Original Article

J Strain Analysis
1–9
Ó IMechE 2018
Statistical analysis of yarn to metal Reprints and permissions:
sagepub.co.uk/journalsPermissions.nav
frictional coefficient of cotton spun DOI: 10.1177/0309324718786373
journals.sagepub.com/home/sdj

yarn using Taguchi design of

experiment

Amir Shahzad1, Naseer Ahmad1, Zulfiqar Ali1, Ali Afzal1, Muhammad

Bilal Qadir1,2 , Zubair Khaliq1,2 and Muhammad Qamar Khan3

Abstract
Yarn’s surface to metal friction is an important consideration in the subsequent process of knitting and weaving as it influ-
ences mainly the ends down rate, fly generation, process efficiency, wear and tear of machine parts, and production rate
of the process. These frictional properties are measured in terms of the coefficient of friction of yarn. The effect of cot-
ton type, yarn twist, yarn linear density, process type, and finishing treatment was studied on the surface to the metal
friction coefficient of cotton spun yarn using Taguchi experimental design. The experiments were conducted with
Pakistani and Indian cotton using combed and carded ring spinning processes. Using Taguchi design of experiment, a total
of 36 samples of cotton ring-spun yarns were produced. The coefficient of friction between the yarn’s surface and metal’s
surface is measured in compliance with ASTM D3108. The outcome of the Taguchi model to predict the coefficient of
friction of yarns with a predefined combination of constituting parameters was further confirmed with nine yarn samples.
The frictional characteristics of yarns are found to be influenced by all factors. In addition to the application of wax, the
longer fiber length, lower trash count, lower short fiber index, and the optimum level of twist are found advantageous
to reduce the yarn coefficient of friction.

Keywords
Coefficient of friction, statistical analysis, Taguchi design, spun yarn, cotton, wax

Date received: 20 March 2018; accepted: 8 June 2018

Introduction properties.4–6 Among various properties of yarn, the

COF of yarn to metal becomes more critical in subse-
A textile yarn is an assembly of substantial length and a quent weaving and knitting processes.7 Knapton et al.
relatively small cross-section of fibers and/or filaments have reported that the loop length and knitting tension
with or without twist. Yarn production is a complicated in flatbed knitting machine can vary due to variations
process which involves vast technologies, materials, and in yarn COF. A higher knitting tension develops due to
influencing factors. The potential application of yarns higher friction which may exceed the yarn breaking
is not only in developing woven and knitted fabrics but load.8 Frictional forces between yarns and knitting nee-
also in the production of ropes, sewing threads, and dles caused by inferior yarn quality may result in heat
some special purpose yarns.1
Among various physical and mechanical properties
1
of yarns, the frictional behavior is also considered National Textile University, Faisalabad, Pakistan
2
Hanyang University, Seoul, South Korea
another important yarn quality parameter.2 Friction is 3
Shinshu University, Nagano, Japan
the force to resist any movement between two sliding
surfaces regardless of the matter. It is measured in Corresponding authors:
terms of coefficient of friction (COF), denoted by m Muhammad Bilal Qadir, National Textile University, Faisalabad 37610,
which reflects the ratio of frictional force to normal Pakistan.
Email: [email protected]
load between two sliding surfaces.3 The physical,
mechanical, and comfort properties of the fabric are Zubair Khaliq, National Textile University, Faisalabad 37610, Pakistan.
highly governed by component fiber and yarn Email: [email protected]
2 Journal of Strain Analysis 00(0)

generation, increased knitting tension, and subsequent

breakage of either yarn or knitting needles.9,10
It is found that besides the traditional yarn para-
meters like evenness, imperfections, strength, and elon-
gation, the yarn to yarn and yarn to metal friction play
a vital role in textile processing.11 An increase in yarn
friction while running with machine parts and needles
may result in heat production, increased ends
down rate, and fly generation. Measurement and con-
trol of yarn friction help to optimize settings of waxing
device, pick up of wax by yarn during yarn winding.
Furthermore, it ensures the requirement of yarn fric-
tional properties to achieve better working efficiencies,
reduced waste, and increased production rates of
desired entities.
The previous research has revealed that the fiber
properties, yarn structural and bulk parameters, pro-
cess variables, and yarn finishing treatments are the Figure 1. Steps involved in Taguchi process.
most influential parameters in the optimization of yarn
frictional properties. The relationship between physical
properties and frictional characteristics of cotton fiber
and cotton yarn has been investigated previously. It friction, yarn to yarn and yarn to metal friction, fabric
was observed that the coarser and compressible yarns friction and their relationships to fiber types, yarn
exhibit a higher value of COF.12,13 Ajayi conducted a structure, and tension applied. However, this study was
study on fabric frictional properties in relation to that focused on the optimization of yarn to metal friction
of yarn and reported that fabrics have higher COF val- properties of cotton yarn in relation to cotton types,
ues than that of its component yarns.6 Researchers spinning process, level of twist, fineness of yarn, and
have also measured the yarn to yarn and yarn to metal application of wax using Taguchi design of experiment.
friction for the ring, rotor, air-jet, and open-end fric- Robust design method, also called the Taguchi
tion (OE friction) spun yarns of viscose fibers at differ- method, pioneered by Dr Genichi Taguchi, focuses on
ent levels of relative speeds and input tensions. It was the effective application of engineering strategies.
concluded from results that the yarn to yarn friction Robust Design focuses on improving the fundamental
decreases gradually for OE friction, rotor, air-jet, and function of the product or process, thus facilitating
ring-spun yarn contrary to the entirely reverse order flexible designs and concurrent engineering. Indeed, it
observed in case of yarn to metal friction. The findings is a powerful method available to reduce product cost,
were attributed to the relative area of interaction improve quality, and simultaneously reduce develop-
between two surfaces in contact in relation to the sur- ment interval.19
face morphologies of metal and yarns. The relative Taguchi method is very useful to optimize the
speed and input tension are also found to have a signif- response variable when various factors with different
icant effect on frictional characteristics of spun yarn.11 levels are involved in an experiment. It is a four-step
The effect of cotton, viscose, and polyester fibers yarn process consisting of Planning, Conducting, Analyzing,
structures produced from the ring, rotor, and friction and Validation, as shown in Figure 1.
spinning technique has also been on the frictional prop- In Taguchi method, the major tools used are main
erties of yarn.14 Subramaniam and Natarajan15 investi- effects plot for means, signal-to-noise (S/N) ratio plots,
gated the friction properties of Sirospunä yarn using and results prediction for validation. Main effects plot
yarn to yarn and yarn to metal friction measurement shows the actual output of experiment. S/N ratio is a
methods. Ramkumar et al.16 evaluated that the fric- signal-to-noise ratio expressed in decibels. It is directly
tional properties of friction spun yarn are influenced by related to the quality of the product. Hence, the S/N
types of fiber and tension applied. It is also observed ratio plot indicates optimum level for each factor.
that the application of lubricants on yarn surface rap- However, S/N ratio calculation depends on results of
idly decreases the COF value from 0.65 to a value experiment whether smaller is better or larger is better.
between 0.35 and 0.15, depending on lubricant For smaller is better, S/N ratio is expressed as
applied.17 The application of wax is also found useful !
in the reduction of lint shedding of cotton hosiery spun 1X n
S=N =  10 log y2 ð1Þ
yarn.18 The survey of literature review revealed that n i=1 i
many studies had been done on the fiber to fiber
Shahzad et al. 3

For larger is better, expressed as Table 1. Factors and levels for ring-spun yarn.
!
Parameters Designation Levels
1X n
1
S=N =  10 log ð2Þ 1 2 3
n i = 1 y2i
Cotton type A Pak Indian –
where n indicates the number of trials and yi is the ith Ring spinning B Carded Combed –
value measured. process
In results prediction, Taguchi model predicts the Yarn treatment C Waxed Unwaxed –
most probable outcome under specific levels of each Twist multiplier (TM) D 3.7 3.9 4.1
Count (Ne) E 16 20 24
factor by the formula

X
j
h = hm + ð hi  hm Þ ð3Þ Table 2. Parameters of Pakistani and Indian cotton.
i=1
Parameter Pakistani Indian
where j denotes the number of factors, hm is the aver-
age value of multiple S/N ratios in all experiments, and Spinning consistency index (SCI) 132 146
hi are the multiple S/N ratios for the optimum factors. Micronaire (mg/in) 4.70 3.94
Fiber length (mm) 27.36 29.49
Detailed information about the Taguchi method can
Strength (g/tex) 31.9 31.1
be found in many articles.20–22 Cho and Jeong23 used Short fiber index (SFI) 8.5 6.9
Taguchi approach to predict the tensile properties of Reflectance, Rd % 71.0 79.8
spun-dyed yarn at certain parameters. Cheng and Li24 Color grade 41-2 21-1
applied L16 Taguchi design to study the yarn hairiness Trash count 48 22
under various spinning factors in the Jet Ring spinning
system and predicted the optimum spinning variables
for different fibers. Ishtiaque et al.25–27 applied Taguchi Taguchi design of experiment
design and analysis of variance in their numerous stud-
ies to investigate the effect of spinning process para- Planning
meters on different properties rotor, air-jet, and ring- To observe the effect of variables under consideration
spun viscose yarns. on COF of cotton ring-spun yarn, the selected factors
The Taguchi method has been introduced in the field and their levels are given in Table 1.
of textile study effectively. However, no research has The twist multiplier (TM) is used to describe hard-
been conducted on the optimization of yarn to metal ness of yarn. It measures the amount of twist in the
COF for ring-spun yarn using Taguchi design of the yarn and is proportional to the fiber angle of wrap on
experiment. Thus, in this study, the effect of certain outer yarn surface to the yarn axis. It is calculated by
parameters of yarn and cotton types on the COF of dividing the turns per inch of yarn by the square root
yarn was analyzed using Taguchi’s experimental design of English yarn count
L36. The L36 is a fractional factorial mixed level design
often termed as orthogonal arrays, which is represented TPI
TM =
by the convention La (b^c) where a indicates the num- =Ne
ber of trial runs, b represents the number of levels of
where TM is the twist multiplier, TPI is the turns per
each factor, and c is the number of factors. Such ortho-
inch, and Ne is the English count system.
gonal array is useful when the factors involved in
experimental design are with unequal levels. For exam-
ple, L36 (2^3 3 3^13) is a design which comprises on Materials and methods. The fiber parameters of Pakistani
up to 3 factors at 2 levels each and up to 13 factors at 3 and Indian cotton are in given in Table 2.
levels each giving into 36 number of experimental runs. Pakistani cotton and Indian cotton were spun sepa-
Such Taguchi designs simplify the number of experi- rately to produce 16s Ne, 20s Ne, and 24s Ne carded
ments compared to full factorial designs and facilitate and combed yarn on mini ring frame machine. In the
the analysis and explanation of results. The Taguchi carded process, the cotton is processed through the
design is usually used to estimate the optimum process blow-room, card, drawing, simplex, and ring frame to
conditions and to study the individual effects of each of produce the yarn. In this process, the large amount of
the controllable factors on a particular response. The short fibers and solid particles waste are removed in
controllable factors considered in this study were cot- blow-room and card department only. While in case of
ton types, spinning process (carded, combed), yarn lin- a combed process, to remove an additional percentage
ear density, level of twist, and application of wax on of short fiber, the material is subjected to two addi-
the yarn. tional processes of lap forming and combing after
4 Journal of Strain Analysis 00(0)

blow-room, card, and first passage of draw frame. The

waste removed at combing machine in terms of short
fibers is termed as noil and is responsible for improved
evenness and strength of yarn. The noil percentage for
Pakistani and Indian cotton was set at 18% and 16%,
respectively. The other process parameters were set in a
way to achieve the final yarn count of 16s, 20s, and
24s. Yarn samples of various counts were produced at
three TMs of 3.7, 3.9, and 4.1. The winding of yarn
was done to apply the wax on yarn, clearing of unac-
ceptable yarn faults, and prepare larger size conical
packages.

Conducting
Using Minitab software-17.0, the Taguchi experimental Figure 2. Schematic diagram of yarn friction measuring
apparatus.28.
design L36 (2^3 3 3^2) is generated consisting of two
factors with three levels and three factors with two lev-
els. According to the diverse combination of factors
average value of the COF of each yarn sample along
and levels proposed by the L36 orthogonal array, 36
with its respective S/N ratio, calculated by formula (2)
yarn samples were produced using ring spinning tech-
is given in Table 3.
nique. The yarn samples were conditioned and tested
according to the standard testing procedures described
under testing section. Analysis of results
Testing results are analyzed using Taguchi method in
Testing Minitab. The COF of yarns is analyzed against cotton
The samples were conditioned at standard atmospheric type, spinning process, yarn treatment, level of twist,
conditions of 65% 6 2 relative humidity and tempera- and count of the yarns.
ture of 20 °C 6 2 °C according to ASTM D 1776. The
COF of each yarn was measured on MESDAN Taguchi optimization method
ATTRIFI-II yarn friction tester according to ASTM
D3108 which encompasses the measurement of sliding Taguchi optimization method is composed of following
friction between yarn and solid surface. During the test, steps:
the yarn run under a pretension (usually 20 cN) at a
speed of 150 m/min around a stainless steel pin making Step 1: In this step, it is defined that whether the value
a warping angle of 180°. The average surface roughness of the response is smaller is better or larger is better. In
(Ra) of stainless steel pin is \ 0.5 mm. The COF of this case, the value of COF is smaller is better.
yarn is calculated by measuring the yarn input and out- According to equation (1), The S/N ratio for each
put tensions, using the following equations observation can be calculated using the mathematical
expression for smaller the better scenario: that is, the
T2 lower yarn COF, the better it is.
emu = ð4Þ
T1 Step 2: The second step is to determine the average
  effect of each factor on the multiple quality characteris-
ln TT21
tics at different levels. This corresponds to the average
m= ð5Þ
u S/N ratio of a factor at its specific level 20. For exam-
where m is the COF, T1 is the average input tension, T2 ple, the average S/N ratio for the first level of factor A
is the average output tension, and u is the cumulative is calculated as follows
wrap angle. ð14:99 + 14:97 + 15:53 + 10:62 + 10:67
The schematic diagram of yarn friction measuring
setup is shown in Figure 2. + 10:69 + 15:32 + 15:84 + 15:93 + 10:74
+ 11:04 + 10:60Þ=12 = 13:08

Analysis Step 3: For various factors under observation, the level

In Taguchi design of the experiment, the COF of all corresponding to maximum average S/N ratio is
samples was analyzed for mean values and respective selected as optimum level. The average S/N ratio
S/N ratios. The Taguchi analysis determined the influ- against each level of a specific factor is given in
ence of each variable on response factor and proposed Table 4, and the main effects plots for S/N ratios are
the optimum factors to achieve the best results. The shown in Figure 3.
Shahzad et al. 5

Table 3. Experimental layout using an L36 orthogonal array table and S/N ratio of experimental results.

Experiment no. Factors and levels COF of yarn S/N ratio (dB)
Cotton Process Finish TM Count

1 PAK CD W 3.7 16s 0.1780 14.992

2 PAK CD W 3.9 20s 0.1784 14.972
3 PAK CD W 4.1 24s 0.1673 15.530
4 PAK CD W 3.7 16s 0.1780
5 PAK CD W 3.9 20s 0.1784
6 PAK CD W 4.1 24s 0.1673
7 PAK CD U 3.7 16s 0.2945 10.618
8 PAK CD U 3.9 20s 0.2926 10.675
9 PAK CD U 4.1 24s 0.2922 10.686
10 PAK CM W 3.7 16s 0.1714 15.320
11 PAK CM W 3.9 20s 0.1615 15.837
12 PAK CM W 4.1 24s 0.1597 15.934
13 PAK CM U 3.7 20s 0.2904 10.740
14 PAK CM U 3.9 24s 0.2804 11.044
15 PAK CM U 4.1 16s 0.2951 10.601
16 PAK CM U 3.7 20s 0.2904
17 PAK CM U 3.9 24s 0.2804
18 PAK CM U 4.1 16s 0.2951
19 IND CD U 3.7 20s 0.2886 10.794
20 IND CD U 3.9 24s 0.2914 10.710
21 IND CD U 4.1 16s 0.2886 10.794
22 IND CD U 3.7 20s 0.2886
23 IND CD U 3.9 24s 0.2914
24 IND CD U 4.1 16s 0.2886
25 IND CD W 3.7 24s 0.1689 15.447
26 IND CD W 3.9 16s 0.1883 14.503
27 IND CD W 4.1 20s 0.1523 16.346
28 IND CM U 3.7 24s 0.2968 10.551
29 IND CM U 3.9 16s 0.2867 10.851
30 IND CM U 4.1 20s 0.2851 10.900
31 IND CM W 3.7 24s 0.1568 16.093
32 IND CM W 3.9 16s 0.1658 15.608
33 IND CM W 4.1 20s 0.1509 16.426
34 IND CM W 3.7 24s 0.1568
35 IND CM W 3.9 16s 0.1658
36 IND CM W 4.1 20s 0.1509

PAK: Pakistani; IND: Indian; CD: carded; CM: combed; W: waxed; U: unwaxed; S/N: signal-to-noise; TM: twist multiplier.

The delta value was calculated by subtracting the

Table 4. Response table for the S/N ratio.
minimum value from the largest value among various
levels of each factor. Higher values of the S/N ratio Factors
indicate influencing factor settings that minimize the
effects of the noise factors. Hence, for the lowest value Average S/N Ratios
of COF of yarn, the optimum parameters with highest Levels A Cotton B Process C Treatment D Count E TM
S/N ratios are A2B2C1D3E2. The detail of optimum
parameters is given in Table 5. 1 13.08 13.01 15.58 13.07 12.91
2 13.25 13.33 10.75 13.03 13.34
3 13.40 13.25
Step 4: Prediction and validation experiment Delta 0.17 0.32 4.84 0.38 0.43
Rank 5 4 1 3 2

In the final step, using the model from Taguchi design S/N: signal-to-noise; TM: twist multiplier.
of the experiment, the prediction of the S/N ratio and
response characteristics of selected factor settings can
be calculated by the formula where j represents the number of factors, hm the mean
value of multiple S/N ratios in all experimental runs,
X
j
h = hm + ð hi  hm Þ ð6Þ and hi are the multiple S/N ratios corresponding to
i=1 optimum factor levels.22,29
6 Journal of Strain Analysis 00(0)

Figure 3. Mean effect plot for S/N ratio (smaller is better).

!
1X n
Table 5. Optimum parameter for lowest COF of yarn. S=N =  10 log y2
n i=1 i
Factor (level) Description/value S/N ratio
The predicted value is
A (2) Cotton (Indian) 13.25 !
B (2) Quality (combed) 13.33 1X n
C (1) Finish (wax) 15.58 15:7659 =  10 log y2 ) yi = 0:1628
D (3) Count (24s) 13.40 1 i=1 i
E (2) TM (3.9) 13.34
Now the sample prepared at the proposed levels of
S/N: signal-to-noise; TM: twist multiplier.
each factor, that is, Indian: Combed: Wax: 24s:3.9. The
COF value of yarn is measured and found to be 0.1535,
which is very close to the predicted value by Taguchi
The following expression can calculate the optimum design, that is, 0.1632.
S/N ratio for proposed optimum levels of influencing
factors
Validation
ho = hm + (hA2  hm ) + (hB2  hm ) + (hC1  hm )
To check the validation of Taguchi model, nine sam-
+ (hD3  hm ) + (hE2  hm )
ples were prepared with a random combination of fac-
ð7Þ tor and their levels for each sample. The samples were
where ho is the optimum S/N ratio, hm the overall mean tested for COF values, and the actual values of COF
of S/N values, hA2 the mean value of S/N at level 2 of were compared to that of predicted values for these
cotton, hB2 the mean value of S/N at level 2 of the spin- samples.
ning process, hC1 is the mean value of S/N at level 1 of
yarn treatment, hD3 is the mean value of S/N at level 3 Prediction of the COF for other combinations using
of yarn count, and hE2 is the mean value of S/N at level
Taguchi method
2 of yarn TM.
Putting the values of different terms in equation (7) According to Taguchi design, values for the COF of
ring-spun yarns for random nine conditions are pre-
ho = 13:17 + ð13:25  13:17Þ + ð13:33  13:17Þ dicted using equations (1) and (7). These samples were
+ ð15:58  13:17Þ + ð13:40  13:17Þ then prepared and tested for an experimental value of
+ ð13:34  13:17Þ = 15:7659ðdBÞ COF. Error percentages are also evaluated between
predicted and experimental values. The predicted
Values of COF for other combinations can be values along with their experimental results are given in
derived using the same method. Now, using equation Table 6 and as graphically in Figure 4.
(1), a procedure of reverse calculation is used to find It is evident from Figure 4 that the experimental val-
the COF value of yarn expected for S/N ratio of ues of COF are much closer to the predicted values.
15.7659.21,30 However, the fluctuation in COF between 0.1 and 03 is
Shahzad et al. 7

Figure 4. Comparison of experimental results and values predicted.

Table 6. Comparison of experimental results with predicted values from the Taguchi design.

S. no. A B C D E Experimental values of COF Predicted values of COF Relative error (%)

1 PAK CD U 16 4.1 0.2963 0.2960 0.10

2 PAK CM W 20 3.9 0.1615 0.1638 1.42
3 PAK CD U 24 3.7 0.2994 0.2957 1.26
4 PAK CD U 20 4.1 0.2855 0.2882 0.92
5 IND CM W 24 3.7 0.1568 0.1687 7.03
6 IND CD U 16 3.7 0.2918 0.2984 2.20
7 IND CM W 20 3.9 0.1535 0.1604 4.30
8 IND CD U 24 4.1 0.2897 0.2865 1.13
9 IND CM W 16 3.9 0.1658 0.1682 1.45

PAK: Pakistani; IND: Indian; CD: carded; CM: combed; W: waxed; U: unwaxed; COF: coefficient of friction.

due to the waxed and unwaxed surface of yarns. Hence, which improves the evenness and helps to reduce the
Taguchi design of the experiment is an excellent tool COF of yarns.
for process optimization as well as yield and quality
improvement by selecting the process and material vari-
ables with a maximum contribution to the desired Effect of process on COF
response. During combing process, short fibers are removed from
the cotton which results in improved evenness and
Results and discussion strength of yarn. With the removal of comber noil and
trash particles during the combing cycle, the remaining
In this study, the comparison of frictional properties of relatively longer fibers also become more parallel and
different yarns is made in relation to a particular metal oriented during the subsequent drafting operations.
surface. Since the roughness of metal surface also influ- Contrary to this, in the carded process only, the short
ences the yarn to metal friction,31 hence a polished
fiber and a large amount of dust and tiny trash particles
stainless steel pin with an average surface roughness
are retained in the processing material causing drafting
Ra \ 0.5 mm is used in all tests to compare the yarn to
irregularities, increased imperfection, and hence rela-
metal sliding frictional properties. The effect yarn struc-
tively rough surface of carded cotton spun yarn. Hence,
tural parameters on COF are discussed as follows.
COF of carded yarn is also relatively high.

Effect of cotton type on COF

The overall better results of COF of yarn for Indian
Effect of level of twist (TM) on COF
cotton are due to its better fiber quality parameters Another factor, controlling the COF of yarn is a level
compared to that of Pakistani cotton used in this study. of twist denoted by ‘‘TM.’’ With the gradual increase in
As shown in Table 2 that Indian cotton has relatively the level of twist, the degree of compactness of yarn
larger fiber length (29.44 mm), finer micronaire (3.94), also increases which may lower the total area of contact
better spin-ability (SCI 146), and lower trash content22 of yarn with metal surface during measurement result-
as compared to that of Pakistani cotton. These quality ing in a relatively lower value of COF. The conclusion
parameters contribute to the better quality of yarn is also in agreement with findings of another research
8 Journal of Strain Analysis 00(0)

Figure 5. SEM image of unwaxed yarn.

Figure 7. Normal probability plot of residuals.

that the waxed yarn has much smooth and even surface
than that of unwaxed yarn.

Conclusion
The finding of this study suggests that in order to
reduce the COF of yarn, we can find the optimum yarn
constructional parameters successfully using Taguchi
design of experiment. Depending upon S/N ratio analy-
Figure 6. SEM image of waxed yarn.
sis, the optimum parameters determined in this study
are cotton with better fiber quality parameters, combed
in which the frictional behavior of carbon fiber tows is
process of spinning, fineness of yarn count, optimum
explored.32
TM level, and waxed treatment of yarns. Moreover,
the closeness between predicted and experimental val-
Effect of yarn count (Ne) on COF ues builds a trust that the Taguchi experimental tech-
nique can be used effectively for both prediction and
COF of yarns decreases with increase in the count
optimization of a particular response. Figure 7 repre-
(Ne). As in English system of yarn numbering, the yarn
sents that the normal probability plot of the residuals
becomes finer with an increase in count number.
which reveals that the residuals are located nearest to
Hence, more twist is required to hold the fiber in the
the straight line and errors are normally distributed.
finer yarn which results in more compact structure of
Hence, it is concluded that the Taguchi approach is an
yarn. Reduced yarn thickness and a higher level of
effective tool for optimizing the near optimum design
twist again result in relatively compact structure result-
comprising the main factors influencing the process
ing in lowering the contact area with friction measuring
with only a few distinct sets of experiments.
disk and reduced COF of finer yarn.

Declaration of conflicting interests

Effect of wax applied on COF
The author(s) declared no potential conflicts of interest
Waxing is lubrication of yarn surface which causes a with respect to the research, authorship, and/or publi-
reduction in COF value of cotton spun yarn. The basic cation of this article.
purpose of wax is to reduce friction and heat generation
during processing. Wax is a lubricant and will always
reduce friction. Therefore, yarns surface treated with Funding
wax also have low COF because of more smooth, slip- The author(s) received no financial support for the
pery, and less hairy yarn surface. Figures 5 and 6 show research, authorship, and/or publication of this article.
Shahzad et al. 9

ORCID iD 17. Rubenstein C. Review on the factors influencing the

Muhammad Bilal Qadir https://orcid.org/0000-0003-4571- friction of fibres, yarns and fabrics. Wear 1959; 2(4):
5328 296–310.
18. Ruppenicker GF and Lofton JT. Factors affecting the
lint shedding of cotton knitting yams. Text Res J 1979;
References 49(12): 681–685.
1. Barella A and Manich AM. The influence of the spinning 19. Kackar RN. Off-line quality control, parameter design,
process, yarn linear density, and fibre properties on the and the Taguchi method. J Qual Technol 1985; 17(4):
hairiness of ring-spun and rotor-spun cotton yarns. J 176–188.
Text I 1988; 79(2): 189–197. 20. Taguchi G, Chowdhury S and Wu Y. Taguchi’s quality
2. Saxl IJ. The coefficient of friction for yarns. J Franklin I engineering handbook. Hoboken, NJ: Wiley, 2005.
1936; 221(6): 789–795. 21. Roy RK. Design of experiments using the Taguchi
3. Howell HG and Mazur J. Amontons’ law and fibre fric- approach: 16 steps to product and process improvement.
tion. J Text Inst Trans 1953; 44(2): T59–T69. Hoboken, NJ: Wiley, 2001.
4. Behera BK, Ishtiaque SM and Chand S. Comfort prop- 22. Ross PJ. Taguchi techniques for quality engineering: loss
erties of fabrics woven from ring-, rotor-, and friction- function, orthogonal experiments, parameter and tolerance
spun yarns. J Text I 1997; 88(3): 255–264. design. New York: McGraw-Hill Education, 1996.
5. Huh Y, Young Ryul K and Oxenham W. Analyzing 23. Cho DH and Jeong Y. Analysis of spinning process
structural and physical properties of ring, rotor, and fric- parameters on development of spun-dyed PET yarn
tion spun yarns. Text Res J 2002; 72(2): 156–163. using the Taguchi method. J Appl Polym Sci 2006;
6. Ajayi J and Elder H. Comparative Studies of Yarn and 102(2): 1419–1427.
Fabric Friction. J Test Eval 1994; 22(5): 463–467. 24. Cheng KPS and Li CHL. JetRing spinning and its influ-
7. Jawale S and Patil U. Yarn friction & its importance, ence on yarn hairiness. Text Res J 2002; 72(12): 1079–1087.
theory, factors, measurement. Indian Text J 2011; 122(3): 25. Ishtiaque SM, Salhotra KR and Kumar A. Analysis of
14–16. spinning process using the Taguchi method. Part II: effect
8. Knapton JJF and Munden DL. A study of the mechan- of spinning process variables on fibre extent and fibre
ism of loop formation on Weft-knitting machinery. overlap in ring, rotor and air-jet yarns. J Text I 2006;
Part II: the effect of yarn friction on yam tensions in 97(4): 285–294.
knitting and loop formation. Text Res J 1966; 36(12): 26. Kumar A, Ishtiaque SM and Salhotra KR. Analysis of
1081–1091. spinning process using the Taguchi method. Part III:
9. Lou CW. Production of a polyester core-spun yarn with effect of spinning process variables on migration para-
spandex using a multi-section drawing frame and a ring meters of ring, rotor and air-jet yarn. J Text I 2006; 97(5):
spinning frame. Text Res J 2005; 75(5): 395–401. 377–384.
10. Onofrei E, Rocha AM and Catarino A. The influence of 27. Kumar A, Ishtiaque SM and Salhotra KR. Analysis of
knitted fabrics’ structure on the thermal and moisture spinning process using the Taguchi method. Part IV:
management properties. J Eng Fibers Fabr 2011; 6(4): effect of spinning process variables on tensile properties
10–22. of ring, rotor and air-jet yarns. J Text I 2006; 97(5):
11. Ghosh A, Patanaik A, Anandjiwala RD, et al. A study 385–390.
on dynamic friction of different spun yarns. J Appl Polym 28. Kumar A, Salhotra KR and Ishtiaque SM. Analysis of
Sci 2008; 108(5): 3233–3238. spinning process using the Taguchi method. Part V:
12. Nair AU, Sheela R, Vivekanandan MV, et al. Studies on effect of spinning processvariables on physical proper-
friction in cotton textiles: part I-A study on the relation- ties of ring, rotor and air-jet yarns. J Text I 2006; 97(6):
ship between physical properties and frictional character- 463–473.
istics of cotton fibres and yarns. Indian J Fibre Text Res 29. Dubey AK and Yadava V. Simultaneous optimization of
2013; 38: 244–250. multiple quality characteristics in laser beam cutting using
13. Sülara V, Öner E and Okur A. Roughness and frictional Taguchi method. Int J Precis Eng Man 2007; 8(4): 10–15.
properties of cotton and polyester woven fabrics. Indian J 30. Mavruz S and Oğulata RT. Taguchi approach for the
Fibre Text Res 2013; 38(4): 349–356. optimisation of the bursting strength of knitted fabrics.
14. Chattopadhyay R and Banerjee S. The frictional beha- Fibres Text East Eur 2010; 18(2): 78–83.
viour of ring-, rotor-, and friction-spun yarn. J Text I 31. Mulvihill DM and Sutcliffe MPF. Effect of tool surface
1996; 87(1): 59–67. topography on friction with carbon fibre tows for com-
15. Subramaniam V and Natarajan KS. Frictional properties posite fabric forming. Compos Part A: Appl S 2017; 93:
of siro spun yarns. Text Res J 1990; 60(4): 234–239. 199–206.
16. Ramkumar SS, Shastri L, Tock RW, et al. Experimental 32. Mulvihill DM, Smerdova O and Sutcliffe MPF. Friction
study of the frictional properties of friction spun yarns. J of carbon fibre tows. Compos Part A: Appl S 2017; 93:
Appl Polym Sci 2003; 88(10): 2450–2454. 185–198.