Textile Research Journal

http://trj.sagepub.com

The Strength and Weakness of Cotton Fibers

Helmut Wakeham and Nancy Spicer
Textile Research Journal 1951; 21; 187
DOI: 10.1177/004051755102100401

The online version of this article can be found at:

http://trj.sagepub.com/cgi/content/abstract/21/4/187

Published by:

http://www.sagepublications.com

Additional services and information for Textile Research Journal can be found at:

Email Alerts: http://trj.sagepub.com/cgi/alerts

Subscriptions: http://trj.sagepub.com/subscriptions

Reprints: http://www.sagepub.com/journalsReprints.nav

Permissions: http://www.sagepub.com/journalsPermissions.nav

Citations (this article cites 7 articles hosted on the

SAGE Journals Online and HighWire Press platforms):
http://trj.sagepub.com/cgi/content/refs/21/4/187

Downloaded from http://trj.sagepub.com by satish aparaj on November 19, 2007

© 1951 SAGE Publications. All rights reserved. Not for commercial use or unauthorized distribution.
TEXTILE RESEARCH
JOURNAL

The Strength and Weakness of Cotton Fibers*

Helmut Wakeham and Nancy Spicer
Textile Research Institute,† Princeton, New Jersey

Abstract
The structural reversals in cotton fibers are a preferred location of break when the fiber
is ruptured in tension. With the aid of a polarizing microscope, observations were made of the
fraction of fibers breaking at the reversals under various conditions of moisture content, specimen
length, and chemical treatment. It is concluded that the cellulose in the region of the reversals
is more highly crystalline than the cellulose between the reversals. The cause for breakage at the
reversals is then explained on the basis of internal stresses at these points which cause the fiber
to tear apart when tension is applied. The weak places between reversals which give rise to
fiber rupture are believed to be at structural defects or thin places. The mean breaking stresses
for fibers breaking at the reversals vary significantly from those for fibers breaking between
reversals for some varieties of cotton. The distributions of weak-spot location and strength for
reversals and structural defects are discussed for three varieties.

Introduction ers. It is also necessary in any fundamental study

of the factors contributing to yarn and fabric prop-
A fundamental knowledge of the weak places in
erties. For these reasons the Textile Research In-
textile fibers as well as information on the mecha-
stitute has engaged in an investigation of the ele-
nism of fiber rupture are basic requirements in fiber
ments of strength and weakness in textile fibers.
science and technology. Such information is essen-
Cotton has been selected as an initial example for
tial to an intelligent program for the development of
study because of the many elements of weakness
improved fibers by the cotton breeders, various nat- which each single fiber contains. The present paper
ural-fiber producers, and synthetic-fiber manufactur-
on the nature of weakest links in cotton fibers re-

* Presented at the 21st Annual Meeting of Textile Re- ports some of the preliminary work in a more general
search Institute, New York City, Nov. 16, 1950. investigation of the mechanism of fiber rupture.
t This work, initially sponsored by The Textile Founda-
tion, is now supported by the general research fund of Tex-
Cotton with increasing specimen length exhibits a
tile Research Institute. relative loss in fiber strength which is greater than
187

Downloaded from http://trj.sagepub.com by satish aparaj on November 19, 2007

© 1951 SAGE Publications. All rights reserved. Not for commercial use or unauthorized distribution.
188

FIG. 1. Effect of specinien Levcgth on cotton-fiber

strength..

that of any other textile fiber. Wool and other nat- FIG. 2. Distributions of the length of fiber involved
in the break and of the distance between reversals f or
ural fibers show appreciably lower tensile strengths
iliids cotton.
at 10-mm. than at 1-mm. specimen lengths, in some
cases as much as 10$o lower. Synthetic fibers show rotations complementary, and the point at which
are
a much smaller drop in strength. But the observed
the color changes indicates the reversal [9].
strength of cotton at a specimen length of 10 mm. Cotton-fiber reversals are randomly spaced along
may be more than 50% lower than the strength at the length of the fiber-approximately 100 to the
1-mm. test length (Figure 1). Although evaluations
of this type are frequently made with fiber bundles, inch-although the average distance between rever-
sals appears to vary somewhat with variety. A dis-
single-fiber test results support the observations with tribution of distances between reversals is of the log-
bundles and indicate that the loss in strength is in-
normal type shown in Figure 2. This statement
herently a fiber property. This great decrease in implies only that there is a minimum distance be-
strength with long cotton specimens is undoubtedly tween reversals, probably determined by the length
due to a very high frequency of weak linkages along
of the reversal itself, and that the probability of find-
the length of the fiber.
It is widely believed that the weakest links in
-

ing an adjoining reversal at a distance greater than

the minimum exponentially approaches unity.
cotton fibers occur at the structural reversals first
studied extensively by Balls and Hancock [3] in
1926. In the cotton fiber the cellulose molecules and Experimental Procedure

bundles of molecules, or fibrils, instead of lying In order to determine the part cotton-fiber rever-
parallel to the long axis of the fiber, spiral around sals play in fiber rupture, it must be possible to
this axis in a helical manner. A fiber reversal occurs
distinguish between breaks which do or do not in-
at those points where the direction of rotation of the volve a reversal. As implied above, this differen-
cellulose about the fiber axis changes from a Z-direc- tiation can be accomplished by merely examining
tion to an S-direction, or vice versa. The reversal
corresponding broken ends of the fiber under a mi-
can be readily located in the natural, unswollen fiber
croscope used with a quarter wave red plate so as
by observing it with a microscope using elliptically to produce elliptically polarized light. If the two
polarized light (See Figure 3A.) Under these broken ends of the properly oriented fiber in the
conditions the interference colors for the Z and S
field appear the same color, the break took place
* For a
description of elliptically polarized light, the reader betweei2 reversals. If the two ends appear differ-
is referred to any standard textbook on physical optics, such
as &dquo;Introduction to Physical Optics,&dquo; by J. K. Robertson,
ently colored-e.g., orange and blue-the break oc-
3rd Ed., New York, D. Van Nostrand Co., Inc., 1942. curred at a reversal. Examples of these two types

Downloaded from http://trj.sagepub.com by satish aparaj on November 19, 2007

© 1951 SAGE Publications. All rights reserved. Not for commercial use or unauthorized distribution.
 189

Fm. 3. Photoinicrographs of cotton fibers using elliptically polarized light. (Magnification approximately
600 X.) A (top)-Fiber reversal, showing change of spiral orientation. B (vniddle)-Fiber broken at the re-
versal. C (botto~)-Fiber broken between reversals.

of breaks are shown in Figure 3 (B and C). From of being right, that the fiber breaks preferentially at
these photographs it is quite apparent that not all the reversal.
fibers break at a reversal. It is possible from the information outlined above
The existence of cotton-fiber breaks between struc- to establish a method for the investigation of the
tural reversals does not eliminate the possibility that fiber reversal as a weak link. From determinations
fibers may break preferentially at the reversals. Such of the fractions of fiber breaks taking place at this
a preference can be established by examining many position for various fibers under various conditions,
fiber breaks, but only if some previous knowledge of certain information concerning fiber reversals and
the expectancy of either type of break is available. the contribution they make to fiber strength has been
In the discussion below, all breaks involving a re- obtained. These observations are summarized in the
versal in a~y way have been classified as breaks at Results section.
the reversal. Even if the reversals were not weak Each fraction (percentage) of fibers breaking at
places in the cotton fiber, some of them would be the reversal reported is based on the observation of
involved in the fiber break simply by chance because at least 100 fibers and in some cases as many as 400.
there are so many of them along the length of the The fibers were selected from a standard length ar-
fiber. On this basis the fraction of fibers breaking ray prepared with a Suter-Webb cotton-fiber sorting
at a reversal would be the ratio of the length of fiber apparatus [ 1 ] , the length chosen being that corre-
involved in a broken place to the mean distance be- sponding approximately to the staple length or upper-
tween reversals. From microscopic measurements quartile mean. It was noticed that the sorting opera-
this ratio turns out to be less than 0.15 (see Figure tion yielded a well-blended fiber sample: the random
2) so that an observation of more than 15 % of the selection of 100 fibers from such a sort gave a frac-
fiber breaks involving a reversal may be taken as tion of mature fibers which agreed very well with a
evidence of an unusual weakness in the fiber at the standard maturity count obtained by the preferred
reversal. Furthermore, from a statistical significance A.S.T.M. method in which the fibers are swollen in
test, it can be shown that if half of the breaks in- 18 % sodium hydroxide solution [1].
volve reversals, it is only necessary to examine 50 The selected fibers were broken in one of two ways
fibers in order to conclude, with a 95 % probability -by hand or by a single-fiber stress-strain test. In

Downloaded from http://trj.sagepub.com by satish aparaj on November 19, 2007

© 1951 SAGE Publications. All rights reserved. Not for commercial use or unauthorized distribution.
190

TABLE I. EFFECT OF VARIETY ON LOCATION OF BREAK AND MEAN BREAKING STRENGTHS

(T.R.1.-SCHAEVITZ TESTER, 70°F, 65% R.H., EXCEPT WHERE NOTED)

*
Mean fiber strengths. The coefficient of variation for these breaking-stress values is about 40% of the mean.

both methods the fibers were first mounted at each 2. There is no significant relationship between
end with Duco cement on small film tabs approxi- fibers of different cross-sectional shapes or maturi-
mately 4 in. by ’ 2 in. in size. In the hand method the ties and the fraction of fiber breaks at reversals.
fiber was slowly pulled apart by sliding the tabs The cross-sectional shapes of fibers were obtained by
apart on a microscope slide. In the machine method observing the color of the fiber in longitudinal view
the fiber was mounted in the T.R.I.-Schaevitz tester when properly oriented in a polarizing microscope
previously described [12] and broken at a constant with elliptically polarized light. Classification was
rate of extension of 20% per min. No effect of the made into linear, elliptical, and roundish fibers, as
method of break on the fraction of fibers breaking at defined in the A.S.T.M. specifications [1]. In this
or betweevc the reversals was observed. grouping the linear fibers would ordinarily be called
In those experiments in which the fibers were immature and the elliptical and roundish fibers ma-
broken on the T.R.I.-Schaevitz tester, the breaking ture. Of the more than 1,000 fibers examined in
loads and breaking-stress values of the fibers were all the varietal groups listed in Table I, 43 % of the
also determined. The breaking-stress values were linear, 46 % of the elliptical, and 42 % of the roundish
obtained by dividing the breaking loads by the mean fibers broke at reversals. Although there are ap-
grex values of the individual fibers determined from proximately three times as many elliptical fibers as
a vibroscope measurement which has been described roundish ones in a common cotton sample, there is
elsewhere [4, 5]. The purpose of quoting stress no significant difference in the location of break for

values instead of breaking loads is to permit com- mature and immature cotton fibers.
parison of mean values for fiber groups of different 3. Increasing the specimen length increases the
types and varieties. fraction of breaks involving reversals (Table II).
This observation gives additional evidence that the
Results fiber reversals are the possible locations of weak
The results of observations made in the manner places in the fiber because increasing, the specimen
described above and shown in Tables I-IV permit length directly increases the number of reversals
the following statements concerning the location of available to act as the weakest link. This result also
break in cotton fibers to be made: bears certain implications concerning the distribution
1. Of several fiber varieties investigated at rea- of weak places occurring between the reversals, which
sonably long lengths and 65 % R.H., about half of will be discussed later.
the fibers broke at structural reversals (Table I ) . 4. Increasing the moisture content of the cotton
This fraction is sufficiently greater than the theo- increases the fraction of breaks involving reversals
retical value of 15 % to permit the definite state- (Table III).
ment that reversals constitute a weak link in the 5. Treatments which modify the cellulose structure
chain. or damage the fiber so as to reduce its strength de-

Downloaded from http://trj.sagepub.com by satish aparaj on November 19, 2007

© 1951 SAGE Publications. All rights reserved. Not for commercial use or unauthorized distribution.
 191

6. The mean breaking strength of cotton fibers

TABLE 11. EFFECT OF SPECIMEN LENGTH ON
LOCATION OF BREAK
’

breaking at reversals does not differ greatly from

(70°F, 65% R.H.) those breaking between reversals (Table I). Appar-
ently there is a varietal difference in this respect.

In some varieties the fibers breaking between re-

versals have an average strength higher than the
average for those breaking at the reversals. With
Wilds cotton the converse seems to be the case,
whereas Stoneville shows no difference. Attention
is also called to the observation for Empire cotton
that increasing the moisture content also increases
the mean breaking stress for fibers breaking between
TABLE III. EFFECT OFMOISTURE ON LOCATION reversals much more than that for fibers breaking at
OF BREAK
the reversals.

Discussion
One of the most interesting of the observations
summarized above is that increasing the moisture
content of the cotton fiber increases the fraction of
fibers breaking at the reversal (Table III). Now it
is well known that increasing the moisture content of
cotton in the range between 35% and 65% R.H. also

TABLE IV. EFFECT OF FIBER TREATMENT ON

increases the strength of the cotton fiber [7]. In
LOCATION OF BREAK some way or another this effect takes place by

strengthening the weakest linkages, which have been

shown to involve fiber reversals, at least in some
cases. But if the reversals are preferentially affected
in this manner, the fraction of breaks at the reversals
would be expected to decrease rather than to increase.
The facts thus indicate that the moisture affects pri-
marily those weakest linkages which occur between
the reversals. Furthermore, x-ray and structural
considerations suggest that the moisture effects take
place mainly in the less ordered or noncrystalline
regions because water molecules cannot readily enter
the crystalline cellulose [6]. Hence, it is logical to
conclude that the reversals, which are not preferen-
crease the fraction of fiber breaks involving reversals tially strengthened, are more highly crystalline than
(Table IV). Swelling the fiber in strong caustic is the remainder of the fiber.
greatly reduces the tendency to break at reversals. The above considerations suggest that the increase
Treatment with formaldehyde in the vapor phase to in moisture should preferentially increase the break-
introduce as little as 0.35% fixed
formaldehyde de- ing stress for those fibers breaking between reversals.
creases not only the strength of
the fiber but also Such an effect is observed for Empire fibers broken
the fraction breaking at the fiber
reversals. Mildly at 35% and 65 % R.H. (Table I ) . The increase in
hydrolyzing the fiber for 18 hrs. in 1N hydrochloric breaking stress with increasing moisture content for
acid solution at 60°C reduces the fraction of fibers those fibers breaking between reversals is 2.5 times
breaking at the reversal f rom 40 % to 26 % . On the that for the fibers breaking at reversals.
other hand, treatment with 6.5 % melamine resin in The conclusion of higher crystallinity at the re-
such a way as to produce no appreciable loss in fiber versals is further supported by the experiment in
strength does not affect the location of breakage. which fibers were mildly hydrolyzed. The acid-

Downloaded from http://trj.sagepub.com by satish aparaj on November 19, 2007

© 1951 SAGE Publications. All rights reserved. Not for commercial use or unauthorized distribution.
192

catalyzed hydrolysis reaction presumably takes place Both Balls (2~ and Meredith (8] have suggested
first in the less ordered regions, cutting the cellulose that the cotton fiber may be internally stressed at
chains in two in these areas. If the reversals are the reversals. Meredith proposed further that this
more crystalline than the rest of the fiber, then the possibility may explain the observation that fibers
fibers, being preferentially damaged by the acid hy- with higher helical angles are weaker fibers. In
drolysis between the reversals, should show a lower other words, the tearing tendency at the reversals
number of breaks at the reversals. Such was found will be less in those cases in which the angular change
to be the case. A cotton sample showing 40% fibers in direction at the reversals is also less.
breaking at the reversals before treatment showed But what about the breaks which occur between
only 26% fibers breaking at the reversals after an reversals? What is the nature of these weak links?
18-hr. hydrolysis in 1 ~l hydrochloric acid at 60°C As a result of the observations on reversals, it
(Table IV). seems reasonable that the breaks occurring between

This conclusion also raises the question concerning reversals are those caused by real structural defects
the reason why fibers tend to break a.t reversals. If and thin places in the cotton fiber. Just as there are
the reversals are more crystalline than the rest of the distributions [10] of locations and strengths for the
fiber is, they might be expected to be exceptionally weakest links at reversals, so there are also distri-
strong rather than weak. The evidence in Table I, butions of locations and strengths for the weakest
as well as the fact that fibers do preferentially break links between reversals. The increase of breaks at
at the reversals, precludes the possibility that re- reversals with increasing specimen length (Table II )
versals are exceptionally strong. indicates that on the average the structural weak
The experiment with caustic-swollen fibers (Table links are somewhat farther apart than the reversals
IV) suggests that fibers break at reversals because are. On the other hand, it must be remembered
the structure at these points is already strained, so that the distributions of these two types of weakest
that additional load causes the fibers to tear apart linkages are imposed upon the same fiber length.
at one of these points. It is known that swelling in There is thus a fair probability that the location of
a structural defect may coincide with that of a re-
strong caustic and subsequent removal of the caustic
by washing out changes the structure of the crystal- versal. This possibility greatly complicates the prob-
line portions of native cellulose. It may therefore lem of finding out more about the weak places by the
be assumed that the 18 % sodium hydroxide treat- experimental methods now being used.
ment reduces the number of breaks at the reversals Furthermore, the results of Table I indicate that
by affecting the structure of the reversals themselves. there may be varietal differences in the relative dis-
If the cellulose structure in the region of the reversals tributions of strengths for the two types of weakest
is under an internal stress, swelling the cellulose linkages in the fiber. In some varieties the fibers
would permit the relaxation and removal of these breaking at reversals are on the average weaker than
internal stresses and would eliminate the cause of those breaking betzueeJa; in others the converse is
true.
preferential breakage at the reversals. The observed
fraction of breaks at the reversal (18%) for the
caustic-swollen fiber is not significantly different from TABLE V. COMPARISON OF WILDS, SAK. AND
HOPI ACALA 50 COTTONS
the theoretical value of 15 % based on a chance in-
volvement of the reversal. The hypothesis that cot-
ton fibers break at reversals because these points are

already internally stressed is thus supported by the

experimental evidence.
Although it has not yet been possible to perform
the experiment, one way to check this hypothesis
would be to examine cotton fibers from the unopened
boll which had been broken before drying the first
time. Presumably, these stresses at the fiber re-
versal are set up during the initial drying of the
cotton-fiber cell.

Downloaded from http://trj.sagepub.com by satish aparaj on November 19, 2007

© 1951 SAGE Publications. All rights reserved. Not for commercial use or unauthorized distribution.
 193

other two. Furthermore, the fraction of breaks in-

volving reversals is lower for the Acala cotton by
an amount proportional to the number of reversals

available to break.
From these data the probable arrangement, shown
in Figure 4, of position and strength distributions for
the reversals (solid curves) and for the structural
defects (dotted curves) has been postulated. The
curves in the left half of the figure represent log

normal distributions of the distances between weak

links in the fiber and are similar to the distribution
shown in the right portion of Figure 2. In the in-
terpretation of these curves, the greater the distance
between weak links the farther apart they are along
the length of the fiber and the lower the number
found in any given test length. In the right half of
Figure 4 the curves represent simple frequency dis-
tributions of the strengths of links in which strength
FiG. 4. Probable arrangement off position and
distributions of f weak links in cotton fibers. is plotted in a linear manner along the abscissas.
strength
Solid lines are f or structural reversals; dotted ~~~
lines at
are Only the relative positions (not shapes) of the curves
f or structural defects between the reversals. are significant in the present consideration.

Space does not permit discussion of all the con-

Thus it is that the strength of a cotton fiber is siderations necessary to arrive at this arrangement.
determined by the weakest link available from a com- Suffice it to point out that it is possible from these
bination of no less than four probability factors in- distributions to explain in a qualitative way the
volving the locations and strengths of at least two observed strength relationships of these three varie-
types of weakest linkages. Is it any wonder, then, ties of cotton. For example, the high strength which
that the observed breaking strength of cotton fibers the Hopi Acala 50 gives in the conventional Pressley
varies as greatly as it does from one fiber to the next ? test, with no space between the clamps, is very prob-
It is interesting to speculate somewhat concerning ably due to the much lower chance of finding either
the nature of these four distributions of weak-spot reversals or structural defects in so short a length
locations and strengths. This has been done for of fiber. The Sak cotton has so many more weak
three cottons which are of commercial interest be- places per unit length that even their relatively high
cause of their different behavior in the modified strength fails to place it above the Acala on a
Pressley bundle strength test [ 1 ~ . One of these, Pressley scale. On the other hand, the strengths
the Hopi Acala 50, has an unusually high Pressley of the weakest links in the Acala cotton are so low
strength in the conventional test, with no space be- that whenever the test is made with a relatively long
tween the clamps. (See Table V.) At a clamp specimen length this cotton appears much weaker
separation of 5 mm. [ 11 ] , however, it appeared to than the Sak.
be no better than the Wilds cotton, although the The Wilds cotton has the most reversals per unit
conventional Pressley rating would indicate it to be fiber length of the three cottons compared. But be-
far superior. Single-fiber strength tests confirmed cause the reversals are relatively stronger than those

the observations of the Pressley test. They also of the Acala, the Wilds appears to be just as strong
showed that the Wilds differed from the other two in a long test length. In a short test length, the
in exhibiting higher average strength values for the high frequency of weak places makes it appear to
fibers breaking at the reversals than for those fibers be weaker than the Acala.
breaking between reversals. The experiments herein reported suggest several
An examination of the reversal spacing for these lines for future work, some of which are now under
cottons showed that the reversals in the Hopi Acala investigation. The true nature of the distributions
50 cotton are considerably farther apart than in the of breaking stress for the two types of weakest link-

Downloaded from http://trj.sagepub.com by satish aparaj on November 19, 2007

© 1951 SAGE Publications. All rights reserved. Not for commercial use or unauthorized distribution.
194

ages is of considerable interest in the theory of fiber Acknowledgments

strength. Evaluation of the effect of specimen length The authors are pleased to acknowledge the as-
on the difference between the mean breaking stresses
sistance of Carol Piper in the experimental work
for the two types of breaks should yield some infor-
mation in this direction. The effect of moisture on during the early part of this investigation. Fred
the breaking stresses should be investigated for varie- Bauer prepared the partially hydrolyzed samples,
ties of cotton other than the Empire cotton observed Robert Wagner prepared the formaldehyde-treated
herein. It may be that this same type of selective cotton, and Malcolm Jewell of Bates Manufacturing
action occurs for various chemical treatments which Company furnished the resin-treated fibers.
modify fiber strength as well as the fraction of fibers
breaking for any one type of break. Literature Cited
1. A.S.T.M., "Tentative General Methods of Testing
Summary Cotton Fibers," Standards on Textile Materials,
It has been demonstrated that cotton fibers tend Designation D414-49T, 1950, p. 250 et seq.
to break preferentially at the structural reversals, 2. Balls, W. L., "Studies of Quality in Cotton," Lon-
which may be readily seen when the fiber is placed don, Macmillan Co., 1928.
3. Balls, W. L., and Hancock, H. A., Proc. Royal Soc.
in a polarizing microscope using elliptically polarized
(London) B99, 130-47 (1926).
light. The cellulose at the reversals appears to be 4. Dart, S. L., and Peterson, L. E., TEXTILE RESEARCH
somewhat more highly crystalline than the cellulose JOURNAL 19, 89-93 (1949).
in the remainder of the fiber. It is suggested that 5. Gonsalves, V. E., TEXTILE RESEARCH JOURNAL 17,
the reversals behave as weak places because of in- 369-75 (1947).
ternal stresses produced there during the growth and 6. Hermans, P. H., "Physics and Chemistry of Cel-
lulose Fibers," New York, Elsevier Publishing
production of the fiber. There are also other points 1949.
of weakness in the fiber which are probably due to Co., Inc.,
7. Mann, J. C., J. Text. Inst. 18, T253-64 (1927).
real structural defects in the fiber wall. The ob-
8. Meredith, R., J. Text. Inst. 37, T205-18 (1946).
served strength of the fiber under given conditions 9. Osborne, G. G., TEXTILE RESEARCH 5, 307-25
of test is then a function of the probability factors
(1935).
controlling the location and strength of the weakest 10. Peirce, F. T., J. Text. Inst. 17, T355-68 (1926).
link in the tested length, whether that link be a struc- 11. Phillips, J. K., TEXTILE RESEARCH JOURNAL 18,
tural reversal or a structural defect of some kind 684-6 (1948).
between the reversals. The nature of these probabil- 12. Reichardt, C. H., Schaevitz, H., and Dillon, J. H.,
Rev. Sci. Instruments 20, 509-16 (1949).
ity factors for three commercial cottons has been
considered. (Manuscript received January 18, 1951.)

Downloaded from http://trj.sagepub.com by satish aparaj on November 19, 2007

© 1951 SAGE Publications. All rights reserved. Not for commercial use or unauthorized distribution.