Indian Journal of Fibre & Textile Research
Vol. 16, March 1991,pp. 1-6
Some unsolved problems In the science of nylon and polyester fibres
J W S Hearle
University of Manchester Insti tute of Science and Technology, 'P 0 Box 88, Manchester. U.K .
Received 5 October 1990
Despite the commercial importance of melt-spun fibres there are many aspects of their science that are not
well understood. A common working model of fine structure is a pseudo-fibrillar assembly of crystallites
linked by tie molecules in amo rphous regions, but other possibilities are suggested . Heat-setting is an
accepted industrial practice, but the science is poorly understood: there is a lack of good experimental data and
no adequate theory. Even the nature of the basic mechanisms is not certain. There has been little attempt to
analyze the mechanics of the material. Composite models a re inadequate, but network models show more
promise. There are a number of differences between nylon and polyester fibres, which are unexplained. It will
be necessary to solve these problems to maximize the industrial potential of nylon and polyester fibres .
Keywords: Melt-spun fibre, Nylon fibre, Po lyester fibre
1 Introduction
Melt-spun fibres are one of the scientific marvels of
20th century technology. In many other areas of
man 's material needs, inventions had been made in
earlier centuries, but, in the everyday world of textiles,
it was nylon and polyester, with rayon as a forerunner,
which first denied a substantial share of the market to
natural fibres. But the story is not a pure triumph ,
either in science or in business. In particular, the
decades of polyester history provide a fascinating
commentary on commercial developments. In the
context of rich industrialized countries, the stages
are:
1942 Discovery of poly(ethylene terephthalate)
1952 New high-price 'miracle' fibre
1962 Great commercial profit
1972 Mature growth, replacing cotton as the
general purpose 'textile fibre
1982 Industrial depression : over-capacity,
losses, plant closures, poor image
To some extent, the problems were due to the move
of polyester into the cheapest mass market for textiles.
Nylon, with a longer history and more specialized
markets, and polypropylene, for different reasons,
were not affected to the same extent. But the
competitive position of all the man-made fibres was
weakened by their derogatory designation as "plastic
fibres" and by the promotional efforts of the natural
fibre producers.
What are the prospects for 1992? Part of the need
will certainly be for basic fibres at the lowest price. But,
within the great diversity of textile markets, another
part will be in premium fibres with special properties,
either giving high fashion appeal, or comfort, or
engineering performance. Both the opportunities will
require advances in technology, the first in speed and
production engineering and the second in controlling
fibre structure. In both, the technology needs to be
based on scientific understanding.
On the whole, the chemistry of the fibres is well
known, and also, 'in a more empirical way, are
macroscopic features , such as fibre size, shape and
multicomponent features. The big gap in our
knowledge is in the physical fine structure, how it is
controlled in production, and how it influences the
fibre properties that determine the ease of processing
during textile manufacture and the attraction to the
consumer.
2 Models of Fine Structure
Fig.l (a), the picture selected for the cover of a
recent book 1 on polymer science, illustrates well the
commonly accepted view of the fine structure of
oriented, semi-crystalline polymers that have not
been exposed to the slow cooling and sparse
nucleation of bulk samples leading to spherulitic
texture. Fibres and films are the important examples
of such forms. Similar pictures have been drawn by
several scientists 2 - 4, often with some quirk intended
to illustrate a particular argument. For example,
Fig. 1(b) shows a drawing of nylon structure, in which
I angled the ends of the crystalli te in order to show how
1
INDI AN J. F IBR E TEXT. RES .. MARCH 1991
48.5°
( a)
( b)
Fig. I - (a) A view of polymer structure, as shown on the cover of the book M olec ular Conformation
and Dynamics of Macromolecules in Co ndensed Systems". No te that the crysta lline mate ri a l occ upies
less than 40 ':1., of the two-dimen sional view: th is wo uld become a bo ut 25 % in three dimensions.
(b) Hearle's drawing of a nnea led nylo n structure 2
(e) Hea rl e's drawin g o f an alternative view of dra wn nylon struc tures S
the form of the crysta l lattice would explain the
inclined pl a ne se parations indicated . by the
small-angle X-ray diffraction diagram 2 . The
common featur~ s o f these pictures are that they
consist of a 'two-phase' pse udo-fibrillar assembly of
crystallites, with polymer molecules partly folding at
the ends of the crystallites and partly linking
crystallites
together
through
tie molecules.
Prevorsek el at. 3 have suggested differences between
nylon and polyester in th e stacking of the crystallites
on a superlattice, but thi s is a spec'u lative view.
Fig.1 (c) shows a different idea of structure s. I drew
this to illustrate a more continuous single-phase form
of structure, suggested on thermodynamic arguments
for the unannealed ' Fo rm r of Bell el al. 6 , and later
called a ' dynamic crystalline ge(' 7. Further
confirmation of the current uncertainty is provided by
Fig.2, which consists of two views of the structure of
polyester fibres spun a t 5000 mjmin drawn by
different authors in the sa me book S.
2
a
Fig. 2- Two views of the structure of PET filaments spun at 5000
m/ min. bo th from th e sa me bookS : (a) drawn by Heu vel and
Hui sma n: a nd (b) drawn hy Shimizu ('/ a/.
HEARLE: UNSOLVED PROBLE MS IN SCIENCE OF NYLON AND POLYESTER FIBRES
But we do not need to look for di sagreemen t
between scienti sts to show up the in a dequacy of o ur
understanding. It is there in the pictures. I wrote many
years ago: " Theories of fine structure are not
mathematica l theories: they cannot be precisely
formulated . All that can be done is to give an
inadequate
description ,
in
words,
in
a
two-dimensional diagram , or sometimes in a model,
of a hypothetical three-dimensiona l network of chain
molecules, which is imperfectly visualized in the
author's mind . The form that the rea der visualizes
may be different again and, as time goes on a nd other
authors give their own pictures, a gradual change in
the idea of the structure can occur" 9 . Often the
differences betwen pictures rela te more to the a rti stic
skill and cha racter of the author than to any difference
in scientific intent. There is another difficulty in regard
to nylon and polyester. Simple lines may be adequate
to represent pol ymers with a simple repeat such as
- CH 2 - in polyethylene, but they do not properl y
represent the long repea ts of nylon and pol yester
molecules, with the alternation of different chemical
forms. Thi s fuzzy uncerta inty is very different to the
precision of the c hemica l specifica ti o ns; yet the
physica l fine struct ure has a maj o r influence on
technologica l variety and perfo rmance.
Whil e the sta temen t quoted a bove is still la rgely
true as hi sto rica l fact, I would now say that we should
a im to be more precise, a lth o ugh , until some recent
and only part ly published work 10 of o urs on a
simplifi ed model , there has been no attempt to
determin e a li st of independent parameters to defi ne
the geo met ry qua ntitatively.
There is a more se ri o us problem in regard to
models, whic h may be inte nsified by atte mpts a t
grea ter prec isio n. The model s may co nt a in feature s
which we take for gran ted but which are incorrect a nd
even unrecognized. This leads to th e so rt of
mind-block that orten impedes scient ific progress.
The crystallites a re shown as perfect with a ll the
pol ymer chains runnin g to the ends, where they frin ge
or fold . But is it poss ible that some mo lec ules foll ow
defect pat hs to the sides of the crysta llites. Is the
mi sorientatio n of crys ta llites, a bse nt in Fig. l , loca l
or, as suggested by some recent elect ro n microscopic
studies, in larger zones? The model s help o ur a bility to
discuss st ruc ture, but they may give a greater
impression of knowledge th a n is justified. All we
really know with any confidence is the geometry of the
crystal latti ce, a nd th e fact that the structure li es
somewhere a mong the ma nifold possibilities between
a perfect crys tal a nd a random , but fairly closely
packed, assembl y of molec ules. so me info rma ti o n
about average orientations, and a rough estimate of
the scale of th e fine structure.
These comments show o ne rea so n for our
ignora nce . The structures a re inherently difficult to
describe, and they vary according to the hi story of the
fibre. We need to devote more effort and imagina tion
to ways of describing structures. A complete
specification is impossible. Therefore, we need to
sea rch for para meters to represent fea tures of the
structure that determine impo rt an t properties.
Another reason for the ignora nce is that a lthough the
techniques for structural in vestiga tion have
increased in instrumental sophistication, they all
remain subj ect to a rtifacts and experimental errors,
and even more to uncertainties of interpretation ,
biassed by being related to pa rticula r model s. If the
wrong model is used , ignora nce can be reinforcd by a
seeming exactness in calculation of parameters. Thi s
is especially true when the d a ta is reported not as
direc tly o bt a ined , but as some number calculated
according to a formula .
3 Thermo-mechanical Responses and Heat-setting
It is ge nera ll y recogn ized tha t pol yamides like
nylo n 6 and 66 a nd polyesters like PET have two major
tra nsitions below the melting region, unlike simpler,
more studi ed polymers such as pol yethylene and
polystyrene . In effect, th e glass-to-rubbe r transition
is sp lit int o two parts . In nylo n, the interpretation
seems clear. Th e tra nsition bel ow room temperature
is assoc ia ted with freedom of rotation around
- C H 2 - bonds, and the one above room temperature.
usua ll y called the glass tra nsiti on, is due to the onset of
mo bility in the hydroge n bo nds between - CO.NH gro ups. Definite identification o f the a na logo us
effec ts in PET remains un so lved. T em po rary hea t (or
m ois ture) sett in g must occ ur o n passin g through th e
glass tra nsiti o n.
Wh a t ha ppens near the melting point is le ss clear.
Desp ite th e industrial impo rtance of ' permanent'
a
hea t-settin g at th ese tempera tures (ca. ISOaC to 230 C
for nylon 66 and PET, lower for nylo n 6) th ere is a
scandalous lack o f good ex perim ent a l d ata from
scientific laboratory expe rimen ts. Th is particularl y
applies to multiple settin g trea tments, to the influence
of time a nd levels of deforinati on ;Ind to the difference
between fibres mad e in differen t wa y ~. Pa rt of th e
pro bl em a ri ses from
the
fact
th a t th e
high-temperature se tt ing is ove rlaid by the effects of
temporary se tlin g at the glass tran si ti o n a nd by the
effects of reversible of th erm a l expa nsio n a nd
co ntraction. which, fQrexarnple . ha ve bee n shown to
ca use anomal o us overtwi stiIlg in heat-setting l l .
C i rcul11s ta n t ia l evide nce fr om ind ustri a I teC hn o logy
3
INDIAN J. FIBRE TEXT. RES., MARCH 1991
suggests that polyester fibres are more easily set
repeatedly at the same or lower temperatures than
nylon and require more cooling, but these
phenomena are not well documented scientifically.
The experiments of Buckley a nd Salem 12 show a clear
time-temperature interdependence in polyester, but
the. situation' in nylon is ambiguous lJ . Even in
phenomenological terms, there is no clear and simple
theoretical representation of heat-setting, though
Buckley and Salem have made a start in what is
possibly an unduly so phisticated mathematical
form.
The commonly accepted explanation of
heat-setting is that it is due to progressive
enhancement of crystalline perfection, some form of
annealing with melting of small imperfect crystals and
growth of larger and better ones. However, other
physical variants are possible: the effects of multiple
melting, with changes between the forms I and II
denoted by Bel1 6 ; mobil ity of <;:rystal defects, as in the
o bserved transitions in polyethylene near the melting
point ; crystal plasticity, the influence of yielding
deformation. Surprisingly, most of the reports of
dynamic mechanical analysis of nylon and polyester
stop at about 160°C. Furthermore, heat.setting may
not be due to a ph ysi cal mechanism, or more correctly
physical mechanism s may not be the only ones
involved. Fakirov l" and others have suggested that
so lid-stat e c hemical reac tion s may occur at a
substantial
ra te
at
these
temperatures.
Tra nsesterification ha s been postulated as a
mechani sm by which polyester films can be welded
together. In heat-se tting, thi s would mean that
stressed tie mo lecul es could be broken and the
crystallites linked together in a new network.
Heat-se tting, in the st rict sense of sta bilizing a
particular geometric form, is only one of the
thermo-mechanica l properties of nylon and
polyester fibres . In addi ti o n, there are heat shrinkage
or sometimes spo ntaneo us elongation , change in
mech a nica l properties such as strength and modulus,
changes in diffus io n rates which affect dyeing, a nd
other effects resultin g from str uctural changes.
Technologically, these are of grea t importance, but
there is no coherent body of scientific knowledge and
understanding about them.
4 Analysis of Mechanics
There has ben little a ttempt to develop any
theoretical analysis of the structural mechanics of
nylon and polyester fibres. It is assumed, with good
reason , that deformation is easiest in amorphous
region s. Below the glass transition the hydrogen
bonding in nylon , and some other interaction in
4
polyester, will cause the amorphous polymer to act
like a rather highl y crosslinked rubber. Above the
glass transition the chains will be free between the
crystalli tes.
It has been shown 15 that a simple composite model
is inadequate as a way of treating the problem. A wide
range of predictions can be obtained depending on the
degree of parallel or series influence. The fact that only
two structural parameters, not the same two in
different theorjes , are involved means that obviously
different structures lead to the same predicted values.
Possibly, a more exact stress analysis would give
better results, but in a new approachl o, still to be
completed and fully published, we have preferred to
consider a more explicit model of crystalline blocks
linked by molecular chains. This seems to be the right
way forward , but more fundamental work remains to
be done. It is not yet clear whether predictions of too
low a value for moduli are due to an incorrect
formulation of the geometry of the model, to some
simplifications in the computer model , or to basic
features of the way in which polymer chains behave in
confined spaces. One interesting feature , which
should have been obvious before, is that eve n under
zero applied stress, the ti e molec ules will be under
substantial tensi o n as they try a nd pull the network
into a smaller vo lume.
5 Differences between Nylon and Polyester
1n the above discussion , there has been Ii ttle
di stinction between nylon and polyester fibres , and
this is true of most models in the literature. Apart from
qualitative suggestions that the benzene rings must
stiffen up the amorphous regions and so cause a higher
modulu s in polyester fibres , and the greater effects of
moisture on nylon , there are no reasons given for
expecting differences.
But there are other differences in behaviour. The
rather poorly understood effects in heat-setting have
been mentioned . A much clearer difference occurs in
the morphology of fatigue behaviour I 6. Both nylon
and polyester show tensile fatigue when cycled
between zero load (or sometimes a very small positive
load) and about half the normal breaking load. But
the form of break, as shown in Fig.3, is different.
Nylon fibres have a short tail due to a crack that runs at
an angle across the fibre under the influence of the
shear stress at the tip of an initial transverse crack . In
polyester, the crack runs almost parallel to the fibre
axis and the breaks are extremely long. What
structural feature causes this difference?
Similar effects have been observed in internal
abrasion in ropes. In nylon, the shear stresses cause
angled cracks across the fibres and fail rapidly . In
HEARLE : UNSOLVED PROBLEMS IN SCIENCE OF NYLON AND POLYESTER FIBRES
established in practical industrial and marketing
operations, a drive for scientific understanding may
be dismissed as special pleading by academic
scientists. I do not believe this is so. There are several
reasons and , in my opinion , those countries and
companies that appreciate them will be the industrial
and financial leaders in the next century.
Fig. 3- Form of ten sile fati gue break s: (a) nyl o n; and (b)
polyester
contrast to this, polyester ropes have been found to
perform much better in conditions of cyclic loading.
6 Conclusion
Despite their commercial success and effective
industrial production and use , there is much that
remains unknown about nylon and polyester fibres in
sCientific terms. As in much of textile technology , the
understanding of chemical aspects is much greater
than that of mechanical aspects. However, in order to
show that situations can change, it is worth quoting
the statement made in 1928 by W L Balls, the greatest
fibre scientist of his day. He referred to " the empirical
basis on which the artificial silk (rayon) industry at
present rests. Not that its technique is not already
specialized, and its control eminently scientific; it is
only permissible thus to describe it as empirical
because its raw material (cellulose) is largely an
unknown substance. The chemical nature and
physical structure of cellulose, or rather of the
different cell uloses, are now only beginning to be
unravelled"l? Within a few years, the polymer
hypothesis had become accepted fact, and the
chemistry of cellulose had been unequivocally
established. This led to the major advances in
high-strength rayons for tyre cords, and later in high
wet modulus rayons, through reserch that depended
on the knowledge of the chemistry. Again, only a few
years later, another variety of ' artificial silk', nylon,
resulted from Carothers' elucidation of the chemistry
of condensation
polymerization.
A
real
understanding of the molecular physics of
melt-spinning of nylon and polyester fibres and the
resulting physical fine structure, which are only now
beginning to be unravelled, might lead to similar
advances in our future years.
Unfortunately, for myself, whose career has been
divided between Britain and the USA , there now
seems to be in these countries a lack of enterprise in the
long and hard pursuit of these goals. With so much
Industrial needs are continually changing.
Advances in engineering lead to changes in machines
to give improved production economics, but result in
different fibres from those produced by older
methods. Unless the scientific base is improved, the
same old way of long empirical trials will be needed, in
a more competitive environment, in order to establish
satisfactory operation . On-line measurement would
lead to possibilities for automation but without
understanding it is difficult to know what to measure
or what control strategy to adopt.
The production of melt-spun fibres is derived from
the simple experiment of pulling out a molten
thread-line, letting it cool and then drawing it. Later,
the effects of hot stretching or relaxation of the drawn
fibre, and then of high-speed attenuation of the
molten thread-line became appreciated, but
generally the impetus for change has been engineering
ecoriomy. The therl}1o-mechanical sequence remains
predominantly monotonic. However, the sensitivity
of polymer assemblies to thermo-mechanical
influences sug6ests th a t vther sequences would give
different structures and properties, but without
guidance from understanding any trials would be
wandering in the dark. Science is needed to throw light
on the right direction to move industrial progress.
Finally, the diversity of the textile fibre market,
which ranges from high fashion to powerful
engineering; requires the ability to tailor structure to
give properties, which match product needs. Fibre
structures should not just happen as a result of
empirical manipulations but should be designed. In
order to be able to do so, the unsolved problems in the
science of nylon and polyester fibres must be solved.
References
1 Molecular conformation and dynamics of macromolecules in
condensed systems. edited by M Nagasawa (Elsevier, Amsterdam) 1988.
2 Hearle J W S & Greer R, J Text Inst. 61 (1970) 243 .
3 Prevorsek D C, Harget P J, Shanna R K & Reimschussel A C, J
Macromol Sci, B-8 (1973) 127.
4 Peterlin A, J Mat er S ci, 6 (1971) 490.
5 Hea rle J W S & Greer R. Text Prog. 2(4) (1970) 53 .
6 Bell J P. Slade P E & Dumblcton J H. J Polym Sci. A-2, 6 (1 %8)
1773.
7 Hearle J W S. J Appl Polym Sci (Al'pl Pnlym Symp) . 31 (1977)
137.
5
INDIA
1 J.
FIBR E TEXT. RES .. MARCH 1991
8 H igh-speedjihre spillllillg. edi ted by A Zia hi cki and H Kawai
(Wiley, New York) 19X5: Heu vel H M & Hui sman. P. 3 10:
Shimizu J: Okui N & Kikutani T. P. 444.
9 Hea rl e J W S. in Film' Sfmeflln'. edited by J W S Hearle & R H
Pete rs (The Textile Institute. Manchester) 1963.209.
10 HearJe J W S. Prakash R. Wilding M A & Da\'is H A. in IlIIegmfion oflimdalll('lIIal pO/rIll<'l" sciellce 1I1It! 1('c!/l/ %gl' - 2 (2//(/ RoIdlle Po/rill COIlf) . edited by P J Lemstra & LA Kleintj cns
(E lsevier. London) 19XK . 540.
.
II Buck ley C P. Hearle J W S & Man da i R. J 1'('.\"1 11/.1'1. 76 (1 985)
264.
6
12 Buckley C P & Salem D R. PII/rllll'r. 2R (l9!1 7) 69.
1.\ Hea rle .l W S. Wilding M A. AuyeungC & Ihmayed R.J Texf
IlISf . 8 1 ( 1990) 2 14.
14 Fakirov S. in So/idS{(If(' hehm'iollr ofIiI/ear poiresfers alld Jio/rIIlI lii/es. edi ted by J M Sch ult z & S Fakirov (Pre ntice Hall.
Englewood Cliffs. New Jersey) 1990. Chap. I.
15 Hea rl e J W S. Prakash R & Wildi ng M A. Poivl1ler . 28 ( 1987)
441.
16 Bun sell A R & H earle J W S, J Appl PO/I'm Sci. 18 ( 1974) 267.
17 Ba ll s W L. Sflldies ol l/ua/if,)' in CO /fOIl ( Macmi llan . London )
1928. 303.