxation times of most non-polymer fibers are significantly
greater than the time scale of normal observation, and hence it is hard to observe
their viscoelastic behavior at room temperature. On the other hand, polymer fibers
have relaxation times that are comparable to the time scale of observation, and
then they easily display viscoelastic behavior. This chapter, therefore, focuses on
the viscoelastic properties of polymer fibers.
16.1 MOLECULAR MECHANISMS OF VISCOELASTIC

BEHAVIOR
Under applied stress, polymer fibers may deform by either or both of two funda-
mentally different molecular mechanisms.
• In the first mechanism, the bond lengths and angles are altered, moving the
atoms to new positions with higher internal energy. This is a small-scale
motion and occurs in a time scale of around 10–12 seconds.
309
310 Viscoelastic Properties of Fibers
• The second mechanism is related to larger-scale rearrangements of the

atoms. For example, the rotation around the carbon-carbon single bonds on
the main chains changes the polymer chain conformations, leading to the
development of molecular orientation and crystallinity, the slide of polymer
chains over one another, etc. Some of the conformational changes (e.g., the
development of molecular orientation) are recoverable, but some other con-
formational changes (e.g., the slide of polymer chains) are not. However, all
conformational changes can decrease the polymer conformational entropy
by making the system less “disordered”. Compared with the changes in
bond lengths and angles, the large-scale rearrangements of atoms associated
with the second mechanism require longer time scale.
According to the combined first and second laws of thermodynamics, the me-
chanical work (f dx) done on a fiber by applied stress produces an increase in the
internal energy (dU) or a decrease in the entropy (dS):
f � dx = dU − � T dS (16.1)
Equation 16.1 provides a convenient means of determining whether the fiber’s

stiffness is energetic or entropic in origin.
For polymer fibers, the internal energy-controlled elasticity has an instan-
taneous nature since it involves the rapid changes of bond lengths and angles.
However, the conformational or entropic changes are processes that are sensitive
to the local molecular mobility. This molecular mobility is affected by many fac-
tors, such as polymer chain structure, temperature, or absorbed moisture or other
small molecules, etc. The rates of conformational changes can be described by the
Arrhenius equation:
*
Rate = Ae − ∆E / RT
(16.2)
where A is a structure constant, ΔE* the activation energy of the conformation

change process, R the gas constant, and T the absolute temperature of the poly-
mer. Figure 16.1 shows the temperature dependence of the conformational change
rates. When the temperature is significantly greater than the glass transition tem-
perature (Tg), the rates of conformational changes are so fast that polymer fibers
are in the rubbery state and exhibit large, instantaneous, and reversible deforma-
tion when the stress is applied. Elastomer fibers, such as polyurethane, are used in
the rubbery state. However, when the temperature is significantly lower than Tg,
the rates are so slow and are negligible. As a result, the polymer chain conforma-
tions are frozen, and the fibers are in the glassy state and are able to deform only
through the changes in bond lengths and angles. Polymer fibers at the glass state
generally are brittle and break easily at low strains. When the temperature is not
too far away from Tg, polymer fibers are midway between the glassy and rubbery
Phenomenological Aspects of Viscoelastic Behavior 311
Figure 16.1. Relationship between conformational change rate and temperature.
states, and their response to applied stress is a combination of viscous fluidity and
elastic solidity, i.e., in the viscoelastic state. Polymer fibers are often used in their
viscoelastic state.
In both glassy and rubbery states, the mechanical properties, such as modulus,
of polymer fibers have no or weak dependence on time. But in the viscoelastic
state, the mechanical properties of polymer fibers depend strongly on time. For
example, the tensile tests take some time, and hence the stress (σ)-strain (ε) be-
havior of polymer fibers is affected by strain rate (Figure 16.2). At a low strain
rate, there is more time available for the polymer chains to change conformations
and fibers exhibit higher strain at a given stress. However, at a high strain rate,
there is less time for conformational changes and fibers present lower strain at the
same stress.
16.2 PHENOMENOLOGICAL ASPECTS OF VISCOELASTIC

BEHAVIOR
The viscoelastic behavior of polymer fibers is complex. Experimentally, it is im-
portant to perform simple laboratory tests from which information relevant to ac-
tual in-use conditions can be obtained. The viscoelastic characterization of poly-
mer fibers often consists of conducting mechanical tests that are similar to those
discussed in Chapter 15, but are modified so as to enable the observation of the
time dependency of fibers’ response. Three most important viscoelastic tests are:
creep, stress relaxation, and dynamic mechanical testing.
Figure 16.2. Effect of strain rate on the stress-strain behavior of polymer fibers.
16.2.1 Creep
The creep test consists of observing the change of strain with time under a con-
stant stress (Figure 16.3). The strain of viscoelastic polymer fibers changes with
time when a constant stress is applied. Figure 16.4 shows the creep curves of a
typical polymer fiber under different constant stresses. On the application of a
constant stress, the polymer fiber instantaneously deforms an amount determined
by the stress, and then it exhibits the creep behavior, i.e., a delayed deformation
that increases gradually with time. Higher constant stress leads to greater creep.
It is important to understand the creep behavior shown in Figure 16.4 is not a
simple superposition of linear elastic and viscous responses. Figure 16.5 shows
the typical strain-time curves of ideal elastic material, ideal viscous material, and
viscoelastic polymer fibers under constant stress. The ideal elastic material de-
forms instantaneously as the stress is applied and the stain remains constant with
time. The removal of the stress causes the ideal elastic material to return to its
original dimension. For the ideal viscous material, the strain increases linearly
with time as long as the stress is applied. The removal of the stress does not return
the ideal viscous material to the original dimension. This is because the energy
introduced by the work of the external stress is dissipated in the flow, leading to
a permanent deformation. Both the ideal elastic and viscous responses contrib-
ute to the creep-recovery curve of the viscoelastic polymer fibers. However, the
creep-recovery curve of viscoelastic polymer fibers is not a simple superposition
of these two ideal behaviors. In addition to the ideal responses, the creep-recov-
ery curve of the polymer fibers also includes retarded elastic response, in which
Figure 16.3. Creep test of a viscoelastic fiber.
Figure 16.4. Creep of a typical polymer fiber under different constant stresses.
viscous and elastic responses are coupled so the fibers gradually deform in a non-
linear fashion with time.
As discussed above, the creep-recovery curve of viscoelastic polymer fibers
includes: elastic, retarded, and viscous responses. Figure 16.6 shows the elastic,
retarded, and viscous components of a typical creep-recovery curve of viscoelas-
tic polymer fibers. The contributions of different responses to the creep-recovery
curve can be described as below:
• When the time is very short, e.g., shorter than the relaxation time of the
polymer chain segments, the polymer chains are “frozen” and the only
Figure 16.5. Typically strain-time curves of (A) ideal elastic material, (B) ideal viscous
material, and (C) viscoelastic polymer fibers under constant stress.
Figure 16.6. (A) Creep-recovery curve of a viscoelastic polymer fiber, and the correspond-
ing (B) elastic, (C) retarded, and (D) viscous components.
possible motions are the changes of bond lengths and angles. As a result,
the deformation at the initial stage of the creep test is the result of elastic
response. This elastic deformation (ε1) is recoverable instantaneously after
the removal of the applied stress (Figure 16.6B).
• When the time becomes longer and is comparable to the relaxation time of
polymer chain segments, the larger-scale rearrangements of the atoms be-
come possible by changes in polymer chain conformations, e.g., molecular
orientation, strain-induced crystallization, etc. As discussed in section 16.1,
these recoverable conformational changes are rate-controlled and cause re-
tarded deformation (ε2), which increases gradually with time (Figure 16.6C).
Since the elastic deformation remains constant during creep, the total fiber
deformation is: εtotal = ε1 + ε2. When the applied stress is removed, the re-
tarded deformation is recovered gradually.
• When the time continues to increase and becomes comparable to the relax-
ation time of the entire polymer chain, the polymer starts to flow and the
viscous deformation (ε3) makes significant contribution to the total deforma-
tion of the polymer fibers (Figure 16.6D). The total fiber deformation is:
εtotal = ε1 + ε2 + ε3. The viscous deformation is associated with unrecoverable
conformational changes, e.g., the slide of polymer chains over one another,
and hence such deformation is permanent and cannot be recovered after the
removal of applied stress.
Creep can be a serious problem for polymer fibers. In general, the fibers should
have good resistance to creep if they are used in applications that require a contin-
uous load. Crosslinking is an effective means to increase the fibers’ resistance to
creep. Crosslinked polymer fibers do not exhibit permanent viscous deformation
since a crosslinked network cannot flow. Crosslinking cannot eliminate the elastic
and retarded deformations, but still can reduce them. In general, crosslinking is
considered to be effective in reducing the creep behavior as long as over-cross-
linking is avoided. In addition to crosslinking, many other factors affect the creep
behavior of polymer fibers. In general, the creep of polymer fibers decreases with
increases in molecular weight, molecular orientation, molecular polarity, crystal-
linity, and glass transition temperature, since they all reduce the molecular mobil-
ity of polymer chains.
16.2.2 Stress Relaxation

In a stress relaxation test, the polymer fiber is deformed to a given value of strain
and the stress needed to maintain that strain is monitored as a function of time
(Figure 16.7). Stress relaxation tests can be easily conducted on Instron or any
displacement-controlled testing instruments, and Figure 16.8 shows typical
stress relaxation curves. At a constant strain, the polymer chains change their
Figure 16.7. Stress relaxation test of a viscoelastic fiber.
Figure 16.8. Stress relaxation of a typical polymer fiber under different constant strains.
conformations, slide over one another, and so on. As a result, the stress decreases
with time. Higher strain leads to higher initial stress, but faster stress relaxation.
Stress relaxation is not desirable in applications where constant strains are re-
quired. Crosslinking is effective in increasing the fibers’ resistance to stress relax-
ation. Figure 16.9 shows the stress relaxation of crosslinked and uncrosslinked
polymer fibers. Crosslinked polymer fibers do not exhibit permanent viscous flow,
and hence the stress does not relax to zero under a constant strain. However, the
stress of uncrosslinked polymer fibers could reduce to zero under strain when
given a sufficiently long time.
Temperature also affects the stress relaxation of polymer fibers. Figure 16.10
shows the effect of temperature on the stress relaxation of polymer fibers. When
the temperature is significantly greater than the glass transition temperature, the
Figure 16.9. Stress relaxation of crosslinked and uncrosslinked polymer fibers.
Figure 16.10. Effect of temperature on the stress relaxation of viscoelastic polymer fibers.
polymer chain mobility is high and the stress is relaxed rapidly. When the tem-
perature is significantly lower than the glass transition temperature, the polymer
chains are “frozen” and the stress relaxation is slow, and sometimes is negligible.
When the temperature is not too far away from the glass transition temperature
(e.g., tens of degrees above or below Tg), the stress relaxation is apparent due to
the moderate molecular mobility.
Other factors that affect the stress relaxation behavior of polymer fibers in-
clude, but are not limited to molecular weight, molecular orientation, molecular
polarity, crystallinity, and moisture or other additives. In general, the stress re-
laxation of polymer fibers decreases with increases in molecular weight, molecu-
lar orientation, molecular polarity, crystallinity, and glass transition temperature.
However, the introduction of moisture or other small molecules into polymer fi-
bers can facilitate faster stress relaxation since these small molecules can improve
the molecular mobility of polymer chains.
16.2.3 Dynamic Mechanical Properties

Creep and stress relaxation tests are convenient for studying fibers’ responses in a
large time scale (e.g., from minutes to days), but they are less accurate when the
time scale is short (e.g., seconds or less). The most useful method for determining
the viscoelastic behavior of polymer fibers in short time scales is the measure-
ment of dynamic mechanical properties. In a typical dynamic mechanical test, a
sinusoidally varying stress is applied to the fiber sample and the resultant strain is
monitored. In a steady state, the strain also is sinusoidal, having the same angular
frequency but retarded by a phase angle δ. This is analogous to the retarded strain
observed in creep tests.
The sinusoidally varying stress, σ(t), in a dynamic mechanical test can be writ-
ten as:
σ ( t ) = σ 0 sin ωt (16.3)
where σ0 is the maximum stress, and ω the angular frequency of the applied stress
in radians per second. The angular frequency ω is equal to 2πf, where f is the
frequency in cycles per second. If this stress is applied to an ideal elastic material,
the resultant strain, ε(t), would be completely in-phase with the applied stress and
would be written as:
σ (t )
ε (t ) = = ε0 sin ωt (16.4)
E
where E is the modulus of the material, and ε0 the maximum strain. Based on
Equations 16.3 and 16.4, the theoretical stress-time, strain-time, and stress-
strain curves of an ideal elastic material can be obtained, and they are shown in
Figure 16.11. From Figures 16.11A and B, it is seen that the strain can follow the
varying stress without any delay. The normalized stress-strain curve shown in
Figure 16.11B is linear since an ideal elastic material obeys the Hooke’s law. The
network (ΔW) done by the varying stress in a loading cycle (e.g., 0 – 2π) can be
given by:
2π/ ω
∆W = ∫ σde = ∫ σ 0 sin ωtd ( ε0 sin ωt ) = 0 (16.5)
0
Figure 16.11. (A) stress-time, (B) strain-time, and (C) normalized stress-strain curves of
an ideal elastic material.
This indicates the energy consumed during the dynamic mechanical test is com-
pletely stored as the internal energy by the ideal elastic deformation, and the en-
ergy can be completely recovered.
If the sinusoidally varying stress is applied on an ideal viscous material, the
strain rate can be written as:
d ε (t ) σ (t ) σ 0 sin ωt
= = (16.6)
dt η η
where η is the viscosity of the material. The resultant strain can then be expressed
by:
σ 0 sin ωt  π
ε (t ) = ∫ dt = ε0 sin  ωt −  (16.7)
η  2
Figure 16.12 shows the stress-time, strain-time, and stress-strain curves of an ide-

al viscous material, which are based on Equations 16.3 and 16.7. The strain of an
ideal viscous material is exactly π/2 out-of-phase with the stress (Figures 16.12A
and B). The normalized stress-strain curve is a perfect circle (Figure 16.12C). The
net work done in a loading cycle is equal to the area surrounded by the circle:
∆W = ∫σd ε = πσ 0 ε0 (16.8)
Figure 16.12. (A) stress-time, (B) strain-time, and (C) normalized stress-strain curves of
an ideal viscous material.
Since the viscous deformation of an ideal viscous material is unrecoverable, the

energy consumed in a cycle is released in the form of heat.
Polymer fibers are viscoelastic. If the sinusoidally varying stress is applied on
a viscoelastic polymer fiber, the resultant strain can be given as:
ε ( t ) = ε0 sin ( ωt − δ) (16.9)
where the phase angle δ is between 0 and π/2. Figure 16.13 shows the correspond-
ing stress-time, strain-time, and stress-strain curves. The strain of a viscoelastic
polymer fiber lags behind the sinusoidally varying stress by a phase angle δ. The
normalized stress-strain curve has an oval shape. The net work done in a loading
cycle is the area surrounded by the oval:
2π/ ω
∆W = ∫ σde = ∫ σ 0 sin ωtd  ε0 sin ( ωt − δ) = πσ 0 ε0 sin δ (16.10)
0
In the dynamic mechanical test of a viscoelastic polymer fiber, the work is done
to: (i) alter the bond lengths and angles, (ii) induce recoverable changes in poly-
mer chain conformations, e.g., molecular orientation, strain-induced crystalliza-
tion, etc., and (iii) introduce pure viscous flow. The strain caused by changes
of bond lengths and angles is a pure elastic response, and the strain involved in
the recoverable conformational changes is a retarded response. These two types
of responses are reversible and they produce no net work over a loading cycle.
However, the work done for the viscous flow is used to overcome the polymer
chain friction and is converted irreversibly to heat.
Figure 16.13. (A) stress-time, (B) strain-time, and (C) normalized stress-strain curves of a
viscoelastic polymer fiber.
Equations 16.3 and 16.9 are very useful in treating the viscoelastic behavior of
polymer fibers. However, in practical applications, they often are rewritten to the
following two equations by “resetting the zero” and considering the stress “lead-
ing” the strain by a phase angle:
ε ( t ) = ε0 sin ( ωt ) (16.11)
and
σ ( t ) = σ 0 sin ( ωt + δ) = ( σ 0 cos δ) sin ωt + ( σ 0 sin δ) cos ωt (16.12)
Equation 16.12 expresses the stress in terms of an in-phase component and an out-

of-phase component with respect to the strain. The in-phase component, (σ0 cosδ)
sin ωt, is the driving force the elastic response, and the out-of-phase component,
(σ0 sinδ) cos ωt, is associated with the viscous response. The stress-strain relation-
ship now can be defined in terms of these in-phase and out-of-phase components
by:
σ ( t ) = ε0 ( E 'sin ωt + E "cos ωt ) (16.13)
where E’ and E” are the storage modulus and loss modulus, defined as:
σ0
E'= cos δ (16.14)
ε0
and
σ0
E" = sin δ (16.15)
ε0
The storage modulus is a measure of the elasticity of the fiber, and it indicates the
fiber’s ability to storage energy. Energy storage occurs as polymer chains are dis-
torted from their equilibrium positions by small-scale motions, such as changes
of bond lengths and angles. On the other hand, the loss modulus represents the
capability of a fiber to dissipate energy as heat, owing to large-scale viscous mo-
tions of polymer chains.
Since the modulus of a viscoelastic fiber has two components, it can be written
as a complex quantity E*:
E * = E ' + iE " (16.16)
where i = −1 . In addition, the loss factor, tan δ, also can be obtained:
E"
tan δ = (16.17)
E'
Tan δ is a useful quantity and it is a measure of the ratio of energy lost to the
energy stored or recovered. As a result, tan δ is an important indication of the
viscoelasticity of fibers.
Figure 16.14 shows the idealized frequency dependence of E’ and tan δ for
an amorphous polymer fiber. At low frequencies, the polymer fiber has sufficient
time to exhibit the viscous deformation and is in a rubbery state. As a result, the
storage modulus E’ is low at low frequencies. When the frequency increases and
the time scale of experiment (or observation) becomes comparable to the polymer
relaxation time, the polymer fiber enters the viscoelastic state and the storage
modulus increases rapidly. Eventually, the storage modulus increases by several
orders of magnitude and levels off after the polymer fiber enters the glassy state.
From Figure 16.14, it also is seen tan δ goes through a maximum and exhibits a
peak in the frequency range where the storage modulus is changing its value from
one characteristic of the rubbery state to that of the glassy state.
The curves shown in Figure 16.14 are idealized for amorphous polymer fibers.
Most commercial polymer fibers are semicrystalline and are more complex, due
to the superposition of the behaviors of the crystalline and amorphous phases. In
general, with increase in degree of crystallinity, the storage modulus increase, but
the tan δ value decreases. However, the effect of crystallinity is different in differ-
ent frequency ranges, and the superposition of the behaviors of the crystalline and
amorphous phases is not linear, especially when the degree of crystallinity is high
and the amorphous regions are constrained by the crystalline phase. Because of
Time-Temperature Equivalence 323
Figure 16.14. Frequency dependence of E’ and tan δ for an amorphous polymer fiber.
these factors, the behavior of semi-crystalline polymer fibers is much less uniform
than amorphous fibers, and often presents individual idiosyncrasies that have to
be treated separately.
In addition to crystallinity, polymer chain rigidity, side groups, molecular
weight, molecular weight distribution, and crosslinking also affect the dynamic
mechanical properties of polymer fibers. In general, fibers made of rigid polymer
chains have smaller tan δ than those of flexible polymer chains. However, flex-
ible polymer chains with side groups can lead to fibers with high tan δ if the side
groups increase the intermolecular friction. The tan δ value typically increases
by increasing molecular weight and molecular weight distribution, but it often
decreases with crosslinking. These are just general trends. In reality, the effects of
these structural parameters are complex and may vary from fiber to fiber.
16.3 TIME-TEMPERATURE EQUIVALENCE

16.3.1 Time-Temperature Equivalence
One important feature of the viscoelastic behavior of polymer fibers is that time
and temperature are inextricably intertwined. Taking the stress relaxation test as
an example, the fiber modulus can be measured as a function of temperature. In
such a test, the time scale of experiment is important, and hence the modulus
needs to be measured over some arbitrary short time period, e.g., 1 minute. The
polymer fiber is, therefore, stretched instantaneously to a given value of strain
and after 1 minute, the stress is measured to calculate the modulus. Figure 16.15
shows a plot of this modulus as a function of temperature. When the temperature
is lower than Tg, the polymer fiber is at its glassy state and the deformation mainly
is caused by the changes in bond lengths and angles, or other local motions. The
Figure 16.15. Modulus as a function of temperature for a polymer fiber while holding the
time constant.
large-scale conformational changes are “frozen”, at least at the time scale of ob-
servation. As a result, the fiber has very high modulus. As the temperature in-
creases to the vicinity of Tg, there now is sufficient thermal energy that enables the
movement of polymer chain segments. The polymer fiber still does not have time
to display viscous response, but the retarded deformation becomes possible, lead-
ing to a rapid decrease in fiber modulus. At this stage, the polymer fiber presents
clear viscoelastic behavior. As the temperature goes beyond Tg, but still is lower
than the melting temperature (Tm), the time scale of conformational relaxation
processes becomes shorter and the polymer chains can adjust their conformations
to orient themselves. The polymer chains in the amorphous phase can stretch out
between the entanglement points or crystallites and the fiber exhibits a rubbery
plateau, in which the modulus remains constant or decreases slowly. The rubbery
plateau could be short if the fiber is highly crystallized. As the temperature con-
tinues to increase and reaches the vicinity of Tm, the polymer fiber starts to melt
and can no longer hold its shape due to the viscous response. At this stage, the
modulus decreases rapidly. The melting of the polymer fiber can be avoided by
introducing a chemically crosslinked polymer structure.
Similar results can be obtained if the test is conducted by making time a vari-
able and holding the temperature constant. To be more specific, the polymer fi-
ber now is stretched to a chosen value of strain and the stress is measured as a
function of time. The calculated modulus is plotted against time, as shown in
Figure 16.16. At short time periods, the modulus is high and the fiber is in the
glassy state. As the time becomes longer and reaches a critical value, the modu-
lus decreases rapidly, corresponding to the Tg in the modulus-temperature curve
shown in Figure 16.15. As the time goes on, the rubbery plateau is encountered.
Time-Temperature Equivalence 325
Figure 16.16. Modulus as a function of time for a polymer fiber while holding the tem-
perature constant.
When the time reaches the second critical value, the modulus again decreases
sharply, corresponding to the Tm in the modulus-temperature curve. It must be
noted if the test is conducted at room temperature, the time needed for obtain-
ing the entire modulus-time curve shown in Figure 16.16 could be very long, of
the order of hundreds or even thousands of years. The test can be accelerated by
increasing the temperature; however, the beginning part of the curve may be lost
at higher temperatures. As a matter of fact, most polymer fibers already are in the
transition or rubbery state at room temperature.
Comparing Figures 16.15 and 16.16, it can be concluded that increasing the
time has the same effect as increasing the temperature. However, increasing the
test time is not practical in many applications. Sometimes, it might take hundreds
of years to produce the entire stress relaxation curve shown in Figure 16.16, espe-
cially at low temperatures. In cases like this, the viscoelastic tests can be acceler-
ated by using the time-temperature equivalence, expressed in terms of a superpo-
sition principle. The time-temperature superposition principle is discussed in the
following section.
16.3.2 Time-Temperature Superposition Principle

Figure 16.17 illustrates the time-temperature superposition principle by using
stress relaxation curves. To obtain Figure 16.17A, the stress relaxation tests can
be carried out in a relatively short time scale at different temperatures: T0 < T1 < T2
< T3 < T4 < T5. The stress relaxation becomes faster as the temperature increases.
As shown in Figure 16.17B, the effect of increasing the temperature is simply to
Figure 16.17. Illustration of the time-temperature superposition principle: (A) stress relax-

ation curves obtained at different temperatures, (B) shifting of stress relaxation curves at
different temperatures to the reference temperature T0 and a plot of the shift factor αT, and
(C) master stress relaxation curve obtained after superimposing the curves obtained at dif-
ferent temperatures.
shift the stress relaxation curves along the horizontal axis without changing the
shape. To complete the shifting, a time-temperature shift factor (αT) is defined
as the horizontal shift that must be applied to a response curve measured at an
arbitrary temperature (e.g., T1, T2, T3, T4, or T5) in order to move it to the curve
measured at the reference temperature (e.g., T0). By shifting the curves obtained
at different temperatures, the entire master stress relaxation curve at the reference
temperature T0 can be constructed, as shown in Figure 16.17C. In addition to
stress relaxation, creep tests can be superimposed in the same fashion.
For time-temperature superposition, the key is to determine the shift factor
required to superimpose stress relaxation or creep curves obtained at one tempera-
ture to those obtained at another temperature. If the relaxation time of the polymer
fibers obeys the Arrhenius relation, the shift factor may be expressed as:
∆E *  1 1 
log αT = − (16.18)
2.303R  T Tref 

Models of Viscoelastic Behavior 327
where ΔE* is the activation energy of the process, R the gas constant, and T the
temperature of interest, Tref the reference temperature, and the factor 2.303 = ln 10
is the conversion between natural and base-10 logarithms.
Another popular approach to obtain the shift factor is to use the empirical WLF
equation developed by Williams, Landel, and Ferry:
−C1 (T − Tref ) (16.19)

log αT =
C2 + T − Tref
where C1 and C2 are constants whose values depend on the polymer structure and
choice of reference temperature. For many polymers, the C1 and C2 values are
17.44 and 51.6, respectively, if the reference temperature is Tg. In this case, the
WLF equation can be rewritten as
−17.44 (T − Tg )
log αT = (16.20)
51.6 + T − Tg
16.4 MODELS OF VISCOELASTIC BEHAVIOR

One convenient way to describe the viscoelastic behavior of polymer fibers and
to visualize the corresponding molecular motions is to employ “spring-dashpot”
mechanical models.
16.4.1 Ideal Spring and Ideal Dashpot

Mechanical models consist of various combinations of ideal springs and ideal
dashpots. Springs show linear elastic behavior (Figure 16.18A):
σ s = E εs (16.21)
where σs is the stress applied on the spring, εs the resultant strain, and E the spring
modulus. The spring models the instantaneous bond deformation in the polymer
fiber, and its magnitude is related to the fraction of energy stored reversible in the
fiber. On the other hand, dashpots show linear viscous behavior (Figure 16.18B):
σ d = η ε d (16.22)
where σd is the stress applied on the dashpot, ε d the strain rate (i.e., d εd / dt )
produced by the stress, and η the dashpot viscosity. The dashpot models the entro-
pic conformational changes in the polymer fiber.
Figure 16.18. Schematic of (A) ideal spring and (B) ideal dashpot.
16.4.2 Maxwell Model

The Maxwell model consists of an ideal spring and an ideal dashpot connected in
series (Figure 16.19). The spring can be visualized as representing the elastic or
energetic component of the viscoelastic behavior, while the dashpot represents the
conformational or entropic component. In the Maxwell model, the stresses on the
spring and dashpot are the same, and they equal to the applied stress (σ), while the
resultant total strain (ε) is the sum of the strains of the spring and dashpot:
σ = σ s = σ d (16.23)
and
ε = εs + εd (16.24)
The stress and strain can be related by differentiating the strain equation and writ-
ing the spring and dashpot strain rates in terms of the stress:
d ε d ε s d εd 1 d σ σ
= + = + (16.25)
dt dt dt E dt η
This is the so-called “governing equation” for the Maxwell model.

Figure 16.19. Schematic of the Maxwell model.
In a stress relaxation test, the polymer fiber is stretched to a constant strain ε0, i.e.,
dε
= 0 (16.26)
dt
The governing equation then can be integrated to obtain:
t
E dσ
σ
∫0 η dt = −σ∫ σ (16.27)
0
where σ0 is the initial stress at t = 0. As a result, the stress needed to maintain the
constant strain can be written as a function of time:
 E   t
σ ( t ) = σ 0 exp  − t  = σ 0 exp  −  (16.28)
 η   τ
where τ is equal to η/E and is called the relaxation time. The relaxation time is
strongly dependent on the temperature and other factors that affect the molecular
mobility. In general, with increase in temperature, the relaxation time decreases.
Based on Equation 16.28, the modulus of the fiber during the stress relaxation test
also can be obtained:
σ (t ) σ0  t  t
E (t ) = = exp  −  = E0 exp  −  (16.29)
ε0 ε0  τ  τ
where E0 is the initial modulus at t = 0.

Figure 16.20A shows the stress-time curve for the Maxwell model in a stress
relaxation test. At t = 0, the initial stress σ0 is the result of the elastic response
(represented by the spring). As time goes on, the stress decreases due to the contri-
bution of viscous response (represented by the dashpot). Therefore, the Maxwell
model gives a reasonable approximation of the stress relaxation behavior of vis-
coelastic polymer fibers in a relative short period of time. However, when the
time scale of experiment become longer, the Maxwell model is no longer able to
reproduce the entire stress relaxation curve (e.g., Figure 16.16). In a real polymer
fiber, many different types of conformational changes could occur at different
temperatures, and each conformational change is characterized by a different re-
laxation time. The Maxwell model only contains one relaxation time, and hence
it cannot reproduce the entire stress relaxation curve over a long period of time.
Nevertheless, the Maxwell model is a classical initial treatment of the viscoelastic
behavior and often is used to describe the stress relaxation of polymer fibers in a
relatively short time period.
Another major drawback of the Maxwell model is that it cannot be used to
describe the creep behavior. In a creep test, the constant stress σ0 is applied to the
polymer fiber and the governing equation becomes:
dε σ
= (16.30)
dt η
As a result, the strain-time curve of the Maxwell model is a straight line

(Figure 16.20B) and this does not represent the creep behavior of polymer fibers.
16.4.3 Kelvin-Voigt Model

The Maxwell model is incapable of describing the creep behavior because it can-
not model the retarded response under a constant stress. The Kelvin-Voigt model
(Figure 16.21), consisting of an ideal spring and an ideal dashpot connected in
Figure 16.20. Behavior of the Maxwell model in (A) stress relaxation and (B) creep tests.
Figure 16.21. Schematic of the Kelvin-Voigt model.
parallel, can represent the retarded response and hence it can describe the creep
behavior of polymer fibers. In the Kelvin-Voigt model, the total stress is the sum
of the stresses of the spring and dashpot, while the strains on the spring and dash-
pot are the same and they equal to the applied strain:
σ = σ s + σ d (16.31)
and
ε = εs = εd (16.32)
This gives the following governing equation:
dε
σ = Eε + η (16.33)
dt
In a creep test, a constant stress σ0 is applied onto a fiber at t = 0. The governing
equation can be rewritten as:
d ε ε σ0
+ = (16.34)
dt τ η
Solving this linear differential equation, the strain during creep can be described
as a function of time:
σ0   t 
ε (t ) = 1 − exp  − τ   (16.35)
E   
Here, τ is called retardation time, instead of relaxation time, since this is a creep
test. If the constant stress is released at t = t’, the elastic component (i.e., the
spring) would retard the fiber and the strain decreases according to the following
equation:
 t −t'
ε ( t > t ') = ε ( t ') exp  −  (16.36)
 τ 
Figure 16.22A shows the strain-time curve of the Kelvin-Voigt model. After
applying a constant stress σ0, the strain increases exponentially and approaches
to the maximum value (σ0/E) at a rate that depends on the retardation time τ.
However, if the stress is released at t = t’, the strain starts to decrease exponential-
ly and the rate of decrease also is dependent on the retardation time. The Kelvin-
Voigt model does not include the contribution of permanent viscous deformation.
As a result, the Kelvin-Voigt model gives a reasonable approximation of the creep
behavior only in a relatively short time period, when the permanent viscous defor-
mation is not apparent. In addition, like Maxwell model, the Kelvin-Voigt model
also is a single relaxation (or retardation) time type of model.
Another disadvantage of Kelvin-Voigt model is that it cannot be used to de-
scribe the stress relaxation behavior of polymer fibers. Under a constant strain ε0,
the governing equation of the Kelvin-Voigt model becomes:
d ε0
σ ( t ) = E ε0 + η = E ε0 (16.37)
dt
This means in the Kelvin-Voigt model, the stress is constant during the stress re-
laxation test (Figure 16.22B), which is not true for real polymer fibers.
16.4.4 Four-Element Model

The Maxwell model does not represent the creep behavior, while the Kelvin-Voigt
does not describe the stress relaxation behavior. This problem can be addressed by
Figure 16.22. Behavior of the Kelvin-Voigt model in (A) creep and (B) stress relaxation tests.
combining more springs and dashpots in various ways to create models that can
describe both creep and stress relaxation behaviors. One example is the so-called
“four-element model”, consisting of one Maxwell sub-model and one Kelvin sub-
model connected in series, as shown in Figure 16.23A. The four-element model
has four parameters: Maxwell spring modulus EM and dashpot viscosity ηM and
the Kelvin-Voigt spring modulus EK-V and dashpot viscosity ηK-V. This simple four-
element model can be used to describe both the creep and stress relaxation behav-
iors of polymer fibers. Taking the creep test as an example, the strain produced
by the constant stress σ0 can be obtained by simply summing the terms for the
Maxwell and Kelvin sub-models:
σ0 σ   t  σ 0
ε (t ) = + 0 1 − exp  − τ   + η t (16.38)
EM EK −V    M
Equation 16.38 is basically the sum of elastic (ε1), retarded (ε2), and viscous (ε3)
responses, as illustrated in Figure 16.23B.
16.4.5 Multi-Relaxation Time Model

The mechanical models discussed above are based on single relaxation (or retar-
dation) time. Real polymer fibers have a spectrum or distribution of relaxation and
retardation times due to the existence of different types of conformational chang-
es. One convenient way to introduce a range of relaxation times into the problem
is to construct models consisting of a number of Maxwell and/or Kelvin-Voigt
sub-models connected in parallel and/or series. Figure 16.24 shows a Maxwell-
Wiechert model, which is constructed by connecting an arbitrary number of
Figure 16.23. Schematics of (A) the four-element model, and (B) the corresponding creep-
recovery curve.
Figure 16.24. Schematic of the Maxwell-Wiechert model.
Maxwell sub-models in parallel. In this model, the total stress is the sum of the
stresses experienced by each Maxwell sub-model, while the total strain is the
same in each of the sub-models, i.e.:
n
σ = σ1 + σ 2 +…+ σ n = ∑σ i (16.39)
i =1
and
ε = ε1 = ε2 = … = εi = … = εn (16.40)
where σi and εi are the stress and strain of the ith Maxwell sub-model, respectively.
The governing equations for each sub-model are:
d ε 1 d σ1 σ1
= +
dt E1 dt η1
dε 1 dσ2 σ2
= +
dt E2 dt η2
d ε 1 d σi σi
= +
dt Ei dt ηi
…
dε 1 dσn σn
= + (16.41)
dt En dt ηn
where Ei and ηi are the modulus and viscosity of the ith Maxwell sub-model, re-
spectively. In a stress relaxation test, the strain is constant (dε/dt = 0) and hence
the stress-time relationship for each sub-model can be given as:
 t 
σ1 ( t ) = ε0 E1 exp  − 
 τ1 
 t 
σ 2 ( t ) = ε0 E2 exp  − 
 τ2 
…
 t 
σ i ( t ) = ε0 Ei exp  − 
 τi 
…
 t 
σ n ( t ) = ε0 En exp  −  (16.42)
 τn 
where τi is the relaxation time of the ith Maxwell sub-model. The total stress of the
Maxwell-Wiechert model now can be written as:
n
 t 
σ ( t ) = ε0 ∑Ei exp  −  (16.43)
i =1  τi 
The total modulus of the Maxwell-Wiechert model also can be obtained:
σ (t ) n
 t 
E (t ) = = ∑Ei exp  −  (16.44)
ε0 i =1  τi 
Research shows by selectively adjusting the relaxation times of appropriate

number of sub-models, Equations 16.43 and 16.44 can give a reasonably good
representation of the stress relaxation behavior of polymer fibers over a wide time
period. This indicates the stress relaxation of polymer fibers can be described
as a superposition of a large number of independent modes of relaxation. Each
relaxation mode is associated with a certain type of conformational change and

is represented by a characteristic relaxation time τi. If there is a continuous dis-
tribution of relaxation times in the studied polymer fibers, the summations in
Equations 16.43 and 16.44 can be replaced with integrals, i.e.:
∞
 t
σ ( t ) = ε0 ∫E ( τ ) exp  −  d τ (16.45)
0  τ
and
∞
 t
E ( t ) = ∫E ( τ ) exp  −  d τ (16.46)
0  τ
The Maxwell-Wiechert model also can be used to describe the creep behavior
of polymer fibers. However, for the creep behavior, it is mathematically more
convenient to create a model involving a range of retardation times by connected
a number of Kelvin-Voigt sub-models in series.
The simple mechanical models discussed above only allow us to describe the
viscoelastic behavior phenomenologically. The in-depth fundamental understand-
ing of the viscoelastic behavior of polymer fibers still requires more complex
microscopic or molecular models. The interested reader is referred to the works
by De Gennes (1971), Ferry (1980), and Rouse (1953).
REFERENCES
[1] Aklonis, J.J., and MacKnight, W.J., Introduction to Polymer Viscoelasticity,
Second edition, Wiley Interscience, 1983.
[2] Boresi, A.P., Schmidt, R.J., and Sidebottom, O.M., Advanced Mechanics of
Materials, Fifth Edition, John Wiley & Sons, 1993.
[3] De Gennes, P.G., “Reptation of a Polymer Chain in the Presence of Fixed
Obstacles”, Journal of Chemical Physics, 55, 572–579, 1971.
[4] Elias, H.G., An Introduction to Polymer Science, VCH Publishers, 1997.
[5] Ferry, J.D., Viscoelastic Properties of Polymers, Third Edition, John Wiley
& Sons, 1980.
[6] Leaderman, H., Elastic and Creep Properties of Filamentous Materials and
Other High Polymers, the Textile Foundation, 1943.
Problems 337

[9] Rouse, P.E., “A Theory of the Linear Viscoelastic Properties of Dilute
Solutions of Coiling Polymers”, Journal of Chemical Physics, 21, 1272–
1280, 1953.
[11] Sperling, L.H., Introduction to Physical Polymer Science, Fourth Edition,
[12] Ward, I.M., and Sweeney, J., Mechanical Properties of Solid Polymers,
Third Edition, John Wiley & Sons, Ltd., 2013.
PROBLEMS
(1) Why do polymer fibers show viscoelastic behavior?
(2) Describe creep, stress relaxation, and dynamic mechanical testing.
(3) Is the creep behavior of a polymer fiber a simple superposition of linear
elastic and viscous responses? Why?
(4) Discuss and explain how you would expect the creep behavior of a polymer
fiber to vary with:
(i) Crystallinity
(iii) Molecular polarity
(iv) Glass transition
(v) Molecular weight
(vi) Crosslinking
(5) Discuss and explain how you would expect the stress relaxation behavior of
a polymer fiber to vary with:
(i) Crystallinity
(vi) Crosslinking
(6) What are storage modulus, loss modulus, and tan δ? What are their
relationships?
(7) Discuss and explain how you would expect the dynamic mechanical proper-
ties of a polymer fiber to vary with:
(i) Crystallinity
(vi) Crosslinking
(8) Describe the time-temperature superposition principle.

(9) Can Maxwell model be used to predict stress relaxation? How about creep?
Why?
(10) Can Kelvin-Voigt model be used to predict stress relaxation? How about
creep? Why?
(11) The four-element model shown below is subject to a creep experiment.
Show how the length increases with time. At time = t, the stress is removed.
Show how the length recovers.
CHAPTER 17
Thermal Properties of Fibers
Thermal properties are important in many applications. For example, the main
function of textile fabrics is to protect the wearer from cold or heat, and to ensure
appropriate heat transfer between the human body and the environment in order
to maintain the physiological thermal balance of the wearer. Composites also need
appropriate thermal properties so they can be useful in aerospace and space indus-
tries. The thermal properties of fibers are the starting point for understanding the
final properties of these products, although many other factors also play important
roles. This chapter deals with the most basic thermal properties of fibers, includ-
ing heat capacity, specific heat, thermal conductivity, thermal expansion and con-
traction, glass transition, melting, and degradation and decomposition.
17.1 HEAT CAPACITY AND SPECIFIC HEAT

17.1.1 Heat Capacity and Specific Heat
Heat capacity is the amount of energy required to raise the temperature of a
substance by one degree. Specific heat, or specific heat capacity, is the heat ca-
pacity divided by the mass of the substance. The unit of specific heat is J/kg·K.
According to the definition of specific heat, the relationship between heat (Q) and
temperature change (ΔT) can be expressed as:
Q = Cm∆T (17.1)
where C and m are the specific heat and mass of the material. It must be noted
this equation does not apply when a phase change, such as melting, is encoun-
tered since the heat added or removed during a phase change does not change the
temperature.
339
340 Thermal Properties of Fibers
Specific heat indicates the ability of a material to store energy in the form of
heat. The specific heats of some fibers and other materials are shown in Table 17.1.
In general, polymer fibers have higher specific heats than non-polymer fibers and
most other materials (except for water, ice, and organic solvents) because a large
amount of energy is needed to excite the vibrational, rotational, and translational
motions of polymer chains in order to raise the temperature. As a result, textile
fabrics could give serious burns when they melt and are in contact with skin.
17.1.2 Effects of Temperature and Humidity

The specific heat of polymer fibers typically increases with increase in temper-
ature. When the temperature approaches the melting temperature of a polymer
fiber, a large amount of heat is needed to melt the polymer, and hence the fiber
exhibits a significantly increased specific heat at the vicinity of the melting tem-
perature (Figure 17.1).
As shown in Table 17.1, liquid water has an extremely high specific heat of
4180 J/kg·K, and hence the absorption of moisture increases the specific heat of
fibers. During the absorption of moisture, the specific heat of fibers is very high
due to the heat of sorption associated with the moisture absorption. If the mois-
ture regain of fibers remains constant, the specific heat (Cwet) of wet fibers can be
related to the moisture regain (MR) by a simple mixture law:
Cdry + 4.2 MR
Cwet = (17.2)
1 + MR
Figure 17.1. Specific heat as a function of temperature for a polymer fiber.

Heat Capacity and Specific Heat 341
Table 17.1. Specific heats of fibers and non-fibrous materials.

Fibers were dried and tested at room temperature.
Material Specific Heat (J/kg·K)

Polyester 1030
Nylon 6,6 1460
Nylon 6 1430
Kevlar 1465
Natural Polymer Fibers
Cotton 1220–1350
Wool 1360
Silk 1380
Rayon 1350–1590
Inorganic Fibers
Carbon 710
Glass 800–850
Other Non-Fibrous Materials
Aluminum 903
Steel 434
Acetone 2180
Ethanol 2720
Glycerol 2427
Water 4180
Ice 2050
Snow 240
Air 1007
Publishing Limited, 2008.; Warner, S.B., Fiber Science, Prentice Hall, 1995.
where Cdry is the specific heat of dried fibers. Equation 17.2 does not consider the
state of water and the structural change of polymer fibers caused by the absorp-
tion of moisture. In reality, the absorbed water molecules exist in different states
and some of them have a specific heat that is similar to that of ice (2050 J/kg·K).
In addition, the absorbed water molecules can increase the intermolecular dis-
tance between polymer chains, which also affect the specific heat of wet fibers.
Therefore, corrections are needed for Equation 17.2 to give more accurate specific
heat values for wet fibers.
17.2 THERMAL CONDUCTIVITY

17.2.1 Longitudinal and Transverse Thermal Conductivities
Thermal conductivity is a measure of the ability of a material to transfer heat.
Figure 17.2 shows the transfer of heat through a material between a cold surface
(T1) and a hot surface (T2). The material thickness, i.e., the distance between the
hot and cold surfaces, is Δx. According to the law of heat conduction, also known
as Fourier’s law, the thermal conductivity (K) is the proportionality constant be-
tween heat flux and temperature gradient:
T2 − T1 ∆T
Q=K =K (17.3)
∆x ∆x
where Q is the heat flux, and ΔT/Δx is the temperature gradient. Equation 17.3
can be rewritten to give:
∆T
K =Q/ (17.4)
∆x
Table 17.2 shows thermal conductivities of some fibers and other materials.
Except for glass fibers, all polymer and carbon fibers are anisotropic due to molec-
ular orientation. For these anisotropic fibers, the thermal conductivity (KL) along
the longitudinal direction is greater than that (KT) on the transverse direction. The
ratio of KL to KT is a good indication of the degree of molecular orientation. For
the same type of fibers, higher KL/KT ratio indicates greater molecular orientation.
Figure 17.2. Transfer of heat through a material between a cold surface and a hot surface.
Thermal Conductivity 343
Table 17.2. Thermal conductivities of fibers and non-fibrous materials.
Thermal Conductivity (W/m·K) Anisotropy

Material
Longitudinal KL Transverse KT KL/KT
Polyester 1.26 0.157 8.01
Acrylic 1.028 0.172 5.93
Nylon 1.43 0.171 8.38
Kevlar
Kevlar 29 3.05 0.192 15.9
Kevlar 49 3.34 0.212 15.8
Kevlar 149 4.74 0.230 20.6
Natural Polymer Fibers
Cotton 2.88 0.243 11.85
Flax 2.851 0.344 8.23
Wool 0.48 0.165 2.91
Silk 1.49 0.118 12.64
Inorganic Fibers
Carbon
Torayca T-300 6.69 0.530 12.6
Torayca M-308 18.33 0.667 27.5
Torayca M-408 58.81 1.215 48.4
Glass 0.04
Other Non-Fibrous Materials*
Aluminum 205
Steel 16–60
Water 0.58
Ice 2.22
Snow 0.05–0.25
Air 0.024
* Most non-fibrous materials are isotropic and only have one thermal conductivity value.
Publishing Limited, 2008.; Warner, S.B., Fiber Science, Prentice Hall, 1995.
From Table 17.2, it also is seen that carbon fibers are thermally conductive,
but polymer and glass fibers are good thermal insulators. Among the polymer
fibers shown in Table 17.2, Kevlar fibers have the highest thermal conductivities
probably because of the close sequence of benzene rings along the polymer main
chains.
17.2.2 Thermal Conductivities of Fiber-Based Products

Polymer and glass fibers are thermal insulators. They often are assembled to form
products, such as bats in sleeping comforters, winter coats, and building insula-
tion, for thermal protection. The thermal conductivity (Ktotal) of these fiber assem-
blies can be estimated by:
K total = f fiber K fiber + f air K air (17.5)
or
1/ K total = f fiber / K fiber + f air / K air (17.6)
where Kfiber and Kair are the thermal conductivities of fiber and air, and ffiber and fair
the volume fractions of fiber and air, respectively. Equation 17.5 gives the upper
limit of the thermal conductivity of fiber assemblies, while Equation 17.6 gives
the lower limit. According to Table 17.2, air has a significantly lower thermal con-
ductivity than polymer and glass fibers, and hence the actual thermal conductivity
of fiber assemblies is largely determined by the amount of air entrapped within
them. In general, the assembly that can trap the most air provides the best thermal
insulation.
Fibers also can be used in composites to adjust their thermal properties. Similar
to fiber-based thermal protection products, the thermal conductivity (Kcomp) of fi-
ber-added composites is between the upper and lower limits determined by the
following two equations:
K comp = f fiber K fiber + f matrix K matrix (17.7)
and
1/ K comp = f fiber / K fiber + f matrix / K matrix (17.8)
where Kmatrix and fmatrix are the thermal conductivity and volume fraction of the
matrix, respectively. Due to fiber’s anisotropic nature, the length and orientation
of fibers also influence the actual thermal conductivity of composites.
17.3 THERMAL EXPANSION AND CONTRACTION

17.3.1 Coefficient of Thermal Expansion
Most materials expand when heated. However, many fibers contract when heated.
This fiber behavior is called thermal contract, or negative thermal expansion.
Thermal Expansion and Contraction 345
The coefficient of linear thermal expansion, αL, of a fiber is defined as:
1 dL
αL = (17.9)
L0 dT
where L0 is the initial length of the fiber, and dL/dT the rate of change of fiber
length per unit temperature change. Similarly, the area thermal expansion coef-
ficient (αA) and volumetric thermal expansion coefficient (αV) of a fiber are defined
as:
1 dA
αA = (17.10)
A0 dT
and
1 dV
αV = (17.11)
V0 dT
where A0 is the initial fiber cross-sectional area, V0 the initial fiber volume, dA/dT
the rate of change of fiber cross-sectional area per unit temperature change, and
dV/dT the rate of change of fiber volume per unit temperature change, respectively.
For isotropic materials, these three thermal expansion coefficients are related
to each other by:
α A = 2 α L (17.12)
and
αV = 3α L (17.13)
However, most fibers are anisotropic and they do not follow these two relationships.
For most fiber applications, the linear thermal expansion coefficient of fibers is
the most important one among the three expansion coefficients. Table 17.3 shows
the linear thermal expansion coefficients of some fibers and non-fibrous materials.
Polymer fibers have negative coefficients of linear thermal expansion although
their corresponding non-fibrous materials have positive coefficients. The negative
coefficients of polymers fibers are caused by the molecular orientation of polymer
chains. Oriented polymer chains have the tendency to increase the entropy of the
system and return to coiled conformations. However, oriented polymer chains
are “frozen” at room temperature. When heated, the increased thermal vibrations
allow the polymer chains to coil on themselves, resulting in negative coefficients
Table 17.3. Linear thermal expansion coefficients of fibers and non-fibrous materials.
Material Linear Thermal Expansion Coefficient (1/°C)
Polymer Fibers
Polyethylene (Spectra® 1000) –0.1 × 10–4
Polyester –10 × 10–4
Nylon –3 × 10–4
Kevlar® 49 –0.04 × 10–4
Inorganic Fibers
Carbon –1.2 × 10–4
Glass 0.05 × 10–4
Other Non-Fibrous Materials
Polyethylene 2 × 10–4
Polyester 0.2 × 10–4
Nylon 1 × 10–4
Sources: Clark, J.F., et. al., Journal of the Textile Institute Transactions, 44, T596-T608, 1953.;
Kumar, S., Indian Journal of Fiber and Textile Research, 16, 52–64, 1991.; Roff, W.J., et. al.,
Fibres, Films, Plastics and Rubbers: Handbook of Common Polymers, Butterworths, 1971.;
Warner, S.B., Fiber Science, Prentice Hall, 1995.
of linear thermal expansion. Although most polymer fibers have negative coef-
ficients of linear thermal expansion, their area and volumetric thermal expansion
coefficients still are positive since polymer chains take up more space at higher
temperatures when they are vibrating more vigorously.
Like polymer fibers, carbon fibers have a negative coefficient of linear thermal ex-
pansion since these fibers also have an oriented structure. However, glass fibers do not
have an oriented structure and have a positive coefficient of linear thermal expansion.
17.3.2 Dimensional Stability and Heat Setting

Dimensional stability of fibers is critically important in many applications. The
linear thermal expansion coefficient discussed above can be used to describe the
dimension stability of fibers. However, in practical applications, the thermal sta-
bility of fibers is commonly evaluated by boiling water shrinkage (BWS) measure-
ment. During the measurement, fibers are placed in boiling water and the BWS
value is obtained at equilibrium by using the following equation:
L − L0
BWS = − (17.14)
L0
Glass Transition 347
where L0 is the initial fiber length, and L the final fiber length in boiling water.
This is a simple and fast measurement of fiber dimensional stability. The presence
of water accelerates the equilibrium of fiber shrinkage since water molecules can
penetrate into the fiber structure and serve as a plasticizer. In general, tension is
not applied during the measurement. However, in some cases, tension is applied
to control the relaxation of polymer chains.
Most fibers shrink when heated, and hence their BWS values are positive. A
higher BWS value indicates greater fiber shrinkage. To minimize the shrinkage of
fibers, it is a common practice to intentionally expose the fibers to at least their
use temperature and allow a small amount of relaxation to occur. This process is
called “heat setting” and has been discussed in Chapter 9. Heat setting of fibers
can be carried out after the fiber drawing process by using dry air, water vapor,
solid heating elements, or heated liquid baths. Figure 17.3 shows an example of
drawing and heat setting processes conducted with heated rolls. The speed of
draw roll is greater than that of feed roll so the fibers can be drawn to obtain high
molecular orientation. The speed of the relax roll is slightly lower (or comparable)
to that of draw roll, allowing the relaxation (i.e., heat setting) of the fibers. During
heat setting, a small portion of the orientation developed during spinning and
drawing is lost. However, heat setting is a necessary process for achieving good
thermal stability for orientated fibers.
17.4 GLASS TRANSITION

In addition to heat transfer and dimensional changes, fibers may experience glass
transition, melting, degradation, and decomposition when they are heated. For
non-polymer fibers, these phenomena do not occur or only occur at relatively
high temperatures. However, these phenomena can be easily observed in polymer
fibers. Sections 17.4–17.6 discuss the glass transition, melting, degradation, and
decomposition of polymer fibers, respectively.
17.4.1 Glass Transition Temperature

Polymer fibers typically are semicrystalline. The amorphous phase of polymer
fibers exhibits glass transition, i.e., a reversible transition from a hard, glassy state
Figure 17.3. Schematic of drawing and heat setting processes of fibers.

to a soft, rubbery state. At the glass transition temperature (Tg), there is large-scale
movement of polymer chain segments in the amorphous phase of polymer fibers.
Figure 17.4 shows the modulus (E)-temperature (T) relationship of polymer fi-
bers. When the temperature is lower than Tg, the relaxation time of the polymer
chain segments is greater than the time scale of observation and the polymer chain
segments do not have sufficient time to move. As a result, the fibers have high
modulus and are in the glassy state. When the temperature is in the vicinity of Tg,
the relaxation time is comparable to the time scale of observation and polymer
chain segments start to move, and as a result, the fibers start to enter the rubbery
state and their modulus decreases rapidly. The modulus of some fibers decreases
by three orders of magnitude in the vicinity of Tg. When the temperature is greater
than Tg, the structural integrity of polymer fibers mainly is maintained by the
physical crosslinking effect of the crystalline phase.
In addition to fiber modulus, polymer fibers experience many other structure
and property changes at Tg. Figure 17.5 shows the typical volume (V)-T and en-
thalpy (H)-T relationships of polymer fibers. Both the volume and enthalpy of
polymer fibers increase with increase in temperature. The slopes of V-T and H-T
curves become larger at temperatures greater than Tg. It must be noted there is no
discontinuity at Tg for both curves. This is an important characteristic of a second-
order transition. In section 17.5, we will discuss the melting of polymer fibers,
which is a first-order transition. In a first-order transition, both the volume and
enthalpy change rapidly at transition temperature, and there are discontinuities in
the V-T and H-T curves.
Figure 17.4. Modulus-temperature relationship of polymer fibers.

Figure 17.6 shows the variations in volumetric thermal expansion coeffi-

cient (αV) and specific heat (C) as a function of temperature for polymer fibers.
Volumetric thermal expansion coefficient and specific heat are, respectively, the
first derivatives of volume and enthalpy with respective to temperature. Both the
expansion coefficient and specific heat increase rapidly and are discontinued at Tg.
17.4.2 Free Volume Theory

Many theories have been developed to explain the glass transition phenomenon.
Among them, the free volume theory proves to be very useful and applicable for a
theoretical description of the glass transition of polymer fibers. According to this
theory, a polymer fiber can be considered to be made up of occupied volume and
free volume. Glass transition is a point at which there is sufficient free volume for
polymer chain segments to move and change their positions (Figure 17.7).
Figure 17.8 shows occupied volume (Vo) and free volume (Vf) as a function
of temperature for the amorphous phase of polymer fibers. The occupied volume
increases with increase in temperature, and the expansion rate of occupied volume
Figure 17.5. Variations in (A) volume and (B) enthalpy in the vicinity of the Tg of polymer
fibers.
Figure 17.6. Variations in (A) volumetric thermal expansion coefficient and (B) specific
heat in the vicinity of the Tg of polymer fibers.
Figure 17.7. Schematic of free volume in the amorphous phase of polymer fibers.
Figure 17.8. Schematic of the free volume theory.
remains constant (i.e., dVo/dT = constant) both below and above Tg. As a result, the
occupied volume can be written as:
dV0
Vo (T ) = Vo (T = 0 ) + T (17.15)
dT
where Vo(T = 0) is the occupied volume at absolute zero.
The temperature dependence of free volume is more complex than that of oc-
cupied volume. Below Tg, the polymer chain segments are “frozen” and the free
volume does not change with temperature, i.e., dVf/dT = 0 (T < Tg). As a result, the
free volume below Tg and at Tg can be written as:
V f (T ≤ Tg ) = V f (T = 0 ) = constant (17.16)
where Vf(T = 0) is the free volume at absolute zero. Above Tg, polymer chain seg-
ments are able to move and the free volume increases with a fixed expansion rate,
i.e., dVf/dT > 0 (T > Tg). The free volume becomes:
dV f
V f (T > Tg ) = V f (Tg ) + (T − T ) (17.17)
g
dT
where Vf(Tg) is the free volume at Tg. The fraction of free volume (f) can then be
expressed as:
Vf
f (T ) = = f (Tg ) + α f (T − Tg ) (17.18)
Vo + V f
where f(Tg) is the fraction of free volume at Tg, and αf the thermal expansion coef-
ficient of free volume. αf is equal to the difference between the thermal expansion
coefficient above and below Tg. For most polymer fibers, f(Tg) and αf are around
0.025 and 4.8×10–4 1/K, respectively.
Many fiber properties are related to the free volume available in the amorphous
phase of the fibers. For example, according to free volume theory, viscosity can be
related to the fraction of free volume by the Doolittle equation:
B
ln η (T ) = A + (17.19)
f (T )
where A and B are constants. The Doolittle equation states when the fraction of
free volume increases, the viscosity decreases. This equation can be rewritten to
give:
η (T )  1 1 
ln = B −  (17.20)
η (Tg )  f (T ) f (Tg ) 
 
Substituting Equation 17.18 into Equation 17.20 leads to:
η (T )  1 1 
ln = B − 
η (Tg )  f (Tg ) + α f (T − Tg ) f (Tg ) 
 
 T − Tg 
B  
=− (17.21)
{ }
f (Tg )  f (Tg ) / α f + T − Tg 
 
Equation 17.21 also can be written as:
 B 
− ( )
 T − Tg
log
η (T )
=
( )
 2.3 f Tg 
(17.22)
( )
η Tg  f (T ) 
 + (T − T )
g

 α 
 f 
g
or
log
η (T )
=
(
−C1 T − Tg ) (17.23)
( )
η Tg C2 + (T − T )g
where C1 and C2 are constants. Equation 17.23 is the famous WLF equation.

Universal constants (C1 = 17.4 and C2 = 51.6) often are used for polymer fibers.
Using the universal constants, Equation 17.23 can be rewritten as:
log
η (T )
=
( )
−17.4 T − Tg
(17.24)
η (T ) 51.6 + (T − T )
g g
Figure 17.9 shows the log η(T)/η(Tg)−T relationship predicted by Equation 17.24.

With increase in temperature, the free volume expands and the viscosity decreases.
The universal C1 and C2 constants are obtained by ﬁtting the WLF equation to a
wide variety of polymers. It must be noted the variation from polymer to polymer
Figure 17.9. Log η(T)/η(Tg)−T relationship predicted by WLF equation.

could be signiﬁcant. In practice, it is better to determine the C1 and C2 constants

experimentally. The universal constants should only be used as a last resort when
there are no sufficient experimental data for the specific fiber of interest.
17.4.3 Factors Affecting the Glass Transition Temperature

Table 17.4 shows the Tg values of some fibers. It has been discussed that at Tg,
the large-scale movement of polymer chain segments becomes possible. It is,
therefore, expected any structural or external factors that influence polymer chain
mobility also will affect the value of Tg. Structural factors include chain flexibil-
ity, intermolecular interaction, molecular weight, crystallinity, and crosslinking.
External factors include additives, pressure, and tensile stress.
Chain flexibility affects the ease of rotation around the primary bonds of the
polymer main chains. Polyethylene fibers are made of flexible polymer chains,
which rotate easily and have excellent mobility. Hence, polyethylene fibers have
a low Tg of around –80°C. The main chains of nylon fibers contain amide groups,
which form strong intermolecular hydrogen bonds. The formation of hydrogen
bonds significantly limits the rotation of primary bonds and leads to increased Tg
values of 75 and 57°C, for nylon 6 and nylon 6,6 fibers, respectively. The bulky
benzene rings on the main chains also can introduce a high energy barrier to bond
rotations, and as a result, polyester fibers have a relatively high Tg of 80°C.
Side groups attached to the main chains have important impacts on the Tg
of polymer fibers. Flexible side groups, including long branches, push polymer
chains further apart, which in turn increases the free volume and lowers the Tg.
Table 17.4. Glass transition temperatures of fibers.
Fiber Glass Transition Temperature (°C)
Polyethylene –80
Polypropylene –20
Polystyrene 100
Polyester 80
Nylon 6 75
Nylon 6,6 57
Polyacrylonitrile 90
Cotton 230
Wool 170
Silk 175
Glass 630 – 785
Sources: Kumar, S., Indian Journal of Fiber and Textile Research, 16, 52–64, 1991.; Warner, S.B.,
Fiber Science, Prentice Hall, 1995.
However, polymers with long flexible side groups or branches are difficult to form
molecular orientation, and hence they are seldom used to make polymer fibers.
Side groups found in polymer fibers typically are rigid and they can raise the
Tg through steric hindrance to bond rotations. For example, polypropylene fibers
have the same type of main chains as polyethylene, but their methyl side groups
hinder the rotation of primary bonds and reduce the mobility of main chains. As
a result, polypropylene fibers have higher Tg (–20°C) than polyethylene fibers.
The size and polarity of the side groups play important roles in determining the
Tg. Compared with polypropylene fibers, polystyrene contains larger side groups
(i.e., benzene rings), which are rigid and have better ability to hinder the move-
ment of polymer chain segments, leading to a higher Tg of 100°C. The nitrile side
groups attached to polyacrylonitrile main chains are polar and can form strong
intermolecular bonds, and hence polyacrylonitrile fibers also have a relatively
high Tg of 90°C. Natural polymer fibers, such as cotton, wool and silk, have large
numbers of side groups that can form strong intermolecular bonds and they have
high Tg values of 230, 170, and 175°C, respectively. Among these fibers, cotton
also has rigid glucose rings on the main chains, and hence it has the highest Tg.
The chain segments on the two ends of the main chain are restricted only at one
end, and they have more freedom of motion than the internal segments, which are
constrained at both ends. With increase in molecular weight, the number of chain
ends per unit volume decreases and the free volume of the system decreases,
which in turn increases the Tg. Figure 17.10 shows the effect of molecular weight
on the Tg of polymer fibers. For relatively short chains, the Tg of polymer fibers
increases rapidly with increase in molecular weight, but it eventually levels off
Figure 17.10. Tg as a function of molecular weight for polymer fibers.

and asymptotically approaches a limiting value at high molecular weights.

According to the free volume theory, the following equation can be used to de-
scribe the relationship between Tg and number average molecular weight ( M n ) of
polymer fibers:
K
Tg = Tg∞ − (17.25)
Mn
where K is a constant related to the excess free volume per chain end, and Tg∞ the
glass transition temperature obtained from a polymer of infinite molecular weight,
i.e., the limiting value shown in Figure 17.10.
Most polymer fibers are semicrystalline. The presence of the crystalline phase
limits the mobility of the polymer chain segments in the amorphous region. As
a result, the Tg increases with increase in degree of crystallinity. There also is a
simple empirical relationship between the Tg and the melting temperature (Tm):
1
Tg = Tm , for symmetrical polymers (17.26)
2
or
2
Tg = Tm , for unsymmetrical polymers (17.27)
3
where Tg and Tm are in degrees Kelvin.
The Tg of polymer fibers can be reduced by the addition of low-molecular-
weight organic additives, such as plasticizers. This is because these organic addi-
tives increase the intermolecular distance and weaken the intermolecular bonds
between neighboring polymer chains. On the other hand, inorganic additives (e.g.,
TiO 2 and SiO2) can increase the Tg of polymer fibers since they reduce the mobil-
ity of polymer chain segments.
The effect of pressure on the Tg is straightforward. The mobility of polymer
chain segments is reduced at high pressures, and hence the Tg of polymer fibers
increases with increase in pressure. In general, Tg increases with pressure at a
rate of 20°C/1000 atm. However, the effect of tensile stress is complex. In some
cases, applying a tensile stress can increase the free volume and reduce the Tg. The
relationship between the Tg and the tensile stress (σ) can be simply expressed as:
Tg = A − B σ (17.28)
where A and B are constants. In some other cases, the tensile stress increases the
Tg because of the development of stress-induced molecular orientation. Whether
the applied tensile stress increases or decreases the Tg depends on the initial struc-
ture of polymer fibers.
17.4.4 Other Transitions

In addition to glass transition, polymer fibers exhibit a number of other transitions
that are associated to different types of molecular motions. Figure 17.11 shows
examples of small-scale molecular motions of polymer fibers. The crankshaft mo-
tion involves the rotation of a small number of carbon atoms around the polymer
main chain (Figure 17.11A). The transition associated with main chain crankshaft
motion often is referred to as the “glass II transition”. This transition occurs at
around –128°C for polystyrene, –120°C for polyethylene, and 5°C for polyvinyl
chloride. In addition to the small-scale motion along the main chain, the side
groups also display rotational vibration around the main chain and/or oscillation
around its bond with the main chain (Figure 17.11B). For example, the transitions
associated with the rotation and oscillation of the benzene rings in polystyrene
are at around 50 and –215°C, respectively. These small-scale molecular motions
provide polymer fibers with good toughness since a significant amount of energy
is consumed to excite these motions.
17.5 MELTING
At temperatures greater than Tg, the integrity of polymer fibers is largely main-
tained by the physical crosslinking effect of the crystalline regions. When the
Figure 17.11. Examples of small-scale molecule motions in polymer fibers: (A) crankshaft

motion of main chain, and (B) rotation and oscillation of side groups.
Melting 357
temperature continues to increase and becomes higher than the melting tempera-
ture (Tm), the crystalline phase of polymer fibers melts. The polymer fibers be-
come sticker, and eventually lose the integrity and contract to molten globules.
Therefore, the melting of polymer fibers is an “irreversible” process. The structure
of polymer fibers can only be reproduced by repeating the fiber formation process.
17.5.1 Melting Temperature

The melting of polymer fibers is a first-order transition, which involves a latent
heat. During such a transition, the crystalline phase absorbs a fixed amount of
energy. Because energy cannot be instantaneously transferred between the crys-
talline phase and its environment, the melting process is associated with “mixed-
phase regions”, in which some parts of the crystalline phase have completed the
transition and others have not.
As a first-order transition, the melting of polymer fibers can be character-
ized by discontinuities in properties such as volume, enthalpy, transparency, etc.
Figure 17.12 shows V-T and H-T relationships of polymer fibers in the vicinity
of Tm. Both the volume and enthalpy of polymer fibers undergo discontinuous
changes at Tm, which are completely different from the continuous V and H chang-
es at Tg (Figure 17.5). The αV-T and C-T curves of polymer fibers in the vicinity of
Tm are shown in Figure 17.13. As mentioned in section 17.4.1, volumetric thermal
expansion coefficient and specific heat are, respectively, the first derivatives of
volume and enthalpy with respective to temperature. It is seen from Figure 17.13
that both the αV-T and C-T curves exhibit peaks at Tm, which also is a characteristic
of first-order transition.
As discussed above, the melting of polymer fibers is a first-order transition and
is associated with discontinuities in many fiber properties. As a result, any prop-
erty whose values are different below and above Tm provides a potential method
for measuring the melting temperature of polymer fibers. In reality, methods that
Figure 17.12. Variations in (A) volume and (B) enthalpy in the vicinity of the Tm of poly-
mer fibers.
Figure 17.13. Variations in (A) volumetric thermal expansion coefficient and (B) specific
heat in the vicinity of the Tm of polymer fibers.
have been used to measure the melting temperature of polymer fibers include
dilatometry, calorimetry, dynamic mechanical testing, stress relaxation, creep,
nuclear magnetic resonance, etc. Compared with the Tm of low-molecular-weight
materials, the observed melting temperature of polymer fibers has some unique
features:
• Low-molecular-weight materials often melt at a single well-defined temper-

ature. During the entire melting process, the temperature of these materials
does not change until a fixed amount of energy is absorbed. However, poly-
mer fibers generally melt over a range of temperatures. The main reason is
that the crystals in polymer fibers span a range of size and perfection. The
smaller and less perfect crystals melt at lower temperatures, and the larger
and more perfect crystals melt at higher temperatures. As a result, the tem-
perature of polymer fibers continues to increase during the melting process.
• The crystalline structure of polymer fibers is imperfect, and hence the ex-
perimentally measured melting temperature of polymer fibers is lower than
the equilibrium melting temperature. For example, the equilibrium melting
temperature of polypropylene is 187°C, but the measured melting tempera-
ture of polypropylene fibers is around 170°C.
• The melting of polymer fibers is more rate sensitive than that of low-molec-
ular-weight materials.
17.5.2 Factors Affecting the Melting Temperature

Table 17.5 shows the experimentally-measured Tm values of some fibers. One
simple way to understand the effects of different factors on the Tm of polymer
fibers is to consider the change of Gibbs free energy (Gm) upon melting. The Gm
Melting 359
Table 17.5. Melting temperatures of polymer fibers.
Fiber Melting Temperature (°C)
Polyethylene 130
Polypropylene 170
Polyester 260
Nylon 6 215
Nylon 6,6 260
Polyacrylonitrile none
Cotton none
Wool none
Acetate 250
Triacetate 300
Sources: Kumar, S., Indian Journal of Fiber and Textile Research, 16, 52–64, 1991.; Morton,
W.E., et. al., Physical Properties of Textile Fibres, Fourth Edition, Woodhead Publishing
Limited, 2008.; Warner, S.B., Fiber Science, Prentice Hall, 1995.; Wiedemann, H.G., et. al.,
Thermochimica Acta, 169, 1–13, 1990.
value of polymer fibers is related to the enthalpy (Hm) and the entropy (Sm) by the
following equation:
Gm = H m − TS m (17.29)
Thermodynamically, the equilibrium Tm is the temperature at which the Gm values

in the crystalline and molten states are the same:
∆Gm = ∆H m − Tm ∆S m = 0 (17.30)
where Δ refers to the difference between the crystalline and molten states.
Equation 17.30 can be rewritten to give the equilibrium Tm value:
∆H m
Tm = (17.31)
∆S m
This indicates the Tm of polymer fibers is related to the ratio of the enthalpy change
ΔHm to the entropy change ΔSm upon melting. It must be noted the equilibrium Tm
obtained by Equation 17.31 is always higher than the experimentally-measured
Tm. However, this equation still is useful since it gives important qualitative in-
sight on the measured Tm of polymer fibers, as discussed below.
The enthalpy change ΔHm is related to the difference in the forces of attraction
between polymer chains in the crystalline phase and the forces between those
same chains when randomly intertwined in the melt. The closely packaged poly-
mer chains in the crystalline phase allow the attractive forces to be maximized.
As a result, polymer fibers with stronger intermolecular bonds between neighbor-
ing polymer chains would have higher Tm than those with weaker intermolecular
bonds. For example, polyethylene fibers only have weak van der Walls forces
between the polymer chains, and hence they have a relatively low Tm value of
around 130°C. In contrast, nylon fibers have amide groups on the main chains and
these groups form strong intermolecular hydrogen bonds. As a result, nylon fibers
have much higher Tm than polyethylene fibers. Attaching polar side groups onto
the main chains also can lead to the formation of strong intermolecular bonds,
which in turn results in high Tm values. One good example is polyacrylonitrile
fibers, in which the nitrile side groups form strong dipole-dipole forces between
polymer chains. The equilibrium Tm of polyacrylonitrile fibers is greater than
300°C, but various chemical reactions, including cyclization and dehydrogena-
tion, occur below 300°C. Hence, the melting of polyacrylonitrile fibers cannot be
observed in common conditions. Wool fibers have various side groups that can
form strong intermolecular bonds. Among these groups, the cysteine side groups
can form covalent disulfide bonds that have similar strength to the primary bonds
in the main chains, and as a result wool fibers do not melt.
The entropy change ΔSm represents the difference in the degree of disorder
between polymer chains in the crystalline and molten states. The ΔSm term can be
written as:
( )
∆S m = k ln Ω melt − ln Ωcrystal (17.32)
where k is the Boltzmann constant (1.38×10–23 J K–1), and Ωmelt and Ωcrystal are,
respectively, the numbers of conformations available in the molten and crystalline
states. In the crystalline state, polymer chains are in a single ordered conforma-
tion (e.g., the planar zig-zag conformation of polyethylene) and hence the Ωcrystal
value is very small. Upon melting, the polymer chains are able to change their
conformations through bond rotations, and the number of available conforma-
tions becomes significantly greater (i.e., Ωmelt >> Ωcrystal). Therefore, the ΔSm term
is mainly determined by the number of conformations available in the molten
state. In general, compared with flexible polymer chains, rigid polymer fibers
have a smaller number of conformations available in the molten state, and hence
they have smaller ΔSm. As a result, rigid polymer chains have higher Tm than flex-
ible polymer chains. For example, polypropylene fibers have methyl side groups,
which make the main chains less flexible than those in polyethylene fibers due to
their steric hindrance to bond rotations. As a result, polypropylene fibers have a Tm
of 170°C, which is much higher than that (130°C) of polyethylene fibers.
Both the ΔHm and ΔSm terms are affected by introducing rigid benzene rings
or glucose rings. For example, the benzene rings on the main chains of polyes-
ter fibers can increase the ΔHm value by forming intermolecular aromatic ring
Degradation and Decomposition 361
association and decrease the ΔSm value by introducing energy barrier to bond rota-
tions. As a result, polyester fibers have a relatively high Tm of 260°C. For cotton
and acetate fibers, the glucose rings decrease the ΔSm value by limiting the bond
rotations and increase the ΔHm value by forming intermolecular hydrogen bonds
through hydroxyl groups. The theoretical Tm of cotton fibers is so high these fibers
decompose before melting. Acetate and triacetate fibers have some of the hy-
droxyl groups acetylated and can form fewer hydrogen bonds than cotton fibers.
Hence, acetate and triacetate fibers have observable Tm values of 250 and 300°C,
respectively, which still are relatively high as compared to many other polymers.
Molecular weight is an important factor affecting the Tm of polymer fibers. In
general, molecular weight does not influence the ΔHm term, which is related to
the attractive forces between polymer chains. However, increasing the molecular
weight can reduce the ΔSm value by decreasing the number of chain ends per
volume. As a result, with increase in molecular weight, the Tm of polymer fibers
increases mainly because of the reduced ΔSm value. It must be noted the effect
of chain ends on the freedom of bond rotations becomes insignificant at very
high molecular weights. Therefore, when the molecular weight exceeds a criti-
cal value, the Tm of polymer fibers becomes relatively constant with continuous
increase in molecular weight.
17.6 DEGRADATION AND DECOMPOSITION

17.6.1 Degradation vs. Decomposition
The degradation of polymer fibers involves several physical and/or chemical pro-
cesses accompanied by small structure changes, which could lead to significant
deterioration of the fiber properties. The decomposition of polymer fibers, on
the other hand, defines the chemical process occurring under the action of heat,
oxygen, chemical agents, etc., and results in the formation of a new structure
completely different from the original. Decomposition can be considered as the
advanced stage of degradation. In many literature reports, the distinction between
the degradation and decomposition is not made. This section focuses on the de-
composition of fibers.
Non-polymer fibers typically are more thermally stable than polymer fibers.
For example, glass fibers can maintain their structure integrity until the Tg, which
ranges from 630 to 785°C. Above Tg, glass fibers lose their physical shape al-
though their chemical structure does not change. Therefore, for most applications,
the maximum use temperature of glass fibers is their Tg. Carbon fibers can keep
their mechanical properties unchanged upon heating to 1600 – 2000°C in the ab-
sence of oxygen. This allows carbon fibers to be used in many high-temperature
applications. The presence of oxygen significantly reduces the thermal stability
of carbon fibers. In air or oxygen, carbon fibers can be oxidized and eventually
burned into carbon dioxide gas. The maximum use temperature of carbon fibers in
air medium typically is 300–350°C, which is still higher than regular polymer fi-
bers. Compared with non-polymer fibers, most polymer fibers start to decompose
at much lower low temperatures, typically ranging from 150 to 250°C.
17.6.2 Decomposition of Polymer Fibers

The decomposition process of polymer fibers is complex. In most cases, polymer
fibers break down into a variety of smaller molecular fragments made up of differ-
ent chemical species. The lighter molecular fragments vaporize immediately upon
their formation, but heavier molecules remain in the condensed phase. While re-
maining in the condensed phase, these heavier molecules undergo further decom-
position to lighter fragments that are more easily vaporized. Some polymer fibers
break down completely so that eventually no solid residue remains. However,
some polymer fibers produce solid residues which can be carbonaceous (char),
inorganic (originating from heteroatoms contained in the polymer chains or ad-
ditives), or a combination of both. Carbonaceous chars can be burned by oxida-
tion in air at higher temperatures. Figure 17.14 show possible mechanisms of the
decomposition of polymer fibers. Decomposition mechanisms can be divided into
reactions involving atoms in the polymer main chains and reactions involving
side groups. The decomposition of most fibers involves the combination of two
or more reactions.
The most common decomposition mechanism involves the breaking of pri-
mary bonds in the polymer main chains. Such chain scission reactions may occur
at chain ends or at random locations along the main chains. End-chain scissions
often result in the formation of monomers, and the process often is known as un-
zipping. Random chain scissions inside the main chains generate monomers and
oligomers, as well as a variety of other chemical species. Crosslinking is another
important reaction involving the polymer main chains and creates bonds between
two adjacent chains. Crosslinking is important for the formation of chars because
it generates a network structure that is less easily volatilized.
The main mechanisms involving side groups are elimination and cyclization
reactions. In elimination reactions, the bonds connecting side groups to the main
Figure 17.14. Possible mechanisms of the decomposition of polymer fibers.

Degradation and Decomposition 363
chains are broken, with the side groups often reacting with other eliminated side
groups. The products of elimination reactions typically are small enough to be
volatile. In cyclization reactions, adjacent side groups react to form covalent
bonds between them, resulting in the formation of a cyclic structure. This process
also is important for char formation.
17.6.3 Thermally Stable Polymer Fibers

There are several approaches to fabricate thermally stable polymer fibers. One
traditional approach to increase the thermal stability of polymer fibers is to ap-
ply coatings of retardants and stabilizers, such as halides and hydrates. Most
flame retardant/stabilizer coatings have one or more of the following functions:
(i) directing the thermal decomposition reactions to low-rate paths, (ii) coating
the fiber surfaces to exclude oxygen, (iii) forming a barrier to prevent release
of small molecules that are combustible, (iv) releasing inert gasses to dilute
combustible molecules, and (v) lowering the temperature by absorbing a large
amount of energy.
Another approach to obtain thermally stabile polymer fibers is to form lad-
der-type structures through high temperature treatment. One good example is the
high temperature treatment of polyacrylonitrile fibers between 200 and 260°C in
air. This high temperature treatment also is called the stabilization process. As
discussed in Chapter 11, the stabilization of polyacrylonitrile fibers can convert
the linear polymer chains into ladder structures that are non-meltable and flame-
resistant. Figure 17.15 shows one possible ladder structure that is formed through
the stabilization process of polyacrylonitrile fibers.
Thermally-stable polymer fibers also can be obtained directly from polymers
that have high decomposition temperatures. Table 17.6 shows a few examples of
such polymer fibers. The aromatic and/or heterocyclic structures on the polymer
main chains make these fibers thermally stable. The maximum use temperatures
of these thermally stable polymer fibers are greater than 300°C, or even 400°C.
Some of these fibers only lose 20 to 30% of their original weight at temperatures
up to 900°C.
Figure 17.15. Possible ladder structure of stabilized polyacrylonitrile fibers.

Table 17.6. Chemical structures of thermally stable polymer fibers.
Fiber Chemical Structure
Polyparaphenylene terephthalamide (Kevlar®)
Polymetaphenylene isophthalamide (Normax®)
Polybenzimidazole (PBI)
Polyparaphenylene benzobisoxazole (PBO)
Polyparaphenylene benzobisthiazole (PBT)
Polyhydroquinone-diimidazopyridine (M-5)
REFERENCES
[1] Beyler, C.L., and Hirschler, M.M., “Thermal Decomposition of Polymers”,
in SFPE Handbook of Fire Protection Engineering, Editor-in-Chief
DiNenno, P.J., Third Edition, National Fire Protection Association, 2002.
[2] Clark, J.F., and Preston, J.M., “Thermoelastic Properties of Synthetic
Fibres”, Journal of the Textile Institute Transactions, 44, T596–T608, 1953.
[3] Ebewele, R.O., Polymer Science and Technology, CRC Press, 2000.
[5] Fu, S.Y., and Mai, Y.W., “Thermal Conductivity of Misaligned Short-Fiber-
Reinforced Polymer Composites”, Journal of Applied Polymer Science, 88,
1497–1505, 2003.
[6] Fujishiro, H., Ikebe, M., Kashima, T., and Yamanaka, A., “Thermal
Conductivity and Diffusivity of High-Strength Polymer Fibers”, Japanese
Journal of Applied Physics, 36, 5633–5637, 1997.
References 365
[7] Hsu, C.K., “Thermal Decomposition Properties of Polymer Fibers”,

Thermochimica Acta, 392–393, 163–167, 2002.
[9] Losi, G.U., and Knauss, W.G., “Free Volume Theory and Nonlinear
Thermoviscoelasticity”, Polymer Engineering and Science, 32, 542–557,
1992.
[10] Mei, Z., and Chung, D.D.L., “Thermal History of Carbon-Fiber Polymer-
Matrix Composite, Evaluated by Electrical Resistance Measurement”,
Thermochimica Acta, 369, 87–93, 2011.
[12] Nechitailo, V.S., “About the Polymer Free Volume Theory”, International
Journal of Polymeric Materials, 16, 171–177, 1992.
[14] Roff, W.J., and Scott, J.R., Fibres, Films, Plastics and Rubbers: Handbook
of Common Polymers, Butterworths, 1971.
[15] Rosenbaum, S., “Polyacrylonitrile Fiber Behavior. I. Mechanisms of Tensile
Response”, Journal of Applied Polymer Science, 9, 2071–2084, 1965.
[16] Rosenbaum, S., “Polyacrylonitrile Fiber Behavior. I. Dependence on
Structure and Environment”, Journal of Applied Polymer Science, 9, 2085–
2099, 1965.
[18] Shlenskii, O.F., Goncharuk, N.I., and Gal’tsov, V.Y., “Determination of the
Coefficient of Thermal Conductivity of Glass and Polymer Fibers”, Glass
and Ceramics, 26, 523–526, 1969.
[19] Sperling, L.H., Introduction to Physical Polymer Science, Fourth Edition,
[21] Wiedemann, H.G., McKarns, T., and Bayer, G., “Thermal Characteristics of
Polymer Fibers”, Thermochimica Acta, 169, 1–13, 1990.
[22] Wróbel, G., Pawlak, S., and Muzia, G., “Thermal Diffusivity Measurements
of Selected Fiber Reinforced Polymer Composites Using Heat Pulse
Method”, Archives of Materials Science and Engineering, 48, 25–32, 2011.
PROBLEMS
(1) What are heat capacity, specific heat, thermal conductivity, coefficient of
thermal expansion, glass transition temperature, melting temperature, and
degradation and decomposition?
(2) Compare the longitudinal thermal conductivity with transverse thermal
conductivity. Why are they different from each other?
(3) How does moisture affect the heat capacity of polymer fibers?
(4) How can you design a textile fabric that can effectively keep the wearer
warm with a given fiber?
(5) Describe the free volume theory.
(6) What are first-order and second-order transitions?
(7) Discuss and explain how you would expect the glass transition temperature
of a polymer fiber to vary with:
(i) Chain flexibility
(ii) Intermolecular interaction
(iii) Molecular weight
(iv) Crystallinity
(v) Crosslinking
(8) In addition to glass transition, there are other types of important second-
order transitions. Give two examples and explain why these second-order
transitions are important.
(9) Discuss and explain how you would expect the melting temperature of a
polymer fiber to vary with:
(i) Chain flexibility
(ii) Intermolecular interaction
(iii) Molecular weight
(iv) Crystallinity
(v) Crosslinking
(10) What are the differences between degradation and decomposition?

(11) What are the possible mechanisms of the decomposition of polymer fibers?
(12) What are the common features of the thermally-stable polymer fibers shown
in Table 17.6?
CHAPTER 18
Electrical Properties of Fibers
The electrical behavior of non-polymer fibers is varied from excellent electrical

conductors (e.g., carbon fibers) to good insulators (e.g., glass fibers). With only a
few exceptions, pure polymer fibers are insulators with electrical conductivities
in the order of 10–16 S cm–1. Static charges easily can be generated and accumu-
lated on the surface of polymer fibers. This could lead to serious consequences,
such as handling problems during fiber processing, breaking down of sensitive
electronic devices, ignition of flammable vapors and dusts in certain environ-
ments, and clinging tendency and annoying electrical shocks during consumer
use. This chapter focuses on the electrical conductivity and static charging of
polymer fibers.
18.1 ELECTRICAL CONDUCTIVITY

18.1.1 Basic Concepts
While applying a voltage to a polymer fiber, the current can be carried by either
electrons or ions. Normally, the conductivity associated to the movement of elec-
trons is called electronic conductivity. The conductivity caused by the movement
of ionic species is called ionic conductivity. Both electronic and ionic conductivi-
ties contribute to the so-called electrical conductivity.
A simple way to understand the electronic conductivity is to examine the avail-
able energies for electrons in the materials. Instead of having discrete energies,
the available energy states form bands. Figure 18.1 shows the schematic of the
electronic band structures of conductors, semiconductors, and insulators. The
lower band is called the valence band, by analogy with the valence electrons of
individual atoms. Electrons in the valence band are bound to individual atoms.
The upper band is called the conduction band because only when electrons are
367
368 Electrical Properties of Fibers
Figure 18.1. Schematic of the electronic band structures of conductors, semiconductors,

and insulators.
excited to the conduction band can current flow in the material. In conductors like
carbon fibers, the valence band overlaps the conduction band and at least a frac-
tion of the valence electrons can move through the material. The conductivities
of conductors are in the range of 10–3 to 106 S cm–1. In semiconductors, there is
a small gap between the valence and conduction bands. This gap can be bridged
by thermal or other excitations. The conductivities of semiconductors range from
10–8 to 10–3 S cm–1. In insulators like polymer fibers, the electrons in the valence
band are separated by a large gap from the conduction band, and the gap is so
large that it is difficult to excite the valence electrons to the conduction band.
Insulators have conductivities ranging from 10–18 to 10–8 S cm–1. Pure polymer
fibers are excellent insulators and their electronic conductivities are in the order
of 10–16 S cm–1 when dried.
Since polymer fibers have low electronic conductivities, their ionic conductivi-
ties cannot be ignored. Depending on the chemical structure of polymer fibers,
ions can be dissociated from the polymer chains by different mechanisms. For
example, in nylon fibers, the momentary ionization of the hydrogen bonds can
generate protons to serve as charge carriers (Figure 18.2). In addition to the ions
generated from the polymer chains, impurities, residues (e.g., catalysts) and ad-
ditives all contribute to the ionic conductivity of polymer fibers. This is why pure
polymer fibers typically have conductivities in the order of 10–16 S cm–1, but im-
pure polymer fibers can have higher conductivities. However, the mobility of ions
typically is low, and as a result most polymer fibers still are insulators even with
the presence of impurities, residues, and additives.
In theory, the overall conductivity (σ) of a polymer fiber is governed by the
following equation:
σ = nq µ (18.1)
where n is the number of charge carriers per unit volume, q the charge per charge
carrier, and μ the mobility of the charge carrier. This equation states conductivity
Electrical Conductivity 369
Figure 18.2. Ionization of hydrogen bond in nylon fibers.
is associated to the concerted motion of charge carriers. In polymer fibers, the

charge carriers can be either electrons or ionic species.
In practice, it is difficult to calculate the conductivity of polymer fibers by
using Equation 18.1. The conductivity of polymer fibers can be obtained by car-
rying out the electrical resistance measurement. According to the Ohm’s law, the
electrical resistance (R) of a fiber can be obtained by:
U
R= (18.2)
I
where U is the voltage in volts, and I the current in amperes. The fiber resistivity
(ρ), a geometry-independent material property, can then be calculated by:
A
ρ= R (18.3)
L
where A is the cross-sectional area of the fiber, and L the length of the fiber. The
fiber conductivity is the reciprocal of resistivity:
1
σ= (18.4)
ρ
18.1.2 Electrical Conductivities of Polymer Fibers

In most polymer fibers, all electrons are bound to the atom nuclei or shared in the
covalent bonds. These electrons are not free to move without being excited over
the large band gap shown in Figure 18.1. Indeed, even if the electrons are excited
to the conduction band, they still are not perfectly free since the trap density along
the polymer chain can be sufficiently large to inhibit the motion of the excited
electrons.
Figure 18.3 shows the conductivity-temperature relationship of polymer fi-
bers. Below the glass transition temperature (Tg), the fiber conductivity increases
slowly with increase in temperature. However, above Tg, the fiber conductivity
increases much faster with temperature. This probably is because the mobility of
Figure 18.3. Electrical conductivity-temperature relationship of polymer fibers.
the entire system is enhanced when the temperature is greater than Tg. Both below
and above Tg, the conductivity-temperature relationship follows the Arrhenius
equation:
σ = σ 0 e − ∆E / kT (18.5)
where ΔE is the activation energy, k the Boltzmann’s constant, and T the absolute
temperature. For many polymer fibers, the activation energies lie between 5 and
15 kcal/mole below Tg and between 30 and 80 kcal/mole above Tg. The high ac-
tivation energies indicate the conductivities in polymer fibers may occur mainly
by the ionic mechanism, especially above Tg. However, evidence also shows the
contribution of electronic conductivity, especially at low temperatures.
As discussed above, the electrical conductivities of pure polymer fibers are
in the order of 10–16 S cm–1, which is approximately 108 times lower than that of
pure water and 1011 times lower than that of water containing impurities capable
of transporting charges such as ions. The relatively high conductivity of water ex-
plains why the introduction of even small amount of water molecules can result in
a significant conductivity increase for many polymer fibers, especially those with
hydrophilic nature. The presence of water molecules provides transport carriers
for electrical charges by facilitating the migration of ions in moisture-absorbed
fibers. In general, hydrophobic fibers, such as polyethylene and polypropylene
fibers, do not absorb moisture and their conductivities show little or no moisture
sensitivity. However, the conductivities of hydrophilic cotton, wool and rayon
fibers could increase by more than 10 orders of magnitude through moisture
Electrical Conductivity 371
absorption. For many hydrophilic polymer fibers, the conductivity between 30

and 90% relative humidity can be related to the moisture content (MC) by:
log σ = n log MC − log K (18.6)
where n and K are constants.

The moisture sensitivity of hydrophilic fibers is affected by the degree of
crystallinity. In general, with increase in crystallinity, the moisture sensitivity of
hydrophilic polymer fibers decreases. This is because the amorphous phase of
polymer fibers is more amenable to both penetration and motion of ions than the
crystalline phase.
18.1.3 Electrically Conductive Polymer Fibers

Although most polymer fibers are insulators, electrically conductive polymer fi-
bers can be obtained by different approaches. One of the simplest approaches is
to incorporate conductive materials, such as carbons, metals, and metal oxides,
directly into polymer fibers. One advantage of this approach is that the conduc-
tivities of the resultant fibers are independent of humidity. The introduction of car-
bons, metals and metal oxides can be realized by: (i) dispersing their particulates
directly into synthetic fibers before or during the fiber extrusion, and (ii) using
them to coat the fiber surface. The challenge of extruding conductive particulate-
added synthetic fibers is the difficulty in achieving homogenous dispersion of
particulates. However, this could be solved by selectively controlling the extru-
sion condition and/or appropriately modifying the particulate surface. The coat-
ing of fiber surface with conductive materials may involve hazardous chemicals,
depending on the coating method used, and this problem is being addressed by
new research and development.
Another approach to obtain conductive polymer fibers is to spin them directly
from electronically conductive polymers. Most conductive polymers have extend-
ed π-bonded, fully conjugated structures, such as fused aromatic rings, double
bonds, lone electron pairs, etc. These polymers in their pure state still are electri-
cal insulators since they do not have intrinsic charge carriers. The required charge
carriers can be provided by partial oxidation (p-doping) of the polymer chains
with electron acceptors (e.g., AsF5, I2) or by partial reduction (n-doping) with
electron donors (e.g., Li, Na, K). Through the doping process, charged defects
(e.g., polarons, bipolarons and solitons) are introduced and are available as the
charge carriers. Table 18.1 shows examples of conductive polymers and the con-
ductivities that they can achieve after being doped. So far, fibers made from these
conductive polymers are in their early research stage and significant efforts are
needed before they can be developed into commercial products.
Table 18.1. Structures and conductivities of conductive polymers.
Conductivity
Polymer Chemical Structure Doping Materials
(S cm–1)
Polyacetylene AsF5, I2, Br2, Li, Na 103 – 104
Polypyrrole BF4−, ClO4−, tosylate 500 – 7.5×103
BF4−, ClO4−, FeO4−,

Polythiophene 10 – 103
tosylate
Polyphenylene AsF5, Li, K 102 – 103
Polyphenylene
AsF5 3 – 5×103
vinylene
Polyaniline HCl 30 – 200
Source: Kumar, D., and Sharma, R.C., European Polymer Journal, 34, 1053–1060, 1998.
18.2 STATE ELECTRICITY

18.2.1 Basic Concepts
All materials are composed of atoms containing positive charges in the nuclei and
an equal amount of negative charges orbiting the nuclei in the form of electrons.
The contact of two materials allows the transfer of electrons across the interface.
The re-distribution of electrons causes the formation of charges on both materials
after they are separated. One material has negative charge (i.e., excess electrons)
and the other material has positive charge (i.e., insufficient electrons). If the ma-
terials are conductive, the charges can equalize themselves by the backflow of
electrons immediately after the separation of two surfaces. However, for most
polymer fibers, electrons are not free to flow because they are bound to the nuclei
or shared in the covalent bonds. As a result, charges can survive on polymer fibers
for a measurable length of time and they are named static charges.
State Electricity 373
The effects of static charges are familiar to most people because we can feel,
hear, and even see the sparks as excess charges are neutralized when brought close
to a grounded conductor, or a region with excess charges of the opposite polarity.
The familiar phenomenon of a static “shock” is caused by the neutralization of
charges. In reality, the generation of static charges on the surface of polymer fibers
can lead to serious consequences. For example, charged fibers may repel from
each other, which makes further processing difficult or even impossible. Charges
generated on fibers may interfere with the operation of sensitive electronic devic-
es, and sometimes overload and break down these devices. In addition, discharges
of sufficient magnitude may cause fires or explosion in an operating room since
the spark generated during discharging can ignite flammable vapors and dusts.
Therefore, in many applications, it is important to take actions to either avoid the
formation of static charges or dissipate them quickly during the processing and
consumer use of polymer fibers.
18.2.2 Mechanisms of Static Charge Generation

Many mechanisms can lead to the formation of static charges. The following
are some common mechanisms. Any one of these mechanisms can lead to static
charging on polymer fibers, but in many cases, more than one mechanism may
work together to generate static charges.
The most important mechanism is contact-induced charge separation (i.e., tri-
boelectric effect). As mentioned in section 18.2.1, electrons can be exchanged
between materials on contact. Materials with sparsely filled outer shells tend to
gain excess electrons, while materials with weakly bound electrons tend to lose
them. This causes one material being negatively charged and the other positively
charged. In addition to electrons, other charged species also can be exchanged
between materials. For example, in fibers made of acidic or basic polymers, or
polymers with space charge layers, ions can be exchanged at the interface of two
contacted fibers. During the contact-induced charge separation, charges redistrib-
ute according to Boltzmann statics, i.e., charges move between the two contact
materials in numbers (n) that depend on the activation energy ∆E:
n = n0 e − ∆E / kT (18.7)
where n0 is the pre-exponential constant, k the Boltzmann’s constant, and T the

absolute temperature.
In contact-induced charge separation mechanism, the polarity of static charges
generated on the materials depends on their relative positions in the triboelec-
tric series. The triboelectric series is an empirically compiled list where materi-
als are arranged from top to bottom depending on their relative ability to lose or
gain electrons, beginning with the most positively charged material and ending
with the material carrying the most negative charge. Table 18.2 shows several
triboelectric series reported by different researchers. Although these series do not
agree with each other completely, there are some similarities. For example, wool
and nylon fibers typically appear at the positive end of the series, cellulose fibers
in the middle, and the synthetic fibers other than nylon at the negative end.
According to the triboelectric series shown in Table 18.2, the polarity of static
charge generated on fibers can be determined. For example, when wool fibers
contact cotton fibers, wool acquires positive charge and cotton negative charge
because cotton has better ability to gain electrons than wool. On the contrary, the
same cotton fibers acquire positive charge when contacting polyethylene fibers be-
cause cotton has a greater tendency to lose electrons as compared to polyethylene.
The contact-induced charge separation does not directly involve relative mo-
tion of the two contact materials. However, motion certainly can enhance the for-
mation of static charges. In many cases, rubbing two polymer fibers produce tem-
perature gradients. Hot spots often develop due to the friction effect, and charges
can move from hot spot to cold surrounding area.
The contact-induced charge separation is the main cause of static charges dur-
ing the processing and consumer use of polymer fibers. However, other static
charge generation mechanisms, such as pressure-induced charge separation (i.e.,
piezoelectric effect), also play important roles. Pressure-induced charge separa-
tion is determined by the ability of polymer fibers to generate static charges in re-
sponse to applied mechanical stress or strain. The nature of the pressure-induced
charge separation is closely related to the formation of electric dipole moments
in fibers. The electric dipole moments can either be induced in crystal lattice sites
with asymmetric charge surroundings or can be carried directly by molecular
groups. The dipole density or polarization can be calculated for crystals by sum-
ming up the dipole moments per unit cell volume. The change of polarization
under a mechanical stress or strain can either be caused by the re-configuration
of the dipole-inducing surrounding or by the re-orientation of molecular dipole
moments under the influence of the external stress or strain. As a result, pressure-
induced charge separation can occur when there is a variation in the polarization
strength, polarization direction, or both, depending on: (i) the orientation of po-
larization within the crystal, (ii) crystal symmetry, and (iii) the applied mechani-
cal stress. Although the change in polarization appears as a variation of surface
charge density upon the crystal faces, pressure-induced charge separation is not
caused by the change in charge density on the surface, but by the change of dipole
density in the bulk. The pressure-induced charge separation often is observed in
natural fibers, such as wool and silk. In synthetic fibers, polyvinylidene fluoride
(PVDF) exhibits piezoelectricity several times greater than other polymer fibers.
Like stress or strain, heating also can generate a separation of charges in poly-
mer fibers. This is so-called heat-induced charge separation (i.e., pyroelectric ef-
fect). The atomic or molecular properties of heat and pressure response are closely
related. Therefore, heat-induced charge separation also is a mechanism that is
Table 18.2. Triboelectric Series.
Hersh and
Lehmick 1949 Henry 1953 Ballou 1954 Montgomery Henniker 1962 Adams 1987
1955
More Positive (+)

Glass Platinum Wool Wool Nylon 11 Nylon
Human hair Formvar Nylon Nylon Nylon 6,6 Wool
Nylon yearn Filter paper Silk Viscose Wool, knitted Silk
Nylon polymer Cellulose Viscose Cotton Silk, wove Paper

acetate
Wool Cellulose Cordura® Silk Polyethylene Cotton
triacetate glycol
Silk Polyethylene Human skin Acetate Cellulose acetate Wood
Viscose Aluminum Fiber glass Lucite® Cotton, woven Hard rubber
Cotton Polystyrene Cotton Polyvinyl Polystyrene Acetate, rayon

alcohol
Paper Copper Glass Dacron® Polyisobutylene Polyester
Ramie Nature rubber Ramie Dynel® Polyvinyl Styroform®

butyral
Steel Dacron® Velon® Natural rubber Orlon®
Hard rubber Chromium Polyethylene Polyacrylonitrile Saran®
Acetate Orlon® Teflon® Polyethylene Polyethylene
Synthetic Polyethylene Polyvinyl Polypropylene

rubber chloride
Orlon® Polyvinyl
chloride
Polyethylene Teflon®
More Negative (-)
Sources: Lehmicke, D.J., American Dyestuff Reporter, 38, 853–855, 1949.; Henry, P.S.H., British
Journal of Applied Physics Supplement No. 2, 4, S31-S36, 1953.; Ballou, T.W., Textile Research
Journal, 24, 146–155, 1954.; Hersh, S.P., et. al., Textile Research Journal, 25, 279–295, 1955.;
Henniker, J., Nature, 196, 474, 1962.; Adams, C. K. Nature’s Electricity, Tab Books, Blue Ridge
Summit, 1987.
related to the change of dipole density or polarization in fibers. All pyroelectric

materials also are piezoelectric.
Charge-induced charge separation (i.e., electrostatic induction) is another im-
portant mechanism for the static charge generation. In insulating polymer fibers,
although the electrons are bound to atoms and are not free to flow between atoms,
they can move within the atoms. For example, if a positively charged object is
brought near a polymer fiber, the electrons in each atom in the fiber are attracted
toward the charged object and move to the side of the atom facing the charge,
while the positive nucleus is repelled and moves slightly to the opposite side of
the atom. Since the negative charges now are closer to the external object than
the positive charges, their attraction is greater than the repulsion of the positive
charges, resulting in a small net attraction toward the inducing charge. This effect
is microscopic, but since there are so many atoms, it could add up to significant
static charges.
18.2.3 Antistatic Treatments

The basic principle for addressing the static charging problem of polymer fibers
is to increase either the ionic or electronic conductivity of the fibers. As discussed
in section 18.1.2, the introduction of water molecules can increase the fiber con-
ductivity by providing transport carriers to facilitate the migration of ions in mois-
ture-absorbed fibers. Therefore, humidity control is a common practice in many
fiber processing facilities. Spraying a mist of water or aqueous solution to the
local surrounding of the fibers also is a common practice. Humidifiers also can be
added to the central heating and air conditioning system. However, even for hy-
drophilic natural polymer fibers such as cotton, wool, and silk that are capable of
absorbing large amount of water molecules, their conductivities still may not be
sufficient in some occasions. In addition, when the relative humidity is reduced,
these hydrophilic fibers become highly insulating again and static charges can be
built up easily. Therefore, it often is necessary to introduce antistatic functionality
to polymer fibers, depending on the intended use.
The application of antistatic agents is the most common approach to obtain
antistatic polymer fibers. Most of the known antistatic agents are hydroscopic
compounds that can improve the ionic conductivity of polymer fibers by ensur-
ing the presence of water molecules and dissolved ions. Antistatic agents can be
ionic or nonionic; however, ionic agents generally are more effective. Nonionic
agents serve mainly as vehicles for carrying ions of other sources, such as impuri-
ties. Ionic agents typically are ionic group-containing molecules that can establish
their functionality without relying on other ion sources. Both ionic and nonionic
agents typically are hygroscopic, and they need water to realize their antistatic
functionality. In this case, the main roles of water molecules are to: (i) lower the
activation energy for the ionization of ionic groups, (ii) increase the number of
the resultant ions, and iii) facilitate the transportation of ions. Many researchers
believe that, for maximum efficiency, the water must be present as a continuous
phase, preferably on the surface, rather than in the bulk. Therefore, most anti-
static agents are applied on the surface of polymer fibers to increase the surface
conductivity rather than the volumetric conductivity. However, in some stages of
processing, antistatic agents can be absorbed into the polymer fibers and they are
no longer just on the surface. In addition, internal antistatic agents that are added
to the bulk of the polymer also are used in some applications.
In addition to ionic conductivity, antistatic polymer fibers also can be prepared
by enhancing the electronic conductivity. As discussed in section 18.1.3, elec-
tronically conductive polymer fibers can be obtained by: (i) dispersing conductive
carbon or metal particles into synthetic polymers before or during fiber spinning,
(ii) coating the fiber surface by a thin layer of electronically conducting materials
such as silver, copper, aluminum, or gold, and (iii) producing fibers directly from
electrically conductive polymers. The advantages of these methods are that the
resultant antistatic properties are durable and do not relay on the presence of water
molecules. However, the conductive component added typically is either black or
colored, and affects the appearance of the final product.
18.2.4 Utilization of Static Charging

There are applications in which the static charging phenomenon of polymer fibers
can be utilized. One such example is filtration. Polymer fibers have large surface
areas and often are used to make dust filters or masks. Statically charging the
polymer fibers in these filters and masks can significantly increase their capability
to capture the dust particles.
Another application that takes advantage of static charges is in the production
of spunbond fibers. Spunbonding is a process for making nonwoven fabrics by
depositing as-spun filaments directly onto a collecting belt in a uniform random
manner, followed by bonding the fibers. During that process, the as-spun fibers
tend to clump together, leading to poor uniformity. This problem can be solved
by charging the fibers prior to the deposition on the collecting belt. Statically
charged fibers repel one another, and as a result, the obtained nonwovens have
better uniformity.
REFERENCES
[1] Adams, C. K. Nature’s Electricity, Tab Books, Blue Ridge Summit, 1987.
[2] Ballou, T.W., “Static Electricity in Textiles”, Textile Research Journal, 24,
146–155, 1954.
[4] Henniker, J., “Triboelectricity in Polymers”, Nature, 196, 474, 1962.
[5] Henry, P.S.H., “Survey of Generation and Distribution of Static Electricity”,

British Journal of Applied Physics Supplement No. 2, 4, S31–S36, 1953.
[6] Hersh, S.P., and Montgomery, D.J., “Static Electrification of Filaments.
Experimental Techniques and Results”, Textile Research Journal, 25, 279–
295, 1955.
[7] Holme, I., McIntyre, J.E., and Shen, Z.J., “Electrostatic Charging of
Textiles”, Textile Progress, 28, 1–85, 1998.
[8] Kumar, D., and Sharma, R.C., “Advances in Conductive Polymers”,
European Polymer Journal, 34, 1053–1060, 1998.
[9] Lehmicke, D.J., “Static in Textile Processing”, American Dyestuff Reporter,
38, 853–855, 1949.
[11] Pionteck, J., and Wypych, G., Handbook of Antistatics, Chemtec Publishing,
2007.
[12] Rebouillat, S., and Lyons, M.E.G., “Measuring the Electrical Conductivity
of Single Fibres”, International Journal of Electrochemical Science, 6,
5731–5740, 2011.
[13] Saito, S., Sasabe, H., Nakajirza, T., and Yada K., “Dielectric Relaxation
and Electrical Conduction of Polymers as a Function of Pressure and
Temperature”, Journal of Polymer Science: Part A-2, 6, 1297–1315 (1968).
[14] Schindler, W.D., and Hauser, P.J., Chemical Finishing of Textiles, Woodhead
Publishing Ltd, Cambridge, 2004.
[15] Smith, F.S., and Scott, C., “The Electrical Conductivity of Poly(ethylene
terephthalate) in the Temperature Range 180–290°C”, British Journal of
Applied Physics, 17, 1149–1154, 1966.
Wiley & Sons, Inc., 2006.
[17] Uchida, E., Utama, Y., and Ikada, Y., “Antistatic Properties of Surface-
Modified Polyester Fabrics”, Textile Research Journal, 61, 483–488, 1991.
[18] Valko, E.I., and Tesoro, G.C., “Polyamine Resins for Finishing of
Hydrophobic Fibers”, Textile Research Journal, 29, 21–31, 1959.
Problems 379
PROBLEMS
(1) What is the difference between electronic and ionic conductivities?
(2) Describe the electronic band structure of polymer fibers.
(3) What is the conductivity-temperature relationship for polymer fibers?
(4) What are the main approaches to obtain electrically conductive polymer
fibers?
(5) Why can static charges be generated easily on polymer fibers?
(6) What are the potential problems if static charges are generated on polymer
fibers?
CHAPTER 19
Frictional Properties of Fibers
Frictional properties of fibers are important because they affect the processing,
structure and properties of all fiber-based products. For example, friction is the
force that holds the fibers together in yarns and fabrics. If there is not sufficient
friction, the strength and structural integrity of yarns and fabrics will be lost.
However, if the friction is too high, it could cause equipment failure, fiber surface
damage, and even fiber breakage. In addition to these two examples where friction
is clearly an advantage or a disadvantage, there are many other properties of fiber-
based products that are affected by friction. Examples are the wear resistance of
fibers and fabrics, the draping behavior of fabrics, the formation of nonwovens,
etc. This chapter addresses the basic principles associated with the frictional prop-
erties of polymer fibers.
19.1 BASIC CONCEPTS

19.1.1 Amontons’ Law
Friction is the force that resists the motion or attempted motion of one surface
against another. Figure 19.1 shows the loading force applied to move one sur-
face over another, the normal force pressing the two surfaces together, and the
resultant friction force. To understand the friction of materials, it is important to
establish the relationship between these forces.
About 500 years ago, Leonardo da Vinci recognized two basic laws of friction:
• The friction force is independent of the contact area between two surfaces.
• The friction force is proportional to the normal force between the two sur-
faces in contact.
381
382 Frictional Properties of Fibers
Figure 19.1. Schematic of loading force, normal force, and the resultant frictional force.
These two laws were later rediscovered by Amontons in 1699 and are ex-
pressed mathematically by using the so-called Amontons’ law:
F = µN (19.1)
where F is the externally applied loading force, N the normal force, and μ the
coefficient of friction.
The friction behavior of polymer fibers does not follow the Amontons’ law.
The deviations from the Amontons’ law will be discussed in sections 19.3 and
19.4. Despite the deviations, the Amontons’ law lays the foundation for under-
standing the friction behavior of many fibers.
19.1.2 Static Friction vs. Kinetic Friction

It is also important to point out the distinction between static friction and kinetic
friction. Static friction is the force that must be overcome to begin sliding. Static
friction force is equal and opposite to the applied loading force up to a maximum
static friction force, and no acceleration or movement occurs below the maximum
force. According to the Amontons’ law, static frictional force is proportional to the
normal force pushing the two surfaces together:
Fs ,� max = µs N (19.2)
where Fs,max is the maximum static friction force, and μs the coefficient of static
friction. On the other hand, kinetic friction is the force resisting continued sliding.
Kinetic frictional force is opposite to the direction of motion, and is also propor-
tional to the normal force:
Fk = µk N (19.3)
where Fk is the kinetic friction force, and μk the coefficient of kinetic friction. For
most materials, the kinetic friction coefficient is independent of the sliding speed
and is smaller than the static friction coefficient.
Basic Concepts 383
When two objects are in contact and a loading force is applied to slide one object
again the other, the relationship between the friction force and the applied loading
force will determine whether the object will slide or not move at all. Figure 19.2
shows the friction forces generated while sliding an object against a flat surface
under different conditions. At all these conditions, the normal force does not change
since the mass of the object remains constant. When the object is at rest and there
is no applied force, the friction force (FA) is zero since there is no attempted mo-
tion (Figure 19.2A). When an external force is applied to push the object to the
verge of sliding, a friction force (FB) is generated to resist the attempted motion
(Figure 19.2B). In this case, FB is equal to the Fs,max described in Equation 19.2.
The object starts to slide when the applied force is greater than FB. The resultant
friction force (FC) is determined by the kinetic friction coefficient and normal force,
as described by Equation 19.3. FC is smaller than FB since the kinetic friction coef-
ficient is smaller than the static friction coefficient. At this condition, the sliding
speed of the object continues to increase since the applied force is greater than FC
(Figure 19.2C). To keep a constant speed, the applied force must be reduced so it is
equal and opposite to the friction force (FD) (Figure 19.2D). If the applied force con-
tinues to decrease and becomes smaller than the friction force (FE), the object starts
to slow down (Figure 19.2E). Both FD and FE are determined by Equation 19.3, and
they are equal to FC since the kinetic friction coefficient is independent of the sliding
speed. Therefore, under these five sliding conditions, the friction forces follow the
following order: FB > FC = FD = FE > FA = 0.
Figure 19.2. Friction forces under different sliding conditions.

19.2 NATURE OF FRICTION

Friction is a complicated phenomenon and is not fully understood yet. One com-
mon belief is the friction force is related to the irregular structure of material
surfaces. On the microscale, almost all surfaces are irregular, and true surface
contacts only occur at the tips of the asperities (Figure 19.3). When a loading
force is applied to slide one surface again another, considerable compressive and
shear stresses are developed locally at the contacting tips. These stresses lead to
the plastic deformation of contacting points. The plastic deformation is the ma-
jor cause of the friction between two surfaces. In addition, the energy associated
with the plastic deformation also facilitates the diffusion of atoms or molecules
between the surfaces. This welding or adhesion process also contributes to the
friction. Therefore, the total frictional force is the sum of two terms: the force of
deformation and the force of adhesion.
The magnitude of deformation depends on the mechanical properties of the
surface materials. Under a fixed normal force, the contacting tips will continue
to deform until the pressure at the points of contact is reduced to the yield stress.
The contacting tips then will be able to support the normal force without further
deformation. The yield pressure (Py) can be expressed:
N
Py = (19.4)
A
where N is the normal force, and A the true contact area. Equation 19.4 can be
rewritten to give the true contact area of the two surfaces:
N
A= (19.5)
Py
Figure 19.3. Schematic of surface contact.

Fiber-on-Fiber Friction 385
Equation 19.5 indicates the true contact area is proportional to the normal

force. When the deformation occurs, the generated energy allows the diffusion
of atoms or molecules between the surfaces, creating welded junctions. To allow
sliding, these joints must be broken by shear stress. To resist the shear stress, the
friction force (F) is generated and can be calculated by:
F = τ* A (19.6)
where τ* is the shear strength of the weaker material of the two surfaces.
Substituting Equation 19.5 into Equation 19.6 gives:
τ*
F= N (19.7)
Py
Since both τ* and Py are constants, Equation 19.7 is the Amontons’ law with:
τ*
µ= (19.8)
Py
Therefore, the coefficient of friction is determined by the yield pressure and shear
strength of the weaker material.
19.3 FIBER-ON-FIBER FRICTION

19.3.1 Friction Coefficient
The friction properties between fibers are important. The friction between fibers is
essential to keep the strength and integrity of yarns and nonwovens. Experimental
research has found marked deformation and welding at points of contact on dam-
aged fiber surfaces. This demonstrates the fiber-on-fiber friction force is the sum
of the force of deformation and the force of adhesion. Table 19.1 shows typical
static and kinetic friction coefficients of polymer fibers. Like most other materials,
the static friction coefficients of fibers are greater than their kinetic counterparts.
The difference between the static and kinetic friction coefficients affects the feel
of the fibers. Fibers with large differences have coarse, crunchy feel. Sometimes,
the large difference between the static and kinetic friction coefficients causes the
“stick-slip” motion. For rayon and nylon fibers, their friction coefficients are in-
dependent of the direction of sliding. However, when wool fibers are involved,
the friction becomes directional due to the presence of scales on the wool fiber
surface.
Table 19.1. Static and kinetic friction coefficients between fibers.
Static Friction Coefficient, μs Kinetic Friction Coefficient, μk
Nylon-on-nylon 0.47 0.40

Rayon-on-rayon 0.35 0.26
Wool-on-wool
with scales 0.13 0.11
against scales 0.61 0.38
fibers in same direction 0.21 0.15
Wool-on-rayon
Wool-on-nylon
Source: Olofsson, B., et. al., Textile Research Journal, 20, 467–476, 1950.
19.3.2 Factors Affecting the Friction Coefficient

Many factors affect the friction coefficient of polymer fibers. Friction coefficients
shown in Table 19.1 are just some typical values. They can be used to provide
some basic information on fiber-on-fiber friction, but they cannot be expected to
have validity in all circumstances.
One important factor that affects the fiber-on-fiber friction is the normal force.
According to the Amontons’ law, the friction coefficient should be a constant and
is independent of the normal force. However, the fiber-on-fiber friction does not
obey the Amontons’ law. Figure 19.4 shows the effect of normal force on the fric-
tion coefficient between fibers. The friction coefficient of fibers decreases with
increase in normal force. This behavior may be explained by the elastic deforma-
tion of surface asperities on fibers. The relationship between fiber-on-fiber friction
force and normal force can be described by using an empirical equation:
F = aN n (19.9)
where a and n are constants. The n value depends on the type of fiber and direction
of sliding, generally ranging from 3/4 to 1.
As discussed in section 19.1.2, the friction coefficient of most materials is in-
dependent of the sliding speed. However, polymer fibers are different. The fric-
tion coefficient of polymer fibers is affected by the sliding speed. In general, at
low speeds, a decrease in friction coefficient can be observed as the speed in-
creases, but at high sliding speeds, the friction coefficient increases with increase
Fiber-on-Other-Material Friction 387
Figure 19.4. Effect of normal force on the friction coefficient between polymer fibers.
in sliding speed. This behavior is related to the viscoelastic properties and physi-
cochemical activities of polymer fibers. It should be recognized the relationship
between friction coefficient and sliding speed is affected by the test temperature.
When the temperature is comparable or greater than the glass transition tempera-
ture, the friction coefficient exhibits a strong dependence on the sliding speed.
However, below the glass transition temperature, the friction coefficient is less
dependent on the sliding speed.
Humidity is another factor that affects the friction coefficient between fibers.
Figure 19.5 shows the effect of relative humidity (% RH) on the friction coef-
ficient between fibers. With increase in relative humidity, the friction coefficient
between fibers increases. One possible reason for increased friction coefficients
at higher relative humidities is that water molecules penetrate the contact points
and increase the contact areas. In addition, while sliding the fibers at high relative
humidities, the shear of the water surface also may contribute to the increased
friction coefficient.
The friction coefficient between fibers also increases with increase in tempera-
ture. As a matter of fact, temperature effect is one of the causes for the increased
friction coefficients at high sliding speeds. At very high speeds, the temperature of
fibers may increase, and this could contribute to the increased friction coefficients.
19.4 FIBER-ON-OTHER-MATERIAL FRICTION

19.4.1 Friction Coefficient
The friction of fibers on other materials also is important. For example, to convert
fibers to final products, they often need pass around cylinder surfaces, such as
Figure 19.5. Effect of relative humidity on the friction coefficient between polymer fibers.
guides, pulleys, etc. When a fiber is pulled over a cylinder surface, a tension must
be developed to overcome the friction force (Figure 19.6). The development of
the tension can be described by the Capstan equation:
T2
= exp ( µθ ) (19.10)
T1
where T1 is the incoming tension, T2 the outgoing tension, μ the friction coefficient
between the fiber and cylinder surface, and θ the angle of contact.
Figure 19.7 shows the theoretical T2/T1 ratios calculated using the Capstan
equation. The tension on the fiber develops quickly while the fiber is being pulled
over a cylinder surface. The development of fiber tension can be accelerated by
increasing the contact angle or the friction coefficient. In practice, when the ten-
sion is too high, the fiber could be broken while passing over the cylinder surface.
In addition, the surface of the fiber could be damaged at high friction, which in
turn makes the breakage of the fiber much easier.
The Capstan equation can be rewritten to give the friction coefficient of fibers
on other materials:
T 
ln  2 
T
µ=  1  (19.11)
θ
Equation 19.11 often is used for measuring the friction coefficient of fibers on oth-
er materials. Table 19.2 shows the experimentally measured friction coefficients
Fiber-on-Other-Material Friction 389
Figure 19.6. Schematic of a fiber passing over a cylinder surface.
Figure 19.7. T2/T1 ratio as a function of θ angle for a fiber being pulled over a cylinder
surface with different friction coefficient μ values.
Table 19.2. Friction coefficients of fibers on steel and porcelain guides.
Steel Guide Porcelain Guide

Nylon 0.32 0.43
Cotton 0.29 0.32
Rayon 0.39 0.43
Acetate 0.38 0.38
Linen 0.27 0.29
Source: Buckle, H., et. al., Journal of the Textile Institute Transactions, 39, T199–T210, 1948.
of fibers on steel and porcelain guides. Both steel and porcelain guides are used
routinely in the industry. For most fibers shown in Table 19.2, their friction coef-
ficients on porcelain guide are greater than those on the steel guide. This suggests
the damage to fiber surface and the wear of the guide will be greater when the
porcelain guide is used to pass fibers. However, the friction coefficient values
vary with many experimental conditions, such as normal force, sliding speed,
humidity, temperature, etc., and hence the data reported in Table 19.2 may not
be valid in many circumstances. In addition, lubricants often are applied to both
steel and porcelain guides, and this changes the friction coefficients significantly,
which will be discussed in section 19.5.
19.4.2 Factors Affecting the Friction Coefficient

Factors that affect the fiber-on-other-material friction coefficient include but are
not limited to normal force, sliding speed, humidity, and temperature. With a few
exceptions, the effects of these factors on fiber-on-other-material friction are the
same as those on fiber-on-fiber friction. Normally, the friction coefficient of fibers
on other materials increases with increase in normal force, sliding speed, humid-
ity, and temperature.
19.5 LUBRICATION
While sliding one dry surface over another, the surface asperities rub against each
other, leading to damage or wear of the surfaces. This can be addressed by intro-
ducing a third material (i.e., lubricant) between the two contacting surfaces. The
lubricant tends to fill the gaps between the asperities and acts as a fluid bearing
layer to allow smoother movement of the two surfaces. A good lubricant for fibers
is a material that has relatively low shear strength and adheres strongly to the fiber
surface. In addition, the lubricant also should be thermally stable since the energy
generated during high speed processing could lead to high temperatures.
Two most important types of lubrication conditions for polymer fibers are:
boundary lubrication and hydrodynamic lubrication. The boundary lubrication
occurs when the amount of lubricant is not sufficient to mask the asperities on
the surface. In this case, the lubricant acts by forming a thin layer on the surface
and preventing the adhesion of the two surfaces at the contact points. This sig-
nificantly reduces the force of adhesion, and the force of deformation becomes
the main source of friction. In boundary lubrication, the heat developed by the
local deformations often causes a condition, called stick-slip. In addition, since
there are fewer contacts between the two surfaces, the force needed to shear the
lubricant layer itself also makes important contribution to the friction. The hy-
drodynamic lubrication occurs when a greater amount of lubricant is present and
Lubrication 391
forms a relatively thick film between the surfaces. In this case, the friction mainly
results from the viscous resistance to the flow of the lubricant.
The coefficient of friction between lubricated surfaces can be described by the
following equation:
µ = f (uη / P ) (19.12)
where u is the relative speed of the two surfaces, η the viscosity of the lubricant,
and P the normal pressure (or load) on the surface. Figure 19.8 shows friction co-
efficient between lubricated surfaces. At low speeds (or high loads), the lubricant
film cannot mask the surface asperities and boundary lubrication dominates. The
resultant friction coefficient decreases with increased speed or decreased load. At
high speeds (or low loads), the lubricant can mask the surface asperities and hy-
drodynamic lubrication dominates. In this case, the friction coefficient increases
with increased speed or decreased load. From Figure 19.8, it also can be con-
cluded that to reduce friction coefficient, the key is to reduce boundary friction at
low speeds and reduce the hydrodynamic friction at high speeds.
The effect of friction coefficient on the damage or wear of lubricated surfaces
is complex. A higher friction coefficient does not always means a higher wear rate
of the surface. For example, the friction coefficient of fibers in the hydrodynamic
lubrication condition can be very high at high speeds. However, the wear rate of
the fiber surface in the boundary lubrication condition still can be much higher
than that in the hydrodynamic lubrication condition even if the friction coefficient
of the boundary lubrication is lower. This is because in a boundary lubrication
condition, the asperities of the two surfaces contact each other and can be broken
off by shear. On the other hand, the two surfaces do not have direct contact in the
hydrodynamic lubrication condition.
Figure 19.8. Friction coefficient between two lubricated surfaces.

REFERENCES
[1] Braun, O.M., and Naumovets, A.G, “Nanotribology: Microscopic
Mechanisms of Friction”, Surface Science Reports, 60, 79–158, 2006.
[2] Brostow, W., Deborde, J.L., Jaklewicz, M., and Olszynski, P., “Tribology
with Emphasis on Polymers: Friction, Scratch Resistance and Wear”,
Journal of Materials Education, 24, 119–132, 2003.
[3] Buckle, H., and Pollitt, J., “An Instrument for Measuring the Coefficient of
Friction of Yarns Against Other Materials”, Journal of the Textile Institute
Transactions, 39, T199-T210, 1948.
[4] Kilic, G.B., and Sular, V., “Frictional properties of Cotton-Tencel Yarns
Spun in Different Spinning Systems”, Textile Research Journal, 82, 755–
765, 2012.
[5] Mogahzy, Y.E.E., and Gupta, B.S., “Friction in Fibrous Materials: Part II:
Experimental Study of the Effects of Structural and Morphological Factors”,
[7] Myshkin, N.K., and Kovalev, A.V., “Adhesion and Friction of Polymers”, in
Polymer Tribology, Editors Sinha, S.K., and Briscoe, B.J., World Scientific,
2009.
[8] Olofsson, B., and Gralen, N., “Measurement of Friction between Single
Fibers V. Frictional Properties of Viscose Rayon Staple Fibers”, Textile
Research Journal, 20, 467–476, 1950.
[9] Olofsson, B., “Measurement of Friction Between Single Fibers VI. A
Theoretical Study of Fiber Friction”, Textile Research Journal, 20, 476–
480, 1950.
[10] Schwarz, U.D., Zworner, O., Koster, P., and Wiesendanger, R., Physical
Review B, 56, 6987–6996, 1997.
[11] Shooter, K.V., and Tabor, D., “The Frictional Properties of Plastics”,
Proceedings of the Physical Society. Section B, 65, 661–671, 1952.
Problems 393
PROBLEMS
(1) Describe the Amontons’ law.
(2) Compare static friction with kinetic friction. Which one is greater?
(3) What is the nature of friction?
(4) Discuss how you would expect the friction coefficient of a polymer fiber to
vary with:
(i) Normal force
(ii) Sliding speed
(iii) Temperature
(iv) Humidity
(5) Compare hydrodynamic lubrication with boundary lubrication.

CHAPTER 20
Optical Properties of Fibers
When light encounters a fiber, it is either transmitted, reflected or absorbed, de-

pending on the structure of the fiber and the wavelength of the light. The transmis-
sion, reflection and absorption of light determine the visual appearance of an indi-
vidual fiber. The appearance of fiber assembles then is the result of the combined
effects of individual fibers, and also is affected by the arrangements of fibers. In
addition, optical properties of fibers provide a convenient measure of many struc-
tural characteristics, especially the molecular orientation. Theoretical treatment of
optical properties of fibers is complex. This chapter discusses the practical aspects
of the optical properties of fibers.
20.1 POLARIZATION AND LIGHT

Polarization refers to the separation of charge, either momentarily or permanently.
There are four mechanisms of polarization (Figure 20.1).
• Electronic polarization arises from the temporary distortion of electron

clouds with respect to the nuclei with which they are associated, upon the
application of an external electric field. Electronic polarization can occur
to all materials and also is called atom or atomic polarization. Electronic
polarization induces dipole moments.
• Ionic polarization occurs in materials that have ionic characteristic. These
ionic materials have internal dipoles, but these built-in dipoles cancel each
other and are unable to rotate by themselves. During ionic polarization, the
external electric field displaces the cations and anions in opposite directions,
leading to net dipole moments.
395
396 Optical Properties of Fibers
Figure 20.1. Schematic of electric, ionic, molecular and space-charge mechanisms.
• Molecular polarization occurs in polar molecules that can rotate freely. In

thermal equilibrium, these natural dipoles are randomly oriented. The exter-
nal electric field aligns the dipoles to some extent and induces a polarization
of the material.
• Space-charge polarization, also known as interface polarization, involves
the redistribution of ionically charged layers. In many cases, surfaces,
grain boundaries, interphase boundaries of materials may be charged. The
charged layers contain dipoles that may become oriented to some degree in
an external electric field and leads to the polarization.
Light basically is an electromagnetic wave, which attempts to polarize the ma-

terial as it passes through. When light travels, it typically propagates as a transverse
wave, i.e., the electric field of the wave is perpendicular to the wave’s direction
of travel. Since the frequency of the waves of visual light is very high (400–790
trillion Hz), only the polarization of the electron clouds around the nuclei of atoms
is fast enough to respond to the electric field of the light. As a result, the optical
properties of materials are determined by the electronic polarization mechanism.
The polarization of larger-scale charges, such as the permanent dipoles, cannot
take place fast enough, and hence ionic, molecular and space-charge polarization
mechanisms are not activated when light passes the materials.
Refractive Index and Birefringence 397
20.2 REFRACTIVE INDEX AND BIREFRINGENCE

20.2.1 Refractive Index
The velocity of light often changes with the medium through which the light is
transmitting. This property can be used to give the most fundamental definition of
refractive index (n), which is the ratio of the velocity of light in a vacuum (vvacuum)
to the velocity of light in the medium (vmedium):
vvacuum
n= (20.1)
vmedium
Since light travels slower in a medium than in a vacuum, the refractive index
typically is greater than 1. One consequence is the direction of light is refracted
or bent while passing from one medium to another (Figure 20.2). As a result, an
alternative definition of refractive index is:
sin θi
n= (20.2)
sin θ r
where θi is the angle of incidence, and θr the angle of refraction.

Refractive index is a basic optical property of fibers that is directly related to
other optical properties. In general, the refractive index of fibers varies with tem-
perature and wavelength. The standard conditions for refractive index measure-
ment involve the use of specific wavelength (589 nm) at a specific temperature
(20°C).
Figure 20.2. Refraction of light in a medium.

The refractive index of fibers also is affected by the fiber density. This is be-
cause electronic polarizability increases when the number of electrons per unit
volume increases. For many fibers, the relationship between refractive index and
fiber density (ρ) can be expressed by Gladstone and Dale’s law:
n −1
= constant (20.3)
ρ
Most fibers are anisotropic and the refractive index value is directional. When the
average refractive index is used, the constant in Equation 20.3 is around 0.3570.
A similar relationship can be obtained between the refractive index (nm) and
volume (vm) of a mixture of different components:
vm ( nm − 1) = v1 ( n1 − 1) + v2 ( n2 − 1) + v3 ( n3 − 1) +… (20.4)
where v1, v2, v3, … and n1, n2, n3, … are the refractive indices and volumes of the
individual components. This relationship can be used to describe the effect of
moisture on the refractive index of fibers. The refractive index of water is 1.333,
and hence the effect of moisture on the refractive index can be described by:
vMR ( nMR − 1) = v0 ( n0 − 1) + 0.333MR (20.5)
where v0 is the volume of 1 gram of dry fiber, vMR the volume of the sample fiber
at a fractional moisture regain of MR, and n0 and nMR the refractive indices of the
same fiber when it is dry and moisturized, respectively. The refractive index val-
ues of most fibers are greater than that of water, and hence they decrease with in-
crease in moisture regain. Figure 20.3 shows schematically the effect of moisture
on the average refractive index of fibers. For some fibers, an increase in refractive
index is observed at low moisture regains and this probably is caused by the filling
up of voids or defects in the fibers by water.
20.2.2 Birefringence
When light is transmitted through a fiber, the change of velocity is associated
with the electronic polarization. In general, the electrons in the inner shells are
not easily displaced, and hence only the outer electrons, which are involved in
the covalent bonds, are affected by light waves. Figure 20.4 shows the electronic
polarization process while the electric field is applied on a covalent bond through
different directions. The electronic polarization is the greatest when the electric
field is applied along the direction of the covalent bond. In isotropic polymer ma-
terials, the covalent bonds are arranged randomly in all directions, and hence the
refractive index is identical in all directions and is the sum of the polarizabilities
Figure 20.3. Effect of moisture regain on the refractive index of fibers.
Figure 20.4. Schematic of electronic polarization of two atoms linked by a covalent bond.

(A) Electron distribution around the atoms without the presence of electric field, (B) polar-
ization while applying an electric field along the covalent bond, and (C) polarization while
applying an electric field perpendicular to the covalent bond.
of all covalent bonds. However, polymer fibers are anisotropic and the polymer
chains are largely aligned along the fiber axial direction. As a result, the refractive
index of oriented polymer fibers is directional.
Most oriented fibers can be characterized by two principal refractive indices,

one parallel to the fiber axis, n� , and one perpendicular to the fiber axis, n⊥ . Here,
the subscripts � and ⊥ indicate the vibration direction of the light waves. The
two principle refractive indices obey the conversation law:
n� + 2n⊥
niso = (20.6)
3
where niso is the refractive index of the unoriented, isotropic sample made from
the same material.
The birefringence (Δn) of a fiber then can be defined as the difference between
the two principal refractive indices:
∆n = n� − n⊥ (20.7)
Table 20.1 shows the two principal refractive indices and birefringence values
of fibers. The values given in Table 20.1 are only typical examples. For example,
both the refractive indices and birefringence of fibers change with increase in
molecular orientation. Table 20.2 shows the effect of drawing on the refractive
indices and birefringence values of polypropylene fibers. With increase in draw
ratio, the birefringence of polypropylene fibers increases due to the increased mo-
lecular orientation. In practice, the measurement of birefringence is an important
method for studying the molecular orientation in polymer fibers.
The birefringence of polymer fibers is associated to the orientation of the crys-
tal axes in the crystalline phase and the orientation of the individual polymer
chains or chain segments in the amorphous phase. As a result, the birefringence of
a polymer fiber can be described as:
∆n = Vc ∆nc + (1 − Vc ) ∆na (20.8)
where Δnc is the birefringence of the crystalline phase, Δna the birefringence of the
amorphous phase, and Vc the volumetric degree of crystallinity. The birefringence
of crystalline and amorphous phases can be written as:
∆nc = ∆nc ,intri f c (20.9)
and
∆na = ∆na ,intri f a (20.10)
where Δnc,intri is the intrinsic, or maximum possible, birefringence of the crystal-

line phase when all polymer chains are perfectly oriented along the filament axis,
Δna,intri the intrinsic, or maximum possible, birefringence of the amorphous phase,
Table 20.1. Refractive indices and birefringence of fibers.
Fiber n� n⊥ Δn

Polyethylene 1.550 1.510 0.040
Nylon 1.582 1.519 0.063
Polyester 1.720 1.540 0.180
Acrylic (Orlon®) 1.500 1.500 0.000
Acrylic (Terylene®) 1.520 1.524 –0.004
Kevlar 2.267 1.606 0.662
Cotton 1.578 1.532 0.046
Ramie 1.596 1.528 0.068
Flax 1.596 1.528 0.068
Wool 1.553 1.542 0.010
Silk 1.591 1.538 0.053
Rayon 1.539 1.519 0.020
Acetate 1.476 1.470 0.006
Triacetate 1.474 1.479 –0.005
Non-Polymer Fibers
Glass 1.547 1.547 0.000
Source: Hamza, A.A., et. al., Journal of Optics A: Pure and Applied Optics, 9, 820–827, 2007.;
Heyn, A.N.J., Textile Research Journal, 22, 513–522, 1952.; Kumar, S., Indian Journal of Fiber
and Textile Research, 16, 52–64, 1991.; Preston, J.M., Modern Textile Microscopy, Emmott,
1933.; Preston, J.M., Transactions of the Faraday Society, 29, 65–71, 1933.
Table 20.2. Refractive indices and birefringence of polypropylene

fibers prepared with different draw ratios.
Draw Ratio n� n⊥ Δn
6:1 1.518 1.494 0.024

7:1 1.521 1.493 0.028
8:1 1.523 1.492 0.031
9:1 1.524 1.491 0.033
10:1 1.526 1.490 0.036
11:1 1.527 1.489 0.038
Source: Hamza, A.A., et. al., Journal of Optics A: Pure and Applied Optics, 9, 820–827, 2007.
fc the orientation factor of the crystalline phase, and fa the orientation factor of
the amorphous phase. Substituting Equations 20.9 and 20.10 into Equation 20.8
gives:
∆n = Vc ∆nc ,intri f c + (1 − Vc ) ∆na ,intri f a (20.11)
In this equation, the Δn, Vc, Δnc,intri, Δna,intri and fc values can be either measured
experimentally or calculated theoretically. Only the fa value is difficult to measure
directly or calculate by other means. Therefore, Equation 20.11 often is used to
calculate the fa value, which is an important property of fibers. High fa is beneficial
for improving some of the mechanical properties, but it could cause large thermal
shrinkage and poor dye penetration.
Both Equations 20.8 and 20.11 work for dry fibers. When fibers absorb a large
amount of moisture, the birefringence may be greater than the value given by
these two equations. This is because these two equations only consider the orien-
tation of polymer chains in the crystalline and amorphous phases. Research has
shown that when non-spherical particles are dispersed with a preferred orienta-
tion in a medium of different refractive index, a birefringence of greater than zero
can be observed even if both the particles and the medium are isotropic. This
is called form birefringence. For dry fibers, the difference between the refrac-
tive indices of the crystalline and amorphous phases is relatively small and the
contribution of form birefringence is negligible. When fibers absorb moisture,
the water molecules mainly enter the amorphous phase and change its refractive
index. Therefore, in wet fibers, the refractive index difference between the crys-
talline and amorphous phases become larger and the form birefringence starts to
make contribution to the total fiber birefringence. The effect of form birefringence
become significant only when the moisture regain is large, e.g., greater than 15%.
20.3 REFLECTION AND LUSTER

When the light falls on a fiber, part of the light is reflected. The amount of light
reflected and how it is reflected are important optical properties of fibers. The
amount of light reflected can be estimated by:
2
 n −1 
R=  (20.12)
 n +1
where R is the reflectivity or the fraction of light reflected, and n the refractive
index. Figure 20.5 shows three major types of surface reflection. Light may be
reflected specularly, leaving the surface with an angle θr = θi (Figure 20.5A), or re-
flected diffusively, with equal intensity in all directions (Figure 20.5B). Most real
surfaces show a combination of specular and diffuse reflections (Figure 20.5C).
Reflection and Luster 403
The luster of a single fiber is determined by total visual appearance of these reflec-
tions from the fiber surface.
The reflection of flight from a fiber surface is affected by how it is incident
upon the fiber. Figure 20.6 shows the surface reflection when the light falls on
a smooth, cylinder-shape fiber along two different directions. When the light is
incident upon the fiber along the longitudinal direction, it can be predominantly
reflected at a constant angle as long as the fiber surface is sufficiently smooth.
However, when the light falls across the fiber on the transverse direction, it is re-
flected at various angles. This is an important feature of the surface reflection on
fibers. However, many fibers do not have smooth surface or circular cross-section.
Rough surface and irregular cross-sectional shapes increase the diffuse reflection.
Fibers with rough or irregular cross-sectional shapes, such as wool and cotton, are
not lustrous even when the light is incident along the fiber longitudinal direction.
Lustrous fibers are those with smooth surface and regular cross-sectional shape,
such as silk and some synthetic polymer fibers.
In addition to the primary reflection from the outer surface of the fiber, light
can reflected from the internal surfaces. Figure 20.7 shows the primary and sec-
ondary reflections from a cylinder-shape fiber. When light falls on the surface of
the fiber, some of it is transmitted through the fiber. A portion of the transmitted
Figure 20.5. Types of surface reflection. (A) Specular reflection, (B) diffuse reflection, and
(C) combination of specular and diffuse reflections.
Figure 20.6. Reflection of light from the surface of a smooth, cylinder-shape fiber. Light
is incident upon the fiber along: (A) the longitudinal direction, and (B) the transverse
direction.
Figure 20.7. Primary and secondary reflections.
light can be reflected from the internal surfaces and join the light reflected from
the outer surface. This is called secondary reflection. The secondary reflection is
enhanced in fibers that have hollow tunnels (e.g., lumen), cavities or particulate
fillers (e.g., TiO2). Figure 20.8 shows the secondary reflection on these fibers.
Hollow tunnels, cavities and particulate fillers scatter the transmitted light and
cause apparent diffuse reflection. This effect masks the specular reflection and can
be used to produce delustered fibers.
20.4 ABSORPTION AND DICHROISM

When light falls on a fiber, it is transmitted, reflected, or absorbed. Sections 20.2
and 3 discussed the transmission and reflection of light, respectively. This section
addresses the absorption of light by fibers.
The absorption of light is the result of the interaction between the electromag-
netic waves and the electron clouds in the fiber. The color of a fiber is determined
by the selective absorption of light wavelengths (or frequencies). For example,
a yellow, opaque fiber has a yellow color because it absorbs all wavelengths in
the visual spectrum except yellow, and the yellow light is reflected. A yellow,
transparent (or translucent) fiber is yellow when viewed in transmission because
it allows yellow light to be transmitted while absorbing other wavelengths in the
spectrum.
The absorption of light in a fiber obeys the Lambert’s law:
I = I 0 exp ( −kx ) (20.13)
where I is the intensity of light after passing for a distance x through a fiber, I0 the
intensity of the incident light, and k the absorption coefficient.
Polymer fibers are semicrystalline, and they typically are translucent. Most
of them are either colorless or slightly colored in neutral shades. In order to pro-
duce color, fibers often are dyed. The absorption coefficient k of a dyed fiber is
References 405
Figure 20.8. Secondary reflection on fibers with (A) hollow tunnel, (B) cavities, or (C) par-
ticulate fillers.
proportional to the density or concentration of the dye molecules added to the

fiber. The Lambert’s law then can be rewritten to give the Beer-Lambert equation:
I = I 0 exp ( −ck ' x ) (20.14)
where c is the concentration of dye molecules, and k’ the extinction coefficient or

absorption per unit of dye concentration.
For a dyed fiber, the absorption of light by dye molecules may vary with the
direction of polarization of the light. This effect is called dichroism and it often
leads to differences in the depth of shade or even in actual color. For dichroism to
occur in a dyed fiber, the dye molecules must have an asymmetrical structure so
that the absorption of light can vary with the direction of polarization. In addition,
the dye molecules must be absorbed onto polymer chains with a particular angle
and the polymer chains need be oriented. When polarized light is used to examine
a dyed fiber that exhibits dichroism, the following intensity ratio often is assessed:

(
log I � / I 0 )=k �
= θ ≡ dichroic ratio (20.15)
log ( I ⊥ / I 0 ) k⊥
where I � and I ⊥ are the intensities of the light polarized along the longitudinal
and transverse directions of the fiber, respectively, and k� and k⊥ are the corre-
sponding absorption coefficients. The dichroic ratio, θ, is independent of the con-
centration of dye molecules and can be used to assess the orientation in fibers. In
general, the dichroic ratio increases with increasing orientation in fibers.
REFERENCES
[1] Duckett, K.E., and Cheng, C.C., “The Detection of Cotton Fiber
Convolutions by the Reflection of Light”, Textile Research Journal, 42:
263–268, 1972.
[2] Duckett, K.E., and Tripp, V.W., “X-Ray and Optical Orientation
Measurements on Single Cotton Fibers”, Textile Research Journal, 37,
517–524, 1967.
[3] El-Dessouky, H.M., Mahmoudi, M.R., Lawrence, C.A., Yassien, K.M.,
Sokkar, T.Z.N., and Hamza, A.A., “On the Physical Behavior of Isotactic
Polypropylene Fibers Extruded at Different Draw-Down Ratios. I. Optical
Properties and Cold-Drawing”, Polymer Engineering and Science, 49,
2216–2224, 2009.
[4] Fouda, I.M., El-Farahaty, K.A., and El-Tonsy, M.M., “Optical Anisotropy
in Mohair Fibers”, Textile Research Journal, 59: 506–512, 1989.
[5] Fouda, I.M., and Seisa, E.A., “Birefringence and Orientation Parameters
of Cold-Drawn Viscose Fibers”, Journal of Applied Polymer Science, 106,
1768–1776, 2007.
[6] Hamza, A.A., “A Contribution to the Study of Optical Properties of Fibers
with Irregular Transverse Sections”, Textile Research Journal, 50, 731–734,
1980.
[7] Hamza, A.A., and Abd El-Kader, H.I., “Optical Properties and Birefringence
Phenomena in Fibers”, Textile Research Journal, 53, 205–209, 1983.
[8] Hamza, A.A., Belal, A.E., Sokkar, T.Z.N., El-Bakary, M.A., and Yassien,
K.M., “Interferometric Detection of Structure Deformation Due to Cold
Drawing of Polypropylene Fibres at High Draw Ratios”, Journal of Optics
A: Pure and Applied Optics, 9, 820–827, 2007.
[9] Hamza, A.A., and El-Farahaty, K.A., “Optical Properties and Geometrical
Parameters of Poly(p-phenylene Terephthalamide) Fibers”, Textile Research
Journal, 56, 580–584, 1986.
[10] Hamza, A.A., Fouda, I.M., El-Farhaty, K.A., and Badawy, Y.K., “Production
of Polyethylene Fibers and their Optical Properties and Radial Differences
in Orientation”, Textile Research Journal, 50: 592–596, 1980.
[11] Hamza, A.A., and Sokkar, T.Z.N., “Optical Properties of Egyptian Cotton
Fibers”, Textile Research Journal, 51: 485–488, 1981.
[12] Heyn, A.N.J., “Observations of the Birefringence and Refractive Index of
Synthetic Fibers with Special Reference to Their Identification”, Textile
Research Journal, 22, 513–522, 1952.
Problems 407
[14] Lynch, L.J., “A Study of the Relationship Between Geometry and Light
Scattering Profiles of Single Wool Fibers”, Textile Research Journal, 44:
446–448, 1974.
[15] Lynch, L.J., and Thomas, N., “Optical Diffraction Profiles of Single Fibers”,
[17] Preston, J.M., Modern Textile Microscopy, Emmott, 1933.
[18] Preston, J.M., “Relations between the Refractive Indices and the Behaviour
of Cellulose Fibres”, Transactions of the Faraday Society, 29, 65–71, 1933.
[19] Rudolf, A., and Smole, M.S., “Structure–Properties Relations of the Drawn
Poly(ethylene terephthalate) Filament Sewing Thread”, Journal of Applied
Polymer Science, 110, 2641–2648, 2008.
[21] Wei, W., Qiu, L., Wang, X.L., Chen, H.P., Lai, Y.C., Tsai, F.C., Zhu, P.
and Yeh, J.T., “Drawing and Tensile Properties of Polyamide 6/Calcium
Chloride Composite Fibers”, Journal of Polymer Research, 18, 1841–1850,
2011.
[22] Yeh, J.T., Lin, S.C., Tu, C.W., Hsie, K.H., and Chang, F.C., “Investigation
of the Drawing Mechanism of UHMWPE Fibers”, Journal of materials
Science, 43, 4892–4900, 2008.
PROBLEMS
(1) What type of polarization mechanism is involved when the light encounters
a fiber? Why?
(2) What is refractive index? How does the fiber density affect the refractive
index of a fiber?
(3) What is birefringence? Why is birefringence so important in fiber science?
(4) Compare specular reflection with diffuse reflection.
(5) What are the possible approaches to produce delustered fibers?
(6) What is dichroism?
Index
Absorption, 404 Ionic bond, 29

Acetate, 57 Preliminary bonding, 25
Acrylic, 284 Secondary bonding, 28
Activation energy, 134, 373 Van der Waals force, 31
Air frictional force, 234, 239 Boundary friction, 390
Alkalization, 191 Branched polymer, 20
Ammonia, 194 Buckling, 292
Ammonium thiocyanate, 194 Bulk modulus, 270
Amontons’ law, 381 Bushing, 211, 213
Amorphous phase, 36
Antistatic treatment, 376 C curve, 165
Applied force, 115, 266 Calcium thiocyanate, 194
Arrhenius equation, 134, 310, 326 Capillary flow, 127
Aspect ratio, 251 Capstan equation, 388
Atactic polymer, 18 Carbamate process, 194
Atomic order, 83, 87 Carbon disulfide, 192
Average molecular weight, 12 Carbonization, 208
Avrami equation, 166 Carreau equation, 125
Cellulose, 53
Beer-Lambert equation, 405 Cellulose fiber, 53, 56, 65, 71, 187, 191
Bembergy process, 193 Cellulose process, 209
Bending property, 301 Cellulose pulp, 191
Bicomponent fiber spinning, 240 Centrifugal spinning, 305
Bingham fluid, 115 Charge carrier, 368
Biotech cotton, 190 Chemical structure, 9, 53
Birefringence, 163, 398 Chiral center, 17
Boiling water shrinkage, 346 Cholesteric order, 175
Bombyx mori, 197 Coagulation bath, 57, 193, 199
Bonding Coefficient of thermal expansion, 344
Aromatic ring association, 30 Cohesiveness, 254
Covalent bond, 25 Color, 260, 404
Dipole-dipole force, 30 Complex quantity, 322
Hydrogen bond, 29 Composition, 88
409
410 INDEX
Compressive property, 292 Direct method, 57, 194

Conduction, 159 Disentanglement, 121
Conductivity, 222 Displacement, 266
Conductor, 367 Doolittle equation, 351
Configuration Drag force, 128, 153
Head-to-head, 16 Draw ratio, 174, 178
Head-to-tail, 16 Drawing, 178, 348
Skeletal structure, 20 Drying tower, 172
Tacticity, 17 Dynamic mechanical property, 318
Conformation, 22
Continuous cooling transformation Elastic effect, 142
diagram, 166 Elastic recovery, 256, 289
Continuous filament process, 211 Electrical conductivity, 367
Convection, 159 Electrical property, 259, 367
Copolymer, 21 Electrically-conductive polymer fibers, 371
Core-sheath structure, 50 Electronic polarization, 395
Cortex, 74, 196 Electrospinning, 95, 217
Cotton, 55, 65, 187 Elongation, 256
Cotton boll, 188 Elongational viscosity, 144
Crankshaft motion, 356 Energy balance, 158, 172
Creep, 312 Entanglement, 123, 139
Cross-sectional effect, 169 Enthalpy, 359
Crystallization, 164 Entropy, 340, 359
Crystalline model Environmental condition, 187
Folded-chain model, 43 Environmental property, 260
Fringed fibril model, 44 Epidermal cell, 189
Fringed micelle model, 43 Epidermis, 195
Modified fringed micelle model, 44
Switchboard model, 43 Fiber formation, 151, 217, 231
Crystalline phase, 36 Fiber property, 251
Cuprammonium process, 173 Fiberglass, 210
Cupro fiber, 57 Fibroin, 77
Cuticle, 67, 74, 196 First-order transition, 348
Flash vaporization, 172
Decomposition, 361 Flax fiber, 68
Degradation, 361 Flexibility, 254
Degree of polymerization, 12 Flow behavior, 117
Denier, 252 Flow curve, 125
Density, 254 Flow rate, 227
Derivative method, 57, 191 Follicle, 195
Dermal papilla, 195 Force balance, 153, 172
Dermis, 195 Form birefringence, 404
Desiccation, 190 Four-element model, 332
Dichloroethane, 194 Free volume theory, 349
Dichroism, 404 Friction coefficient, 382, 385, 387
Die swelling, 142 Frictional property, 262, 381
Dimensional stability, 346 Funiculus, 189
Dimethylacetamide, 194
INDEX 411
Fuzz fiber, 188 Lint fiber, 188

Lithium chloride, 194
Genetic code, 187 Loss modulus, 321
Gibbs free energy, 358 Lower Newtonian region, 125
Ginning, 191 Lubrication, 390
Glass transition, 347 Lumen, 67
Glass wool, 210 Luster, 402
Glucose ring, 65 Lyocell, 57
Gossypieae, 187 Lyotropic liquid crystal, 176
Gossypium, 187
Graphitization, 208 Malvaceae, 187
Gravitational force, 153 Malvales, 187
Manufactured cellulose fiber, 59, 71, 191
Heat capacity, 339 Manufactured protein fiber, 62, 78, 198
Heat flux, 38 Marble process, 212
Heat setting, 182, 346 Mass transfer, 172, 173
Heat transfer, 171 Maxwell model, 328
Heat treatment, 181 Mechanical property, 256, 265
Hemp fiber, 69 Mechanical work, 310
Homopolymer, 21 Medulla, 74
Hooke’s law, 265 Melt blowing, 239
Hydrocarbon gas process, 210 Melt fracture, 144
Hydrodynamic lubrication, 390 Melting, 357
Hydrostatic force, 128 Mesophase pitch process, 208
Micropyle, 189
Inertial force, 153 Mixed solvent, 225
Inertial mass, 300 Modulus, 256, 267
Inner root sheath, 196 Moisture regain, 262, 340
Insulator, 367 Molecular mechanism, 309
Interface polarization, 396 Molecular orientation, 162
Internal energy, 310 Molecular polarization, 396
Ionic polarization, 395 Molecular weight, 12
Islands-in-the-sea, 50, 240 Moment of inertia, 300
Isokinetic approximation, 166 Monomer, 10
Isotactic polymer, 18 Monomeric unit, 10
Isothermal crystallization rate, 167 Morphology
Cross-Sectional morphology, 48
Kapok fiber, 68 Crystalline morphology, 46
Kelvin-Voigt model, 330 Fibrillar morphology, 46
Kenaf, 70 Longitudinal morphology, 50
Kevlar, 11 Multifilament effect, 171
Kinetic friction, 382 Multi-relaxation time model, 333
Lambert’s law, 404 Nanofiber

Length-to-diameter ratio, 252 Carbon nanofiber, 98
Light, 395 Ceramic nanofiber, 98
Linear polymer, 20 Composite nanofiber, 99
412 INDEX
Core-Sheath nanofiber, 103 Polyamide, 10

Hollow nanofiber, 103 Polybenzimidazole, 12
Metal nanofiber, 99 Polybutylene terephthalate, 11
Multichannel nanofiber, 104 Polycarbonate, 12
Polymer nanofiber, 96 Polyester, 11
Porous nanofiber, 101 Polyethylene, 11, 40
Nanofiber assembly Polyethylene naphthalate, 11
Aligned nanofiber sheet, 107 Polyethylene terephthalate, 11, 40
Nanofiber array, 108 Polyolefin, 11
Nanofiber filament, 107 Polyphenylene sulfide, 12
Nanofiber nonwoven, 106 Polypropylene, 11, 40
Nanofiber yarn, 109 Polystyrene, 11
Nanoparticle filler, 223 Polytetrafluoroethylene, 11
Natural cellulose fiber, 53, 65, 187 Polytrimethylene terephthalate, 11
Natural protein fiber, 59, 74, 195 Polyvinyl alcohol, 11
Nematic order, 175 Polyvinyl chloride, 11
Network structure, 20 Polyvinylidene chloride, 11
Newtonian fluid, 119 Polyvinylidene dinitrile, 11
Non-isothermal crystallization rate, 166 Polyvinylidene fluoride, 11
Non-Newtonian fluid, 119 Potential energy, 275
Non-polymer fiber Power supply, 217
Carbon fiber, 83, 205 Pre-aging, 192
Glass fiber, 86, 210 Precursor, 205
Normax, 11 Primary property, 249
Nozzle-collector distance, 226 Primary reflection, 403
Nozzle diameter, 226 Principal refractive index, 400
N-methylmorpholine-N-oxide, 194 Protein, 59
Nylon, 10, 40
Rabinowitsch equation, 131
Operational condition, 226, 236 Rayon, 57
Optical property, 260, 395 Radiation, 159
Orientation factor, 162 Reflection, 402
Out-of-phase, 319 Reflectivity, 402
Outer root sheath, 196 Refractive index, 397
Ovule, 188 Relative humidity, 284
Repeating unit, 9
PAN process, 205 Resilience, 256
Pascal, 252 Restoring force, 265
Phase separation, 241 Reynolds number, 154
Physical shape, 255 Rheological behavior, 117, 231
Physical structure, 35, 65 Rheological force, 153
Poise, 119 Rheopectic fluid, 123
Poisson’s ratio, 269 Ripening, 193
Polarization, 395
Polydispersity index, 15 Salt additive, 224
Polymer Second-order transition, 348
Polyacrylonitrile, 11 Secondary property, 255
INDEX 413
Secondary reflection, 403 Tan δ, 322

Segmented pie, 50, 240 Taylor cone 217, 219
Self-assembly, 243 Temperature gradient, 342
Semiconductor, 367 Template synthesis, 244
Sericin, 77 Tensile modulus, 267
Shear modulus, 272 Tensile property, 273
Shear rate, 118, 129, 272 Tensile strain, 267
Shear stress, 117, 127, 272 Tensile stress, 267
Shear-thickening fluid, 120 Tex, 252
Shear-thinning fluid, 120 Thermal conductivity, 342
Side-by-side structure, 50 Thermal contraction, 344
Silk, 60, 76, 197 Thermal expansion, 344
Sisal fiber, 69 Thermal property, 259, 339
Smectic order, 175 Thermally-stable polymer fiber, 363
Sodium cellulose xanthate, 192 Thermotropic liquid crystal, 176
Sodium hydroxide, 194 Thixotropic fluid, 123
Solid phase, 37 Tie molecule, 179, 181
Solution Property, 221, 235 Time-dependent fluid, 123
Space-charge polarization, 396 Time-temperature equivalence, 323
Spandex fiber, 177 Time-temperature superposition, 325
Specific heat, 339 Torque, 305
Specific stress, 269 Torsion, 296
Spherulitic morphology, 46 Torsion pendulum, 299
Spinning Torsional property, 296
Dispersion spinning, 177 Transgenic cotton, 188
Dry spinning, 171 Triacetate, 52
Gel spinning, 173 Triboelectric series, 373
Liquid crystal spinning, 175 Trifluoroacetic acid, 194
Melt spinning, 151 Trouton ratio, 145
Reaction spinning, 178 True strain, 268
Solution spinning, 171 True stress, 268
Wet spinning, 173
Spinning bath, 172, 173, 193, 206 Uniformity, 255
Spinning gland, 197 Unit cell, 37
Stabilization, 206 Upper Newtonian region, 125
State charge generation, 373
State electricity, 372 Velocity gradient, 118
Static friction, 382 Velocity profile, 129
Storage modulus, 321 Viscoelastic property, 309
Strain-induced crystallization, 181 Viscose process, 191
Strength, 252, 273 Viscosity, 118, 133, 134, 144
Stress relaxation, 315 Voltage, 226
Stress-induced crystallization rate, 167
Stress-strain curve, 273 WLF equation, 327, 352
Surface tension, 153, 223 Wool, 60, 74, 195
Syndiotactic polymer, 18 Wool process, 210
Synthetic fibers, 9, 35, 151 Work recovery, 291
Syringe, 217
414 INDEX
Xanthation, 192 Zero-shear viscosity, 125

Zinc chloride, 194
Yield, 276
Yield pressure, 384
About the Author
Professor Xiangwu Zhang received his B.S. in Polymer

Materials and Engineering in 1997 and Ph.D. in Materials
Science and Engineering in 2001, both from Zhejiang
University, China. Following his graduation from Zhejiang
University, Zhang joined the Center for Electrochemical
Systems and Hydrogen Research at Texas A&M University
as a Postdoctoral Research Associate in 2001. During
2002–2006, he was a Postdoctoral Research Associate in
the Department of Chemical and Biomolecular Engineering
at North Carolina State University. He joined the faculty in the Department of
Textile Engineering, Chemistry, and Science at North Carolina State University
in 2006. Zhang’s research interests focus on nanostructured and multifunctional
polymer, composite, fiber, and textile materials with an emphasis on practical ap-
plications: (i) energy storage and conversion, (ii) chemical and biological protec-
tion, and (iii) composites. His research encompasses both fundamental materials
studies such as synthesis and physical characterization, as well as system design
and fabrication. Widely published, he is frequently an invited speaker at major
conferences in his field of expertise. Zhang is an inaugural University Faculty
Scholar and has received the Alumni Association Outstanding Research Award
and Outstanding Teacher Award from North Carolina State University.
Zhang currently serves as the President and Executive Committee Chair of
the NCSU Chapter of Sigma Xi, the Scientific Research Society. Zhang is an
active member of the Minerals, Metals & Materials Society (TMS), American
Association of Textile Chemists and Colorists (AATCC), the Fiber Society, the
American Society of Mechanical Engineers (ASME), the American Chemical
Society (ACS), Materials Research Society (MRS), and the Electrochemical
Society (ECS). Zhang is also an Executive Editor, Associate Editor or Editorial
Board Member of Advances in Nanoparticles, Journal of Energy Storage and
Conversion, Journal of Membrane Science & Technology, Journal of Nanoparticles,
Journal of Textile Science and Engineering, Textiles and Light Industrial Science
and Technology, Global Journal of Physical Chemistry, and International Journal
of Nano Science, and Nano Engineering and Nanotechnology.
415

Zhang, Xiangwu - Fundamentals of Fiber Science-DeStech Publications (2014)

Uploaded by

Copyright:

Available Formats

Zhang, Xiangwu - Fundamentals of Fiber Science-DeStech Publications (2014)

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Zhang, Xiangwu - Fundamentals of Fiber Science-DeStech Publications (2014)

Uploaded by

Copyright:

Available Formats

of

Xiangwu Zhang, Ph.D.

DEStech Publications, Inc.

Copyrig ht © 2014 by DEStech Publications, Inc.

No part of this pub licati on may be reprodu ced, sto red in a

Prin ted in the United States of Ameri ca

Library of Congress Contro l Nu mber: 20139 56174

HOW TO ORDER THIS BOOK

DEStech Publications, Inc.

Part I FIBER STRUCTURE

Chapter 2. Chemical Structure of Synthetic Polymer Fibers.................. 9

Chapter 3. Physical Structure of Synthetic Polymer Fibers.................. 35

Chapter 4. Chemical Structure of Natural Polymer Fibers................... 53

Chapter 5. Physical Structure of Natural Polymer Fibers..................... 65

5.4 Manufactured Protein Fibers.......................................... 78

Chapter 6. Structure of Non-Polymer Fibers........................................ 83

Chapter 7. Structure of Nanofibers....................................................... 95

Part II FIBER FORMATION

Chapter 8. Flow Behavior of Polymers...............................................117

Chapter 9. Formation of Synthetic Polymer Fibers............................ 151

9.2 Solution Spinning......................................................... 171

Chapter 10. Formation of Natural Polymer Fibers............................... 187

Chapter 11. Formation of Non-Polymer Fibers.................................... 205

11.2.3 Marble Process............................................. 212

Chapter 12. Formation of Nanofibers by Electrospinning.................... 217

Chapter 13. Formation of Nanofibers by Other Methods..................... 233

Part III FIBER PROPERTIES

Chapter 14. Primary and Secondary Properties.................................... 251

Chapter 15. Mechanical Properties of Fibers........................................ 265

15.3 Compressive Properties................................................ 292

Chapter 16. Viscoelastic Properties of Fibers....................................... 309

Chapter 17. Thermal Properties of Fibers............................................. 339

17.4 Glass Transition............................................................ 347

Chapter 18. Electrical Properties of Fibers........................................... 367

Chapter 19. Frictional Properties of Fibers........................................... 381

Chapter 20. Optical Properties of Fibers.............................................. 395

About the Author 415

materials, such as petroleum-based chemicals or petrochemicals. In general, syn-

does not depend on their shape or morphology, either. In addition to chemical

Non-polymer fibers can be produced by many different methods, depending

Chemical Structure of Synthetic Polymer Fibers

The fundamental knowledge on the chemical structure of synthetic polymer fibers

2.1 REPEATING UNITS

2.1.2 Repeating Units of Synthetic Fibers

Table 2.2. Repeating units of some important polymers.

Polyvinyl chloride (PVC)

Polyvinylidene chloride (PVDC)

Polyvinylidene fluoride (PVDF)

Vinyl Polytetrafluoroethylene (PTFE)

Polyethylene terephthalate (PET)

Polytrimethylene terephthalate (PTT)

Polyesters Polybutylene terephthalate (PBT)

Polyethylene naphthalate (PEN)

Type Polymer Name Repeating Unit

Polyphenylene sulfide (PPS)

Polyparaphenylene benzobisthiazole (PBT)

2.2 MOLECULAR WEIGHT

where M0 is the molecular weight of the repeating unit.

2.2.1 Average Molecular Weights

Figure 2.2. A distribution of molecular weight along with various average molecular

&XELF 6LPSOH %RG\&HQWHUHG )DFH&HQWHUHG