Fibres and Textiles 29(4), 2022, 3-8 DOI: 10.

15240/tul/008/2022-4-001

TEAR AND TENSILE STRENGTH OF 100% COTTON WOVEN

FABRICS’ BASIC STRUCTURES: REGRESSION MODELLING

HOSSAIN MOHAMMAD MOBARAK1*, ALIMUZZAMAN SHAH2, AHMED DEWAN MURSHED2

1
Department of Textile Engineering, Jashore University of Science and Technology (JUST), Jashore, Bangladesh
2
Department of Fabric Engineering, Bangladesh University of Textiles (BUTEX), Dhaka, Bangladesh

ABSTRACT
This research paper aims to estimate the tear and tensile strength of woven fabrics while considering a
number of construction factors. Construction variables include ends per cm (EPCm), picks per cm (PPCm),
an overall configuration of yarn, and fabric’s areal density or grams per square meter (GSM). While the
statistical relationship in deciding the fabric strength is very complicated considering all variables, the
correlation-regression model is used to explain the influence of structural parameters on the tear and tensile
strength of various fundamental fabrics’ designs. With different thread densities varying reed counts, and
heald count using 100 percent cotton yarn having 36.9 tex, eight different designs of plain, twill, and sateen
are prepared for the study. Four regression models, built to predict the tear and tensile strength of the sample
woven fabrics, are vital components of this research. It is noticed that the setting of yarn affects the tensile
strength of the fabrics, and the fabric pattern determines the tear strength of the fabrics. For higher tear
strength, matt weave, and tensile strength, a twill structure is desired within this scope of the fabric structures.
KEYWORDS
Tear; Tensile; Strength; Woven; Regression; Modelling, Prediction.

INTRODUCTION the parameters before manufacturing fabrics to

ensure the tremendous loss of supplies, time,
Strength is the first property that has the most electricity, labor, and money [2-6].
significant effect when choosing the necessary
fabrics to produce clothing or apparel. These are Different fabric manufacturing processes, such as
based mainly on the expected end-use [1]. The spinning, knitting, non-woven, and braiding, are
tensile and tear behavior of the fabric depends not available [7]. Strong dimensional stability and good
only on the strength of the yarn alone but also on cover are seen between them by woven fabric.
other variables, including the use of fiber or blend Strength is one of the most significant features of
form, twist amount, twist angle, yarn count, spinning woven fabric [8]. In stretching a test piece to a
systems, yarn bending behavior, frictional properties, breaking point, the highest tensile force measured is
interlacement pattern, fabric construction called tensile strength or breaking strength. Tearing
parameters, series of warp and weft, finishing intensity is the average force needed to continue a
treatment etc. [2]. The geometry of the fabric, thread tear already begun in a fabric [9]. During the tearing
density, and weaving design also have a significant test, threads break singly or in small clusters. Threads
effect on the strength of the fabrics. The strength also that have been twisted skew and slide. The strain is
be influenced based on the production state during first carried by a few strands. The adjacent yarn also
wet processing and finishing treatment [3-5]. Also, fails. Yarns may cluster around the tear due to
test conditions such as temperature, humidity, increased extensibility or fewer frictional restrictions,
loading time, loading quantity, jaw distance, and however yarn strength may not improve tear strength.
measuring methods often influence the difference in Again, breaking force is determined by area. Tensile
the intensity of the fabric value. For all these force at the point of rupture is referred to as tenacity.
controllable and uncontrollable variables relevant to Tensile strength of fabrics is affected by yarn density
fiber, yarn, and fabric during production cycles, and cross-section of fabrics itself and yarn from which
creating a simple direct quantitative connection it is composed of. Thread density and oblique yarn
between the yarn's intensity and the fabric's orientation need more efforts to balance the weight.
corresponding strength is quite complicated. In this Crimping causes yarn to stretch. More crimp extends
respect, it is crucial to identify the right ways to decide the cloth. Floats increase thread density while

*
Corresponding author: Hossain M.M., e-mail: [email protected]
Received October 8, 2022; accepted February 3, 2023
 Fibres and Textiles 29(4), 2022, 3-8

decreasing extension. Weaving decreases fabric • To explore the effect of loom settings on output
extensibility by tightening the warp and weft and variables warp way (Y1) and weft way (Y2) tear
tensile strength fell [10]. strength and warp way (Y3) and weft way (Y4) tensile
strength.
These two strengths (tear and tensile) had different
application in the textile field to produce fabrics • To explore the effect of design types on output
considering wear and tear resistance and produce variables warp way (Y1) and weft way (Y2) tear
higher tensile strength in technical textiles strength and warp way (Y3) and weft way (Y4) tensile
manufacturing. However different research strength.
approached are done in predicting these strength
• To investigate the effect of “EPCm, PPCm,
properties in different way. Multiple linear regression
Law’s maximum thread density and GSM with output
(MLR), artificial neural network (ANN), Automated
variables warp way (Y1) and weft way (Y2) tear
Machine Learning (AutoML), and Fuzzy techniques
strength and warp way (Y3) and weft way (Y4) tensile
are some preferred tools that have been used to
strength.
predict the strength of the fabric of composite fabric
considering different variations in count, thread Due to the difficulty of integrating all those variables
density (warp and weft), inter- yarn friction, float to represent the configuration of the weave, the
length, interlacing points, design (plain, twill, matt), correlation-regression tool and ANOVA are used to
loading direction (uniaxial or biaxial), fiber explain the influence of structural parameters of
composition (single or blends), fiber composition fabrics on the tear and tensile strength of fabrics with
(single/blends), spinning (ring/rotor), yarn strength various fundamental designs. Regression tools are
transfer efficiency etc. [11-14]. When utilizing hidden used to evaluate study theories that have been
layers in prediction models, machine learning established from literature reviews [18]. And after
approach outperformed regression method in checking approaches [19], the impact of yarn
prediction using learnt data [15]. However, the input environment and design styles on the intensity of
parameters consideration and design variation fabrics is observed. Finally, predictors such as
cannot be included as a whole in modeling for PPCm, EPCm, GSM, and Law's maximum yarn
incapability in numerical or weighted value setting of the sample woven fabrics are provided by
expression. Because various weave designs might four regression models. Here following effects in
have the same float length or yarn densities [10]. alternate hypothesis statements are analyzed
statistically-
To know the effect of design variation (basic plain,
twill and sateen) on tensile and tear strength of woven H1-H8: There is a significant difference in yield of
fabrics having similar loom setting and yarn warp way tear strength (Y1), weft way tear strength
properties within the smallest repeat sizes, this (Y2), warp way tensile strength (Y3), and weft way
experiement is performed. Although machine tensile strength (Y4) for different loom settings
learning approach is better this experiment only used (H1,H2, H3, H4 respectively) and similarly across
multiple linear regression tool primarity to investigate design types (H5, H6, H7, H8 respectively).
the found result from testing the sample fabrics
produced in this study with the relation of considered H9-H12: EPCm (H9), PPCm (H10), Law’s max yarn
set (H11), and GSM (H12) have significant effect on
parameters. Here, the primary objective of this
research is to forecast woven fabrics considering warp way tear strength (Y1).
various construction variables. Ends per cm (EPCm), H13-H16: EPCm (H13), PPCm (H14), Law’s max
picks per cm (PPCm), maximum yarn setting, and yarn set (H15), and GSM (H16) have considerable
cloth areal density in gram per square meter (GSM) impact on weft way tear strength (Y2).
are considered to be primary factors. At the same
time, numerous researchers have established H17-H20: EPCm (H17), PPCm (H18), Law’s max
different fabric weaving factors [10, 16]. But there are yarn set (H19), and GSM (H20) have significant effect
too many considerations linked to the strength of the on warp way tensile strength (Y3).
fabric, such as yarn count, twist, fiber fineness, H21-H24: EPCm (H21), PPCm (H22), Law’s max
rigidity, fiber density, the shape of the fabrics, cover, yarn set (H23), and GSM (H24) have considerable
yarn density, layer number, tightness factor, and so impact on weft way tensile strength (Y4).
on [17].
The objectives of the study are to develop four EXPERIMENTAL METHODOLOGY
regression models to predict both warp and weft way
tear strength as well as warp and weft tensile strength
Materials
of woven fabrics having eight different weave 36.9 Tex, 100% cotton rotor yarn is used to
structures of plain, twill, and sateen. The objectives of manufacture woven fabrics. It has a lea strength of
the study are mentioned below- 81.93 kg at 25 0C with 78% relative humidity in the
testing lab. Single yarn strength is 5.6 N, with 6.9%
 HOSSAIN M.M., ET AL.: TEAR AND TENSILE STRENGTH OF 100 % COTTON WOVEN FABRICS’ BASIC STRUCTURES: REGRESSION MODELLING

breaking extension and tenacity is 15 cN/Tex. As the a dent (40 heald count) and 60 Stockport reed
interlacement ratio of plain, twill, and satin fabric drawing 1 in a dent (30 heald count), and 2 in a dent
differs in various way, and their stiffness are affected (60 heald count) in Automated Sampling Rapier Loom
by the angle of weaving, in this experiment, fabrics (CCI Tech Inc.). For collecting tear (5×2×24=240) and
𝟏 tensile strength (5×2×24 =240) in both warp and weft
structures for eight woven designs, namely
𝟑 direction “240 + 240 =480” specimens are produced,
𝟐 𝟑 𝟏 𝟐 tested and average values are recorded in Table 2.
sateen, z-twill, z-twill, plain, plain warp rib,
𝟐 𝟏 𝟏 𝟐
𝟐 𝟑 Instruments
𝟐 plain regular matt, plain warp rib and
𝟐 𝟏
𝟑
𝟑 𝟏 plain irregular matt are prepared. For data Lea strength tester, single yarn tester (Pytan),
𝟏 Automated Sampling Rapier Loom (CCI Tech Inc.),
processing in the statistical tools following universal tensile tester (Titan 500 Newton) and Elma
descriptions are used to represent types of fabrics tear tester (133 N), wrap reel and weight balance
(Fig. 1). (yarn count test method), and twist tester are used in
this experiment. All data is recorded in a single
working sheet then the relation among the factors of
fabrics is analyzed using a graphical method, SPSS
23, and R-Studio software.
Standards and methods for testing
Single yarn strength, fabrics’ tear, and tensile
strength are tested as ASTM D2256/D 2256M:2010
[20], Elma Tear with max load 133 N as ISO 13937-
1:2000 [21], and Universal Tensile Tester as ISO
13934-2:2014 [22] respectively. 5 breaks are
performed both for warp and weft direction with
similar setting and speed. The machine speed of
tensile testing is 50 mm/min. After enzymatic
Figure 1. Fabrics’ design repeats used in this experiment. desizing, washing for 25 min at 50 0C temperature
Finally, 8 (Designs)×3 (Loom Setting) = 24 samples with 1 ml detergent, the samples were dried and
are being tested according to the testing methods. To relaxed for 72 hours before testing.
produce these fabrics, 40 Stockport reed drawing 2 in

Figure 2. Fabric tear strength testing specimen size and testing [21].

Figure 3. Fabric tensile strength testing specimen size and testing [22]
 Fibres and Textiles 29(4), 2022, 3-8

Table 1. Some symbols and their meanings in this experiment.

𝑌 Warp way tear strength (Y1), laws S(yN)1/2 F t-test value
Weft way tear strength (Y2), Where, S = setting ratio
Warp way tensile strength (Y3), varying with weave upto 4
Weft way tensile strength (Y4) float, y = cloth setting
constant depends on yarn
numbering system, N indirect
count of yarn
𝛽 Constant for regression models gsm Gram per squar meter (GSM) p Probability value

𝜖 Error term R and Reliability of the beta values H1-H24 Research Hypothesis
Adjusted
2
R
EPCm Ends per centimeter W Shapiro-Wilk values
PPCm Picks per centimeter df Degree of freedom

indicates the model’s significance and significant

Variables individual effect of each predictor variable. For
Dependent variables are warp way (Y1) and weft way four response variables, four separate regression
(Y2) tear strength and warp way (Y3) and weft way models were developed:
(Y4) tensile strength of the fabrics. Again, 𝑌 𝛽 𝛽 𝐸𝑃𝐶𝑚 𝛽 𝑃𝑃𝐶𝑚
independent variables are EPCm, PPCm, GSM, and 𝛽 Law s maximum yarn setting (1)
𝟏 𝟐 𝟐 𝛽 GSM 𝜖 ; 𝑖 1,2,3,4
Law’s max yarn set. plain, plain warp rib, 𝟐
𝟏 𝟐 𝟐
𝟑 𝟑 where in each model, the error term ϵ is assumed to
plain regular matt, plain warp rib, 𝟑 𝟏 plain be normally distributed with mean 0 and constant
𝟏 𝟏
𝟐 𝟑 𝟏 variance 𝛔𝟐𝐞 .
irregular matt, z-twill, z-twill, and sateen
𝟐 𝟏 𝟑
structure of fabric 1/1P, 2/2P, 2/2(2)P, 3/1 P,
3/1(3+1)P, 2/2T, 3/1T, and 1/3S are used respectively
RESULTS AND DISCUSSION
for software processing in R-studio 3.6 and SPSS 23. In this section effect of the design or changing of
Some symbols and their meaning are enlisted in fabrics’ structures and loom settings on tear and
Table 1. tensile strength are tested by two-way ANOVA. Then,
by regression analysis four specific models are
Statistical methods prepared using EPCm, PPCm, Law’s maximum yarn
Following statistical analyses were conducted setting, and GSM as influencing factors of the fabrics
throughout this study: where other factors assumed constant during
production of the fabrics samples. Considering
1. The level of significance for all hypotheses was constant yarn properties like yarn count, fiber
taken as α=5%. A p-value less than α indicates the consumptions, twist per inch the loom setting
test's significance (i.e., reject the null hypothesis). changed within Stockport reed count 30 with 1 and 2
2. Shapiro -Wilk test was performed to check yarn per dent, and Stockport reed count 40 with single
whether response/output variables and residuals yarn drawn per dent. The data set is summarized in
of the regression model are normally distributed or Table 2.
not.
The hypothesis statement of the Shapiro -Wilk test is Two-way ANOVA of Strength (Y1-Y4) for
given below: loom setting and design name
H0: Variable or data is normally distributed
A two-way ANOVA was conducted to determine if
Ha: Variable or data is NOT normally distributed
there is a difference in average “OUTPUT” with loom
3. Two-way ANOVA was conducted to understand setting and design type (derivatives of plain, twill and
the effect of fabric setting and design type on sateen) which is mentioned as ‘design_name’
output variables. A p-value less than α indicates a variable here from Table 3 to Table 6. Since there are
significant impact. four output variables warp way tear strength (Y1),
4. Multiple linear regression techniques were applied weft way tear strength (Y2), warp way tensile strength
to assess the effect of five predictor variables like (Y3), and weft way tensile strength (Y4) in this study,
EPCm, PPCm, Laws maximum yarn setting, and so four two-way ANOVA have been performed
GSM on output variables. A p-value less than α separately.
 HOSSAIN M.M., ET AL.: TEAR AND TENSILE STRENGTH OF 100 % COTTON WOVEN FABRICS’ BASIC STRUCTURES: REGRESSION MODELLING

Table 2. Data summary for tear and tensile strength in warp and weft way due to different design and loom setting

Law’s Tear Strength (N)* Tensile Strength (N)*

Setting
Loom

Max
Design Name EPCm PPCm GSM
Thread Warp
Density Weft Way Warp way Weft Way
way
Plain 15 14 22 132.47 56.390 64.011 158.678 159.551
Reed Count (Stock Port) 60 and Heald

Plain
14 14 30 128.95 92.132 78.373 144.548 130.681
Warp Rib
(2) Plain
16 13 30 135.49 100.950 103.010 141.525 111.855
Regular Matt
Count 30

Plain
15 14 33 143.98 62.296 76.838 155.143 115.397
Warp Rib
(3+1) Plain
17 14 33 147.29 75.424 70.724 159.277 104.566
Irregular Matt
Z-twill 16 14 30 143.28 80.386 88.242 142.465 132.514

Z-twill 15 14 33 134.27 84.252 88.754 138.579 123.679

Sateen 17 14 33 146.97 81.405 92.152 142.964 136.200

Plain 19 17 22 158.68 39.681 48.406 252.439 239.639
Reed Count (Stock Port) 40 and

Plain
21 18 30 179.08 105.640 85.288 238.676 187.382
Warp Rib
(2) Plain
20 17 30 169.26 106.390 102.505 209.917 194.409
Heald Count 40

Regular Matt
Plain Warp Rib 20 18 33 181.52 108.470 84.684 242.757 184.091
(3+1) Plain
21 18 33 178.73 93.060 95.304 209.777 184.762
Irregular Matt
Z-twill 20 18 30 171.35 96.719 100.558 233.029 208.370

Z-twill 20 18 33 179.17 83.534 90.806 230.675 216.376

Sateen 21 18 33 183.73 72.418 85.810 207.945 238.204

Plain 25 24 22 218.11 18.858 32.798 405.833 338.846

Reed Count (Stock Port) 60 and Heald

Plain
26 25 30 219.36 63.872 56.667 428.512 341.938
Warp Rib
(2) Plain
25 25 30 210.38 100.408 104.193 371.480 333.181
Regular Matt
Count 60

Plain
26 25 33 215.67 122.200 64.117 388.311 330.500
Warp Rib
(3+1) Plain
27 26 33 220.32 132.810 129.400 357.948 357.381
Irregular Matt
Z-twill 26 26 30 224.45 59.613 72.977 392.879 366.607

Z-twill 26 25 33 223.05 47.421 48.136 319.733 440.507

Sateen 27 26 33 229.65 41.407 71.366 374.405 360.847

Note:* mean values are taken from 5 tests.

Table 3 indicates that the loom setting's main effect is From Table 4 it is seen that as weft way tear strength
insignificant, F(2,14)= 0.92, p=0.42, which means that (Y2) the main effect of loom setting is not significant,
the average warp way tear strength (Y1) is almost the F(2,14)=1.59, p=0.24, implying that average weft way
same for the three loom setting. The main effect of tear strength (Y2) does not differ significantly across
the fabric design is significant, F(7,14)=2.89, p=0.04; loom setting. Whereas the main effect of the fabric
that is average warp way tear strength (Y1) differs design is significant, F(7,14)=3.20, p=0.03. So, it can
significantly from design to design, and the highest be said that design has an impact on weft way tear
𝟐
average of warp way tear strength (Y1) was yielded strength (Y2) and design 𝟐 plain matt had the
𝟐
𝟐
for design 𝟐 plain matt. highest average yield.
𝟐
 Fibres and Textiles 29(4), 2022, 3-8

Table 3. Response variable warp way tear strength (Y1). tensile strength change. The more yarn present in the
Source of
Df
Sum Mean
F p fabric tensile strength will be high. Again, whether the
variation Square Square number of yarns more or less design variation
Loom Setting 2 903.99 451.99 0.92 0.42 influences its tear strength. The more floats of yarn
design_name 7 9963.02 1423.29 2.89 0.04 are present in the fabrics, the more the tear strength.
Residuals 14 6890.9 492.21 It also supports the result found by Eryuruk S.H. and
Total 23 17757.9
Kalaoğlu F. [23] that weft tearing strength values
found greater than the warp tearing strength and
Table 4. Response variable weft way tear strength (Y2). change in tear strength as the number of threads per
Source of Sum Mean meter in either direction. Again the thread densities
Df F p
variation Square Square improve the tensile strength [24-25].
Loom Setting 2 862.7 431.34 1.59 0.24
As a result of the inability to present fabric design in
design_name 7 6064.3 866.33 3.20 0.03 numerical numbers such as EPCm, PPCm, and
Residuals 14 3791.0 270.79 GSM, Law's maximum thread density in ‘table 1’ is the
term used in regression models due to its
Total 23 10718.0
consideration of fabric design in its formula to limit the
Table 5. Response variable warp way tensile strength (Y3). maximum yarn density for specific yarn count and
Source of Sum Mean fabric design. A drawback of the research is that the
Df F p
variation Square Square fundamental design and its derivatives cannot be
Loom Setting 2 222090.7 111045.3
316.
0.00 separated with distinct justified numerical values to
9
consider in the regression models which is a scope of
design_name 7 5084.9 726.4 2.07 0.12
further research. Thus, design parameters and their
Residuals 14 4906.3 350.5 numerical expression needs to be further analyzed.
Total 23 232081 Law’s max yarn set is the term that consider both the
yarn representing unit and basic fabric design
Table 6. Response variable weft way tensile strength (Y4). variation that improves the model prediction strength
Source of Sum Mean in this regard [10].
Df F p
variation Square Square
Loom Setting 2 222103 111052
210.
0.00
Regression Models with ANOVA
45
design_name 7 7252 1036 1.96 0.13
Regression Model 1: Dependent variable warp way
tear strength (Y1)
Residuals 14 7388 528
Y 16.367 20.629 ∗ EPCm 2.46 ∗
Total 23 236743 PPCm 4.695 ∗ laws 3.095 ∗ gsm (A)

In the same procedure, two-way ANOVA was carried Regression Model 2: Dependent variable weft way
out, and the result is displayed in Table 5. The main tear strength (Y2)
effect of the loom setting is very significant on the
yield of Y3, F(2,14)=316.9, p=0.00. The highest Y 33.258 13.816 ∗ EPCm 0.754 ∗
PPCm 3.509 ∗ laws 2.015 ∗ gsm (B)
average of warp way tensile strength (Y3) is obtained
for loom setting 2. On the other hand, the fabric
design effect is insignificant; the average yield of warp Regression Model 3: Dependent variable warp way
way tensile strength (Y3) is almost the same for eight tensile strength (Y3)
designs with same loom setting, F(7,14)=2.07, Y 20.290 2.102 ∗ EPCm 16.877 ∗
p=0.12. PPCm 4.549 ∗ laws 0.755 ∗ gsm (C)

Lastly, Table 6 depicts that loom setting has a very

significant impact on an average yield of weft way Regression Model 4: Dependent variable weft way
tensile strength (Y4), F(2,14)=210.45, p=0.00, and tensile strength (Y4)
the highest average yield of weft way tensile strength Y 84.04 1.665 ∗ EPCm 18.872 ∗
(Y4) is obtained for loom setting 2. The fabric design PPCm 2.729 ∗ laws 0.419 ∗ gsm (D)
has no significant effect on an average yield of weft
way tensile strength (Y4), F(7,14)=1.96, p=0.13. Regression Model 5: Dependent variable log weft
way tensile strength (log(Y4))
Thus, from Table 3 to Table 6 hypotheses H3, H4, H5
and H6 are accepted. On the other hand, hypotheses log Y 1.753 0.006 ∗ EPCm 0.022 ∗
H1, H2, H7, and H8 are rejected. Thus it is proved PPCm 0.009 ∗ laws 0.002 ∗ gsm (E)
that loom setting variation in yarn setting influences
the tensile strength of fabrics and design variation From Table 7, the overall impact of independent
influences the tear strength of the fabrics. Thus, an variables significantly impacts warp way tear strength
increase in the yarn setting greatly affects the fabric's (Y1). However, the estimated model can explain a
 HOSSAIN M.M., ET AL.: TEAR AND TENSILE STRENGTH OF 100 % COTTON WOVEN FABRICS’ BASIC STRUCTURES: REGRESSION MODELLING

Table 7. ANOVA for regression models.

ANOVA for
Source of variation Df Sum Square Mean Square F value Model Summary
regression
Regression 4 8729.966 2182.491 4.593
R=0.701, Adjusted R2=0.385,
Model 1 of Y1
Residuals 19 9027.946 475.155 W=1.880, p-value=0.009
(A)
Total 23 17757.912
Regression 4 4813.391 1203.348 3.872
Model 2 of Y2 R=0.670, Adjusted R2= 0.333,
Residuals 19 5904.616 310.769
(B) W=2.238, p-value=0.018
Total 23 10718.007
Regression 4 223540.839 55885.210 124.32
Model 3 of Y3 R=0.981, Adjusted R2=0.955,
Residuals 19 854.991 449.526
(C) W=2.036, p-value=0.000
Total 23 232081.830
Regression 4 222607.584 55651.896 74.804
Model 4 of Y4 Residuals 19 14135.373 743.967 R=0.970, Adjusted R2 =0.928,
(D) W=1.066, p-value = 0.000
236742.957
Total 23

Regression 4 0.821 0.205 75.45

Model 5 of Y5 R=0.97, Adjusted R2= 0.928,
Residuals 19 0.052 0.003
(E) W = 1.427, p-value = 0.000
Total 23 0.873

79.7% variation in warp way tear strength (Y1). The p=0.002. The effect of GSM on warp way tear
model residuals are also normally distributed, strength (Y1) is negative and highly significant,
W=1.880, p=0.009, which indicates the excellent fit of t(19)=-2.892 p=0.009. Moreover, PPCm has positive
the model. impact but is not statistically significant. Thus,
hypotheses H9, H11, and H12 are accepted, whereas
Again, The overall model 2 (model B) is significant,
F(4,19)=4.579, p=0.009. The estimated model can H10 is rejected. But their overall effect made the
model satisfactory with the goodness of fit.
explain a 70.1% variation of weft way tear strength
(Y2). The model residuals are also normally Again, EPCm, Law's maximum yarn setting, and
distributed, indicating a well-fitted model, W=2.238, GSM significantly affect weft way tear strength (Y2)
p=0.018. (p=0.048, 0.004, 0.031, respectively). Thus,
On the other hand, to investigate the overall impact of hypotheses H13, H15, and H16 are accepted,
whereas H14 is rejected. But their overall effect made
EPCm, PPCm, Law's maximum yarn setting, GSM on
warp way tensile strength (Y3); model 3(C) has been the model satisfactory with the goodness of fit.
fitted. The fitted model is highly significant, On the other hand, it also indicates that PPCm has a
F(4,19)=124.32, p=0.000. The R=0.981 implies that significant positive effect, and Law's maximum yarn
the fitted model can predict a 98.1% variation in warp setting significantly negatively impacts warp way
way tensile strength (Y3) . The model's residual (C) is tensile strength (Y3) where p=0.003, p=0.003,
normally distributed, W=2.036, p=0.000. respectively. Other variables’ impacts are not
Though, model 4 (D) is overall significant, significant. Thus, hypotheses H18 and H19 are
accepted, whereas H17 and H20 are rejected. But
F(4,19)=74.804, p=0.000, but the residuals are not
normally distributed, W=1.066, p=0.000. This may be their overall effect made the model satisfactory with
the goodness of fit.
due to functional misspecification of the model. So,
another model is developed taking the logarithm of The interpretation of the estimated coefficient of the
weft way tensile strength (Y4). Now, semi-log model semi-log model differs from the linear model. Table 8
5 (Model E) is highly significant, F(4,19)=75.45, implies that the yield of weft way tensile strength (Y4)
p=0.000 and the model’s residuals are also normally increases at a rate of 0.022 for a unit of PPCm, or
distributed, W=1.427, p=0.000. The estimated model 2.2%. Interpretation of other coefficients is in the
can explain 97.0% variation in weft way tensile same manner. Again, the yield of weft way tensile
strength (Y4). strength (Y4) decreases at a rate of 0.009 for a unit
of Law’s max yarn setting or 0.9%. But they have a
Table 8 indicates that EPCm has a positive and very
significant effect on warp way tear strength (Y1), very significant effect on the yield of weft way tensile
t(19)=2.547, p=0.020. The impact of the law's strength (Y4). Thus, hypotheses H22 and H23 are
accepted, whereas hypotheses H21 and H24 are
maximum yarn setting on warp way tear strength (Y1)
is positive and highly significant, t(19)=3.490, rejected. But their overall effect made the model
satisfactory with the goodness of fit.
 Fibres and Textiles 29(4), 2022, 3-8

Table 8. Regression coefficients for models.

Regression coefficients Variables/predictor Estimate Std. Error t value Pr(>|t|)
(Intercept) 16.367 46.293 0.354 0.728
EPCm 20.629 8.101 2.547 0.020**
Model Y1 (A) PPCm 2.460 5.037 0.488 0.631
laws_max_yarn_set 4.695 1.345 3.490 0.002***
gsm -3.096 1.071 -2.892 0.009***
(Intercept) 33.258 37.438 0.888 0.385
EPCm 13.816 6.551 2.109 0.048**
Model Y2 (B)
PPCm 0.754 4.073 0.185 0.855
laws_max_yarn_set 3.509 1.088 3.225 0.004***
gsm -2.015 0.866 -2.327 0.031**
(Intercept) -20.290 45.027 -0.451 0.657
EPCm -2.102 7.879 -0.267 0.793
Model Y3 (C) PPCm 16.877 4.899 3.445 0.003***
laws_max_yarn_set -4.549 1.308 -3.477 0.003***
gsm 0.755 1.041 0.725 0.478
(Intercept) -84.040 57.926 -1.451 0.163
EPCm -1.665 10.137 -0.164 0.871
Model Y4 (D) PPCm 18.872 6.302 2.994 .007***
laws_max_yarn_set -2.729 1.683 -1.621 0.121
gsm 0.419 1.340 0.312 0.758
(Intercept) 1.753 .111 15.823 .000
EPCm .006 .019 .319 .753
Model Y5 (E) PPCm .022 .012 1.805 .087*
laws_max_yarn_set -.009 .003 -2.713 .014**
gsm .002 .003 .647 .525
Note:*p<0.1; **p<0.05; ***p<0.01
[26] due to the change is weave design of plain
Fractional changes of strength based on structure.
design
For different heald counts, it is seen that rib, matt,
According to float length, it can be said that the float twill, and sateen showed greater weft way tear
length of the fabric significantly affects tear strength. strength than plain fabric. And the strength increases
With the increased float length, more yarns are close with the increase of heald counts. The tear strength
to each other, thus increasing tear strength. But, it is of matt fabric is higher than that of rib, twill, and
not clearly said from the data the effect of change in sateen, with a similar float length (Fig. 5). These also
float length how the tensile strength changes as it supported by result found in work of Asaduzzaman M.
didn’t show any clear indication due to the fluctuation et al. and Malik Z.A. et al. [25-26]. Thus contact
of the tensile strength much with the change of the between warp and weft yarn or interlacement pattern
design of the fabrics and yarn set [23-25]. or variation in weave design plays a vital role in warp
and weft tear trength which is also supported by
For distinct heald counts, rib, matt, twill, and sateen
Radwan S.S. [27].
show greater warp way tear strength than plain fabric
in Fig. 4. Similar result found for twill weave by It is found in Fig. 6 that matt, twill, and sateen show
Asaduzzaman M. et al. [26], and the strength lower warp way tensile strength than that of plain
increases with the increase of heald counts that fabric except plain warp rib and z-twill. But, the
supports the result found by Malik Z.A. et al. [25]. Matt
fabric with a similar float length shows greater change is not very continuous except z-twill
strength than rib, twill, and sateen (Fig. 2) which structures in this experiment which are similar found
further shows different from Asaduzzaman M. et al. by Asaduzzaman M. et al. [26].
 HOSSAIN M.M., ET AL.: TEAR AND TENSILE STRENGTH OF 100 % COTTON WOVEN FABRICS’ BASIC STRUCTURES: REGRESSION MODELLING

Figure 4. Fractional change in warp way tear strength from the plain weave.

Figure 5. Fractional change in weft way tear strength with respect to plain weave.

Figure 6. Fractional change in warp way tensile strength with respect to plain weave structure.
 Fibres and Textiles 29(4), 2022, 3-8

Figure 7. Fractional change in weft way tensile strength from the plain weave structure.

Fig. 7 demonstrates that the tensile strength of the rib, design variation significantly influences tear strength,
matt, twill, and sateen fabric is lower than that of the and fabric setting or yarn density greatly varies the
plain structure. However, the strength of these fabrics tensile strength of the fabric. Thus, hypothesis testing
becomes greater with the increase of the heald count shows changing yarn density in fabric structure does
and demonstrates more than that of plain not influence tear strength, but yarn floating or weave
construction. Concerning plain fabric construction, design significantly influences the tear strength of the
twill and sateen fabric change is more regular than rib fabrics. On the other hand, design change or weave
and matt. Here with the increase of yarn densities interlacement does not considerably change the
below the optimum level the weft way tensile strength tensile strength of the fabrics. Other hypothesis
of twill and sateen was found lower than plain weave testing shows that EPCm, Law’s maximum yarn
that supports the result found by Asaduzzaman M. et setting, and GSM significantly influence warp and
al. [26] but at the optimum level the strength become weft tear strength. On the contrary, PPCm and Law’s
1 maximum yarn setting significantly affect warp and
higher than the plain weave which is again support
1 weft way tensile strength.
the theoretical context of Gokarneshan N. [28].
This research's hardest part was considering weave
The four regression models can precisely predict the factors in prediction models. As most of the weave
values using EPCm, PPCm, and Law’s maximum factors have high inter-correlation. They cannot be
yarn setting as influencing factors. As the yarn used viewed in a single model; only Law’s maximum yarn
in this research is one thus, fiber composition and setting is considered in prediction models. To
yarn properties remained constant in this study. It was understand effect of float length and comparison with
most challenging to find out the relation of these basic plain structure for tear and tensile strength, a
factors with the tear and tensile strength and consider graphical representation was the only option that
design variation in the prediction model. Law’s authors found during this study. The float length of the
maximum yarn setting parameters use the float length fabric significantly affects tear strength. With the
and yarn count during measurement and are increased float length, more yarns are close to each
statistically significantly related to tear and tensile other, thus increasing tear strength. But, it is not
strength. As float length is still the same for different clearly said from the data the effect of change in float
weave designs thus, there are still spaces to improve length how the tensile strength changes as it didn’t
this model. This study also uses graphical show any clear indication due to the fluctuation of the
representations to understand the variation of tear tensile strength much with the change of design of the
and tensile strength with design variation. fabrics and yarn set.
This study found that all models can predict the tear It is also found that the tear strength increases with
and tensile strength with accuracy from 70.1% to the increase of heald counts. The tear strength of
97.3% at a 5% significance level considering EPCm, matt fabric is higher than that of rib, twill, and sateen,
PPCm, Law’s maximum yarn setting, and GSM as with a similar float length. At the same time, twill and
independent variables. Among these parameters, the sateen show lower warp way tensile strength than
most significant factor for tear strength is EPCm and
PPCm for tensile strength. It is also observed that
 HOSSAIN M.M., ET AL.: TEAR AND TENSILE STRENGTH OF 100 % COTTON WOVEN FABRICS’ BASIC STRUCTURES: REGRESSION MODELLING

2 2 positively on tear and tensile strength also might be

that of plain fabric, except warp rib, z-twill and
2 2 balancing the negative influence of those parameters
3
z-twill structures in this experiment. in respected cases. As, design of the fabric cannot be
1
expessed in the numerical value as like other
The tensile strength of the rib, matt, twill, and sateen parameters (EPCm, PPCm, Law’s Max Yarn setting
fabric is lower than that of the plain structure. and GSM) the models did not consider the design
However, the strength of these fabrics becomes variables. Design variables as categorical or ordinal
greater with the increase of the heald count and variables are needed to adjust the models with
demonstrates more than that of plain construction. various designs. Again, Choosing fibers other than
With regard to plain fabric construction, the change in 100% cotton and yarn count other than 36.9 tex will
twill and sateen fabric is more regular than rib and enhance the possibility of versatile technical
matt. applications from lower strength to higher strength.
Thus, stiffness and other properties will be another
CONCLUSION option in proper selection of the fabric design to meet
the technical requirements.
Mathematical models concentrating on the
fundamentals of woven textiles may struggle to Funding: The work is financed by author throughout the
provide adequate findings since all of the model's experiment, sample preparation, raw material purchase,
and manpower recruitment. Authors took help from
uncertainties cannot be accounted for. Due to the
numerous skilled colleagues and took assistance without
accuracy required, mathematical modeling of fabric financial benefits. This complete work is done without any
constitutive equations necessitates the use of a organizational funding.
specialist approach. Certain variables are used as
assumptions for prediction in predictive, descriptive, Acknowledgement: We like to thank Bangladesh
University of Textiles (BUTEX) and Bangladesh University
and computational models. Mathematical
Business and Technology (BUBT) for doing the test
approaches are also case-specific. Four variables, required for the experiment and all its staff. Special thanks
EPCm, PPCm, GSM, and Law's maximum yarn to Prof. Engr. Masud Ahmed, Kazi Sayadul Alam, Rina
setting, are utilized to generate four regression Alam, Sheuli Aktar, Salma Akter, Kazi Md Mosharrof
models to predict fabric strength in the sample to Hossain, Md. Mozzammal Hossain, Md. Nazim Uddin, Md.
understand the effect of these parameters on the tear Shahin Alam, Md. Ali Akbar, Md. Shofiqul Islam, Md. Abul
and tensile strength of the 100% cotton woven Hasan and Md. Enamul Hasan.
fabrics. Combining yarn composition, textile structure,
twisting, and test techniques in a single formula is REFERENCES
difficult. This research determines the optimum
1. Realff M.L., Boyce M.C. and Backer S.: A micro-mechanical
parameters and statistical representation to predict model of the tensile behavior of woven fabric, Textile
fabric strength. Floating parameters impact test Research Journal, 67, 1997, pp. 445-459.
performance and parameters like elongation. Such https://doi.org/10.1177/0040517597067006
2. Chattopadhay R.: Design of apparel fabrics: Role of fiber,
variables are omitted from regression models to avoid yarn and fabric parameters on its functional attributes,
ambiguity. Research shows that thread setting and Journal of Textile Engineering, 54, 2008, pp. 179-190.
fabric design influence tear and tensile strength. As https://doi.org/10.4188/jte.54.179
3. Sharma I.C., Deshpande S.D., Jaiswani O.P., et al.: Effect of
yarn properties such as single yarn power, fiber, twist, wet processing on the tearing strength of polyester/viscose
count, lea strength, etc., are the same, they rayon blended fabrics, Indian Journal of Textile Research, 9,
dramatically alter the design result's tear strength. 1984, pp. 106-111.
Thread density affects tensile strength more than 4. Rafique S., Khattak S.P., Hussain T., et al.: Impact of
functional finishes on the tensile and tear strength properties
design, however. Four regression models might be of pigment dyed P/C fabrics by post & meta finishing modes
used within the scope of the design and yarn of application, Journal of Science and Technology University
properties, and more research could be done with Peshawar, 37(2), 2013, pp. 71-83.
5. Gupta P., Roy M.D., Ghosh S.: Effect of finishing chemicals
more yarn variation. Despite certain disadvantages, on tearing strength of plain-woven cotton fabric, Research
four regression models fit well (67%-98.1%) in this Journal of Textile and Apparel, 24(3), 2020, pp. 229-243.
case. Here, the matt and twill weave structure https://doi.org/10.1108/RJTA-09-2019-0043
showed more tear strength and tensile strength 6. Gabrijelcic, H., Cernosa, E., Dimitrovski, K.: Influence of
weave and weft characteristics on tensile properties of
respectively. So, these structures can be the best fabrics, Fibres & Textiles in Eastern Europe, 2(67), 2008, pp.
choice for specific strength in technical textile 45-51.
production for higher tear and tensile strength. The 7. Lopez V.M., Carou D., Cruz S.F.A.: Feasibility study on the
use of recycled materials for prototyping purpose: A
prediction models can be adjusted with consideration comparative study based on the tensile strength,
of other parameters and more design structures. Proceedings of the Institution of Mechanical Engineers, Part
Therefore in models, GSM had a negative impact on B: Journal of Engineering Manufacture, 46(19), 2022, pp.
weft way tear strength, EPCm and PPCm on warp 9187-9193.
https://doi.org/10.1177/09544054221113378
way tensile strength, and PPCm on weft way tensile 8. Zhang X., Wu M.: Modified stress field model for critical
strength, which might be due to woven fabric tearing strength of architectural coated fabrics, Journal of
relaxation and design variation or floating of warp and Industrial Textiles, 51(4), 2022, pp. 5560S-5591S.
https://doi.org/10.1177/15280837221106232
weft threads. The impact of other parameters
 Fibres and Textiles 29(4), 2022, 3-8

9. Sinclair R.: Textiles and Fashion: Materials, Design and 19. Morino H., Matsudaira M., Furutani M.: Predicting
Technology, Woodhead Publishing, 2014, pp. 705-737. mechanical properties and hand values from the parameters
https://doi.org/10.1016/C2013-0-17410-7 of weave structures, Textile Research Journal, 75(3), 2016,
10. Hossain M.M., Datta E., Rahman S.: A Review on different pp. 10-82.
factors of woven fabrics’ strength prediction, Science https://doi.org/10.1177/004051750507500
Research, 4(3), 2016, pp. 88-97. 20. ASTM D2256/D 2256M:2010. Standard Test Method for
https://doi.org/10.11648/j.sr.20160403.13 Tensile Properties of Yarns by the Single-Strand Method.
11. Ahirwar M., Behera B.K.: Prediction of tear strength of bed 21. ISO 13937-1:2000. Textiles - Tear Properties of Fabrics -
sheet fabric using machine learning based artificial neural Part 1: Determination of Tear Force Using Ballistic Pendulum
network, The Journal of The Textile Institute, 2022. Method (Elmendorf)
https://doi.org/10.1080/00405000.2022.2150960 22. ISO 13934-2:2014. Textiles - Tensile Properties of Fabrics -
12. Ribeiro R., Pilastri A., Moura C., et. al: Predicting the tear Part 2: Determination of Maximum Force Using the Grab
strength of woven fabrics via automated machine learning: Method.
An application of the CRISP-DM methodology. In 23. Eryuruk S.H., Kalaoğlu F.: The effect of weave construction
Proceedings of the 22nd International Conference on on tear strength of woven fabrics, AUTEX Research Journal,
Enterprise Information Systems, 1, 2020, pp. 548-555. 15(3), 2015, pp, 207-214.
https://doi.org/10.5220/0009411205480555 https://doi.org/10.1515/aut-2015-0004
13. Hossain K., Anwar M., Samani S.: Regression and artificial 24. Nasrun F.M.Z., Yahya M.F., Ghani S.A., et al: Effect of weft
neural network models for strength properties of engineered density and yarn crimps towards tensile strength of 3D angle
cementitious composites, Neural Computing and interlock woven fabric, AIP Conference Proceedings, 1774,
Applications, 29, 2018. 2016, 020003.
https://doi.org/10.1007/s00521-016-2602-3 https://doi.org/10.1063/1.4965051
14. Singh N.: Weave factors of different woven constructions, 25. Malik M.H., Hussain T., Ali, Z.: Effect of fabric count on the
Delhi: IIT, 2007. B. Tech Project Report. tensile strength of blended woven fabrics, Journal of
15. Elkateb S.N.: Prediction of mechanical properties of woven Engineering and Applied Science, 28(2), 2009, pp. 23-29.
fabrics by ANN, Fibres & Textiles in Eastern Europe, 30(4), https://doi.org/10.25211/JEAS.V28I2.297
2022, pp. 54-59. 26. Asaduzzaman M., Hasan A.K.M.M., Patwary M.M., et. al:
https://doi.org/10.2478/ftee-2022-0036 Effect of weave type variation on tensile and tearing strength
16. Shahpurwala A., Schwartz P.: Modeling woven fabric tensile of woven fabric, Technium, 2(6), 2020, pp. 35-40.
strength using statistical bundle theory, Textile Research https://doi.org/10.47577/technium.v2i6.1409
Journal, 59(1), 1989, pp. 26-32. 27. Radwan S.S.: Effect of plain rib direction on fabric properties,
https://doi.org/10.1177/004051758905900104 International Design Journal, 4(4), 2014, pp. 179-185.
17. Hu J., Fabric Testing. Cambride: Woodhead Publishing 28. Gokarneshan N.: Fabric structure and design, New Delhi:
Limited, 2008, pp. 92-104. New Age International Publishers, 2020, pp. 1-184.
18. Mishra, S.: Prediction of yarn strength utilization in cotton
woven fabrics using artificial neural network, Journal of the
Institute of Engineers, Series E, 96(2), 2014, pp. 151-157.
https://doi.org/10.1007/s40034-014-0049-6
 
.