4

Stress-Strain Diagram, Young's Modulus

and Poisson's Ratio of Textile Fibers

By Kenji Higuchi and Hideo Takai, Members,TMSJ

Faculty of Textile Industry, Tokyo University of Agriculture and Technology, Koganei, Tokyo

Abstract

Dynamic properties represented by curved load-elongation diagrams of textile fibers

have been re-examined to obtain suitable values for their elastic properties. For
this purpose, we have used the logarithmic strains In(l/lo) for longitudial and

ln(r/ro) for lateral, and Poisson's logarithmic ratio pt=-(dr/r)/(dl/1). The authors

also defined Young's logarithmic modulus, Et=(W/s)/((l/1o)-2ƒÊt ln(1/l)l, which are

based on the logarithmic strains. The decrease in the cross-sectional area accom

panying a longitudinal extension of a textile fiber has been studied together with its
Poisson's ratio. The results :

1) A testing machine based on the principle of an air-micrometer makes it

possible to record continuously both longitudial and lateral contractions.
2) Nylon-6 fishing lines show an almost constant Poisson's ratio of ƒÊ=0.385
0.386 until the longitudial strain reaches 11~13%. Beyond that point, the lateral

strain changes its trend and Poisson's ratio reaches as much as ƒÊ=0.446 at the

breaking point.
3) Fluctuations in the strain diagram are presumably due to the instability of
a lateral -vs.-longitudial strain diagram.

1. Introduction 2. Definition of Young's Modulus

Young's modulus, or the modulus of elasticity, Take an ideally elastic material satisfying Hooke's

is one of the most important measures of the me law and let ƒÐt be the stress and ƒÃt the strain at any

chanical properties of a material. However, it is point (i) in the straight-line region of the stress
difficult to obtain an exact stress-strain diagram on strain diagram. Then, Young's modulus E is

textile fibers even if we use the load-elongation dia defined as: [2]

gram as a substitute for the stress-strain diagram.

E=ƒÐt/ƒÃt ...... (1)
This is because the load-elongation diagram does
not make straight lines and because the percentage If the initial length of a material is lo, and the
of elongation is so high that many difficulties occur length with load W added is 1, then sectional area
in determining Young's modulus. Generally, his
s and the strain are definable thus:
modulus is measured while elongation is kept very
ƒÃ= (l-lo)/lo=‡™l/lo ......(2)
small.
Hitherto the load-elongation diagram has usually where l-lo=dl is the elongation length.
been explained by various visco-elastic models. [11
E=ƒÐ/ƒÃ= (W/s)/ ((l-lo)/lo} = (W/s)/(‡™l/lo)
We have inquired into the modulus of elasticity by
observing the relation between stress-strain diagrams If the sectional area remains unchanged in the
and load-elongation diagrams. process of stretching, then:
The decrease in the cross-sectional area accom
s=so (=constant) ...... (3)
panying a longitudinal extension of a textile fiber
has been studied together with its Poisson's ratio. where so is the initial sectional area
 5

and E=(W/s)/{(l-l0)/lo} =K(W/‡™l) (4)

lim El=(W/so)/ƒÃ=E ......(9)
where K=1o/so is a constant.

In a general case where the stress-strain diagram where e=(l-lo)/lo. Young's logarithmic modulus

is not a straight line we can apply equation (4) to Et here is equal to his ordinary modulus E.

any small section of the stress-strain diagram. Then 4-2. When the decrease in sectional area is not

the numerator and the derominator of equation (4) negligible but Poisson's ratio ƒÊ is constant.

become incremental changes in stress ƒÐ and strain If Poisson's ratio ƒÊ stays constant in the process

ƒÃ respectively. Therefore, dƒÐ/dƒÃ=E ...... (5) of elongation

s=s0 {1-2/z,(l-lo)/lo}
As in equation (7), Young's logarithmic modulus
3. Logarithmic Strain in this case is
El= (W/so) / {1-21ƒÊ(1-lo)/lo} 1n (1/1o)

...... (10)
Equation (2) holds good for a material like
metal whose extension is negligible, but not for Therefore,
materials such as textile fibers which have an elon
{1+ (1-lo)/lo) {1-2ƒÊ(1-lo)/10
1=exp(W/soEl)
gation as high as 10-40%. For such materials it ...... (11)
is better, in our opinion, to define the strain as the
intergral of incremental strain. Incremental strain Equation (11) shows that when the stress-loga
is the ratio of incremental elongation dl to existing rithmic strain diagram is a straight line, the load
length l. elongation diagram is curved upward.
In the initial stage of elongation,
dƒÃt=dl/l and ƒÃl= ?? (dl/l)=In(l/lo) ...... (6)
lim El=E (12) ??
We call ln(l/lo) the logarithmic strain ƒÃl.
In this case, too, Young's logarithmic modulus Ez
Elastic modulus Et which is calculated from El=
is equal to his ordinary modulus E.
dƒÐ/dƒÃl by using equation (6), is called Young's loga
4-3. When the decrease in sectional area is not
rithmic modulus by us. In what follows we discuss
negligible and Poisson's logarithmic ratio is constant.
the relation between Young's logarithmic modulus
Consider a wire-like material which, with load
and the load-elongation diagram.
W added, is 1 in length and r in sectional radius.
Assume that this material has then been elongated
4. Relation between Stress-Strain and Load by dl due to the increase in load by W; and that
its sectional radius has decreased by dr. Then,
elongation Diagrams
by the definition of Young's modulus and the loga
rithmic strain, we define pi, which is represented
Under this heading we discuss the load-elonga by
tion diagram when logarithmic strain and stress have -(dr/r)=
,ƒÊt(dl/l) ...... (13)
a straight-line relation, i.e., where Young's logarithmic
We call ƒÊt Poisson's logarithmic ratio. When
modulus is a constant.
integrated on the initial condition l = to at r=ro the
4.1. When decrease in sectional area is negligible.
above equation becomes
From equation (6) is obtained El = dƒÐ/dƒÃ=
r=ro(l/lo)-ƒÊl and s=ƒÎr2=so(1/1o)-2ƒÊ1
(dW/s0)/(dl/l)
By integration on the initial condition l=lo at W=O, Now the question is whether this y, exists as a

we get; constant value. Studies by F. I. Frank, A. L. Ruoff,

et al.,[31 confirm that for a large elongation, it is bet

El=(W/so)/ln(l/lo)......(7)
ter to use this Poisson's logarithmic ratio ƒÊz, rather
from which we obtain
than his ordinary ratio ƒÊ. It is recognized that,
ƒÃ =(l-lo)/lo= exp(W/soEl)-1 ...... (8)
for nylon, there is a nearly constant value of ƒÊt=

Equation (8) shows that the load-elongation dia 0.392•}0.025.

gram is a convex curve when the stress-strain diagram Now, Young's logarithmic modulus Et in this
is a straight line. case is

In the initial stage of elongation, El=(W/so)/{(l/lo)-02ƒÊ1n(1/lo)} ...... (14)

 6

and from this we obtain den.) manufactured by the Nippon Rayon Co. The
initial length of the specimens was 20 cm. The
{1+ (1-10)/10} {-1+Cl-to)lt0} 2ƒÊl =exp(W/soEl) testing machine was a continuous self-recording
...... (15) universal tensile tester[41 of the air-micrometer
type as shown in Figure 8. The tensile speed of
Equation (15) shows also that the load-elongation
the machine was set at 20 mm per minute.
diagram is convex when the stress-strain logarithmic
The results of the tests are given in Figure 1.
diagram is a straight line.
The original curve is obtained from a continuous
In the initial stage of elongation,
recorder chart. Two other curves are calculated

lim El=E ...... (16) from the original curve by reading elongation dl

per unit load 1 g on the original curve. Figure I

In this case, too, Young's logarithmic modulus Ez is shows that the curved original line can be approxi
equal to his ordinary modulus E. mately changed to a straight line by using logarith
mic strain ƒÃl and modifying the variations in the
sectional area by Poisson's logarithmic ratio ƒÊz.

5. Relation between Young's Ordinary Modu soE and soEl in the three cases cited in Figure
lus E and his Logarithmic Modulus El 1 are calculated from each equation in chapter 4.

Assume that so in each equation in the preceding

chapter is equal to so in equation (4). Then, the
relation between Et and E in the following three
cases can be obtained.
5-1. Where the decrease in sectional area is negli
gible. From equations (4) and (7):
El=({(1-10)/lo}/ln(l/lo))E ...... (17)
In this case, the two modulus differ owing to the dif
erence in the definition of strain. The ratio of the
two modulus is equal to the inverse ratio of strain,
and apparently El> E.
5-2. Where Poisson's ratio ƒÊ is constant.

From equations (4) and (10):

El-({(l-lo)/lo}
/ {1-2ƒÊ(l-l0/lo} ln(1/lo)E
......(18)
Equation (18) differs from (17), in that it includes
the effect of changes in the sectional area.
5-3. Where Poisson's logarithmic ratio is constant.
Fig. 1 Load-elongation curve
From Eqs. (4) and (14):
El=({(1-10)/l0}/(1/lo)"' ln(1/lo)lE ...... (19)
Here again equation (19) is more general than equa
tion (17), in that it includes the effect of changes in
the sectional area.

6. Some Examples of Poisson's Ratio and

Young's Modulus Obtained by Tensile Tests
of Textile Fibers

To see how far the assumptions made thus far

agree with tests, we made tensile tests of specimens. Fig. 2 Young's ordinary modulus E and Young's logarithmic
The specimens used were Nylon mono-filaments (4.2 modulus Ez
 7

So is then determined from the values 4.2 den. and 8-1. Elongation up to the breaking point
1.14 g/cm8 (fineness and density of the specimens). First, we made an experiment up to the breaking
Young's modulus E and his logarithmic modulus Ea point. The test specimen was a Nylon-6 mono
are obtained by dividing soE and soEl by so. See filament shown in Table 1 as Nylon A. Figure 5
Figure 2. shows the relation between lateral strain and longi
tudinal strain for five repeated test measurements.
All of the five lines except curve No. 3 fall within
7. Testing Apparatus for Measurement of Cross a narrow range indicating a consistent relation be

Sectional Area tween the two strains.

It has been found that the nearly straight line

relation lose their smoothness in the range of high

A yarn evenness recording tester based on the elongation of s = (‡™l/lo) •~ 100> 1113% and become

principle of an air-micrometer was altered for the zig-zag waves in shape.

measurement of the cross sectional area of test 8-2. Elongation up to 12.5%

specimen. A stretching chain was installed in the In the light of the results mentioned in (8-1)

running section of the tester for the purpose of above, we examined the phenomena up to the elon

holding the specimen. Figures 3 and 4 show the gation (‡™l/lo) •~ 100=12.5%. The specimen used was the
apparatus. same as in (8-1), but we tested it six times each at

This apparatus was designed in such a way that two different tensile speeds, 0.14 and 0.097 mm/sec.

the variations in the back pressure Pa (Fig. 3) One of the diagrams obtained is shown in Figure 6.

due to the decrease in the cross sectional area of

the specimen fiber were converted into an extention
of the bellows, which, in turn, was magnified on a
lever mechanism to operate the recording pen placed
on a drum.

Fig. 5 Longitudinal and lateral strain diagram to breaking

Fig. 3 Apparatus to measure longitudinal and lateral strain elongation

so=7.42•~1O4ƒÊ2, d„=349 p, 1,010 den.

Fig. 4 Nozzle for measuring cross sectional area of test piece

Fig. 6 Longitudinal and lateral strain diagram to 12.5

8. Experiment Concerning Diagrams of Longi elongation
tudinal and Lateral Strain su=7.42•~ 104ƒÊ2, d0=349ƒÊ, 1,010 den.
8

Table 1 Sectional diameter, area and original length of Table 2 Poisson's ratio of synthetic filaments
tested filaments

By the graph it is established again that a

longitudial and lateral strain diagrams below this
assume that the diagrams of strain below 12.5% in
level of elongation are curves bent very slightly
elongation are straight lines, although they are
downward.
actually slightly bent. By measuring the tangent
8-3. Effect of tensile speed on lateral and longi
of such straight lines, we obtain, as an average
tudinal strain diagrams
value, ƒÊ=0.385-386 as shown in Table 2.
To see the effect of tensile speed on the relation
Beyond this elongation level, the ratio of lateral
between lateral and longitudinal strain we made 28
to longitudinal strain starts to fluctuate and in
experiments at four different tensile speeds, 0.25, 0.17,
crease, and the tangent of the line which connects
0.11 and 0.08 mm/sec, and at an elongation level of
the origin and a point on the curve reaches 0.446.
12.5%-with seven experiments at each speed. The
For the purpose of comparison, we give here
specimen used is shown in Table 1 as Nylon B.
Poisson's ratio as given by F. I. Frank and A. L.
All of the 28 diagrams obtained from these tests
Ruoff, who obtained the average of 0.392 as a Poisson's
showed nearly the same tendency, thus indicating
logarithmic ratio which is defined as
that the differences in tensile speed in these ex
ƒÊl =-ln (s/so)/2 In (l/lo)
periments had no bearing on the relation between
longitudinal and lateral strains, i.e., Poisson's ratio. and at a logarithmic elongation of ƒÃl =(l/lo) •~ 100=

6%0. Usually Poisson's logarithmic ratio exceeds his

ordinary ratio obtained directly from strain values.

9. Poisson's Ratio Obtained in these Ex In the light of this, the value of Poisson's ratio ƒÊ=

periments (‡™s/2so) / (‡™l/lo) = 0.385--0.386 which we obtained

at an elongation of ƒÃ = (‡™l/lo) •~ 100=12.5%, may be

thought to agree with Frank and Ruoff's results.

We discuss in this chapter Poisson's ratio[5]

calculated from the strain diagrams in the preceding

chapter.

We assume the test specimen to be cylindrical

10. Rebuilt Testing Apparatus and Strain
in shape and define the following notations: initial Diagram
length lo; axial elongation by tension ‡™l; initial

sectional diameter do; initial sectional area so; de

crease in sectional diameter by axial elongation ‡™d;

decrease in sectional area by axial elongation ‡™s,

Then Poisson's ratio is given as :

ƒÊ= (‡™d/do)/(‡™l/lo)

Therefore, ƒÊ=(‡™ /so) /(‡™l/lo)

Poisson's ratio in each experiment referred to in

chapter 8 was calculated by the equation shown

above. The value of ƒÊ thus obtained are shown

in Table 2.

Comparing Table 2 with the diagrams in Figures

6-8, we can grasp more easily the strain accompany Fig. 7 Apparatus for recording longitudinal load elongation
ing elongation. In calculating Poisson's ratio, we and lateral strain curve
 9

A feeling nozzle (Fig. 9) to measure the sectional The longitudinal and lateral strains were calculated
area of the specimen was attached to the universal from a recorded diagram and was plotted into diag
tensile tester (Figs. 7 and 8) used in our experiment rams as shown in Figure 10.
referred to in chapter 6, so that the load and longi
As to the results of the experiment with Nylon
tudinal and lateral strains could be masured and
6. The diagrams are approximately straight lines,
recorded simultaneously.
though they are slightly curved downward toward
Since the originally built nozzle (Fig. 4) for
the left end.
measuring the sectional area was too short, fluctu
Because the nozzle for feeling sectional area
ations in diagrams were produced by partial deform
was made longer in the rebuilt apparatus, the small
ations of the specimen. Therefore, we substituted
fluctuations noticed in the experiments referred to
a long feeling nozzle (Fig. 9) which would draw
in chapter 8, which were due to partial deformation
smooth curves.
of the specimens, disappeared and the diagrams
The specimens used in the experiments using
were smooth in the whole range up to a high elon
the long nozzle were of three kinds: Nylon, Envilon
gation percentage. The experiments with Envilon
(PVC fiber) and Polyethylene filaments (Table 1). and Polyethylene filaments showed the same results.
We experimented at five tensile speeds in a range
of 0.1612.00 mm/sec (Table 2).

Fig. 10 Longitudinal and lateral strain diagram of Nylon

filaments

do= 308ƒÊ, so= 7.12•~104ƒÊ2

Fig. 8 Recording apparatus

5: Nozzle for measuring sectional area

6: Apparatus for measuring tensile load 11. Calculating Poisson's Logarithmic Ratio
7: Recording drum with Rebuilt Testing Apparatus
8: Apparatus for measuring sectional area

As explained in chapters 4 and 9, Poisson's ratio

for each diagram was obtained from equation (13):

ƒÊl =-(1/2)ln(s/so)/ln(1/1o)

as Poisson's logarithmic ratio.

Poisson's logarithmic ratios in various ranges of

elongation of Nylon-6 are given in Table 3. The

higher the tensile speed, the lower the Poisson's ratio,

as this table shows. This tendency is expressed

in graphs in Figures 11-12.

The fact that Poisson's ratio decreases with an

increase in elongation is presumably because, as

J. Kaufman and W. George have said,[61 a Nylon

Fig. 9 Nozzle for measuring cross sectional area filament stretched longitudinally shrinks laterally at

first, but with the progress of elongation the mole

From left: 1: for Nylon filaments; 2: for Envilon and
Polyethylene filaments; 3: for Partial deformation; 4: cules of the fiber approach one another and thereafter

Nozzle in Fig. 4 a slip takes place only in the longitudinal direction

 10

with no more change in the lateral distance be 12. Young's Logarithmic Modulus and his
tween molecules. Ordinary Modulus
The decrease in Poisson's ratio with an increase
in tensile speed is presumably because of a defor The ratio of stress a and strain a in the stress.
mation of that part of a specimen which is outside strain diagrams is Young's modulus by secant mod
the feeling range of the measuring nozzle-partial ulus. Young's logarithmic modulus corresponding
deformation near the clamped point of the specimen to the broken line in Figure 13 is obtainable thus:
where stress is too large-and because of a slight El (W/s)/ln(l/lo)
time delay in the air-micrometer circuit.
Young's ordinary modulus corresponding to the solid
Poisson's logarithmic ratios calculated on Envilon
line Figure 13 is obtainable thus :
and Polyethylene are in Table 3.
E= (W/so)/ {(l-lo)/lo}

As reported in chapter 5, there is, between these

two formulas, this relation (Eq. 19): El= {(1-10)/l0}

/{(l-lo)_2ƒÊ1 ln(l/lo)}E where Poisson's logarithmic

ratio j u, constant and El> E.
Although stress-strain diagrams are slightly
curved (as in Fig. 13), the gradients of straight lines
connecting the points in the curve and the origin
may be used to calculate Young's logarithmic modulus
EL and his ordinary modulus E (see Table 4).

Fig. 11 Relation between longitudinal strain level and

Poisson's ratio of Nylon filaments

Fig. 12 Relation between stretching speed and Poisson's

ratio of Nylon filaments Fig. 13 Stress-strain diagram of Nylon filaments

Table 3 Poisson's logarithmic ratio of tested filaments

 11

Both values of Young's modulus are large at In the light of the curves in Figure 14, we may
the initial point of elongation and decrease with the say that Young's modulus EE decreases less severely
progress of elongation. The decrease in Young's with strain than does his ordinary modulus E.
logarithmic modulus El is smaller than the decrease That is to say, our experiments have proved that
in his ordinary modulus E, as is clear from Figure Young's logarithmic modulus can be approximated
14. by a constant in a wider range than his ordinary
In Figure 14 Young's logarithmic modulus and modulus can.
his ordinary modulus are plotted against logarithmic
strain as the ratio over their respective value at 5%
strain. 13. Conclusions

The experiments described in the foregoing article

have made it clear that:
13-1. The fact that the load-elongation diagrams
of textile fibers do not make straight lines has
generally been considered a problem bearing on the
visco-elastical properties of the internal structure.
Our experiment has shown that there is another
reason for load-elongation diagrams of textile fibers
not making straight lines. This reason lies in the
way stress-strain and Young's modulus have been
traditionally defined.
The traditional definition of Young's modulus
is all right if strain is small, but where strain is
large, it is better to use logarithmic strains because
they are compatible with a wider range of strains.
13-2. It is not always correct to say Young's
modulus are not constant just because load-elongation
diagrams do not make straight lines. The reason
Fig. 14 Decrease in Young's modulus in Nylon filaments for non-straight lines lies not only in plasticity but
with elongation also in the properties of elasticity. We have found,

Table 4 Young's ordinary and logarithmic modulus (E and Et)

 12

too, that load-elongation diagrams do not necessarily in the measuring nozzle and Poisson's ratio declines
make straight lines even if the stress-strain relation 13-6. By taking the variations in the sectional
is linear and the properties of elasticity are constant. area, into account, stress-strain diagrams can be
13-3. By applying the air-micrometer principle straightened in a wider range than the load-elon
in measuring the sectional area of a test specimen, gation diagram, and Young's logarithmic modulus
the decrease in the sectional area and the elonga can be approximated by a constant in a wider range
tion of a textile fiber can be measured simultane of strain and stress than his ordinary modulus can.
ously. A long nozzle for the measurement of the
sectional area eliminates minor fluctuations in the
recording diagram and makes it smooth.
References cited
13-4. The ratio of increase in longitudinal and
lateral strains, i.e., Poisson's ratio, is 0.30-0.41 for
Nylon-6, 0.210.31 for Envilon (PVC filaments) and [1] S. Watanabe; J. Textile Machinery Soc., Japan, 12, 163
0.33-0.37 for Polyethylene filaments, on the average . (1959). S. Yamaguchi; J. Textile Machinery Soc., Japan
In Nylon and Polyethylene Poisson's ratio declines 10, 116 (1957). 10, 348 (1957) and 10, 786 (1957).
as elongation progresses. The reverse is true with [2] S. Timoshenko; Strength of Materials. 3rd. Ed. Part
Envilon. 1, Norstrand, New York (1953).
13-5. With an increase in tensile speed and , theref [3] F. I. Frank and A. L. Ruoff; Textile Rec. J., 3, 902 (1958)
ore, an increase in load speed , a partial defor [4] K. Higuchi, K. Kawada and M. Tanaka; J. Soc. Textile
and Cellulose Industry, Japan, 13, 225 (1957) .
mation of the specimen made in or near the clamped
[5] M. Reiner; Deformation and Flow, Lewis, London (1949).
position where the stress is heavy starts to remain [61 J. Kaufman and W. George; J. Appl. Phys., 21 (1950).