New possibility oF objective evaluatioN oF yarN appearaNce: part ii

AUTEX Research Journal, Vol. 13, No 1, March 2013, DOI: 10.2478/v10304-012-0016-6 © AUTEX 

New possibility oF objective evaluation of Yarn appearAnce: part II 

Eva Moučková, Petra Jirásková 

Technical University of Liberec, Faculty of Textile Engineering, Department of Textile Technologies, Liberec, 

The Czech Republic Studentská 2, 461 17 Liberec 1 

E-mail: eva.mouckova@tul.cz; petra.jiraskova@tul.cz 

Abstract: 

Keywords: 

In this article, the possibility of new objective evaluation of yarn appearance in area employing selected spatial 

statistical functions is presented. The yarns wound on the boards were used for the experiment. Appearance of 

standard yarn boards from the standard CSN 80 0704 and real yarn boards with faultless and faulty yarns were 

converted into grayscale images. Fluctuation in degrees of grayness was evaluated between square fields in the 

image using statistical function called the area variation curve. In addition, the method was applied to the simulated 

yarn board appearance generated by the Uster Tester apparatus. Behavior of area variation curves in dependence 

on the result of visual evaluation of yarn board appearance was discussed. The generated appearances from the 

Uster Tester device were also evaluated by other statistical function called semivariogram. It was found out that the 

area variation curve is not a suitable tool for evaluation of yarn board appearance. The semivariogram seems to be 

a more suitable tool. The paper extends the knowledge on the issue of objective evaluation of yarn appearance and 

directly follows the author’s work [1]. 

Yarn appearance, Evaluation, Area variation curve, Semivariogram, Yarn board, Yarn grades 

Introduction 

This paper is a direct follow-up of the authors’ work [1], where 

the spatial statistical function called semivariogram was 

introduced as a possible tool for the objective evaluation of 

yarn appearance. 

Yarn quality is judged according to the achieved values of 

tested yarn properties that are selected in dependence on 

yarn utilization. Quality level of the yarns can be assessed 

using the USTER® STATISTICS [2]. This tool is composed 

of graphs enabling users to compare their measured results 

of yarn (sliver, roving, fiber) with the corresponding worldwide 

established fibrous product quality reference values; in the 

case of yarns, it includes count variation, mass variation, 

imperfection, hairiness, diameter variation as well as tensile 

properties. 

Yarn count variation, mass and diameter variation, 

imperfection and hairiness mainly affect the yarn appearance. 

In practice, the yarn appearance is usually evaluated 

subjectively by comparing a yarn board of defined winding 

density with a standard yarn board according to ASTM D 

2255-90 [3], or according to standard CSN 80 0704, which is 

used in the Czech Republic [4]. The evaluation is dependent 

on the reviewer and their ability to visually compare yarns. 

The company Lawson-Hemphill in cooperation with the 

United States Department of Agriculture and Cotton Inc. has 

developed the automated yarn grading system according 

to ASTM for several counts [5,6]. This system also known 

as EIB (Electronic Inspection Board) or YAS (Yarn Analysis 

System) uses optical digital technology and replaces 

human inspection grading [6]. The other method simulating 

the procedure of subjective visual assessment of yarn 

appearance is introduced in work [7]. 

Some devices for measuring yarn mass irregularity (e.g. Uster 

Tester) allow users, among others, to display the appearance 

of measured yarn wound on the board (yarn taper board), but 

they do not grade the yarn. Generally, on the yarn board, it is 

possible to identify the characteristic expression of yarn mass 

irregularity – the moiré effect (short-term periodical irregularity), 

stripiness (very long-term periodical irregularity). The nonperiodical 

irregularity of yarn expresses itself in the area (on 

the board) as unsettled appearance of textile. Furthermore, the 

appearance of yarns in the area is influenced by other random 

exposures, which are determined by the structure of yarn mass 

irregularity, yarn faults, impurities and yarn hairiness. Yarn 

irregularity can be described by parameters (CV, U, DR), which 

indicate a level of irregularity, and characteristic functions 

(a spectrogram, a variance-length curve), which can help in 

indicating the cause of irregularity in yarn. On the basis of these 

functions, we can predicate the appearance of future plain 

textile. In the literature, the interrelation between the course of 

spectrogram (the periodical irregularity) and the moiré effect as 

well as the stripiness [8] and between the course of variancelength 

curve and unsettled appearance of plain textile [9] is 

mentioned. 

Several research works have been carried out to present the 

objective evaluation of yarn appearance. For example, the work 

by Kim et al. [10] described a developed quantitative method 

for grading spun yarn appearance derived from optical yarn 

diameter measurements. Semnani et al. [11,12] introduced a 

http://www.autexrj.com 

1


method for grading the appearance of various types of yarn 

using image analysis and artificial neural network. The work 

by [13], described a new device for evaluating a yarn on the 

surface as an independent formation. Recently, Liang et al. 

[14] presented intelligent characterization and evaluation of 

yarn surface appearance using saliency map analysis, wavelet 

transform and fuzzy neural network. Another method presented 

by Rong et al. [15] graded yarns by cluster analysis. 

The derived statistical function [16,17] can be used for the 

objective evaluation of surface unevenness. For the evaluation, 

both generated images of the appearance of the textile in area 

(yarn taper board, woven fabrics, knitted fabric), and images 

of real fabrics or yarn board can be used. Simulated image 

of the appearance is in grayscale with different levels of 

gray, the real image of gray textile is converted into grayness 

degrees. Grayness degrees reflect the unevenness of textile 

and in the case of yarn, its faults. Thus, surface unevenness 

of a textile can be converted into the unevenness of color 

of fabric images. The yarn mass irregularity and yarn faults 

also present non-uniformity in a color image. The fluctuation 

of average grayness degrees in the image can be evaluated 

by means of statistical functions. Essentially, a sample of 

the flat textile is divided into square fields, where individual 

properties (grayness degrees) are measured. The so-called 

area-variation curves can be constructed as a parallel of a 

variation-length curve. The area variation curve is also used 

in works [18,19] as a quantitative evaluation of the quality of 

predicated image of the plain textile. It has been suggested 

as a new evaluation method of woven fabric unevenness [20]. 

Surface variability can also be described by other statistical 

functions, for example the so-called directional semivariograms 

[17,21,22]. The semivariogram was used for evaluation of the 

surface variability of woven and nonwoven fabrics [23-25] and 

it was also applied on evaluating the appearance of standard 

yarn boards from the standard CSN 80 0704 as well as real 

yarn boards with faultless and faulty yarn [1]. 

The new suggested methods for the objective evaluation of 

yarn appearance are analyzed in this paper. Considering 

the same yarn count and constant winding density, the 

appearance of yarn wound on the board is influenced mainly 

by variations in the yarn diameter, yarn hairiness, number 

of yarn faults, remains of impurities as well as yarn mass 

irregularity. We suppose that the appearance of yarn transfers 

itself into fluctuation in degrees of grayness after digitizing 

and converting the yarn board to a grayscale image. In the 

presented work, the variation of degrees of grayness in an 

obtained image of the yarn board appearance is evaluated 

using a spatial statistical function called the area variation 

curve. This method is applied to the same yarn boards 

(standard and real) as in work [1]. In addition, the simulated 

appearance of yarn taper boards generated by the Uster 

Tester IV-SX device is used. The behavior of the constructed 

area variation curves in dependence on the results of visual 

evaluation of the yarn board appearance is discussed. For the 

verification of results described in [1], the appearance of the 

generated yarn boards is evaluated by semivariograms too. 

Employing both these functions for the objective evaluation of 

the yarn board appearance is discussed here. 

Tools for yarn board appearance evaluation 

Area variation curve 

The area variation curve describes the variability of 

grayness degrees in dependence on the square field area. 

It can be expressed as an external or an internal curve. 

The internal area variation curve records the variation 

coefficient of grayness degrees inside the square area in 

dependence on the area of the observed square fields. This 

curve increases with the growing area of square fields. The 

external variation curve shows the variability of grayness 

degrees between the square field areas of an image. The 

curve slopes down with the growing area of square fields. 

In this work, we used the external area variation curve 

which is calculated as: 

S( 

A) 

CVB ( A) 

= (1) 

X ( A) 

where CVB(A) is the external variation coefficient of average 

grayness degrees between the square fields of area A in the 

fabric image; S(A) is the standard deviation of the mean 

values of grayness degrees in the square fields of area A 

included in a fabric image; X (A) is the mean value from all 

mean values of grayness degrees in the square fields of 

area A. 

Semivariogram 

The statistical function semivariogram can be used for 

evaluating the variability of random field properties. In this 

case, the yarn board was converted into grayness degrees and 

divided into square fields like a net. The centers of the fields 

are the locations x. The average value of grayness degree 

in the given square field is assigned to location x (z(x i 

)). The 

semivariogram expresses spatial dissimilarity between the 

values of grayness degrees at points x i 

and x j 

, and its general 

definition is mentioned in works [23-26]. We used the so-called 

centered sample semivariogram [22]: 

1 

G( lag ) = 

2N( lag ) 

N( lag ) 

∑ 

i= 

1 

( zc ( xi 

) − zc 

( xi 

+ lag ) 

2 

(2) 

where z c 

(x i 

) is the centered average grayness degree defined 

as: 

z ( x 

c 

i 

) = 

n( xi 

) 

∑ 

i= 

1 

z( x 

n( x 

N(lag) is the number of pairs of observations separated by 

distance lag and z(x i 

) is the grayness degree in location x i 

. 

Four types of semivariograms can be constructed in direction 

of columns, rows, diagonals and omni-semivariogram as an 

average of the mentioned three types of semivariogram. 

i 

i 

) 

) 

(3) 

http://www.autexrj.com 2


Experimental part 

For the experiment we used: 

a) Appearance of yarn taper boards and appearances of 

magnified yarn boards generated by the Uster Tester IV-SX 

apparatus based on measurements of yarn mass irregularity. 

These boards are one of the outputs of this device; the 

magnified yarn board shows a zoomed part of the regular taper 

board, with a different pitch [27] (for example see Figures 1a 

and 1b). 

Basic characteristics of used 100% CO single yarns of various 

types of periodical faults identifiable from spectrograms and 

yarns without periodical fault are mentioned in Table 1. Their 

spectrograms and generated yarn taper boards as well as 

magnified yarn boards are mentioned in Table 2. 

b) Real standard yarn boards – (called etalons) from CSN 80 

0704 - grade A to F (for example see Figures 2a and 2b). 

c) Real yarn boards of two qualities – 100% CO rotor spun 

yarns of fineness 30 tex without faults and with short-term 

irregularity caused the moiré effect (see Figures 3a and 3b) 

were wound on the black board with the same winding by the 

Planiscop device. 

The images of yarn appearance generated by the Uster Tester 

apparatus were printed and scanned. The real yarn boards were 

scanned. The scanning resolution was 300 dpi and the obtained 

images were saved in a non-compressed tiff format. The 

images were treated in the script “Fabric unevenness” written 

by Militký, J. (Technical University of Liberec) in programming 

environment Matlab. The program converts images to grayscale. 

The area variation curve and semivariograms are the output 

Figure 1a. Yarn taper board generated by Uster Tester apparatus - 

example, 100% CO yarn - fineness 30 tex - real size 1741 x 

1048 pxl, resolution 300 dpi. 

Figure 1b. Magnified yarn board generated by Uster Tester apparatus 

- example - 100% CO yarn - fineness 30 tex, real size 1763 

x 1049 pxl, resolution 300 dpi. 

Table 1. Short description of yarn used for experiment. 

Yarn 

No. 

Yarn count 

T [tex] 

CV m 

[%] 

CV(1m) 

[%] 

CV(3m) 

[%] 

CV(10m) 

[%] 

Thin place 

-50% [1/km] 

Thick place 

+50% [1/km] 

Neps +200% 

[1/km] 

1 30 18,29 5,2 4,42 3,98 350 510 0 

2 30 17,54 7,5 6,98 6,28 47,5 277,5 35 

3 25 15,76 4,6 3,54 2,68 111 300 191 

4 20 14,3 6,62 5,96 4,42 2,5 77,5 47,5 

5 29,5 11,73 3,17 3,12 1,36 0 10 45 

Figure 2a. Standard yarn board (CSN) – grade A, 100% CO yarn - 

fineness 30 tex; original size 1547 x 2677 pxl, resolution 

300 dpi. 

Figure 2b. Standard yarn board (CSN) – grade F, 100% CO yarn - 


300 dpi. 



Table 2. Spectrograms and generated yarn boards of yarn used for experiment. 

Yarn 

No. 

Spectrogram 

Yarn board and its subjective 

evaluation 

Magnified yarn boards 

1 

CV m 

= 18,29 % 

Combination of cumulous spectrum and 

characteristics spectrum (a chimney) on 

short wavelength 

Moiré effect and short-term stripiness 

2 

CV m 

= 17,54 % 

Typical cumulous spectrum on short 

wavelengths 

Short-term stripiness 

3 

CV m 

= 15,76 % 

Characteristic spectrum on short 

wavelength 

Clear moiré effect 

4 

CV m 

= 14,3 % 

Cumulous spectrum on long wavelengths 

Stripiness 

5 

CV m 

= 11,73 % 

Spectrum without fault 

Figure 3a. Real yarn board – yarn without fault, 100% CO yarn – 


300 dpi. 

Figure 3b. Real yarn board - yarn with moiré effect, 100% CO yarn – 


300 dpi. 



of the program. The area variation curve is constructed from 

external variation coefficients of grayness in dependence on 

observed square areas according to formula (1). All presented 

yarn boards are used. The image is divided into squared fields 

with the consequently growing area. Minimum number of 

observed square fields into sample was 100; maximum square 

field area was 2.19 cm 2 . In the case of semivariograms, applied 

to yarn board generated from Uster Tester here, the image is 

divided into square fields of selected size step x step pixels. 

The average grayness degree (z(x i 

)) is calculated in each 

field. From the obtained values, the centered semivariogram in 

the given direction is calculated according to formula (2). We 

used the step 3 pxl and step 30 pxl. We selected the step 3 

pxl because it corresponds to the yarn width in the image, and 

step 30 because it corresponds to 0.25 cm in the image; in our 

opinion it is the smallest possible area which a human eye can 

see and evaluate. 

Results and its discussion 

Yarn taper boards and magnified yarn boards generated 

by Uster Tester 

The area variation curves of grayness degrees constructed 

from the yarn taper boards and the magnified yarn boards are 

mentioned in Figures 4a and 4b. 

The courses of area variation curves of grayness degree 

constructed from the yarn taper board generated by the Uster 

Tester IV-SX are very similar (see Figure 4a). It is obvious, from 

this figure, that the curve for faultless yarn (No. 5) lies lower 

than others. In the case of regular yarn, the yarn taper board 

has the best appearance; grayness degrees fluctuate less 

compared to the yarn board with irregular yarn. Thus, the area 

variation curve for this yarn taper board lies lowermost. The 

curves for faulty yarns (faulty yarn taper boards) overlap and 

there is not much difference between them. It means the area 

variation curve is not a suitable tool for recording these faults in 

the yarn board generated by Uster Tester IV-SX. 

The area variation curves of grayness degrees constructed 

from the magnified yarn boards fluctuates periodically. The 

reason is that the magnified yarn board is an enlarged part of 

the yarn taper board and thus has more visible yarn winding on 

the board compared to the non-magnified one. Therefore, we 

can say the curve records the yarn winding on the board. There 

is no difference among curves in terms of yarn irregularity. In 

neither of these cases is the area variation curve a suitable tool 

for the evaluation of yarn taper board generated by the Uster 

Tester IV-SX. 

Directional semivariograms of grayness degrees constructed 

from yarn taper boards and magnified yarn boards are 

mentioned in Figures 5a, 5b, 6a and 6b. From each image, 

the area of size 650 x 650 pixels was evaluated in the case of 

the yarn taper board, and the area 1000 x 1000 pixels in the 

case of the magnified yarn board. The observed area was in 

the center of the image. 

The courses of semivariograms of grayness degrees constructed 

from the yarn taper boards with step 3 pxl are also similar (see 

Figure 5a). In works [28] and [1], it was found out that increasing 

the linear character of the course of semivariogram curve in the 

direction of the columns in combination with periodic course 

of semivariogram in the direction of row records longitudinal 

stripes in the image of textiles. In this case, semivariograms 

with step 3 pxl has a similar character (see Figure 5a). It can 

be said, stripes in the yarn boards generated by Uster Tester 

were also recorded. Stripes express winding of the yarn on 

the board, but the yarn irregularity was not probably recorded. 

This holds also in the case of semivariograms constructed from 

magnified yarn boards (see Figure 6a). 

Curves of semivariograms with step 30 pxl have a little different 

behavior compared to semivariograms with step 3 pxl; they also 

differ in their location in the graphs. In this case, the position of 

the curves of semivariograms of grayness degrees from yarn 

taper boards (see Figure 5b) corresponds to the quality of yarn 

on the board. The curve of yarn without faults lies lowest in 

the case of all types of semivariograms, whereas the curve of 

yarn containing cumulous spectrum and short stripiness has 

the highest values. As already mentioned, the irregularity of 

yarn on the board is converted to the grayscale. In the case of 

irregular yarn, the color image of yarn board is unbalanced. It 

contains large number of areas with white color representing 

Figure 4a. Area variation curve – yarn taper board generated by Uster 

Tester IV-SX. 

Figure 4b. Area variation curve – magnified yarn board generated by 

Uster Tester IV-SX. 



Figure 5a. Semivariograms – yarn taper board generated by Uster Tester IV-SX– step 3 pxl - zoomed. 

Figure 5b. Semivariograms – yarn taper board generated by Uster Tester – step 30 pxl. 

the yarn and its irregularity. The fluctuation of average grayness 

degrees in observed squared fields is higher compared to yarn 

without faults. Therefore, the curves of semivariograms lie 

higher in the graph. This fact is most visible on semivariogram 

in the direction of rows, columns and omni-semivariogram. 

In the case of semivariograms with step 30 pxl constructed 

from the magnified yarn boards (see Figure 6b), the 

curves of semivariograms in direction of columns have a 

growing character, their course diverges from the linear 

one with increasing irregularity of yarn. Therefore, curves of 



Figure 6a. Semivariograms – magnified yarn board generated by Uster Tester IV-SX – step 3 pxl. 

Figure 6b. Semivariograms – magnified yarn board generated by Uster Tester IV-SX – step 30 pxl. 

semivariograms from the magnified yarn board No. 4 and No. 

5 (the lowest value of yarn irregularity) show the best course, 

while curves from the yarn boards No. 1 – No. 3 (irregular yarns) 

have higher values with increasing distance between squares 

(the parameter lag). Their course is not linear. This is because 

the magnified yarn boards with higher yarn unevenness 

exhibit more unsettled appearance in comparison with the 

yarn taper board. Due to higher irregularity of yarn displayed 

on the board, the variation of grayness degrees in individual 

square fields in direction of columns is higher in comparison 



with the regular yarn. Therefore, the course of semivariogram 

in direction of columns deviates from the ideal one. The curves 

of semivariogram from the magnified yarn boards in direction 

of rows have the same periodical shape with different height 

of peaks and lows. With the increasing yarn irregularity, the 

curves achieve greater differences between peaks and lows. 

The curves belonging to the yarn with higher irregularity have 

higher values too. It is due to the fact that the zoomed part 

of the yarn taper board is recorded on the magnified yarn 

board. Thus, the individual yarns are much more marked 

in the magnified yarn boards in comparison to yarn taper 

boards. Thus, semivariograms partially recorded winding the 

yarn on the board. However, based on the courses of these 

semivariograms it is not possible to identify the type of yarn 

irregularity. From the curves of semivariograms mentioned 

above, it is evident that the size of the observed square fields 

(the parameter step) influences the number of points of the 

curve and the course of the curve. With increasing step size, 

the curve is smoother. This corresponds to a well-known fact 

that the variability of properties decreases with increasing area. 

Real standard yarn boards (etalons) 

Area variation curves of grayness degrees constructed from 

real standard yarn boards (etalons) from CSN 80 07 04 (yarn 

fineness 30 tex) grade A to grade F are shown in Figure 7a. 

The curve from etalons (grade A) of all fineness is shown in 

Figure 7b. 

Figure 7a Area variation curves – etalons (A-F) - yarn fineness 30 tex. 

The curves from etalons (grades A to F, see Figure 7a) are 

similar. The positions of the curves are nearly identical, with the 

exception of etalon F (the worst appearance), where the curve 

has the highest values. Comparing curves from etalons (grade 

A, Figure 7b), we can see that the behavior of the curves is 

dependent on yarn fineness on the etalon as well as yarn 

winding. 

The results showed that the area variation curve is not a 

suitable tool for objective evaluation of the appearance of the 

yarn wound on the board even in this case. The curves from 

etalons (grades A to F, see Figure 7) are similar. The positions 

of the curves are nearly identical, with the exception of etalon 

F (the worst appearance), where the curve has the highest 

values. 

Real yarn boards 

The area variation curve of grayness degree constructed from 

the real yarn board image is mentioned in Figure 8 (red and 

blue curves). 

The area variation curve from the real yarn board with the yarn 

without fault has lower values and shows regular fluctuation 

compared with the curve from the real yarn board with moiré 

effect (see Figure 8). The regular fluctuation of the curve is 

caused by individual winds of regular yarn (winding), which 

the curve records. In the case of irregular yarn (moiré effect), 

this fluctuation is subdued due to irregularity of yarn. Thanks 

to the moiré effect, the appearance of yarn on the board is 

unbalanced it contains a lot of periodically repetitive thick and 

thin places. Therefore, average grayness degrees vary more 

between the observed square fields, and due to this the area 

variation curve has higher values compared to the curve from 

the yarn without fault. But, we must say that the difference 

between the positions of both area variation curves is not too 

significant. 

We can compare the behavior of the area variation curve with 

curve of ideal yarn board. For this reason, the simulated ideal 

yarn board was constructed (see Figure 9a). The white stripes 

in the image represent a yarn. It is considered the absolute 

ideal state; it means neither irregularity nor hairiness of the 

yarn is taken into account. The stripe width corresponds to the 

Figure 7b. Area variation curves – etalons (grade A). 

Figure 8 Area variation curves – real yarn boards – yarn fineness 30 

tex. 



Figure 9a. Zoomed part of the simulated ideal yarn board - 242 x 204 

pxl, resolution 300 dpi. 

Figure 9b. Zoomed part of the real yarn board of yarns without faults - 

242 x 204 pxl, resolution 300 dpi. 

yarn width on the real yarn board – i.e. 5 pxl. The width of black 

stripes corresponds to the distance between winds of yarns; 

it is 7 pxl. The colors were gained from the image of the real 

yarn board. 

The area variation curve of grayness degrees constructed from 

a simulated ideal yarn board is mentioned in Figure 9 (black dot 

curve). The periodical course of the curve records individual 

winds of yarn on the board. The period nearly corresponds 

to the period of area variation curves of grayness degrees 

constructed from the real yarn board with the faulty yarn and 

yarn without fault. 

The courses of directional semivariograms of grayness degrees 

for these real yarn boards together with semivariograms of 

real standard yarn boards were presented and discussed in 

work [1]. The behavior of these curves constructed from the 

simulated ideal yarn boards was also analyzed. The course of 

semivariogram was influenced by yarn fineness and winding 

density in combination with the parameter step. It was found 

out that a small size of the parameter step is less suitable for 

evaluation of the yarn appearance on used yarn board. The 

course of semivariogram recorded more the structure of yarn 

winding on the board – regularity of winding. With deteriorating 

grade of yarn appearance, the curves of semivariograms with 

used step 15 pxl showed higher values and fluctuating course. 

Conclusion 

A study of using the area variation curve for the evaluation of 

the appearance of yarn wound on the board was presented 

with the aim to extend the obtained piece of knowledge. For this 

reason, we used the same etalons of yarns from CSN 80 0704, 

the same real yarn boards with yarn of various quality as in 

work [1] for which this article follows. To enlarge and confirm the 

results, we also used images of the yarn taper board in addition 

to the magnified yarn taper board appearance generated by 

the Uster Tester IV-SX instrument based on measurement of 

yarn with various irregularities. These generated boards were 

also evaluated by means of directional semivariogram from the 

point of view of yarn irregularity. All yarn boards were scanned 

and then, using the script in Maltab, the area variation curves 

and directional semivariograms (only in the case of generated 

yarn board) were constructed. These functions express 

fluctuation of grayness degrees of image in dependence on the 

size of evaluated square fields or on various distances between 

square fields of the pre-set size. Even though in some cases 

the area variation curve can be used for evaluation of surface 

unevenness of woven fabric [20], based on these results we 

can say that the area variation curve is not a suitable tool for the 

evaluation of yarn board appearance. The difference between 

curves constructed from the boards of deteriorating quality 

is not much significant. Semivariograms seem to be a more 

suitable tool for the evaluation. The position and behavior of the 

curve of semivariogram corresponds to quality of yarn on the 

board, but the type of irregularity is not possible to be identified. 

Also the piece of knowledge about behavior of semivariograms 

presented in work [1] was confirmed here. 

Acknowledgment 

This work was supported by institutional research program 

initiated by the Faculty of Textile Engineering, Technical 

University of Liberec. 

References 

[1] Moučková, E., Jirásková, P.: New possibility of objective 

evaluation of yarn appearance, AUTEX Research Journal, 

Vol. 12, No. 1, p. 7 – 13, ISSN 1470-9589, 2012. 

[2] USTER® STATISTICS 2013, http://www.uster.com, 

Accessed 28. 1. 2013. 

[3] ASTM Standard Test Method for Grading Spun Yarns for 

Appearance, 2255-90, 1990. 

[4] ČSN 80 0704 Determination of yarn appearance, 1993. 

[5] Bragg, C.K, Wessinger, J.D.: Instrument measurements of 

yarn appearance, Beltwide cotton conference, Vol. 2. p. 

1432-1434, ISSN 1059-2644, 1995. 

[6] Lawson-Hemphill: YAS – Electronic inspection board, Yarn 

analysis system, http://www.lawsonhemphill.com/LH-481- 

yarn-analysis-system.html. Accessed: 2012-07-18. 

[7] Wang, J., Yunming, L.: Evaluation of Yarn Regularity Using 

Computer Vision. 9th International conference on Pattern 

Recognition, vol. 2, p. 854-856, 1988. 

[8] Uster News Bulletin, Ed. by K. Douglas, B.Sc., C. Eng., M. 

I. Mech. E., C. Text., A. T.I., 1971-, No.15. Zellweger Uster, 

Switzerland, 1971. 

[9] Uster News Bulletin, Ed. by K. Douglas, B.Sc., C. Eng., M. 

I. Mech. E., C. Text., A. T.I., 1988-, No 35. Zellweger Uster, 

Switzerland, 1988. 

[10] Kim, Y. K.; Langley, K. D.; Avsar, F.: Quantitative grading of 

Spun Yarns for appearance, Journal of Textile Engineering, 

vol. 52, No. 1, p. 13-18, 2006. 



[11] Semnani, D.; Iatifi, M.; Tehran, M. A.; Pourdeyhimi, B.; Merati, 

A. A.: Development of appearance grading method of cotton 

for various types of yarns, RJTA, Vol. 9, No. 4, 2005. 

[12] Semnani, D., Latifi, M., Tehran, M.A., Pourdeyhimi, B., 

Merati, A. A.: Grading yarn apperance using Image Analysis 

and Artifical Inteligence Technique, Textile Research 

Journal, Vol. 76(3), p. 187-196, ISSN 1746-7748, 2006. 

[13] Beran, L.; Šrámek, R., Vyšanská, M., Horčička, J., et all: 

Optical evaluation of a yarn being continuously wound 

on the surface, 16th International Conference Structure 

and Structural Mechanics of Textiles, December 2009, 

Technical University of Liberec, Liberec, 2009. 

[14] Liang, Z., Xu, B., Chi, Z. and Feng, D.D.: Intelligent 

characterization and evaluation of yarn surface appearance 

using saliency map analysis, wavelet transform and fuzzy 

ARTMAP neural network. Proceedings of Expert Syst. 

Appl.. p. 4201-4212, 2012. 

[15] Rong, G.H., Slater, K., Fei, R.:The use of Cluster Analysis 

for Grading Textile Yarns, Journal of the Textile Institute, 

Vol. 85 (3), p. 389-396, ISSN 1754-2340, 1994. 

[16] Militký, J.: Probability model of nonwovens unevenness, 

Proceedings of 7th international conference Structure 

and Structural Mechanics of Textile, p.193-198, Technical 

University of Liberec, Liberec, 2000, ISBN 80-7083-442-0. 

[17] Militký, J.; Klička, V.: Characterization of textiles mass 

variation in plane, Proceedings of 5th world Textile 

Conference Autex 2005, p. 750-755, ISBN 86-435-0709-1, 

University of Maribor, Faculty of Mechanical Engineering, 

Department of Textiles, Maribor, 2005. 

[18] Suh, M., W.: An electronic Imagining of Fabric Qualities 

by on-line yarn data, Available from www.ntcresearch.org/ 

pdf-rtps/AnRp01/I01-A1.pdf Accessed: 2005-02-01. 

[19] Suh, M., W.: An electronic Imagining of Fabric Qualities by 

on-line yarn data, Available from www.ntcresearch.org/pdfrtps/AnRp01/S01-NS12-A2.pdf 

Accessed: 2005-02-01. 

[20] Jirásková, P., Moučková, E.: New method for the evaluation 

of woven fabric unevenness, AUTEX Research Journal, 

Vol. 10, No. 2, p. 49 – 54, ISSN 1470-9589, 2010. 

[21] Březina, M.; Militký, J. Complex characterization of textile 

surface, Robust’2002 – proceeding of twelfth winter school 

JČMF, p. 50 –58, Hejnice, January 2002, Union of Czech 

Mathematicians and Physicists, Prague, 2002. 

[22] Militký, J.; Rubnerová, J.; Klička, V.: Spatial statistics 

and unevenness of surface mass of non-woven textiles, 

Proceeding of 7th national conference Strutex, p. 199- 

203, ISBN 80-7083-668-7, Technical University of Liberec, 

Liberec, 2002. 

[23] Jirásková, P.; Moučková, E.: Unevenness of flat textiles 

and its quantification, Proceedings of 3rd International 

Textile, Clothing & Design Conference – Magic World 

of Textiles, p. 612-617, ISBN 953-7105-12-1, Faculty of 

Textile Technology, University of Zagreb, Zagreb, 2006. 

[24] Moučková, E., Jirásková, P., Ursíny, P.: Surface 

unevenness of fabric, in Woven Fabric Engineering, Sciyo, 

Rijeka, Croatia, 2010. 

[25] Militký, J., Klička, V.: Some tools for Nonwovens uniformity 

description, Proceedings of 4th International Textile, 

Clothing & Design Conference – Magic World of Textiles, 

p. 842-848, ISBN 978-953-7105-26-6, Faculty of Textile 

Technology, University of Zagreb, Zagreb, 2008, 

[26] Cressie, N.A.C.: Statistics for spatial data, J. Wiley, ISBN 

0-473-00255-0, New York, 2002. 

[27] Application manual of Uster Tester IV, Uster Technologies, 

Uster, Switzerland, 2002. 

[28] Moučková, E.; Jirásková, P.: Utilization of semivariogram 

for evaluation of surface unevenness. Book of Proceedings 

of 4nd international Textile, Clothing & Design conference 

- Magic World of Textile, p. 848-853, Dubrovník, Croatia, 

ISBN: 978-953-7105-26-6, Faculty of Textile Technology, 

University of Zagreb, Zagreb, Croatia, 2008. 



PREDICTION OF WOVEN FABRIC PROPERTIES USING SOFTWARE PROTKATEX 

Brigita Kolčavová Sirková, Iva Mertová 

Technical University of Liberec, Faculty of Textile Engineering, Department of Textile Technologies, Liberec, Czech Republic 

Studentská 2, 461 17 Liberec 1 

E-mail: brigita.kolcavova@tul.cz, iva.mertova@tul.cz 

Abstract: 

Keywords: 

Fabric properties and fabric structure prediction are important in each industry domain. Generally all professional 

CAD packages for woven textiles system will be able to achieve basic fabric simulation and production output. A 

good CAD system should enable you to create design (dobby and jacquard woven fabric) ideas quickly and easily 

to enhance the way you work. The differences among competing systems fall mainly into the following categories: 

ease of use; speed of operation; flexibility of operation; advanced features; technical support; and ongoing software 

development. Computer simulation or prediction is oriented on standard woven fabrics, technical textiles, and 

composites. This article focuses on the presentation of software ProTkaTex and its use in the prediction of woven 

fabric properties. The software implements a generalized description of the internal structure of woven fabric on 

the unit cell level, integrated with mathematical models of the fabric relaxed state. User can calculate selected 

mechanical and end-use properties of dobby and jacquard woven fabric as well as can evaluate fabric behavior 

before real weaving. The major challenge is to develop software that industry will use in design centers for creation 

and development of new fabric structures for technical as well as clothing application. 

Fabric geometry, weave, warp, weft, property, simulation, prediction 

Introduction 

Computer-aided design (CAD) is the use of computer technology 

for the process of design and design documentation. CAD 

software, or environments, provides the user with input-tools 

for the purpose of streamlining design processes, i.e. drafting, 

documentation, and manufacturing processes. CAD output is 

often in the form of electronic files for print or manufacturing. 

CAD environments often involve more than just shapes. 

As in the manual drafting of technical and engineering 

drawings, the output of CAD must convey information, such as 

materials, processes, dimensions, and tolerances, according 

to application-specific conventions. CAD is an important 

industrial art extensively used in many applications, including 

automotive, shipbuilding, and aerospace industries, industrial, 

textile and architectural design, prosthetics, and much more. 

CAD is also widely used to produce computer animation for 

special effects in movies, advertising, and technical manuals 

[1]. All professional CAD packages for woven textiles will be 

able to achieve basic fabric simulation and production output. 

The difference among competing systems falls mainly into 

the following categories: ease of use, speed of operation, 

flexibility of operation, advanced features, technical support, 

and ongoing software development [12-15]. These systems 

are compatible with electronic dobby or jacquard shedding 

mechanism; production data can be sent directly to electronic 

loom controllers. 

A good CAD system should enable you to create design ideas 

quickly and easily to enhance the way you already work. 

It should be able to fit seamlessly into your current working 

practices and become a valuable timesaving, creative tool that 

works with you [14]. 

CAD systems support the creation of everything, from simple 

single cloth structures to complex layered architectures, 

containing stacking orders and variable densities as well as 2D 

and 3D visualization of fabrics. 3D simulation module creates 

fabric visualization on the basis of 3D weave, yarn thickness, 

warp and weft density. Warp and weft parameters may be set 

and displayed individually. Separate layers can be isolated 

and edited on its own, or the whole fabric can be viewed and 

manipulated as a whole [12]. 

CAD systems for prediction of fabric properties 

The aim of CAD system for fabric properties prediction is to 

optimize the construction of textiles on the basis of virtually 

created fabrics [20]. Prediction of fabric properties is based 

on a combination of mathematical modeling and experimental 

research, including development and application of nonstandard 

methods for definition and measuring of the fabric 

structure [16,19,22,24]. Mathematical models are possible 

to use for modeling the geometry of textile structures as 

well as properties, including textile mechanics, permeability, 

drape, roughness, and composite mechanical behavior 

[24-27]. 

At present, the software packages implement a generalized 

description of the internal structure of textile reinforcements 

on the unit cell level, integrated with mechanical models of 


11

auteX research journal, vol. 13, No 1, March 2013, Doi: 10.2478/v10304-012-0017-5 © auteX 

the relaxed and deformed state of 2D- and 3D-woven (see 

Figure 1), two- and three-axial braided, weft-knitted and 

non-crimp warp-knit stitched fabrics [16,18,38]. Realistic 

models have application in the design and manufacture of 

textile composites. A composite material is one that is made 

by combining two existing materials: in a fi ber-reinforced 

composite, stiff, strong fi bers form one part of the composite, 

reinforcing the other. In manufacturing, it is common for this 

reinforcing material to be supplied in textile form, woven from 

‘yarns’ made from the fi bers [22,23]. 

Some modules of CAD system are focused on simulation of 

3D-woven fabrics with consideration of weaving technological 

boundary conditions: quick check up of the feasibility of 

geometry from 3D-woven fabrics; prediction of woven 3D 

structure and quality for each position of geometry; optimization 

of parameters of the 3D-woven fabric (thread sizes, weave, 

and thread distances) by simulation [19]. The packages are 

dedicated for the design and manufacture (CAD/CAM) of 

advanced textile structures based on the use of conventional 

weaving technology. It has been used for the design and 

manufacture of 3D textile structures, with both solid and hollow 

architectures, and non-crimp composite reinforcement [22,23]. 

Software packages can be used for the prediction of standard 

gray fabric properties for technical and clothing applications. 

In this case, the software can predict selected properties and 

fi ber parameters in module fi ber, selected yarn properties and 

parameters in module yarn, selected fabric properties and 

parameters in fabric module. The system contains databases of 

fi ber properties and fabric weaves, and the prediction is based 

on the complex of theoretical and regression models. The 

material and technological parameters for different materials, 

yarns and fabrics are included. The system is mainly used for 

the optimization of fabric design based on virtually created 

fabric [20]. 

and evaluation of individual interlacing pores in repeat defi nes 

the fabric structure and distribution of interlacing points, which 

infl uence mechanical and end-use properties. The software 

consists of three modules: yarn defi nition, fabric properties, 

and weaves (design). For the prediction of selected woven 

fabric properties, it is necessary to know basic input yarn 

properties. Warp and weft threads are defi ned on the basis 

of: type of yarn, fi ber packing density μ [], yarn count T [tex], 

specifi c density r [kg/m 3 ], yarn strength Fpr [N or N/tex], yarn 

elongation Epr [%], and yarn irregularity CV [%]. Defi nition of 

fi ber packing density: inside of the textile fi brous assembly or 

inside of some spatial part of them, there lies fi ber volume V. 

Total volume of this body is called Vc. The compactness of this 

body is characterized by the ratio between these two volumes 

and is known as fi ber packing density. (Note: alternatively, this 

value is called the packing ratio or volume fraction.) Evidently, 

the fi ber packing density value must lie in the interval from 0 

to 1 [4]. Type of yarn is evaluated by technology as well as 

yarn structure (staple yarn, multifi lament yarn, etc.). Software 

ProTkaTex distinguishes the technologies: combed, carded, 

and open-end. Based on experimental and theoretical research 

work [4,32,35], different types of yarn were analyzed (on the 

basis of threads cross-section evaluation, see Figures 2 and 3) 

and their construction parameters, and mathematical equations 

were prepared for yarn parameters prediction. Theoretical 

expressions of individual parameters of yarn (fi ber packing 

densities, compression of yarn in relaxing state before weaving 

as well as in fabric, diameter of threads) were compared with 

experimental parameters that followed from real cross-section 

of yarn and woven fabric [5-7]. 

Prediction of woven fabric properties: software 

Protkatex 

ProTkaTex software can be applied for the prediction of 

selected mechanical and end-use woven jacquard fabric 

properties with knowledge of the yarns’ properties and fabric 

interlacing. ProTkaTex was developed at the Faculty of Textile 

Engineering, Technical University of Liberec. The software is 

compatible with common CAD jacquard and dobby weaving 

design systems. ProTkaTex enables 3D visualization of fabrics 

based on the input yarn parameters. It is used for properties, 

prediction on the basis of combination of mathematical 

modeling and experimental research [2,4,31]. The major 

challenge is to develop a software package that industry will 

use in design centers for the creation and development of new 

fabric structures for technical as well as clothing application. 

In comparison with the above-mentioned software packages, 

this software is focused on the determination of standard dobby 

fabrics as well as jacquard fabrics, and is not focused on the 

description of textile composite. At this stage, the weave (small 

and big patterns) covers the structure of single-layer woven 

fabrics. It is able to evaluate the jacquard pattern with unlimited 

number of warp and weft threads. On the basis of defi nition 

Figure 1. Modeled deformed twill weave unit cell in tension [17]. 

Figure 2. Defi nition of yarn diameter in cross-section. 

Figure 3. Modeling of cross-section in fabric. 


12


Weave definition and prediction of woven fabric properties 

Woven fabric weave determines the manner of the thread’s 

interlacing in the fabric [2]. Defi nition of individual pores in 

weave – expression of pores frequency in weave (in rows and 

in columns) – is used for description of threads interlacing by 

mathematical equation in this software. In the woven fabric 

weave, exists only four structural models of interlacing [33], 

see Figure 4. Dobby as well as jacquard pattern is possible 

to create by various combinations of these structural pores in 

repeat. Jacquard design is not possible to create in ProTkaTex. 

The software is compatible with CAD jacquard design systems 

(EAT, NedGraphics, Arachne, ScotWeave, etc.), where pattern 

has to be saved in one of the following formats: TIFF, BMP, 

JPEG (see Figure 5). 

Jacquard weave in the above-mentioned format (see Figure 6) 

can be opened in module “weave” and then converted into a 

format required for fabric properties prediction in ProTkaTex. 

This software uses its own format, VZB. 

Description of woven fabric properties and parameters is based 

on the complex of theoretical and regression models. Prediction 

is based on the knowledge of yarn parameters and interlacing 

of individual threads in dobby and jacquard pattern. Calculation 

of individual fabric properties and mathematical formulations 

were created on the basis of analysis of areal and spatial fabric 

geometry. As mentioned above, the software calculates selected 

properties for dobby as well as jacquard woven fabric. It is 

possible to predict the following properties of fabrics: relative 

wave height [] (separately for warp and weft system), warp and 

weft density [treads/100mm], weight of fabric [g.m -2 ], cover of 

fabric [%] (separately for warp and weft system too), crimp of 

threads in fabric [%] (warp and weft), fabric thickness [mm], 

fabric roughness [μm], fabric strength [N/5cm] (for warp and weft 

direction), and fabric elongation [%] (for warp and weft direction). 

For presentation of individual results of prediction, the abovementioned 

properties were selected. In the above-mentioned 

graphs we can see fabric properties behavior and comparison of 

theoretical values with experimental values. Parameters of fabric 

samples: All fabrics are in plain weave. Material composition: 

100%PP yarn and 100%CO yarn in three counts 20tex, 29,5tex 

and 45tex in both directions. 

Weight of fabric 

Calculation of weight of fabric in this software is based on the 

warp and weft sett and on the yarn count as well as yarn crimp 

[31]. The software distinguishes two kinds of fabric weight: 

weight of linear meter of fabric [g.bm -1 ] and weight of square 

meter of fabric [g.m -2 ]. 

Fabric thickness 

Fabric thickness is defi ned as perpendicular distance to the 

fabric, which determines the dimension between the upper and 

lower side of the fabric. Prediction of fabric thickness depends 

on fabric geometry description. Mathematical description is 

based on defi nition of binding wave in repeat and threads’ 

waviness in interlacing [37]. Threads’ deformation in interlacing 

Figure 4. Principle of X-Ray Tomography. 




13


is dependent on the yarn structure and fabric input parameters 

(weave, warp and weft waviness, warp and weft sett) [4,33]. 

In the following model (1), the main infl uence is from yarn 

diameter and weft and warp waviness. 

thickness 

⎡ 

⎢ 

⎣ 

o u 

o u 

m 

[ m ] = ( d + d ) + .e1 

− ( . 1− 

e1) .f . β 

where d o 

– warp diameter, d u 

– weft diameter, e1 – warp 

waviness, f m - interlacing coeffi cient, β – yarn compression in 

fabric (the yarn deformation in interlacing). 

Fabric elongation 

o 

u 

⎡d 

+ d 

⎢ 

⎣ 2 

d + d 

2 

Prediction of fabric elongation in the warp and weft direction 

is defi ned as a fabric extension for maximum strength (the 

breakage) to the original length of fabric. Fabric elongation in 

warp and weft direction depends on yarn elongation and yarn 

interlacing in fabrics. Calculation of fabric elongation is based 

on defi nition of threads’ crimp and level of threads’ interlacing 

(which is necessary to evaluate the number of interlacing parts 

as well as fl oat part in repeat [31]). 

⎤ ⎤ 

⎥ ⎥ 

⎦ ⎦ 

Fabric strength 

Prediction of fabric strength in warp or weft direction depends 

on warp (weft) yarns and warp (weft) sett. The infl uence 

of second system on fabric strength value is neglected. 

Woven fabric strength does not correspond to the sum of 

yarn’s strength per fabric width unit in straining direction 

only. Relation between fabric and yarn strength is corrected 

by coeffi cient of utilization of yarn in fabric in warp (weft) 

direction. It is assumed that coeffi cient includes infl uence of 

material and fabric weave. 

Fabric roughness 

Roughness is a surface micro-geometry and is defi ned as 

the sum of unevenness (geometric deviations) of the surface 

with relatively small distances. It is an important parameter 

infl uencing subjective hand feeling and is connected with 

behavior of textiles’ layers in mutual contact. Calculation of 

roughness is based on the defi nition of structural pores in 

repeat, fabric geometry description, threads’ interlacing as well 

as yarn irregularity [37]. 






14


∑ 

⎛ p1 p2 

⎞ 

⎜ repeat 

repeat 

⎟ 

⎜ 

. Roughness p1 + . Roughness p2 

+ 

no. nu no. 

n 

⎟ 

u 

Fabric Roughness[ µ m] = ⎜ 

⎟ 

⎜ ∑ ( p3 A + p3B 

) 

∑ p4 

⎟ 

⎜ repeat 

repeat 

⎟ 

⎜ + . Roughness p3 A, B 

+ . Roughness p4 

no. nu no. 

n 

⎟ 

⎝ 

u 

⎠ 

100 100 3 

. . .10 

Do 

+ Du 

100 − CV 

2 

∑ 

where D u,o 

[threads / 100mm] – weft, warp density, 

p 1-4 

[] – pores in weave (definition of interlacing cell in repeat), 

CV [%] – yarn irregularity. 

Conclusion 

The ProTkaTex software provides an integrated description of 

the internal geometry of dobby and jacquard fabric and their 

properties. It is one of the few software packages that are able 

to characterize and determine the behavior of complicated 

jacquard patterns in selected properties. The major challenge 

is to develop a software product that industry will use in 

design centers for the creation and development of new fabric 

structures for technical as well as clothing application. 

At present are elaborated additional properties and their 

mathematical formulation that will be implemented in software 

and extend the existing possibilities of prediction. Software 

is open to be used in further developments of creation and 

prediction of new fabric structures and their properties. 


This work was supported by the project Textile Research 

Centre 1M0553 and by the project GACR 106/09/1916. 

References 

[1] Vijay Duggal. “CADD Primer”. Mailmax Publishing, 

http://www.caddprimer.com/cadd_primer_chapters/ 

[2] Nosek, S.: The structure and geometry of the woven 

fabrics, Liberec 1996, 

[3] Behera, B.K., Hari, P.K.: Woven textile structure, Theory 

and applications, Woodhead Publishing Limited, ISBN 

978-1-84569-514-9 (book), 2010 

[4] Neckář, B.: Compression and Packing Density of Fibrous 

Assemblies. Textile Research Journal, pp. 123-130 Vol. 67 

No. 3. 

[5] Plívová, H., Influence of interlacing and sett of threads 

on the yarns shape in the weave for the cotton woven 

fabric – only in Czech, Diploma thesis Faculty of Textile 

Engineering, Technical University of Liberec 2001 

[6] Nový O., Spectral analysis of binding waves in the woven 

fabric with carbon fibres, – only in Czech, Diploma thesis 

Faculty of Textile Engineering, Technical University of 

Liberec 2004 

[7] Čaprnková N.: Analysis of the woven fabric cross-section 

produced from twisted yarns – only in Czech, Diploma 

thesis Faculty of Textile Engineering, Technical University 

of Liberec 2007 

Figure 11: 11. Comparison Principle of of X-Ray experimental Tomography. and predicted fabric roughness 

[8] Kašparová K.: Design possibilities of woven jacquard 

fabric, Bachelor work, Faculty of Textile Engineering, 

Technical University of Liberec 2009 

[9] Peirce, F.T.: The Geometry of Cloth Structure, Journal of 

Textile Institute, Vol.28, 1937 

[10] Kemp, A. J.: Textile Institute 49, T44, 1958 

[11] Duckett, K. E., Cheng C. C.: A Discussion of The Crosspoint 

Theories, Journal of the Textile Institute, Volume 69, 

Issue 2 & 3 February 1978, pp. 55-59. 

[12] CAD/CAM systems for weaving - ScotCad Textiles Limited, 

www.scotweave.com. 

[13] CAD/CAM systems for weaving EAT - 

www.designscopecompany.com. 

[14] CAD/CAM systems for weaving -NedGraphics, 

www.nedgraphics.com. 

[15] CAD /CAM systems for weaving - Arachne, www.arahne.si 

[16] Lomov, S., Verpoest, I., Modelling of the internal 

structure and deformability of textile reinforcements: 

WiseTex software, In Proc. of 10th European Conf. 

Composite Materials (ECCM-10) (Brugge, Belgium, 

June 3--7, 2002) 

[17] Lomov, S. V., G. Huysmans, and I. Verpoest: Hierarchy 

of Textile Structures and Architecture of Fabric Geometric 

Models, Textile Research Journal, Jun 2001; vol. 71 

[18] Košek M., Mikolanda T., Košková B.: Ideal, Real and Virtual 

Textile Structure Modeling and Visualization, Afriograpf 

2004, Proc. Of 3rd International conference on computer 

graphics, South Africa 2004 

[19] CAD-Simulation of 3D woven shapes, Department of 

Textile and Clothing Technology, Niederrhein University of 

Applied Sciences, Germany 

[20] Křemenáková D., Kolčavová S. B., Mertová I.: Libtex 

software package, Technical University of Liberec,Textile 

Faculty, National research Center TEXTIL I, Czech Textile 

Seminar Greece May 2005 

[21] Szosland J.,: Modelling the structural barrier ability of 

woven fabrics, Textile research, Department of Textile 

Architecture, Technical University of Łódź, Poland 

[22] Hearle, J.W.S., Grosberg P., Backer, S.: Structural 

Mechanics of Fibers, Yarns, and Fabrics, (book) Wiley- 

Interscience 1969 


15


[23] Hearle, J.W.S.: Engineering design of textiles, Indian 

Journal of Fibre and Textile Research, Special Issue, 2005 

[24] http://www.texeng.co.uk 

[25] Long A.: TexGen – Open Source Software for Modelling 

of Technical Textiles, Transfer Summit/UK, Keble college, 

Oxford, 2011 

[26] H. Lin, X. Zeng, M. Sherburn, A. C. Long and M. J. Clifford. 

“Automated geometric modelling of textile structures”, 

Textile Research Journal, in press, May 2011 

[27] http://texgen.sourceforge.net/index.php/Main_Page 

[28] Masajtis, J.: Analiza strukturalna tkanin, Polska Akademia 

Nauk Oddzial w Łodzi, Komisja Włokiennictwa, Łodź 1999 

[29] Barburski M., Masajtis J.; Modelling of the Change in 

Structure of Woven Fabric under Mechanical Loading. 

FIBRES & TEXTILES in Eastern Europe, January/March 

2009, Vol. 17, No. 1 (72) pp. 39-44. 

[30] Szosland J., ‘Kształtowanie własności tkanin poprzez 

kształtowanie fazy ich struktury’ (in Polish, ‘Designing 

of woven fabric features by designing the phase of their 

structure’), Architektura Tekstyliów, No. 1-3, 1999 

[31] Keefe, M.: Solid Modeling Applied to Fibrous Assemblies 

PartII: Woven Structure, Journal of Textile Institut, Vol.85 

No.3, 1994 [350-358] 

[32] S. Backer. The relationship between the structural 

geometry of textile and its physical properties, I: Literature 

review. Text. Res. J. 1948, 18: 650-658. 

[33] Sirková, B.: Theses, Mathematical model for description of 

thread’s interlacing in fabric using Fourier series, Liberec 2002 

[34] Stepanović, J., Milutinović, Z., Petrović, V., Pavlović, M.: 

Influence of relative density on deformation characteristics 

of plain weave fabrics, Indian Journal of Fibre & Textile 

Research Vol. 34, March 2009 

[35] Milašius V.: Woven Fabric’s Cross-Sec¬tion: Problems, 

Theory, and Experimental Data, Fibres and Textiles in 

Eastern Europe No 4(23)/98, pp. 48-50. 

[36] Oloffson B.; „A general model of a fabric as a geometric 

mechanical structure” J. Textiles Isnt. Nr11,55, pp. 541- 

557; 1964. 

[37] Ozgen, B., Gong, H.: Yarn geometry in woven fabrice, 

Textile Research Journal, May 2011; vol. 81, 7: pp. 738- 

745, 

[38] Ozgen, B., Gong, H.: Modelling of yarn flattening in woven 

fabrice, Textile Research Journal, September 2011; vol. 

81, 15: pp. 1523-1531. 

[39] Kolčavová S. B.: Description of fabric thickness and 

roughness on the basis of fabric structure parameters, 

18th International conference Strutex 2011, Liberec, Czech 

Republic 2011 

[40] Kurashiki, T., H. Nakai, S. Hirosawa , M. Imura, M. Zako, 

S.V. Lomov, and I. Verpoest, Mechanical behaviors for 

textile composites by FEM based on damage mechanics, 

Key Engineering Materials, 2007, 334-335 (Advances in 

Composite Materials and Structures): 257-260. 


16


EVALUATION OF YARN LATERAL DEFORMATION 

Krupincová G. 1 , Drašarová J. 2 , Mertová I. 1 

1 

Technical University of Liberec, Department of Textile Technology 2 Technical University of Liberec, Department of Textile Design 

Studentská 2, 461 17 Liberec 1, Czech Republic, Tel.: +420 48 535 3424, Fax: +420 48 535 3542 

E-mail: gabriela.krupincova@tul.cz, jana.drasarova@tul.cz, iva.mertova@tul.cz 

Abstract: 

Keywords: 

The article focuses on a new approach for characterization and evaluation of lateral yarn deformation. A small review 

about theoretical description and measurement possibilities will be introduced. The evaluation of yarn compression 

will be done by three innovative methods (lateral deformation of yarn between two parallel plates, simulation of 

binding point of fabric, cross-sectional analysis of real fabric). The analysis of yarn deformation will be carried out 

for a set of samples in combination of fiber material, yarn count and given fabric structure. 

Yarn diameter, lateral deformation of yarn, innovative measurement principals. 

1. Introduction to fabric structure 

The woven fabrics structure is complicated due to its complex 

hierarchy. There are no models, which are able to describe the 

structure from the fibers through the yarns, and to the fabric. 

Usually the yarn is used as an elementary building unit of 

the structural model of the fabric. The fineness T, twist Z and 

diameter d are the basic geometrical characteristics describing 

a yarn. The diameter is considered only as a theoretical idea. 

For evaluation of the yarn diameter, it is necessary to know the 

packing density μ [1,9]. 

The simplified assumption, that the yarn is compact, solid 

and circular cross-section, is implemented for a description 

of binding point geometry. The troubles with establishing the 

yarn diameter originate in the incompactness of yarn structure. 

Some air gaps are found between the fibers; the yarn crosssection 

(especially in binding point) also is not a circle. In the 

binding points, the deformation of the cross-section and the 

compression of fibers are considered. 

In the stretched state of fabrics, yarns compress each other at 

their cross‐over points. The lateral compression force at the 

cross‐over points is generated by the yarn tension. Internal 

tension in fabric structure is given by the balance of forces, which 

depends on different types and levels of deformation during 

fabric production stages and its use. The typical deformation 

of the yarn cross-section is generally caused by the combined 

effect of compression, extension, bending and torsion. 

Main aim of this work is to report about possibilities, which 

can be used for measuring the lateral deformation of yarn. 

The evaluation of yarn compression will be done by three 

innovative methods (lateral deformation of yarn between two 

parallel plates, simulation of binding point of fabric, crosssectional 

analysis of real fabric). The experimental results for 

the selected material, yarn count and fabric structure will be 

presented. 

2. Yarn cross-section deformation 

The nearly circular yarn cross-section is due to change of 

compression to a more flat profile. A lot of geometrical models 

do not include this phenomenon and the so-called free yarn 

geometry is assumed. Yarn flattening is important from the point 

of view of selected fabric parameters evaluation, modeling and 

design. It is, for example, fabric porosity, air permeability and 

mechanical characteristics in terms of ultimate and user range 

loading. 

2.1 Models 

The width a and height b are defined for the description of 

yarn cross-section deformation. The circular cross-section 

of the “free” yarn (Figure 1a) is changed into a shape, which 

can be substituted by Kemp´s cross-section (oval of two halfcircles 

with semi-diameter b and two abscissas of length a-b, 

Figure 1b), elliptical or lens shape (Figure 1c, d) [1]. 

Figure 1. Shape of yarn cross-section: 

a) “free” yarn, b) Kemp, c) ellipse, d) lens. 

It is possible to define relative enlargement and relative 

compression: 

( − ) 

ε 

1 

= b d d 

ε 

2 

= ( a − d ) d 

The area S and the perimeter L of these shapes can be easily 

calculated. 

(1) 

(2) 


17


2.2 Geometrical hypothesis 

For the relationship between the shapes, two alternative 

hypotheses about constant area and constant perimeter are 

proposed. This hypothesis can be described as a function of 

relative enlargement e 1 

and relative compression e 2 

. 

Constant area (cross-sections before and after deformation 

have the same area) subsequently holds: 

From the second hypothesis of “constant” perimeter, the 

area of the deformed cross-section must be decreasing; the 

packing density increases, because the number of interπd 

S = = S 

(3) 

deformed yarn 

4 

The relation between relative enlargement and relative 

compression is derived: 

2 

ε 

2 

= ( ε1 ( 1 − π 4 ) + ε1 ( 1 − π 2 )) ( ε1 

+ 1 ) 

Kemp 

, (4a) 

ε = − ε ε + 

2 1 1 

1 . (4b) 

ellipse 

( ) 

It is impossible to express the relative enlargement explicitly for 

lens; therefore, the equation was solved numerically: 

ε 

2 ( ε1 

) 

lens 

1,0 6 

= 1,1 1 + 1 − 1 . (4c) 

Constant perimeter (cross-sections before and after 

deformation have the same perimeter) subsequently holds: 

L = πd = L . (5) 

deformed yarn 

The relation between relative enlargement and relative 

compression is derived: 

ε 

2 

= ε1 ( − π ) 

Kemp 1 2 , (6a) 

2 

ε 

2 

= 2 − ( ε1 

+ 1) 

− 1 

ellipse 

, (6b) 

2 2 

ε 

2 

= ( π 2 ) − 4 3 ( ε1 

+ 1) 

− 1 

lens 

. (6c) 

Lomov [7] proposes the relation between relative enlargement 

and relative compression empirically: 

( ε ) 

ε = 1 + 1 − 1 

2 1 

n 

for (n=1, 2...). (7) 

The estimation of these hypotheses is based on the assumption 

of circular cross-section of “free” yarn. The circle has minimal 

perimeter for the same area and maximal area for the same 

perimeter, compared to other plane figures. The cross-section 

changes are caused not only by changes in shape but also due 

to relaxation of radial forces. These forces originate from the 

helix structure of the fibers in the yarn. 

From the first hypothesis of “constant area”, we conclude that 

the perimeter of the deformed cross-section must be increasing; 

the volume of inter-fibers pores does not change, it means that 

the packing density decreases. It would be a particular effect, 

which eliminates the action of radial forces relaxation. 

fiber pores decreases and the contacts of fibers increase. It 

means the destruction of the original (primary) yarn structure 

turn up. 

3. Experimental methods 

There exist various methodologies for evaluation changes in 

yarn cross-sections. A change of yarn diameter under tension in 

the yarn-axis direction was studied in many pieces of research. 

They are mostly based on optical system or mechanical 

detection. Only few of them are described and their results are 

shown in this article. 

There is a group of methodologies, which is based on fabric 

analysis. One of them is the judgment of fabric thickness before 

and after biaxial tension, which is described in [4]. The crosssectional 

analysis of fabric in freeze state is another approach 

to gain information about yarn’s deformation. Fabric can be 

fixed by soft or hard methods. The experiment is based on the 

analysis of frozen fabric structure in terms of cross-sectional 

analysis or surface analysis in the third main fabric direction 

[5]. The improved possibility of novel stress-freezing technique 

for studying the compression behavior of fabrics is described 

in [10]. The biggest advantage of this modified method is the 

possibility of deformed fabric fixing. 

Fixing of fabric in stressed state is limited and therefore the 

simplified approaches were found to see the influence of 

selected factors and forces. Methodologies, which take not 

only compression but also bending into consideration, are 

wire method, V wire method, three-rod unit and simulation of 

binding point in hollow block [2] and [4]. 

The deformation of yarn between two parallel plates is the 

highest simplification of a real state in balance of forces at a 

cross‐over point. In this case, only complexional forces cause 

the deformation of yarn. There are several methods by which 

the thickness of yarn can be measured. Using of a rotation 

drum and feeler, in which the yarn thickness is measured 

by passing the yarn around the drum’s circumference with 

the feeler pressed very gently against the yarn, has been 

mentioned in [8]. The KES F3 system allows the measurement 

of yarn compression in terms of yarn thickness as a function 

of compression load [3,4]. In case, the information about 

yarn thickness (minor yarn diameter b) is not enough and 

the knowledge of yarn widening (major yarn diameter a) is 

demanded then it is possible to use special equipments, 

which offer measuring the change in yarn diameter in both 

main directions under higher pressure [2]. 

3.1 Analysis of weave cross-sections 

The method used for the detection of the internal weave 

structure is based on the analysis of the weave cross-sections 

[4,5,7]. The measuring parameters are shown in Figure 2. The 

fabric cross-sections were prepared by the method of “soft” 

cross-sections, where the blend of bee wax and paraffin, as 

fixing medium, was used [1,9]. These cross-sections usually 

have a thickness of 30 mm. 


18


3.2 Assessment of yarn flattening caused by compression 

and bending 

The wire and V wire method is the simplification of real state in 

fabric binding point. One of the cross‐over yarns is substituted 

by absolute stiff wire because of the factors influencing 

elimination. A steal wire is usually fixed on a horizontal base 

and yarn is hanged on the wire with tension. The angle 

between the yarn and the horizontal level is 30°, which is 

approximately equal to the averaged yarn-intersecting angle 

in various weaves. A V-shaped groove at the cross‐over area 

is more close to the real intersecting state of the yarn in fabric 

binding point. The yarn thickness is measured by a needle 

sensor contacting at the top of the yarn surface with a small 

compression force [4,9]. 

Figure 2. Measured values. 

The alternative possibility based on the V wire method is using 

the three-rod unit mounted, for example, on the Instron Tensile 

Tester. The three-rod unit consists of a rod fixed to a horizontal 

base and two rods of the same diameter placed parallel to the 

bottom one. It is spaced at the center of the three rods that 

form an equilateral triangle. The unit enables the measurement 

to be made of the changes in both minor and major diameters 

of a yarn bent over the three-rod unit subjected to increasing 

extension [4]. 

Simulation methodology of real fabric binding points goes 

out from idea that the yarns are crossed in a hollow block 

and various forces realize their deformation. Arrangement of 

experiment is shown in Figure 3. The hollow block is placed 

under a macroscope and the measurement of change in yarn’s 

diameters is realized in the system of image analysis [2]. Yarn 

is guided in between two opposite corners that are placed in 

position of block diagonals. One end of the yarn is fixed by 

clumps and the other is guided over a small ideal pulley with a 

small pretension. The simulated binding point is placed in the 

hollow block center. The loading of the yarn sample is realized 

by various weights. 

Figure 3. Hollow block: 

1 camera 

2 macro-scope objective 

3 tripod 

4 pulley 

5 weight 

6 simulated binding point 

3.3 Yarn compression between two parallel plates 

The other method for the simulation of stress in the binding point 

is based on the yarn compression between two parallel plates 

[5,6]. The first prototype is shown in Figure 4. This device is 

placed under a macroscope holder. The yarn is guided though 

the measuring zone between two glass parallel plates and is 

pretended proportionally to the yarn count. The deformation of 

yarn is the result of loading realized by an upper frame. The 

upper frame can be connected with various pieces of defined 

weight and a seven level of loading is available. Sequence 

of yarn longitudinal views before and after deformation at 

same place is scanned and the absolute values on scales of 

two contact thickness meters are read. Sequences of yarn 

macro‐images before and after deformation in terms of yarn 

diameter d and yarn characteristic proportion a are evaluated. 

Characteristic proportion b, in other words yarn thickness, 

after deformation is equal to the value, which describes the 

difference between the absolute value on the scale of thickness 

meter without and with deformed yarn between parallel plates 

under pressure. 

Figure 4. Measuring device for yarn compression: 

1 main frame 

2 yarn 

3 guide 

4 brake 

5 straightening load 

6 contact thickness meter 

7 lighting system 

8 lower plate with calibrated ruler 

9 frame with upper glass plate 

10 measuring area 


19


4. Experimental material and results 

The main idea of this experiment was to use the selected 

methodologies used for the evaluation of yarn deformation 

and apply them to a set of fabrics and yarns, which were 

used for fabric weaving. The results should be discussed and 

compared with expectations and the influencing factors should 

be identified. 

a) 

0,5 

0,4 

0,3 

0,2 

0,1 

2 [-] 

A set of experimental gray relaxed fabric in plain weave was 

used for the experiment. Fabrics were produced from 100% CO, 

100% PP and 50CO/50 PP 29,5tex staple single ring yarns with 

a given set of warp (25 thread cm -1 ) and three level of set of weft 

(8,8 thread cm -1 , optimum 13 thread cm -1 and 17 thread cm -1 ). 

One gray relaxed fabric produced with comparable geometrical 

structure in plain weave from 100%PET 16,5tex staple 

single yarn was added for the experiment. Analysis of yarn 

deformation was realized according to the cross-sectional 

technique described in section 3.1 for both main directions of 

the fabric (warp and weft). Experimental results are shown in 

Figure 5a. 

Estimation of the level of yarn deformation based on the 

evaluation of simulated binding point described in section 3.2 

was realized for a set of yarn used for fabric production. Yarns 

were spun by classical ring spinning technology with 29,5tex 

yarn count from 100% CO, 100% PP and 50CO/ 50 PP staple 

fiber material. Three levels of loading force were applied to 

simulate yarn deformation (1,8 g – 0,07N, 6,8 g – 0,26N and 

11,8 g – 0,44N). The obtained data are presented in Figure 5b. 

Yarn deformation was also simulated by deformation between 

two plates mentioned in section 3.3. Hundred percent CO 

29,5tex staple single ring spun yarn, which was used for fabric 

production, was analyzed and an experimental set of 100% CO 

single staple yarn was added to the lab measurements. It was 

a set of typical ring spun yarn that was produced with 16,5tex, 

20tex and 38tex yarn count. Seven levels of deformed forces 

were used for the simulation of yarn deformation (10N, 15N, 

20N, 25N, 30N and 40N). Summarization of data is given in 

Figure 5c. 

5. Discussion 

b) 

c) 

-0,7 -0,5 -0,3 -0,1 

 

constant area Kemp 

constant area lentil 

constant perimeter elipse 

Lomov 

50%CO/50%PP 29,5tex 

100% PET 16,5tex 

0 

constant area elipse 

constant perimeter Kemp 

constant perimeter lentil 

100%PP 29,5tex 

100% CO 29,5tex 

-0,7 -0,5 -0,3 -0,1 



constant perimeter elipse 

Lomov 

strength -0,26 [N] 

0,5 

0,4 

0,3 

0,2 

0,1 

0 

2 [-] 



constant perimeter lentil 



0,5 

0,4 

0,3 

0,2 

0,1 

2 [-] 

Figures 5 a, b, c show relationships between relative 

enlargement and relative compression. The hypothesis for 

constant perimeter and constant area are compared with 

experimental data. 

The generally known results could be expected. Experimentally 

analyzed yarn deformations in a fabric binding point are located 

in the low levels of relative yarn enlargement and compression. 

It is not possible to decide, if the deformations follow the 

hypothesis of constant perimeter or constant area. Calculated 

curves describe that the limited cases are very close to each 

other in this range of deformations. Experimental data should 

be placed in a delimited area of both hypotheses. Most of 

them are under hypothesis of constant perimeter. It is probably 

caused by the precision of original yarn diameter estimation. 

0 

-0,7 1 [-] -0,5 -0,3 -0,1 





constant perimeter elipse constant perimeter lentil 

Lomov 

100%CO 16,5tex 

100%CO 20tex 

100%CO 38tex 

100%CO 29,5tex 

Figure 5. Experimental results: 

a) method 3.1 

b) method 3.2 

c) method 3.3 

The influence of evaluated factors was in accordance with 

previous experience. Higher deformations in terms of relative 

enlargement and compression were found for the main fabric 

warp dimension. From the point of view of weft analysis, it 

can be said that the deformation increases when the weft set 


20


increases. The behavior of fabric produced from blended yarn 

is more close to the behavior of 100% PP fabric. Differences 

among experimental data were very small. The verification of 

significance power of these factors is limited, because the data 

are very sensitive to original yarn determination. 

Using of hollow block for the analysis of yarn deformation 

is limited from the point of view of estimation of relative 

compression only. In other words, it is not possible to measure the 

second parameter describing the level of relative enlargement. 

Therefore, the second coordinate of the experimental data 

is equal to zero in Figure 5b. An increased loading force 

causes the increase of yarn relative compression in contact 

point. The statistical significance of the fiber material used 

for yarn production is very low but for verification repeating of 

measurements should be realized. The blended yarn behavior 

is much closer to the behavior of 100%PP yarn. The reason 

can be what’s hidden in yarn production because of used 

mass fibers mixing. There is a higher number of polypropylene 

fibers in yarn volume than cotton fibers. It confirms the outputs 

from the methodology given in section 3.1. The interesting 

conclusion arises from the comparison of relative deformation 

obtained form fabric analysis and this method. It seems that the 

level of normal force in the gray relaxed fabric is very close to 

the first level of loading force realized here by the weight 1.8 g. 

Simulation of yarn between two parallel plates is the most 

simplified process of real deformation in real binding point. 

Based on previous experiments and solving force balance, 

it can be expected that for the description of deformed yarn 

the cross-section Kemp model and hypothesis of constant 

perimeter will be desirable. It means, in other words, that the 

yarn is due to deformation compressed and the pacing density 

of yarn increases. It can be also expected, that the deformation 

will increase with the increase in yarn count or loading force. 

The obtained experimental results are in a good agreement 

with our expectations. Differences in experimental data results 

are higher for the higher applied forces. Moreover, the statistical 

significance of data differences was not verified. 

6. Conclusion 

In this investigation, the yarn lateral deformation was studied. The 

experimental data obtained from three selected methodologies 

were compared. The influence of selected factors was roughly 

evaluated. There was for analysis used methodology, fiber 

material, yarn count, applied level of deformation force and 

fabric structure in terms of set of warp and weft. 

It can be concluded (thanks to the realized experiment) that the 

methodologies give us comparable results, which can be the 

background for precise modeling of structure and mechanical 

parameters of fabric. Yarn flattening is important, for example, 

for estimation of fabric porosity, air permeability and mechanical 

characteristics in terms of ultimate and user range loading. 


This work was supported by the research project GACR 

106/09/1916. 

References 

[1] Drašarová, J.: Analysis of fabric cross-sections. PhD. 

Thesis, FT TUL, Liberec 2004, in Czech. 

[2] Drašarová, J., Jamborová, J., Dzurindak, P.: Lateral 

deformation of yarn. FRVŠ 2000 2001, in Czech. 

[3] Göktepe, E., Lawrence, C. A.: Deformation of Yarn Crosssection 

in Relation to Yarn Structure. Fibers & Textile in 

Eastern Europe. April/ June Vol. 2 No. 33, 2001, ISSN: 

1230-3666. 

[4] Kawabata, S., Niwa M., Matsudaria, M.: Measurement of 

Yarn Thickness Change Caused by Tension and Lateral 

Pressure by Wire Method. Journal of Textile Machinery 

Society of Japan, Vol. 31, No. 1, 1985, ISSN: 1883-8723. 

[5] Křemenáková, D., et all: Internal Standards. Textile 

Research Center, Technical University of Liberec 2003. 

[6] Křemenáková, D., Krupincová, G.: Lateral compression of 

free yarn. International conference ARCHTEX 2003, ISBN 

83 917808 0 5. 

[7] Lomov, S.V., Peeters, T.: Integrated textile preprocessor 

WiseTex. Version 2.3. Specific software User´s guide. 

[8] Mahmoudi, M. R., Oxenham, W.: A new electro-mechanical 

method for measuring yarn thickness. Autex Research 

journal, Vol. 2., No. 1, March 2002, ISSN: 1470-9589. 

[9] Neckář, B.: Yarn. Creation, structure, properties. SNTL 

Praha 1990, ISBN: 80-03-00213-3, in Czech. 

[10] Potluri, P., Wilding, M. A. and Memon A.: A Novel stress- 

Freezing Technique for Studying and Compressional 

Behavior of Woven Fabrics. Textile Research Journal, Vol 

72, No. 12, Dezember 2001, ISSN: 00405175. 


21


TESTING OF YARN ABRASION 

Krupincová, G. 1 , Hatipoglu, J. 2 

1 

Technical University of Liberec, Department of Textile Technology, Liberec, Czech Republic 

Tel.: +420 48 535 342474, Fax: +420 48 535 3542, E-mail: gabriela.krupincova@tul.cz, 

2 

Ege University, Textile Engineering Department, Izmir, Turkey, E-mail: yakuphatipoglu@gmail.com 

Abstract: 

Keywords: 

There exist a lot of methodologies, which can be used for yarn quality testing. Abrasion resistance and its 

measurement for raw and sized yarn can help in the judgment of yarn weaving-ability. This article concentrates on 

the possibility of yarn abrasion expression and testing. Relation among fiber material characteristics, selected yarn 

structural, and mechanical parameters is discussed and a few experimental results are shown. 

Structural and mechanical yarn characteristics, yarn abrasion resistance. 

1. Introduction 

Higher production and higher demands of customers on fabric 

utility value means higher requirements on yarn production 

quality. Selected fiber material, chosen spinning technology, 

and all previous operations are necessary to be applied 

before weaving, since their correct application influences 

fabric and end-product quality. Performance of warp yarns 

on a loom during weaving is affected by a number of factors 

as it is subjected to complex deformation including abrasion, 

cyclic bending together with tension and impact loading. 

Controlling yarn’s structural characteristics and examining of 

level of mechanical parameters together with evaluation of 

yarn’s weaving‐ability is essential. Abrasion resistance and its 

measurement for raw and sized yarn can help in the judgment 

of yarn’s weaving-ability. 

2. Yarn abrasion 

2.1 Testing possibilities 

Methodologies used for yarn abrasion resistance testing 

can be divided into two groups. First group uses a defined 

abrasion material. Results from this test are comparable [7]. 

An example of this instrument is Zweigle G 552 tester [2], 

Wira tester [5] or CTT yarn abrasion tester [9]. Simulation of 

mechanical behavior on laboratory loom or on its function parts 

can give results more close to real weaving. On the other hand 

simulation is limited too. Loom settings, its speed, all interaction 

among yarns and guiding places are different for various kinds 

of looms. Representative instrument is Reutlinger Webtester 

[1]. Simulation of “yarn on yarn” abrasion can help us in 

understanding the mechanism of yarn damage during yarn on 

yarn contact. Staff tester of Zweigle [5] simulates the running 

characteristics of spun yarns and smooth plied yarns. Special 

method “yarn on yarn” is used for testing of rope or specific 

yarns made from special synthetic fibers. Measurement can be 

realized in normal or wet conditions [8]. 

Zweigle G 522 method was used during the experiment and 

therefore few things need to be addressed. Usually up to twenty 

threads are placed in the abrasion tester; thereafter pre-tension 

is applied (usually 20g or 30g per thread). Everything proceeds 

automatically: An abrasion roller covered with emery paper 

traverses in constant rhythmic motion and constant pressure 

at right angles to the direction at which the test threads are 

tensioned. The abrasion roller continuously rotates about its 

own axis so that an abrasive action is not impaired by abraded 

yarn residue in the emery paper. The computer controls the test 

procedure and logs the yarn breaks. Special optical sensors are 

activated when a weight drops down, and number of strokes for 

all samples is recorded in a database [2]. 

2.2 Approaches to yarn abrasion description 

Abrasion resistance is usually expressed as a number of strokes 

to yarn destruction. A criterion based on weight reduction is a bit 

problematic, because limited yarn length is possible to weight. 

A weight reduction due to abrasion can be easy described 

thanks to yarn diameter changes. The diameters of original 

yarn samples and yarns after extension of 50% of the number 

of strokes for yarn destruction can be observed according to 

internal methodology IN 32‐102‐01/01 [10]. 

The method is based on scanning and processing images 

of yarn longitudinal views. Color images are transformed 

through gray-scale to binary images by using Otsu’s method 

[4]. Fibers that belong to the hairiness sphere are eliminated 

by morphological operation (erosion, dilatation, opening, and 

closing of image). All image rows are processed step by step. 

Number of pixels belonging to yarn is counted. Their original 

length is recalculated by used calibration. Evaluation of each 

image row – potential yarn diameter (original yarn diameter 

D, yarn diameter after abrasion Da), outlier values exclusion, 

statistical yarn diameter finding – follows. Yarn diameter as a 

mean value itself cannot describe diameter change completely. 

Minimum Da min 

, maximum Da max 

, and mean value Da of yarn 

rows’ length can qualify yarn dimensions after abrasion. 


22


Minimum diameter means the shortest row length of imaginary 

yarn cross-section. Maximum diameter means diameter 

of cylinder, which can cover the yarn. In other words, it is a 

difference between the smallest coordinate and the highest 

coordinate, where black pixel that belongs to yarn is placed in 

hole image and not only in actual image row. The percentage 

change of yarn diameter before and after abrasion can express 

abrasion resistance. 

3. EXPERIMENT 

The influence of selected factors on abrasion resistance is 

investigated. The level of influencing factors is evaluated thanks 

to correlation and ANOVA analysis. Pair and partial correlation 

coefficient was used for expression of strength and direction 

of a linear relationship between two random variables. It was 

calculated according to eq. (1a,b). Pair correlation measures 

the strength of the relationship between two random variables. 

It does not take other influencing factors into consideration. 

Partial correlation is more sufficient, because it measures the 

degree of association between two random variables, with the 

effect of a set of controlling random variables removed. ANOVA 

analysis enables dividing of variance into different components 

due to explanatory variables. Two-dimensional ANOVA analysis 

with fixed-effects model was used for data processing. 

R 

R 

12 

= 

1 i (2,3,... k ) 

3.1 Experimental material 

cov( X1, X 

2) 

var X var X 

= 

1 2 

i 

( −1) det( R1, 

i 

) 

. 

det( R ) det(R ) 

(1a,b) 

The idea of this experiment is to prepare yarns from various 

fiber materials with similar yarn geometrical parameters under 

comparable condition. A set of one component and blended 

single and two-ply ring spun yarns was used for the experiment. 

Step 1: One-component single ring spun yarns were spun 

in five levels of yarn count and three levels of Phrix twist 

coefficient from 100% PET, PAN, VS fibers. The information 

about the fiber material is shown in Table 1. One-component 

single ring spun yarns are described in Table 2. 

Step 2: Set of blended single and two ply yarns were produced 

with typical yarn count level 29,5tex, respectively 2x29,5tex and 

optimal Phrix twist coefficient in five level of blending portion of 

polypropylene PP and cotton CO fibers. The information about 

the fiber material is shown in Table 1. The description of this set 

of yarns is given in Table 3. 

, 

11 i,i 

Table 1. Selected fiber parameters. 

PET PAN VS CO (Egypt Giza 70) PP 

t n 

/ t v 

[dtex] 1,3/ 1,4 0,9/ 1,17 1,3/ 1,34 1,65 1,88 

(1,36; 1,45) (1,13; 1,21) (1,30; 1,37) (1,53; 1,77) (1,80; 1,95) 

r v 

[kgm -3 ] 1360 1170 1520 1520 910 

l [mm] 38 38 38 31,13 50 

RF [mNmm 2 tex -1 ] 0,3 0,33-0,48 0,19 0,19 0,51 

f v 

[cNtex -1 ] 53,32 33,97 17,56 40,5 40,42 

(51,67; 54,98) (32,85; 35,08) (16,77; 18,34) (36,6; 44,4) (39,35; 41,08) 

e v 

[%] 17,51 31,86 30,05 5,95 63,35 

(16,27; 18,74) (30,63; 33,08) (29,11; 30,99) (5,41; 6,48) (58,13;68,56) 

Table 2. Description of one-component single ring spun yarns. 

material PET, PAN, VS PET, PAN, VS PET, PAN, VS PET, PAN, VS PET, PAN, VS 

T [tex] 16,5 20 29,5 35,5 42 

a [ktex 2/3 m -1 ] 50 56 62 50 56 62 50 56 62 50 56 62 50 56 62 

Table 3. Description of blended single and two-ply yarns. 

material 100%PP 35%CO/65%PP 50%CO/%PP 65%CO/35%PP 100%CO 

single two-ply single two-ply single two-ply single two-ply single two-ply 

T [tex] 29,5 2 x 29,5 29,5 2 x 29,5 29,5 2 x 29,5 29,5 2 x 29,5 29,5 2 x 29,5 

Z [m -1 ] 601 350 643 359 621 357 632 380 625 371 

(593; 610) (340; 360) (637; 648 ) (350; 368) (615; 626) (350; 365) (624; 639) (370; 389) (616; 633) (360; 383) 


23


3.2 Testing conditions 

Testing of selected fiber and yarn parameters was 

realized according to the Czech European International 

Standard. Conditioning of samples was made in respect to 

ČSN‐EN‐ISO‐2061. Fiber fineness and mechanical parameters 

were measured according to ČSN‐EN‐ISO‐1973 by Vibroskop 

& Vibrodyn instruments (gauche length 10mm, pretension 

selected according fiber material, 50 measurements). Controlling 

of yarn count and yarn twist was realized in agreement with 

ČSN‐EN‐ISO‐2060, ČSN‐EN‐ISO‐2061. Yarn mechanical 

parameters were tested according to ČSN‐EN‐ISO‐2062 

on Instron tester (gauche length 50mm, pretension selected 

according yarn count, time of correct test up to 20s ± 3s, 50 

correct measurements). Zweigle G 552 instrument was used 

for measuring number of stokes to destruction (pretension 20g, 

emery paper P 800 with abrasive grain alpha Al 2 

O 3 

and loom 

reed, 60 measurement). Yarn diameter change due to abrasion 

was studied only for the set of blended yarns. Diameters of 

original yarn sample and yarn after extension of 50% of 

number of strokes to yarn destruction were observed according 

to internal methodology IN 32‐102‐01/01 [10] (image resolution 

548pxl x 704pxl, calibration 2,23mmpxl -1 ). 

3.3 Discussion of experimental results 

The relation among fiber parameters, yarn count, yarn twist, 

and mechanical parameters is well known. Theoretical 

presumptions were confirmed by many experiments. It is 

generally accepted that yarn strength is related with fiber 

characteristics (mechanical parameters, stiffness, friction, 

and flexural rigidity), yarn structural characteristic (count, twist 

coefficient, and packing density), and technology of production. 

Production technology selection influences the level of fiber 

arrangement. Higher degree of fiber arrangement in yarn 

means better yarn mechanical properties. Correct level of 

yarn count and twist is important from the point of view of yarn 

packing density. Higher compactness of fiber in yarn causes 

stronger utilization of fiber strength and inter-fiber slippage. 

Using of stronger, stiffer fiber material with higher yarn count 

and higher twist leads to stronger yarn with lower elongation. It 

can be expected that similar assumptions will be valid for yarn 

abrasion. 

Simulation of yarn abrasion straining on a loom was realized 

using Zweigle G 552 tester. Yarn structure was opened 

during yarn abrasion and twists were pushed to the ends of 

the testing zone. Bundle of fibers and its damaged segments 

created entanglements on the sample body (see Figure 1). 

Inter-fiber slippage supported by pretension and structure 

opening tended to break the sample. Knowledge of dimension 

and the occurrence of thin and thick places on yarn’s body is 

important. Thick places create trouble while passing through 

all the guiding places on a loom. Thin places are a risky part 

of the chain that increases the warp breakages. It is necessary 

to mention that the images of yarn are transformed from 3D 

to 2D and therefore dimension of thin or thick places is not 

fully described. An example of yarn longitudinal view of 29,5tex 

yarn before and after abrasion due to emery paper is shown in 

Figure 2 (Figure 2a shows the original yarn sample, Figure 2b 

the thin place, and Figure 2c thick place of the sample). 

Multivariate data analysis (Correlation and ANOVA analysis) 

was used for data processing. Question of significance of 

the power of various factors on yarn abrasion resistance was 

solved in two steps. A set of single yarn samples allows us 

to judge the influence of fiber material, and structural and 

mechanical characteristics on abrasion resistance (step 1). 

Figure 1. 100% cotton single yarn 29,5tex after 50% of the number of strokes to yarn destruction (calibration 4,72µmpxl -1 , resolution of image 

548pxl x 704pxl). 

a b c 

Figure 2. 100% cotton single yarn 29,5tex before and after 50% of the number of strokes to yarn destruction (calibration 2,23µmpxl -1 , resolution 

of image 548pxl x 704pxl). 


24


The effect of blending portion together with plying technology 

on abrasion resistance can be studied, thanks to comparison of 

blended single and two-ply yarn results (step 2). 

Step 1: Influence of selected fiber (fineness t v 

, diameter d v 

, 

mass density r v 

, flexural rigidity RF, strength f v 

, and elongation 

e v 

) and yarn characteristics (nominal count T jm 

, Phrix twist 

coefficient a, experimental count T exp 

, twist number Z, and 

mechanical parameters F, e) on yarn abrasion resistance was 

investigated for a set of one component single yarns defined in 

Table 2. Yarn abrasion resistance was expressed as a number 

of strokes to yarn destruction a 1 

. Number of strokes is highly 

correlated with yarn count. Therefore, the ratio between the 

number of strokes and yarn count was added to data analysis 

a 2 

. Correlation map for paired correlation coefficients is shown 

in Figure 3a and for partial correlation coefficients in Figure 3b. 

Thanks to multivariate data analysis, it was found that fiber 

strength, yarn count, yarn strength, and elongation are 

significantly related to abrasion resistance (paired and partial 

correlation coefficients higher than 0,5). It was verified that 

number of strokes to yarn destruction a 1 

is positively correlated 

with yarn count (partial correlation coefficient 0,64). Influence of 

twist level is not as significant as we expected. This approach is 

limited because of a mutually connected factor (multicolinearity), 

factor’s limited range, and proper selection of technological yarn 

creation parameters (interdependence yarn count, yarn twist). 

Step 2: Set of blended single and two-ply yarns is very 

interesting, since polypropylene is normally not used for blending 

with cotton fibers. These two kinds of fibers are dissimilar in 

many characteristics, but an otherness in mass densities and 

mechanical parameters give blended yarns attractive parameters. 

When we use well-known relation among fiber mass density, 

fibers characteristics (e.g., fineness t v 

, strength f v 

, elongation 

e v 

), and yarn’s parameters (e.g. count T, twist Z, tenacity F, and 

elongation e), we can find interesting connection. Same fineness 

of cotton and polypropylene fibers means higher diameter of 

polypropylene fibers. Polypropylene fibers have higher tenacity 

and elongation than cotton fibers, because of their chemical 

nature. Same yarn count and same level of twists means higher 

number of polypropylene fibers in yarn cross-section. This 

phenomenon leads to higher diameter and higher tenacity of 

polypropylene yarn. Dependence of yarn tenacity on blending ratio 

can be predicted by Hamburger’s theory [10]. Modeling of other 

mechanical parameters (e.g., elongation, abrasion resistance) is 

problematic and exists only in regression equation. 

Figure 4a, b, c shows relationships between yarn tenacity, 

yarn elongation, abrasion resistance, and blending portion of 

cotton fibers for single and two-ply yarns. The obtained results 

a 

F [Ntex - 1 ] 

0,3 

0,25 

0,2 

0,15 

0,1 

0,05 

0 

0 0,2 0,4 0,6 0,8 1 

Ble n d in g p o r t io n o f c o t t o n f ib r e s [ -] 

s ingle y arns tw o-ply y arns 

a 

b 

3 0 

2 5 

2 0 

e [%] 

1 5 

1 0 

5 

0 

0 0 ,2 0 ,4 0 ,6 0 ,8 1 


s ingle y arns tw o-ply y arns 

b 

c 

a* [strokes tex 

- 1 ] 

8 0 

7 0 

6 0 

5 0 

4 0 

3 0 

2 0 

1 0 

0 

0 0 , 2 0 , 4 0 , 6 0 , 8 1 


s ingle y arns 

tw o-ply y arns 

Figure 3. Correlation map for paired correlation coefficients and for 

partial correlation coefficients (step 1). 


25 

Figure 4. Comparison of mechanical parameters and blending portion 

of cotton fibers.


of mechanical parameters correspond with our expectation. It 

is evident, that two-ply yarns have higher tenacity, comparable 

elongation, and higher abrasion resistance than single yarns. 

Tenacity of yarns decreases in respect to increasing blending 

portion of stronger component (PP); and that after reaching 

critical blending portion, point increases. Higher portion of 

stiffer stronger polypropylene fiber in yarn leads to lower 

elongation. General conclusion for dependence of abrasion 

resistance on blending portion was not found. Only semi-linear 

function can be suitable for trend description. Blending portion 

35PP/65CO seems to be optimal from the point of view of 

abrasion resistance and other mechanical parameters. 

Analysis of yarn diameters before and after abrasion shows 

interesting results. Percentage of yarn diameter change 

describes yarn’s abrasion resistance in terms of weight loss. 

Diameter comparisons (D‐Da)/D 10 2 , (D‐Da min 

)/D 10 2 , and 

D‐Da max 

)/D 10 2 show, that the change due to abrasion is 

more significant in case of single yarn. Typical trend of yarn 

diameter change in terms of blending portion can be found 

only for comparison of mean diameters D and Da. Diameter D 

decreases if blending portion of PP fibers decreases. Diameter 

change increases if blending portion of PP fibers decreases. 

Diameter reduction due to abrasion is more obvious in case 

of single 100%CO yarn and is about 20% (for 100% PP only 

10%). Two-ply yarns show lower reduction of yarn diameter 

and is in a range from 1% to 3%. Evaluation of diameter D and 

Da min 

indicates that 40% reduction of diameter in case of single 

and two-ply yarn is possible. Yarn diameter enlargement is not 

so significant and is maximally 8% of original diameter. Twodimensional 

ANOVA analysis (factors: blending portion, plying) 

a 

b 

d iam et er c h an g e [ -] 

d iam et er c h an g e [ -] 

0,5 

0,4 

0,3 

0,2 

0,1 

0,0 

-0,1 0 0,5 1 

-0,2 


tw o-ply y arns 1-Da/D 

tw o-ply y arns 1-Damin/D 

tw o-ply y arns 1-Da max /D 

0,5 

0,4 

0,3 

0,2 

0,1 

0,0 

-0,1 0 0,5 1 

-0,2 


s ingle y arns 1-Da/D 

s ingle y arns 1-Da min/D 

s ingle y arns 1-Da max /D 

Figure 5. Comparison of yarn diameters before and after abrasion. 

confirms that plying technology is a significant factor, which 

influences mechanical parameters F, e, abrasion resistance a1, 

a2, and diameter change. Two-ply yarns made of 100%PP and 

65PP/35CO are the best one in terms of strength, elongation, 

and abrasion resistance characteristics. 

4. Conclusion 

The main aim was to report about approaches to yarn resistance 

evaluation. Two sets of yarn were selected for experiment. Only 

the row of single and two-ply yarns was tested. The selected 

structure and mechanical parameters of fiber and yarn were 

analyzed. Zweigle G 552 was used for measuring yarn 

abrasion resistance. Number of strokes to yarn destruction 

and diameter change due to abrasion were observed. The 

generally known relationships were confirmed. Experimental 

results are in a good agreement with expectation. Increase 

of yarn count causes increase of yarn strength and abrasion 

resistance. Increase in number of twists is followed by increase 

of yarn strength and abrasion resistance. It is the reason of 

yarn elongation decrease. The influence of yarn count is in 

connection with number of fibers in yarn cross-section. Effect of 

twist is related to the level of fiber compactness. Two-ply yarn 

is more resistant to abrasion and is the reason of their using as 

warp yarn. Using of higher blending portion of a stiffer stronger 

component can improve yarn characteristics in terms of 

mechanical behavior and yarn abrasion. New approach to yarn 

abrasion evaluation was introduced. Interesting information 

was obtained thanks to comparison of yarn diameters before 

and after abrasion. Dimension of potentially thin and thick 

places is helpful for the assessment of weaving‐ability. None 

of the parameters on its own provided a reliable method for 

establishing a definitive correlation between measurements 

made in laboratory and actual performance of yarns during 

weaving, because the process of mechanical yarn deformation 

on a loom is very complex and cannot be simulated absolutely 

during laboratory analysis. 


This work was supported by the Textile Center II of Czech 

Ministry of Education 1M0553 and was published in AUTEX 

2009 WORLD TEXTILE CONFERENCE May, 26-29, 2009 

Çesme, Izmir, TURKEY. 

References 

[1] CTT yarn abrasion tester, Lawson Hemphil. http://www. 

lawsonhemphill.com/. Last Update 20.10. 2007. 

[2] Goswami, B., C., Anandjiwala, R., D. and Hall, D. M.: 

Textile Sizing. Marcel Dekker, Inc., ISBN 0 8247-5053-5, 

USA - New York - Basel, (2004). 

[3] Hearle, J., W., S., Grosberg, P., Backer, S.: Structural 

Mechanics of Fibers, Yarns and Fabrics. Wiley, New York 

196. 

[4] Křemenáková, D. and all: Internal Standards. Textile 

Research Centre Textile. Faculty of textile engineering. 

Technical University of Liberec 2004. 


26


[5] Operating instructions Zweigle G 567. www.zweigle.com. 

Last Update 20.10. 2007. 

[6] Otsu, N.: A Threshold Selection Method from Gray-Level 

Histograms, IEEE Transactions on Systems, Man, and 

Cybernetics, Vol. 9, No. 1, 1979, pp. 62-66. 

[7] Oxenham, W., Brzan, E. and Yu, C.: The abrasive 

properties of yarn. Department of Textile and Apparel, 

Technology and Management, College of Textiles, North 

Carolina State University web site. 

[8] Webtester Reutlingen, ITV Denkendorf. www.itvdenkendorf.de. 

Last Update 20.10. 2007. 

[9] Yarn abrasion tester Wira. www.wira.com. Last Update 

20.10. 2007. 

[10] Yarn on yarn abrasion test. Quantitative Measure of yarn 

durability. Tension Technology International, Technical 

Notes 18, January 2005. www.vectranfiber.com. Last 

Update 20.10. 2007. 


27

AUTEX Research Journal, Vol. 13, No 1, March 2013, DOI: 10.2478/v10304-012-0020-x © AUTEX 

Computed Microtomography in the analysis of fiber migration in yarn 

Marta Toda, Katarzyna Ewa Grabowska 

The Technical University of Lodz, The Institute of Architecture of Textiles, 

Zeromskiego street 116, 90-543 Lodz, Poland 

E-mail: marta.toda@p.lodz.pl katarzyna.grabowska@p.lodz.pl 

Abstract: 

Keywords: 

This study is a short analysis of the use of computer microphotography in fiber migration testing as a modern nondestructive 

testing method. Microtomography operates similarly to X-ray computed tomography systems used in 

medicine, but with much better resolution owing to the use of a smaller radiation spot. The internal structure is 

reconstructed as a series of two-dimensional cross-sections that are then used to create 2D and 3D morphological 

objects. This process is non-destructive and does not require special preparation of a testing material. 

Computed tomography, Microtomography, Yarn structure 

Introduction 

X-ray computed tomography (CT) is a medical imaging method 

using computer processing of obtained images (scans). Digital 

processing of geometry is used to create three-dimensional 

(3D) image inside the object from large series of twodimensional 

x-ray images taken around one axis. Although the 

computed tomography is most often used in medicine, it is also 

more and more often used for testing in materials engineering. 

Another example is the use of CT in archaeology for imaging of 

sarcophagus content or e.g. DigiMorph project of the University 

of Texas at Austin that uses a CT scanner to study biological 

and paleontological specimens. X-ray tomography offers a 

powerful tool that enables internal structure of textiles to be 

explored before and after deformation and provides information 

on their geometry. 

Computed Tomography 

Using a high resolution CT scanner is a non-destructive 

technique that may be used to obtain internal images of materials. 

Schematic operation principle is presented in Figure 1. An X-ray 

beam passing through a rotating sample gives a two-dimensional 

projection that is recorded by the CCD detector. The purpose of 

the detector scintillator (for example monocrystal of cadmium) 

is to convert the x-ray energy into visible light to protect CCD 

matrix against radiation. In classical tomography, 2D projection 

is made up of attenuation coefficients of each stage of object 

scanning. Data collected are then used for numerical volume 

reconstruction by using an algorithm filtering out rear projection. 

As a final result, 3D model is obtained [1]. 

Industrial CT 

Industrial computed tomography is a process that uses X-ray 

equipment to produce a 3D model of components both in outer 

and internal structure. Industrial CT has already been used in 

many areas of industry to monitor internal components. 

The conversion of CT data into CAD models by using tools 

available on market is still quite difficult, and therefore this 

area offers large potential of development. In future, 3D-CAD 

rendering of data series from 3D tomography for simulation 

and analysis with the finite element method will be even more 

important. The reason for this is the fact that 3D model instead 

of theoretical model will be used as a basis for calculations for 

geometry of objects [2]. 

Figure 1. Principle of X-Ray Tomography 


28


CT in textiles 

Toshihiro Shinohara [3] proposed to extract the positional 

information of its yarns from a 3D image obtained from its X-ray 

CT images. For extracting the yarn positional information, 

filament directions are first estimated by correlating the 3D 

image with a filament model. Then, by averaging the estimated 

filament directions in a calculating region, the yarn direction is 

estimated. 

Knowledge on the micro-structure of composites is important 

to investigate their mechanical properties such as the tensile 

stiffness. X-ray computer tomography, Pandita S.D., is used to 

characterize the micro-structure of rib weft knitted fabrics. This 

tomography technique provides three-dimensional images 

of rib weft knitted fabrics [4]. CT allows analyzing the spatial 

distribution of fibers, for example in textiles. The high resolution 

images for plain weave give interesting information in terms 

of spatial fiber distribution for both deformed and undeformed 

configurations [5]. 

Fiber migration 

The existing strength analyses of twisted yarns showed a 

significant impact of fiber movement phenomenon, fiber 

migration. The last research shows that yarns are not ideally 

twisted in a helical structure and it is believed that fibers in 

yarn have stochastic distribution with variable distance of a 

fiber trajectory from the yarn axis. This phenomenon is closely 

related to tension of component fibers in the yarn twisting 

process. 

In a yarn drawing process, there are some phenomena that 

must be taken into consideration in theoretical analysis, that is: 

• Sliding the fibers apart and their breaking during drawing; 

• Fiber migration in the process of twisting them into yarn; 

• Compression of fibers and yarn; 

• Unevenness of yarn. 

The yarn structure is described by three main parameters: 

twist, fiber density, and fiber migration. The yarn structure may 

be divided into areas with larger or smaller radial coordinate. 

Packing density of fibers across the yarn cross-section is 

different. Its significant reduction occurs rather on the outer 

surface and in the core than between those areas. So far, the 

idealized helical yarn geometry according to Hearle J.W.S. 

theory (Figure 2a) has been taken in theoretical assumptions. 

In this theoretical model, it is assumed that yarn is of circular 

section and fibers form helical trajectories around the concentric 

cylinders with constant radius. Each fiber has a helical 

trajectory with constant pitch h and radius r, and helix angle 

increasing from 0 for r=0 to α for r=R. In reality, fiber trajectories 

in yarn are of very complicated shapes; one of such trajectories 

indicated with thick line is shown in Figure 2b. Therefore, a fiber 

trajectory is often questionable [6,7] 

Materials and methods 

Tests of five samples of multiple threads with a length of 5 mm 

were performed. Measurements of thread microtomography 

were taken with the best possible resolution of 2.5 µm. 

X-ray computed tomography is a non-destructive testing 

method of materials that makes it possible to obtain flat or 

spatial distribution of the selected physical quantity. It utilizes 

object projections taken from different directions to produce 

2D sectional images or 3D spatial images. As a result of 

measurements, precise images of internal details of an object 

tested are obtained. Image scanning was performed by using 

SkyScan 1174 micro-CT scanner. An example 3D model of the 

sample D5 (1100 twists/m) is presented in Figure 3. 

Results 

As a result of imaging with CT, a set of images (scans) of the 

yarn cross-sections was obtained for each yarn sample as 

a bitmap that was then assembled into 3D model. Figures 4 

- 8 present yarn in a longitudinal section for seven different 

distances from the yarn axis. 

Discussions 

As a result of CT scanning analysis of cotton yarn, it was found 

that the regularity of fiber distribution in section increases with 

increasing the number of yarn twists. Packing density of fibers 

across the yarn cross-section is different. Reduction occurs 

on the outer surface. Considering the helical arrangement of 

fibers in yarn, it was found that fibers, in the core area, are in a 

straightened form or twisted one with a small helix radius. The 

fibers, located further from the yarn axis, increase helix radius 

r and relocate toward the outer surface. 

As a result of scanning analysis of the samples of obtained 

yarns made of staple fibers, it was found that the regularity 

of fiber distribution in section increases with increasing the 

number of yarn twists. Packing density of fibers in the yarn 

longitudinal sections is different depending on the distance of 

section from the yarn axis. Considering the helical arrangement 

Figure 2. a) Idealized twisted yarn geometry [7]; b) fiber trajectory in 

yarn [6] 


29

auteX research journal, vol. 13, No 1, March 2013, Doi: 10.2478/v10304-012-0020-x © auteX 

Figure 3. An example of a three-dimensional model of the sample D5 – 1100 twist/m 

Figure 4. The longitudinal section yarn in seven different distances from the axis of the yarn - cotton yarn 25tex, 700 twist/m 



30






31


of fibers in yarn, it was found that fibers, in the core area, are in 

a straightened form and/or twisted one with a small helix radius 

and their uniform distribution. The fibers, located further from 

the yarn axis, increase their helix radius and are pushed toward 

the outer surface of the yarn. 

Conclusion 

There are many advantages to using CT scanning over 

traditional. The main points include: 

• A non-destructive test for inspection and metrology; 

• Design requirements for both internal and external components 

are validated quickly and accurately. Development costs are 

reduced in creating the first CAD model; 

• Product quality is improved to reduce the risk of recalls; 

• Internal complex features can be precisely measured 

without destructive testing; 

• Parts are scanned in a free state environment with no 

fixtureing applying stresses which could damage delicate 

part or display warping that is not present in the part; 

• For the first time rapid prototyping of the internal 

components can be completed without the daunting task 

of creating the CAD file from scratch. 

References 

[1] Merle P., et al. X-Ray Computed Tomography on MMS. 

MMS Assess Thematic Network. 

[2] Flisch, A., et al. Industrial Computer Tomography in 

Reverse Engineering Applications. DGZfP-Proceedings 

BB 67-CD Paper 8, Computerized Tomography for 

Industrial Applications and Image Processing in 

Radiology, March 15–17, 1999, Berlin, Germany. 

[3] Shinohara T, Takayama J., Shinji Ohyama S. 

KobayashiA.: Extraction of Yarn Positional 

Information from a Three-dimensional CT Image of 

Textile Fabric using Yarn Tracing with a Filament 

Model for Structure Analysis. Textile Research 

Journal 2010; 80; 623-630. 

[4] Pandita S.D., Verpoest I.:Prediction of the tensile 

stiffness of weft knitted fabric composites based on X-ray 

tomography images. Composites Science and Technology 

63 (2003) 311–325. 

[5] Badel P. , E. Vidal-Sallé E., Maire E., Boisse P.: 

Simulation And Tomography Analysis of Textile Composite 

Reinforcement Deformation at the Mesoscopic Scale. Int J 

Mater Form (2009) Vol. 2 Suppl 1:189–192. 

[6] El-Behery H.M.: Study of Theories of Fiber Migration— 

Need for More Fundamental Approach and Further 

Studies. Textile Research Journal April 1968 38: 321- 

331. 

[7] Hearle J.W.S., Gupta B.S., Merchant V.B.: Migration of 

Fibers in Yarns. Part I: Characterization and Idealization 

of Migration Behavior. Textile Research Journal. Kwiecień 

1965. pp329-334. 


32

New possibility oF objective evaluatioN oF yarN appearaNce: part ii

You also want an ePaper? Increase the reach of your titles

Delete template?

Save as template?