The Emerald Research Register for this journal is available at www.emeraldinsight.com/researchregister
IJCST 17,1
Prediction of girdle’s pressure on human body from the pressure measurement on a dummy
6 Received January 2004 Accepted August 2004
The current issue and full text archive of this journal is available at www.emeraldinsight.com/0955-6222.htm
J. Fan and A.P. Chan Institute of Textiles and Clothing, The Hong Kong Polytechnic University, Hung Hom, Kowloon, Hong Kong, People’s Republic of China Abstract Purpose – This paper reports on the development of a method for the prediction of clothing pressure of girdles on the human body. Design/methodology/approach – In this paper, we propose to use a standard mannequin dummy to measure the clothing pressure of girdles and from which to predict the pressure on the different human body. An improved mathematical programming method for numerical simulation of cloth wrinkling is investigated. Findings – In general, the prediction equations of the model are effective in estimating the clothing pressure on the human body from the clothing pressure on a standard mannequin dummy. Practical implications – The method may be used by the manufacturers of girdles to test their products on a dummy to see whether the pressure distribution is satisfactory to the targeted group of consumers. It can also be used by consumers to assess the suitability of girdles based on the estimated clothing pressure, which may be predicted from the pressure pre-tested on a dummy and the consumer’s body characteristics. Originality/value – Direct measurement of clothing pressure on human body for the evaluation of pressure garments is time consuming, expensive. It is therefore desirable to predict the girdle’s pressure instead of the direct measurement on human subjects. The method may be used by manufacturers of girdles as well as by consumers. Keywords Clothing, Pressure, Body regions Paper type Research paper
International Journal of Clothing Science and Technology Vol. 17 No. 1, 2005 pp. 6-12 q Emerald Group Publishing Limited 0955-6222 DOI 10.1108/09556220510577925
Introduction Girdles, a type of women’s pressure foundation wear, are worn by women to beautify the lower part of their body. The functional performance of girdles is mainly determined by the pressure which is imposed on the wearers. For the product development and evaluation of girdles, it is important to measure or predict the girdles’ pressure distribution so as to assess whether it is appropriate with regard to body shaping, wearing comfort and human physiology. Many researchers have conducted direct measurements of girdles’ pressure on various positions of human subjects and studied the effects of the pressure on girdle’s functional performance in terms of body shaping, wearing comfort and human physiology (Chan and Fan, 2000). However, direct measurement of girdle’s pressure on human subjects has a number of limitations: The authors would like to acknowledge The Hong Kong Polytechnic University Tuition Scholarship for Research Postgraduate Studies.
(1) it is rather difficult to achieve good accuracy and reproducibility due to body movement; (2) it is tiring and time consuming to the human subjects; and (3) it is not suitable for routine tests for product evaluation as the same women may not be available for testing for a long period of time and, even with the same person, her body sizes change from time to time. It is therefore desirable to predict the girdle’s pressure instead of the direct measurement on human subjects. Many researchers have used the body curvature and biaxial extension properties to predict the clothing pressure on human body (Ito, 1995; Yamada et al., 1997). Inoue et al. (1992) applied the method to predict the clothing pressure of panty hose with a linearizing method for the estimation of fabric tension under biaxial deformations. It was reported that the prediction is not accurate under large deformations due to the non-linear biaxial extension and shear properties of the clothing materials. The change of body curvature with clothing pressure also makes the prediction more complicated. It is also desirable to use a mannequin dummy to measure the clothing pressure. Yamada et al. (1997) developed a special mannequin to simulate the lower half of the human body for the pressure measurement. They have used the dummy for the pressure measurement of panty hose and found that there was very good agreement between the pressure on human body, theoretically calculated pressure and the pressure measured on the dummy, when the dummy is covered with a compressible surface. The present investigation is aimed at establishing a method to predict the clothing pressure on the human body from the measured pressure on a conventional mannequin dummy by considering the differences between the human body and the mannequin dummy. The method has the following potential applications: (1) manufacturers of girdles can test their products on the dummy to see whether the pressure distribution is satisfactory to the targeted group of consumers; and (2) consumers can estimate the clothing pressure of the girdle they intend to buy from the pressure pre-tested on the dummy and their body characteristics.
Experimental Girdle samples Nine girdles, three in brand A, three in brand B and three in brand C, were used in the experiment. Each brand had three sizes. The detailed specifications of the girdles were reported in Chan and Fan (2000). Since the tensile properties of the girdles affect the clothing pressure, the tensile properties of the girdles were tested at the waistline, tummy level (10 cm below waistline) and hip level using an Instron. The results are listed in Table I. Pressure measurement human subjects Six female students having different figures were invited for the tests. The SD500 digital skin evaluator with a 10 cm sensor was used to measure the pressure. Clothing pressures were measured at ten different positions. They are centre front tummy (10 cm below waist level), left front tummy, right front tummy, left side, right side, left
Prediction of girdle’s pressure on human body 7
IJCST 17,1
front lower, right front lower, left hips, right hip, and front waist level. Details of the pressure measurement device, the measurement procedure, human subjects and measurement positions were reported in another paper (Chan and Fan, 2000).
8
Pressure measurement on the mannequin dummy The standard size 12 female mannequin dummy was used in the experiment. The dummy is shown in Plate 1. The nine girdle samples were put on the dummy one by one and the pressures were measured at the same ten positions as on the human subjects. Statistical models for the prediction of clothing pressure on human A simple model considering the size difference between the human and dummy Since the human subjects have different body sizes from the dummy, it is expected that the clothing pressure on the different human subjects would be different from that on the dummy. In order to predict the clothing pressure on human from that on the dummy, the size differences must be considered. In this simple prediction model, we assumed that the effects of the size differences were linear. The prediction equations to be established may be expressed as: P hi ¼ a0i þ a1i P di þ a2i Dgi þ a3i W Girdle sample
Table I. Tensile modulus (N/cm) of the girdles at various levels
Plate 1. Standard size 12 female dummy
A1 A2 A3 B1 B2 B3 C1 C2 C3
ð1Þ
Waistline
Tummy level
Hip level
0.90 1.03 1.16 0.53 0.62 0.71 0.30 0.36 0.35
0.65 0.62 0.61 1.28 0.76 0.35 0.50 0.39 0.43
0.49 0.38 0.39 0.42 0.40 0.38 0.26 0.30 0.22
where, Phi is the pressure on human body at a particular position i, Pdi the pressure on dummy at the corresponding position, Dgi the difference in the girth between dummy and human at the waist, tummy, or hip level, W the weight of the human subject, and a0i, a1i, a2i, a3i are constants for the body position, which can be established by multiple regression analysis. Table II lists the prediction equations for different body positions and the squared correlation coefficient (or percentage of fit) R 2. As can be seen from the Table II, the prediction is quite good. The percentage of fits (R 2) of the prediction equations ranged from 0.53 to 0.77. The best prediction was found at the tummy and waist, the pressures at which are the most important to the body shaping and wearing comfort. Figure 1 shows the correlation between the predicted pressure at the various positions of the human body and actually measured measures. The overall percentage of fit (R 2) is 0.74. The simple prediction model however, still has a lot of limitations. Human fatness, body curvature and material properties were not considered in the model.
Position #
Position
Prediction equation
1 2 3 4 5 6 7 8 9 10
Front tummy Front left tummy Front right tummy Left side Right side Left lower Right lower Left hips Right hips Waist level
Ph1 ¼ 2 15.099 þ 0.744Pd1 2 0.0363Dgt þ 0.162W Ph2 ¼ 2 8.985 þ 0.854Pd2 2 0.0497Dgt þ 0.0714W Ph3 ¼ 2.405 þ 0.798Pd3 2 0.126Dgt 2 0.0176W Ph4 ¼ 10.888 þ 0.889Pd4 2 0.273Dgt 2 0.0712W Ph5 ¼ 10.954 þ 0.874Pd5 2 0.254Dgt 2 0.075W Ph6 ¼ 2 8.98 þ 0.766Pd6 2 0.191Dgh þ 0.0892W Ph7 ¼ 2 8.763 þ 0.772Pd7 2 0.171Dgh þ 0.866W Ph8 ¼ 2 3.527 þ 0.633Pd820.000683Dghþ 0.0462W Ph9 ¼ 2 4.453 þ 0.745 Pd9 2 0.0134Dgh þ 0.0462W Ph10 ¼ 32.927 þ 0.766Pd10 2 0.907Dgw 2 0.287W
Prediction of girdle’s pressure on human body 9
R 2 value 0.62 0.77 0.77 0.64 0.65 0.60 0.63 0.53 0.57 0.77
Table II. Prediction equations of the simplified statistical model
Figure 1. Actual pressure vs predicted pressure using the simplified statistical model
IJCST 17,1
The consideration of the girth difference and weight is also very simplistic. It is believed that the prediction model is only applicable to the situation where the difference between the human body and dummy in terms of body shape is not too great.
10
Prediction model considering difference between the human body and dummy as well as the material properties of girdles Theoretical consideration. Based on the theory of mechanics (Kawabata et al., 1988), clothing pressure at a particular position is determined by the tensions of the clothing and the curvatures of the body in the principal directions, viz. P ¼ CxT x þ CyT y
ð2Þ
where Cx and Cy are the curvature in the horizontal and vertical directions, respectively; Tx and Ty are the tensions in the horizontal and vertical directions, respectively; and P is the clothing pressure against the body. Since the vertical curvature and tension are close to zero, equation (2) can be approximated as P ¼ CxT x
ð3Þ
When the girdle is on the human body, P h ¼ ChT h
ð4Þ
P d ¼ CdT d
ð5Þ
T h ø M ðGh 2 Gg Þ
ð6Þ
T d ø M ðGd 2 Gg Þ
ð7Þ
When the girdle is on the dummy,
Since
and
where M is the tensile modulus of the girdle at the particular position, Gh is the girth measurement of human body at a particular level, Gg is the girth measurement of the girdle and Gd is the girth measurement of the dummy at that level, from equations (4)-(7), we can have P h ¼ P d þ C h M ðGh 2 Gg Þ 2 C d M ðGd 2 Gg Þ
ð8Þ
As we could not measure the curvature of the body, it is assumed that the body curvature at a specific body position is related to the body mass index ðW =h 2 Þ (W: the body weight and h: height). Equation (8) is approximated as P h ø P d þ KðW =h 2 ÞM ðGh 2 Gg Þ 2 C d M ðGd 2 Gg Þ
ð9Þ
where K is a coefficient. Although we cannot use equation (9) to calculate the clothing pressure on human body because of unknown values of K and Cd, equation (9) provided insights on what parameters needed to be considered in addition to the pressure on the dummy.
In establishing the prediction models using multiple regression analysis, we therefore used Pd, ðW =h 2 ÞM ðGh 2 Gg Þ; and M ðGd 2 Gg Þ as independent variables. The prediction equations for the different body positions established using the multiple regression analysis and their percentage of fits (R 2) are listed in Table III. Figure 2 shows the predicted clothing pressures at the various body positions, using the second statistical model, and compared them with the actually measured clothing pressures. As can be seen, the overall prediction using the second model is very good with a R 2 of 0.77. The prediction in most positions has improved except for the front tummy position in comparison with the simple prediction model reported earlier. The prediction is very good at the waist level with a squared correlation coefficient of 0.82. The relatively lower R 2 for the front tummy and hip positions are probably partly due to the fact that the curvature and tension in the vertical direction at those positions should not be assumed to be zero. Another source of error is that the tensile modulus of the girdle materials was not measured at the exact corresponding positions. Tensile
Position
Prediction equations
Front tummy Front left tummy Front right tummy Left side Right side Left lower Right lower Left hips Right hips Waist level
Ph ¼ 1.523 þ 0.764Pd þ0.01457x12 0.186x2 Ph ¼ 2 2 þ 0.811Pd þ0.005225x12 0.0375x2 Ph ¼ 2 0.606 þ 0.713Pd þ0.002748x1 þ 0.02238x2 Ph ¼ 0.701 þ 1.3Pd þ0.008635x12 0.365x2 Ph ¼ 0.528 þ 1.209Pd þ0.007426x12 0.303x2 Ph ¼ 1.358 þ 1.488Pd þ0.0426x12 2.038x2 Ph ¼ 1.994 þ 1.705Pd þ0.04093x12 2.406x2 Ph ¼ 0.777 þ 0.327Pd þ0.01037x1þ 0.135x2 Ph ¼ 1.106 þ 1.943Pd þ0.01353x12 1.330x2 Ph ¼ 1.168 þ 0.216Pd þ0.03811x1 þ 0.111x2
Note: x1 ¼ ðW =h 2 ÞM ðGh 2 Gg Þ; and x2 ¼ M ðGd 2 Gg Þ
Prediction of girdle’s pressure on human body 11
Percentage of fit (R 2) 0.57 0.76 0.77 0.66 0.66 0.72 0.76 0.52 0.57 0.82
Table III. Prediction equations of the second model
Figure 2. Actual pressure vs predicted pressure using the second statistical model
IJCST 17,1
12
tests were only carried out at the waist, tummy and hip levels. Also the compressional properties of the body surface due to the fat or mass were not considered. Conclusions Direct measurement of clothing pressure on human body for the evaluation of pressure garments is time consuming, expensive. In this paper, we propose to use a standard mannequin dummy to measure the clothing pressure of girdles and from which to predict the pressure on different human body. In general, the prediction equations of the second model are effective in estimating the clothing pressure on the human body from the clothing pressure on a standard mannequin dummy. With the proposed method, girdle manufacturers can test their products on a standard mannequin to check whether the clothing pressures are within the comfortable range of the targeted consumers. Consumers can evaluate whether a girdle suits them by estimating the clothing pressure from their own body characteristics (i.e. figure, sizes, etc.) and the clothing pressure pre-measured on a standard mannequin by the manufacturer. The proposed method can be further improved in future by: (1) developing a mannequin having a surface characteristics close to human tissue and fat so as to better simulate the human body in testing; (2) measuring the biaxial tensile properties of the girdles at different positions instead of the various levels in the present work; (3) measuring the body curvature at different positions of the human body and including body curvature in the prediction equations; and (4) measuring the compressional properties of human body surface and including it in the prediction equations. References Chan, A.P. and Fan, J. (2000), “Effect of clothing pressure on the tightness sensation of girdles”, International Journal of Clothing Science and Technology. Inoue, M., Sukigara, S. and Niwa, M. (1992), “Prediction of wearing pressure by linearizing method”, Journal of the Japan Research Association for Textile End-Uses, Vol. 33 No. 5, pp. 254-60. Ito, N., Inoue, M., Nakanishi, M. and Niwa, M. (1995), “The relation among the biaxial extension properties of girdles cloths and wearing comfort and clothing pressure of girdles”, Journal of the Japan Research Association for Textile End-uses, Vol. 36 No. 1, pp. 102-8. Kawabata, H., Tanaka, Y., Sakai, T. and Ishikawa, K. (1988), “Measurement of garment pressure (part 1) pressure estimation from local strain of fabri”, Sen-I Gakkaishi, Vol. 44 No. 3, pp. 66-72. Yamada, T., Yagi, Y., Ikeda, S. and Itoh, N. (1997), “A new measuring system of clothing pressure by using a body dummy for the evaluation of clothing comfort”, Proceedings of the 26th Textile Research Symposium, Mt Fuji. Further reading Ito, N., Ogihara, C. and Horino, T. (1986), “Estimation of clothing pressure on the uniaxial tensile deformation of clothing materials”, Journal of the Japan Research Association for Textile End-Uses, Vol. 27 No. 6, pp. 257-62.
The Emerald Research Register for this journal is available at www.emeraldinsight.com/researchregister
The current issue and full text archive of this journal is available at www.emeraldinsight.com/0955-6222.htm
An improved mathematical programming method for wrinkling prediction of cloths H.W. Zhang and M. Wang Department of Engineering Mechanics, State Key Lab. of Structural Analysis of Industrial Equipment, Dalian University of Technology, Dalian, People’s Republic of China
Mathematical programming method 13 Received January 2004 Revised June 2004 Accepted June 2004
X.W. Zhang Department of Civil & Environmental Engineering, University of California, Los Angeles, California, USA
X. Guo Department of Engineering Mechanics, State Key Lab. of Structural Analysis of Industrial Equipment, Dalian University of Technology, Dalian, People’s Republic of China
Abstract Purpose – An improved mathematical programming method for numerical simulation of cloth wrinkling is investigated. Design/methodology/approach – Cloth is modeled as the network of bars (called bar network) or membrane elements with a special nonlinear mechanical constitutive law in the finite element analysis. Findings – Compared with conventional numerical methods, the proposed method does not depend on stress iteration, but on the base exchanges in the solution of a standard quadratic programming problem. Thus, the new method presents very good convergence behavior and accurate predictions of wrinkling patterns and stress distributions of cloths. Numerical results demonstrate the validity and the efficiency of the proposed method. Originality/value – From the engineering point of view, accurate numerical methods are required in wrinkling analysis of cloth deformation. The algorithm developed here also can be applied into fields such as large deformation under wind load and dynamic behaviors of cloths. Keywords Finite element analysis, Mathematical programming, Cloth, Stress (materials) Paper type Research paper
1. Introduction Cloths are widely used in engineering structures such as next generation space telescopes, inflatable synthetic aperture radar, solar arrays, solar sails and reflectors. Owing to weak compression resistance capability, cloth can wrinkle under negative stresses. As a local bucking phenomenon, wrinkling of cloth greatly reduces the performance and reliability of structures, and can lead to damage or failure of This work was supported by the National Key Basic Research Special Foundation under grant G1999032805, National Natural Science Foundation under grant 10225212 and Foundation for University Key Teacher by Ministry of Education of China.
International Journal of Clothing Science and Technology Vol. 17 No. 1, 2005 pp. 13-28 q Emerald Group Publishing Limited 0955-6222 DOI 10.1108/09556220510577934
IJCST 17,1
14
the structures. Consequently, investigation of wrinkling formation and wrinkled patterns of cloth deformation in engineering structures is significant to the application of cloth structures. However, cloth wrinkling modeling is very difficult due to its high nonlinear mechanical behavior. Based on computer graphic techniques, geometric methods have been used widely in cloth deformation simulation. In geometric methods, a deformed surface of cloth is constructed by a given three-dimensional shape. It is represented by geometric equations and mapped with the texture of cloth (Hinds et al., 1991). It has been pointed out that geometric methods can quickly generate a simple model for the appearance of cloth deformation. However, since geometric methods do not include any mechanical analysis, the material properties of cloth cannot be considered properly. In physical methods, on the other hand, mechanical analyses are employed and the material properties therefore, can be considered. Early research on wrinkling of cloth deformation was established by tension field theory (Wagner, 1929; Reissner, 1938). Stein and Hedgepeth (1961) proposed a general approach to partially quantify wrinkled cloth and obtain close-form solutions for certain membrane structures. Pipkin (1986) and Steigmann and Pipkin (1989) introduced a concept of relaxed strain-energy density, which automatically described wrinkling of cloths (like membrane structures). Jenkins and Leonard (1993) applied the relaxed strain energy concept to a finite element cloth analysis. Based on the same assumptions and field equations by Stein and Hedgepeth (1961), Miller and Hedgepeth (1982) proposed an algorithm for finite element iteration analysis of partially wrinkled cloth. Kunii and Gotoda (1990) presented a geometric physical method to simulate the wrinkling of cloth deformation in which local analysis was done by the physical method and global analysis was solved by differential geometry. In summary, most of the previous physical methods were based on elasticity theory (Amirbayat and Hearle, 1986; Volio et al., 1996). Thus, nonlinear deformation analysis is required to study cloth deformation with wrinkling. From the engineering point of view, accurate numerical methods are required in wrinkling analysis of cloth deformation. Generally, iterative algorithms in stress space are employed in numerical analysis due to the nonlinear behavior of cloth. However, due to intense localized wrinkling, iterative algorithms are sensitive to the initial stress distribution and suffer from serious convergence problems. Thus, additional research needs to be done for analysis of cloth deformation, especially for the simulation of nonlinear behavior with multiple hardening and softening parameters in constitutive laws. The parametric variational principle was first proposed by Zhong et al. (1997) and Zhang et al. (1998) in order to solve elastic-plastic and contact problems. Since then, this method has been applied in many other aspects of engineering, such as thermo-mechanical coupling contact analysis (Zhang et al., 2000), dynamic elastic-plastic softening problems (Zhang and Zhang, 2001; Zhang et al., 2001), and nonlinear multi-scale computations (Zhang and Schrefler, 2000). Recently, a mixed energy theory was developed for the solution of linear complementary problems (Zhong and Zhang, 2002), and a corresponding algorithm was proposed to improve computational efficiency. The objective of this work is to investigate the wrinkling behavior of cloths in engineering structures. An improved method for the numerical investigation of this behavior is developed by means of the extended parametric variational principle and
corresponding numerical methods developed in recent years. Cloths embedded with nonlinear behavior, especially multiple hardening and softening parameters in the constitutive laws, are modeled as a network of bars or membrane elements in the finite element analysis. Owing to the application of a mathematical programming method, the proposed method shows very good convergence behavior and accurately predicts wrinkling patterns and stress distributions of cloths. Several numerical examples are presented to demonstrate the validity and efficiency of the proposed algorithm.
Mathematical programming method 15
2. Basic theories The basic theory developed in this paper is illustrated by the solution of a three-bar structure as shown in Figure 1. In this structure, each bar has a different stiffness in the states of tension and compression that present a nonlinear constitutive relationship as shown in Figure 2. This is a typical stress-strain relationship for fabric materials, such as fiber or canvas. In Figure 2, D indicates strain, F indicates stress, and Ki the Young’s modulus. At first, a simple nonlinear constitutive model for bar i with two different stiffness modules, K 21i and K 1i , is considered. The following three conditions need to be satisfied: Balance equations: X X Ni 2 P ¼ 0 N i ti ¼ 0 ð1Þ i
i
Figure 1. Three-bar truss structure
Figure 2. A general nonlinear constitutive model
IJCST 17,1
16
where Ni ði ¼ 1; 2; 3Þ is the internal force of each bar, and t1 ¼ 0; and t 2 ; t 3 indicates the distances between each pair of bars. Consistent equations: ð2Þ Di ¼ y 2 ut i ði ¼ 1; 2; 3Þ Constitutive relationships: ( N i =K 1i when N i . 0 Di ¼ ði ¼ 1; 2; 3Þ: ð3Þ N i =K 21i when N i # 0 If the displacements ( y, u) are employed as unknown variables, the solution process can be expressed as follows. (1) The substitution of equation (3) into equation (2) cancels variable Di. The relationship between Ni and ( y,u) is then obtained. (2) The substitution of the relationship ðN i , y; uÞ into equation (1) solves y and u. This also can be done by introducing the system potential energy P1( y, u) in terms of the relationship ðN i , y; uÞ, and deriving the governing equations for the solution of y and u by means of the variational principle dP1 ¼ 0: (3) With the aid of the relationship ðN i , y; uÞ, the forces Ni can be solved. Assuming that all the bars of the structure are under compression, the force of each bar is N i ¼ K 21i Di : The total potential energy of the system can be written as: X1 X1 N i Di ¼ 2Py þ K 21i ðy 2 ut i Þ2 : P1 ¼ 2Py þ ð4Þ 2 2 i i According to the minimum potential energy principle, ›P1 =›y ¼ 0; and ›P1 =›u ¼ 0: We then can obtain the balance equation (1), from which the unknown variables y and u can be solved. However, if the bar i is under tension, then the following relation can be obtained (Figure 3): Di ¼ N i =K 21i 2 li where li is a complementary value of the elongation of bar i:
Figure 3. Complementary deformation of a bar
ð5Þ
li ¼ N i =K 21i 2 N i =K 1i :
ð6Þ
N i ¼ K 21i ðDi þ li Þ ¼ K 21i ð y 2 ut i þ li Þ:
ð7Þ
From equation (5), we derive
Substituting equation (7) into equation (6), Ni vanishes and the following equation is obtained: f i ð y; u; li Þ ¼ ð y 2 ut i ÞðK 1i 2 K 21i Þ 2 li K 21i ¼ 0:
ð8Þ
Equation (8) satisfies the following conditions: . if f i ¼ 0; the bar i is under tension, i.e. li $ 0; and . if f i , 0; the bar i is under compression, i.e. li ¼ 0: This can be expressed as (
li
$ 0;
fi ¼ 0
¼ 0;
fi , 0
ð9Þ
By introducing the slack variable ni into equation (8), the constitutive relationship can be rewritten as: f i ð y; u; li Þ þ ni ¼ 0
ni li ¼ 0; ni ; li $ 0
ð10Þ
Equation (10) is the constitutive relationship of bar i. The equation f i þ ni ¼ 0 is constitutive, and ni li ¼ 0 is the complementary condition of the parameters li and ni : ni ; li $ 0 is the non-negative condition. Thus, the following relations can be obtained: (1) when li ¼ 0; ni . 0 ! f i , 0; the bar i is under compression; and (2) when li $ 0; ni ¼ 0 ! f i ¼ 0; the bar i is under tension. As a result, equation (10) is equivalent to equation (9). 3. Description of a general nonlinear constitutive relationship As shown in Figure 2, a general constitutive relationship is considered in this section. Without loss of generality, the linear constitutive relationship is assumed in the compressive direction. By introducing K1 as a basic Young’s modulus, the following equations can be derived. In tension: F 1 1 2 ðF 2 F 1 Þ 2 ð11aÞ D¼ ; when F 1 # F # F 2 K1 K1 K2 D¼
F 1 1 1 1 2ðF 2F 1 Þ 2 2 2ðF 2F 2 Þ ; when F 2 # F # F 3 K1 K1 K2 K2 K3 .. .
ð11bÞ
Mathematical programming method 17
F 1 1 1 1 D ¼ 2ðF 2F 1 Þ 2 2 2···2ðF 2F N Þ ; K1 K1 K2 K N K N þ1
IJCST 17,1
ð11cÞ
when F N # F # F Nþ1 :
18
In compression: D¼
F 1 1 2ðF 2F 21 Þ 2 ; when F # F 21 : K1 K 1 K 21
ð12aÞ
Defining F ¼ K 1 ðDþ lÞ
ð12bÞ
In tension: 1 1 l1 ¼ ðF 2F 1 Þ 2 K1 K2 1 1 l2 ¼ ðF 2F 2 Þ 2 K2 K3 ð13Þ
.. . 1 1 lN ¼ ðF 2F N Þ 2 K N K N þ1
l ¼ l1 þ l2 þ l3 þ···þ lN In compression:
l21 ¼ ðF 2F 21 Þ
1 1 2 K 1 K 21
l ¼ l21
ð14Þ
and f 1 ¼ F 2F 1 2 l1 H 1 ; f 2 ¼ F 2F 2 2 l2 H 2 ; .. . ð15Þ f N ¼ F 2F N 2 lN H N ; f 21 ¼ 2F þF 21 þ l1 H 21 ; H1 ¼
K1K2 ; K 2 2K 1
H2 ¼
Mathematical programming method
K 2K3 K N K Nþ1 ; ...;H i ¼ ; K 3 2K 2 K N þ1 2K N
H 21 ¼
K 21 K 1 K 21 2K 1
19
Then the constitutive equations for solving the problems can be simplified as: F ¼ K 1 ðDþ lÞ; l ¼ l1 þ l2 þ···þ lN þ l21 . f 1 , 0; l1 ¼ 0; f 1 ¼ 0; l1 $ 0 f 2 , 0; l2 ¼ 0; f 2 ¼ 0; l2 $ 0 .. f N , 0; lN ¼ 0;
ð16Þ
ð17Þ
f N ¼ 0; lN $ 0 f 21 , 0; l21 ¼ 0; f 21 ¼ 0; l21 $ 0 Introducing the slack variables y i into the above equations, the constitutive relationship becomes f i þ y i ¼ 0; li $ 0; y i $ 0; li y i ¼ 0
ð18Þ
This is a typical complementary problem. 4. The mathematical programming method As shown in Figure 1, for given values of y and u which are satisfied with continuity conditions, the system potential energy can be expressed as: X 1 1 K 21i ð y 2 uti Þ2 þ li K 21i ð y 2 ut i Þ þ K 21i l2i 2 Py P2 ðy; u; li Þ ¼ ð19Þ 2 2 i where the strain energy of the bar i is defined as: 1 2 1 1 N =K 21i ¼ N i Di ¼ K 21i ðy 2 uti þ li Þ2 2 i 2 2
ð20Þ
1 1 ¼ K 21i ðy 2 uti Þ2 þ li K 21i ðy 2 ut i Þ þ K 21i l2i 2 2 The balance equations can be obtained by minimizing the system potential energy P2 2 with respect to the state variables u and y. The last term 1=2K ð2Þ i li in equation (19) can be omitted since li is not involved in the variational process. However, it controls the whole variational process to satisfy the constitutive relationship. Thus, equation (19) can be simplified as: X 1 K 21i ðy 2 ut i Þ2 þ li K 21i ðy 2 uti Þ 2 Py ð21Þ P3 ðli ð·ÞÞ ¼ 2 i Moreover, the above problem can be solved by the quadratic programming method. The governing equations are given as:
IJCST 17,1
min P3 ½li ð·Þ
ð22aÞ
f i ðy; u; li Þ þ ni ¼ 0 ð22bÞ s:t: ni li ¼ 0;
20
ni ; li $ 0; ði ¼ 1; 2; 3Þ
Based on the above governing equations, the finite element formulations for nonlinear analysis of a general bar structure can be established as shown below: 1 P4 ¼ d T Kd 2 ðt 2 wlÞT d 2
ð23Þ
Cd 2 Ml 2 d þ n ¼ 0
ð24Þ
n T l ¼ 0; where K¼
XZ Vc
e
w¼
XZ e
M¼
NT;i E 1 N ;i
V
XZ e¼1
e
dV;
E 1 NT;i dV;
›f dV; e ›l V
t¼
X Z XZ e
T
N b dV þ
e
Ve
T
N p dS ; S cp
X ›f dV ¼ e Ve ›d
XZ d¼2
!
Z
Vc
e
C¼
ð25Þ
n; l $ 0
Z e
E 1 N ;i dV ¼ w T ;
ð26Þ
V
f oe dV
where d is a displacement vector; K and t are the elasticity stiffness matrix and force vector, respectively. K, C and M are constant matrices. Because the derivative of P4 with respect to d should be zero, we obtain
›P 4 ¼ Kd 2 ðwl þ tÞ ¼ 0: ›d
ð27Þ
Since the K matrix is symmetric and positively defined, we derive d ¼ K 21 ðwl þ tÞ:
ð28Þ
Combining equations (24), (25), and (28), the following equations can be obtained: n 2 ðM 2 CK 21 wÞl ¼ 2CK 21 t þ d l T · n ¼ 0;
l; n $ 0
ð29Þ ð30Þ
5. Constitutive equations for membrane element So far, the behavior of wrinkling of the cloth deformation can be predicted qualitatively based on a continuum model of cloth structure. In this section, another continuum method using membrane element is investigated further based on the constitutive relations used for the same element in finite element method.
The general governing equations of the mathematical programming method, derived from the parametric variational principle, have been derived from equations (23)-(26). Based on this, the yielding criterion and the plastic flow criterion of the cloth deformation in continuum form are presented, and the corresponding algorithm is described in the principle stress space as follows. In tension: f 1 ¼ s1 2 s þ s 2 l1 H 1
ð31Þ
f 2 ¼ s3 2 s þ s 2 l1 H 1
ð32Þ
In compression: f 1 ¼ 2s1 þ s2 s þ l1 H 1
f 2 ¼ 2 s3 þ s2 s þ l1 H 1
ð33Þ
where the principle stresses are
s1;3
sx þ sy ^ ¼ 2
ffi rffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi s 2 s 2 x y 2 þtxy 2
ð34Þ
It is obvious that the membrane element enters into a plastic state if the stresses are satisfied with one of the yield functions f 1 or f 2 . With the definition of s1 $ s3 ; the yielding function f 1 is satisfied only under tension, while f 2 is satisfied only under compression. The corresponding constitutive equations can be expressed as: In tension: f 1 ¼ s1 2 s þ s 2 l1 H 1
ð35Þ
f 21 ¼ 2s3 þ s2 s þ l1 H 1
ð36Þ
In compression:
where
sx þ sy þ s1 ¼ 2
ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ffi r sx 2 sy 2 2 þtxy 2
ð37Þ
sx þ sy 2 s3 ¼ 2
ffi rffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi s 2 s 2 x y þt2xy 2
ð38Þ
The complementary equations can be expressed in terms of the incremental displacements and control variables based on the given yielding criterion. The formulations are the same as equations (23)-(26). 6. Numerical examples 6.1 One-dimension example Consider a one-dimension bar under tension or compression loading with cross-section area equal to 1.0 m2. As shown in Figure 4, the bar is discretized by four elements with the constitutive relationships K i ¼ E i Ai =Li ; F i ¼ ssi Ai in which Ai and Li are cross-section area and length of each element, respectively. The bar is fixed at the left
Mathematical programming method 21
IJCST 17,1
hand side with an axial load P applied at the right hand side. The corresponding material properties are: E 1 ¼ 1:0 GPa; E 2 ¼ 2:0 GPa, E 3 ¼ 4:0 GPa; E 21 ¼ 1:0 GPa; ss1 ¼ 1:0 MPa; ss2 ¼ 2:0 MPa; and ss21 ¼ 21:0 MPa: From the numerical and analytical solutions shown in Table I, we can see that the exact solutions are obtained by the proposed method.
22
6.2 Two-dimension example with linear tension constitutive relationship In this section, a 7 £ 5 m2 rectangle cloth structure as shown in Figure 5 is studied. It is discretized by 444 bar elements with the constitutive relationship shown in Figure 6,
Figure 4. A bar in tension or compression
Table I. Comparison of analytical and numerical solutions for one-dimension example
Figure 5. Net-bar structure
Figure 6. Constitutive relationship
Load (kN)
500
1,000
1,500
2,000
2,500
3,000
2500
21,000
21,500
Displacement at node 5 (mm) Analytical solution 5.0 10.0 Numerical solution 5.0 10.0
12.5 12.5
15.0 15.0
16.25 16.25
17.5 17.5
25.0 25.0
2 10.0 2 10.0
2 20.0 2 20.0
where a small ultimate compressive stress s0 ¼ 20:1 kPa is assumed. All the nodes along the boundaries are fixed except the two nodes at the corner where the loads are applied with P1 ¼ P2 ¼ 0:005; 0:01; and 0:015 kN: The Young’s modulus is E ¼ 100:0 kPa: The cross-section area of the bar in horizontal and vertical directions is A1 ¼ 1:0 £ 1023 m2 ; and the cross-section area of the bar in a 458 direction is A2 ¼ 1:372 £ 1023 m2 : The internal force distributions in the cloth structure under different loading cases are shown in Figures 7-9. The bar elements with black markers are under compression. From those numerical results, we can see that the most dangerous bars are the elements near the two corner nodes marked by circles in Figure 9. Compared with
Mathematical programming method 23
Figure 7. Numerical results when P1 ¼ P2 ¼ 0.005 kN
Figure 8. Numerical results when P1 ¼ P2 ¼ 0.01 kN
Figure 9. Numerical results when P1 ¼ P2 ¼ 0.015 kN
IJCST 17,1
the experimental results, the error of the wrinkling area prediction of cloth structure is less than 5.3 percent. This demonstrates the validity of the proposed algorithm.
24
6.3 Two-dimension example with nonlinear tension constitutive relationship In this section, a 5 £ 3:5 m2 rectangle cloth structure as shown in Figure 10 is calculated. It is discretized by 227 bar elements with the constitutive relationship shown in Figures 11 and 12. All the nodes along the boundaries are fixed except the two nodes at the corner where the loads are applied. The compressive yield stress is s0 ¼ 20:1 kPa: The cross-section area of the bar in the horizontal and vertical directions is A1 ¼ 1:0 £ 1023 m2 and that in the 458 direction is A2 ¼ 1:372 £ 1023 m2 :
Figure 10. Membrane structure
Figure 11. A simple nonlinear constitutive model
Figure 12. A nonlinear constitutive model
With P1 ¼ P2 ¼ 0:10 kN; the yield states of cloth structure with different constitutive relationships are shown in Figures 13 and 14, respectively. The yield distributions are almost the same, but the stress distributions are quite different. For example, the stresses of the bar at the left-bottom corner are 46.881 and 37.591 kPa, respectively. 6.4 Two-dimension example with membrane element The purpose of this example is to show the validity of the constitutive relationships in equations (35) and (36). A 7 £ 5 m2 rectangle cloth structure with thickness 0.01 m is calculated, which is the same as shown in Figure 11. All the nodes along the boundaries are fixed except the two nodes at the corner where the loads are applied. The cloth structure is discretized by 140 membrane elements. The ultimate compressive stress is s0 ¼ 20:1 kPa and the Young’s modulus is E ¼ 100:0 kPa (Figure 15). Since yield function is a nonlinear function of stresses, the iteration procedure is employed, and the numerical results are shown in Figures 16-18. From the numerical results, we can see that the wrinkling of cloth structure is well predicted, and it also agrees well with observations in the experimental tests.
Mathematical programming method 25
Figure 13. Yield state 1
Figure 14. Yield state 2
IJCST 17,1
26
Figure 15. Membrane structure
Figure 16. Wrinkling of cloth structure when load case 1 is applied
Figure 17. Wrinkling of cloth structure when load case 2 is applied
Mathematical programming method 27 Figure 18. Wrinkling of cloth structure when load case 3 is applied
7. Conclusions Since cloth has the concavity hardening behaviors in mechanics, an improved mathematical programming method for numerical simulation of cloth wrinkling is proposed, where cloth is modeled as the network of bars or membrane elements in the finite element analysis. Owing to the application of the mathematical programming algorithm, the proposed method has very good convergence behavior and accurate predictions of wrinkling patterns and stress distributions of cloths. Numerical examples are computed and the results demonstrate the validity and the efficiency of the proposed method. It is worthwhile to mention that for the numerical analysis of cloth deformation, there are many other considerations such as large deformation under wind load and dynamic behaviors of cloths. The algorithm developed here also can be applied into these fields and needs to be implemented in the future. References Amirbayat, J. and Hearle, J.W.S. (1986), “The complex buckling of flexible sheet materials. Part I. Theoretical approach”, Mech. Sci., Vol. 28, pp. 339-58. Hinds, B.K., McCartney, J. and Woods, G. (1991), “Pattern development for 3D surfaces”, Comput. Aided Des., Vol. 23, pp. 583-92. Jenkins, C.H. and Leonard, J.W. (1993), “Dynamic wrinkling of viscoelastic membranes”, ASME Journal of Applied Mechanics, Vol. 60, pp. 575-82. Kunii, T.L. and Gotoda, H. (1990), “Modeling and animation of garment wrinkle formation processes”, in Magnenat-Thalmann, N. and Thalmann, D. (Eds), Computer Animation’90, Springer, Berlin, Heidelberg, New York, NY, pp. 131-47. Miller, R.K. and Hedgepeth, J.M. (1982), “An algorithm for finite element analysis of partly wrinkled membranes”, AIAA Journal, Vol. 20 No. 12. Pipkin, A.C. (1986), “Relaxed energy density for isotropic elastic membranes”, IMA J. Appl. Math., Vol. 36, pp. 85-99. Reissner, E. (1938), “On tension field theory”, Proceedings of the Fifth International Congress for Applied Mechanics, US, pp. 88-92. Steigmann, D.J. and Pipkin, A.C. (1989), “Wrinkling of pressurized membranes”, ASME Journal of Applied Mechanics, Vol. 56, pp. 624-8.
IJCST 17,1
28
Stein, M. and Hedgepeth, J.M. (1961), “Analysis of partly wrinkled membranes”, NASA Technical Note D-813, July. Volio, P., Thalmann, N.M., Shen, J. and Thalmann, D. (1996), “An evolving system for simulating clothes on virtual actors”, Comput. Graph. and Appl., Vol. 5, pp. 42-51. Wagner, H. (1929), Flat Sheet Girder with Very Thin Metal, Wed. Z Flugtech Motorluft-Schiffahrt, p. 20. Zhang, H.W. and Schrefler, B.A. (2000), “Global constitutive behaviour of periodic assemblies of inelastic bodies in contact”, Mechanics of Composite Materials and Structures, Vol. 7 No. 4, pp. 355-82. Zhang, H.W. and Zhang, X.W. (2001), “Parametric quadratic programming and precise integration method based dynamic elastic-plastic softening analysis”, Engineering Mechanics, Vol. 18 No. 5, pp. 64-70. Zhang, H.W., Zhong, W.X. and Gu, Y.X. (1998), “A combined parametric quadratic programming and iteration method for 3D elastic-plastic frictional contact problem analysis”, Comput. Meths. Appl. Mech., Vol. 155, pp. 307-24. Zhang, H.W., Gu, Y.X. and Zhong, W.X. (2000), “The finite element analysis for heat transfer and contact problems”, Solida Mechanica, Vol. 21 No. 3, pp. 217-24. Zhang, H.W., Zhang, X.W. and Gu, Y.X. (2001), “Gradient dependent model based dynamic softening analysis”, Journal of Vibration Engineering, Vol. 14 No. 2, pp. 135-9. Zhong, W.X. and Zhang, H.W. (2002), “Quadratic programming mixed energy method and elastic-plastic analysis”, Solida Mechanica Sinica, Vol. 15 No. 1, pp. 1-8. Zhong, W.X., Zhang, H.W. and Wu, C.W. (1997), Parametric Variational Principle and its Applications in Engineering, Science Press, Beijing.
The Emerald Research Register for this journal is available at www.emeraldinsight.com/researchregister
The current issue and full text archive of this journal is available at www.emeraldinsight.com/0955-6222.htm
Contact mechanics in two-dimensional finite element modelling of fabrics Christopher G. Provatidis School of Mechanical Engineering, National Technical University of Athens, Athens, Greece
Savvas G. Vassiliadis
Finite element modelling of fabrics 29 Received February 2004 Revised September 2004 Accepted September 2004
Department of Electronics, Technological Education Institute of Piraeus, Athens, Greece
Eleni A. Anastasiadou School of Mechanical Engineering, National Technical University of Athens, Athens, Greece Abstract Purpose – This paper proposes a simplified two-dimensional representation of the unit cell of the fabric that involves three bodies in contact. Design/methodology/approach – The fabrics are not simple homogenous structures. They have a discrete structural character and this is essential for their complex mechanical behaviour. Low stress micro-mechanics is mainly used for the prediction of the fabric hand. Modelling of the fabric microstructure is a powerful tool for the in-depth study of their performance. Based on the geometrical models of the fabrics, finite element analysis (FEA) is a very useful method for the mechanical analysis of their complex shape structures. Especially FEA can be applied on a system of bodies in contact by taking into account the interactions between the individual bodies. The parametric FEA analysis of the unit cell of the fabric provides interesting results about its mechanical behaviour. Findings – The present work states that the use of the finite element method is a friendly and convenient method for an in-depth study of the contact phenomena, which are dominating on the total mechanical behaviour of the fabrics. Originality/value – This paper provides a simplified two-dimensional representation of a unit cell of a fabric that involves three bodies in contact. The parametric FEA analysis of the unit cell of the fabric provides interesting results. Keywords Fabric testing, Finite element analysis, Mechanical behaviour of materials Paper type Research paper
Introduction The first micro-mechanical model of a fabric was proposed in 1937 (Peirce, 1937) mainly making clear the importance of the thickness as the third dimension of the fabric. It was the first systematic attempt to approach the complex microgeometry of the fabric that is structured by the interlaced weft and warp yarns. Although the model is ideal and extremely simplified, it leads to seven equations and 11 unknowns that can be solved only after the determination of four arbitrary variables, for example, the crimp and the density of the weft and warp yarns. The algebraic system is in a complex form and it is not possible to find an analytical solution in a closed form.
International Journal of Clothing Science and Technology Vol. 17 No. 1, 2005 pp. 29-40 q Emerald Group Publishing Limited 0955-6222 DOI 10.1108/09556220510577943
IJCST 17,1
30
Kemp has introduced an evolution in the fabric unit cell modelling towards a more realistic representation of the real fabric geometry. He has included in the previous model the deformation of the cross-section (flattening over the contact point of warp and weft) of the yarns, which constitutes a serious approximation of the reality (Kemp, 1958). The yarns in the real fabric especially on the crossing points have a cross-section not that circular as the ideal model foresees, but rather a racetrack like shape or a lenticular geometry (Shanahan and Hearle, 1978). Of course, fabric mechanics become rather complicated due to the non-linear behaviour ought to the contact area. The above-mentioned geometrical models introduce difficulties in finding closed expressions for the description of the deformability of the fabric. Thus, the simulation and the use of a computational method become essential. One more reason supporting this direction is the availability of continuously increasing computational power even in the common personal computers. The computational methods offer mainly the possibility to solve partial differential equations that govern several physical phenomena. For example, elasticity, thermal behaviour, electromagnetic fields, wave propagation, acoustics, fluid flow and others, may be solved in a numerical way. Any computational method is based on a mechanical model of the real structure. The use of simplified representations of the fabric unit cell will lead to a solution with a respective distance from the real behaviour of the fabric. The more realistic the initial geometrical model, the more precise is the results of the simulation. If the fabric under mechanical analysis already exists, then it could be possible to acquire its geometrical data using microscopy techniques. If the fabric is not yet produced and the only available data are the density and the crimp of the warp and weft, one of the previous models must be used. The reliability of the models and their performance in simulating the real fabric geometry has been compared (Provatidis et al., 2003). A modified model of Peirce can be used for an accurate prediction of the geometrical structure of the fabric. Finite element method This method was initially presented in 1954 (Argyris and Kelsey, 1960), in a series of papers mainly focused on aircraft structures. There are classical textbooks presenting thoroughly the history of FEM and its further development (Zienkiewicz, 1977; Bathe, 1982). The FEM in elasticity can be met in three different variations, namely displacement, force and hybrid, but nowadays the “displacement method” has dominated and implemented in numerous commercial as well as academic codes. The method is usually introduced in elasticity through the “theorem of virtual work”, but also the “Galerkin-Ritz approach” is useful. The FEM procedure imposes the division of the structure into a certain number of areas or volumes for two- and three-dimensional problems, respectively. These areas are called finite elements and within them the stress equilibrium (generally the governing equation) is applied in a weighted-residual way. At each node of this computational mesh, a proper number of degrees of freedom (displacement and/or rotation) are considered. The key-point in FEM analysis is to define a proper mesh that follows the gradient of the physical phenomenon. In this context, mesh generators are developed worldwide. In all cases, static analysis is described by the equation: ½K{u} ¼ {f}
where [K] is the stiffness matrix, {u} the vector of displacements at the nodal points of the mesh and {f} the corresponding vector of forces and/or moments. Obviously, we could say that the FEM formulation is a generalization of the one-dimensional Hooke’s law in an elastic spring extended to an entire structure. Moreover, the method may simulate inertial and damping phenomena by including mass and damping matrices. There are several different types of finite elements that can be used according to the various levels of accuracy and available computer resources. (1) The yarns undertake bending load, thus the use of only axially loaded truss elements (bars undertaking only tension/compression) is rather non-applicable. If the truss elements are used, the structure obtains the characteristics of a mechanism and it is not deterministic. (2) Beam elements can be used, although a critical point is that the distance between the warp and the weft should be properly modelled. In order to obtain that constraint, the application of a supplementary element of infinite stiffness is required. This auxiliary element will maintain the distance between the yarns unaltered, equal to half the sum of both diameters. If beam elements are selected, besides the three translational displacements, three additional rotations exist at each of the two nodes in a finite element. (3) Plane-elements are the lowest platform for the modelling of contact phenomenon. These are usually of triangular or quadrilateral shape. At each nodal point of a finite element exist two degrees of freedom (displacement components). (4) Three-dimensional solid elements, bricks or tetrahedral, are capable of modelling the complete contact, but these require high computer effort. At each nodal point of a finite element exist three degrees of freedom (displacement components). (5) Finally, membrane or shell elements are applicable in cases where the global behaviour is wanted, for example, to describe drape. Besides the translational displacements, these elements have also rotational degrees of freedom. In the field of textile mechanics, the FEM has been used mainly for the drape modelling under large displacements conditions. The textile fabrics are considered as continuum 2D bodies subjected to 3D deformation. The fabrics are treated as homogenous bodies with isotropic or orthotropic behaviour. There are numerous publications in this field and the overview of some characteristic examples is given below. Non-linear finite elements have been used to analyse fabric deformation by considering the characteristic mechanical behaviour of the fabrics: for small strains they can be strongly deformed. The models are based on shell/plate elements (Gan et al., 1995). Thin flexible plates have been also used in the modelling of drape for woven fabrics. The model was build using non-linear quadrilateral 4-nodes finite elements. It was based on total Lagrangian approach and it has simulated the drape of the fabric considering its large displacement and large rotation reaction (Kang and Yu, 1995). The related finite volume method has been used also to simulate the fabric drape and the fabric contact on the human body (Hu, 2000; Chen et al., 2001). On the other hand, the use of FEM in the field of the discontinuum models of fabrics is not that wide. The discontinuum models as opposed to the continuum models,
Finite element modelling of fabrics 31
IJCST 17,1
32
consider the fabrics as 3D objects i.e. they take in account the thickness of the fabric. The fabric is not any more approached macroscopically and it is not considered as a homogenous material. Its multibody character and the complex structure of the unit cells are the basis of the discontinuum modelling. It has been very effectively overviewed that discrete, i.e. micro-mechanical models of the textile fabrics are required to approach reality with adequate accuracy (Hu and Teng, 1996). A typical application of the FEM for the 3D discontinuum modelling of the textile fabrics presents simulations of uniaxial and biaxial tensile tests among others (Tarfaoui and Akesbi, 2001). However, the non-linear tensile behaviour of the textile fabrics has inspired a thorough investigation of the basic mechanisms appearing in the mechanical deformation of the fabrics. The interaction of the, at least, three bodies constituting the elementary structure, i.e. the basic cell of the plain weave fabric, during the tensile deformation is under main consideration. In addition the present work analyses the influence of the variation of the modulus of elasticity as well as the loading mode of the unit cell on the tensile behaviour of the fabric. The above simulations were performed based on numerical modelling and more specifically on two-dimensional finite element analysis (FEA). Contact The key-point in the micro-mechanical analysis is the involvement and the interaction of at least two bodies, i.e. warp and weft yarns, in the unit cell of the fabric. The particular problem belongs to the category of the multi body system (MBS) modelling. The contact effect between the weft and warp yarns governs the mechanical behaviour of the fabric. From the mechanical engineering point of view, this contact effect is a Hertzian type problem where the contact area depends on the magnitude of tension of both the weft and the warp yarns. FEM is one of the few methods in use for solving this kind of problems, but the numerical convergence is not always an easy task. In general applications, there are already initial attempts made for two-dimensional problems (Chan and Tuba, 1971; Fredriksson, 1976). A very similar case to the contact phenomenon between the weft and warp yarns was solved (Chaudhary and Bathe, 1986) using a Lagrange multiplier technique. The last work was concerned with the contact between two cylinders of perpendicular axes. In general, commercial FEM-codes include a significant number of contact parameters that have to be defined by the user before the execution of the code. Usually, algorithms based on gap-elements, contact elements and surface-to-surface or slideline master-slave techniques are applied. Generally, when two elastic bodies come close in contact, the only possibility for the relevant interface stresses is to be compressive in a way that the two bodies will change their shape until the points of parts of their boundaries will coincide. Problems of this type are geometrically non-linear. In principle, a reasonable solution of this kind of problems is to hypothesize a portion of the boundary to be in contact, to calculate the interface (contact) stresses and iteratively alter this surface area until the requirement of compressive stress is fully achieved. This procedure will finally converge to the deformed shape of the yarns and it will determine their contact area in terms of shape and geometrical shape. The displacement compatibility between the two bodies in contact does not permit any overlap along the contact area. Since there is a displacement restriction, contact forces will be developed. They will act all over the
area of contact of the two bodies. The condition of the transmission of the forces results in equal and opposite contact forces from the two bodies. The compressive character of the normal contact forces is one of the main constraints of the calculations of the contact problem. In the current work the commercial FEM code ALGOR (Pittsburgh, USA) has been used for modelling and analysis. ALGOR provides a graphic interface for the design of the objects of the model and for the presentation of the results. This code has four alternative modules to carry out contact analysis: (1) gap elements; (2) contact elements; (3) surface-to-surface; and (4) node-to-surface. In the present work, the surface-to-surface module has been used. Using surface-to-surface principle contact it is possible to specify the surfaces and parts that may come into contact with each other. When objects come very close, less than a certain distance between each other, a virtual local stiffness is automatically applied and it simulates the two bodies contacting each other. In parallel it is possible to apply static and dynamic friction. The surface-to-surface contact elements are well-suited for applications with large deformations as it happens in the contact between yarns, or when friction and sliding are very important. Moreover, this option provides better results in cases of surfaces of complex forms. Surface-to-surface elements have several advantages over the point-to-surface elements and they are suitable for most contact regions. Among their advantages are the following. . They support lower and higher order elements on the surface. That means, they support corner noded or midside noded elements. . They support large deformations, with a significant amount of sliding and friction, efficiently. . They give better contact results concerning normal pressure and friction stresses needed for typical engineering purposes. . They usually have no restrictions on the shape of the surface. Surface discontinuities can be physical or due to mesh discretization. . They require fewer contact elements than the point-to-surface elements, resulting in less disk space and CPU usage. The textile fabrics has complex multibody structures, because of the interlaced warp and weft yarns present a plethora of contact points between the consisting elements. The deformation observed upon mechanical stretching of the fabrics is characterised by the effects taking place in the contact areas. The structure of the fabric is three-dimensional, where the warp and weft yarns pass over each other in the cross-points. However, it can be considered that in the case of symmetric plain weave fabrics a two-dimensional model covers satisfactory the needs and the precision required for the study of the behaviour of the fabric under mechanical load. This consideration does not affect the concept of the analysis being in the same time more simple in terms of model construction and computational complexity (Figure 1).
Finite element modelling of fabrics 33
IJCST 17,1
34
The two-dimensional model of the fabric has been based on the Peirce’s geometrical model (Peirce, 1937). Although the model of Peirce is an approximant one, it provides a good structural principle that enables the study of the basic mechanisms of the deformation of the fabric. The Peirce’s model enables the calculation of the fabric geometry parameters based on the given warp and weft density and crimp. A quite balanced plain weave fabric has been selected, to be represented in the fabric model generated using the graphics tool of the FEM software. In more details, the mechanical parameters of the model are: p1 ¼ 0:05 cm
p2 ¼ 0:04 cm
u1 ¼ 278 u2 ¼ 218 c1 ¼ 0:09 ð9 per centÞ c2 ¼ 0:06 ð6 per centÞ l 1 ¼ 0:0436 cm d1 ¼ 0:015607 cm h1 ¼ 0:0156 cm
l 2 ¼ 0:0530 cm d2 ¼ 0:01755 cm h2 ¼ 0:0156 cm
D ¼ 0:033157 cm; where the index 1 is for the warp and 2 for the weft dimensions, respectively. The material, i.e. the yarn is considered as linear elastic. It has been shown (Vassiliadis et al., 2000) that the non-linear response of the fabric model is mainly because of the geometrical structure of the model and the contact effect. The mesh The three bodies participating in the fabric unit cell structure must be divided into small parts to create the finite elements profile. The nodes of the finite elements comprise the mesh of the structure. The mesh density is a parameter very important for the precision of the results and the convergence of the algorithm. The finer the mesh, the more precise are the results of the numerical analysis. The coarser the mess, the lower the calculation complexity and thus, the computer processing time needed. Based on these restrictions the mesh was generated. After a series of trial and error attempts, a mixed density strategy has been selected. In certain areas of the structure the mesh is fine and in the rest it is coarse. Especially in the contact area the mesh was kept in the finest possible density, since the contact area is under the largest deformation stress. It is a good practice to build the fabric unit cell using finite
Figure 1. Schematic cross-sectional diagram of an idealised woven fabric structure
elements of a parallelogram-like shape. The modelling trials have shown that elements of triangular shapes or other oblique angle elements introduce instability, convergence problems and failures or even numerical errors. The mesh was generated first for the cross-sections of the yarns and then the longitudinal element was adapted to the first ones, so that the grid will be absolutely compatible in terms of density. A special care has been taken in order to have corresponding conjugated nodes in the contact areas, on the two bodies surfaces. It supports the convergence of the algorithm and the higher precision of the results (Figure 2).
Finite element modelling of fabrics 35
Boundary conditions and loading The definition of the boundary conditions of the micro-mechanical model of the fabric is of significant importance since it influences the results of the simulation. In a previous work the unit cell of the fabric was fully restricted on one side on the warp direction and on the adjacent side of the weft direction (Tarfaoui and Akesbi, 2001). In similar cases of modelling of woven fabric composites, although the physical conditions are lightly different, the boundary conditions of the unit cell permit the motion only in one direction for two adjacent sides of the model. Each one of the two adjacent sides is free to move in one direction, parallel to the side itself (Choi and Tamma, 2001). It seems that this second aspect about the boundary conditions tends to approach better the real loading conditions. In a similar model of a non-woven fabric, the boundary conditions are defined into two steps. In the first phase, the unit cell is fully restricted in the side opposite to the loading side. In the adjacent side, the unit cell is restricted to move in one direction parallel to that side. In the second phase, both adjacent sides are restricted to move in one direction parallel to each side (Bais-Singh et al., 1998). In the present model, the selection of the boundary conditions has been made mainly based on the symmetry of the model and the physical and computational restrictions. Since the load applied is symmetrical and the structure is also symmetrical over a central point, this central node of the grid remains in the same position, i.e. it is absolutely fixed. This fact permits the consideration of totally constrained central nodes. The stability of the system under analysis requires more boundary conditions. The extreme nodes of the yarns cross-sections (left top and right bottom of the model) are restricted and they are free to move only in one dimension, vertical to the neutral
Figure 2. Finite element mesh and boundary conditions of the fabric model
IJCST 17,1
36
line of the fabric. This particular definition of boundary conditions seems to be more realistic and much more closer to the real conditions of the tensile loading of the fabric. Especially in the case of the first phase of the tensile loading of a plain weave fabric, the friction force generated in the contact area obliges the yarns, vertical to the load application direction into a light rotation before they will overcome the mobility restrictions and to move according to the load applied. The loading of the model is axial. The forces acting were applied at the ending nodes of the longitudinal yarn body. The axial load is the main deformation cause. The value of the maximum load applied is equal to the magnitude of the load applied by the tensile tester of the Kawabata Evaluation System for Fabrics. KES-F applies 500 g/cm across the fabric sample. The respective maximum load applied per yarn end for the specific sample is 0.25 N. In addition to the axial load there is a symmetric load on the yarns cross-sections. The direction of the forces applied is vertical to the neutral plane of the fabric. The forces on the yarn cross-sections are not initially applied on that point; they are the resultants of the forces applied in the cross direction, the third direction vertical to the one of the plane of the 2D model of the fabric unit cell. The initial forces cannot be presented on the 2D unit cell representation, but their resultants are shown as the vertical load to the fabric neutral plane on the top left and bottom right extreme nodes of the yarn cross-sections. The vertical load (FV) is variable from a zero value (corresponding to uniaxial loading) up to the value F of the axial load (corresponding to the symmetric biaxial loading conditions)(Figure 3). Results The analysis of the fabric model was parametric in order to define the influence of the various parameters in the total behaviour of the fabric. A typical value 80.000 N/cm2 of the modulus of elasticity of the yarn has been considered. That value is a result of laboratory measurements on the yarns of the fabric. In addition to that typical value, the parametric analysis took place for values in the range 40,000-120,000 N/cm2. That wide range of the values of the modulus of elasticity permits an in-depth study of the role of the elastic properties as a structural parameter of the fabric. Figure 4 shows the deformation phases while the load increases gradually.
Figure 3. Schematic representation of the unit cell loading
Finite element modelling of fabrics 37
Figure 4. Subsequent deformation phases of the unit cell (percentage of the maximum load)
As has been described earlier, the role of the forces FV, i.e. the forces on the cross-sections of the yarns (resultants of the forces applied in the missing third direction of the model) representing the degree of the biaxial loading, is essential for the characterization of the behaviour of the fabric under various modes of loading (variation range between uniaxial and biaxial loading conditions). Figure 5 shows that, as expected, the elongation of the fabric cell increases under a non-linear law, from the symmetric biaxial loading towards the uniaxial loading of the fabric. The change of the modulus of elasticity introduces a shift of the characteristic curve. The numerical data of the Figure 5 are given in Table I. They indicate that the elongation (per cent) becomes more than double when passing from the biaxial to the uniaxial loading conditions. The modulus of elasticity influences the change of elongation but with essentially less importance. The KES-F system has been used in order to take, the experimental measurements especially on the tensile deformation of the fabric sample, which has been modelled. The experimental results gave the
IJCST 17,1
38
Figure 5. Elongation of the fabric under uniaxial to biaxial loading conditions Fv/F¼0/1
Table I. Variation of the elongation between the uniaxial and biaxial loading (E ¼ modulus of elasticity)
E (N/cm2)
Elongation under uniaxial loading (per cent)
Elongation under biaxial loading (per cent)
D1 (per cent)
40,000 60,000 80,000 100,000 120,000
11.88 10.87 10.32 9.98 9.75
5.95 4.93 4.39 4.07 3.84
99.6 120.5 135.1 145.2 153.9
characteristic non-linear response and the measured EMT (1 %) value is 5.15 per cent. The experimental value of the elongation is typical and it is included in the range of the values of the elongation predicted by the 2D model. Another part of the analysis corresponds to the examination of the influence of the variation of the modulus of elasticity of the yarns on the stress strain curve of the fabric. It is worth to observe that even under the consideration of a linear elastic behaviour of the yarns, the elastic behaviour of the fabric is essentially non-linear. The obvious elastic non-linearity of the fabric is an indication of the influence of the non-linear topology and the influence of the contact mechanism. Figure 6 shows the non-linear character of the stress-strain curve of the fabric derived from the FEM model. More precisely a bundle of curves are presented, one curve for each different value of the modulus of elasticity of the yarns. The process of the numerical data of the parametric analysis results are shown in Table II. The table contains the values of the elongation for a constant stress value. The value of the modulus of elasticity 80,000 N/cm2 has been considered as the basic one serving as a reference value. The table indicates the trend that the variation of the modulus of elasticity of the yarns does not affect in the same rate the variation of the elongation of the fabric and under certain restrictions, the indicative linearized modulus of elasticity of the fabric.
Finite element modelling of fabrics 39
Figure 6. Stress – strain curve of the fabric unit cell
Conclusions The non-linear mechanical characteristics and the complexity of the structure of the fabrics discourage the use of even more precise analytical models for the prediction of their mechanical properties, especially under low stress. The FEM is a powerful tool for the simulation of the mechanical behaviour of the textile fabrics. FEM provides the capability of simulating elastic multi-body systems like the fabric unit cells. Especially it faces the contact mechanism and interactions between the interlaced yarns. Contact itself is a non-linearity factor. The parametric analysis has been used to define the role of the load in the cross-direction, i.e. tensile properties in the range between uniaxial and biaxial loading. In parallel, the influence of the variation of the modulus of elasticity of the yarns has been studied since it is one of the major factors affecting its final mechanical behaviour. For that reason a parametric analysis with five different moduli of elasticity has been performed to investigate the influence of this factor on the tensile behaviour of the fabric. Furthermore, It has been pointed out that the non-linear character of the fabric depending on the boundary conditions can be also due to the complex geometry and the contact mechanism, in parallel to the tensile properties of the yarns. Of course, the two-dimensional approach gives a very good indication of the roles of the various parameters, however a three-dimensional model, despite its computational complexity, would probably provide more realistic simulation and results. The present work states that the use of the FEM is a friendly and convenient method for an in-depth study of the contact phenomena, which are dominating on the total mechanical behaviour of the fabrics. E (N/cm2) 40,000 60,000 80,000 100,000 120,000
Elongation at s ¼ 1,000 N/cm2 (per cent)
DE (per cent)
D1 (per cent)
5.27 4.42 3.95 3.69 3.50
2 50 2 25 0 þ 25 þ 50
þ33.4 þ11.9 0 2 6.6 211.4
Table II. Variation of the elongation (D1) under different modulus of elasticity, E, of the yarns
IJCST 17,1
40
References Argyris, J.H. and Kelsey, S. (1960), Energy Theorems and Structural Analysis, (a collection of papers published in AIAA Vols 25/27, between 1954 and 1956), Butterworths, London. Bais-Singh, S., Biggers, S. and Goswami, B. (1998), “Finite element modelling of the nonuniform deformation of spun-bonded nonwovens”, Textile Research Journal, Vol. 68 No. 5, pp. 327-42. Bathe, K.J. (1982), Finite Element Procedures in Engineering Analysis, Prentice-Hall, Englewood cliffs, NJ. Chan, S.K. and Tuba, I.S. (1971), “A finite element method for contact problems of solid – bodies Part I. Theory and validation”, Int. J. Mech. Sci., Vol. 13, pp. 519-30. Chaudhary, A.B. and Bathe, K.J. (1986), “A solution method for static and dynamic analysis of three-dimensional contact problems with friction”, Computers & Structures, Vol. 24, pp. 855-73. Chen, S.F., Hu, J.L. and Teng, J.G. (2001), “A finite volume method for contact drape simulation of woven fabrics and garments”, Finite Elements in Analysis and Design, Vol. 37, pp. 513-51. Choi, J. and Tamma, K.K. (2001), “Woven fabric composites – part I. Predictions of homogenised elastic properties and micromechanical damage analysis”, International Journal for Numerical Methods in Engineering, Vol. 50, pp. 2285-98. Fredriksson, B. (1976), “Finite element solution of surface nonlinearities in structural mechanics with special emphasis to contact and fracture mechanics problems”, Computers & Structures, Vol. 6, pp. 281-90. Gan, L., Ly, N.G. and Steven, G.P. (1995), “A study of fabric deformation using nonlinear finite elements”, Textile Research Journal, Vol. 65, pp. 660-8. Hu, J. (2000), “3D skirt simulation”, Proceedings of the 3rd International Conference on Innovation and Modelling of Clothing Engineering Processes – IMCEP 2000, Maribor, Slovenia, pp. 66-72. Hu, J.L. and Teng, J.G. (1996), “Computational fabric mechanics: present status and future trends”, Finite Elements in Analysis and Design, Vol. 21, pp. 225-37. Kang, T.G. and Yu, W.R. (1995), “Drape simulation of woven fabric by using the finite element method”, Journal of the Textile Institute, Vol. 86, pp. 635-48. Kemp, A. (1958), “An extension of Peirce’s cloth geometry to the treatment of non-circular threads”, Journal of the Textile Institute, Vol. 49, pp. 44-8. Peirce, F.T. (1937), “The geometry of cloth structure”, Journal of the Textile Institute, Vol. 28, pp. 43-77. Provatidis, Ch.G., Vassiliadis, S.G. and Livaditi, M.A. (2003), “Estimation of the geometrical characteristics of the fabrics for use in numerical mechanical modelling”, Proceedings of the 4th International Conference: Innovation and Modelling of Clothing Engineering Processes – IMCEP 2003, pp. 17-23. Shanahan, W.J. and Hearle, J.W.S. (1978), “An energy method for calculation in fabric mechanics – Part II. examples of applications of the method to woven fabrics”, Journal of the Textile Institute, Vol. 4, pp. 92-100. Tarfaoui, M. and Akesbi, S. (2001), “Numerical study of the mechanical behaviour of textile structures”, International Journal of Clothing Science and Technology, Vol. 13 No. 3/4, pp. 166-75. Vassiliadis, S., Provatidis, C. and Tsigginos, L. (2000), “A computational approach of the micromechanical structure of the fabrics”, Applied Research Review on Technology and Automation, pp. 455-60. Zienkiewicz, O.C. (1977), The Finite Element Method, McGraw-Hill, London.
The Emerald Research Register for this journal is available at www.emeraldinsight.com/researchregister
The current issue and full text archive of this journal is available at www.emeraldinsight.com/0955-6222.htm
A high pressure bursting strength tester for coated industrial fabrics
A high pressure bursting strength tester
C.J. Lewis and D.W. Lloyd United Kingdom Abstract
41 Received April 2004 Accepted November 2004
Purpose – The design and construction of a high-pressure bursting strength tester is described. Design/methodology/approach – The tester has a large aperture and is intended to operate with coated industrial fabrics. To ensure safe operation, since the tester is capable of operating up to pressures of 5,000 kPa, water is used as the inflation medium and all tests are conducted under water. The tester is designed to operate on both flat specimens and cylindrical specimens. Findings – Sample results are reported to demonstrate the successful operation of the tester on plain specimens and on specimens with welded and taped seams. The tester has demonstrated that it is capable of yielding reproducible and useful results on coated industrial fabrics of the type used to manufacture inflatable boats. Practical implications – The low-cost, high pressure bursting strength tester can be used for lightweight, coated industrial fabrics of the type commonly used in the construction of inflatable boats and liferafts. Originality/value – The tester was developed to support a programme of work to develop a lightweight, collapsible decompression chamber. Although the parameters of the tester were chosen to meet the needs of the decompression chamber project, the tester has wider application and may, therefore, be of wider interest. Keywords Pressure, Fabric testing Paper type Technical paper
1. Introduction The purpose of this paper is to describe the development of a low-cost, high pressure bursting strength tester for lightweight, coated industrial fabrics of the type commonly used in the construction of inflatable boats and liferafts. The tester was developed to support a programme of work to develop a lightweight, collapsible decompression chamber (Lewis, 1988). Although the parameters of the tester were chosen to meet the needs of the decompression chamber project, the tester has wider application and may, therefore, be of wider interest. The design allows both flat sheets and cylindrical bags, closed at one end, to be tested at pressures up to 5,000 kPa. Such high pressures demand that, great care is taken over safety; as a consequence, the tester has no separate diaphragm and operates submerged in water with water as the pressure medium. The authors gratefully acknowledge the support provided by the Clothworkers’ Textile Structures and Mechanics Laboratory in the School of Textile Industries, University of Leeds, and the generous supply of fabrics by Carrington Performance Fabrics and Greengate Polymer Coatings. They also acknowledge the assistance of the Worshipful Company of Clothworkers, the Textile Institute, and the Worshipful Company of Weavers, as well as information kindly provided by Dr D.G. Neilly of the School of Textile Industries (currently at Thomas Ferguson & Co.), and Mr F. Roffee of Humber Inflatables.
International Journal of Clothing Science and Technology Vol. 17 No. 1, 2005 pp. 41-51 q Emerald Group Publishing Limited 0955-6222 DOI 10.1108/09556220510577952
IJCST 17,1
42
Bursting strength and related test methods have a long history (Anon, 1911; Haas and Dietzius, 1917) in connection with engineering application of fabrics that involve biaxial stresses. A full history of the development of test methods for the mechanical properties of industrial textiles is given in Neilly (1986), and of bursting strength test methods in Lewis (1988). “Hemisperical” bursting tests form the basis of a number of international standards, both for knitted fabrics and coated fabrics (BS 3424, 1982; ISO3303, 1979; BS 4F100, 1982; ASTM D2210-64, 1982). Usually, bursting tests make use of small apertures in the clamping plate, typically of the order of 30 mm in diameter. Occasionally, larger apertures are reported, as in the tests reported by Nakahara (1972) and Jagfeld (1970). “Cylindrical” bursting tests also date back to early textiles in engineering applications (Sueter, 1910; Anon, 1910; Burgess, 1930). Such tests also find more modern application (Minster and Berka, 1977; Paute and Segoin, 1977; Van Leeuwen, 1977). 2. Design criteria The present tester was designed to allow both hemispherical and cylindrical types of test to be undertaken. The hemispherical tests were used to assess the suitability of individual fabrics for the proposed application, whilst the cylindrical test was used to assess the relative merits of different detailed designs. The dimensions of the tester were chosen to allow one-third scale models of the proposed design to be tested. This in turn required a test pressure three times the test pressure that would have been required on a full-size prototype, setting a minimum test pressure of 4,150 kPa. The pump used in the present tester easily exceeds this requirement. If it is assumed that the specimen adopts a shape under pressure that is part of the surface of a sphere, it is relatively simple to calculate the extension of the fabric from knowledge of the fabric deflection, i.e. from the “height” of the domed fabric (Lewis, 1988). Consequently, the design of the tester must make provision for measurement of the fabric deflection. The other parameter that requires measurement is the inflation pressure. This is useful in the case of hemispherical tests, but vitally important in the case of cylindrical tests, as the inflation pressure determines the axial and hoop stresses in the fabric. If air or other pressurised gas were to be used as the inflation medium, the release of stored strain energy at the point of fabric failure would constitute a severe hazard to the experimenter and the equipment. The only safe inflation medium is an incompressible fluid that stores little strain energy; of the possible alternatives, water is the cheapest and most convenient. The design of the apparatus, therefore, must allow for both hemispherical and cylindrical specimens, for measurement of the axial deflection of the specimen, for inflation to high pressures with water and for measurement of the inflation pressure. Even with water, safety dictates that the test be carried out behind a suitable safety barrier; again water provides the most convenient material. Thus, the tester must be capable of being set-up in dry conditions on the bench, then charged with water and submerged under a suitable depth of water. It is worth noting that this arrangement precludes determining strain fields in the specimens by the non-contact method described elsewhere (Lloyd et al., 2000). The basic arrangement of the tester body is shown in Figures 1-3. The body was constructed of steel plate 15 mm thick. A short length of tube was closed at one end with a turned disc welded in place. This plate carried a central inlet port with a welded
A high pressure bursting strength tester 43 Figure 1. Bursting strength tester – side elevation
boss that was threaded to match the outlet pipe from the pump used to pressurise the system. A further hole was drilled in the top of the plate and fitted with a valve. This was used to release trapped air and to ensure that the system was fully charged with water. A flange was welded to the other end of the tube. This carried eight bolts used to secure the fabric specimen under a clamping ring. Details of the clamping ring are shown in Figure 4. The ring was made in two parts; used together, they constituted the clamping ring for flat specimens. Separated, they were used to clamp cylindrical specimens. The body of the tester was mounted on a stand that incorporated means of clamping a submersible, long-stroke, linear voltage differential transducer (LVDT). These supports could be moved to allow both types of specimen to be tested. The LVDT was fitted with a low-force return spring that ensured that the transducer probe was always in contact with the specimen. The probe was fitted with a lightweight aluminium disc, to ensure the force of the contact was spread over sufficient area to avoid distorting the test results. The force provided by the spring was negligible in comparison with the test pressures employed, but could have introduced a local distortion had the original small-diameter end been retained. The particular submersible LVDT used was an RDP type D2/2000AW with transducer indicator E307-2. For reasons of economy and ease of control, a hand pump of the type used to pressure-test piping systems was used to pressurise the tester. The particular model used was a Rothenberger model RP 50-60. This is normally fitted only with a dial pressure gauge. To enable a continuous record to be taken of inflation pressure, the pump was fitted with a “Sensotec” model A5/761-29 pressure transducer, with transducer indicator E307-2, both manufactured by RDP Ltd. The assembled tester can be seen set-up on the bench in Plate 1. Finally, the whole system was submerged in a plastic water tank with approximate dimensions 1.75 (length) £ 0.7 (width) £ 1.0 m (depth). The particular tank used was a domestic cold water tank of a pattern widely available in the UK (Plate 2). 3. Operating the tester Initially, fabrics were clamped directly between the two steel faces. This proved to be unsatisfactory, as the fabric was not clamped sufficiently or slight distortion was
IJCST 17,1
44
Figure 2. Bursting strength tester – top elevation
A high pressure bursting strength tester 45
Figure 3. Bursting strength tester – end elevation
Figure 4. Detail of clamping rings
IJCST 17,1
46
Plate 1. Bursting strength tester assembled on the bench, showing the hand pump, LVDT, pressure sensor and electronics
Plate 2. Bursting strength tester set-up in tank; the purpose of the white plastic sheet is to provide contrast
caused around the bolts by tightening them to too high a torque. After the clamping surfaces had been re faced, gaskets of different materials were used to assist the clamping process. The most effective gasket material found in the series of fabrics studied proved to be the material itself. In the results reported below, all the tests were conducted with gaskets made from the test material itself.
A second problem that was experienced was that of premature failure at one of the boltholes cut in the fabric. Such failures always occurred at a hole that cut directly across either the warp or the weft yarns. These failures were avoided by the simple expedient of ensuring that specimens were cut so that the warp and weft directions passed mid-way between boltholes. Some failures were initiated by the inner edge of the clamping ring; these failures were eliminated by turning a small radius on the inside edge of the clamping ring. This radius has the disadvantage of introducing a systematic error in the calculation of the length of the stretched fabric and hence of the extension. However, the error is small compared to the diameter of the aperture. One of the objectives of the test programme was to determine the relative strength of welded and taped seams. Seamed fabrics present a difficult clamping problem because at two points on their circumference, the fabric thickness is doubled. This difficulty was overcome by substituting 2 mm thick neoprene as the gasket material. The pump is hand-operated and is fitted with a valve to prevent water from the high-pressure side leaking back to the low-pressure side during the return stroke of the pump. It was found, however, that at higher pressures, the valve was not completely effective and that minor leakage could occur. BS 4F100 requires pumping to take place at a rate that reaches the minimum bursting pressure in 60 ^ 10 s; but in view of the minor leakage problem, it was found that a more rapid application of pressure was needed. For all the tests, the minimum bursting pressure reached in 30 ^ 10 s: Five replicates were used for each test and the arithmetic means were calculated for the bursting pressure and the deflection (dome height). The time to failure was also recorded. The results for a series of tests on fabrics supplied by Carrington Performance Fabrics and Greengate Polymer Coatings are given in Table I for plain specimens and in Table II for specimens with a seam. Details of the fabrics are given in Table III. Each fabric exhibited a characteristic type of failure, depending on its properties. Plates 3 and 4 show examples of failed specimens.
Mean burst pressure Standard Fabric No. of (kPa) deviation quality replicates TD422 3752 3762 3500
5 5 5 5
1,793 1,676 1,676 1,800
5 5 5 5
1,958 1,613 1,724 1,593
Mean percentage linear extension
Mean percentage area extension
Mean time to break
51 50 49 46
22 21 21 18
35 34 33 29
30 25 24 23
15 10 11 26
Mean burst pressure Standard Fabric No. of (kPa) deviation quality replicates TD422 3752 3762 3500
Mean dome height (mm)
5.8 8.6 7.3 26
Mean dome height (mm)
Mean percentage linear extension
Mean percentage area extension
Mean percentage original strength
Mean time to break
48 45 42 37
20 18 15 12
35 34 33 29
109 96 103 88
30 25 24 23
A high pressure bursting strength tester 47
Table I. Results of burst tests on plain coated fabric specimens
Table II. Results of burst tests on coated fabric specimens containing 20 mm overlap seams
Table III. Details of fabrics
Base Fabric Total Weight Ultimate tensile strength (BS 3424/6) Tear strength (BS 3424/7A)
General description Colour
100 per cent nylon, coated both sides with TPU Orange 2 fold 940 dtex warp & weft 140 £ 75 e & p/m Oxford weave 1.32-1.50 kg/m Warp: 78.5 kN/m Weft: 68.7 kN/m
100 per cent nylon, coated both sides with TPU Orange 940 dtex warp 940 dtex weft 175 £ 175 e & p/m 2 £ 2 matt weave 0.92-1.05 kg/m Warp: 78.5 kN/m Weft: 68.7 kN/m Warp: 294 N Weft: 294 N
Warp: 425 N Weft: 385 N
Warp: 343 N Weft: 294 N
Greengate Polymer Coatings –
3762
Greengate Polymer Coatings –
3752
TD422 Carrington Performance Fabrics Inflatable oil booms 100 per cent high tenacity nylon 6.6, coated both sides with Polyether TPU Orange 950 dtex warp 940 dtex weft 70 £ 70 e & p/m 2 £ 2 matt weave 0.902 kg/m Warp: 102.0 kN/m Weft: 93.1 kN/m
Warp: 809 N Weft: 760 N
100 per cent polyester, coated both sides with neoprene Grey face Olive back 2 fold 1100 dtex warp & weft 90 £ 90 e & p/m 2 £ 2 matt weave 1.860 kg/m Warp: 98.1 kN/m Weft: 98.1 kN/m
Greengate Polymer Coatings Inflatable boats
3500
48
Manufacturer Specified end-use
Fabric
IJCST 17,1
A high pressure bursting strength tester 49
Plate 3. Failed plain specimen, showing primary extension failure in the warp direction with secondary extension failure in the weft direction
Plate 4. Failed specimen with lapped seam, showing failure in the yarns perpendicular to the seam
The bursting strength tester was designed to allow tests to be carried out on cylindrical specimens. This feature was used to study different designs of flexible “bag”, to determine the most effective design for the body of the “hyperbaric stretcher” that formed the focus of the project. In particular, the study focused on different possible
IJCST 17,1
50
Plate 5. Bursting strength test on cylindrical specimen in progress
designs for the closed end of the bag. Plate 5 shows one of these tests in progress. It is worth noting that the design of the tester and the size of the water tank enabled the LVDT to be used to measure the stretching of the bag in the direction of its longitudinal axis. 4. Summary The design and construction of a low-cost, high pressure bursting strength tester has been proposed. The tester is of simple construction and is designed to allow tests to be conducted on both flat specimens (hemispherical tests) and cylindrical specimens. The tester has a large aperture of about 170 mm. It has provision for measuring the deflection of flat specimens or the longitudinal deformation of cylindrical specimens, as well as means of recording inflation pressure. Because of high pressures, up to 5,000 kPa, the tester employs water as its inflation medium and all tests are conducted with the tester submerged in a tank of water. The tester has demonstrated that it is capable of yielding reproducible and useful results on coated industrial fabrics of the type used to manufacture inflatable boats. References Anon (1910), “Report from the National Physical Laboratory on tests of balloon fabrics”, Advisory Committee for Aeronautics, Reports and Memoranda, No. 27, October. Anon (1911), “Experiments on the bursting strength of cylinders made of fabric when the ratio of the hoop tension to the longitudinal tension is varied”, Department of Science and Industrial Research File 23/T107.
ASTM D2210-64 (1982), “Grain crack and extension of leather by the Mullen test”, American Society for Testing and Materials (reapproved 1982). British Standard BS3424 (1982), “Testing coated fabrics”, Part 6, Methods 8A and 8B – Methods for the Determination of Bursting Strength. British Standard 4F100 (1982), “Inspection and testing of textiles”, British Standards Institution Aerospace Series. Burgess, C.P. (1930), “Bursting tests of cylinders of balloon fabrics”, US Navy Department, Bureau of Aeronautics Technical Note No. 10. Haas, R. and Dietzius, H. (1917), “The stretching of the fabric and the shape of the envelope in non-rigid balloons”, National Advisory Council on Aeronautics, Report Number 16. ISO3303 (1979), “Rubber- or plastics-coated fabrics – determination of bursting strength”, 1st ed., International Organisation for Standardisation. Jagfeld, P. (1970), “Ein-und mehrachsige Zugversuche an der Dachfolie des Deutschen Pavillions auf der Weltaus-stellung 1967 in Montreal”, Melliand Textilberichte, Vol. 51, pp. 349-54. Lewis, C.J. (1988), “A study of hyperbaric applications of industrial textiles”, PhD thesis, University of Leeds, Leeds. Lloyd, D.W., Price, D. and Brook, D.B. (2000), “Strain measurement in fabrics. Part III: a non-contact method of determining finite strains – practical implementation”, Research Journal of Textile and Apparel. Minster, J. and Berka, L. (1977), “Strength of technical fabrics under biaxial state of stress”, Stevenbricky Casopis, Vol. 25 No. 1, pp. 37-48. Nakahara, Y. (1972), “The mechanical behaviour of circular membrane fabric under uniform lateral pressure”, paper presented at the IASS Symposium Part 5.9, Vol. 2, pp. 1-16. Neilly, D.G. (1986), “The development of methods for the study of properties and performance in fabric for industrial and engineering end-uses”, PhD thesis, University of Leeds, Leeds. Paute, J.L. and Segoin, M. (1977), “Determination of strength and deformability characteristics of fabrics by dilation of a cylindrical sleeve”, Proceedings of the International Conference on the Use of Fabrics in Geotechnics, Vol. 2, pp. 293-8. Sueter, M.F. (1910), “Report on the details of construction of naval airship no. 1”, Advisory Committee for Aeronautics, Reports and Memoranda, No. 26, September. Van Leeuwen, J.H. (1977), “New methods of determining the stress-strain behaviour of woven and non-woven fabrics in the laboratory and in practice”, Proceedings of the International Conference on the Use of Fabrics in Geotechnics, Vol. 2, pp. 299-304.
A high pressure bursting strength tester 51
The Emerald Research Register for this journal is available at www.emeraldinsight.com/researchregister
IJCST 17,1
The current issue and full text archive of this journal is available at www.emeraldinsight.com/0955-6222.htm
COMMUNICATIONS
Clothing fit preferences of young female adult consumers
52
Marina Alexander East Carolina University, Greenville, North Carolina, USA
Lenda Jo Connell and Ann Beth Presley Auburn University, Auburn, Alabama, USA Abstract Purpose – This paper explores the relationships between body type and fit preferences with body cathexis, clothing benefits sought by consumers, and demographic profiles of consumers. Design/methodology/approach – The survey instrument consisted of a questionnaire with scales assessing fit preference, body type, body cathexis, clothing benefits sought and consumer demographics. Findings – Significant associations were found between body cathexis (satisfaction with head/upper body, lower body, height, weight and torso) and body shape. The degree of satisfaction with different body parts depended on the body type of the individual. The level of satisfaction with head/upper body, height and torso did not vary by body type. No significant differences were found between fit preferences and body type for lower body garments. Research limitations/implications – The majority of respondents were between the ages 18 and 28, affluent Caucasian Americans, with an hourglass body type, who had a family income of $85,000 or more and shopped in department or boutique/specialty stores. Originality/value – Understanding the fit preferences of female consumers could help apparel companies to produce and meet demands for comfortable and well fitting clothes for women. The results of this research may be used as a first step to develop an expert system to correlate body shape and fit preferences of consumers. Keywords Clothing, Human physiology, Customer satisfaction, Women Paper type Research paper
International Journal of Clothing Science and Technology Vol. 17 No. 1, 2005 pp. 52-64 q Emerald Group Publishing Limited 0955-6222 DOI 10.1108/09556220510577961
Introduction In ancient civilizations, clothing was primarily used to cover and protect human bodies. Today, most consumers see clothing as more than just a basic necessity. Research conducted in clothing behavior has shown that consumers differ in attitudes, values and expectations of clothing. People use clothing to identify themselves with a social class, project a positive image and as a means to improve their overall appearance. According to Williams (1974), “the reasons for wearing clothes are numerous, complex and interrelated. The attitudes consumers hold toward clothing influence their choices of clothing even though they may be unaware of the specific reasons for certain clothing choices. These attitudes may also affect the closeness or looseness of fit in clothing”. Fit in a garment is one important factor that contributes to the confidence and comfort of the wearer. Well-fitted clothes are considered vital to an individual’s psychological and social well being (Smathers and Horridge, 1978-79). A garment that
looks good on the wearer is essentially one that fits the wearer well. Dissatisfaction with fit is one of the most frequently stated problems with garment purchases. Studies by Kurt Salmon Associates (2000) reported that more than half of the female population in the US cannot find apparel in the marketplace to fit. Women responding to their surveys indicated that fit was the third most frequent reason for not making an apparel purchase. In other studies, women have reported trying on as many as 20 pairs of jeans before they find a pair that fits (Consumer Reports, 1996). To fit consumers, each manufacturer must successfully interpret body measurements and produce apparel that satisfies their customer’s fit preferences. Fit preference is very subjective and varies from person to person. In addition to body measurements used by manufacturers, patterns have both fit ease and style ease. According to Hudson (1980), fit ease accommodates body functions by allowing body movements, and it prevents garments from binding. Style ease adds fullness garments to create visual effect. The fit preference of two consumers having the same body measurements, height, and weight can be very different and may vary greatly. Observation indicates that people differ in their preference for clothing fit, from garments that cling tightly to the body to garments that barely touch the body. The snugness or ease a person desires in clothing depends on one’s personal preferences, attitudes, or the look consumer’s desire. Body scanning technology has advanced to the stage of being useful in collecting body measurements to be used to produce patterns. However, beyond a notion of body measurements, manufacturers must also understand fit from the consumer’s perspective. Williams (1998) noted that if a company could deliver a product with the right fit, they could be certain of keeping the customer for a long time. A search of the literature revealed no research on fit preference from a consumer perspective. This study is an exploratory study designed to understand fit preferences of college women. Purpose of the study The specific objective of the research was to explore the relationships between body type and fit preferences with body cathexis, clothing benefits sought by consumers, and demographic profiles of consumers. Review of literature Theoretical framework and research model Ashdown’s (2000) model for development of a sizing system (Figure 1) was used as a theoretical framework for this study. The model is complex and this study focuses on one part of Section B which involves understanding fit perceptions from the consumer’s standpoint. Developing measures to understand fit preference from an individual viewpoint could help manufacturers, retailers and standards setting bodies better define sizing systems. Fit issues A search of the literature revealed few studies that look at fit from a consumer perspective. Ashdown and DeLong (1995) noted two issues which consumers’ address relative to a perception of fit. One issue is a personal judgment relative to how the garment looks on the body. The other is the perception of the comfort level of the garment based on both tactile and visual responses from the consumer. Other studies
Clothing fit preferences
53
IJCST 17,1
54
Figure 1. Ashdown’s (2000) model for development of a sizing system
have used demographics, body-cathexis (feelings about one’s body), benefits sought from clothing (e.g. sex appeal, fashion innovativeness, and figure flaw compensation), and body type of the respondent to investigate consumer’s satisfaction with fit. Body cathexis. Body cathexis is defined as the “degree of satisfaction or dissatisfaction with the various parts or processes of the body” (Secord and Jourard, 1953, p. 343). Body Cathexis is a concept closely related to body image and is viewed as part of self-concept. Self-feelings about the body play a major role in clothing preferences and attitudes (Kaiser, 1990). According to Shim et al. (1991), clothing is an extension of the bodily self and has important symbolic meanings in social interactions. Many studies have shown that women, in particular, are dissatisfied with their bodies. According to LaBat and DeLong (1990), one factor that may contribute to women’s dissatisfaction with their bodies is that fashionable clothing reflects a standard sizing, which is not realistic. When a garment does not fit, consumers often blame their own body instead of the garment, which in turn causes a negative body image.
LaBat and DeLong (1990) found that female participants were dissatisfied with the lower body (pant length, crotch, thigh, hip and buttocks) fit of ready-to-wear and were relatively satisfied with the fit of apparel in the upper body. The study concluded that dissatisfaction with the lower part of the body is due to the fact that women are becoming broader hipped and are having more difficulty wearing garments designed for a smaller hipped, hourglass-shaped woman. The authors further suggested that additional studies are needed to confirm the relationship of body cathexis and fit satisfaction with clothing. Hwang (1996) used clothing benefits, clothing attitudes and behavior to study body cathexis and the importance of meeting the ideal body image. Hwang (1996) defined body cathexis as a function which shapes preferences for clothing attitudes and behaviors related to clothing such as fashion innovativeness, satisfaction with ready-to-wear, clothing preferences, and shopping behavior. Results indicated that clothing attitude and the importance of meeting an ideal body image were important moderators for the relationship between body cathexis and clothing benefits sought. LaBat and Delong (1990) in their study on body cathexis and satisfaction with fit of apparel, found that the participants showed the least satisfaction with lower body dimensions. In the LaBat and Delong (1990) study, a correlation was found between body cathexis and the stated body parts: bust, hip, thigh, buttocks, abdomen, arm, back and shoulder. This study indicated a relationship exists between body cathexis and level of satisfaction with fit. Clothing benefits sought. Researchers have proven the construct of symbolic meaning of clothing and its use in social environments (Hwang, 1996; Horn, 1975). Clothing used positively contributes to one’s feelings of self-confidence. Clothing may be viewed as an extension of the physical self and as an integral part of body image (Horn and Gurel, 1981). A study by Davis (1985) found that feelings and self-perceptions of one’s body-build defined by somatographs related to fashion and the use of clothing. The study indicated that female students, regardless of body type, were interested in fashion and clothing. This is contrary to the stereotypic assumptions that people who differed from an ideal body type would be less interested in clothing and fashion. In a study conducted by Swan and Combs (1976) on consumer satisfaction with clothing purchases, two dimensions of satisfaction were identified: (1) instrumental outcomes relating to the physical aspects of product such as durability, construction and fit; and (2) expressive outcomes relating to styling responses of other people and comfort. The results of their study showed instrumental outcomes were more linked to dissatisfaction than expressive outcomes and concluded that durability, construction and fit aspects of clothing must be demonstrated for clothing satisfaction to occur. Consumers seek a variety of benefits from clothing. It would be beneficial to apparel companies and researchers to identify the benefit traits and characteristics consumers want from clothing. Information from a study done by Shim and Bickle (1994), indicated that self-improvement, sex appeal, social status/prestige, figure flaw compensation and fashion image had the highest mean scores on the nine clothing benefits factors sought.
Clothing fit preferences
55
IJCST 17,1
56
Communication of sizing and fit The size of the garments is communicated to the consumers through the size labels on the garment. Consumers select garments according to the size labels. If consumers are satisfied with the purchase, they will keep the garment. If consumers are dissatisfied with the garment, they may return it. Only a fraction of the population has what is culturally considered perfectly proportional bodies. Ready-to-wear clothes are made for consumers with proportional bodies. The majority of people’s body proportions deviate from standard sizes. Voluntary size standards were established in the 1940s by the Federal Trade Commission and the Department of Commerce. According to Tamburrino (1992), after the initial introduction, this system has proved to be a failure for women’s wear. Today, in women’s wear, most apparel companies ignore this system and follow their own sizing standards. In designing women’s wear, one of the primary reasons apparel companies do not want to follow a standard sizing system is because different firms have different target populations of women whose lifestyles, incomes and body shapes differ. Another reason for varying sizing standards is that there is a psychological need to feel slim among consumers in the Western culture. This need has motivated many manufacturers to mark down sizes. Measurements for a size 10, a decade ago might be the measurements used to produce a size 8 today (Workman and Lentz, 2001). Another reason companies prefer not to use the voluntary sizing standards is that sizing is seen as a way to create and maintain brand identities. Sizing has become a selling tool distinguishing one apparel manufacturer from another and building customer loyalty (Workman, 1991). “Vanity sizing” or offering garments with smaller size numbers but larger measurements allows firms to influence the psychological need of consumer’s to fit into a smaller size. Consumers believe that they are wearing, what they consider, a more socially acceptable size. The purpose of sizing has been to fit the majority of the population whose body proportions fall within predetermined standard dimensions (Workman, 1991). Tamburrino (1992) pointed out that designations of women’s apparel sizes with numbers which have no direct relationship with any part of the female body has lead to the present confusion among apparel companies and even more so among consumers. According to Brown (1992, p. 261) “ Personal preferences of fit are shaped by current fashion trends and cultural influences, age, sex, figure type and life style”. Hazen (1998, p. 6) stated that “Fit is an individual preference and there is no way the pattern companies are going to meet the needs of every person”. Hazen (1998) pointed out that clothing manufacturers traditionally use a standard size 12 Wolf dress form as their fit model. A fit model is a stationary three-dimensional form that represents the figure type of the target customer. Very few consumers are built like the perfect body form, and this helps explain the difficulty experienced by consumers in finding well-fitted clothing. Brown (1992, p. 32) discussed visiting a designer’s studio and seeing a line of Wolf dress forms ranging from size 6 to 22. She said, “The thing that fascinates me about these dress forms is that the size 22 didn’t have a tummy. I still haven’t figured out how someone can be a size 22 and not have some kind of tummy – unless she is 7 feet tall”. Apparel companies must incorporate the body changes that take place with weight gain and age.
Method Sampling and data collection This exploratory study focused on understanding the fit preference of young female consumers between the ages of 18 and 29 enrolled in classes in a Southeastern university in the US. A questionnaire was distributed to 232 participants and the purpose of the survey was briefly explained. Respondents took approximately 20 min to complete the self-administered questionnaire. A total of 223 useable questionnaires were returned.
Clothing fit preferences
57
Survey instrument The instrument consisted of a questionnaire with scales assessing fit preference, body type, body cathexis, clothing benefits sought and consumer demographics. Parts of the instrument were adapted and modified from previous research. Fit preferences. The construct of fit preference was defined by line drawings representing six separate garment categories including jackets, skirts, dresses, tops, jeans and pants. Three line drawings in each category represented fit for that category as fitted, semi-fitted, or loosely fitted. Respondents were instructed to select the garments they would buy from these illustrations (Figure 2). Body cathexis. A body cathexis scale initially developed by Secord and Jourard (1953) was used to assess respondent’s satisfaction/dissatisfaction with their bodies. This scale has been used extensively in research on feelings about the self (Hwang, 1996; Shim et al., 1991; Mahoney and Finch, 1976). It consists of 19 body areas with five sub-scales measuring feelings of satisfaction/dissatisfaction with specific body areas including (1) lower body; (2) head/upper body; (3) weight; (4) height; and (5) torso. Clothing benefits sought. One section of the questionnaire included a scale measuring clothing benefits sought segment. The scale was adapted from Shim and Bickle (1994) and Hwang (1996). It consists of 25 statements used to measure the benefits consumers seek in using clothing. The scale consists of six sub-scales measuring
Figure 2. Presentation of career separates
IJCST 17,1
58
(1) (2) (3) (4) (5) (6)
fashion image; figure flaw compensation; sex appeal; clothing preference; fashion innovativeness; and satisfaction with RTW.
Body shapes. Body shape was self-reported by the respondents. They selected from illustrations with descriptions defining rectangular, pear, inverted triangle and hourglass shapes. A body shape was considered as rectangular if the bust-waist, waist-hip measurements were straight up and down with very little waist definition. Respondents classified their bodies as pear if hips were broader than shoulders. An inverted triangular body shape was illustrated with broad shoulders and smaller hips. An hourglass figure was characterized by a full bust and hips with an obvious waist definition. Statistical procedures One way ANOVA was performed on body shape and (1) body cathexis; (2) clothing benefits sought; and (3) fit problems. Pearson’ s correlation was done on fit preference and (1) body cathexis; (2) clothing benefits sought; and (3) fit problems in RTW. Chi squared tests were preformed on fit problems in RTW and body shape to investigate if fit problems were dependent on the body shape. Results Respondents characteristics Results of the survey revealed the majority of respondents were between the ages 18 and 28, affluent Caucasian Americans, with an hourglass body type, who had a family income of $85,000 or more and shopped in department or boutique/specialty stores. The Gap, Esprit, American Eagle, Old Navy, Express and J. Crew were the most popular brands purchased by the respondents. Of those that reported weight, the average was 130.81 pounds with a height was 64.01 in. Bust measurements for the group averaged 35.08 in. The average waist measurement was 27.09 in. and hip measurement was 36.09 in. Slightly over 66 percent of the respondents reported exercising regularly. Fit problems Almost 64 percent of the respondents had to alter RTW apparel, on a regular basis, to achieve the desired fit. As a result, 54 percent of the respondents reported being somewhat satisfied to mostly unsatisfied with the fit of RTW. These figures show that
substantial fit problems exist in the RTW market and are consistent with the Goldsberry et al. (1996) study. In the current study, the percentage of respondents who reported to have fit problems at various body parts was: bust (50 percent), waist (46 percent), hip (46 percent), dress length (46.5 percent), pant length (60.5 percent), thigh (30 percent), sleeve length (31.4 percent) and crotch (26.5 percent). The pear and hourglass body types were more likely than the rectangular and inverted body types to have problems with the fit at the waist, hips and thighs. Fit problems and body shape. Rectangular, pear and hourglass body shapes were more likely to report fit problems at the bust than the inverted triangular body type. Fit problems at the waist, hip, thigh, dress length, and pant length were more likely to be reported by the pear and hourglass body types than the rectangular and inverted triangle body types. The hourglass and the rectangular body types reported problems with the sleeve length. The inverted triangular body types were the least likely to report fit problems at the crotch. Overall, the inverted triangle body types were satisfied with the fit of RTW. Pearson’s correlation showed that respondents who reported fit problems at the bust in RTW did not prefer a fitted top. Similarly, respondents who had fit problems at the hip and armhole did not prefer a fitted jacket and dress. Consumers reporting fit problems in dress length preferred a fitted dress and respondents who reported fit problems in the crotch preferred fitted jeans. Respondents who had fit problems at the neck preferred a loose dress and respondents reporting fit problems at the waist preferred a more fitted pant. Body cathexis A significant association was found between body shape and body cathexis. Subjects with an inverted triangle shape were the more satisfied with weight than the other body shapes. The inverted triangle shape also reported greater satisfaction with their lower body than did subjects with a pear or hourglass shape. No significant associations were found among the four body shapes and their reported satisfaction with (1) head and upper body; (2) height; and (3) torso (Table I). Among the 19 body parts which made up the body cathexis scale, satisfaction of body parts that were significant ðp , 0:05Þ to the respondents were face, hips, knees, legs, waist, bust and thighs. Body parts such as the teeth, hair, nose, voice, height, eyes, feet and muscular strength that did not contribute body shape were not found to be significant. Pearson’s correlation for body cathexis and fit preference revealed that respondents satisfied with their weight preferred a fitted top and dress. Clothing benefits sought One-way ANOVA performed on the clothing benefits factors and body shape revealed that fashion innovativeness and satisfaction with RTW were significant ðp , 0:05Þ: Post hoc tests showed that the inverted triangular body types were different from all other body types for the satisfaction with RTW factor. However, post hoc tests showed that none of the body types differed in the fashion innovativeness factor. Pearson’s
Clothing fit preferences
59
IJCST 17,1
60
Table I. One-way ANOVA body cathexis * body shape
BC subscales Lower body Head/upper body Height Weight Torso Body parts Neck Calves Teeth Face Hair Nose Hips Knee Legs Voice Overall height Leg length Overall weight Feet Waist Bust Eyes Muscular strength Thigh
B
O
P
I
p-level
20.71ab 31.80 10.48 17.11ab 14.60
19.86a 34.07 10.25 15.46a 15.51
22.47b 34.41 11.05 20.82b 16.35
19.21a 32.71 10.48 16.50a 15.34
0.012 0.134 0.797 0.011 0.157
5.80 5.31 5.33 4.87a 5.27 5.13 4.49b 5.36ab 4.96ab 5.24 5.29 5.20 4.11 5.09 4.24a 4.04a 5.96 4.76 4.27a
6.08 5.41 5.74 5.54b 5.72 5.62 3.54a 5.08a 4.45a 5.36 5.33 4.92 3.97 5.08 4.44a 4.77ab 6.03 4.67 3.51a
5.76 5.53 5.47 5.53b 5.41 5.71 5.65c 5.82b 5.35b 6.00 5.59 5.47 4.76 5.76 5.35b 5.29b 6.29 5.29 5.06b
5.95 5.05 5.33 5.38ab 5.33 5.46 4.09ab 4.75a 4.19a 5.21 5.33 5.16 4.12 5.20 4.77ab 4.73ab 6.01 4.67 3.53a
0.586 0.313 0.418 0.041 0.351 0.203 0.000 0.008 0.009 0.160 0.898 0.653 0.501 0.439 0.073 0.032 0.693 0.497 0.001
Note: Means not sharing a superscript are different at p , 0.05 using Duncan post hoc test
correlation done on the clothing benefits scale and fit preference revealed that respondent’s who preferred to emphasis different parts of the body preferred a fitted jacket, dress and skirt. Respondents who preferred a fitted top used clothing to portray a fashionable image, for sex appeal, were fashion innovators, and liked to emphasis different parts of the body through fit. Respondents who did not prefer emphasizing body parts through fit and did not use clothing for sex appeal preferred a loose fitting dress. Discussions and conclusion Significant associations were found between body cathexis (satisfaction with head/upper body, lower body, height, weight and torso) and body shape. Subjects with an inverted triangle body type were more satisfied with their weight and lower body than all other body types. Correlation of body types and satisfaction with the body parts (which comprised the body cathexis scale) were statistically significant ðp , 0:05Þ; confirming that the degree of satisfaction with different body parts depended on the body type of the individual. The level of satisfaction with head/upper body, height and torso did not vary by body type. Pearson’s correlation coefficient was performed on the clothing benefits sought scale (e.g. fashion image, sex appeal, figure flaw compensation, clothing preferences
and fashion innovativeness). Respondents to whom the clothing benefits factors such as fashion innovativeness and satisfaction with RTW were important, preferred a fitted upper body garment (tops and jackets). Respondents who were dissatisfied with their weight preferred a loose fit in dresses. Fashion at the time of data collection for a loose fit in jeans and pants may have been reflected in the fit preferences of all body types. Hence, no significant differences were found between fit preferences and body type for lower body garments. ANOVA test run for the clothing benefits sought scale indicated that fashion innovativeness and satisfaction with RTW was significant ðp , 0:05Þ: Based on the fit preferences mentioned above, it can be concluded that respondents who were satisfied with certain body parts (e.g. thighs, bust, hips, waist, etc.) preferred a closer fit at that area. Post hoc tests revealed that rectangular ðn ¼ 45Þ and inverted triangular body types ðn ¼ 17Þ were similar in their fit preferences and were different from the hourglass ðn ¼ 114Þ and pear body types ðn ¼ 39Þ: The overall goal of the project was to determine the connection between body type and fit preferences. With similar studies across all consumer segments, the apparel industry would be able to better predict fit preferences of female consumers. This initial research would help future researchers develop an expert system to correlate fit preferences of female consumers based on their demographics, body cathexis, clothing benefits sought and body shape. Implications, limitations and recommendation Implications Understanding the fit preferences of female consumers could help apparel companies to produce and meet demands for comfortable and well fitting clothes for women. The fit problems present in the current ready-to-wear market signal to the industry that significant fit problems exist in the sizes available and substantial steps need to be taken to update the outmoded sizing system. Problems encountered with the survey instrument can be corrected and more research could be conducted to provide further insight into the fit preferences of consumers. The results of this research may be used as a first step to develop an expert system to correlate body shape and fit preferences of consumers. Scope and limitations Apparel companies are interested gaining insight into the fit preferences of their consumers. Dissatisfaction with fit and inability to find the right size have been stated as the top reason for the majority of apparel returns (Goldsberry et al., 1996; McVey, 1984). Consumer dissatisfaction with size and fit may explain why many clothes end up on markdown racks. Retailers lose millions of dollars every year because of markdowns. Understanding fit preferences of consumers could help apparel firms provide better fitting apparel products to customers, boost revenues and lower markdowns. Pine (1993) noted that when consumers establish successful relationships with vendors, those companies gain a competitive advantage over other firms. Repeat orders from satisfied customers are often the result. A convenience sample was used for this research. College students between the ages of 17 and 25 participated in this study. The majority of students who participated in this self-reported study were Caucasian. Therefore, the results obtained will not be
Clothing fit preferences
61
IJCST 17,1
representative of the entire female population, and may not be applicable across a similar population of age and culture groups. Fit preference varies across different age groups and from culture to culture. What is considered ideal in one culture, may not be so in another. Today’s sizing is based on an ideal body proportions of the 1970s and provides little room to accommodate people whose bodies deviate from those.
62
Recommendations for future research On the survey instrument many respondents did not attempt the question on sizing. The researcher had difficulty in coding the data that was reported for this question. This confusion about apparel sizes among the respondents strongly suggests that consumers are thoroughly confused about the sizing systems followed by the apparel industry. A study could be done using an older population, as the population of the 50þ will grow to 4 million by the end of 2000. In addition, the percentage of the population between 18 and 34 will decline, creating opportunities for new products and services. People over 50 years old already have as much discretionary income as all the other age groups combined (Hahn, 1992). A study done by Belleau et al. (1994) on the fit preferences of older women show that there was dissatisfaction with types and fit of available apparel. Since, more than half of the sample in this study, reported fit problems relative to length of the garment. Future studies could be done to include the length dimensions of the person also. That is along with weight, height and body measurements, specific body area length such as leg length, length of torso and hand length of the respondents could also be included. These measurements can be easily obtained from body scanners. As the human form is three-dimensional, information on the measurements of all the body parts would help the industry provide better fitting garment to the consumers. This was an exploratory pilot study, the researcher encountered problems that are associated with an initial, self-reported study. A majority of the respondents reported that they did not know their waist and hip measurement. Therefore, while conducting a further study in this area it is recommended that the respondents actual measurements through body scans or physical measurements, be attached and compared to the reported data. From the analysis of the data, it can be concluded that this is an understudied topic. There is tremendous potential for further research in this area as the results have indicated customer satisfaction with fit of the ready-to-wear will have ramifications in markdowns and sales in ready-to-wear. References Ashdown, S. (2000), “Introduction to sizing and fit research”, Clemson Apparel Research, Fit 2000: The Fit Symposium. Clemson Apparel Research Center, Clemson, SC, May, available at: http://car.clemson.edu/fit2000/ Ashdown, S.P. and DeLong, M. (1995), “Perception testing of apparel ease variation”, Applied Ergonomics, Vol. 26 No. 1, p. 47. Belleau, D.B., Broussard, L., Summers, A.T. and Didier, J. (1994), “Attitudes of women over 50 towards apparel and media”, Perceptual and Motor Skills, Vol. 78, pp. 1075-84. Brown, P. (1992), Ready-To-Wear Apparel Analysis, Macmillan, New York, NY.
Consumer Reports (1996), “Why don’t these pants fit?”, May, pp. 38-9. Davis, L.L. (1985), “Perceived somatotypes, body-cathexis, and attitudes toward clothing among college females”, Perceptual and Motor Skills, Vol. 61, pp. 1199-205. Goldsberry, E., Shim, S. and Reich, N. (1996), “Women 55 years and older: Part II. Overall satisfaction and dissatisfaction with fit of ready-to-wear”, Clothing and Textiles Research Journal, Vol. 14 No. 2, pp. 121-32. Hahn, W. (1992), “Aging America”, The Annals of the American Academy of Political and Social Science, Vol. 522, pp. 116-29. Hazen, G.G. (1998), Fantastic Fit for Everybody, Rodale Press Inc., Emmaus, PA. Horn, M.J. (1975), The Second Skin, 2nd ed., Houghton Mifflin, Boston, MA. Horn, M.J. and Gurel, L.M. (1981), The Second Skin, 3rd ed., Houghton Mifflin, Boston, MA. Hudson, P. (1980), “The role of fit and fashion on apparel quality”, Bobbin, July, pp. 108-22. Hwang, J. (1996), “Relationships between body-cathexis, clothing benefits sought and clothing behavior: effects of importance of meeting the ideal body-image and clothing attitudes”, Unpublished doctoral dissertation, Virginia Polytechnic Institute and State University, Blacksburg, VA. Kaiser, S.B. (1990), The Social Psychology of Clothing, Macmillan, New York, NY. Kurt Salmon Associates (2000), “Annual consumer outlook survey”, paper presented at a meeting of the American Apparel and Footwear Association Apparel Research Committee, Orlando, FL. LaBat, L.K. and DeLong, R.M. (1990), “Body cathexis and satisfaction with fit of apparel”, Clothing and Textiles Research Journal, Vol. 8 No. 2, pp. 43-8. McVey, D. (1984), “Fit to be sold”, Apparel Industry, Vol. 45, pp. 24-6. Mahoney, R.E. and Finch, D. (1976), “The dimensionality of body-cathexis”, The Journal of Psychology, Vol. 92, pp. 277-9. Pine, B.J. II (1993), Mass Customization: The New Frontier in Business Competition, Harvard Business School Press, Boston, MA. Secord, P.F. and Jourard, S.M. (1953), “The appraisal of body-cathexis: body-cathexis and the self”, Journal of Consulting Psychology, Vol. 17, pp. 343-7. Shim, S., Kotsiopulos, A. and Knoll, S.D. (1991), “Body cathexis, clothing attitude, and their relations to clothing and shopping behavior among male consumers”, Clothing and Textiles Research Journal, Vol. 9, pp. 35-44, winter. Shim, S. and Bickle, C.M. (1994), “Benefit segments of the female apparel market: psychograpics, shopping orientations, and demographics”, Clothing and Textiles Research Journal, Vol. 2, pp. 1-12. Smathers, D. and Horridge, P.E. (1978-79), “The effects of physical changes on clothing preferences of the elderly women”, International Journal of Aging & Human Development, Vol. 9 No. 3, pp. 273-8. Swan, J.E. and Combs, L.J. (1976), “Product performance and consumer satisfaction: a new concept”, Journal of Marketing, Vol. 40 No. 1, pp. 25-33. Tamburrino, N. (1992), “Apparel sizing issues, Part II”, Bobbin, May, pp. 55-60. Williams, A. (1974), “Fit of clothing related to body-image, body built and selected clothing attitudes”, Unpublished doctoral dissertation, University of North Carolina, Greensboro, NC. Willams, S. (1998), “The fit factor”, Daily News Record, April, p. 9.
Clothing fit preferences
63
IJCST 17,1
64
Workman, J.E. (1991), “Body measurement specifications for fit models as a factor in clothing size variation”, Clothing and Textiles Research Journal, Vol. 10 No. 1, pp. 31-6. Workman, J.E. and Lentz, E.S. (2001), “Measurement specifications for manufacturers’ prototype bodies”, Clothing and Textiles Research Journal, Vol. 18 No. 4, pp. 251-9. Further reading Anderson, L.J. (1990), “A methodology for consumer style preference testing of apparel at the product development level”, Unpublished doctoral dissertation, Auburn University, Auburn, AL. Anderson, L.J., Brannon, E.L., Ulrich, P. and Marshall, T. (1997), “Confluences: towards a consumer driven model for mass customization in the apparel market”, Confluences: Fashioning Intercultural Perspectives, paper presented at the International Textile and Apparel Association Conference, Lyon, July. “Clothes that fit” (1988), Glamour, September, p. 122, 128. Delong, M., Ashdown, S., Butterfield, L. and Turnbladh, F.K. (1993), “Data specification needed for apparel production using computers”, Clothing and Textiles Research Journal, Vol. 11 No. 3, pp. 1-7. Engle, J.F., Blackwell, R.D. and Miniard, P.W. (1990), Consumer Behaviour, 6th ed., The Dryden Press, Chicago, IL. Garner, M.D. (1997), “The 1997 body-image survey results”, Psychology Today, Vol. 30, pp. 30-47. Graeber, M. (1999), “An investigation of store managers’ interest in mass customization”, Unpublished thesis, The Florida State University, Tallahassee, FL. Peavy, K. (1996), “An investigation of expert evaluations of market turbulence in the apparel industry”, Unpublished Master’s thesis, Auburn University, Auburn, AL.
The Emerald Research Register for this journal is available at www.emeraldinsight.com/researchregister
The current issue and full text archive of this journal is available at www.emeraldinsight.com/0955-6222.htm
Effect of sewing and fusing of interlining on drape behaviour of suiting fabrics Kaushal Raj Sharma and B.K. Behera Department of Textile Technology, Indian Institute of Technology, Delhi, India
H. Roedel and Andrea Schenk Institute of Clothing Science and Technology, Technical University, Dresden, Germany
Effect of sewing and fusing
75 Received August 2003 Revised October 2004 Accepted October 2004
Abstract Purpose – Drape of the fabric is its ability to hang freely in graceful folds when some area of it is supported over a surface and the rest is unsupported. When two-dimensional fabrics are converted to three-dimensional garment forms, a number of operations are required which affect drape behaviour of the fabric while present in garment form. In the present study, the effect of sewing and fusing of interlining on drape behaviour of men’s suiting fabrics is investigated. Design/methodology/approach – The effect of sewing and fusing of interlining on drape behaviour of men’s suiting fabrics is investigated. Comparisons were also made between different stitches (chain stitch and lock stitch), different seams for lock stitch and different types of interlinings for their effect on drape behaviour of fabrics. In addition to drape coefficient and number of folds, a new drape parameter – average amplitude to average radius (A/r) ratio – was also defined and calculated for drape image geometry. Findings – Drape coefficient has a good to strong correlation with A/r ratio and number of folds for most of the shell, sewn and interlining fused fabrics except for a few cases. A/r defines image in a more descriptive manner than drape coefficient. Drape coefficient changes with the types of seams and stitches used, as well as with the interlining used. Originality/value – This paper provides information on the effects of sewing (seams and stitch types) and fused interlining on drape behaviour of men’s suiting fabrics. Keywords Garment industry, Men, Fabric testing Paper type Research paper
Introduction Drape is an important property that decides the gracefulness of any garment as it is related to aesthetics and appearance of garments. It describes the way in which fabric hangs itself in specific shape according to its properties when part of it is supported by any surface and rest is unsupported. Drape is of much importance for designing and development of garments and selection of appropriate fabric for intended garment. The research in the area of drape of textile materials started around half a century ago when Chu et al. (1950) first of all invented (Fabric Research Lab.) FRL drapemeter for measurement of fabric drape. He coined the term “Drape coefficient” which is a dimensionless quantity and it is being widely used even until this time. The FRL drapemeter was further modified by Chu himself, but major modifications were subsequently carried out by Cusick (1965). The modifications envisaged the use of point source of light, paper weighing method and different supporting disc diameter
International Journal of Clothing Science and Technology Vol. 17 No. 2, 2005 pp. 75-90 q Emerald Group Publishing Limited 0955-6222 DOI 10.1108/09556220510581227
IJCST 17,2
76
for different stiffness of fabrics. Further modifications of Cusick drapemeter by use of computerised image analysis was also reported by Vangheluwe and Kiekens (1993) and Ro¨del et al. (2000). By such modifications accuracy and efficiency improve and better analysis of drape image geometry can be made. The various components of garment include sewn fabric, interlining, lining and accessories. The components which affect garment appearance due to change in fabric drape are mainly seams and interlinings. So the effects of sewing and fusing of interlinings on fabric drape must be investigated in detail. Hu and Chung (1998) determined the effect of radial and circular seams on drape coefficient and drape profile of fabrics. In the present study, the Cusick drapemeter was modified using digital image processing to determine different drape parameter precisely. To analyse the drape image geometry in a better way, some more drape parameters in addition to conventional drape coefficient were also determined. Drape of the fabrics is generally determined for shell fabrics before their conversion to garments. During garment manufacturing number of operations are carried out on shell fabrics which change their drape behaviour. Such operations are sewing for their conversion from two-dimensional to three-dimensional form, fusing of interlining to improve stiffness, etc. Whole study was conducted for men’s suiting fabrics made of 100 per cent wool and 55/45 polyester wool (PW) blends in different weight/area. Effects of sewing and fusing of interlining on drape behaviour of these fabrics were determined so that drape behaviour of fabrics in actual garment form can be predicted. To observe the effects of sewing, seams were made in warp direction of fabric along the diameter of circular drape specimen. Warp direction of seams was chosen because most of the seams are applied in warp direction in garments. Two types of stitches, lock stitch (LS1) and chain stitch (CS), were applied to observe the effect of type of stitch while within lock stitch four types of seams (LS1, LS2, LS3 and LS4) were applied in order to observe the effect of type of seam for same stitch. Shell fabrics were fused to interlinings to determine the way in which interlinings affect fabric drape after fusing. Comparison was also made between type of interlinings by taking woven base fabric interlining and knitted base fabric interlining. Materials Fabrics Two groups of fabrics covering a range of Areal density (weight per unit area) were chosen keeping the end product as men’s jacket in mind. One group was 100 per cent wool fabrics comprising fabric samples F1-F4 and the other was 55/45 PW blended fabrics from F5 to F7. The fabric details are shown in Table I. Interlinings Two types of interlining were selected, one with woven base fabric and the other with warp knitted weft inserted base fabric. Same woven base fabric interlining was fused to all samples while knitted base fabric interlinings were selected according to their compatibility to respective shell fabrics. The specifications of interlinings are shown in Table II.
Plain 52 48 16 14 166.74
Twill 66 48 11 11 239.60
97 per cent W, 3 per cent EA
100 per cent W
Notes: P¼ Polyester; W ¼ Wool; EA ¼ Elastane Fibre
Weave Ends/inch Picks/inch Warp count(Ne) Weft count(Ne) Weight(g/m2)
Blend
F2
F1
Twill 108 68 14 14 297.30
100 per cent W
F3
Twill 156 108 2/28 2/34 390.20
100 per cent W
Fabric sample F4
Twill 100 60 2/34 2/34 203.30
55 per cent P, 45 per cent W
F5
55 per cent P, 45 per cent W 53 per cent P, 44per cent W, 3 per cent EA Twill 76 60 2/36 21 257.00
Twill 94 100 2/24 2/24 293.80
F7
F6
Effect of sewing and fusing
77
Table I. Specifications of shell fabric samples
IJCST 17,2
78
Table II. Properties of interlinings
Interlining W1
K11
K12
K13
Weight/area(g/m2) Base cloth
70 Woven
Base cloth content
100 per cent polyester
48 Warp knitted, weft inserted 100 per cent polyester
Base cloth structure Longitudinal threads Lateral threads Type of adhesive Adhesive Dots/area(dots/cm2) Fabric to which fused
Cross twill 2/2 Straight fringe
73 Warp knitted, weft inserted 73 per cent viscose, 27 per cent polyamide Staggered fringe
67 Warp knitted, weft inserted 69 per cent viscose, 31 per cent polyamide Staggered fringe
Polyester
Polyester
Polyamide
Polyamide
Polyester Polyamide-3P 52
Polyester Polyamide-3P 66
Viscose Polyamide-3P 52
Viscose Polyamide-3P 52
F3
F6.F7
F1,F2,F3,F4,F5 F1,F2,F4,F5 F6.F7
Methods To observe the effect of sewing on fabric drape the fabric samples had to be sewn first. For sewing, stitches were applied in single straight line along the diameter of circular fabric specimen in warp direction. Two types of stitches viz. CS and LS1 were applied to observe the effect of type of stitch on drape. Within lock stitch four types of seams (LS1, LS2, LS3, and LS4) were applied to observe the effect of type of seam on drape. For determination of effect of fusing of interlinings on drape, interlining was fused to circular fabric specimen before making drape test. Effect of type of interlining on drape was also observed by taking two types of interlining with woven base fabric (WI) and knitted base fabric (KI), respectively. Sewing Textima lock stitch sewing machine model 8332/3005 was used for all types of lock stitches and Textima chain stitch sewing machine model 8431/8410/1 was used for chain stitches. Two ply, 100 per cent Polyester sewing thread and Schmetz sewing needle of fineness 80 Nm were used. The specifications of seams and stitches applied to shell fabrics in warp direction are shown in Figure 1.
Figure 1. Specifications of seams and stitches applied to shell fabrics
Fusing of interlining VEIT-5310 Flat bed fusing press with scissors action with working area 90 £ 48 cm was used for fusing of interlinings to shell fabrics. Time, temperature and pressure were selected according to requirements shown in Table III. Drape testing The Cusick drapemeter was modified using digital image processing. The digital camera used was NIKON-950 with video output. To process the signal captured by the camera a frame grabber card was connected between the camera and the computer. The board contains software that is responsible for capturing still image, converting to video image and editing. It digitizes the image obtained from camera for feeding to monitor. The digitised image consists of a matrix of 256 £ 256 pixels. Each pixel has a value in the range of 0-255 which corresponds to a grey scale observed by camera. There are 256 different grey scales with the zero value representing black and 255 representing white. The pixels are used for the determination of surface area of specimen from the image received through image processing. For each sample ten tests of drape were carried out and all ten individual images were saved as a film in computer. Before taking images of samples individual image of supporting disc (18 cm diameter) and larger disc (30 cm diameter) of drapemeter were saved as initial images of film for each sample. The area of these two discs was also calculated which was essentially required for the calculation of drape coefficient. After storing images in the computer during drape testing these were modified in terms of brightness, darkness and contrast and then converted to Pixel format in files with extension name PCX. The PCX formatted files were then run through Fourier analysis program in computer. The results after Fourier analysis were available in terms of Fourier coefficients and drape coefficient separately for each image present in the file with extension OUT. One such file was for one film. These coefficients were converted to the results in terms of drape coefficient, number of folds, minimum radius, maximum radius, average radius and amplitude in the form of a table for each fabric image when OUT files were processed by a macro written in Visual Basic language, driven in Microsoft Excel. Shell fabric specimen for drape testing was prepared by cutting the fabric as per the dimension of the circular disc of 30 cm diameter provided as the larger disc of drapemeter. For sewn fabric samples, the fabric was cut along warp direction into long strips of 20 cm width. Two such adjacent strips were then sewn together on sewing machine maintaining the grain line of the fabric same. The sewn fabrics were then ironed to set the seam and subsequently conditioned in standard atmosphere for 24 h.
W1 Applicable range of temperature(8C) Applicable range of pressure(p/cm2) Applicable range of time (s) Applied temperature(8C) Applied pressure(Pa/cm2) Applied time(s)
121-132 200-350 12-15 129 250 14
Type of interlining K11 K12 116-132 200-350 12 129 250 12
124-134 200-350 12 129 250 12
Effect of sewing and fusing
79
K13 124-134 200-350 12 129 250 12
Table III. Parameters for fusing of interlinings to shell fabrics
IJCST 17,2
80
Then they were marked according to the periphery of larger circular disc and cut by scissors. For preparing the specimen of interlining fused fabric each fabric and interlining were cut into rectangular pieces of 70 £ 35 cm and 68 £ 33 cm, respectively (the dimensions of rectangular pieces were restricted by the working area of fusing press). Both were then fused into the fusing press at appropriate fusing conditions. After fusing they were allowed for cooling and conditioning and then marked to the shape of larger circular disc and cut by scissors. All types of specimen were conditioned in standard controlled atmosphere of 20 ^ 28C temperature and 65 ^ 2 per cent relative humidity for 24 h before testing. Determination of drape parameters by Fourier analysis Fourier analysis analyse the computer-internal pixel diagram of the fabric drape image to determine its characteristic features. The image pixels are distinctive into white and not white pixels. The focal point of drape image is determined first by searching white pixels within the image centre. White pixels are formed due to light of glowing light emitting diode (LED) placed exactly at the centre of the drape specimen on the drapemeter. Knowing the focal point is necessary for determination of two-dimensional polar co-ordinates system. Outgoing from the focal point in a straight line the drape image is searched radially for the first white pixel. In order to determine the outline of the drape shadow geometry with sufficient accuracy this is scanned in a gradation of 38 of angle, so that n ¼ 120 points along the curve are available. The points of the image with their determined co-ordinates (radius ri and angle wi with i ¼ 1; . . . ; n) are used as basis for Fourier series development. Drape image geometry is shown in Figure 2 in which co-ordinates of boundary point, its radius and angle are also shown. The approximation of the boundary curve is realised with the approach: n n 2 21 an a0 X ½ak cosðkwÞ þ bk sinðkwÞ þ 2 cos w r ¼ f ð wÞ ¼ þ 2 2 2 k¼1
The function f(w) consists of partial sums of the sine and cosine terms, whereby the Fourier coefficients ak and bk are the amplitudes of the partial waves which are also called Fourier orders:
Figure 2. Drape image geometry
ak ¼
n n 2X 2X r i cosðkwi Þ and bk ¼ r i sinðkwi Þ n i¼1 n i¼1
Determination of area for the calculation of drape coefficient By means of double integration of the Fourier series within the boundaries of 0-2p over the angle w, the area of the drape image can be calculated according to the following formula: n21 Z 2 p Z f ðw Þ Z 2p p a2n 2 X a2 1 2 f ð wÞ d w ¼ 0 p þ þ 2p A¼ r dr dw ¼ a2k þ b2k 2 4 2 8 0 0 0 k¼1 Determination of the number of nodes N All Fourier coefficients with their terms collectively represent a complete mathematical description of the analysed boundary curve of the drape image. The terms with Fourier coefficients ak and bk can be represented by means of addition rule: ak cosðkwÞ þ bk sinðkwÞ ¼ ck cosðkw 2 kck Þ whereby ck and ck can be calculated as follows qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi bk 2 2 ck ¼ ak þ bk and kck ¼ arctan ak After sorting the ck-terms of the Fourier series, the Fourier order of the highest influence is defined and according to it the number of folds are determined. Determination of the absolute size for description of the folds shape For the description of the folds shape it is appropriate to introduce dimensionless folds shape relation A/r (amplitude to average radius) ratio. Therefore, the reference amplitude A and the reference radius r are determined from the radius rH (the maximum outward extent of folds from the centre in projected drape picture) and the radius rP (the minimum inward extent of folds from the centre in projected drape picture) as follows: A¼
rH 2 rP 2
and R ¼ a0 ¼ r H 2 A ¼ r P þ A
The value of rH and rP represents the maximum and the minimum value of ri, respectively. Parameters rH, rP, r and A are shown on drape image in Figure 3. Determination of drape coefficient After determining the area of smaller disc (Ad), area of larger disc (AD) and area of draped sample (AS) from their respective drape image geometry, the drape coefficient (D) is calculated for each sample according to the formula: D¼
As 2 Ad AD 2 Ad
Effect of sewing and fusing
81
IJCST 17,2
By digital image processing and using Fourier analysis six drape parameters viz. drape coefficient, number of folds, amplitude, average radius, minimum radius and maximum radius were obtained from the drape image geometry of fabric specimen.
82
Results and discussion Drape parameters Drape parameters such as drape coefficient, A/r (amplitude to average radius) ratio and number of nodes determined by modified drape metre is shown in Tables IV-VI, respectively. The correlation of A/r (amplitude to average radius) ratio and the number of nodes with drape coefficient is discussed for all shell, sewn and interlining fused fabrics and are shown in Figures 4-9. Drape coefficient is the conventional parameter
Figure 3. Determination of rH, rP, r and A from drape image geometry
Drape coefficient
Table IV. Drape coefficient of shell, sewn and interlining fused fabrics
Shell fabrics CS-1 LS-1 LS-2 LS-3 LS-4 WO WK
A/r
Table V. Amplitude to average radius (A/r) ratio for shell, sewn and interlining fused fabrics
Shell fabric CS-1 LS-1 LS-2 LS-3 LS-4 WO WK
F1
F2
F3
F4
F5
F6
F7
0.425 0.559 0.543 0.578 0.565 0.558 0.718 0.733
0.442 0.533 0.522 0.538 0.544 0.534 0.731 0.792
0.472 0.544 0.535 0.606 0.554 0.557 0.757 0.820
0.610 0.766 0.733 0.780 0.775 0.764 0.821 0.849
0.353 0.504 0.465 0.480 0.486 0.482 0.677 0.809
0.426 0.595 0.569 0.587 0.587 0.594 0.783 0.835
0.611 0.703 0.698 0.732 0.733 0.734 0.816 0.863
F1
F2
F3
F4
F5
F6
F7
0.122 0.150 0.158 0.147 0.156 0.168 0.108 0.115
0.118 0.155 0.146 0.150 0.157 0.165 0.077 0.091
0.129 0.154 0.156 0.136 0.152 0.149 0.105 0.086
0.105 0.102 0.096 0.098 0.092 0.106 0.085 0.074
0.158 0.177 0.190 0.168 0.181 0.200 0.098 0.075
0.112 0.134 0.142 0.144 0.140 0.154 0.085 0.073
0.110 0.112 0.113 0.099 0.101 0.117 0.090 0.077
used to define drape behaviour. So correlations of A/r and the number of folds with drape coefficient were determined so that these relatively new parameters could be understood in relation to drape coefficient. The drape coefficient gives an indication about the appearance of the fabric when hanging under its own weight in unsupported region and gives only a non-descriptive idea about fabric behaviour while hanging freely. The A/r is the ratio of average amplitude of the folds to the average radius of the projected drape image of a fabric sample. It is a unit less factor which gives more detailed description of form of fabric draping as compared to drape coefficient since it is more related to drape image geometry. A/r is mainly related to actual fabric appearance during free hanging of fabric. It is related to bending of fabric in curvatures and moreover to the extent of curvature. With fabric appearance point of view in garments this is more important than drape coefficient. It can be observed from Figures 4 to 6 that as the A/r value increases, the drape coefficient decreases in all the cases. It supports the idea that drape coefficient and the reciprocal of A/r work in the same way to define drape behaviour, the difference being only the extent to which they define geometry of drape image. Further it can be observed that for all the seven types of suiting fabrics the coefficient of linear correlation between the drape coefficient and the A/r value is found to be 2 0.739 for shell fabrics, from 2 0.9581 to 2 0.9923 for five different sewn fabrics and 2 0.4025 to 2 0.9052 for interlining fused fabrics, which can be considered as good to strong correlations. For different values of A/r the projected drape images are shown in Figure 10. Analysing A/r ratio it may be perceived that if A/r increases, three conditions can prevail: (1) amplitude increases – the depths in which folds appear in draped fabric increases; (2) average radius decreases – the fabric in draped condition hangs more sharply from the edge of supporting discs of drapemeter so that its projected shadow reduces; and (3) when both the above conditions occur simultaneously, it means the fabric hangs more sharply from supporting disc and also makes deep folds so that fabric characteristics approach towards characteristics of flexible fabric, which is definitely the condition of lower drape coefficient. No. of folds
F1
F2
F3
F4
F5
F6
F7
Shell fabric CS-1 LS-1 LS-2 LS-3 LS-4 WO WK
8.7 7.8 7.7 7.9 7.9 7.2 7.8 6.2
8.7 8.4 8.1 8.0 7.8 7.6 8.1 6.5
8.6 7.5 7.5 8.0 7.8 7.5 7.7 7.2
7.9 7.3 7.5 7.4 7.3 6.8 7.0 6.6
8.6 7.7 7.2 8.4 7.8 7.3 8.1 8.1
8.6 8.1 7.7 7.9 8.1 7.8 7.5 7.5
8.0 7.3 7.6 7.9 7.6 7.5 7.3 7.1
Effect of sewing and fusing
83
Table VI. Number of folds for shell, sewn and interlining fused fabrics
IJCST 17,2
84
Figure 4. Correlation between drape coefficient and A/r
Figure 5. Correlations between drape coefficient and A/r combined for all types of seams/stitches
Figure 6. Correlations between drape coefficient and A/r for interlining fused fabrics
These observations suggest that as the value of A/r increases, drapability of the fabric also increases. The value of A/r is limited by two conditions, they are (r þ A) # 15 and (r þ A) $ 9 by which the value of A/r can be theoretically inbetween 0 and 0.25. Two more limits for A/r exist which are 0 # A # 3 and 9 # (R 2 A). Like A/r, number of folds also helps to understand the geometry of drape image more thoroughly. The number of folds are the nodes that are formed when the fabric hangs in space in a wave form in its unsupported region. With the drape coefficient the number of folds
Effect of sewing and fusing
85
Figure 7. Correlation between drape coefficient and number of folds in drape image
Figure 8. Correlations between drape coefficient and number of folds combined for all types of seams/stitches
Figure 9. Correlations between drape coefficient and number of folds for interlining fused fabrics
IJCST 17,2
86 Figure 10. Drape images for different values of A/r
have strong negative correlation as linear coefficient of correlation for shell fabric is 2 0.909; for five types of sewn fabric it is between 0.0084 and 2 0.7487 and for interlining fused fabrics it is from 0.431 to 2 0.9233. Figures 7-9 are the graphical representations of correlation equations. Except for one type of sewn fabric (LS1) and one type of interlining fused fabric (WI), the correlation between the drape coefficient and the number of folds can be said as good to strong. Effect of sewing on drape coefficient Whenever any type of seam is applied to a fabric, it increases its drape coefficient. This can be observed from the results shown in Table VII and Figures 11 and 12. It is clear from Figure 12 that the drape coefficient for shell fabric is least as compared to that of sewn fabrics. The application of any seam to the fabric increases its resistance to fall
F1 Shell fabric (Plain) 0.425 CS 0.559b,d,e Per cent Difference 31,53 LS-1 0.543a,d,e Per cent Difference 27.76 LS-2 0.578d,e Per cent Difference 36.00 LS-3 0.565a,b,c,e Per cent Difference 32.94 LS-4 0.558a,b,c,d Per cent Difference 31.29
Table VII. Drape coefficient of sewn fabrics
F2
F3
0.442 0.533c,d,e
0.472 0.544b,e
Fabric sample F4
0.61 0.766c,d,e
F5
F6
F7
0.353 0.504d,e
0.426 0.595c,d,e
0.611 0.703b,c,d,e
20.59 0.522e
15.25 0.535a
25.57 0.733
18.10 0.538a,d,e
13.35 0.606
20.16 31.73 0.780a,d,e 0.480b,d,e
21.72 0.544a,c,e
28.39 0.554e
27.87 35.98 37.79 19.80 0.775a,c,e 0.486a,b,c,e 0.587a,b,c,e 0.733a,b,c,e
23.08 17.37 0.534a,b,c,d 0.557a,d 20.81
18.01
42.78 0.465c,d,e
39.67 0.569d,e
15.06 0.698a,c,d,e
33.57 0.587a,d,e
14.24 0.732a,b,d,e
27.05 37.68 37.79 19.97 0.764a,c,d 0.482a,b,c,d 0.594a,b,c,d 0.734a,b,c,d 25,25
36.54
39.44
20.13
Notes: a CS; b LS1; c LS2; d LS3; and e LS4 for same fabric only. (Superfix indicates significance at 95 Per cent level of confidence, e.g. Superfix a on any drape coefficient value indicates significant difference between that drape coefficient and drape coefficient of CS for that fabric)
against gravity. A sewn fabric is not one piece of fabric it is in two parts of same fabric which are joined by an additional sewing thread. This leads to the increase in fabric stiffness which can be due to additional fabric as a result of seam under the face side of drape specimen of sewn fabric. This additional fabric lies under the fabric as second layer and supports the top layer of specimen against hanging which leads to increase in drape coefficient of specimen. It can also be due to the sewing thread used to sew two pieces of fabric. When this thread traverses between the two fabric pieces it acts like forming a bridge between two fabric pieces and provides rigidity in the seam line which therefore increase its bending rigidity. Further it may be seen from Table VII that the extent of increase in drape coefficient of shell fabric after the application of seam is in the range of 13.35-42.78 per cent. From Table VII and Figure 13, comparing the effect of type of stitch it may be seen that the drape coefficient for fabrics with CS is always higher than the drape coefficient
Effect of sewing and fusing
87
Figure 11. Comparison of drape coefficient values for seven types of fabrics in plain and seamed state
Figure 12. Trend of change in drape coefficient for seven types of fabrics in plain and seamed state
Figure 13. Trend of change in drape coefficient values for different kind of seams/stitches applied to seven types of fabrics
IJCST 17,2
88
for respective fabric with lock-stitch-1 (LS1). However, for fabrics F1, F3 and F7 the difference between drape coefficients of the two is insignificant. The higher drape coefficient of cross stitch sewn fabrics than lock stitch sewn fabrics may be due to incorporation of more thread in the stitch line of cross stitch under the fabric. In each single stitch of cross stitch two more thread segments are present than for similar case of lock stitch which provide extra stiffness to the sewn fabric. To observe the difference of effect of type of seam on drape, sewn fabrics LS1, LS2, LS3 and LS4 are compared. From Figure 13 and Table VII it may be seen that no significant difference in drape coefficient has been found among LS2, LS3 and LS4 that seams in all samples except F3. So it can be said that once the drape specimen becomes three layered as for LS2 and LS3 the addition of more fabric layer (as in LS4) or extra stitch (as in LS3 and LS4) does not affect its drape coefficient. From Figures 12 and 13, it can be observed that drape coefficient of fabrics with LS1 is always less than the drape coefficient of respective fabrics with LS2, LS3 and LS4. So the drape coefficient of fabrics with LS1 is minimum among all five seams/stitches applied. In LS2, LS3 and LS4 more fabric is entrapped in the stitch line than in LS1. Therefore, the drape specimen becomes three-layered structure unlike two-layered structure in LS1 which raises their drape coefficient significantly. Effect of fusing of interlinings on drape coefficient Table VIII shows the values of drape coefficient for shell and interlining fused fabrics and the percentage difference between them. KI represents fabric fused with knitted base fabric interlining and WI represents fabric with woven based fabric interlining. The drape coefficients were tested for significance at 95 per cent level of confidence. For all types of fabrics, drape coefficient increases as the interlinings (interlining with woven base fabric as well as knitted base fabric) are fused to them, which can also be clearly seen from Figures 14 and 15. It is in the range of 33.55-129.18 per cent for different fabric-interlining combinations as a whole as shown in Table VIII. When the shell fabric is fused to interlining, it becomes a composite and its bending stiffness and shear stiffness increase in all directions. It leads to reduction in the freedom of yarns in fabric for any type of movement which is required by the yarns to hang freely by gravity while draping which eventually increases the drape coefficient of interlining fused fabric. Comparing woven and knitted base fabric interlinings, it may be seen that for all types of fabrics the drape coefficient when fused with interlining having woven base fabric is more than the drape coefficient when fused with interlining having knitted base fabric. For woven base fabric interlining, the drape coefficient of shell interlining is higher than that of interlining with knitted base fabric as shown in Table IX. Drape coefficient
Table VIII. Drape coefficients of interlining fused fabrics and their percentage difference with respective shell fabrics
Shell fabric (Plain) WK Percentage difference WO Per cent Difference
F1
F2
F3
F4
F5
F6
F7
0.425 0.718 g 68.94 0.733f 72.47
0.442 0.731 65.38 0.792 79.19
0.472 0.757 60.38 0.820 73.73
0.610 0.821 34.59 0.849 39.18
0.353 0.677 91.78 0.809 129.18
0.426 0.783 83.80 0.835 96.01
0.611 0.816 33.55 0.863 41.24
Notes: fWO; and gWK, for same fabric only
The woven structure of base fabric is also more compact and rigid as compared to knitted structure, which is relatively open and flexible. So these factors lead to higher drape coefficient of fabric fused with woven base fabric interlining as compared to fabric fused with knitted base fabric interlining. It may further be seen from Table VIII that the percentage of increase in drape coefficient for the shell fabrics fused with knitted base fabric interlining is in the range of 33.55-91.78 per cent, whereas for the fabrics fused with woven base fabric interlining this range is 39.18-129.18 per cent. For both types of fused fabrics this extent of increase in drape coefficient over shell fabrics decreases as the weight/area of shell fabrics increases. So for light weight fabrics this extent is maximum.
Effect of sewing and fusing
89
Conclusions Drape coefficient has good to strong correlation with A/r ratio and number of folds for most of the shell, sewn and interlining fused fabrics except few cases. On stitching with different types of seams, the drape coefficient increases by 13.35-42.78 per cent
Figure 14. Comparison of drape coefficient values for seven types of fabrics in plain and interlining fused state
Figure 15. Trend of change in drape coefficient values for seven types of fabrics in plain and interlining fused state
Interlining Type Drape coefficient A/r No. of nodes
L1 Woven 0.465 0.145 8.8
L2 Knitted 0.274 0.19 9.3
L3 Knitted 0.309 0.194 8.6
L4 Knitted 0.376 0.203 8.5
Table IX. Drape parameters for shell interlinings
IJCST 17,2
90
depending on stitch-fabric combination. Comparing cross stitch and lock stitch for their effect on fabric drape in sewn state, the drape coefficient of fabrics with cross stitch is higher however, the difference between the two types of stitch is not significant. Comparing various seams of lock stitch for drape coefficient of sewn fabric, it was found that there is no substantial change in drape coefficient for these seams. For LS2, LS3 and LS4, the drape coefficients have insignificant difference with each other. The drape coefficient of LS1 is minimum in all the cases because for LS2, LS3 and LS4 fabric specimen becomes three or four layered in seam zone which leads to little increase in drape coefficient as against LS1 where specimen is two layered in seam zone. When any interlining is fused to shell fabric its drape coefficient increases substantially. The drape coefficient of interlining fused fabrics with woven base fabric interlining is higher than that with knitted base fabric interlining presumably due to higher drape coefficient of shell woven base fabric interlining as compared to that of shell knitted base fabric interlining. References Chu, C.C., Cummings, C.L. and Teixira, N.A. (1950), “Mechanics of elastic performance of textile materials – Part V: A study of the factors affecting the drape of fabrics, the development of drape meter”, Textile Research Journal, Vol. 20, pp. 539-48. Cusick, G.E. (1965), “The measurement of fabric drape”, Journal of the Textile Institute, Vol. 59 No. 6, pp. T253-60. Hu, J. and Chung, S. (1998), “Drape behavior of woven fabrics with seams”, Textile Research Journal, Vol. 68 No. 12, pp. 913-19. Ro¨del, H., Ulbricht, V., Krzywinski, S., Schenk, A. and Fischer, P. (2000), “Simulation of drape behaviour of fabrics”, International Journal of Clothing Science and Technology, Vol. 10 No. 3/4, pp. 201-8. Vangheluwe, L. and Kiekens, P. (1993), “Time dependence of the drape coefficient of fabrics”, International Journal of Clothing Science and Technology, Vol. 5 No. 5, pp. 5-8. Further reading Chu, C.C., Platt, M.M. and Hamburger, W.J. (1960), “Investigation of the factors affecting the drapeability of fabrics”, Textile Research Journal, Vol. 30 No. 1, pp. 66-7. Hearle, J.W.S., Grosberg, P. and Backer, S. (1969), Structural Mechanics of Fibres, Yarns and Fabrics, Chapter 12. Vol. 1, Wiley, New York, NY. Lindberg, J., Waesterberg, L. and Svenson, R. (1960), “Wool fabrics as garment construction materials”, Journal of the Textile Institute, Vol. 51, p. T1475.
The Emerald Research Register for this journal is available at www.emeraldinsight.com/researchregister
The current issue and full text archive of this journal is available at www.emeraldinsight.com/0955-6222.htm
An interactive body model for individual pattern making
An interactive body model
Youngsook Cho and Naoko Okada Department of Textile Science and Technology, Graduate school of Shinshu University, Shinshu University, Nagano-ken, Japan
Hyejun Park Korea Research Institute of Standards and Science, Daejon, South Korea
91 Received July 2004 Accepted October 2004
Masayuki Takatera, Shigeru Inui and Yoshio Shimizu Faculty of Textile Science and Technology, Shinshu University, Nagano-ken, Japan Abstract Purpose – In order to mass-customize clothes, it is essential to consider individual body shape using computerized 3D body models. This paper describes the development of an interactive body model that can be altered with individual body shape for the purpose of computerized pattern making. Design/methodology/approach – For altering perimeter and length for contouring individual body shapes, a cross-sectional line model is proposed arranged at regular intervals. This model is easy for controlling body shape and also for calculating length and perimeters. Shape control lines (SCL) are used to modify the shape of the model in order to adjust the model to represent different body shapes. SCL are used to modify the perimeter of the cross-sectional line by scaling method with different center position and scaling ratio in a horizontal direction. Findings – In order to investigate whether virtual body models can be adequately substituted for real physical models, the perimeter and cross-section areas of shape control lines were compared, which resulted in an agreement ratio of over 93 percent. This fact supports the adaptability and potential usefulness of the body model. Originality/value – This research makes it possible for customers to modify the body model to match their own body shape during internet or catalogue shopping; it can also enable apparel manufacturers to communicate with their customers by describing the body model to fit on the screen while in the ordering process. Keywords Computer applications, Modelling, Physical testing, Shopping, Human anatomy, Clothing Paper type Research paper
Introduction In recent years, the clothing industry has changed due to the industry’s increased awareness of fashion and design. As the industry has changed from mass production to mass customization, it has needed to develop new production techniques, that provide to individual preferences, particularly apparel fitting that must be incorporated into pattern-making systems. However, many surveys have shown that there are still both demands and complaints in the current sizing system; “There are too few sizes and types for obese and slim customers, most designs cater for average body sizes,” “It is difficult to find suitable pants because of the fixed ratio between hip, thigh and waist sizes”
International Journal of Clothing Science and Technology Vol. 17 No. 2, 2005 pp. 91-99 q Emerald Group Publishing Limited 0955-6222 DOI 10.1108/09556220510581236
IJCST 17,2
92
(Hashizume and Nagata, 1999). Therefore, the development of techniques for better fitting clothing for individual body shapes is continuing. By using three-dimensional (3D) human body data measured by 3D scanning technology (Nam and Han, 2001), it is possible to make personalized dummies for use in fitted pattern making. However, it is expensive and difficult to find enough storage for physical dummies. Therefore, computerized 3D body models of the human body (Xu et al., 2002; Inui, 2001; Watanabe, 1999; Jones et al., 1995; Stylios et al., 2001; Minoh, 1997; Kurokawa, 1997) have recently attracted considerable attention in the clothing industry. For instance, in order to express the curved body surface geometrically, many methods are developed such as cross section model, triangle polygon model (Toyoda, 1998), B spline model, superquadics (Horikoshi and Kasahara, 1990), the Delaunay triangulation (Yamamoto and Uchiyama, 1995) and meta-balls (Imaoka, 1997). Although there are a number of research efforts in the field of 3D virtual human body modeling, almost none of them are focused on pattern making they are focused on body mechanics. Also they are not enough to correspond to individual variations in the shape and size of the human body. Even using shape control parameters, existing body models modify the whole body by proportion. There is no investigation into whether virtual body models can be adequately substituted for real physical models. The goal of our work is to develop an interactive body model that can be altered with individual body shape for the purpose of computerized pattern making. Method 3D measurement of body shape and basic body model A female dummy as a basic model is measured using a 3D shape scanner (Minolta Vivid700). Scanned data from different angles were combined to generate a 3D body model. Scanned data have ASC file and DXF file format. The 3D body model thus generated could then be used interactively (Table I and Figure 1). Cross-sectional line modeling In order to alter dimensions for contouring individual body shapes, we propose a cross-sectional line model. It is easy to control the body shape and also easy to calculate length and perimeters because the model consists of cross-sectional lines arranged at regular intervals. In this research, body models consist of cross-sectional lines arranged at 10 mm intervals. To construct the cross-sectional line model, sections 3 mm deep were assigned with a gap of 10 mm between each section. Using sorting method, points in the sections of 3 mm deep are arranged. Adjacent points are linked as a cross-sectional line. The process is shown in Figure 2.
Types
Table I. Used dummy information
Dummy of the upper half the body Dummy of the lower half the body
Product name
Dummy dimension
Maker name
Ladies’
Bust: 83 cm/waist: 60 cm/hip: 90.5 cm
Kiiya
Lirica
Waist: 62 cm/hip: 91 cm
Hosokawa
An interactive body model
93 Figure 1. The female dummy of upper and lower half the body and 3D body shape data
Figure 2. The process of cross-sectional line model construction
However, it is difficult to arrange the points at as adjacent points orderly. First, all of the adjacent points are calculated each of them in an area distance of 3 mm deep and arranged on line in order of the shortest distance. To reduce the complexity of calculating the distance between points, we divide each section into four areas of points group (I, II, III, IV). The areas are divided by two divisional lines (Du, Dd) and a centroid (Gx, Gz) in X and Z coordinates. The divisional lines consisted of middle position of the maximum and minimum Z value in a horizontal line. The formulae of Gx, Gz and Du, Dd are described as follows. Gx ¼
n 1X xi ; n i¼1
Gz ¼
n 1X zi n i¼1
ðn : number of pointsÞ
The upper division lineðDu Þ : Z ¼ ððMax Z Þ þ Gz Þ=2 ðMax Z : the maximum Z valueÞ The lower division lineðDd Þ : Z ¼ ððMin Z Þ þ Gz Þ=2 ðMin Z : the minimum Z valueÞ Furthermore, we create dashed lines at a mm intervals which cross the Z coordinate ðz ¼ a £ nÞ and at b mm interval which cross the X coordinate ðx ¼ b £ nÞ: In each of
IJCST 17,2
94
Figure 3. The division area and point reduction method
Figure 4. Shape control lines
the four areas of point group, extracted intersections crossed the dashed line. Finally, the cross-sectional line consists of intersections. This process cuts the number of points down by 80 percent so that we attained a smooth curve that is similar with original line shape (Figure 3). Assigning shape control lines In constructing the interactive body model, we assigned certain SCL that were used to modify the shape of the model representing different body shapes. We have three main factors, which influence body shape (height, bust, and hip) and add six other SCL; underbust, waist, stomach, thigh, calf and ankle to fine tune the model to specific body shapes. The SCL are shown in Figure 4.
Scaling method The interactive body model is constructed using scaling methods which are applied to the dimension parameters. Existing body models usually modify the perimeter of the cross-sectional line by scaling method based on centroid. Scaling ratio ðP ¼ ðS 2 RÞ=RÞ is described as the perimeter of the cross-sectional line of the basic model (R) and the perimeter of the cross-sectional line of the modified body model (S). However, modifying the whole body by same scaling ratio with same centroid is not suitable to represent real body shapes accurately. Therefore, centroid (Gx, Gz) is moved as center position (Gx, Gzþ Z) and horizontal scaling ratio is different from nine SCL (Figure 5). We had center positions (Gx, Gzþ Z) with maximum Z value and did no scaling in the horizontal direction of the upper bust parts. For underbust, a center position (Gx, Gzþ Z) with 0 of Z value and no scaling in the horizontal direction was used. To retain body shape from bust to underbust, center positions of Z values were changed gradually. For the waist, a center position (Gx, Gzþ Z) with 0 of Z value and with P scaling ratio in the horizontal direction were used. For the stomach, a center position (Gx, Gzþ Z) with 60 of Z value and with P/2 scaling ratio in the horizontal direction were used. For the hip, two scaling methods are required, one for increasing and the other for reducing size. The method for increasing uses a center position (Gx, Gzþ Z) with 60 of Z value and with P/2 scaling ratio in the horizontal direction. The method for reducing uses a center position (Gx, Gzþ Z) with 2 60 of Z value and with P/2 scaling ratio in the horizontal direction. For parts below the crotch, the same method was used as for waist. In the whole body, eight SCL are scaled. For gained smoothing body shape, other lines in the whole body have a need to be scaled. Therefore, eight SCL added four other lines (the highest line of body, the lowest line of body, shoulder and neck) is linked by spline curve. When the perimeter of one part is altered, other parts would be influenced by smoothly spline curve.
An interactive body model
95
Figure 5. Scaling method
IJCST 17,2
96
The vertical scaling method had same scaling ratio (P) and body model is altered based on the lowest line of body model. Result Interactive body model system This dialog represents 3D body models (Figure 6). The interactive body model system constructed with eight slide bars for controlling perimeters and three slide bars for controlling length. Eight slide bars related to SCL. Three slide bars can control length such as from top to waist, from waist to crotch and from crotch to bottom. Especially three length controllers are useful for body model when making skirt, pants and one-piece. Using this system, Figure 7 shows the results for controlling perimeters and length in certain control positions. Comparison of real human body shape with virtual body model shape In order to investigate whether virtual body model can be adequately substituted for real physical models, we compared the perimeter and cross section areas of SCL. Eight numbers of real body data were obtained from the research institute of Human engineering for Quality Life (HQL). A system including adjustment parameters and displaying the virtual body model and the real body model (RBM) side-by-side was developed, thus enabling easy comparison and matching of the virtual model to the real model (Figure 8). Table II shows the results of the comparison of cross section areas ( J(percent)) and perimeter (K(percent)). As a result, J shows over 90 percent agreement at bust, underbust, waist, stomach and hip. The calf and ankle show J over 80 percent because of data loss during 3D measurement. When comparing each, K shows 99 percent agreement rates. This result supports the adaptability and potential usefulness of our body model.
Figure 6. Rendering body model and interactive body model dialog
An interactive body model
97
Figure 7. The result of modified body model shape
Figure 8. The dialog for comparison of the virtual body model and the real body model
Table II. The comparison of J and K with RBM
89.7 86.7 93.0 95.6 95.1 95.2 92.8 93.6
100.0 100.0 99.6 99.9 99.9 99.7 99.9 99.7
K 90.0 91.7 95.5 95.0 95.9 95.0 91.5 93.5
J 100.0 99.1 99.6 100.0 99.9 99.8 100.0 99.9
K 82.3 84.8 92.6 95.5 94.9 95.9 94.6 93.9
J
C 99.5 99.4 99.8 99.9 100.0 100.0 99.7 99.9
K 86.5 84.9 89.6 95.9 97.0 95.6 93.5 94.5
J
D 99.1 100.0 99.6 99.8 99.8 99.7 100.0 100.0
K
J 91.6 92.4 94.0 96.8 95.5 96.3 91.3 94.1
RBM E 99.0 100.0 100.0 100.0 100.0 99.7 100.0 100.0
K
91.2 88.5 92.1 95.9 95.7 94.7 92.9 94.9
J
F 99.1 98.9 100.0 100.0 99.9 99.7 99.9 99.9
K
91.9 84.4 89.1 96.0 97.3 96.5 94.2 94.3
J
G 99.6 98.0 99.5 99.8 99.8 99.6 99.9 100.0
K
82.0 89.5 95.1 93.8 93.7 93.2 93.6 94.1
J
H 100.0 98.7 99.3 99.8 99.6 99.5 99.9 99.9
K
Notes: J(percent): agreement rate of cross section areas to the virtual body model and the real body model; and K(percent): agreement rate of perimeter to the virtual body model and the real body model
Ankle Calf Thigh Hip Stomach Waist Under bust Bust
J
B
98
Parts
A
IJCST 17,2
Conclusion Using 3D body shape data, we developed an interactive 3D body model suitable for pattern making. We presented a cross-sectional line method for body model construction. Also we assigned each SCL which can be altered independently for suitable to accurately real body shapes. Established interactive body model system enables easy and accurate model adjustments. Moreover, from result of investigation of the adaptability and potential usefulness of the virtual body model, there was over 93 percent agreement ratio between the cross section areas and perimeters. It is capable for customers during the buying process to not only modify the body model to match their own body shape, using this system on the internet or catalogs, but also for apparel manufactures to communicate with their customers by describing the body model to fit on the screen while in the ordering process. References Hashizume, S. and Nagata, M. (1999), “Study on the fittability of clothing size”, Sen’I Senihin Shohi Kagaku, Vol. 40 No. 4, pp. 246-54. Horikoshi, T. and Kasahara, H. (1990), “Superquadrics for 3D shape indexing language”, IEICE Transactions, Vol. J73-D-II No. 10, pp. 1716-24. Imaoka, H. (1997), “Human body model represented by meta-balls in apparel industry”, Journal of the Society of Instrument and Control Engineers, Vol. 36 No. 2, pp. 89-94. Inui, S. (2001), “A preliminary study of a deformable body model for clothing simulation”, International Journal of Clothing Science and Technology, Vol. 13 No. 5, pp. 339-50. Jones, P.R.M., Li, P., Brooke-Wavell, K. and West, G.M. (1995), “Format for human body modeling from 3-D body scanning”, International Journal of Clothing Science and Technology, Vol. 7 No. 1, pp. 7-16. Kurokawa, T. (1997), “Measurement and description of human body shape and their applications”, Journal of the Society of Instrument and Control Engineers, Vol. 36 No. 2, pp. 77-83. Minoh, M. (1997), “Methods of generating three-dimensional human shape model”, Journal of the Society of Instrument and Control Engineers, Vol. 36 No. 2, pp. 105-9. Nam, Y. and Han, H. (2001), “Automation human measurement extraction algorithms for the apparel industry”, Journal of the Korean Fiber, Vol. 38 No. 9, pp. 478-86. Stylios, G.K., Han, F. and Wan, T.R. (2001), “A remote online 3-D human measurement and reconstruction approach for virtual wearer trials in global retailing”, International Journal of Clothing Science and Technology, Vol. 13 No. 1, pp. 65-75. Toyoda, H. (1998), “Range scanners for human body and geometric modeling”, Seni’I Gakkaishi, Vol. 54 No. 5, pp. 185-8. Watanabe, Y. (1999), “Ordering your cloth to fit yourself”, The Journal of the Institute of Electronics, Information and Communication Engineers, Vol. 82 No. 4, pp. 404-11. Xu, B., Huang, Y., Yu, W. and Chen, T. (2002), “Body scanning and modeling for custom fit garments”, Journal of Textile and Apparel Technology and Management, Vol. 2 No. 2, pp. 1-11. Yamamoto, H. and Uchiyama, S. (1995), “The Delaunay triangulation for accurate three-dimensional graphic model”, IEICE Transactions, Vol. J78-D-II No. 5, pp. 745-53.
An interactive body model
99
The Emerald Research Register for this journal is available at www.emeraldinsight.com/researchregister
IJCST 17,2
The current issue and full text archive of this journal is available at www.emeraldinsight.com/0955-6222.htm
Prediction of men’s shirt pattern based on 3D body measurements A.P. Chan, J. Fan and W.M. Yu
100 Received August 2004 Accepted November 2004
Institute of Textiles and Clothing, The Hong Kong Polytechnic University, Kowloon, Hong Kong, People’s Republic of China Abstract Purpose – The paper discusses the prediction of shirt patterns for different body sizes using multiple linear regression (MLR). Design/methodology/approach – A total of 29 pattern parameters from men’s tailor-made shirt and 34 body parameters obtained from a body scanner were designed for analysis. MLR has been applied to examine the underlying relationship between shirt pattern parameters and body measurements. Findings – Compared with formulae from the pattern expert, the prediction of shirt pattern from MLR has been improved. Originality/value – The findings could help to predict pattern size with different body sizes more accurately. Keywords Men, Multiple regression analysis, Physical testing, Measurement, Clothing Paper type Research paper
International Journal of Clothing Science and Technology Vol. 17 No. 2, 2005 pp. 100-108 q Emerald Group Publishing Limited 0955-6222 DOI 10.1108/09556220510581245
Introduction One of the problems on garment fitting is the improper apparel sizing. That means the proportion of garments and the existing grading methods are not fit properly to various sizes of customers. This problem may be due to the outdated anthropometric data and the sizes of garments are not appropriate for the full range of variation in body type that exists in the population (Asdown, 1998). Therefore, one of the possible methods is to provide an appropriate sizing of garment and grading rules for producing a range of ready-to-wear apparel by the anthropometric survey. Using the anthroprometric database in terms of various body dimensions for creating optimized sizing will potentially improve garment fit to the population, instead of only evaluating one or two body dimensions of human subject (Asdown, 1998). Physique and body forms are very important in garment design (Iwasaki et al., 1998). Therefore, understanding the physical characteristic of male upper body by Momota and Makabe (1998a, b) and referring to necessary information from anthropometrical survey on each age group of 50-70 years old women by Iwasaki et al. (1998) are also imperative in improving the clothes fitting design. Manley (1997) conducted an anthropometric measurement study to the pregnant women’s protective wear. Fberle et al. (1999) conducted a simple survey for comparing the difference of chest girth and waist girth in different age groups of western adult male. Workman (1991) investigated the body measurements in sizing variation. He especially focused on comparing current standards for size 8 and 10 of the previous decade. The integrated sizing system for The authors would like to acknowledge The Hong Kong Polytechnic University Tuition Scholarship for Research Postgraduate Studies.
men and women was derived mathematically from the US Army’s anthropometric database to solve the problem of size fitting of the female uniform for US Army (Gordon, 1986). Although a lot of comprehensive reports of the anthropometric survey were attempted on western population (Winks, 1998), those surveys might not suit the Chinese people due to different size dimensions between two races. In Hong Kong, Kee (1995) compared the women in China and Hong Kong so as to determine much-fitted garments for them. So far, the updated anthropometric survey, which is conducted for men, is still not available in Hong Kong. On the other hand, a lot of developed grading rules have been widely adopted for a long period of time (Cooklin, 1997). However, it is controversial about the accuracy of those grading rules from using the obsolete anthropometric data. To set-up a huge anthropometric database for designing the garments that fit the wide ranges of variation in body type is time-consuming and uneconomical. The second possible solution is using the prediction model to anticipate the garment sizes from different body dimensions. According to our previous paper (Chan et al., 2003b), there is a remarkable result in prediction of shirt pattern from 3D body measurements by artificial neural network (ANN). In this paper, the underlying relationship between body measurements and garment patterns by multiple linear regression (MLR) will then be discussed. The comparison between tailoring expert and the prediction model from MLR will also be illustrated. Experiment Body parameters The experiment procedures were similar to the previous paper (Chan et al., 2003a), only the number of male subjects has been raised from 19 to 59. Thirty four upper body measurements obtained from Tecmath laser body scanner was designed for analysis and shown in Table I. These 34 upper body measurements might be potentially correlated to the shirt pattern design and were then classified from B1 to B34. Shirt pattern parameters In geometry consideration, 33 shirt pattern parameters were categorized as an important parameter for pattern drafting. These pattern parameters were ranged from G1 to G33 and are shown in Figure 1. Results and discussion The accuracy of the existing drafting formulae According to our previous paper (Chan et al., 2003a), it showed that the correlation of those drafting formulae from four different experts was not appropriate to predict the shirt pattern when different levels of body measurements were presented. The correlation was low whereas the maximum error was high. In this paper, the shirt pattern formulae from tailoring expert Aldrich (1997) was adopted for making comparison. Table II presents the deviation of Aldrich’s formulae and correlation between actual shirt pattern measurements and calculated results from author’s formula.
Prediction of men’s shirt pattern 101
IJCST 17,2
Body parameters
Description
a
Body height Head height Distance neck to hip Neck diameter Mid neck girth Side upper torso length Torso width Total torso girth Cross shoulder over neck Cross shoulder 1/2 shoulder width Shoulder angle L (8) Shoulder angle R (8) Cross front width Breast width Neck R to waist over bust Neck front to waist over bust Bust points around neck Bust point to neck Chest band Midriff girth Cross back width Back width Neck to underarm back Back length Back length over shoulder blade Distance armpit-waist Waist to hip back Waist girth Waist to hip Hip girth Arm length to neck back Arm length to neck Arm length
B1 B2 B3 B4 B5 a B6 B7 a B8 B9 B10 B11 B12 B13 B14 B15 a B16 a B17 a B18 a B19 B20 B21 B22 B23 a B24 a B25 a B26 a B27 a B28 B29 a B30 B31 a B32 a B33 a B34 a a
102
Table I. The classification of body parameters from Tecmath laser body scanner
Note: aVertical body parameters
The highest correlation coefficient R 2 in Table II is 0.8, which are G4 (3-10), G9 (3-15) and G19 (0-15). They are the chest girth, back chest girth and front chest girth, respectively. The percentage of the maximum error divided by the mean of the actual measurements is around 9-12 per cent, the percentage of the deviation in the chest girth level is acceptable. The second high in R 2 is 0.7, which are G11 (8-26), G14 (3-33), G20 (12-40) and G21 (5-39). They are front neck width, nearly 2/3 chest girth, front hip girth and back hip girth, respectively. The percentage of the maximum error divided by the mean of the actual measurements is around 10-15 per cent, it also seems acceptable in these horizontal girth levels. The R 2 of the rest of pattern parameters are below 0.5. Such traditional pattern drafting method for shirt pattern is by mean of the linear relationship from body measurements. The relative high squared correlation coefficient obtained from those pattern parameters showed that the horizontal pattern parameters correlated to body measurements linearly. However, it is still not
Prediction of men’s shirt pattern 103
Figure 1. Pattern parameters
proper to predict the pattern size from various body measurements accurately. Besides, the prediction of other shirt patterns is still inaccurate in this study. Prediction of shirt pattern by ANN In order to improve the results of prediction on pattern parameters from body measurements, a three-layered back-propagation neural network (Fan et al., 2001; Fausett, 1994) has been applied to establish the model. The number of cases was increased from 43 to 143 for ANN training in order to predict the extreme cases more accurately. That means the number of cases of the fattest subjects has been increased from 1 to 50. The number of cases of thinnest subjects has also been enlarged from 1 to 50. From the 143 pairs of data, 133 pairs were used as training pairs and 10 were used as cross-checking pairs so as to ensure that the derived model can provide a generalized mapping of the input and output pattern. Training of ANN stopped when the sum of errors of the checking data set started to increase. Table III lists the model parameters. The ANN model trained based on the above parameters was found to provide flawless prediction of the shirt pattern parameters from the 3D body measurements obtained from the body scanner. Figure 2 shows the predicted pattern parameters and the actual shirt pattern measurements, with the squared correlation coefficient 0.82. Multiple linear regression The ANN model shows that it provides a decent prediction of shirt patterns from 3D body measurements. However, in order to examine the underlying relationship between shirt pattern and body measurements, the MLR has been applied. To select much appropriate and reasonable body parameters from MLR model, both body
IJCST 17,2 Pattern parameter
104
G1 G2 G4 G5 G6 G7 G8 G9 G10 G11 G12 G13 G14 G16 G19 G20 G21 G23 G26 G28
Deviation of Aldrich’s formulae (cm) Mean SDb 95 per cent error Maximum error/ a error (cm) (cm) intervalc (cm) meand (per cent) 9.8 6.2 1.5 1.7 0.8 5.3 2.5 0.9 1.9 0.3 0.6 2.4 1.7 6.3 0.9 1.9 1.1 8.5 3.0 4.1
1.4 4.1 1.4 0.4 0.3 0.6 0.9 0.7 1.5 0.2 0.4 0.8 1.2 1.8 0.8 1.1 0.8 1.6 0.8 1.6
7.2-13 0.1-15.9 0.0-5.7 0.9-2.5 0.2-1.2 4.2-6.2 0.4-4.5 0.0-2.5 0.1-7.6 0.0-0.6 0.0-1.3 0.7-3.9 0.1-5.3 0.1-9.1 0.1-3.3 0.0-4.0 0.0-3.0 5.7-11.9 1.1-4.6 1.2-7.5
44.3 33.0 10.4 29.0 22.6 80.0 19.2 9.0 68.3 10.2 26.7 23.4 15.9 39.0 12.0 14.5 11.0 56.5 18.5 60.5
Correlation coefficient among actual shirt pattern measurements and calculated results from author’s formula (R 2) 0.2 0.0 0.8 0.3 0.0 0.0 0.5 0.8 0.0 0.7 0.4 0.4 0.7 0.0 0.8 0.7 0.7 0.0 0.5 0.0
Table II. The deviation of Aldrich’s formulae and squared correlation coefficient between the actual shirt pattern measurements and calculated results from author’s formula
Notes: aMean error ¼ (actual measurement – calculated result)/number of samples; bSD is the standard deviation; c95 per cent error interval the maximum and minimum deviation based on 95 per cent confident limit; and dMaximum error/mean the maximum deviation/mean of the actual measurements
Table III. ANN model parameters
Activation function at the output layer Activation function at the hidden layer Input parameters Output parameters Number of hidden units Number of training pairs Number of cross-checking pairs
f(x) ¼ 1/(1 þ e2 x) f(x) ¼ 1/(1 þ e2 x) 34 body measurements 33 shirt pattern parameters 80 50 9
parameters and pattern parameters have been assembled into two categories. They were horizontal parameters and vertical parameters, respectively. Table IV presents the corresponding body parameters at each pattern parameter, the equations derived from MLR and its squared correlation coefficient. Referring back to Table II and Figure 1, the Aldrich’s pattern drafting formulae do not include the pattern parameters G3, 15, 17, 18, 22, 24, 25, 27, 29, 30, 31, 32 and 33. In geometry consideration, those parameters are important for pattern drafting. MLR model can predict those pattern parameters with the highest squared correlation coefficient 0.9 and the lowest squared correlation coefficient 0.3. The lowest squared
correlation coefficient at pattern parameters G24, 25, 32 and 33 are the scye depth of front, scye depth of back, front neck to waist and dart length of back shoulder, respectively. There is no direct body measurement at those locations from body scanner. Therefore, the accuracy of prediction from MLR model might be affected. In horizontal pattern parameters, G4 (3-10), G9 (3-15) and G19 (10-15) with the highest squared correlation coefficient 0.9, are the half chest girth, back chest girth and front chest girth, respectively. The correlated body parameters in G4 are B21 (midriff girth), B31 (hip girth), B15 (breast width) and B22 (cross back width). The correlated body parameters in G9 and G19 are B21 (midriff girth), B31 (hip girth), B15 (breast width) and B22 (cross back width). It proves that the prediction results are satisfying in those girth measurements such as chest, waist and hip comparing with traditional formulae. There is no doubt that the body measurements are proportionally correlated to pattern measurements. However, it proves that the relationship between body measurements and shirt pattern is not only correlated with single body parameter, but to several body parameters. Actually, human body is not in circular shape. Therefore, the proportion of girth measurements with constant amount of ease allowance is not accurate enough to determine the pattern size, which fit the customers. The squared correlation coefficient of G23 (angle 27,28,29) is 0.3 and G28 (angle 20,19,21) is 0.4. G23 and G28 are the front shoulder angle and back shoulder angle, respectively. The correlated body parameter in G23 is B12 (left shoulder angle). The correlated body parameter in G28 is B4 (neck diameter). The relative low squared correlation coefficient at these two shoulder angles might be due to the traditional experience of shirt tailors. Tailors do not measure the exact shoulder angle of wearers in practice, but only observe the shoulder slopes of those wearers by their experience. In vertical pattern parameters, G29 (41-43) is the sleeve length with the highest squared correlation coefficient 0.9. The correlated body parameters in G29 are B33 (arm length to neck) and B28 (waist to hip back).
Prediction of men’s shirt pattern 105
Comparison between the existing drafting formulae and prediction from MLR Comparing the squared correlation coefficient between Aldrich’s formulae and prediction from MLR model with actual shirt pattern measurements, all the values
Figure 2. The prediction value of shirt pattern vs the actually shirt pattern measurements
IJCST 17,2
Pattern parameter a
G1 G2 a G3 G4 G5 a G6 a G7 G8 G9 a G10 G11 a G12 G13 G14 a G15 a G16 a G17 a G18 G19 G20 G21 G22 G23 a G24 a G25 G26 G27 G28 a G29 G30 G31 a G32 a G33 a
106
Table IV. Equations of shirt pattern parameters from MLR
Equation from MLR 22.95920.194B6þ 0.06659B8 22.613þ0.215B8þ0.08599B3020.36B2720.253B28 28.063þ0.131B30þ 0.131B8þ 0.304B32 3.461þ0.198B21þ 0.172B31þ 0.233B15þ0.274B22 3.829þ0.06175B9þ0.05957B5 4.73520.0259B17þ 0.09095B332 0.0844B34 8.62þ0.04054B302 0.0792B17 8.165þ0.0566B31þ 0.163B23þ 0.05576B13þ0.07124B14 1.59þ0.09889B21þ 0.09377B31þ 0.114B15þ0.123B22 10.99220.224B17þ 0.05944B30 0.107þ0.166B5 20.597 2 0.0338B30þ 0.203B5þ0.06355B172 0.0321B33 9.699þ0.185B22 5.059þ0.341B21 4.871þ0.167B172 0.0265B302 0.0967B27 9.211þ0.09652B8þ0.153B62 0.239B3þ 0.116B32 20.483þ0.217B8þ0.212B17þ 0.39B34þ 0.06097B3020.279B3 3.059þ0.193B172 0.0246B302 0.093B27 2.413þ0.108B21þ 0.0894B31þ 0.109B15þ0.9646B22 5.036þ0.143B31þ 0.109B21 5.752þ0.118B21þ 0.126B31 0.119þ0.224B520.151B11þ0.02429B20 24.139þ0.179B122 0.0838B31 7.549þ0.323B28þ 0.157B620.0683B18 7.694þ0.12B28þ 0.756B332 0.628B32 12.417þ0.181B22þ 0.05971B31 6.291þ0.172B920.0195B20 24.17820.448B42 0.104B31þ 0.11B14 217.007 þ 0.569B33þ 0.08707B3020.122B17þ0.113B1 6.06þ0.242B29 5.491þ0.248B29 24.749þ0.1B30þ 0.086B8 4.91220.106B6
Correlation coefficient (R 2) 0.3 0.6 0.7 0.9 0.4 0.7 0.7 0.7 0.9 0.6 0.6 0.7 0.3 0.8 0.5 0.4 0.7 0.6 0.9 0.8 0.8 0.7 0.3 0.4 0.4 0.5 0.5 0.6 0.9 0.8 0.8 0.4 0.3
Note: aVertical parameters
have significant improvement on prediction. Table V shows the changes of the squared correlation coefficient among those pattern parameters. According to the highest squared correlation coefficient in Aldrich’s formulae, R 2 of G9 and G19 have been increased from 0.8 to 0.9. The lowest squared correlation coefficient in Aldrich’s formulae is also increased from 0.0 to at most 0.7 and to at least 0.3. Conclusion Anthropometric survey is one of the feasible solutions to develop the relationship between pattern sizes and body dimensions. However, to conduct a huge anthropometric database is considered to be time-consuming and uneconomical. Therefore, prediction on pattern sizes from various body measurements by ANN is
Pattern parameter G1 G2 G4 G5 G6 G7 G8 G9 G10 G11 G12 G13 G14 G16 G19 G20 G21 G23 G26 G28
Correlation coefficient among actual shirt pattern measurements and calculated results from author’s formula (R 2)
Correlation coefficient prediction from MLR model (R 2)
0.2 0.0 0.8 0.3 0.0 0.0 0.5 0.8 0.0 0.7 0.4 0.4 0.7 0.0 0.8 0.7 0.7 0.0 0.5 0.0
0.6 0.5 0.9 0.7 0.7 0.7 0.7 0.9 0.8 0.7 0.6 0.5 0.8 0.5 0.9 0.9 0.9 0.3 0.7 0.4
much appropriate for pattern design. The results show that the underlying relationship between the shirt patterns and body measurements can be examined by the MLR model. When comparing to the tailoring expert, MLR model is found better in prediction on shirt pattern from different body measurements. References Aldrich, W. (1997), Metric Pattern Cutting for Menswear, 3rd ed., pp. 30-1. Asdown, S.P. (1998), “An investigation of the structure of sizing systems: a comparison of three multidimensional optimized sizing systems generated from anthropometric data with the ASTM Standard D5585-94”, International Journal of Clothing Science and Technology, Vol. 10 No. 5, pp. 324-41. Chan, A.P., Fan, J. and Yu, W. (2003a), “Men’s shirt pattern design. Part I: An experimental evaluation of shirt pattern drafting methods”, Sen’I Gakkaishi, Vol. 59 No. 8, pp. 319-27. Chan, A.P., Fan, J. and Yu, W. (2003b), “Men’s shirt pattern design. Part II: Prediction of pattern parameters from 3D body measurements”, Sen’I Gakkaishi, Vol. 59 No. 8, pp. 328-33. Cooklin, G. (1997), “A new approach to ‘making the grade’”, Apparel International, pp. 10-11. Fan, J., Newton, E., Au, R. and Chan, S.C.F. (2001), “Predicting garment drape with a fuzzy-neural network”, Textile Research Journal, Vol. 71 No. 7, pp. 605-8. Fausett, L. (1994), Fundamentals of Neural Networks: Architectures, Algorithms and Applications, Prentice-Hall International, New York, NY. Fberle, H., Hermeling, H., Hornberger, M., Menzer, D. and Ring, E. (1999), Clothing Technology, pp. 192-3.
Prediction of men’s shirt pattern 107
Table V. Comparison between Aldrich’s formulae and prediction from MLR model in terms of squared correlation coefficient
IJCST 17,2
108
Gordon, C.C. (1986), “Anthropometric sizing and fit testing of a single battledress uniform for US army men and women”, Performance of Protective Clothing, pp. 581-92. Iwasaki, K., Miyoshi, M., Hirokawa, T., Saito, K. and Isozaki, A. (1998), “Variation of body forms of middle and old aged women. Part 1: Comparison of body forms of middle and old aged women according to discriminant analysis”, Journal of the Japan Research Association for Textile End-Uses, Vol. 39 No. 5, pp. 318-26. Kee, L. (1995), “Fitting factors”, Proceedings of the 3rd Asian Textile Conference, April, pp. 743-9. Manley, J.W. (1997), “Protective clothing fitting considerations for pregnant women”, Performance of Protective Clothing, Vol. 6, pp. 233-302. Momota, H. and Makabe, H. (1998a), “Conditions of fit in business suits of Japanese adult males”, Journal of the Japan Research Association for Textile End-Uses, Vol. 39 No. 7, pp. 452-61. Momota, H. and Makabe, H. (1998b), “Physical type characteristics of the upper half body of Japanese adult males. Part 1: Examination by body measurement and developed plane figures”, Journal of the Japan Research Association for Textile End-Uses, Vol. 39, pp. 382-91. Winks, J. (1998), “Finding a universal body language: the right fit for clothing standards”, ISO Bulletin, pp. 3-6. Workman, F.E. (1991), “Body measurement specifications for fit models as a factor in clothing size variation”, Clothing and Textile Research Journal, Vol. 10 No. 1, pp. 31-6.
The Emerald Research Register for this journal is available at www.emeraldinsight.com/researchregister
The current issue and full text archive of this journal is available at www.emeraldinsight.com/0955-6222.htm
Evaluation of drape characteristics in fabrics
Evaluation of drape characteristics
Narahari Kenkare and Traci May-Plumlee College of Textiles, North Carolina State University, Raleigh, North Carolina, USA
109
Abstract Purpose – To provide researchers with the details of developments in instruments to measure fabric drape and review the literature related to fabric drape. Design/methodology/approach – In recent years, there has been a renewed interest in investigating the aesthetic behavior of fabrics due to the developments in objective evaluation techniques. To understand drape behavior, it is essential to know how drape is measured quantitatively. This paper reviews research related to drape characteristics of fabrics, two-dimensional instruments and analysis of drape by measuring stiffness, three-dimensional instruments developed to measure drape, fabric mechanical properties and their influence on drape measurement, and the latest developments in the field including image analysis, the dynamic drape tester and other related research. Findings – Many instruments for measuring drape have been developed including the earliest that assessed stiffness of fabrics, later versions of drape meters and recent innovative instruments for capturing complex drape information. Even though extensive detail for simple geometric forms such as circles and squares can be provided by the newest methods, measurement of the drape characteristics of complex forms needs the consideration of researchers to extend the work on drape measurement to garments. It was also noted that there are some contradictory conclusions regarding the properties influencing fabric drape. Originality/value – This paper is offered as a concise reference for individuals beginning research in the area of fabric drape. Keywords Fabric testing, Measurement, Research, Innovation, Garment industry Paper type General review
Introduction Fabric drape is a very important mechanical property due to its influence on the appearance of clothing. Drape is defined as “the extent to which a fabric will deform when it is allowed to hang under its own weight” (BS 5058: 1973, 1974). Drape is an important factor that affects the aesthetics and dynamic functionality of fabrics determining the adjustment of clothing to the human silhouette and providing the description of the fabric deformation produced by gravity when the fabric is partially supported. This unique characteristic provides a sense of fullness and a graceful appearance, which distinguishes fabrics from other sheet materials. When a fabric is draped; it can bend in one or more directions. Curtains and drapes usually bend in one direction, whereas garments and upholstery exhibit a complex three-dimensional form with double curvature. Hence, fabric drape is a complex mathematical problem involving large deformations under low stresses (Postle and Postle, 1993). Initially, fabric is draped rapidly based on its weight overcoming the resistance associated with its stiffness, after which the stiffness of the fabric structure resists further deformation. Fabrics may drape in dramatically different ways, depending on the fiber content, type of yarn, fabric structure and type of finish.
International Journal of Clothing Science and Technology Vol. 17 No. 2, 2005 pp. 109-123 q Emerald Group Publishing Limited 0955-6222 DOI 10.1108/09556220510581254
IJCST 17,2
110
There have been numerous instruments, ranging from a simple cantilever bending tester to a dynamic drape tester developed for measuring fabric drape. “Drape co efficient (DC)” the main parameter used to quantify fabric drape. Though useful, it is insufficient to characterize complex forms such as garments. This paper reviews research in the area of fabric drape. A historical perspective on drape, techniques and instruments used to measure drape, and the scope of recent developments are discussed. Historical perspective The scope of drape and related research presents challenges and opportunities to textile engineers and researchers. To provide the background for further development, it is essential to interpret the drape related work of early researchers. The textile and clothing industries have traditionally used subjective assessments by individual judges or by methods dependent on some manual input as the basis of fabric drape evaluation. Such evaluation methods usually involve placement of a circular fabric specimen on a pedestal or placement of a simple garment (usually a skirt) on a dress form for evaluators to view. The evaluators compare subjectively to yield ranks in terms of draped fabric appearance, or evaluate by paired comparisons. It has been shown (Collier and Epps, 1999) that, although a panel of judges has a reasonably common notion or concept of desirable fabric features, the ability of any individual judge to assess the drape of a particular fabric is subject to large errors in the form of human bias and inconsistency. Studies on subjective evaluation of drape have found that drape preferences are influenced by prevailing fashion. A study done in the 1960s, when the apparel styles were stiffer and more geometric, found that evaluators preferred less drapeable fabrics. Later in the 1990s, very drapeable fabrics suitable for the fluid lines popular in apparel at that time were rated more highly by evaluators. Subjective evaluation of fabric drape can provide some understanding of human perception and fashion trends, but the results are inconsistent because of personal preference and fashion changes. Thus, it is generally accepted that subjective methods can be unreliable, and the development of more reliable and consistent systems for objective evaluation of drape became essential. Development of instruments for drape measurement The literature suggests (Postle, 1998) that the mechanical properties of fabrics were first studied during the late 19th century by German researchers working on developing airships. The study of drape behavior of fabrics dates back to the classic paper published in Journal of the Textile Institute, entitled The handle of cloth as a measurable quantity by Pierce (1930). In this paper, measures of fabric stiffness are used to obtain the flexural rigidity of fabrics. Flexural rigidity is a simple calculation of bending length and weight in combination, and is highly dependent on fabric thickness. So, Pierce (1930) converted flexural rigidity into bending modulus that takes thickness into consideration. Pierce developed the “cantilever method” for measurement of fabric bending properties and used bending as a measure of fabric drape. The Pierce Cantilever Test is performed on the commercially available Shirley stiffness tester (Figure 1(a)). The principle of the cantilever stiffness test is shown in Figure 1(b). A rectangular strip of fabric is allowed to bend under its own weight to a fixed angle when projected as a
Evaluation of drape characteristics 111 Figure 1. (a) Shirley stiffness tester; and (b) principle of cantilever stiffness test
cantilever. The fabric sample in the instrument represents a cantilever beam that is uniformly loaded by its own weight and bends downward until it reaches 41.58. The longer the projected length, the stiffer the fabric (Booth, 1969). To measure bending length, a fabric sample is placed under the metal ruler (Figure 1(a)) and slowly moved forward until the tip of the specimen viewed in the mirror cuts both the index lines, then bending length is read. The mean value for the bending length is calculated by testing the specimen four times on each end and again with the strip turned over. Using the average value of bending length (C), Pierce calculated flexural rigidity (G) and bending modulus (q). From the length (l) and angle (u) bending length (C) can be calculated using the formula: C ¼ lf 1 ðuÞ and 1 13 cos 2 u f 1 ð uÞ ¼ 8 tan u Flexural rigidity of the fabric can be calculated from the bending length (C) and weight using the formula: G ¼ 3:39w1 C 3 mg=cm or w2 £ 103 mg=cm where w1 is the cloth weight in ounces per square yard, and w2 the cloth weight in grams per square yard. Bending modulus (q) is calculated as: q¼
732G kg=cm2 g31
IJCST 17,2
112
or 12G £ 1026 kg=cm2 g 32 where g1 is the cloth thickness in thousands of an inch, and g2 the cloth thickness in centimeters. This method is still considered the standard to measure two-dimensional bending of fabrics. Research in this area was continued by several researchers. Behre (1961), Dahlberg (1961), Grosberg (1966, 1971, 1973), Lindberg et al. (1960, 1961) and Oloffson (1964) reported the nonlinear nature of bending and shear properties. This work contributed to the understanding of drape, but two-dimensional drape assessment cannot capture the complex three-dimensional double curvature deformations of 3D drape. Hearle and Amirbayat (1986) conducted a theoretical investigation of drape by using a different approach to study complex buckling. They found that the geometric form of deformation can be related to two dimensionless energy groups J1 and J2, which relate bending, membrane, and potential energies, and are defined in terms of sheet parameters and size. Their experiments showed that drape is not only a function of J1 and J2, but must also be influenced by other parameters such as the full set of anisotropic in-plane membrane and out-of-plane bending and cross term elastic constants, and perhaps the nonlinearity of response. Niwa and Seto (1986) published a paper about the relationship between drapeability and mechanical properties. They used mechanical parameters in combinations [(B/W)1/3, (2HB/W)1/3, (G/W)1/3 and (2HG/W)1/3] as independent variables, where B, 2HB, W, G, and 2HG are bending rigidity, bending hysteresis, weight per unit area, shear stiffness, and shear hysteresis, respectively. They derived these parameter combinations by applying the heavy elastica theory to analysis of bending of a fabric cantilever with hysteresis in bending and shear. They then obtained an equation to describe DC. Sudnik (1972) correlated fabric drapeability with bending length, and in 1978 (Sudnik, 1978), confirmed the importance of bending length in predicting fabric drape. He determined that shear resistance was also a factor, although not as important. However, research by Collier (1991) found shear properties to be more significant predictors of fabric drape than bending properties and shear hysteresis, and that shear hysteresis was closely related to DC. Other researchers who have contributed to this area include Morooka and Niwa (1976), Gaucher et al. (1983), Collier and Collier (1990) and Collier (1991). Vangheluwe and Kiekens (1993), Jeong (1998), and Jeong and Philips (1998), reported a relationship between fabric drape and the values of mechanical properties. Measurement of bending length A number of stiffness testing instruments have been designed, developed and studied. Table I lists major instruments used to measure the stiffness of fabric, sample size for measurement, and a brief description of each instrument. Further details on measurement procedures can be found in the references provided. Abbott (1951) compared results from subjective evaluation of stiffness with five laboratory methods of measuring stiffness. Analysis of the results indicates that the Pierce cantilever test provided the best correlation, and four of the five instrument methods provided significant correlation, with the subjective evaluation.
Instrument
Sample size
Description
Peirce cantilever tester (Shirley stiffness tester)
6 in. £ 1 in.: 3 samples each way and each sample tested both sides
The sample is allowed to project as a cantilever from a horizontal platform and the angle between the horizontal and the chord from the angle between the horizontal and the chord from the edge of the platform to the tip of the fabric is measured (Peirce, 1930; Booth, 1969) The test method consists of bending the ends of a 1 in. wide strip through 5408, bringing the ends together forming heart loop The length of the loop is measured under the force of gravity (Winn and Schwarz, 1939) The instrument measures the amount of work required to fold a pair of samples The samples are mounted between a fixed and a moving plate in such a way as to form a couple opposing the moving plate The amount of work required to fold the samples to a minimum angle is calculated and is considered as stiffness measurement (Winn and Schwarz, 1940) The sample is mounted in vertical position and measures the force in milligrams needed to bend the strip through a sufficient angle to cause slippage of the specimen over a vane The force is produced by the deflection of a weighted pendulum-type vane that slides over the surface of the specimen as the supporting arm is rotated alternately to each side The reading is obtained from the instrument and is multiplied by a factor to give the stiffness in milligrams (Winn and Schwarz, 1940) The sample is mounted in a frame which permits lateral displacement of one end of the fabric The movement of the fabric distorts the fabric, and is carried on until diagonal wrinkles appear The angle of the frame at the appearance of the wrinkles denotes the measure of stiffness (Dreby, 1942) The instrument is a beam bending type that involves bending of a fabric through an angle of 608 and recording the load applied at the bending plate
The Heart loop tester
8 in. £ 1 in.: 3 samples each way and each sample tested both sides
Schiefer flexometer
1 3/4 in. £ 4 in.: 2 pair of samples cut each way
The Gurley stiffness tester
1/2 in., 1 in., or 2 in. wide; 1, 1 1/2 in., 2 1/2 in., 3 1/2 in. or 4 in. long
Planoflex
3 in. £ 10 in.: 5 samples cut each way
Olsen stiffness tester
2 in. £ 1 in.
(continued)
Evaluation of drape characteristics 113
Table I. List of instruments measuring stiffness
IJCST 17,2
Instrument
Sample size
MIT drape meter
114
Table I.
4.7 in. £ 10 in.: 3 samples cut each way and each samples tested both sides
Description The stiffness is calculated in terms of bending moment (Hynek and Winston, 1953) The instrument designed to measure drape rather than stiffness It was designed to correlate with the other test methods to measure stiffness One end of the sample is strapped to the edge of a circular disc and the fabric is allowed to hang vertically. The curvature of the fabric at any given depth depends on the fabric stiffness and several other parameters The chord length below the support was considered as the measurement, (Winn and Schwarz, 1941)
The instruments listed in Table I are predominantly those measuring two-dimensional distortions of fabrics. These instruments provided an acceptable correlation in terms of measuring stiffness, but due to two-dimensional testing, they are incapable of differentiating fabrics from papers having the same bending length. The inability to discriminate between a piece of paper and a piece of fabric having the same stiffness value was the major disadvantage of characterizing drape using two-dimensional distortions. Development of drapemeter A three-dimensional representation was necessary for making appropriate evaluation of drape characteristics quantitatively. To overcome the limitations of using two-dimensional measurement of stiffness as the estimating parameter of drape, researchers in Fabric Research Laboratories developed the F.R.L Drapemeter (Chu et al., 1950). Later Cusick (1968), developed a drapemeter (Figure 2) based on similar principles. By developing drapemeters, Chu et al. (1950) and Cusick (1961, 1965, 1968) made significant contribution to the practical determination of this fabric property by measuring three-dimensional drape. Plate 1 shows a circular specimen of fabric about 36 cm in diameter supported on a circular disk of 18 cm diameter on the drapemeter. The unsupported area drapes over the edges of the support disk forming the drape configuration of the fabric specimen. The drape coefficient (DC) defined as the fraction of the area of the annular ring covered by the projection of the draped sample, is used to quantify the fabric drape. The DC can provide an objective description of deformation, although not a complete one. A low drape coefficient indicates easy deformation of a fabric and a high DC indicates less deformation. DC ¼
Area under the draped sample 2 Area of support disk Area of the specimen 2 Area of support disk
Evaluation of drape characteristics 115
Figure 2. Cusick’s drapemeter
Plate 1. Drape configuration of fabric on drapemeter
Cusick (1968) introduced a simpler method of calculating DC. In this method, “A circular piece of paper, of radius R, is placed under the center of the tester. The perimeter of the shadow of the draped fabric is then drawn on the paper. The circle of paper is folded and weighed to give W1. The paper is then cut along the perimeter of the shadow, and the paper in the shape of the shadow of the area A is weighed to give W2” (p. 258). DC is expressed as the ratio of W1 and W2. In Cusick’s original work, he measured the DC, bending length, and shear stiffness (the shear angle at which a fabric begins to buckle; fabrics with higher shear stiffness buckle at lower angles). He found that theoretical drape coefficients, when ignoring shear stiffness, were lower than the measured values for the majority of fabrics tested. In his later work, Cusick (1965) used statistical evidence to prove that fabric drape involves curvature in more than one direction, and that the deformation depends on both shear stiffness and bending length. He also formulated an equation for the relationship between DC, bending length and shear angle. Collier et al. (1988) developed a modified version of a drapemeter. The instrument utilizes a bottom surface of photovoltaic cells to determine the amount of light blocked by a fabric specimen draped on a pedestal. The amount of light absorbed by photovoltaic cells is displayed digitally relating the amount of drape of the fabric. Hoffman and Beste (1951) observed that elastic modulus, the lateral dimensions and cross-sectional shape are the three basic fiber properties influencing fabric stiffness. They categorize the fabric stiffness as high, medium and low based on denier and
IJCST 17,2
modulus data from tested fabrics. Their work can be related to the results from Okur and Cihan’s (2002) paper that suggests fabrics having low values of shear and bending stiffness will have low DCs. Frydrych et al. (2003) investigated the influence of finish on selected aesthetic and utility properties. They found that elastomeric finishes had more influence on DC than the starch finish, and that the DC value of finished fabrics was higher than raw fabrics.
116 Recent developments in drape research Drape research has attracted the attention of computer engineers along with textile engineers in the past few decades. This is due to the increasing application in designing, product development, and e-commerce. The research not only has application in the apparel industry but also has in the movie – animation industry. Recent animation movies (Pixar’s Geri’s game, Disney’s Fantasia) provide insight into the impact of drape and drape simulation technologies on improved representation of the characters. Image analysis One recent significant development in the area of drape measurement is the use of an image analysis technique. Typically the image analysis setup (Figure 3) consists of a drapemeter, a digital camera to capture the draped image of the mounted fabric sample, and a computer to analyze the captured image and translate it into appropriate output. Vangheluwe and Kiekens (1993) proved that there is no significant variation in the drape coefficient calculated using conventional and image analysis methods. The results of their study also show an exponential decrease in DC over time with time of up to 300 s. Ruckman et al. (1998) integrated Cusick’s drapemeter principle with the image analysis technique to measure static and dynamic DC of fabrics. The author calls this system static and dynamic drape automatic measuring system (SDDAMS). The DC is calculated by using the number of pixels covered by the fabric samples, determining the fabric area. Robson and Long (2000) evaluated drape by automatic characterization of drape profiles using an image analysis technique. The study establishes a strong correlation
Figure 3. Diagram of image analysis system for the measurement of drapeability
between the traditional cut and weigh method of calculating the DC and the image analysis method. It is noted in the study that the three parameter combination of drape coefficient, number of nodes, mean node severity (a height/width gradient measure), and variation in node severity (standard deviation of gradient) is required for reasonable discrimination of drape profiles of fabrics.
Evaluation of drape characteristics
Dynamic drapemeter Static draping behavior of fabrics was discussed in the previous sections and by several authors. There have been very few researchers involved in investigation of dynamic drape behavior of fabric. Matsudaira, M., Yang, M., Qin, L., and Yan, M., of Kanazawa University, Japan, discuss dynamic drape in a series of papers published in the Journal of Textile Machinery Society of Japan. The researchers found that there existed an inherent node number for any fabric, and a conventional static DC, Ds, for a fabric could be measured with high accuracy and reproducibility by an image analysis system (Matsudaira and Yang, 1997). Then, Yang and Matsudaira (1998a) derived a regression equation for the coefficient for both isotropic and anisotropic fabrics. Further, Yang and Matsudaira (1998b) analyzed the effect of basic mechanical parameters of fabrics on static drape shapes using computer simulation. In papers by Yang and Matsudaira (1999, 2000), the researchers define the revolving DC (Dr). The revolution of a fabric sample increases drape coefficient. The revolving DC at 200 rpm (D200), which is the saturated spreading of fabrics at rapid revolution, was also analyzed using regression analysis. In the final paper of the series, Yang and Matsudaira (2001) defined the dynamic DC with swinging motion (Dd), that is considered to be similar to the human body’s motion when walking. They derived a regression equation from the basic mechanical parameters of fabrics. The testing instrument built to measure dynamic drapeability of fabric included a circular rotating bidirectional supporting stand. A sample is rotated at 100 rpm for 1 min, then, two-dimensional images are captured continuously at 1/30 s time intervals by an image analysis system. To determine DC, an image taken by the charge-coupled device camera (CCD) is put on a 512 £ 512 dots picture plane and the projected area is calculated. From the measured area the dynamic DC is calculated. The dynamic DC (Dd) is the change in draping shape of fabrics at a swinging motion. If Dd is large, the draping shape is changed easily by a small force such as light wind or the swinging motion of human body. 2 3 ðS max 2 S min Þ 5 £ 100ðpercentÞ D d ¼ 4 pR21 2 pR20
117
where Dd is the dynamic DC, Smax the maximum projected area at the turn-round angle, Smin the minimum projected area at the turn-round angle, R0 the radius of the circular supporting stand, and R1 the radius of fabric sample. Measuring parameters of fabric drape Fabric drape is a 3D phenomenon conventionally measured by determining DC using a drapemeter. However, DC is insufficient to completely describe fabric drape, and two fabrics sharing the same DC may have differing draped shapes. So, other parameters
IJCST 17,2
118
such as number of nodes, node dimensions, drape distance ratio, and drape profile values have been used to explain fabric drape. Chu et al. (1963) concluded that a drape diagram provides three parameters; the area of the sample, the number of nodes, and the shape of the nodes, and further that node number is directly related to the DC. Lo et al. (2002) predicted DC, node location, node number and node shape using a model predicting fabric drape profile using polar coordinates. The study also suggested that better prediction of drape profile was achieved by considering mean values of warp, weft and 458 bias directions rather than considering only warp and weft directions. The work highlights the finding that bending and shear hysteresis have strong relationships with the drape profile when analyzed using stepwise regression. Jeong (1998) proposed “drape distance ratio” (Figure 4) as an alternative to DC. The drape distance ratio increases as fabrics become flexible, which is reverse of DC which decreases as fabrics become flexible.
ðr f 2 r ad Þ £ 100 Rd ¼ ðr f 2 r d Þ where rf is the radius of fabric before being draped; rd the radius of disk of drapemeter; rad the average distance to edge of draped fabric; ri and ui the radius and angle at ith point; and Rd the drape-distance ratio: Robson and Long (2000) measured drape profile circularity (CIRC) along with drape coefficient using an image analysis method. Central circle and drape profile area components were analyzed using a particle analysis function. The final image quantified these two components enabling measurement of DC and CIRC. The value of CIRC ranges from 0 to 1, the value of 1 for the perfect circle and value toward zero for complex profiles.
Figure 4. Diagram showing parameters of drape-distance ratio
Evaluation of drape characteristics
CIRC is calculated using the formula: CIRC ¼ 4p
A P2
where A is the area enclosed within the profile outline, and P the length of the outline. Robson and Long (2000) also measured “number of nodes”, “mean node severity” (which is a height/width gradient measure) and “variability of node severity” (i.e. standard deviation of gradient). Fabric properties and drape Objective measurement provides an important set of instrumentally measurable parameters required to specify the fabric quality, tailorability and clothing performance providing a “fingerprint” of a fabric. In objective methods, low stress mechanical and surface properties such as fabric bending, extension, shear, compression, surface friction, relaxation shrinkage and hygral expansion are measured using scientific instruments. There are primarily two systems for evaluating low stress mechanical properties of fabrics. The first is the Kawabata evaluation system (KES) that is very elaborate and accurate in predicting feel, hand and appearance of fabrics. The other is fabric assurance by simple testing (FAST), that is simple and adequately predicts tailorability, though it does not directly predict or define fabric hand. Initially KES was restricted to research applications, but later it was widely used in industry for quality assurance and performance prediction. Hu and Chan (1998) considered the relationship between the fabric drape from the Cusick drapemeter and 16 values from the KES. They used four mathematical models (equations (1)-(4)) based on earlier studies to evaluate the influence of the Kawabata parameters on DC. DC ¼ b0 þ
n X
bi x i
ð1Þ
i¼1
rffiffiffiffiffi rffiffiffiffiffiffiffiffiffi rffiffiffiffiffi rffiffiffiffiffiffiffiffiffi 3 G 3 2HG 3 B 3 2HB DC ¼ b0 þ b1 þ b2 þ b3 þ b4 W W W W DC ¼ b0 þ
n X
bi ln xi
ð2Þ
ð3Þ
i¼1
ln DC ¼ b0 þ
n X
bi ln xi
ð4Þ
i¼1
where b0 and bi are arbitrary constants, n the number of parameters closely related to the DC, xi represents a mechanical property parameter, and DC the drape coefficient. Equation (1) is a simple regression equation. Equation (2) is the derived equation for drape coefficient from Niwa and Seto’s (1986) research. Equation (3) is the extension of equation (1) with logarithmic transformation of mechanical properties. Equation (4) is the logarithmic expression of both mechanical properties and DC. Three to four
119
IJCST 17,2
120
parameters from KES were considered in deriving the multiple regression equation. The researchers found that equations (3) and (4) had regression coefficients around 96 percent compared to 89 percent for equations (1) and (2). The results suggested that “mean deviation of friction coefficient (MMD)” and “tensile linearity (LT)” from the KES, in addition to bending and shear properties, related closely to DC. Okur and Cihan (2002) analyzed the relationship between DC from Cusick’s drapemeter and mechanical properties measured using the FAST system for women’s woven suiting fabrics. The authors proposed a multiple regression model (same as equation (1)) based on stepwise analysis. The results showed significant correlation between DC and shear stiffness, and also suggested high correlation between DC, bending properties and extension at a 458 bias angle. They also found that parameters related to compression characteristics measured on FAST 1 (compression test) had no significant correlation with the DC. Along with the physical and mechanical properties, the direction of fabric also influences garment drape. When draping circular fabric samples on a circular pedestal, the typical method of assessing drape, the draped samples are considered to be axisymmetric. But fabrics are not axisymmetric in that they behave differently in warp and weft directions. Fabric anisotropy is one of the parameters influencing the drape of fabrics. Sidabraites and Masteikaite (2003) studied the effect of anisotropic behavior of woven fabrics on drape. They predicted the bending rigidity in 12 different directions of the fabric. “Distances to the edge of a draped fabric profile in twenty-four different directions were measured in order to get polar diagrams comparable with polar diagrams of bending rigidity” (p. 111) eventually predicting the drape profile of lightweight woven fabrics. Analysis resulted in good correlation values for average DCs and average distances to the edge of the draped profile between the experimental method and the model. The researchers concluded that bending rigidity in 12 directions and bending rigidity in warp and weft directions were good predictors of drape performance as measured with DC and drape profile. Orzada (1999) studied the effect of grain alignment in eight directions on fabric properties measured using the FAST system. Orzada observed increases in bending length, bending rigidity, formability, dimensional stability, and extension with increases in tilt angle of the fabrics. The results show that these increases in the FAST parameters were not significant suggesting that the Kawabata system could be more sensitive to variations resulting from grain alignment in fabrics. Conclusions Several researchers beginning with Pierce (1930) have contributed to the field of drape measurement in the form of designing an instrument, studying the parameters influencing the drape, or by defining the relationship between the mechanical properties and drape. Drape characteristics are influenced by complex interactions between varying factors of bending, shear, fabric history, operating conditions, and finishing. Based on previous studies (Hu and Chan, 1998; Sudnik, 1972; Morooka and Niwa, 1976) it can be noted that there are some contradictory conclusions regarding the properties influencing fabric drape. This may be due to the fact that the kinds of fabrics used in the studies were quite different. It can also be noted that researchers have concentrated primarily on the relationship between drape coefficient and bending
and shear properties, and some on weight and thickness, except Hearle and Amirbayat (1986) who focused on dimensionless quantities. In general, bending, shearing and extension properties were shown to affect DC. Even though extensive detail can be provided by image analysis and measurement for simple geometric forms such as circles and squares, measurement of the drape characteristics of complex forms needs the consideration of researchers to extend the work on drape measurement to garments.
References Abbott, N.J. (1951), “The measurement of stiffness in textile fabrics”, Textile Research Journal, Vol. 21, pp. 435-44. Behre, B. (1961), “Mechanical properties of textile fabrics. Part I: Shearing”, Textile Research Journal, Vol. 64, pp. 346-62. Booth, J.E. (1969), Principles of Textile Testing – An Introduction to Physical Methods of Testing Textile Fibres, 3rd ed., Chemical Publishing Company, Inc., New York, NY. BS 5058: 1973 (1974), “The assessment of drape of fabrics”, BS Handbook 11: British Standards Institution, pp. 4,29-31. Chu, C.C., Cummings, C.L. and Teixeira, N.A. (1950), “Mechanics of elastic performance of textile materials. Part V: A study of the factors affecting the drape of fabrics – the development of a drape meter”, Textile Research Journal, Vol. 20 No. 8, pp. 539-48. Chu, C.C., Hamburger, W.J. and Platt, M.M. (1963), “Determination of factors which influence the draping properties of cotton fabrics”, Agricultural Research Service – US Department of Agriculture, Southern Utilization Research and Development Division, LA. Collier, B.J. (1991), “Measurement of fabric drape and its relation to fabric mechanical properties and subjective evaluation”, Clothing and Textile Research Journal, Vol. 10 No. 1, pp. 46-52. Collier, B.J. and Collier, J.R. (1990), “CAD/CAM in the textile and apparel industry”, Clothing and Textile Research Journal, Vol. 8 No. 3, pp. 7-13. Collier, B.J. and Epps, H.H. (1999), Textile Testing and Analysis, 1st ed., Prentice-Hall, Englewood Cliffs, NJ. Collier, B.J., Collier, J.R., Scarberry, H. and Swearingen, A. (1988), “Development of digital drape tester”, ACPTC Combined Proceedings, p. 35. Cusick, G.E. (1961), “The resistance of fabrics to shearing forces”, Journal of the Textile Institute, Vol. 52, pp. 395-406. Cusick, G.E. (1965), “The dependence of fabric drape on bending and shear stiffness”, Journal of the Textile Institute, Vol. 56, pp. T596-T606. Cusick, G.E. (1968), “The measurement of fabric drape”, Journal of the Textile Institute, Vol. 59, pp. 253-60. Dahlberg, B. (1961), “Mechanical properties of textile fabrics. Part II: Buckling”, Textile Research Journal, Vol. 31 No. 2, pp. 94-9. Dreby, E.C. (1942), “Physical methods for evaluating the hand of fabrics and for determining the effects of certain textile finishing processes”, American Dyestuff Reporter, Vol. 31, pp. 497-504. Frydrych, I., Dziworska, G. and Matusiak, M. (2003), “Influence of the kind of fabric finishing on selected aesthetic and utility properties”, Fibres and Textiles in Eastern Europe, Vol. 11 No. 3, pp. 31-7.
Evaluation of drape characteristics 121
IJCST 17,2
122
Gaucher, M.L., King, M.W. and Johnston, B. (1983), “Predicting the drape coefficient of knitted fabrics”, Textile Research Journal, Vol. 53, pp. 297-303. Grosberg, P. (1966), “The mechanical properties of woven fabrics. Part II: The bending of woven fabrics”, Textile Research Journal, Vol. 36, pp. 205-11. Hearle, J.W.S. and Amirbayat, J. (1986), “Analysis of drape by means of dimensionless groups”, Textile Research Journal, Vol. 56, pp. 727-33. Hoffman, R.M. and Beste, L.F. (1951), “Some relations of fiber properties to fabric hand”, Textile Research Journal, Vol. 21, pp. 66-77. Hu, J. and Chan, Y.F. (1998), “Effect of fabric mechanical properties on drape”, Textile Research Journal, Vol. 68 No. 1, pp. 57-64. Hynek, W.J. and Winston, G. (1953), “A comparative study of the Tinius Olsen and Pierce stiffness testers”, Textile Research Journal, Vol. 23, pp. 743-8. Jeong, Y.J. (1998), “A study of fabric drape behavior with image analysis. Part I: Measurement, characterization, and instability”, Journal of the Textile Institute, Vol. 89 No. 1, pp. 59-69. Jeong, Y.J. and Philips, D.G. (1998), “A study of fabric drape behavior with image analysis. Part II: The effect of fabric structure and mechanical properties on fabric drape”, Journal of the Textile Institute, Vol. 89 No. 1, pp. 70-9. Lindberg, J., Waesterberg, L. and Svenson, R. (1960), “Wool fabrics as garment construction materials”, Journal of the Textile Institute, Vol. 51, pp. T1475-93. Matsudaira, M. and Yang, M. (1997), “Measurement of drape coefficients of fabrics and description of those hanging shapes”, Journal of Textile Machinery Society of Japan, Vol. 50 No. 9, pp. T242-50. Morooka, H. and Niwa, M. (1976), “Relation between drape coefficients and mechanical properties of fabrics”, Journal of Textile Machinery Society of Japan, Vol. 22 No. 3, pp. 67-73. Niwa, M. and Seto, F. (1986), “Relationship between drapeability and mechanical properties of fabrics”, Journal of Textile Machinery Society of Japan, Vol. 39 No. 11, pp. 161-8. Okur, A. and Cihan, T. (2002), “Prediction of fabric drape coefficients from FAST data”, Textile Asia, Vol. 33 No. 7, pp. 28-31. Oloffson, B. (1964), “A general model of a fabric as a geometric-mechanical structure”, Journal of the Textile Institute, Vol. 55 No. 11, pp. T541-57. Orzada, B.T. (1999), “Effect of grain alignment on fabric properties measured with the FAST system”, International Journal of Clothing Science and Technology, Vol. 11 No. 6, pp. 10-11. Pierce, F.T. (1930), “The handle of cloth as a measurable quantity”, Journal of the Textile Institute, Vol. 21, pp. T377-T416. Postle, R. (1998), “The mechanics of woven and knitted fabric”, Textile Asia, Vol. 29 No. 8, pp. 35-6. Postle, R.J. and Postle, R. (1993), “Fabric drape based on objective measurement of fabric bending length”, Textile Asia, Vol. 24 No. 2, pp. 63-6. Robson, D. and Long, C.C. (2000), “Drape analysis using imaging techniques”, Clothing and Textile Research Journal, Vol. 18 No. 1, pp. 1-8. Ruckman, J.E., Cheng, K.B. and Murray, R. (1998), “Dynamic drape measuring system”, International Journal of Clothing Science and Technology, Vol. 10 No. 6, p. 56. Sidabraites, V. and Masteikaite, V. (2003), “Effect of woven fabric anisotropy on drape behavior”, Materials Science, Vol. 9 No. 1, pp. 111-15. Sudnik, Z.M. (1972), “Objective measurement of fabric drape: practical experience in the laboratory”, Textile Institute Industry, Vol. 10 No. 1, pp. 14-18.
Sudnik, Z.M. (1978), “Rapid assessment of fabric stiffness and associated fabric aesthetics”, Textile Institute Industry, Vol. 16, pp. 155-9. Vangheluwe, L. and Kiekens, P. (1993), “Time dependence of the drape coefficient of fabrics”, International Journal of Clothing Science and Technology, Vol. 5 No. 5, pp. 5-8. Winn, L.J. and Schwarz, E.R. (1939), “Technical evaluation of textile finishing treatments: flexibility and drape as measurable properties of fabric”, American Dyestuff Reporter, Vol. 28, pp. 688-94. Winn, L.J. and Schwarz, E.R. (1940), “Technical evaluation of textile finishing treatments. Part IV: A comparison of certain methods of measuring stiffness in fabrics”, American Dyestuff Reporter, Vol. 29, pp. 469-76. Winn, L.J. and Schwarz, E.R. (1941), “Technical evaluation of textile finishing treatments: The effect of relative humidity on flexibility; test methods for the drapemeter”, American Dyestuff Reporter, Vol. 30, pp. 226-30, 238. Yang, M. and Matsudaira, M. (1998a), “Measurement of drape coefficients of fabrics and description of those hanging shapes. Part 2: Description of hanging shapes about anisotropic fabrics”, Journal of Textile Machinery Society of Japan, Vol. 51 No. 4, pp. T65-T71. Yang, M. and Matsudaira, M. (1998b), “Measurement of drape coefficients of fabrics and description of those hanging shapes. Part 3: The effect of fabric parameters on drape shapes”, Journal of Textile Machinery Society of Japan, Vol. 51 No. 9, pp. T182-91. Yang, M. and Matsudaira, M. (1999), “Measurement of drape coefficients of fabrics and description of those hanging shapes. Part 4: Evaluation of dynamic drape behavior of fabrics using a testing device”, Journal of Textile Machinery Society of Japan, Vol. 52 No. 9, pp. T167-75. Yang, M. and Matsudaira, M. (2000), “Measurement of drape coefficients of fabrics and description of those hanging shapes. Part 5: Relationship between dynamic drape behavior of fabrics and mechanical properties”, Journal of Textile Machinery Society of Japan, Vol. 53 No. 5, pp. T115-20. Yang, M. and Matsudaira, M. (2001), “Measurement of drape coefficients of fabrics and description of those hanging shapes. Part 6: Evaluation of dynamic drape behavior of fabrics in swinging motion”, Journal of Textile Machinery Society of Japan, Vol. 54 No. 3, pp. T57-T64. Further reading Lindberg, W.M., Hu, J.L. and Li, L.K. (2002), “Modeling a fabric drape profile”, Textile Research Journal, Vol. 72 No. 5, pp. 454-63. Lo, J., Behre, B. and Dahlberg, B. (1961), “Mechanical properties of textile fabrics. Part III: Shearing and buckling of various commercial fabrics”, Textile Research Journal, Vol. 31 No. 2, pp. 99-122. Postle, R.J. and Postle, R. (2000), “Fabric drape”, Textile Asia, Vol. 31 No. 1, pp. 30-2. Schwarz, E.R. (1939), “Technical evaluation of textile finishing treatments”, Textile Research Journal, Vol. 9, pp. 216-29.
Evaluation of drape characteristics 123
The Emerald Research Register for this journal is available at www.emeraldinsight.com/researchregister
The current issue and full text archive of this journal is available at www.emeraldinsight.com/0955-6222.htm
New measurement technologies for textiles and clothing
New measurement technologies
George K. Stylios Heriot-Watt University, Galashiels, Selkirkshire, UK
135
Abstract Purpose – To establish new measurement technologies in textiles and clothing. Design/methodology/approach – Three areas are covered, diverse in methodology and approach: measurement of fabric mechanics, measurement of seam quality, and measurement of human size and shape. Findings – Three new measurement technologies have been established. Research limitations/implications – The mechanics are limited to the measurement of lightweight fabrics. The human measurement needs clear photographs. Practical implications – All techniques can make measurement more accurate and efficient in all three areas. Originality/value – All techniques are original and have a major contribution to textile and clothing science and technology. Keywords Fabric production processes, Measurement, Fabric testing, Lasers, Anatomy Paper type Research paper
1. General introduction Despite the popular research in nanotechnologies, in smart and interactive materials, and in multifunctional textiles, measurement technologies are still at the heart of textile products and processes, important for product quality and production efficiency. This paper presents three new measurement technologies, one in fabric mechanics, second in seam quality and third in body size and shape. All techniques have unique attributes and are trying to improve problems associated with the existing measurement techniques. 2. The measurement of textile mechanics Textile materials are of particular interest since they neither behave as solids nor as liquids, i.e. they are said to be limp and they possess viscoelastic properties which enable them to take up any 3D configurations by wrapping or hanging around solid bodies. 2.1 Literature considerations The interest in the behaviour of textile materials has been with us since the ancient Greek times, and although not cited in literature the conventional way, it is certainly demonstrated by the breathtaking accuracy of the way that the sculptures are clothed indicating knowledge by their creators of the behaviour of these textile materials. The author wishes to thank the European Union for partly funding the 3D body simulation and KyberStyl Ltd for their permission to using materials in this manuscript. FAMOUS is subjected to International Patents.
International Journal of Clothing Science and Technology Vol. 17 No. 3/4, 2005 pp. 135-149 q Emerald Group Publishing Limited 0955-6222 DOI 10.1108/09556220510590830
IJCST 17,3/4
136
This is illustrated in the realistic hanging, creasing and draping depicted in the sculptures (National Museum, Athens, Greece). The first scientific citation related to this field is a paper by Peirce (1930) in which he sets out the importance of mechanical properties of textile materials in an attempt to predict their handle. Another landmark is the important research work carried out at TEFO in Sweden in the 1960s ,where the fundamentals of actual measurement methods and magnitudes were established and reported (Lindberg et al., 1960). In the 1970s the Japanese having realised the importance of routine fabric measurement developed the “Kawabata Evaluation System for Fabrics” (KESF) series of instruments , which has been regarded as an answer to the imposed questions made by Peirce. These series of measurements were focused in the accurate prediction of fabric handle (Kawabata, 1982). In the late 1980s, CSIRO in Australia realised the importance of commercial measurement for wool fabrics and tried to offer a simpler and cheaper alternative to KESF, which was named “Fabric Assurance by Simple Testing” (FAST) (Biglia, 1994). The equipment was promoted for predicting the tailorability of wool fabrics, which was demanded by the industry. Although the area of fabric measurement has been established and recognised under the generic name of “fabric objective measurement” and is being used by many companies as best practise, nevertheless industry is expected to have fully adapted this technology by now and the main reason is attributed to existing measurement devises having not been widely accepted in industry. In the last few years, there has been a general feeling of a need for developing a better alternative method of fabric measurement. 2.2 Problems associated with the existing methods of fabric measurement The general understanding about the current provision of equipment after extensive scientific and industrial use over the last 20 years is that the KESF is regarded as a scientific device for research and FAST as a simplified alternative devise for industrial use. Both instruments offer mechanical measurements by using a number of different devises. For example, in the case of KESF there is one devise that measures tensile and shear properties, another for measuring bending, one for compression and the other for surface. Each devise may have attachments, which need to be used for specific measurements as in the case of measuring surface irregularity and friction. The measurement is accurate using sensors and the movement is smoothly achieved by efficient use of lead screws, gears and motors. The sample handling and feeding on each devise is manual and the operator needs to calibrate each devise before each measurement. In the case of FAST thickness and compression are measured in one devise at two positions of a movable head. Bending is measured in another devise by the cantilever principle. For tensile strength, the fabric is placed between two clamps one of which is loaded by a known weight in another devise. The devise registers the load in the fabric by release of the weighted clamp. The results of the KESF are very accurate and the instrument accumulates enough data for a complete test, but a fabric sample test is completed in 1-2 h. The FAST results are limited to the measured loads only and may not be able to provide sufficient data for complete stress/strain profiles in some fabric samples. Difficulties may have been experienced in the reproducibility of the measurements of both devises due to the need of different sample sizes and the manual handling for each test.
Cost is another important consideration. The case of the KESF equipment is that it is expensive and out of the reach of SMEs, which is the majority of textile enterprises today. The FAST is considerably cheaper on the other hand. Another added cost for both systems is the skill of the operator for measurement and interpretation of the test results which is crucial for the effective implementation of the technology. In the last 3 years, the KESF has been made in a totally automated version although yet to be fully commercialised. The system is said to consist of at least two modules and although connected during feeding these modules are independent of each other. The cost of this system is even higher than its manual replacement, which hinders even more the wider use of this technology. It is therefore, understandable that despite the efforts of the textile industry to use those techniques for defining their products, the above shortcomings prevent them from doing so. Recently and under the pressure of the industry, a new concept was established which was developed into a new devise for measurement of limp materials and received a SMART award in the UK. This devise is called FAMOUS; and stands for fabric automatic measurement and optimisation universal system. 2.3 The concept and principle of famous fabric measurement To fulfil simplicity all tests had to be made using one sample only and without the need for human intervention, a new concept of measurement in two planes had to be invented; one for tensile, shear and flexural rigidity and another for compression and surface. The first aim of this concept is to provide a single apparatus for the measurement of the mechanical properties of a single sample of a limp sheet material in order to reduce the equipment costs compared with that of the number of existing devises required. The second aim of the present invention is to reduce the time and complexity of making such measurements and to increase the accuracy and reproducibility of such measurements compared with the existing methods. To achieve these aims the equipment uses state-of-the-art stages, slides, motors and sensors in an ingenious way which produce motion in six axis and measure property changes in five different positions. Plate 1 shows a photograph of the equipment, which is bench top based and portable. A fabric sample is cut 20 £ 20 cm and placed on the machine. The order of measurement is as follows: flexural rigidity, shear, surface, compression and tensile properties. All magnitudes and measured parameters are the same as the existing equipment so that there is direct correlation and continuation of measured data, with the exception of flexural rigidity, which is being derived by a different mode of measurement. A complete suite of measurement takes less than 5 min. The measured data are interpreted into an ideal fabric chart automatically. 2.4 Implementation of famous in the textile, clothing and retailing industries The equipment is currently undergoing industrial trials. Figures 1-5 show a tensile, shear, buckling, compression and surface force graphs produced after fabric testing trials. A fabric sample is cut 20 £ 20 cm and placed on the machine. The order of measurement is as follows: flexural rigidity, shear, surface, compression and tensile properties. All magnitudes and measured parameters are the same as the existing equipment so that there is direct correlation and continuation of measured data, with the exception of flexural rigidity, which is being carried out by a different mode of
New measurement technologies 137
IJCST 17,3/4
138
Plate 1. FAMOUS equipment
Figure 1. Typical tensile curves of a suiting plain weave fabric
measurement. A complete suite of measurement takes less than 5 min, and the measured data are interpreted into a snake chart automatically. Figure 6 shows the output of the ideal snake chart program produced to work with data received during measurement. The ideal woollen chart is based upon extensive work done using FAMOUS for the upholstery industry.
New measurement technologies 139
Figure 2. Typical shear curves of a suiting plain weave fabric
Figure 3. Typical buckling curves of a suiting twill fabric
3. 3D measurement of body size and shape 3.1 Literature considerations Realistic simulation of garments on real humans is one of the most fascinating areas of computer graphics and engineering. It has wide applications in computer graphics, the garment industry, the cartoon, film and virtual environments. Over the past several years, mathematicians, physicians and engineers with the aim of finding algorithms with which to model clothes or other kinds of textiles have carried out a considerable amount of research. Such research has usually been focused on simple objects like curtains or flags, while limited work has been reported on computing the behaviour of clothes for virtual catwalks or virtual actors. Many efforts have been directed on this
IJCST 17,3/4
140
Figure 4. Typical compression curves of a suiting twill fabric
Figure 5. A typical surface curve of a suiting twill fabric
problem; however, modelling real humans for virtual garments still remains a challenge at large (Huh et al., 2001; Volino et al., 1995; Chittaro and Corvaglia, 2003). The polyhedral representation is capable of displaying a 3D object in any degree of accuracy. However, the requirement of a vast amount of polygons overwhelms in practical terms any shape reconstruction capability for individual human modelling. Very recently, there has been a growing interest in using the concept of implicit surfaces for reconstruction of humans and thus, complex surfaces can be modelled using relatively few primitives that can easily be animated (Fun et al., 1998; Thalmann and Thalmann, 1991; Komatsu and Thalmann, 1988; Komatsu, 1988). However, not all of these techniques can combine reliable shape reconstruction with seamless, high-resolution texture generation and mapping. So far, although there are some
New measurement technologies 141
Figure 6. Automatic measurement ideal fabric chart
IJCST 17,3/4
142
advances in this field, they have not been widely applied because their processes are too complex. This paper highlights the integration of four important research areas attempting to introduce new ways of designing, selling and producing garments; hence exploiting global internet retailing is an area of enormous interest because it challenges the conventional way of buying, selling, producing and distributing clothes. The four areas are as follows: (1) geometrical reconstruction of real humans; (2) digital cloning of 3D face and body (3) virtual human locomotion; and (4) cloth simulation. Our approach emphasises simplicity, low-cost and accessibility by anybody without special equipment. We recognise the potential application of such a system in many industrial applications including garment-retailing applications via interactive web-based PC systems (Stylios et al., 2000; Stylios and Wan, 1997). This paper outlines the results we have achieved in the four directions. 3.2 Geometrical reconstruction of a real human In order to be able to extract reliable body size and shape data, a reconstructed 3D human has to be presented. To achieve this, their body shape has to be reconstructed and hence we have a hybrid approach of our so-called “synthetically real human” concept. To begin with, a scheme with an initial control mesh is generated to construct a hierarchical multi-resolution curved surface that represents the geometrical silhouette of the human, in relation to position variation of a limited number of feature points from ordinary 2D photographs. Then, details are added with Trigonometry and Bicubic Spline incorporation (Bicubic Spline, http://mathworld.wolfram.com/ BicubicSpline.html) on and between cross sections separately. The surface patches based on these 3D curves are represented by n X m X Bi;j ðt; sÞP i;j ð1Þ Qðt; sÞ ¼ i¼0 j¼0
where Pi,j is an array of control points and Bi,j (t, s) is a basis function, which are evaluated by Bi; j ðt; sÞ ¼ Bi ðtÞ · Bj ðsÞ
ð2Þ
where Bi(t) and Bj(s) are defined as Bicubic Spline or Hermite and Trigonometry functions in the current modelling. However, the requirements for the regulation of vertex distribution and the simplification in the reconstruction impose additional restrictions on the chosen feature control points. That is, not all feature points may be arranged as control points on the parametric surface. In this case, as an important complement, extra-localized deformation for some important but complex features should be exerted to fit an individual. This is achieved by free deformation techniques based on a line deformation algorithm, in combination with appearance feature measurement and a curve-fitting algorithm presented in our work (Stylios and Han, n.d.).
Plate 2 shows the procedure for creating various feature details of a generic human head by changing the position of the control point and exerting a local deformation nearby the corresponding feature position. From three sets of human photographs of any individuals, several modelling examples of the human shape have been originated using the described algorithm, as shown in Plate 3. Such a parametric based modelling approach leads to an individualized model, based on a generic head model that can easily incorporate specific features of any human, without additional complex manipulations. As a result, a generic parametric human model may be deformed into any specific human model with regular vertex distribution, which is very significant
New measurement technologies 143
Plate 2. The creation of various features on a human head
Plate 3. The reconstruction of two individual human body models; derived from photographs in the front, side and rear views
IJCST 17,3/4
144
for face and body cloning as well as animation. Other advantages of this modelling approach are simplicity and accuracy. 3.3 Digital cloning of 3D human face and body Starting with a generic human model, face and body cloning use a set of 2D photographs of a real person. The procedure of surface modification to clone an individual human is described in the geometrical reconstruction of a real human. Further detail fitting is achieved by combining a local deformation and mapping procedure. The detail process of the texture image generation is provided elsewhere (Stylios and Han, n.d.). Guided by the merging lines for the textured image generation, all vertices of an individualized 3D head surface are projected onto one of the following three planes: YZ plane on the left view, XY plane on the front view and YZ plane on the right view. Subsequently, any point on the 3D head is projected to the corresponding region of the 2D texture image and finally, the mapping coordinates of the points on a texture image are generated. Accordingly, an algorithm is presented to calculate the texture coordinate of these vertices by using a different plane texture-mapping algorithm in the corresponding sub-domain of the texture image, in terms of the cast direction of the photograph. If we take a point on the front face as (Xi, Yi, Zi) on the 3D head model, the corresponding texture mapping coordinates in the texture space are calculated by the following: uij ¼
X ij 2 H imin H imax 2 H imin
vij ¼
Y ij 2 V jmin V jmax 2 V jmin
ð3Þ
ði ¼ 0; 1; . . . ; level 2 1; j ¼ 0; 1; . . . ; point 2 1Þ where H imin ; V jmin and H imax ; V jmax stand for the minimum and maximum of unfold value along ith horizontal and jth vertical direction on the 3D head model surface, respectively. Similarly, instead of Xij with Zij, the formula can be used to calculate the 2D texture-mapping coordinate for one point on the side of the face. Instead of a pair of photographs from the front and side views in the face cloning, two body photographs from the front and rear views are chosen to generate texture images that are cloned on different body parts of the virtual human with the multiple branch body-cloning algorithm. As shown in Plates 2 and 3 each skin part with regular vertex distribution of a parametric body model is reconstructed in local coordinates and is then integrated smoothly into a complete human model in global coordinates by transformation and rotation with referencing to the corresponding joints. In this case, the texture coordinate of each pixel in every human body part in the texture image has to match with the position of the relative vertex in the human body components to be cloned. Thus, different parts of the body textures can be cloned directly on the corresponding parts in the front and rear views by computing and regulating of the texture coordinates in the global coordinate system. Plate 4 shows some examples of real human face cloning generated from real human head photographs in the front and side views. These figures are shown with the corresponding individual human head photographs, it can be clearly observed that the presented face cloning techniques produce a lifelike human head. Plate 4 also shows
New measurement technologies 145
Plate 4. Examples of human cloning with individual heads
several examples of complete cloning of the whole human including face and body. As a result, a lifelike virtual human with satisfactory cloned features is displayed without complexity, which may further be transformed on line and animated in virtual environments. From the 3D reconstructed human body, size and shape data can be extracted and used for a variety of end uses. 4. The Pucker laser measurement system 4.1 Introduction of the PLMSe The Pucker laser measurement system (PLMS) is an advanced instrument for the measurement of seam Pucker and other irregularities it utilises a laser-based vision technique, which simulates human assessment of seam Pucker. This technique overcomes problems with fabrics of low reflectance and patterns as well as dark and light fabric samples that other techniques cannot. The instrument is simple to use, portable, maintenance free and robust enough for use in laboratory and factory conditions. Seam Pucker is one of the most demanding and serious problems in garment manufacture. Its measurement is important because it determines whether the
IJCST 17,3/4
146
stitching in a fabric is acceptable or not. Judges who grade stitched fabric samples and compare them by photographic standards have traditionally carried out this measurement, the most common used ones are the AATCC. Measurement by using judges is subjective and highly dependent on personal interpretation. In 1992, Stylios et al., stated that seam Pucker was an aesthetic problem and had to be measured objectively, hence the first pure vision based system was developed that used a camera to grade seam Pucker by measuring the amplitude and length of the wave created by fabric deformation along the stitch line (National Museum, Athens, Greece; Peirce, 1930; Lindberg et al., 1960). The system was efficient in the laboratory but had some difficulty due to ambient light, patterned and light and dark samples. In 1993 Stylios et al. improved this system by replacing the camera with a distance laser sensor which improved the situation and the instrument was used and is still used for industrial measurement of seam Pucker at the research centre of excellence COMIT (Kawabata, 1982). This instrument needed some care and fabrics of low reflectance; patterns, dark etc. need special attention. Recently, ITMA’97 showed and attempt by a Korean company to commercialise this instrument but without success due to the stated reasons. These difficulties have now been overcome in a leading edge high-tech instrument that utilises a combination of laser and camera in a simple, robust design. The PLMSe is therefore, the product of not less than 15-year research with the pioneering study of seam Pucker by Stylios (1982) (Kawabata, 1982). 4.2 Description of the PLMSe The PLMSe consists of a box-like cabinet (Figure 7) inside which the laser arrays, the camera and the sample holder are situated. The box has an opening from where a sample slider (ruler-like) can be pulled out and inserted in for testing. The slider is also used as a fabric sample template for cutting fabric samples (see fabric sample preparation for testing). The box at its opening has a micro switch so that the lasers are switched off when opened for inserting the fabric for measurement. The lasers are of class 2 and should not be looked directly by the human eye. The box is connected via two cables with a vision interface card, with a laser control card (outside the PC) and an input/output card, which is installed inside the PC and is occupying two of its expansion slots.
Figure 7. The PMLS showing the PC with its interface and control cards
4.3 Fabric sample testing procedure The sample testing procedure consists of three parts. Fabric sample preparation (Figure 8). The box door is opened; the slider is pulled upwards and outwards, and placed on top of a laid on a table fabric for sample cutting for measurement. A pen should carefully outline the fabric all the way around the template and as close to it as possible, so that the sample is not smaller or larger than the template. The slider has two holes for marking the centre of the sample stitch line for sewing. Mark the two holes and draw a straight line by joining them. Cut carefully on the outside rectangular lines so that the size of the fabric is exactly the same as the template; this is very important for sample preparation standardisation; if not the measured results may be affected. Two rectangular fabric shapes should be carefully stitched by a sewing machine with a straight seam on the marked line (there can be one fabric or more than two). The instrument is capable of measuring samples that are stitched nearer to the edge of the fabric. These fabric samples can also be measured provided that the stitched line is in the same position. In this case, only the left side of the stitch line will be measured. The stitched sample should be carefully placed on the template (slider) and the upper side should be clamped by turning the eccentric wheel so that it gently clamps
New measurement technologies 147
Figure 8. Fabric sample preparation, cutting and mounting
IJCST 17,3/4
148
the sample. Care should be taken for the sample to be contained within the edges of the slider. The sample should not be longer and not allowed to overhang. The slider with the sample clamped on it should be slide back through the door of the box and must be guided down the guide, which engages with the slider easily. The door of the box should then be closed to perform measurement of the sample. If the cabinet door is not closed, the lasers are inoperative and measurement will not be made. Please note that the sample preparation procedure: sample cutting and sample holding are very important for consistent measurement and have to be standardised. The instrument is very accurate and can like the human eye detect differences in fabric sample presentation (Figure 9).
5. Discussion and conclusion Three diverse new measurement technologies have been described and discussed. A new fabric mechanics devise can measure the mechanical properties of limp sheet materials very accurately and in less than 5 min. It consists of art stages, slides, motors and sensors in an ingenious way, which produces motion in six axis and measure property changes in five different positions. The measurement of seam Pucker and other irregularities can be achieved by a newly devised laser-based technique quickly, reliably and automatically without the need of judges. Body size and shape of humans can now be reconstructed from simple photographs and without the need of complex and costly scanners. From this reconstruction reliable body size and shape can be determined.
Figure 9. Screen graphical illustrations
References Biglia, U. (1994) in Stylios, G. (Ed.), Textile Objective Measurement and Automation in Garment Manufacture, Ellis Horwood Ltd, Chichester, pp. 139-44. Chittaro, L. and Corvaglia, D. (2003), “3D virtual clothing: from garment design to web3D visualization and simulation”, Proceedings of Web3D 2003: 8th International Conference on 3D Web Technology, March 2003, ACM Press, New York, NY, pp. 73-84. Fun, P. et al. (1998), “Human body modelling and motion analysis from video sequences”, International Archives of Photogrammetry and Remote Sensing, Vol. 32, pp. 866-73. Huh, S., Metaxas, D. and Badler, N. (2001), “Collision resolutions in cloth simulation”, IEEE Computer Animation, pp. 122-7. Kawabata, S. (1982) in Kawabata, S., Postle, R. and Niwa, M. (Eds), Objective Specification of Fabric Quality, The Textile Machinery Society of Japan, Tokyo, pp. 31-59. Komatsu, K. (1988), “Human skin model capable of natural shape variation”, The Visual Computer, Vol. 3, pp. 265-71. Komatsu, K. and Thalmann, D. (1988), “Interactive shape design using metaballs and splines”, Proc. Eurog. Skin Model Capable of Natural Shape Variation, Virtual Computer, Vol. 3, pp. 265-71. Lindberg, J., Waesterberg, L. and Svenson, R.J. (1960), J. Text. Inst, pp. T1475-93. Peirce, F.T. (1930), J. Text. Inst., Vol. 21, p. T337. Stylios, G. (1982), “Seam Pucker and structural jamming in woven textiles”, MSc thesis, The University of Leeds,. Stylios, G.K. and Han, F. (n.d.), “Parametric reconstruction and virtual cloning of the human head” (in press). Stylios, G.K. and Wan, T.R. (1997), “Concept of global retailing; the integration of new technologies in the fashion”, Textile and Apparel Industries, Refereed Paper, Proceedings, The 78th World Conference of The Textile Institute in Association with The 5th Textile Symposium of SEVE and SEPVE, Thessaloniki, Vol. I, pp. 45-54. Stylios, G.K., Han, F. and Wan, T.R. (2000), “A remote, online 3D human measurement and reconstruction approach for virtual wearer trials in global retailing”, in Gersˇak, J. (Ed.), paper presented at the 3rd International Conference Innovation and Modelling of Clothing Engineering Processes – IMCEP, October 2000, Faculty of Mechanical Engineering, Maribor. Thalmann, N.M. and Thalmann, D. (1991), “Complex models for animating synthetic actors”, Computer Graphics and Applications, pp. 32-44. Volino, P., Courchesne, M. and Thalmann, M. (1995), “Versatile and efficient techniques for simulating cloth and other deformable objects”, Computer Graphics Proceedings, Annual Conference Series, SIGGRAPH, pp. 137-44. Further reading Stylios, G. and Sotomi, J.O. (1993a), “A new instrument for routine objective assessment of seam deformations in limp materials”, in Hope, D. and Smith, G.T. (Eds), Laser Metrology and Machine Performance, Lamdamap 93, Computational Mechanics Publications, Blackshaw, pp. 233-8. Stylios, G. and Sotomi, J.O. (1993b), “Investigation of seam Pucker as an aesthetic property using computer vision. Part 1: a cognitive model for the measurement of seam pucker”, Journal of the Textile Institute, Vol. 84 No. 4. Stylios, G. and Sotomi, J.O. (1993c), “Investigation of seam Pucker as an aesthetic property using computer vision. Part 2: model implementation using computer vision”, Journal of the Textile Institute, Vol. 84 No. 4.
New measurement technologies 149
The Emerald Research Register for this journal is available at www.emeraldinsight.com/researchregister
The current issue and full text archive of this journal is available at www.emeraldinsight.com/0955-6222.htm
IJCST 17,3/4
Ceremonial academic gowns of the University of Zagreb – idea to a finished product path
150
Zvonko Dragcˇevic´, Slavica Bogovic´ and Edita Vujasinovic´ Department of Clothing Technology, Faculty of Textile Technology, University of Zagreb, Zagreb, Croatia
Tomislav Bakran GRIP, d.o.o., Zagreb, Croatia Abstract Purpose – To design, develop and construct specific garment designs for use in Croatia as academic gowns using advanced engineering principles. Design/methodology/approach – The synergism of historians, designers, engineers, clothing technologists and textile finishers using specialised equipment has been employed in this project. Findings – Using interdisciplinarity can yield good results. Research limitations/implications – The research targets specific products, but its methodology may be used for any other products/end users also. Practical implications – Gowns have been designed, made and used in academic ceremonies successfully. Originality/value – Design/technology approach to new product development. Keywords Clothing, Modelling, Measurement, Croatia, Universities Paper type Research paper
International Journal of Clothing Science and Technology Vol. 17 No. 3/4, 2005 pp. 150-160 q Emerald Group Publishing Limited 0955-6222 DOI 10.1108/09556220510590849
1. The history of the academic gown Academic dress is the insignia of a particular university. In the past, it was an everyday, obligatory uniform used at work, while today it is worn on formal occasions only. In early days of constituting European Universities (lat. studia generalia) in the XII century, the teachers and students wore a usual clerical gown (Hargreaves-Mawdsley, 1963; Bazala, 1942), since they were all clergy members. Even in the countries of southern Europe ordinary students, for the sake of the spiritual discipline, used gowns as those of the clergy. The robe they wore – it was called pluvial – in general marquee did not differ from the clothes of the other layers of society: simple wide gown with the hood that was protecting them from bad weather, with an opening for the head and a slit in the front for arms. This kind of clothes is used even today and is called cape. There is only one robe that is meant exclusively for clergy (except the ones used in liturgy) – cappa clausa – which developed from the pluvial. Cappa clausa is a robe with wide folds, sleeveless and reaching the feet. It was worn as a recommended Part of the experimental measurements (FAST and KES) has been executed as a part of CEEPUS program SI-007 at Textile Department, Laboratory for Clothing Engineering, Physiology and Construction of Garments, Faculty of Mechanical Engineering, University of Maribor, Slovenia.
academic gown at the entire tree of the most important medieval universities – in Bologna, Paris and Oxford, as well as at the other universities (Hargreaves-Mawdsley, 1963). Later, this robe was less worn by clergy, so cappa clausa has become a predominantly academic gown. In the same way pileus, that was a head cover common in those times, was a remarkable clerical robe, and in the same evolution path as cappa clausa has become a head covering at academic institutions. After some time, the officials of all of the universities accepted changes in the academic clothing that moved towards everyday fashion and the real academic gown of the XVI century. Today, every university has its own cut and shape of the academic gown. In the XVII century, the custom of wearing university clothes was neglected everywhere except in England, France, Portugal, Spain and at a couple of universities in the South of Germany and Austria. In other countries, because of the university opening towards everyday mundane influences, university members did not cling to wearing academic gowns, although they existed as such. The University of Zagreb was founded at times when the importance of wearing academic gowns was already reduced, and since in 1784 the emperor Joseph II banned wearing academic gowns in the whole of the Austro-Hungarian empire, the academic gown was not in use and the only symbol of academic status was the sceptre and the chain (Dragcˇevic´ and Potocˇic´, 1999; Horvat, 1942). Today, in almost every part of the world all the most important ceremonies at the universities are directed, recorded and equipped with an adequate academic gown, the purpose of which is to contribute to the dignity of the occasion and to create an impression of traditional culture and wisdom. Academic dresses are loose, preventing sudden movements, forcing the wearer into a particular posture, particular manner of walking, keeping the arms apart from the body, which, on a symbolic level, signifies eminence, seriousness and dignity. 2. Experimental part 2.1 Methodology Since the academic gown was not used nor did the tradition of its use exist at all, during its creation it was essential to avoid the use of marquee and looks of the gowns of the other universities. Actually, the ceremonial academic gown should incorporate recognisable marquee, class, beauty and academic dignity. It should represent the University of Zagreb, as well as the Republic of Croatia in all of the formal occasions that Rector and other university and faculty officials attend. With this goal in mind, the experts from different scientific fields where summoned for the project of creating ceremonial academic gowns (Figure 1). These were the experts like historians, fashion designers, experts in textile materials, embroidery, construction and modelling. Ceremonial academic gowns for honorary officials were realised in the Department of Clothing Technology at the Faculty of Textile Technology, University of Zagreb (Croatia), based on the general idea of MSc Vesna Marija Potocˇic´. To transform the general idea of the academic gown creation into a finished product, one of the important conditions was a correct selection of the fabric to be used to produce the ceremonial academic gown. With this in mind, while employing the basic methods of textile fabric characterisation (Bona, 1994), the modern automatic
Ceremonial academic gowns
151
IJCST 17,3/4
152
Figure 1. The plan of creating ceremonial academic gown of the University of Zagreb
measuring systems for fabric quality assurance, fabric assurance by simple testing (FAST) and Kawabata evaluation system (KES-FB) (Kawabata and Niwa, 1991) were used. Basic modelled cutting parts were created on the modern system for clothing design Lectra Systems, while the creation and execution of decorative embroidery of
braid ornaments was done in company of GRIP Zagreb, using the embroidery CAD system on modern multi-head embroidery machines TANG-JINGWEI, China.
2.2 Results Based on the data obtained by investigating the history of academic dresses, and with the aim of obtaining an academic dress arising from the tradition of the first days of the University of Zagreb, comparable with the contemporary formal academic dresses of the European universities, the designers have chosen the Jesuit cassock belonging to the founders of the university, Jesuits, as their basic model. The basic idea has been modified as follows: the cassock has been shortened a bit to make walking easier, has become more bulky, sleeves have been given a full bell-shaped form. Embroidered tapes have been inserted onto the front and back side, having a multiple function of giving additional bulkiness to the dress, while the embroidery is decorative and is used at the same time to denote the academic status and the faculty of the wearer. Those embroidered tapes are decorated with braid ornaments (Figure 2). The braid has been chosen as one of the visually most representative Croatian motifs, beautiful in its geometrical simplicity and mysterious rhythm of interlocking strands. Apart from the ornamental motives, the other characteristic used was colour. Basic colour is black and the colours of the sleeves differ based on the academic status. For the sleeves of the rector, as well as for the post honorary, different shades of red were used, white-candidus, is the colour of the sleeves of the one that will change his academic status (colour of the candidate), whilst the sleeves on the gowns of the deans and vice deans are blue and dark blue colours (colours of the city of Zagreb). To create the ceremonial academic gowns, several different types and constructions of the main fabric were selected. To exclude subjective quality evaluation of each and every fabric and to ensure the scientific approach to designing and creating ceremonial academic gown, detailed research of the quality of the fabric selected were conducted. It was concluded, based on research, that the fabric Cordial, Varteks has the best property values for ceremonial gowns. The main construction characteristics of the fabric selected: Cordial, Varteks (HR) are shown in Figure 3. Figure 4 shows the FAST control card for Cordial fabric, whilst Figure 5 shows the results of distinguishing its mechanical-physical properties and the overall THV values by FES-FB system. The values of tensile properties of the chosen fabric and its resistance to wrinkling are shown in Table I. Figure 4 shows visible that in the process of academic gown creation potential difficulties can occur when sleeves are inserted, because of the relatively low value of the warp and weft formability, possible problems with overfed seams caused by low warp extensibility as well as possible sizing problems, since the warp relaxation shrinkage is relatively high. It is necessary, in designing and cut sheet creation of the ceremonial academic gown, to pay special attention and to match the basic cut in with fabric characteristics. Based on the data shown in Figure 5, the overall assessment of the touch for the fabric selected ðTHV ¼ 2:91Þ is given. The fabric itself is of good quality, usual average formability, good fit and fall, comfortable to wear and adequate for different weather conditions. Based on these quality properties, and also on the high values of wrinkle recovery
Ceremonial academic gowns
153
IJCST 17,3/4
154
Figure 2. The motives of decorative ornamental braid on ceremonial academic gowns of the University of Zagreb, which represents honorary status of: (a) rector; (b) vice rector; (c) dean/vice dean of the university
(71:30 , K , 90:82; Table I), the tested fabric is assessed to be fully adequate for ceremonial academic gown creation. Main cut (Figure 1) of the ceremonial academic gown is based on the construction of a gown and is created on the breast circumference 100 cm and the overall body height of 172 cm. It was tested and corrected in match with general design and the quality of the fabric used. The model of the gown is lightly bell-shaped. The sleeves in the front and back part are profound and the front and back parts are additionally widened during the construction, so as to yield to the general design and to ensure additional comfort in wears. There are openings in side seam that permit access to the pockets of the clothes worn under the ceremonial gown. The neck opening in the front and back part is profound. Since the stiff collar is used for ceremonial academic gowns, it was necessary to widen the neck opening in comparison with classic construction, to prevent the ceremonial academic gown from wrinkling because of the clothes worn under it. The sleeve that was used for ceremonial academic gown was constructed based on the front and back part of sleeve itself, during this process of construction a fold was added, based on the model. The sleeve is constructed of two parts and its width is the same as length, measured from the armhole. The sleeve curve and armhole are lightly curved to yield to the general design and to avoid deformations due to the low stability of the fabric by the weft. Modelling the main cut of the ceremonial academic gown is done on the computer construction system by company Lectra, in the program module Modaris. The front part was split vertically at the middle of the shoulder line downward in two parts, where a fold is added to the armhole. In the front part, a band with embroidered braid ornamentals is added. The band is on its top, at the shoulder line, shaped to ensure good fall of the embroidered part. Since the shoulder line is curved, it is important to shape the embroidered band to avoid deformations in the finished ceremonial academic gown. In case of adjustment of the front part of the gown to a female body, the breast insert is added under the embroidered band, to keep the looks of the ceremonial academic gown the same. The back part of the ceremonial academic gown is also split vertically downwards at the same point as the front part. The bands for the embroidered braid ornamentals are added also, but it is different from the front part, since the curves of the shoulder lines differ. Two additional folds are added to the sleeve and while joining all of the parts a single sheet cut is created, as shown in Figure 6. In the single sheet cut, the seam of the sleeve is merged with the side seam of the robe. The main cut of the sleeve is shortened for the width of cuff, which represents different academic status. The cut of the sleeve is changed varying the width of cuff. After modelling of the ceremonial academic gown, grading of the cut was executed (based on the age, bodily constitution and static anthropometric ratio) and the cut sheets for gowns are created in the CAD system Lectra, program module Diamino. Academic gowns are created in series of four (rector þ 3 vice-rectors; dean þ 3 vice
Ceremonial academic gowns
155
Figure 3. Construction characteristics of the Cordial fabric
IJCST 17,3/4
156
Figure 4. FAST control chart
deans). During the creation of the cut sheets, the usage of the main fabric is calculated for each single academic gown, depending upon the physical constitution of the person wearing it. Variations are from 4.5 to 6.5 m length. Because of the ceremonial and dignifying character of the academic gowns, it is decided that traditional Croatian braid (symbol of belonging) should be embroidered to the whole length of the front and back part of the gown. Since the imperative of the whole ceremonial academic gown is academic dignity and eminence, the embroidery is chosen because of its three dimensionality and a shiny thread for embroidery is used in colour and tone of the main fabric of the gown (ISACORD No. 40; 100 percent PES; Amann, DE). Thus embroidered ornamentals can be seen on the dull black fabric as a play of light only.
Ceremonial academic gowns
157
Figure 5. The properties of the Cordial fabric obtained by KES-FB
Fabric Cordial
Tensile properties Warp Weft Fb/N 1p/percent Fb/N 1p/percent
a0/8
xs s
493.75 45.08
138.04 2.80
57.25 1.55
317.77 29.30
36.99 7.43
Resistance to creasing Warp Weft K/percent a0/8 K/percent 71.30 1.57
164.08 17.45
90.82 10.40
During machine embroidery of the braid ornamentals, special attention is paid to the size of the braid motif, as the embroidery machine is imprecise while embroidering large motives. The embroidered braid ornamentals dimension is 80 £ 1,400 to 1,600 mm (based on the clothing size) and the standard embroidery machine (TANG-JINGWEI, China) has the work field of 400 mm. The whole braid ornament is divided into four equal parts that are all embroidered separately, under each head of embroidery machine. The problem of impreciseness of the embroidery machine is solved by calibrating the exact embroidery machine at the spot where the embroidery process is executed. In the CAD system for digitising the embroidery patterns Stilista (GMI, Italy) the rectangles of given dimension are created in the size of the embroidery field of the machine. They are embroidered on the machine, and using a scaled magnifier glass
Table I. Tensile properties and resistance to creasing
IJCST 17,3/4
158
Figure 6. The cut sheet of ceremonial academic gown after modelling and grading
the difference from given sizes are noted and the embroidery pattern is then scaled to yield a perfect joining of the connecting parts of the braid ornament. Since the ceremonial academic gown has almost one million machine stitches, it is necessary to fix each cut band that is embroidered, before embroidering it on the embroidery machine. Each cut band is fixed with a PE coated non-woven fabric, which is removed after the process of machine embroidery. The values of the upper and lower thread tension were measured with Coats meter (Coats) and reset to optimal values before the embroidery process, due to the characteristics of the basic fabric and high quality of the embroidered braid ornament. Figure 7 shows the motif of the embroidered decorative ornamental braid for the Faculty of Textile Technology dean’s gown (Rogale et al., 2001). 3. Conclusion remarks The project of creating ceremonial academic gowns from the idea to the finished product is based on the scientific approach and a joint work of fashion designers, textile technologists in area of textile materials, clothing engineering and machine embroidery. This type of approach results in the creation of academic gowns that will possess aesthetics properties resulting from merging modern design and traditional Croatian ornamentals, and at the same time be comfortable and functional and show recognisable marquee, class, beauty and academic dignity at all of the occasions the university and faculty officials attend. Academic gown is also aimed at a long-term use, its design, material choice and construction execution yields its easy maintenance. Since the academic gown is rather long (150-180 cm) its execution with a special folding method allows for easy transport assuring that its shape and form are kept immaculate on later wearing occasions. The project of making ceremonial academic gowns, although started in 2001, is still going on. Ceremonial academic gowns for the high-ranking honourable officials of the Rector’s Office of the University of Zagreb have been made in the course of the three years (Plate 1), as well as for the faculty deans and vice-deans of the 19 faculties in Zagreb, and for the rectors and vice-rectors of the Polytechnic of Zagreb and
Ceremonial academic gowns
159
Figure 7. Motive of decorative embroidery of braid ornament for Faculty of Textile Technology dean’s gown
Plate 1. Honourable high-ranking officials of the University of Zagreb conferring doctor’s degree to the candidates (Photo by IPIK, Zagreb)
IJCST 17,3/4
160
Polytechnic of Karlovac. The project also included ceremonial academic gowns for the graduates of the University of Zagreb, to be worn by them on the graduation day at the faculties. References Bazala, V. (1942), “Hrvatsko sveucˇilisˇno drusˇtvo”, Alma Mater Croatica, Vol. 5, p. 349. Bona, M. (1994), Textile Quality – Physical Methods of Product and Process Control, Nuova Oflito, Torino. Dragcˇevic´, Z. and Potocˇic´, V.M. (1999), “Svecˇane akademske odore sveucˇilisˇta u Zagrebu”, Sveucˇilisˇni vijesnik, Vol. 45 Nos 3-4, pp. 135-40. Hargreaves-Mawdsley, W.N. (1963), A History of Academical Dress in Europe, Oxford University Press, London. Horvat, R. (1942), “Prosˇlost grada Zagreba”, Kulturno historijsko drusˇtvo Hrvatski rodoljub, Zagreb. Kawabata, S. and Niwa, M. (1991), “Objective measurement of fabric mechanical property and quality: its application to textile and clothing manufacturing”, International Journal of Clothing Science and Technology, Vol. 3 No. 1, pp. 7-17. Rogale, D., Dragcˇevic´, Z. and Bakran, T. (2001), “The application of centrally expanding pattern on sewing automata for decorative embroidery”, in Katalinic, B. (Ed.), Proceedings of the 12th DAAAM Symposium, Jena, October, DAAAM International, Vienna, pp. 405-6.
The Emerald Research Register for this journal is available at www.emeraldinsight.com/researchregister
The current issue and full text archive of this journal is available at www.emeraldinsight.com/0955-6222.htm
The advance engineering methods to plan the behaviour of fused panel Simona Jevsˇnik and Jelka Gersˇak
The advance engineering methods 161
Faculty of Mechanical Engineering, Textile Department, University of Maribor, Maribor, Slovenia, Europe, and
Ivan Gubensˇek Higher profesional school Celje, Celje, Slovenia, Europe Abstract Purpose – The purpose of this paper is to analyse some mechanical properties and parameters of drapability using different methods from two different points of research area: knowledge bases and numerical modelling using the finite element method. Design/methodology/approach – The approach consists of analysing some mechanical properties and parameters of drapability using different methods from two different points of research area: knowledge bases and numerical modelling using the finite element method. The knowledge bases, named FP_B-1 and FPO_B-2, were used to analyse the bending rigidity of fused panels. The numerical model of fused panel NMFP is used to analyse parameters of drapability. Findings – Based on the analyses of bending rigidity and draping of fused panels the conclusions indicate the significance of interaction between mechanical properties and parameters of drapability of the fused panel to garment appearance. Furthermore, the methods used present a computer approach to the study of the fused panel properties important for the computer-based engineering and the presentation of real behaviour of all aspects of clothes. Practical implications – This numerical model of a fused panel enables a 3D observation of this aspect of clothes’ which is a behaviour, very important contribution to the computer planning of the behaviour of produced clothes. Originality/value – A better understanding of how to construct fused panels in clothing. Keywords Finite element analysis, Knowledge management, Clothing, Modelling Paper type Research paper
1. Introduction In the past extensive research work has been done in connection with analysing the fabric behaviour of cloth (Collier and Collier, 1991; Breen et al., 1994; Jeong and Phillips, 1998). The main eternising orientation has been the investigation of fabric drape of produced cloths because its construction is very specific and unpredictable. The behaviour of the fused parts of clothes has not been particularly investigated. In this case there is a minimum of two non-homogeneous and anisotropic textile materials usually connected using adhesive. In the electronic business of clothes sales, the progress ahead can be by the simulated virtual presentation of clothing. Most problems, the experts meet when planning a virtual presentation at a fashion show are how to assimilate the real behaviour of all incorporated materials into working clothes model (Breen et al., 1994; Stylios et al., 2000).
International Journal of Clothing Science and Technology Vol. 17 No. 3/4, 2005 pp. 161-170 q Emerald Group Publishing Limited 0955-6222 DOI 10.1108/09556220510590858
IJCST 17,3/4
There are a lot of different ideas and models of how to treat or present a fabric. All have one common aim; to get closer to the real fabric behaviour clothes. The purpose of this paper is to present two different methods for analysing the bending rigidity and drapability of a fused panel. The achieved results of both methods present one important contribution when planning the behaviour of a fused panel in produced clothes.
162 2. Evaluating the behaviour of fused parts of clothes The behaviour of textile materials clothes depends on the interactions of the construction parameters and mechanical properties of all incorporated components i.e. shell fabrics as well as auxiliary materials. The shell fabric has a more or less similar behaviour over the whole area; therefore, it does not usually assure suitable 3D form in all parts of the clothes. It is necessary to fuse with fusible interlining in order to get a suitable form of clothing. If the shell fabric is a component that has a direct influence on the appearance of clothes, then the fusible interlining is the component that improves the aesthetic and applicable properties of the clothes. 2.1 Bending properties of fused panel Arising from the fact that the kinds and quality of fabrics as a basic element of clothes is determined by the planning and modelling could be considered as constant (Gersˇak, 1997). This means that the required properties of the produced garment depend on the mechanical properties of the shell fabric as well as the fusible interlining. Previous investigations of fused panels have shown that it is not possible to obtain by summing the mechanical properties of fusible interlining and shell fabric, but it is necessary to investigate every fused panel as a separate unit (Jevsˇnik, 1999; Shishoo et al., 1971; Jevsˇnik and Gersˇak, 2000). Fused panels, fused with fusible interlining, applying thermoplast in dots, exhibit lower bending rigidity values than those, where fusible interlining is used with randomly applied thermoplast. Likewise, more thermoplast on the backing fabric means higher bending rigidity (Jevsˇnik and Gersˇak, 2004).The bending hysteresis of fused panels shows clearly defined asymmetry, bending in the direction of warp and weft, as well as the fact that fused panels are harder to bend when fusible interlining lie on their convex sides (Figure 1). The bending rigidity of a fused panel is, on average, 4-10 times higher than the bending rigidity of individual components (Jevsˇnik, 1999). 2.2 Drapability of fused panel Drapability is a phenomenon which arises when a fused panel or shell fabric hangs down over a circular pedestal without the use of external force. After some time the bending and shear deformation appears that results to series the folds. The characteristic examples in practice when describe drape are: how the curtain hangs, how the fully flared skirt looks, how the tablecloth drapes over the table and so on. The drapability of textile materials can be established using subjective procedures, where experts’ abilities, and the experience of professionals is essential and objective when measuring the drape parameters like drape coefficient, depth of folds, number of folds and distribution of folds. Drape coefficient describes any deformation between deformed and non-deformed fabrics (Figure 2).
The advance engineering methods 163
Figure 1. Characteristic bending hysteresis of fused panel
Figure 2. Determination of drape coefficient
It is the ratio of the ring, between radius R1 of the fabric and radius R2 of the disc holding the fabric which is covered by the projected shadow and it can be determined by Leung et al. (2000): CD ¼
Sp 2 pR21
pR22 2 pR21
ð1Þ
where CD is the drape coefficient, Sp the projection area of draped specimen including the part covered by the horizontal disk in mm2, R1 the radius of horizontal disk in mm and R2 the radius of non-deformed specimen in mm. The interpretation of drape coefficient value is connected with the number, form, amplitude, and distribution of folds and their positions according to the weft and warp directions. High drape coefficient value means that the fabric is stiff, and therefore, it could be difficult to re-form. Alternatively, low drape coefficient value means easier reform and at the same time, better adaptation of the fabric to the shape of the clothes. The shape and number of folds are depending on fullness and fabric stiffness. A fabric with higher stiffness has larger and wider folds and less stiff fabrics have narrower folds. The effect of various properties, related to both yarn and fabric, were linked to drapability. The attempt to use Kawabata evaluation system for fabrics to obtain data of fabric and calculate drape properties for writing software for a 3D simulation of garment are done by Marooka and Niwa. There is equation for determining the drape coefficient from the bending rigidity and the weight of the fabric, is shown as (Leung et al., 2000):
IJCST 17,3/4
164
DC ¼ 5:1 þ 115:0
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ffi 3 3 3 B90 =W þ 131:1 B0 =W þ 1:2 B45 =W
ð2Þ
where B90 is the bending rigidity along warp, B0 along weft and B45 in bias directions, and W is the fabric weight. Because the equation has incorporated only bending rigidity and weight, it is not useful for a wide range of fabric. The drapability of fused penal is amore complex phenomena than drapability of fabric. The end effect of fused panel drape is dependent on orientation of shell fabric and of fusible interlining as well as the constructional and mechanical properties of shell fabric and fusible interlining separately. The number, shape, and distribution of folds of fused panels in fitting, e.g. draping, are defined by the impact of fabric construction parameters and parameters of fusible interlining, as well as by associated mechanical properties of both of them (Jevsˇnik and Gersˇak, 2003). 3. Experimental work The investigation is divided into two groups. First the bending rigidity of fused panel were predicted in the warp and weft directions with two knowledge bases named FP_B-1 and FP_B-2 and secondly, for the equal fused panel the drapability with different drape parameters was determined using the finite element model NMFP. The following drape parameters were determined by means of a drape meter: drape coefficient, number of folds, minimum and maximum amplitude, and length between folds. A research model to plan the bending rigidity and drapability of a fused panel is shown in Figure 3. The knowledge-bases named FP_B-1 and FP_B-2 for predicting the bending rigidity of a fused panel were constructed using 480 examples. The fused panel composing the learning set was made from a shell fabric suitable for upper-clothes and they were fused with different, but suitable fusible interlining. The bending rigidity
Figure 3. Research model to plan the bending rigidity and drapability of fused panels
was determined using FAST measuring system. The regressive trees inductive system (RETIS) program package for machine learning from examples was used for predicting the bending rigidity of a fused panel (Karalicˇ, 1999; Jevsˇnik, 2000). The numerical model of the fused panel drape NMFP was constructed using shell finite elements. The fused panel is modelled as a two-layer laminated fabric, with the bond supposed to be uniformed all over the surface. The fused panel draping model is monitored, in the same way as the experiment, under the loading of its own mass. The following suppositions were considered: that the fabric and fusible interlining are a continuum with homogeneous and orthotropic properties, and the fabrics behaviour within low loading is linear. The numerical model for the numerical analysis of drape was constructed on basis of the Cusik drape meter method (Operatior’s guide, 1993). The drape model had been discretisized with 240 elements. Thin 3D shell elements marked S9R5 were used for this purpose. Those parts of the samples draped over the pedestal were described as 120 and the remaining 120 finite elements described the sample lying on the pedestal. The ABAQUS program package for numerical analysis was used (ABAQUS, 2000). The Newton-Raphson’s iterative method was used for equation solving. The achieved results of knowledge-bases named FP_B-1 and FP_B-2 for predicting the bending rigidity of a fused panel and the NMFP numerical model of the fused panel drape was compared with the experimental data.
4. Results According to the above mentioned the results are presented in the following ways: (1) bending rigidity of fused panel using the knowledge bases named FP_B-1 and FP_B-2; and (2) drapability of fused panel using the numerical model NMFP.
4.1 The results of fused panel bending rigidity The knowledge-bases named FP_B-1 and FP_B-2 were presented in the form of regression trees, which could be used for predicting the bending rigidity of a fused panel, as well as analyzing the dependence between particular parameters. The interdependence analyses between particular parameters in regression trees are performed using graphical and textual forms (Figure 4). It commences at the root of the regression tree using a graphical interpretation of the rules and then we simply follow the value in the nodes until the value in a leaf is reached. The predicted value was then compared with the measured value of bending rigidity of the fused panel in a warp and weft direction (Figure 4). 4.2 The results of fused panel drape The experimentally obtained drape shapes of the fused panel carried out by the testing equipment drape metre Cusik are shown in Figure 7, while Figure 6 shows the vertical projection, as well as lateral sight, of the numerically simulated drape shapes of the fused panels.
The advance engineering methods 165
IJCST 17,3/4
166
Figure 4. The part of regression tree for a bending rigidity in a warp direction
5. Discussion The bending rigidity and drapability of fused panel depends on the interaction between the fusing parameters, and the construction parameters of a fabric and the fusible interlining as well as its mechanical properties. The method of machine learning from examples using the RETIS programme package provides a successful technique for predicting the bending rigidity of new fused panels examples, because there is very good agreement (Figure 5) between the measured and predicted values. Furthermore, from regression trees which predict bending rigidity in the warp and weft directions of the fused panels also carry out
Figure 5. Correlation between predicted and measured values of bending rigidity
The advance engineering methods 167
Figure 6. Numerically obtained drape shape of fused panels
IJCST 17,3/4
168
Figure 7. Experimentally obtained drape shape of fused panel
analysis of shell fabric construction parameters, as well as fusible interlining that has a significant influence on the properties of fused panels (Figure 5). Every fused panel could be analyzed individually from a regression tree, and so it is possible to get information which parameters have an influence on bending rigidity (Figure 4). The results obtained with numerical simulation of fused panel drape are shown in Figure 6, while Figure 7 shows the obtained drape shape of the fused panel given by experiment, using a Cusik drape metre. Analysis of the achieved results (Figures 6 and 7) shows the similarity in drape form between the fused panels according to the experimental drape using Cusik drape metre and the numerical simulation of the ABAQUS program package. 6. Conclusion Using this knowledge basis and numerical method for analyzing the bending rigidity and drapability of a fused panel presents many advantages over the classical manner. Using regression trees it is possible at the same time, to analyse the influential parameters on bending rigidity of fused panels, as well as the prediction. The numerical drapability model presents beside 2D what is characteristic for an experiment as well as 3D observation of fused panel drape. This numerical model of a fused panel enables a 3D observation of fused part clothes’ behaviour, which is a very important contribution to computer planning the behaviour of produced clothes. References ABAQUS (2000), User Manuel Version 6.2, Hibbitt, Karlsson & Sorensen, Volume I, II, III Inc.,Plymouth, MN. Breen, D.E. et al., (1994), “A particle-based model for simulating the draping behaviour of woven cloth”, Textile Research Journal, Vol. 64 No. 1, pp. 663-85. Collier, R. and Collier, B.J. (1991), “Drape prediction by means of finite element analysis”, Journal of Textile Institute, Vol. 82 No. 1, pp. 96-107. Gersˇak, J. (1997), “Objektivno vrednovanje fiksiranih dijelova odjec´e”, Tekstil, Vol. 46 No. 4, pp. 193-203. Jeong, Y.J. and Phillips, D.G. (1998), “A study of fabric drape behaviour with image analysis. Part II: the effects of fabric structure and mechanical properties on fabric drape”, Journal of Textile Institute, Vol. 89 No. 1, pp. 70-9. Jevsˇnik, S. (1999), “The selection of fusible interlining and prediction of the properties of fused garment part with knowledge base system”, Master thesis, Faculty of Mechanical Engineering, University of Maribor, Maribor. Jevsˇnik, S. (2000), “Predicting mechanical properties of fused panel”, Fibres and Textiles in Eastern Europe, Vol. 8 No. 4, pp. 54-6. Jevsˇnik, S. and Gersˇak, J. (2000), “Use of a knowledge base for studying the correlation between the constructional parameters of fabrics and properties of a fused panel”, International Journal of Clothing Science and Technology, Vol. 13 Nos 3/4, pp. 186-97. Jevsˇnik, S. and Gersˇak, J. (2003), “The analysis of fused panel drape using the finite element method”, in Gersˇak, J., Stjepanovicˇ, Z., Zˇunicˇ Lojen, D. and Rudolf, A. (Eds), paper presented at the 4th International Conference IMCEP’2003, Innovation and Modelling of Clothing Engineering, Faculty of Mechanical Engineering, Maribor, pp. 78-85. Jevsˇnik, S. and Gersˇak, J. (2004), “Modelling a fused panel for a numerical simulation of drape”, Fibres and Textiles in Eastern Europe, Vol. 12 No. 1, pp. 47-52.
The advance engineering methods 169
IJCST 17,3/4
170
Karalicˇ, A. (1999), “Avtomatsko ucˇenje regresijskih dreves iz nepopolnih podatkov”, Master thesis, University of Ljubljana, Ljubljana. Leung, K.Y.C. et al. (2000), “Draping performance of fabrics for 3D garment simulation”, paper presented at the 29th Textile Research Syposium, Mt Fuji, in Kawabata, S. (Ed.), pp. 169-75. Operatior’s guide (1993), “Cusik drape tester”, Model 165, James H. Heal&Co. Ltd, Halifax. Shishoo, R. et al. (1971), “Multilayer textile structures. Relationship between the properties of a textile composite and its components”, Textile Research Journal, Vol. 12, pp. 669-79. Stylios, G. et al. (2000), “An investigation into engineering of the drapability of fabric”, in Gersˇak, J. (Ed.), paper presented at the 3rd International Conference IMCEP’2000, Innovation and Modelling of Clothing Engineering, Faculty of Mechanical Engineering, Maribor, 1997, pp. 88-95.
The Emerald Research Register for this journal is available at www.emeraldinsight.com/researchregister
The current issue and full text archive of this journal is available at www.emeraldinsight.com/0955-6222.htm
Investigation of the performance of linings
Performance of linings
Fatma Kalaoglu and Binnaz Meric Textile Engineering Department, Istanbul Technical University, Istanbul, Turkey
171
Abstract Purpose – To investigate the performance of linings in clothing. Design/methodology/approach – A total of 24 lining fabrics were produced in different constructions. 150 denier 350 twist/cm filament polyester warp yarn was used for all of the fabrics. Two different weft yarns (textured, filament) were used to produce lining fabrics in three different densities. Findings – In the garment sector, lining performance is highly important for the manufacture of proper quality garments. The main problem of linings during usage is seam slippage for some constructions. Research limitations/implications – Fabric constructions were chosen as warp rips, weft rips, ripstop (rips both in warp and weft direction) and plain weave. Seam slippage, bending behaviour, crease recovery angle and comfort properties of the linings were measured and the results evaluated. Originality/value – The paper contributes to understanding the performance of linings. Keywords Fabric testing, Yarn preparation processes, Measurement Paper type Research paper
1. Introduction Linings are functional parts of garment, being used to maintain the shape of the garment, to improve the hang and comfort by allowing it to slide over other garments, to add insulation and to cover the inside of the garment of complex construction, to make it neat. They are usually made from polyester, polyamide, acetate and viscose for the use where slippery fabric is required for suits, jackets, dresses or skirts. For the garments where decoration or warm handle is required cotton, polyester/cotton, wool and wool mixtures can also be used (Ruth et al., 1995). Performance expectations of linings vary with product type and end use. The factors that affect the quality and performance of linings include fabric properties, design and structure of the lining, compatibility with other materials and garment structure. The factors that contribute to the performance of linings are fibre content, fabrication, finishes hand and drape. The choice of lining materials depends upon the intended end use, the characteristics of the piece goods, and the performance requirements for the product, the quality level, and cost. Properly designed and constructed linings may extend the wear life of garments by absorbing the stress of the body movement and activity, preventing stretching and overextension of the shell fabric. Linings also extend garment life by preventing hanger stress and body contact with the shell fabric. Linings make a barrier between the body and the garment, therefore, preventing absorption of perspiration and body oils that may deteriorate or stain shell fabrics (Carr and Lutham, 1994).
International Journal of Clothing Science and Technology Vol. 17 No. 3/4, 2005 pp. 171-178 q Emerald Group Publishing Limited 0955-6222 DOI 10.1108/09556220510590867
IJCST 17,3/4
172
The hand of the linings should complement the aesthetics of the shell fabric and provide comfort to the wearer. Linings need to be flexible and soft unless a firmer fabric is needed for support. Linings with a stiff heavy hand may be strong and provide support, but they may alter the drape of the shell fabric and the fit of the garment. A stiff garment tends to push up instead of flexing with body movement. Linings may provide tactile comfort, garment fit and thermal comfort for a garment. Tactile comfort is provided by the lining fabric if it is smooth, absorbent or of pleasant hand. Linings improve garment fit by absorbing the stress of tight fit and movement, thus allowing the outer garment to hang freely and relaxed. Linings of smooth filament yarns provide the ease of movement by reducing friction with other garments. With slip ease, dressing is easier and garments are less restrictive. Linings may also be used to provide thermal comfort. Lining materials may be selected for breathability or insulative properties. The comfort of water resistant garment may depend on the absorbency and breathabilty of the lining material used. For example, nylon mesh may be used for the lining to allow for more air circulation. Fibre content is a major determinant of the type of processing, aesthetics, performance and durability of the linings. The synthetic fibre nylon and polyester are used for durable lightweight linings. Nylon and nylon/spandex blends are used in the linings for active sportswear (football pants, swimsuits, jogging shorts etc.) providing high strength, elasticity, and minimum weight and bulk. Synthetic fibres are clammy, do not breathe, and allow the build up of static electricity. Fibre modifications with wicking properties, which increase comfort and reduce static build up, are sometimes used for lining fabrics. The weight of the lining fabric affects wearing comfort, thermal comfort, compatibility, opacity, hand, and drapeability. Generally, lighter weight linings may be selected for their support and shaping characteristics. Linings made from polyester, polyamide or viscose can, in some constructions, which have long floats, such as satin, be prone to seam slippage. Since these fabrics also tend to fray easily, the loss of seam allowance by fraying can contribute to seam slippage. Fabric construction and yarn material are important parameters that affect the performance of linings. Plain, warp rip, weft rip and warp-weft rip constructions are chosen for this study. The formability of the lining fabrics is also very important, as fabric bending rigidity is thus evaluated. Air permeability and water vapour resistance properties of the fabrics are important for comfort properties. This is why seam slippage, seam strength, bending rigidity and comfort properties of polyester linings are analysed in this paper. 2. Experimental A total of 24 lining fabrics were produced in different constructions. A 150 denier 350 twist/cm filament polyester warp yarn was used for all of the fabrics. Two different weft yarns of 100 denier (filament and texturised) were used to produce lining fabrics in three different weft densities, while the warp density was kept constant at 64 warp/cm. Fabric constructions were chosen as warp rips, weft rips, ripstop (rips both in warp and weft direction and plain weave). The properties of the fabrics are shown in Table I. Five samples were prepared in both warp and weft direction according to the BS 3320 for seam slippage. Eight kilograms of force was applied to the sewn samples on the Instron apparatus, and seam slippage of the fabrics was measured. Afterwards, seam strength and elongation properties were measured.
Fabric code 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
Weft yarn count (denier)
Weft yarn type
Weave type
150 150 150 150 150 150 150 150 150 150 150 150 150 150 150 150 150 150 150 150 150 150 150 150
Textured Textured Textured Filament Filament Filament Textured Textured Textured Filament Filament Filament Filament Filament Filament Textured Textured Textured Textured Textured Textured Filament Filament Filament
Warp rips Warp rips Warp rips Warp rips Warp rips Warp rips Weft rips Weft rips Weft rips Weft rips Weft rips Weft rips Ripstop Ripstop Ripstop Ripstop Ripstop Ripstop Plain Plain Plain Plain Plain Plain
Weight (g/m2)
Fabric thickness (mm)
Weft cover factor
Weft density picks/cm
179.7 162.1 165.6 180.4 163.9 156.5 177.2 167.7 162.7 172.4 162.7 155.6 172.5 161.2 154.6 177.4 165.2 157.3 178.5 167.1 162 177 162.5 154.8
0.51 0.42 0.52 0.45 0.34 0.34 0.44 0.46 0.46 0.35 0.35 0.34 0.41 0.36 0.35 0.53 0.50 0.46 0.38 0.39 0.40 0.31 0.31 0.31
9.04 6.26 6.96 9.04 5.57 4.87 6.96 5.57 4.87 6.96 5.57 4.87 6.96 5.57 4.87 6.96 5.57 4.87 8.67 6.96 6.26 8.67 6.96 6.26
26 18 20 18 20 26 20 14 16 20 16 14 20 16 14 20 16 14 25 20 18 25 20 18
The bending length of the fabrics was measured using the Fast tester and bending rigidity of the fabric was calculated. Air permeability of all the fabrics was measured by WIRA Instrumentation Air Permameter, according to the BS5636. For water vapour resistance and heat resistance measurements, the fabrics of 20 weft/cm density was chosen. Water vapour resistance and heat resistance were measured using the MTNV Sweating hot plate tester, according to ISO 11092. 3. Results The results of seam slippage and seam strength of the fabrics are shown in Table II. Due to high seam slippage, seam strength values in weft direction cannot be measured. The results are analysed according to the variance analysis and the most important parameter is found to be weft density. Yarn type and weave type are also important. The results are shown in a graphics form, in from Figures 1-8. It can be easily seen that weft seam slippage is higher than warp seam slippage for all of the fabric types, because warp density is higher than weft density. Among the four weave types, warp rip, warp-weft rip and plain weaves, the lowest seam slippages in warp direction is recorded for the weft rip fabrics. In the case of the weft rip fabrics, as the float on the weave construction is in the weft direction, the seam slippage in the weft direction is more prominent than with the warp rip fabrics. Therefore, weft rip fabrics show maximum slippage and warp rip fabrics minimum slippage in the weft direction.
Performance of linings
173
Table I. The properties of lining fabrics
IJCST 17,3/4 Fabric code Weave
174
Table II. Seam slippage test results
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
Warp rip
Seam slippage Seam strength Seam elongation Weft Warp Weft Warp Weft Warp Weft Weft yarn (thread/cm) (mm) (mm) (kN) (kN) (per cent) (per cent) Textured Filament
Weft rip
Textured Filament
Warp-weft rip Textured Filament Plain weave
Textured Filament
26 20 18 26 20 18 20 16 14 20 16 14 20 16 14 20 16 14 25 20 18 25 20 18
1.96 2.78 4.2 4.42 4.06 2.24 3.28 5.18 4.1 2.4 4.26 5.42 3.02 3.82 4.2 3.78 3.08 3.08 2.04 3.32 3.9 3.44 3.92 4.0
2.44 6.9 5.78 12.6 8.1 3.0 6.7 18.24 10.85 6.84 31.24 34.7 5.14 13.16 16.7 6.5 15.24 23.56 2.9 6.48 12.58 3.34 8.52 15.3
0.6289 0.5420 0.5318 0.6523 0.5426 0.5491 0.5850 0.5247 0.5579 0.6070 0.5325 0.5443 0.5588 0.5851 0.4737 0.5617 0.5329 0.5208 0.6059 0.5578 0.4640 0.6084 0.5875 0.5147
0.5575 – – 0.5786 – – – – – – – – – – – 0.484 – – 0.526 – – 0.557 – –
29.71 22.38 24.18 29.56 22.97 21.87 28.53 29.28 32.28 23.86 21.76 22.99 21.96 23.17 20.31 24.82 24.85 23.79 27.41 24.60 22.91 23.27 20.40 21.50
38.92 – – 41.68 – – – – – – – – – – – 51.08 – – 38.55 – – 40.74 – –
Figure 1. Seam slippage of the warp rips
The figures also show that seam slippage is higher with filament weft yarns, due to their slippery surface. The results of measuring bending rigidity, air permeability, water vapour resistance and heat resistance are shown in Table III. The formability of the fabric depends on the bending rigidity and extensibility. Formability may be increased by increasing the bending rigidity of the fabric or its extensibility. Low bending rigidity leads to poor shape retention. Bending rigidity has
Performance of linings
175 Figure 2. Seam slippage of the warp rips
Figure 3. Seam slippage of the plain weaves
Figure 4. Seam slippage of the plain weaves
an important effect on handle, appearance in wear and tailoring performance of the fabric. Bending rigidity is related to fibre type and fineness, fibre interaction, yarn count/twist, cover factor, yarn crimp. Bending rigidity values in Table III, indicate that the fabrics with filament weft yarns have higher bending rigidity than the fabrics woven with texturised yarns, as the bending rigidity of a fabric depends upon the bending rigidity of the threads used in its manufacture. The bending rigidity of all of the fabrics increases with the increase in weft density and therefore, corresponds to the cover factor of the fabrics.
IJCST 17,3/4
176 Figure 5. Seam slippage of the weft rips
Figure 6. Seam slippage of the weft rips
Figure 7. Seam slippage of the warp-weft rips
Thermal comfort is important for lining fabric, as is for the shell fabric of the garment. Four properties are suggested as critical for the thermal comfort of a clothed body: thermal resistance, air permeability, water vapour permeability and liquid water permeability. The first three are measured in this study. Air permeability results in Table III show that the results are related to the cover factors of the fabric. When the cover factor increases, the air permeability of the fabric is decreased. Although texturised yarns are bulkier than filament yarns, the fabrics with texturised weft yarns exhibit higher air permeability than the fabrics with filament weft yarns. This is true for all of the weave types. Among the four weave types, warp-weft rip (ripstop) fabrics
Performance of linings
177 Figure 8. Seam slippage of the warp-weft rips
Fabric code 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
Bending rigidity (mg-cm) weft
Air permeability (Lt/m2 s)
97.97 47.41 49.45 202.76 123.37 97.33 91.75 61.25 42.76 247.38 334.69 258.01 186.19 147.3 96.5 90.27 51.41 46.97 111.2 67.17 36.42 563.67 156.55 103.97
24.93 77.82 92.85 9.51 36.74 61.68 75.79 213.25 363.52 40.35 98.43 199.47 39.69 132.87 235.25 92.85 218.83 336.29 23.29 71.19 114.83 11.48 34.78 42.65
Water vapour resistance m2 Pa/Watt
Heat resistance m2 C/W
3.611
0.011
3.914
0.002
3.107
0.010
2.951
0.009
3.518
0.013
2.823
0.006
3.059
0.004
have the highest air permeability, due to longer floats in both warp and weft directions. However, in plain fabrics, due to high crossover points of the warp and weft threads, minimum air permeability values are recorded. Water-vapour resistance generally depends on the air permeability of the fabrics and represents its inability to transfer perspiration coming out of the skin. Water vapour resistance is the most important parameter in determining thermal comfort. Heat resistance of the fabrics is higher for the fabrics with textured weft yarns than for the ones with filament weft yarns, due to their bulky construction. From the results in Table III, it can be seen that fabric
Table III. Bending rigidity, air permeability and water vapour resistance test results
IJCST 17,3/4
178
thickness influences heat resistance of the lining fabrics. This result is confirmed by the findings of other researchers. 4. Conclusion In the garment sector, lining performance is highly important for the manufacture of proper quality garments. The main problem of linings during usage is seam slippage for some constructions. In this study, the effects of weft density, yarn construction and weave type on seam slippage performance of lining fabrics are analysed. A total of 24 different lining fabrics were produced in different constructions. Seam slippage results are parallel to the findings in literature (Wojonska and Filipowska, 1973; Burtonwood and Chamberlein, 1967; Galuszynski, 1985; Garner, 1955). Weft density is the most important parameter that influences seam slippage of the lining fabrics. Filament yarns cause more slippage opening than texturised yarns, due to their smooth surface. Weave structure also impacts seam slippage of the fabrics, which is related to the floats in the weave construction. The bending rigidity of all of the fabrics increases with the increase in weft density and can be associated with the cover factor of the fabric. The fabrics woven with filament weft yarns have higher bending rigidity than the fabrics woven with texturised yarns, primarily due to the high bending rigidity of the filament yarns used. Comfort properties are related to the construction of the fabrics. When the cover factor increases, the air permeability of the fabrics decreases. Fabric thickness influences the heat resistance of the lining fabrics. This result is confirmed by the findings of other researchers (Behera et al., 1997). References Behera, B.K., Ishtiaque, S.M. and Chand, S. (1997), “Comfort properties of fabrics woven from ring-rotor- and friction-spun yarns”, Journal of the Textile Institute, Vol. 88 No. 3, pp. 255-64. Burtonwood, B. and Chamberlein, N.H. (1967), “The strength of seams in woven fabrics, Part II”, Clothing Institute Technological Report, No. 18. Carr, H. and Lutham, B. (1994), The Technology of Clothing Manufacture, Blackwell Scientific Publisher, London. Galuszynski, S. (1985), “Some aspects of the mechanism of seam slippage in woven fabrics”, Journal of the Textile Institute, Vol. 76, p. 425. Garner, W. (1955), Seam Slippage of Linings Manufacturing Clothings. Ruth, E., Glock, G. and Kunz, I. (1995), Apparel Manufacturing, Prentice-Hall, Hemel Hempstead. Wojonska, J. and Filipowska, B. (1973), “Slippage of threads in the seams of cotton fabrics”, Przeglad Wlokienniozy, Vol. 27 Nos 7/8, pp. 364-9.
The Emerald Research Register for this journal is available at www.emeraldinsight.com/researchregister
The current issue and full text archive of this journal is available at www.emeraldinsight.com/0955-6222.htm
Equipment and methods used to investigate energy processing parameters of sewing technology operations
Equipment and methods used
179
Dubravko Rogale, Igor Petrunic´, Zvonko Dragcˇevic´ and Snjezˇana Firsˇt Rogale Department of Clothing Technology, Faculty of Textile Technology, University of Zagreb, Zagreb, Croatia Abstract Purpose – The equipment for computerised measuring of electrical power and energy is presented, adapted to the needs of investigating processing parameters of garment sewing operations. Design/methodology/approach – The method of measuring the energy necessary to run the sewing-machine driving electrical motor is also presented, correlated to the stitching speed in joining a straight seam in a single, two, or three, segments. Electrical energy consumption is analysed as dependent on the stitching speed, varying the number of stitches in the seam. Findings – The investigations described have shown the impact of the method of work applied and the effect of the changes in garment sewing operation in processing parameters on the level of electrical energy consumed by the sewing-machine drive electrical motor. A new measuring method has been introduced in garment engineering, aimed at predicting electrical energy consumption in garment sewing operations, thus opening a completely new field of investigation in the area of garment technologies. Originality/value – A method of calculating the energy processing parameters of sewing operations. Keywords Garment industry, Electrical measurement, Energy consumption Paper type Research paper
1. Introduction One of the key tasks of garment engineering is to find and establish new processing parameters, important for the analysis of the technological operation structure and to optimise manufacturing processing in garment technology. Energy consumption is an important field of optimisation, as, besides varying costs (material, labour), the costs of energy are one of the key factors in calculating manufacturing costs, thus determining the price of any article of clothing. It is well known that energy costs constitute some 10-15 per cent of the overall manufacturing costs in garment industry. However, energy consumption has been neglected until now in the attempts to analyse and optimise garment manufacturing processes. Consumption of electrical energy is of high importance in garment sewing processes and selection of a proper method of work can result in reducing the time necessary to perform the operations, simplification of the operation structure, higher average stitching speed and higher sewing machine utilisation (Firsˇt Rogale et al., 2003). All these yield higher productivity and directly impact optimal and efficient electrical energy consumption. Contemporary sewing machines, with driving electrical motors of
International Journal of Clothing Science and Technology Vol. 17 No. 3/4, 2005 pp. 179-187 q Emerald Group Publishing Limited 0955-6222 DOI 10.1108/09556220510590876
IJCST 17,3/4
180
a new design, also significantly contribute to rational electrical energy consumption in the process of sewing. Investigations and rationalisation of electrical energy consumption in garment manufacturing processes, especially from the point of view of performing particular operations, as well as proper workplace design and calculating needs for electrical energy in new and restructured manufacturing processes gain in importance, particularly as the global tendencies are aimed at rational usage of energy and energy savings. Determining energy consumption in individual operations has not been possible until now, as there has been no adequate measuring equipment available, suited for garment manufacturing processes. This is why the Department of Clothing Technology, Faculty of Textile Technology, has designed, manufactured and calibrated new computerised equipment for measuring the parameters of electrical energy consumption on sewing machines. Apart from the application in research, and under laboratory conditions, this measuring system can be used, thanks to its design characteristics, under real in-plant conditions as well, where electric motor drives are actually used. 2. Equipment used to measure electrical energy processing parameters The new computerised measuring system KVP-1, and the software package JNJ-1, used with it, has been developed for the purpose of investigating electrical energy consumption in garment manufacturing processes. It is able to measure electrical energy parameters of the sewing machine electrical motors, linked to a single-or three-phase electrical system. The new measuring equipment can determine electrical energy consumption at all the three-phase conductors’ linked, electrical power at each phase, electrical current at each phase, tension at each phase and the simultaneity coefficient (Rogale and Petrunic´, 2002). The measuring system KVP-1, as shown in Figure 1(a), consists of the following: . a digital three-phase for measuring power and energy consumption – Energy Control EC400; . three digital wattmeters – SA9106A; . three current measuring transformers – TZ77, with adequate measuring rectifiers and operational amplifiers – TL064; . three tension measuring transformers; . an A/D converter – BMC mmeter-4; and . a personal computer. Configuration of the measuring system is highly complex, consisting of 11 individual measuring instruments, so it is necessary to calibrate and synchronise their joint work, perform measurements, present measuring results and store the data into data bases (Tilli, 1995; Schweitzer, 2001). An integral part of the KVP-1 measuring system, apart from the hardware described, is also the JNJ-1 software package, which has two functions: (1) to synchronise and measure; and (2) to analyse.
Equipment and methods used
181
Figure 1. Measuring equipment used to determine electrical energy processing parameters: (a) the KVP-1 measuring system; and (b) processing window of the software package JNJ-1
IJCST 17,3/4
182
NextView, a basic software package by BMC, is used to synchronise, measure and measuring data acquisition. For the purposes of the system described the software has been adapted slightly (JNJ-1) (Shoup, 1984). Figure 1(b) shows the screen with the measuring processing window with three analogous instruments, the purpose of which is to measure sewing machine stitching speed, driving electrical motor tension and its current. Such a representation of the analogous instruments is especially adequate for visual monitoring of the changes in energy processing parameters, related to stitching speed reached in performing machine-hand sub-operations of sewing. The digital part is positioned below the analogous one, and is used to calibrate and check correctness and precision of the measuring system. Measuring data from both analogous and digital part are jointly presented continuously on-screen, as a graphic presentation of the changes in the driving electrical motor current. Right of it is a vertical column representing power of the current in the motor, with borderlines that indicate minimum and maximum current power reached in the motor. 3. Measuring system and experimental work Practical measurements of energy parameters for a sewing machine driving electrical motor are presented, as dependent upon the processing parameters and the method of work applied. The measurements were performed on the Brother universal sewing machine, designated DB2-B755-403A Mark III, equipped with a processing microcomputer F-40, which enables pre-programmed complex segments of the sub-operations to be precisely performed. The machine was driven by a single-phase Brother AC servo motor, series MD 601 type LD 471C, power 400 W. Its key property is that it idles when the sewing machine main shaft idles and starts the rotation at the moment of actuating the machine mechanism by pressing the pedal regulator. Since the electrical motor starts only when the main shaft is actuated (when performing the machine-hand sub-operations), and is idle for the rest of the operation, considerable savings of electrical energy can be achieved. This type of drive electrical motors enable high degree of automation, highly precise speed regulation (using the microcomputer), which results in much more optimal sewing operation structure, with considerably less energy spent for its work. The necessary power of sewing machine drive electrical motor, as dependent upon the rotational speed of the main shaft, was measured using a digital wattmeter SA 9106A, a part of the KVP-1 measuring equipment, infrared reflexive measuring transformer, a part of the system for automatically measuring processing parameters and garment manufacturing operation structures (MMPP system) (Rogale and Dragcˇevic´, 1998), tachogenerator HAROWE SERVO CONTROLS and microprocessor digital tachometer Mayer & Wonisch (Rogale et al., 2003). Figure 2 shows the arrangement of the measuring equipment used in determining energy parameters for sewing machine electrical motors. Measuring samples used in laboratory measurements were made of cotton/PES (70/30) blend, in twill weave K 2/1. Warp yarn density was 396 threads/10 cm, weft density 246 threads/10 cm (HRN F.S2.013). Warp yarn count was 39.3 tex, weft yarn count 37.9 tex. The fabric was 0.40 mm thick (HRN F.S2.021). Surface mass of the absolutely dry sample was 248.65 gm2 2, while the cured sample mass was 261.08 gm2 2 (HRN F.S2.016).
Equipment and methods used
183
Figure 2. Linking the KVP-1 measuring system and the MMPP with the sewing machine and computer
Electrical energy consumption for the sewing machine electrical motor was measured under laboratory conditions of sewing straight seams in the range of lengths from 20 to 600 mm. The laboratory samples were prepared so that the fabric was cut parallel with the warp. Sample lengths were 20, 50, 100, 200, 300, 400, 500 and 600 mm, while all the samples were 100 mm wide. The samples were joined by a seam, following the direction of the warp, at 7 mm from the sample selvedge. Specific stitch density was 10 cm2 1. Each sample length was measured 20 times consecutively, and the samples were stitched using a double machine lockstitch of the type 301. 4. Results and discussion The result obtained through systematic investigations and varying factors (nominal sewing machine stitching speed, number of sewing segments and seam length) are shown in the Figures 3 and 4. They deal with the energy parameters and are used as a basis to establish a new method of predicting electrical energy consumption in sewing straight seams (RAVEN). Figure 3 shows a diagram depicting correlation of the energy spent and stitching speed in joining a straight seam 300 mm long, employing various methods of seam joining in a single, in two or three sewing segments, with stitching speeds ranging from 1,000 to 4,000 rpm.
IJCST 17,3/4
184
Figure 3. Correlation of the energy consumption (E/Ws), nominal stitching speed (vn/rpm) and the number of seam segments sewn
Figure 4. Correlation of electrical energy consumption (E/Ws) and nominal stitching speed (vn/rpm) at varying number of stitches in the seam (Ns/n)
The least amount of energy (495 Ws) is spent in joining the seam in a single sewing segment, at stitching speed of 1,000 rpm, while energy expenditure is at its highest at the stitching speed of 4,700 rpm, in joining the same seam, but in three segments (1,576 Ws). The results show that when joining a seam at lower stitching speeds and in a single segment, energy consumption is 3.18 times lower than in joining the same seam, at top stitching speed and in three segments. Measurements performed show that electric energy consumption for driving electrical motor on a Brother sewing machine is considerably affected by: method of work; level of skill on the part of the operator; number of segments in joining the seam and stitching speed while joining the seam in question. Graph depicting the correlation of energy consumption (E/Ws) and nominal stitching speed in the range between 1,000 and 4,700 rpm, at varying number of stitches in the seam (Ns/n), can be seen in Figure 4.
It is quite obvious that increasing the number of stitches in the seam results in higher electrical energy consumption for the workings of the sewing machine drive electrical motor. To join a seam of 40 stitches, at the sewing speed of 3,000 rpm, 439 Ws is spent, while it is necessary to spend 1,161 Ws to join a seam of 200 stitches, at the same nominal stitching speed, which means 2.64 times more energy is necessary to do the job. Higher stitching speed also means more energy spent at the same number of stitches in the seam. When joining a seam of 80 stitches at the nominal stitching of 1,000 rpm, 330 Ws is spent, while the same number of stitches at 4,700 rpm asks for 736 Ws. The measuring data obtained are exposed to multiple regression analysis and the following equation is obtained, which can be used to calculate electrical energy consumption (E/Ws) for the sewing machine drive electrical motor in sewing straight seams. E ¼ e½0:44158320:000070·vn þ0:003719·N s þ0:624983 · lnðvn Þþ0:198207 · lnðN s Þ
ð1Þ
The mathematical expression presented, equation (1), is the initial step of establishing the method of predicting electrical energy consumption for the sewing machine drive motor in sewing straight seams. The method is called the EN method. It can be seen that the same input parameters are used to predict the need for electrical energy as for the calculation of normal times of machine-hand sub-operations in sewing straight seams, as determined by previous investigations. This is why the two methods are joined into a new RAVEN method ðRAV þ ENÞ: Figure 5 shows the scheme of the software programme with the input and output parameters of the RAV and EN method. The number of stitches in the seam and nominal stitching speed are common input parameters, while the output parameters for the RAV method is normal time for machine-hand sewing sub-operation and for the EN method the electrical energy spent. 5. Conclusions Systematic investigations of electrical energy consumption in garment sewing operations have shown that joining the same seam, when using different methods of work, can result in the increase of electrical energy consumption by as much as 55 per cent. It can also be seen that continuous joining of seams at lower stitching speed, as compared to joining in more segments at higher machine speeds, results in more than three times higher electrical energy consumption. It is necessary to pay special attention in designing the processes to matching the method of work applied and stitching speed used. Measuring results obtained indicate that electrical energy consumption for the sewing machine drive electrical motor depends in a considerable degree upon the changes of garment sewing operation processing parameters. In the case described, these parameters include the number of stitches in the seam and nominal stitching speed. More stitches in the seam result in a linear increase of electrical energy consumption for the sewing machine drive electrical motor. It means that joining the seams with the least possible number of stitches, and at the same nominal stitching speed, as compared to joining the same seam with the maximum number of stitches possible, can result in savings of electrical energy by as much as 78 per cent. The investigations of electrical energy processing parameters in garment sewing operations included designing a mathematical model and software package aimed at
Equipment and methods used
185
IJCST 17,3/4
186
Figure 5. The scheme of the RAVEN software programme
faster and simpler application of the results obtained under real in-plant conditions. This is why the previously established RAV method for determining normal times of machine-hand sewing sub-operation of sewing straight seams has been joined with the EN method of calculating the predicted electrical energy consumption in sewing straight seams. Both methods use the same input parameters, i.e. the number of stitches in the seam and nominal stitching speed and can be joined into the unique RAVEN method that can be used to determine normal machine-hand times and electrical energy consumption for a particular technological operation. The investigations described have shown the impact of the method of work applied and the changes of garment sewing operation processing parameters on the level of electrical energy consumed by the sewing machine drive electrical motor.
New measuring method has been introduced in garment engineering, aimed at predicting electrical energy consumption in garment sewing operations, opening thus a completely new field of investigation in the area of garment technologies. References Firsˇt Rogale, S. et al. (2003), “Determining reaction abilities of sewing machine operators in joining curved seams”, International Journal of Clothing Science and Technology, Vol. 15 Nos 3/4, pp. 179-88. Rogale, D. and Dragcˇevic´, Z. (1998), “Portable computer measuring systems for automatic process parameter acquisition in garment sewing processes”, International Journal of Clothing Science and Technology, Vol. 10 Nos 3/4, pp. 283-92. Rogale, D. and Petrunic´, I. (2002), “The method of determining sewing machine energy consumption coefficient”, in Dragcˇevic´, Z. (Ed.), Proceedings of 1st International Textile, Clothing & Design Conference – Magic World of Textiles, Faculty of Textile Technology, University of Zagreb, Zagreb, October, pp. 389-94. Rogale, D. et al. (2003), “Determining energy parameters in garment sewing operations”, in Gersˇak, J. (Ed.), Proceedings of 4th International Conference IMCEP 2003, Faculty of Mechanical Engineering, Institute for Textile and Garment Manufacture Processes, Maribor, October, pp. 115-24. Schweitzer (2001), Meßdatenanalysen mit Excel, Fotosatz Pfeifer, Gra¨felfing. Shoup, T. (1984), Applied Numerical Methods for the Microcomputer, Prentice-Hall, Englewood Cliffs, NJ. Tilli, T. (1995), Messen, Steuern und Reglen mit Visual Basic, Franzis-Verlag GmbH, Poing.
Equipment and methods used
187
The Emerald Research Register for this journal is available at www.emeraldinsight.com/researchregister
The current issue and full text archive of this journal is available at www.emeraldinsight.com/0955-6222.htm
IJCST 17,3/4
A study of the relaxation phenomena in the fabrics containing elastane yarns
188
Jelka Gersˇak Textile Department, Faculty of Mechanical Engineering, University of Maribor, Maribor, Slovenia
Dunja Sˇajn and Vili Bukosˇek Department of Textiles, Faculty of Natural Science and Technology, University of Ljubljana, Ljubljana, Slovenia Abstract Purpose – In this paper, special attention is focused on the study of the relaxation phenomena of fabrics containing elastane yarn. Design/methodology/approach – For this purpose, the relaxation phenomena of wound fabric under constant deformation, as the consequence of accumulated stress during winding, were analysed. Maxwell’s model and the modified standard linear solid model are used for explaining the relaxation. Findings – The results of the study of the relaxation phenomena of fabrics containing elastane yarn show a close connection between stress relaxation under constant deformation in the fabric roll and the degree of deformation with manual unwinding. Expert knowledge of the relaxation phenomena in fabrics containing elastane yarns has a big influence on explaining the problem of dimensional changes and instability in such fabrics. Originality/value – A better understanding of the relaxation phenomena in fabrics containing elastane yarns. Keywords Relaxation theory, Modelling, Deformation, Fabric testing Paper type Research paper
International Journal of Clothing Science and Technology Vol. 17 No. 3/4, 2005 pp. 188-199 q Emerald Group Publishing Limited 0955-6222 DOI 10.1108/09556220510590885
1. Introduction The fabrics containing elastane yarns have wide application value, especially because of their increased extensibility, elasticity, high degree of recovery, good dimensional stability and simple care. In the apparel industry, this group of fabrics is used for sportswear and leisure garments, hosiery, underwear, swimwear, therefore, for the body confirming garments, which ensure stable shape under loads in wearing. This group of fabrics is characterised by the elastane yarns, which may be either incorporated into the fabric in the pure state or wrapped with relatively inextensible fibres. Wrapping is done by covering, core spinning or twisting. These structures are also known as wrapped, core spun yarns, spin-twisted or Siro-Spun yarns or air-twisted yarns (Malej-Kveder and Nikolic´, 1992). Exceptional elasticity of the fabrics containing elastane yarns is obtained in manufacturing process as a result of technologically based forces, resulting in various types of fabric deformation. The degree of deformation depends on the chemical composition, construction of the fabric, mass and thickness. Since in the production process lower loads are used, resulting deformations are not permanent. It means that the initial deformation of the fabric after loading consists of two elements, the
recoverable deformation (elastic deformation) and the deformation, which is recoverable in time (primary creep) (Morton and Hearle, 1993). Having the above facts in mind, the first part of this paper deals with the relaxation phenomena in the fabric roll, where the fabric is exposed to permanent deformation. A mechanical model is used to describe the relaxation phenomena, as it explains the stress decrease with time under a permanent deformation. In the other part of the paper, the deformation of the layers of fabric are explored after manual unwinding from the roll and spreading the fabric to form the lays. 2. Elastic properties of the fabrics containing elastane yarns Elastic properties of the fabrics containing elastane yarn result in the exceptional extensibility, since the elastane filament has the extensibility of 300 to as high as 700 per cent. The extensibility and elasticity of the fabrics depends on the percentage of the elastane filament content, which is usually between 2 and 8 per cent (Rupp and Bo¨hringher, 1999). This group of fabrics possess low elasticity modulus E0 and offer lower resistance to acting forces. Apart from the modulus, the breaking load is also lower than the values for ordinary fabrics. On the contrary, the breaking extension of the fabrics containing elastane yarns is higher than for ordinary fabric. As a typical viscoelastic property, extension under moderate load can be either completely recoverable, or partially and time-dependent. The first part of the overall deformation is known as elastic deformation and the other as plastic or permanent part of overall deformation. When the fabric is stretched by applying the load below the yield point, the result is elastic deformation. With the values of the stress and extension in the yield point, extensibility is again increased, as is the field of the elastic deformation of the fabric on the stress-extension curve. The yield point on the curve presents the limit of proportionality, where the extension ceases to be proportional to the stress. Permanent deformation starts above the yield point. It means unrecoverable deformation occurs, which is undesirable in the manufacturing process (Rupp and Bo¨hringher, 1999; Bukosˇek, 1983). As already mentioned, the fabrics containing elastane yarns have high elasticity, which means the ability of the fabric to recover (elastic recovery or extension recovery) and return into the initial shape. Elastic recovery depends on the elastic component of the deformation, which is reduced with increased deformation. 2.1 Relation between loading and deformation Tensile forces cause fabric deformation, which can be seen as the change of fabric length as compared to the length prior to stretching. Fabric deformations under the load below 490.5 Nm2 1 are characteristic for the evaluation of tensile properties with the KES-FB measuring system, for the fabrics containing elastane yarn from 10 to 30 per cent. This is ten times more than the deformation of conventional fabrics (Gersˇak, 1997). During winding on the roll and spreading the fabric is exposed to tensile forces, which cause deformation. The deformation mentioned is characterised by the elastic (instantaneous recoverable) component and the deformation, which is recoverable with time. The knowledge of the tensile loads and the relaxation phenomena, which the fabrics are exposed to during winding and spreading, is of theoretical and practical importance in planning further processes.
A study of the relaxation phenomena 189
IJCST 17,3/4
190
Figure 1. The cross section of the roll with the compressive forces acting inside the roll
Knowing the relation between the tensile loads and deformation it can be used to predict the stress relaxation of the fabric under constant deformation in the roll, based on the mechanical models with different combinations of basic elements, the spring and the dashpot. At unwinding and spreading the fabric, the response to the loads results in the deformation or the elongation of the fabric. On the removal of the load, the fabric relaxes, which can be seen as the change of deformation in a particular time. 2.1.1 Relaxation phenomena in the roll. Radial forces ðT dx=xÞ act from the outside (with the radius rl) to the inner part (with the radius rx) of the roll during winding, as a consequence of previous layers in the roll. The conditions of winding create a state of stress in the wound roll, which is higher in the radial direction of the roll than in the tangential (Figure 1) (Ghosh et al., 1991). In making the roll, the fabrics containing elastane yarns exhibits viscoelastic properties of a viscoelastic solid, with the elastic properties of a solid, and respond to the Hook’s law and as a viscous liquid as stipulated by the Newton’s law. Due to the typical viscoelastic properties of the wound fabric, the stresses in the roll in tangential direction are lower than in radial one. This can be seen after manually unwinding the first layer of the fabric roll, as there is no change of length of the cut layer after particular relaxation time (Sˇajn et al., 2003). On the contrary, the stresses in the radial direction from the outside to the inner part of the roll increase (Ghosh et al., 1992). The behaviour of the fabric in the roll as a viscoelastic material is described using the mechanical models, which consist of the basic model of the spring and the dashpot. The spring presents the elastic properties of the fabric according to the Hook’s law, while the dashpot presents the viscose component of the deformation, which is not completely recoverable and timely dependent. The behaviour of the viscose component is described by the Newton’s law. The mechanical models are constructed from the basic model of the spring and the dashpot, in different combinations. Literature mentions mechanical models for explanation of the relaxation phenomena in the fabrics as Maxwell’s model, used to explain the changes of stress inside a fabric roll (Ward and Hadley, 1993; Barnes et al., 1989).
2.1.1.1 Maxwell’s model. The Maxwell’s model consists of a series of connections of the spring and the dashpot. For the model described, the total stress s is equal to the stresses of the spring sH, as well as of the dashpot sN. The following equation describes the situation: s ¼ sH ¼ sN : The overall deformation is the sum of the deformations of the spring 1H and the dashpot 1N, 1 ¼ 1H þ 1N : The equation for the stress-extension relation for the Hook’s spring is: 1H ¼
s E
ð1Þ
where s is the stress and E the elasticity modulus of the spring. The equation (2) is valid for the dashpot: 1_N ¼
s m
ð2Þ
The differential equation (3) describes the stress-extension behaviour of the Maxwell’s model: 1_ ¼
s_ s þ E m
ð3Þ
Solution for the equation (3) depends on the velocity of the extension or stress s(t). The stress-extension relation with the determined value of the stress can be expressed as (Vangheluwe, 1992): Z t 1ð sðtÞ ¼ e2ðE=mÞðt2t0 Þ sðt 0 Þ þ E ð4Þ _ tÞeðE=mÞðt2t0 Þ dt t0
The expression for the Maxwell’s model under constant deformation ð1ðtÞ ¼ 1ðt0 Þ; t $ t 0 Þ can be written as:
sðtÞ ¼ sðt 0 Þe2ðE=mÞðt2t0 Þ
ð5Þ
The equation shows that the stress disappears exponentially with the characteristic time constant, which is the relaxation time t ¼ m=E: It means the time in which the stress disappears with the factor e ¼ 2:71828; due to the relaxation. The equation (5) can also be written as
sðtÞ ¼ sðt0 Þe2ðt2t0 Þ=t
ð6Þ
2.1.2 Loading and deformation in spreading of the fabrics containing elastane yarns. The fabrics are exposed to tensile and shear loads in the course of fabric laying. It causes more or less pronounced deformation of the fabric exposed. The deformations are caused by the tensile forces acting on the part of the fabric, from feed to guide rollers of the spreading machine. On the other hand, the shear deformation of the fabric is caused by the helix-shaped rubber lining of the guide roller. The deformations occurring in the form of fabric elongation depends upon the sort of fabrics, construction parameters, mechanical and physical properties, technology-caused loads during spreading and on the spreading machine equipment. After spreading, the fabric relaxes, which is shown as the change of the dimension of the layers in the warp and
A study of the relaxation phenomena 191
IJCST 17,3/4
192
weft directions (Blekacˇ et al., 2003). The degree of the fabric relaxation depends on the degree of the load during laying up the fabric and the number of the laid plies in the lay. The laid plies in the lay are loaded by compression forces, higher in the inner part of the lay. Due to the impact of the mass of the upper plies, the inner parts do not have a chance to relax. 3. Methods Based on the theoretical elements given, the investigation of the relaxation phenomena was in the first part focused on the stress relaxation under a constant deformation with the initial load ð490:5 ^ 100 N=mÞ: The initial or maximum load was applied at the intersection of the specimen as the initial or maximum stress. The stress decrease is expressed as the percentage of the initial stress after 15, 30, 45 and 60 min. The experimental stress relaxation curve was fitted with a model based on the Maxwell’s element (6). The other part of the research describes the monitoring of the relaxation of the layers at manual unwinding of the central and the last layer, immediately after the manual unwinding and after 15, 30, 45 and 90 min, as well as after 2, 3, 4, 5, 6, 7 and 24 h. The relaxation during spreading was analysed for the time intervals mentioned before. The research was focused on the three different kinds of fabrics of twill weave with the same fibre composition of the blended yarn from Wo/El, but including different content of elastane in the yarn. Table I shows the general properties of the fabrics used. 4. Results and discussion The results of the stress relaxation measurements under constant deformation and the relaxation of the fabric after manual unwinding and spreading the fabric are discussed. 4.1 The results of stress relaxation measurements under constant deformation The results of stress relaxation measurements after 15, 30, 45 and 60 min are listed in Table I, while Figure 2 shows the experimental relaxation curve and the calculated curve of the Maxwell’s model (6). The values of relaxation time from the equation (6) are listed in Table II. The analysis of the results of the relaxation phenomena indicates that, with the growing content of the elastane, the degree of stress relaxation under constant deformation also increases. It means that higher elasticity and higher deformation under lower loads results in faster stress decrease. The analysis of the fabrics designated TK-3, those with the highest percentage of elastane in the yarn (6.67 per cent), shows the fastest decrease of stress in the first 15 min, under constant deformation (Table III). The results of the stress relaxation also show that the Maxwell’s model, which explains the stress relaxation curve, is not convenient for the experimental curve of the fabrics analysed. The differential equation of the Maxwell’s model shows that the stress decreases exponentially to zero, which is not valid for the fabrics analysed (Figure 2). Furthermore, from the results calculated for the relaxation time, it is clear that longer time, which is between 119.79 and 168.61 min, means slower decrease of stress under constant deformation (Table II). The longest relaxation time is calculated for the fabric designated TK-3, which has the highest percentage of elastane in the yarn.
TK-1 TK-2 TK-3
Fabric designation Weft 22 24 29
Warp
33 35 31
Yarn density per 10 mm
133.2 123.6 229.8
Mass (m/gm2 2) 0.65 0.51 0.47
Thickness (mm) Two-way stretch One-way stretch (weft) Two-way stretch
Fabric type 0.75 0.96 1.04
Warp
0.75 0.96 1.04
Weft
Stress s24.5 (Nmm2 2)
28 8.32 27.3
Warp
24 34.5 26.4
Weft
Extension 124.5 (per cent)
Stress and extension under maximum load of 490.5 N/m or 24.5 N/5 cm
Weft 3.27 5.4 6.39
Warp 3.46 0 6.67
Percentage of the elastane in the yarn (per cent)
A study of the relaxation phenomena 193
Table I. General properties of the fabrics
IJCST 17,3/4
194
Figure 2. Results of the stress relaxation of the fabrics tested, designated TK-1, TK-2 and TK-3, under constant deformation and the fitted values from the Maxwell’s model: (a) warp direction; and (b) weft direction
4.2 The results of the fabric relaxation measurements after manual unwinding and spreading The results of the fabric relaxation measurements immediately after manual unwinding and spreading, after 15, 30, 45 and 90 min and after 2, 3, 4, 5, 6, 7, 24 h are shown in Figure 3 and 4. The results of the analysis of the deformation after manual unwinding and spreading the fabric show high correlation with the results for stress relaxation under constant deformation. The analysis of the results shows the relaxation of the fabric after manual unwinding as the consequence of the stress accumulated inside the roll. The relaxation is seen as the deformation with the highest value in the first 15 min after unwinding. The degree of deformation depends on the content of elastane in the yarn. The highest value of deformation is recorded for the fabric designated TK-3, 0.08 per cent after 15 min. The highest value of deformation 15 min after spreading (0.16 per cent) is also measured for the fabric mentioned. Higher deformation is recorded in spreading than in manually unwinding, due to the higher tensile forces acting in this sub-operation.
A study of the relaxation phenomena 195
5. Conclusion The investigation of the relaxation phenomena for the fabrics containing elastane yarns indicates that: . the stress relaxation under constant deformation depends on the content of elastane in the fabric; the highest value is reached in the first 15 min of constant deformation; . the fabrics with the highest content of elastane need the longest relaxation time; . the Maxwell’s model cannot adequately explain the stress relaxation of the fabrics containing elastane yarns; . the highest deformation is recorded in the first 15 min after the manual unwinding and spreading; and . the degree of deformation in winding and spreading depends on the elastane content in the yarn. The knowledge of the relationship between the degree of deformation and the stress relaxation is essential for the procedure of engineer planning the fabric responses, due to the technology-caused forces that act in the production process. Future investigations will be focused on the study of the mechanical models and the search for the appropriate model that could precisely determine the relaxation phenomena of the fabrics containing elastane yarns.
Relaxation time t (min) Fabric designation
Warp
Weft
TK-1 TK-2 TK-3
147.44 159.34 168.61
119.79 126.49 144.68
Table II. The results calculated for the relaxation times of the fabrics used (equation (6))
TK-1 TK-2 TK-3
1.2 0.89 1.05
1.07 1.15 1.95
Fabric designation
Table III. The results of stress relaxation measurement of the maximum stress, after 15, 30, 45 and 60 min Initial or maximum stress (Nmm2 2) Warp Weft 21 15.6 19.4
24 19.3 21.4
After 15 min Warp Weft 22.2 16.7 21.4
26 23.4 24.5
After 30 min Warp Weft
23.4 18.8 22.4
27.2 24.8 25.5
After 45 min Warp Weft
Percentage of the initial stress (per cent)
24.7 18.8 23.5
28.3 25.7 26.5
After 60 min Warp Weft
IJCST 17,3/4
196
A study of the relaxation phenomena 197
Figure 3. The relaxation of the fabrics laid by manual unwinding of the central and the last layer from the fabric roll: (a) central layer of the fabric designated TK-1, 503.6 cm long, the fabric designated TK-2, 495 cm long and the fabric designated TK-3, 492 cm long; and (b) the last layer of the fabric of the mentioned length
IJCST 17,3/4
198
Figure 4. The relaxation of the fabric layers immediately after spreading and after pre-determined periods of time: (a) the change of length of the fabric designated TK-1, with the length of lay 501.0 cm, TK-2, 556 cm long and TK-3 508.5 cm long; and (b) the change of the width for the fabric designated TK-1 with the length of lay 152.2 cm, TK-2 153.7 cm long and TK-3 147.9 cm long
References Barnes, H.A., Hutton, J.F. and Walters, K. (1989), An Introduction to the Rheology, Department of Mathematics, University College of Wales, Aberystwyth. Blekacˇ, R., Gersˇak, J. and Gubensˇek, I. (2003), “Model simulacije obremenitev in deformacij tkanin pri polaganju”, Tekstilec, Vol. 46 Nos 11/12, pp. 348-53. ˇ Bukosek, V. (1983), “Racˇunalnisˇko vrednotenje viskoelasticˇnih lastnosti vlaken”, Tekstilec, Vol. 26 No. 12, pp. 24-9. Gersˇak, J. (1997), “Objektivno vrednovanje fiksiranih dijelova odjec´e”, Tekstil, Vol. 46 No. 4, pp. 193-203. Ghosh, T.K., Peng, H. and Banks-Lee, P. (1992), “Analysis of fabric deformation in a roll-making operation. Part II: effect of viscoelasticity”, Textile Research Journal, Vol. 62 No. 10, pp. 669-76. Ghosh, T.K., Peng, H., Banks-Lee, P., Hamouda, H. and Shin, D.H. (1991), “Analysis of fabric deformation in a roll-making operation. Part I: a static case”, Textile Research Journal, Vol. 61 No. 3, pp. 153-61. Malej-Kveder, S. and Nikolic´, M. (1992), “Lycra core spun preja v predelavi do koncˇnega izdelka”, Tekstilec, Vol. 35 No. 4, pp. 241-3. Morton, W.E. and Hearle, W.S. (1993), Physical Properties of Textile Fibres, The Textile Institute, Manchester. Rupp, J. and Bo¨hringher, A. (1999), “Elastanhaltige garne und stoffe”, International Textile Bulletin, Vol. 35 No. 1, pp. 10-30. ˇSajn, D., Gersˇak, J. and Bukosˇek, V. (2003), “Sˇtudij odnosa med obremenitvijo in relaksacijo tkanin z dodanim elastanom”, Tekstilec, Vol. 46 Nos 9/10, pp. 274-81. Vangheluwe, L. (1992), “Influence of strain rate and yarn number on tensile test results”, Textile Research Journal, Vol. 62 No. 1, pp. 586-9. Ward, I.M. and Hadley, D.W. (1993), An Introduction to the Mechanical Properties of Solid Polymers, Wiley, Chichester. Further reading Vangheluwe, L. and Kiekens, P. (1995), “Test method for cloth fell displacement caused by relaxation”, Journal of the Textile Institute, Vol. 65 No. 9, pp. 540-4.
A study of the relaxation phenomena 199
The Emerald Research Register for this journal is available at www.emeraldinsight.com/researchregister
IJCST 17,3/4
200
The current issue and full text archive of this journal is available at www.emeraldinsight.com/0955-6222.htm
Numerical optimisation of mechanical reinforcements of weak spots in textiles and garment Zˇeljko Sˇomodi, Anica Hursa and Dubravko Rogale Department of Clothing Technology, Faculty of Textile Technology, University of Zagreb, Zagreb, Croatia Abstract Purpose – The purpose of this paper is to investigate and illustrate the possibilities of a systematic engineering approach in the design of mechanical reinforcements in garments. A mechanical reinforcement can be designed using optimisation strategies. Design/methodology/approach – In this work, an iterative algorithm for minimum search based on parabolic approximations is proposed and applied in the optimisation of mechanical reinforcement in a selected model problem of a buttonhole type. Findings – Optimisation algorithm based on parabolic approximations, in conjunction with the finite element analysis, offers some promising possibilities as support for the decision-making process in the design of mechanical reinforcements. The selection of optimisation criteria – influence parameters and corresponding weight factors – remains of course to be studied and implemented by the clothing engineering experts. Research limitations/implications – The intention in future work should be to optimise two or more geometric parameters simultaneously (multidimensional optimisation), and to produce a computer program for an automated optimum search. Practical implications – The contents of the paper could be useful for the experts in clothing engineering in the process of design or selection of the reinforcements of weak spots in textiles and garments. Originality/value – This paper provides optimisation routes to the weak sports of mechanical reinforcements in textiles and garments. Keywords Clothing, Mechanical testing, Optimization techniques, Numerical analysis, Finite element analysis Paper type Research paper
International Journal of Clothing Science and Technology Vol. 17 No. 3/4, 2005 pp. 200-208 q Emerald Group Publishing Limited 0955-6222 DOI 10.1108/09556220510590894
1. Introduction It has always been one of the basic tasks in any field of engineering: to answer in the best possible way to a number of demands, often divergent and sometimes directly opposed, when designing certain product. Some typical examples: high strength and rigidity – low weight and cost, attractive look – ease of manufacturing and maintenance. In mathematical sense, the optimal (best possible) solution corresponds to the extreme of an objective function – the quantity containing numerical representations of two or more criteria (Donald, 1986). If the objective function represents whatever is unwanted, then it needs to be minimised, for example, sum of production costs.
In the recent decades, a strong and fast development of numerical methods of engineering analysis has opened new possibilities of optimisations in various problems of structural and production design (Lankalapalli, 2002). The logical step in the development of application of various codes for numerical analysis, in particular finite element codes, is to incorporate them into programmes for optimum search. We believe that this fusion of numerical analysis and optimisation can also be of interest in textile and clothing engineering, in particular in the “structural” type of situations such as stress concentration in the vicinity of geometric irregularities. A typical representative example is buttonhole reinforcement – a problem that has been so far dealt with, at least to our knowledge, on an ad hoc basis, following the feeling, instinct and experience, rather than a rational engineering approach. The aim of this paper is to offer a possible way of thinking and acting in the design process of mechanical reinforcements of such weak spots in textiles and garment. 2. A method of iterative extreme search It is well known from the basic differential calculus that vanishing of the first derivative is the condition of existence of an extreme (here our attention is restricted to the functions of a single variable – the generalisation to multidimensional problems should be of no conceptual difficulty and is beyond the scope of this work). If however, the function is not analytically defined in the domain of interest, but can only be numerically evaluated at certain point, then it is not possible to enforce the condition of zero first derivative. In such case, a search for the extreme can be done in the following way. (1) Select the starting point i (initial guess) and compute the values of function and its first and second derivatives. (Remarks: the derivatives can be evaluated by computing the values of the function in two close points and using the finite difference scheme: f 0i <
f ðx þ DxÞ 2 f ðx 2 DxÞ f iþ1 2 f i21 ¼ ; 2Dx 2Dx
f 00i <
f iþ1 2 2f i þ f i21 ð1Þ ðDxÞ2
If a minimum is sought, f 00 must be positive). (2) Approximate the function by the quadratic polynomial (parabola) at the point i: y ¼ ax 2 þ bx þ c;
1 a ¼ f 00i ; 2
b ¼ f 0i 2 2axi ;
c ¼ f i 2 ax2i 2 bxi :
ð2Þ
(3) Compute the coordinates of the apex as the first approximation of the extreme to be found: x0 ¼ 2
b ; 2a
y0 ¼ c 2
b2 : 4a
ð3Þ
(4) Put xi¼ x0 and repeat steps 1, 2 and 3. (5) In each iteration check the convergence; exit the iteration loop when the two successive approximations coincide to within the prescribed tolerance. In order to examine the behaviour of the described iterative scheme, let us illustrate its performance by the following two examples.
Numerical optimisation
201
IJCST 17,3/4
202
2.1 Example 1: the catenary The hyperbolic cosine function y ¼ coshðxÞ is represented by the curve known as catenary with the minimum min y ¼ 1 at x ¼ 0: In Figure 1 both graphic representation and numerical results of the iteration are given for the case of extreme search starting from the point xi ¼ 3:5: In each step only the arc of the approximating parabola from the starting point to the apex is drawn. In the numerical results for each step the abscissa of the starting point is given followed by the x and y coordinates of the apex of the approximating parabola. Note that once the stating x is under unity, the progressive vanishing of error occurs, known as the quadratic rate of convergence. 2.2 Example 2: first half period of the sine curve As we know, the sine curve y ¼ sinðxÞ has the extreme Max y ¼ 1 at x ¼ p=2 ¼ 1:570796 . . . : The iterative extreme search is shown in Figure 2 for the case of the initial guess at x ¼ 0:5: This example is instructive for the possible danger of divergence. See the case in Figure 3, where the point of initial guess is slightly removed to x ¼ 0:4: As can be seen in the example of Figure 3, the approximations may escape from the extreme – let us say that generally such divergence may occur if the second
Figure 1. Iterative extreme search in the case of the catenary
Numerical optimisation
203
Figure 2. Sine curve – iterative extreme search starting from x ¼ 0.5
IJCST 17,3/4
204
Figure 3. Sine curve – divergence of iteration starting from x ¼ 0.4
derivative is insufficient at the starting point. A simple analysis of the parabolic approximation in the case of the sine curve reveals that in this case the limit of the attraction zone is defined by the condition x þ cotðxÞ ¼ p=2 or x < 0:40525:
Numerical optimisation
3. Application to a buttonhole type reinforcement In this section, an example of optimisation of the mechanical reinforcement for the reduction of stress concentration is presented. The computer programme for the plane stress analysis of materials weak in compression (Sˇomodiet al., 2003a, b) has been modified to enable the analysis of problems with the reinforced zone of variable thickness. In this example, a one-dimensional optimisation is done, in the sense that the only variable is the factor of reinforcement fR, defining the ratio of additional thickness in the reinforced zone over the basic material thickness, while other geometric parameters (shape and size of the reinforced zone) are kept constant. The problem and the load case representative for the buttonhole type situations are defined as follows: a quadratic strip with transverse opening is loaded by uniformly distributed tensile stress at the ends. The opening has semi-circular ends reinforced in the zone of the adopted geometry. Due to symmetry, one quarter of the strip is modelled by the finite element mesh (Figure 4). The finite elements in the reinforced zone are drawn with thick lines. Let us mention that the effect of sharp step in the transition from the basic to increased thickness in the reinforced zone is neglected. This should be acceptable for the reasonably low values of the reinforcement factor.
205
Figure 4. Finite element model of the tension strip with opening and reinforcements
IJCST 17,3/4
206
The analysis has been done for the range of reinforcement factor fR from 0 (no reinforcement) to 4 (total thickness of material in the reinforced zone five times higher than the basic thickness). Several resulting parameters are computed: maximum principal stress in a finite element, maximum von Mises equivalent stress at a node, maximum nodal displacement (elongation), total volume of built – in material, and volume of the material added for reinforcement. After certain insight into the results some of the parameters are eliminated and objective functions suitable for minimisation are constructed in the following way. It appears that the parameters expressing the opposed requirements of high strength and low cost are the maximum element principal stress s and the volume of added material V. The objective function is constructed from their relative values with respect to the adopted reference point (here we adopt it at the reinforcement factor f R ¼ 2) in the form: F ¼ ws
s sref
2
þwV
V V ref
2 ð4Þ
Since the sum of the weight (influence) factors ws and wV is equal to unity, it follows that the objective function takes the value of unity at the reference point. In Table I, two objective functions are listed: F1 with ws ¼ wV ¼ 0:5 and F2 with ws ¼ 0:8; wV ¼ 0:2 (higher influence of stress). The same results are shown graphically in the diagram of Figure 5. Relative volume (curve o), relative stress (curve þ ) and objective functions F1 (curve x) and F2 (spotted curve) are shown in terms of the reinforcement factor fR ranging from 0 to 4 (abscissa). Note that both objective functions are well suited for minimum search. Let us finally demonstrate how the minimum of the objective functions, hence the accordingly optimal values of the reinforcement factor, can be found using the proposed minimisation algorithm without all these computations. Starting from the reference point f R ¼ 2 (initial guess), the results of the minimum search for the objective functions F1 and F2 are given in the Tables II and III, respectively. These tables indicate that only two to three steps of iteration lead to the required optimum. 4. Conclusion Optimisation algorithm based on parabolic approximations, in conjunction with the finite element analysis, offers some promising possibilities as support for the decision making process in the design of mechanical reinforcements. The selection of optimisation criteria – influence parameters and corresponding weight factors – remains of course to be studied and done by the clothing engineering experts. It appears that the position of the minimum of an objective function of the form (equation (4)) depends to some extent on the selection of the reference point. Such a lack of objectivity could be eliminated by an adaptive redefinition of the reference point in each iteration step. This would lead to the final minimum of the
Reinforcement factor, fR 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2 2.1 2.2 2.3 2.4 2.5 2.6 2.7 2.8 2.9 3 3.1 3.2 3.3 3.4 3.5 3.6 3.7 3.8 3.9 4
Relative volume, V
Relative stress, s
0 0.279 0.558 0.837 1.116 1.395 1.674 1.953 2.232 2.511 2.79 3.069 3.348 3.627 3.906 4.185 4.464 4.743 5.022 5.301 5.58 5.859 6.138 6.417 6.696 6.975 7.254 7.533 7.812 8.091 8.37 8.649 8.928 9.207 9.486 9.765 10.044 10.323 10.602 10.881 11.16
6.239 5.935 5.662 5.417 5.194 4.991 4.806 4.635 4.478 4.332 4.2 4.072 3.954 3.844 3.741 3.645 3.553 3.467 3.386 3.309 3.236 3.166 3.1 3.037 2.977 2.92 2.865 2.813 2.762 2.714 2.668 2.624 2.581 2.54 2.5 2.462 2.425 2.39 2.355 2.322 2.29
Objective function F1 F2 1.718919 1.683128 1.535709 1.412355 1.308121 1.220651 1.14786 1.087026 1.037461 0.997295 0.9672703 0.9429643 0.926494 0.9167868 0.9132336 0.9156278 0.9227586 0.9351823 0.9524277 0.9740632 1 1.029852 1.063856 1.101645 1.143166 1.188366 1.236924 1.289077 1.344251 1.40295 1.464879 1.530011 1.598075 1.669299 1.743423 1.820671 1.900787 1.98399 2.06981 2.158691 2.250394
2.750271 2.691504 2.451135 2.246268 2.068994 1.915541 1.782577 1.665741 1.563938 1.474172 1.397632 1.327243 1.26639 1.213359 1.167174 1.127505 1.092414 1.062792 1.037884 1.017001 1 0.9862636 0.9761695 0.9691322 0.9650654 0.9638864 0.9650789 0.9690224 0.9748011 0.9832203 0.9938067 1.006518 1.020919 1.037379 1.055478 1.075573 1.097259 1.120884 1.145697 1.172405 1.20063
objective function being equal to unity, hopefully without any numerical instability. The intention in the future work should be to optimise two or more geometric parameters simultaneously (multidimensional optimisation), and to produce a computer programme for an automated optimum search.
Numerical optimisation
207
Table I. Results of computation and objective functions F1 and F2
IJCST 17,3/4
208
Figure 5. Graphic representation of influence parameters and objective functions
Step Table II. Iterative minimum search for F1
1 2 3
Step Table III. Iterative minimum search for F2
1 2
Starting fR
Approx. min F1
Approx. opt fR by F1
2 1.3 1.4
0.900631187 0.913215176 0.913205367
1.287536 1.407738 1.409744
Starting fR
Approx. min F2
Approx. opt fR by F2
2 2.5
0.9638245207 0.9638863904
2.470768 2.499715
References Donald, A.P. (1986), Optimisation Theory With Applications, Dover Publications, Inc., New York, NY. Lankalapalli, S. (2002), “Optimal pick-up locations for transport and handling of limp materials”, PhD thesis, North Carolina State University. Sˇomodi, Zˇ., Hursa, A. and Rogale, D. (2003a), “Finite element analysis of textile in plane stress”, in Gersˇak, J. (Ed.), Proceedings of 4th International Conference Innovation and Modelling of Clothing Engineering Processes – IMCEP 2003, October 2003, Faculty of Mechanical Engineering, Maribor, Slovenia, pp. 108-14. Sˇomodi, Zˇ., Hursa, A. and Rogale, D. (2003b), “Nonlinear modelling of flexible materials in plane stress with application to stress concentration in textile”, in Matejicˇek, F. (Ed.), Proceedings of 4th International Congress of Croatian Society of Mechanics, September 2003, Croatian Society of Mechanics, Croatia, pp. 517-22.
The Emerald Research Register for this journal is available at www.emeraldinsight.com/researchregister
The current issue and full text archive of this journal is available at www.emeraldinsight.com/0955-6222.htm
Determining temperature regulating factor for apparel fabrics containing phase change material
Determining temperature regulating factor 209
Wiesława Bendkowska Instytut Wlo´kiennictwa (Textile Research Institute), Lodz, Poland, and
Janusz Tysiak, Leszek Grabowski and Albert Blejzyk Przedsiebiorstwo Inzynierskie Kontech (Kontech Enterprise), Kontech, Lodz, Poland Abstract Purpose – In order to characterize the temperature regulating ability of fabrics containing phase change material (PCM), the test instrument has been designed and built. Design/methodology/approach – To assess temperature regulating ability, temperature regulating factor (TRF) is determined. TRF is defined by Hittle as a quotient of the amplitude of the temperature variation of the hot plate and the amplitude of the heat flux variation divided by the steady state heat resistance of the fabric. Findings – The test instrument presented here is intended to be used for testing steady state and transient state characteristics of the apparel fabrics containing the PCMs. Practical implications – This test instrument can be used in quality control during the manufacture of fabrics containing PCMs. TRF can be used in clothing industry to establish the criteria for comfort parameters of textiles. Originality/value – The instrument can provide information for the fabric and garment designers and be useful in quality control during the manufacture of fabrics with the microPCMs. The TRF can be used in the clothing industry to establish the criteria for the comfort parameters of textiles. Keywords Temperature measurement, Dynamics, Thermal efficiency, Textile industry Paper type Research paper
1. Introduction The manufacture and properties of intelligent textiles, containing phase change materials (PCM), have been extensively studied at the Textile Research Institute (Ło´dz´) since 1999 (Bendkowska, 2000). Phase change material is able to absorb, store and release large amounts of latent heat over defined temperature range when the material changes phase or state. A fabric containing PCM can act as a transient thermal barrier which regulates the heat flux (Cox, 2000; Hittle and Andre, 2001; Magill, 2000; Nuckols, 1999; Pause, 2000). The heat absorption by PCM results in a delay in microclimate temperature and hence a substantial decrease of the sweat amount produced by the skin of the wearer. Both lead to an enhancement of the wearing comfort and prevent heat stress (Pause, 2000). The authors wish to acknowledge Polish State Committee of Scientific Research for supporting this work (research project no. 4 T08E 063 23).
International Journal of Clothing Science and Technology Vol. 17 No. 3/4, 2005 pp. 209-214 q Emerald Group Publishing Limited 0955-6222 DOI 10.1108/09556220510590902
IJCST 17,3/4
210
In the manufacture of thermoregulating textiles, the most important PCMs are linear chain hydrocarbons (octadecane, nonadecane, hexadecane, eicosane, etc.) called paraffin waxes. The use of the paraffin waxes to improve thermal performance of textile materials became possible only by entrapping them in microcapsules with the diameter of 5-60 mm (Magill, 2000). Paraffin waxes possess very high heat storage capacity. They can be mixed in order to realize desired temperature ranges in which the phase change takes place. They are nontoxic, noncorrosive and nonhygroscopic, and their thermal behavior remains stable under the conditions of permanent use as well. The shell material of the microcapsules must be durable and safe through finishing process, laundering and dry cleaning. The commercial production of fabrics containing PCMs has already begun. In the past few years, apparel fabrics containing microencapsulated PCMs have appeared in outdoor garments, particularly sportswear (parkas, gloves, helmets, sky boots). Microencapsulated PCMs are incorporated into acrylic fibers (Cox, 2000), or polyurethane foams, or coated on the surface of the fabric substrate (Pause, 2000; Shim et al., 2001). Figure 1 shows a scanning microscope image of microcapsules PCM incorporated into a needled nonwoven. In the case of traditional fabrics, the thermal properties are investigated by standard steady-state procedures involving the use of guarded hot plate apparatus. Since the PCM is a highly productive thermal storage medium, steady-state procedures are inadequate for testing fabrics containing it. Hence, it is necessary to work out a test method and the instrument for assessing the temperature regulating ability of these fabrics. The test instrument has been designed and built based on a model of heat transfer through textile containing the PCM, formulated by Hittle and Andre (2001). 2. Heat transfer through textiles containing the PCM Nuckols (1999) has developed finite difference model that accounts for energy stored and released in fabrics containing PCM. Another approach was presented by Hittle and Andre (2001).
Figure 1. SEM image of polyester nonwoven containing PCM microcapsules
Heat conduction through one-dimensional homogenous material is governed by the following second-order partial differential equation:
›2 Tðx; tÞ 1 ›Tðx; tÞ ¼ ›x 2 a ›t
ð1Þ
where T is the temperature at position x; t the time; a the thermal diffusitivity, a ¼ k=ðrC p Þ; k the thermal conductivity (W/m K); r the density (kg/m3); and Cp the specific heat (J/kg K). The heat flux at the position x and time t is given by equation: qðx; tÞ ¼ 2k
›Tðx; tÞ ›x
ð2Þ
Hittle and Andre (2001) assumed that in both the above equations k, r, i, Cp are constant. Additionally, they assumed that a constant and comparably large Cp in the temperature region of the phase change is a reasonable approximation for the energy storage of the PCM in the fabric. This assumption leads to convenient metric called temperature regulating factor (TRF). Hittle and Andre offered the solution for the above equations for sinusoidal boundary conditions. In order to characterize thermoregulation effect they have proposed to use the quotient of the amplitude of the temperature variation and the amplitude of the heat flux variation, present here. The smaller the quotient the better the regulation effect. Dividing this quotient by the value of steady-state thermal resistance of fabric (R), they obtained the TRF value: TRF ¼
ðT max 2 T min Þ 1 ðqmax 2 qmin Þ R
Determining temperature regulating factor
ð3Þ
The TRF is a dimensionless number of the range (0, 1). The TRF shows how well a fabric containing the microPCM mode rates the hot plate temperature. The TRF value of 1 means the fabric has no capacitance and poor temperature regulation. The TRF equals zero means that the fabric has infinite capacitance and that a body being in contact with it will remain at constant temperature. It is obvious that all the fabrics fall somewhere between these extreme values. 3. The measuring principle Determination of the TRF of apparel fabrics is done by means of the instrument, which uses a dynamic heat source. This instrument simulates an arrangement: skin – apparel – environment. Fabric sample is sandwiched between a hot plate and two cold plates, one on either side of the hot plate. These cold plates at a constant temperature simulate the environment outside the apparel. Sinusoidally varying heat input to the hot plate simulates human activity. To measure the steady-state thermal resistance of the fabric (R) the controlled heat flux is constant and the test proceeds until a steady-state is reached. To assess temperature regulating ability, the heat flux is varied sinusoidally with time and the TRF is determined.
211
IJCST 17,3/4
212
4. Test instrument description The most important element of the test instrument (Plate 1) is the hot plate, which has a sandwich structure. The heater layer is placed between two insulation layers, made of epoxy-glass laminate. The construction of the plate provides the uniformity of temperature distribution over its surface and as small heat capacitance as possible. A thin, flexible hot plate ð0:25 m £ 0:25 mÞ is situated in the centre of the test stand. Directly above the hot plate there is a rod from which the fabric sample is hung. The sample is large enough to cover the hot plate on both sides. Controlled heat flux, either constant or varying, is maintained for the hot plate. Cold plates are situated on both sides of the hot plate. The cold plates are designed as the thinwalled tanks, cooled to specified temperature by constantly circulating water from the thermostat. The temperature of the cold plates is measured by thermoresistant sensor. The hot and cold plates can be displaced on guide bars. The cold plates can be pressed against the fabric sample at the constant pressure (200 Pa), provided by a compression spring. All control and registration functions are realized by the computer provided with the measuring chart and the application software LabView (National Instruments). 5. Experimental The TRF is the function of the frequency of the sinusoidal variation of the heat flux into the hot plate. The temperature regulation increases with increasing frequency. Hence the TRF increases with increasing cycle time of the sinusoidal variation, going exponentially from 0 to 1. Figure 2 shows the TRF as the function of cycle time for several nonwovens manufactured at the Textile Research Institute. The parameters of nonwovens are shown in Table I. Figure 3 shows cycle time – TRF curves for fabric assemblies. The whole assembly is in both cases composed of three different fabrics: outer fabric þ interlining þ lining.
Plate 1. General view of apparatus for testing thermoregulation properties of the fabrics containing the PCM
The outer layer is of woven fabric laminated microporous membrane PTFE. The interlining is of needled nonwoven 100 per cent PET. The lining is of hydroentangled nonwoven. In the case U4, the microPCM is applied to the hydroentangled nonwoven by printing. In the case U8, the lining is hydroentangled nonwoven with no PCM. During the measurements, the lining comes into contact with the hot plate, thus simulating real wearing conditions.
Determining temperature regulating factor 213
6. Summary The test instrument presented here is intended to be used for testing steady-state and transient state characteristics of the apparel fabrics containing the PCMs. The instrument can provide information for the fabric and garment designers and be useful in quality control during the manufacture of fabrics with the microPCMs. The TRF can be used in clothing industry to establish the criteria for the comfort parameters of textiles.
Figure 2. The TRF as the function of cycle time for apparel fabrics containing the PCM. PAN – needled nonwoven of 90 per cent fibers Outlast; NA – needled nonwoven of 100 per cent PET fibers; PU – knitted fabric laminated PU foam containing PCM
Fabric
Composition
NA
Needled nonwoven of 100 per cent PET fibers with microPCM added to nonwoven Needled nonwoven of: 90 per cent Outlast* fibers and 10 per cent bicomponent PET fibers Knitted fabric laminated PU foam containing PCM
PAN PU
Area weight g/m2
Thickness (mm)
Thermal resistance, R (m2 K/W)
243
5.17
0.161
170 335
6.13 3.34
0.156 0.102
Note: *Outlast is name of PAN fibers with microcapsules PCM in polymer matrix
Table I. The parameters of apparel fabrics
IJCST 17,3/4
214 Figure 3. The TRF as the function of cycle time for the assemblies of fabrics. U 5-lining with PCM; U 8-lining with no PCM
References Bendkowska, W. (2000), “Thermal insulation of apparel textiles with PCM”, paper presented at the Frankfurt am Main, November 2000, International Symposium Avantex. Cox, R. (2000), “Outlast – thermal regulation were it is needed”, paper presented at the 39th International Man-Made Fibres Congress, Dornbirn, September 2000. Hittle, D.C. and Andre, T.L. (2001), “A new procedure for evaluation of fabrics containing phase change materials”, available at: www.engr.colostate.edu (accessed January 2001). Magill, M.C. (2000), “An overview of the outlast temperature regulation technology”, available at: www.outlast.com (accessed June 2000). Nuckols, M.L. (1999), “Analytical modeling of a diver dry suit enhanced with microencapsulated phase change materials”, Ocean Engineering, Vol. 26, pp. 547-64. Pause, B. (2000), “Textiles with improved thermal capacities through the application of phase change materials (PCM) microcapsules”, Melliand Textilberichte, Vol. 81, pp. E179-80. Shim, H. et al. (2001), “Using phase change materials in clothing”, Textile Research Journal, Vol. 71, pp. 495-502.
The Emerald Research Register for this journal is available at www.emeraldinsight.com/researchregister
The current issue and full text archive of this journal is available at www.emeraldinsight.com/0955-6222.htm
Analysis of 3D woven structure as a device for improving thermal comfort of ballistic vests Eva Kunz and Xiaogang Chen
Analysis of 3D woven structure as a device 215
Department of Textiles and Paper Science, UMIST, Manchester, UK Abstract Purpose – This study presents the design, manufacture and evaluation of a type of 3D hollow woven structure, as a mean for improving ventilation underneath ballistic body armour and thus, thermal comfort. Design/methodology/approach – By means of a computational fluid dynamic package, fluid flows through different cross-sectional tubular geometries were simulated in order to predict, which structural parameters of the 3D hollow fabrics are optimal to support ventilation. Findings – As the result of the computational analysis four optimised 3D hollow woven structures were selected and generated on a standard weaving loom. Originality/value – Investigation of thermal comfort of 3D ballistic vests. Keywords Protective clothing, Computational geometry, Garment industry, Simulation Paper type Research paper
1. Introduction In some profession, like in police and security, the persons are required to wear ballistic body armour in order to minimize the risk of injuries and even death. In the USA, a study was conducted, which underlined the importance of the routine use of ballistic vests while on duty, stating that a policeman’s fatal risk was 14 times higher when not wearing a ballistic vest (National Law Enforcement and Correction Technology Centre, 2004). Despite this proven effectiveness, though many policemen tend not to put on an armoured vest, which they explain by rather uncomfortable wear conditions (Lorei, 2004). The uncomfortable clothing conditions are partly caused by the weight and bulkiness of a ballistic vest. However, a fundamental criterion for clothing comfort of body armour is to avoid the accumulation of sweat. Owing to the almost impermeable nature of ballistic fabrics, evaporation of sweat is impeded because ventilation is restricted in the area covered by the vest. Common single layered or multi-layered textile solutions for assisting thermal comfort of an armoured vest are based on the principle of transporting the moisture away from the skin. However, as moisture cannot be transferred to the outside because it is blocked by the ballistic fabric, such textiles are only partially suited for the use under a body armour. More beneficial for clothing comfort of an armoured vest is the improvement of ventilation under the ballistic inserts. This can be achieved by insertion of 3D structures between the skin and the ballistic fabric. Such constructions keep a distance between the skin and the outer layer and so facilitate ventilation within the microclimate, which allows evaporation of sweat. The study described in this paper implements a 3D hollow woven structure as a mean to improve ventilation and thus, thermal comfort of ballistic vests. It reports how computational fluid dynamics (CFD) can be used for selection of structural parameters of
International Journal of Clothing Science and Technology Vol. 17 No. 3/4, 2005 pp. 215-224 q Emerald Group Publishing Limited 0955-6222 DOI 10.1108/09556220510590911
IJCST 17,3/4
216
the 3D woven structure. Moreover, the study introduces the application of a thermal camera and thermocouples in order to evaluate thermal condition and heat development underneath a ballistic vest under wear performance tests and investigates efficiency of the new 3D woven structures assisting the body to dissipate heat. 2. Theoretical analysis For an analysis, a section of the 3D hollow woven structure is isolated and modelled as two 3D straight tubes with a uniform cross-section linked to each other. The programme we used for calculations, “FLUENT”, needs a cell structure of the space, where the flow takes place as input information. Because of the symmetry along the vertical tube axis, only one half-section of each tube is designed. Figure 1(a) shows the half cross-section of a double tube design and Figure 1(b) shows an example of the cell structure. Here the z-coordinate denotes the direction of fluid flow through the tube design. The airflow at the inlet of the tube is of a uniform velocity (w) and temperature (Ti). Since the three-dimensional fabric is heated on one side by the skin, this temperature influence is taken into account by employing constant temperature at one wall (TW), leaving the other walls insulated. The numerical model is built up on the following assumptions: (1) the fluid is treated as incompressible Newtonian Fluid, which is specified as laminar and steady state; and (2) the 3D flow is considered as axis-symmetric along the vertical tube axis. Based on the above assumptions, the equations for the conservation of heat energy and mass balance are solved by means of the FLUENT code, which is based on the control volume method (Fluent Europe Ltd, 2002). The boundary conditions are considered to be: (1) no-slip conditions at tube walls; (2) fixed velocity and temperature profile at inlet; (3) constant temperature at one wall; and (4) the walls are modelled porous. For modelling laminar fluid flow through the porous tube walls Darcy’s Law (1) is implemented with: m ð1Þ p ¼ 2 v·n a where p is the pressure loss, v the velocity, a the permeability, n the thickness of porous medium and m the dynamic viscosity (Scheidegger, 1974). In order to determine the viscous loss coefficient 1=a – a parameter describing the flow through the porous wall, a sample from a preliminary weaving range is selected to conduct air permeability test in accordance to the BS EN ISO 9237 (1995). The data gathered by the air permeability test are extrapolated to the Darcy’s law to specify the viscous loss coefficient, necessary for modelling with the “FLUENT”. The porous wall boundary conditions, as shown in Figure 1(a), can only be employed between the fluid cells. Therefore, a double tube design comprises four fluid zones: two of them represent the two half-sections of the tube design, the third and fourth are associated with the top and bottom sections of the double tube design, as shown in Figure 1(a).
Analysis of 3D woven structure as a device 217
Figure 1. Schematic illustration: (a) showing the boundary conditions of computational model; and (b) an example of meshed cross-section belonging to a trapezoidal tube design
IJCST 17,3/4
218
It is the aim of the CFD analysis to determine cross-sectional tube designs with minimal flow resistance (good flow features). The selection of the tube design is limited by the geometry. It can be woven under given conditions and offers good resilience properties after being compressed. In order to assess flow and heat transfer features of the tube design, the velocity field, pressure difference along the tube length and the heat transfer rate from the heated wall, are also recorded. In addition, the contours of the longitudinal velocity, plotted on a cross-sectional plane at the tube centre, are examined for the evaluation of the fluid dynamics. The squared, trapezoidal with 608, 768 and 308 at the baseline and triangular tube designs with 458 and 608 are compared. For comparison, all the tube designs have the same volume and in parts the same heated area. After conducting the analysis, a 768 trapezoidal and a 608 triangular tube design with a height of 5 and 10 mm, respectively, are woven. Both designs show good flow characteristic in the numerical analysis, and offer good resilience behaviour due to their sidewall position. 2.1 Manufacturing The 3D-hollow woven structures used in this study are composed of three sets of warp yarns and one set of weft yarn inserted horizontally at 908. Each series of warp yarns interweave with the set of weft yarns to form plain woven fabric layers, namely top, centre and bottom, interlinked with each other by alternately stitching the centre fabric layer to the top and bottom layer. The centre layer warp is stitched to the bottom layer by lifting all the top warp yarns and opening the centre and bottom shed simultaneously, while inserting a weft yarn from the centre layer. The structures are thus generated with the desired cross-sections, triangular or trapezoidal, at right angles to the warp direction. The 3D hollow woven structures are woven on a conventional shuttle loom. On the loom, the design is 2D and only after removal from the loom the structure is converted into a 3D design through an opening up process. A three-folded cotton yarn (50,3 tex) is used as yarn material. Weave characteristics, like warp and weft density, are derived from the preliminary weaving range, used to determine the permeability value for the computational analysis. Since the creation of the weaving instructions for such 3D structures is quite complex and time-consuming, the Windows-run CAD/CAM software “HollowCAD” is employed. This program was developed at the Department of Textiles and Paper Science at the UMIST, for the automatic generation of weave instructions for multilayer woven structures. Three warp beams provide the warp yarn supply, one for each layer, in order to set lower tension to the warp yarns for the centre and higher for the top and bottom yarns, as well as to avoid yarn friction due to the high warp density of the 45 warp yarns per cm. After weaving the four different 3D hollow woven structures, a vest design is manufactured from each structure. The vest design comprises a sleeveless vest with the front panel made out of a 3D hollow woven structure, fitted to the body at shoulder and back by means of elastic and velcro tape. The 3D structure is processed to the vest front panel with the tubular structure running across the chest width. 2.2 Testing The testing consisted of investigating the temperature changes underneath body armour during exercise. The exercise was to run at a constant pace. A single test cycle
consisted of measuring body’s surface temperature underneath the body armour before, after 10 min of exercise and finally, after additional 10 min of exercise. The temperature measurements were conducted with the Infrared Imaging System “Thermovisionw 880 AGEMA” and standard thermocouples (type K). Thermocouples readings were taken at the skin, one at the stomach and the other at the chest. Thermal shows were made from the nude upper torso, one from the front and the other from the back, before and at the end of the entire running cycle; the pictures after 10 min were omitted in order to avoid that the wearer cools down too much. Four students volunteered to take part in the study and were involved in the wear trials. The clothing ensembles were: (No) body armour and T-Shirt, (Tri-5) body armour, vest made of 3D fabric with triangular cross-section and 5 mm thickness, (Tri-10) body armour, vest made of triangular cross-section and 10 mm thickness and T-Shirt, (Tra-5) body armour, vest made of trapezoidal cross-section and 5 mm thickness, and T-Shirt and (Tra-10) body armour, vest made of trapezoidal cross-section and 10 mm thickness and T-Shirt. All participants wore a white, short-sleeved 100 per cent cotton T-Shirt and the same standard ballistic body armour for testing. As one of the students was a female, she wore additional underwear. An example of the thermal pictures made before and after the exercise is shown in Figure 2. 3. Results and discussion As mentioned earlier, all the 3D structures could be woven with standard shuttle loom, however, after take-off from the loom, none of the fabric structures did unfold satisfyingly. It was observed that the trapezoidal designs seemed to open up slightly better, which could be explained by the shorter and steeper arrangement of the centre wall layers. We tried to enhance the stability of the fabric by steam heat-setting, but with no effect. The application of a starch solution was also tried, but the improvements were only temporary. The first visual observation of the thermal pictures taken under unclothed conditions indicates temperature increase in some parts at the back and in the chest centre, over the duration of exercise. It is especially true for the pictures taken from the students 1 and 3. From the visual analysis, no difference in temperature could be observed among the trials for different clothing ensembles. For a more accurate analysis of the thermal pictures, the mean surface temperature was taken from the five points selected on the thermal pictures by means of a template. Mean temperature was determined from the chest and the stomach area. The positions of the five measurements are shown in Figure 2, for chest and stomach areas, respectively. The graphs in Figure 3(a) and (b) shows temperature differences of the mean measurements taken from the thermal pictures at the chest (Figure 3(a)) and stomach (Figure 3(b)), after the 20 min exercise, wearing various clothing ensembles. The mean skin temperature at the chest increased by up to 18C for the students 1 and 3, while for the students 2 and 4 the skin temperature partly decreased after the 20 min run. Apart from the student 4, all the others had lower skin temperatures by the average of 18C, or even up to 28C at the stomach, by the end of exercise. These results suggest that the design and fit of the ballistic vest affects the temperature development underneath the ballistic fabric. In the areas, where
Analysis of 3D woven structure as a device 219
IJCST 17,3/4
220
Figure 2. Thermal pictures: (a) taken before the exercise; and (b) 20 min after the exercise
Analysis of 3D woven structure as a device 221
Figure 3. Mean skin temperature difference after 20 min exercise, measured with thermal camera: (a) chest; and (b) stomach
IJCST 17,3/4
222
the ballistic vest is worn thigh-fitting, which was the case for the chest of the students 1 and 3, the increase in temperature could be clearly detected during exercising. The lower skin temperatures on the chest level of the students 2 and 4 could be explained by the fact that the body armour used did not fit properly to the chest. While exercising, the pumping actions due to the wearer motion caused ventilation between the ballistic vest and the skin, thus lowering skin temperature. Similar explanation could be employed for the temperature observation at stomach level. When the vest was looser fitting, as in the case of the students 1-3 in Figure 3(b), the gaps between the ballistic vest and the skin allowed ventilation to take place, which resulted in the skin temperature decrease during the exercise. The higher mean temperature at the stomach level of the student 4 could be explained by the fact that the female student 4 wore additional underwear during imaging. The histograms in Figure 4 show skin temperature differences measured using the thermocouples at chest (Figure 4(a)) and stomach (Figure 4(b)) after the 20 min run. The results in Figure 4(a) show temperature increase at the chest in the case of the students 1-3. The lower skin temperatures of the student 4 after exercise can be explained by the fact that the student 4 was a female, and the body armour used did not fit properly around the chest, but around the waist. Among the results taken at the stomach, the temperature measurements from the student 3 are striking. They show that the student 3 had noticeably lower skin temperature after the exercise. The reason was that the body armour fitted quite well around the test person 3’s chest, whereas it was loose fitting around the stomach, since the person 3 was a trained athlete. Furthermore, the results suggest that the temperature rose after 20 min exercise was significantly higher when measured with thermocouples than when detected with a thermal camera. The mean temperatures taken from the thermal pictures were scattered around the initial mean temperatures. One reason for this was the tape, which was used to fix the thermocouples to the skin, and which caused thermal insulation. The other is that the thermocouple measurements were made first with clothing, whereas the thermal pictures were taken after the test person took off the clothing ensemble and thus, cooled down. From the results shown in Figures 3 and 4, it is almost impossible to determine the clothing ensemble, which would favourably improve ventilation underneath the ballistic vest. The testing clearly proves that the ballistic body armour blocks the exchange of air and thus of heat and moisture. Indeed, all the test persons perspired after finishing the test run and reported chilly perception after taking off the ballistic body armour. This cooling effect is caused by the evaporation of sweat; therefore, great care should be taken to keep the time intervals between the end of the run and the taking of the equal. The measurements indicate that the ventilation within the investigated 3D hollow woven structures does not significantly alter heat exchange. It seems that the contribution of the air circulation in the gap between the different fabric layers is of the same order of magnitude as the pumping action of the 3D hollow woven structure. 4. Conclusion The paper presents engineering, production and evaluation of the 3D hollow woven structures, as a device for improving ventilation underneath ballistic body armour.
Analysis of 3D woven structure as a device 223
Figure 4. Skin temperature difference after 20 min exercise measured with thermocouple at: (a) chest; and (b) stomach
IJCST 17,3/4
224
Within the engineering stage, the CFD software “FLUENT” is used to determine optimal designs of the 3D structures. In preparation for weaving, the CAD/CAM software “HollowCAD” is used to design the weave instructions. The desired 3D hollow woven structures can be produced on conventional shuttle looms, though further work needs to be done to enable the woven structures to unfold more properly. It could be advantageous for the unfolding of the structure if at least the centre layer is generated out of synthetic yarn material, possessing higher stiffness, as, for example, nylon or polyester. It would further be beneficial to the functioning of the vest if the 3D hollow woven structure could be firmly connected to the body armour, as this would prevent the 3D hollow woven structure from crinkling and destabilising. The advantageous use of the thermal camera to measure the complete set of skin temperatures is demonstrated as well. Theoretically, the set comprises around 64,000 temperature values taken with the camera used for this work. It should be compared to only a few temperature values received by the thermocouples. References BS EN ISO 9237 (1995), “Textiles-determination of the permeability of fabrics to air”, BSI, London. Fluent Europe Ltd (2002), FLUENT Manual 6.0, Sheffield, 2002, available at: www.fluent.com Lorei, C. (2004), Das Tragen von Schutzwesten bei der Polizei (in German), available at: www. schusswaffeneinsatz.de/Schusswaffeneinsatz/literatur.htm (accessed 1 March 2004). National Law Enforcement and Correction Technology Centre (2004), Selection and Application Guide to Police Body Armor, National Institute of Justice, available at: www.nlectc.org/ txtfiles/selectapp.html (accessed 1 March 2004). Scheidegger, A.E. (1974), The Physics of Flow Through Porous Media, University of Toronto Press, Buffalo.
The Emerald Research Register for this journal is available at www.emeraldinsight.com/researchregister
The current issue and full text archive of this journal is available at www.emeraldinsight.com/0955-6222.htm
Fabric design considering the optimisation of seam slippage Rui Alberto Lopes Miguel, Jose´ Mendes Lucas, Maria de Lurdes Carvalho and Albert Maria Manich
Fabric design
225
R&D Unit of Textile and Paper Materials – Textile Department, University of Beira Interior, Covilha˜, Portugal Abstract Purpose – The dependence of seam slippage values on fabric construction parameters makes this property an interesting case for study. Design/methodology/approach – In this study, made on a significant wool and blended fabrics sample, the seam slippage was measured, either in warp direction (weft yarns slip), or in weft direction (warp yarns slip), using a specially equipped load-elongation tester. Testing was done following the TM 117 Woolmark Company test method. Findings – For most fabrics, the conventional variables that impact seam slippage most seriously are opacity, polyamide content, finish type and cover factor. Research limitations/implications – Since this research does not deal with the variable of yarn crimp in fabrics, it is the cover factor that plays the central role, as the property determining seam slippage. The yarns with lower cover factor (less crimped) are in less danger of slipping between the perpendicular yarns (more crimped) and vice versa. Practical implications – Based on the equations given, and changing the most relevant variables concerning the explanation of the fabric seam slippage property, the fabric properties can be optimised for specific end-uses. Originality/value – Optimisation of seam slippage in fabric design. Keywords Wool fabric, Fabric testing, Modelling, Statistics Paper type Research paper
1. Introduction Garment products are under a constant pressure of designing towards the aim of humanising the environment around us – the concept centred on the user – primarily because of the proximity of garments and human body. The properties that influence garments performance related to the end-use, like seam slippage, have been on top of textile research and development efforts. The force of seam slippage, generally known as seam slippage, measures the ability of the warp yarns to slip over the weft ones near the seam, which extends in the direction of the warp, when the fabric is subjected to a given load in the weft direction (and vice versa). This load is applied so as to separate the two pieces of the fabric joined by the seam, and thus an opening, which is the result of yarn slippage, appears near the seam. It is an important parameter for fabric characterisation, especially for garment making, and its value depends on the fabric construction and finishing applied. Wool and wool blend fabrics samples are used in this research, with the aim to measure seam slippage, either in the warp direction (weft yarns slip), or in the weft direction (warp yarns slip), using a specially designed load-elongation tester. The testing is done following the TM 117 Woolmark Company test method. The instrument applies
International Journal of Clothing Science and Technology Vol. 17 No. 3/4, 2005 pp. 225-231 q Emerald Group Publishing Limited 0955-6222 DOI 10.1108/09556220510590920
IJCST 17,3/4
226
a constant load, which causes slippage of yarns placed across to the direction of the load. The resulting opening between yarns, measured in millimeter, represents seam slippage. As fabric seam slippage is a function of its technical construction, the objective of this study is to correlate the “seam slippage” property with its structural characteristics, in order to find the best statistical models to explain the relationship, as well as the fabric construction variables which influence the behaviour of the fabric in the view of seam slippage most precisely. The idea is to find the means to optimise, as a function of wear, the performance of fabrics in relation to seam slippage, so as to correctly manipulate the proper structural characteristics (Miguel and Lucas, 2002). 2. Materials and methods 2.1 Selected fabrics The investigation (Miguel, 2000) is done using 82 wool and blended fabrics, 51 worsted and 31 woollen ones, having different structural characteristics. The only property common to all the fabrics is the wool in their compositions. Thus, selected fabrics are characterised by different composition, weight per square meter, yarn type, opacity, weave and finish, which makes them representative of most industrial fabrics in the wool industry. 2.1.1 Fabric characterisation. The evaluation of fabric structural characteristics is done by laboratory testing and according to current standards. The following are the characteristics measured and their corresponding units. . composition (La, PE, PA, EA) – in per cent – standard NP 2248; . average fibre fineness (Fi) – in m – standard NP 3160; . average fibre length (Co) – in mm – standard BS 6176:1981(1991) – WIRA apparatus; . yarn count and number of yarn plies (Tex, NC) – in tex – standard NP 4105; . twist of yarns and plies – turns/m – standard NP 4104; . weight per square meter (PM2) – in g/m2 – standard NP 1701; . thickness (ESh) – in mm – standard UNE 40-224-73 – Louis Schopper apparatus; . yarn density – in yarns/cm – standard NP EN 1049-2; and . weave – standard NP 4114/1700. Based on the structural characteristics measured and their mathematical correlation, the parameters that impact the complete definition of fabric construction are established: twist factor (CTt), weave coefficient (CL), weave type (TL), average float (Amed), cover factor (FaC), opacity (CoT) and porosity (Po). The type of finish (TA) is also determined, according to a 1-5 scale, as a function of the fabric felting degree. 2.2 Measuring techniques applied Seam slippage was measured, either in the warp direction (weft yarns slip), or in the weft direction (warp yarns slip), using specially designed load-elongation tester. The testing was done following the TM 117 Woolmark Company test method. The instrument applied constant load, causing slippage of the yarns placed under the right angle to the direction of the load. The total opening between the yarns, measured in millimeter, was established to be the seam slippage.
3. Results and discussion 3.1 Equations that explain the properties most precisely The best equations found to explain the fabric seam slippage property, respective determination coefficient (R 2) and the significance level (P) are shown in Table I (Miguel, 2000).
Fabric design
3.2 The importance of equation variables The relative weight of equation variables, which explain fabric seam slippage property most precisely, is shown in Table II.
227
3.3 Discussion of the equations 3.3.1 Discussion of the equation for the worsted and woollen fabrics. The analysis of the equation that explains the property under study lead to the conclusion that seam slippage (DC) decreases with the increase of opacity (CoT), polyamide composition (PA) and finish type (TA). The warp (or weft) seam slippage decreases as the weft (or warp) cover factor decreases (IFaC) (Miguel, 2000). Fabric opacity describes the relationship between the actual and maximum cover factors of a fabric. Thus, it allows evaluating the fact that the actual density, for the same yarns, is near or not near to the maximum one. In the first case, the corresponding fabric opacity is close to 100 per cent. Maximum fabric density depends on raw material content, yarn count and fabric weave. This means that, in order to have high fabric opacity, the following conditions should be fulfilled: (1) The same raw material and weave should be kept, in order to maintain yarns as close as possible to each other yarn density yarn count in tex, or both, should be increased.
Fabric group
Accuracy R 2 (per cent) P (per cent) level
Equation
Woollen and worsted DC ¼ 17:8 2 1:4ESh 2 0:3TA 2 0:1PA 2 0:2CoT 2 1:2NC þ 0:03ICTt þ 0:2IFaC þ 0:5TL Woollen DC ¼ 2:6 2 1:0TAm þ 0:01Lam2 2 0:3FaCm þ 0:8IAmedm þ 0:03ITexm þ 1:0TLm 2 0:2Com 2 0:5IAmedm2 Worsted DC ¼ 3:7 þ 1:6NCm2 2 0:1Texm 2 0:003Texm2 2 0:2FaCm þ 0:5TLm 2 0:01ITexm2 þ 0:2Com þ 0:04Com2 þ 0:03Pom2
Fabric group Woollen and worsted Woollen Worsted
64.2
0.1
Medium
90.7
0.1
High
59.9
0.1
Medium
Variables (per cent) CoT 2 22 TA 2 52 FaC 2 23
PA 2 21 La 2 24 Co 2 11
IFaC and TA 2 6 IAmed 2 7 Tex 2 9
Table I. Equations explaining the fabric seam slippage property
Not explained (per cent) 35 9 40
Table II. Relative weight of the most important equation variables
IJCST 17,3/4
228
(2) If the yarn count in tex and raw material are constant, to keep yarns as close as possible, either the yarn density should be increased, or the weave average float should be decreased. It is evident that fabrics having high opacity are high cohesion structures, with compact and/or strongly interlaced yarns. Fabrics possessing these characteristics present higher difficulty to yarn separation during seam slippage. Internal cohesion of the fabrics can also be increased by increasing the finish type applied. As a matter of fact, the increase of finish type means higher wool felting degree and, consequently, stronger fibre entanglement. It results in the inability to separate yarns during seam slippage. The influence of polyamide composition on seam slippage, especially in the case of woollen fabrics in the example studied, confirms the above. Fabrics with polyamide in its composition generally exhibit high felting and raising degree, resulting in high finish types. The influence of weft cover factor on the warp seam slippage (and vice versa) can also be explained. Fabrics with high warp cover factor generally show a lower weft cover factor, for the overall factor to be balanced. Thus, the decrease of the weft cover factor probably indicates an increase of the warp cover factor and a consequent increase of friction between the warp and weft yarns, this making seam slippage difficult. For the same reasons, and for the fabrics having these characteristics, the weft seam slippage increases (warp yarn slippage becomes easier). The change of friction between the warp and weft yarns described before is basically associated with the contact points between the yarns and with yarn crimp. Low-density weft yarn combined with high-density warp yarn creates a number of contact points with the warp yarns. On the other hand, the weft becomes slightly crimped because high-density warp does not allow the existence of spaces inside it, in order that the weft could pass through. In this way, so that the weft could go from one selvage to the other, the warp yarns should be placed into two levels, one above and other below the weft. High-density warp and low-density weft generate a few contact points between yarns. The warp gets a high crimp, since the weft is quite straight; in order to reach the necessary crimp required by the weave. In short, fabrics having much higher warp cover factor than the weft one, exhibit limited slippage of the weft against the warp, as there is a number of contact points between them, and the adjacent warp yarns are highly crimped. On the other hand, warp slippage through weft is easier when there are few contact points and the weft crimp is lower. This phenomenon is often recorded in the “gabardine” fabrics, made with a 2/1 twill 3 weave and having considerably higher warp cover factor than the weft one, as a consequence of higher warp density and frequently coarser warp yarns as compared to the weft. These fabrics have higher seam slippage (sometimes critical) in the weft direction than in the warp direction. 3.3.2 Discussion of the equation for the woollen fabrics. The analysis of the equation that explains the property studied leads to the conclusion that seam slippage (DC) decreases with the increase of fabric finish type (TA). Seam slippage has a maximum value when the wool composition (La) is near to 75 per cent. The warp (or weft) seam slippage exhibits a maximum value when the warp (or weft) average float (IAmed) is near 2.8 (Miguel, 2000).
The influence of the finish type has already been explained. For the woollen fabrics, the finishing range is higher, as compared to the worsted ones and, for this reason, this characteristic can be used to explain the seam slippage in these fabrics. Speaking of wool composition, it is also under the influence of finish type. Thus, seam slippage reaches its minimum (lowest value) for the wool composition near 75 per cent, meaning that the best fabrics are made of a wool polyamide blend (Figure 1). Consequently, these fabrics generally exhibit high felting and raising degree, that is to say, high finish type. On the other hand, and for this type of fabrics, an increase in wool composition results in the increase of warp and weft yarn friction. The influence of the inverse average float on seam slippage is not very strong and, given its high correlation with the corresponding direct variable, its presence in the model should be attributed to a better statistical fit (Figure 1). For this reason, the influence of the average float is also represented in the influence of fabric opacity. As a matter of fact, seam slippage decreases with the increase of opacity, as explained before. For the fabrics with an average float between 1 and 2, the increase of the average float corresponds to a decrease of opacity (Section 3.3.1). These are fabrics where weave is changed, which consequently changes the maximum cover factor, without a significant alteration of the overall and actual cover factors, due to the type of application. This change of fabric opacity, caused by the decrease of the number of weaving points, results in the increase of seam slippage. As a result, the increase of the average float in the warp results in a decrease of the number of weaving points of the weft, lower friction and easier seam slippage. For the fabrics with an average float equal or above 2 (2 and 4), the increase of the average float corresponds to the increase of opacity. Even with an identical average float increase, as in the case described, (from 1 to 2 and 2 to 4), which implies the same weave coefficient reduction, the decrease takes place in a more critical region of the fabric consistency. This means that the reduction of the normal construction fabric rigidity, associated with the increase of the average float, impacts the fabric consistency less when the increase is from 1 to 2 than from 2 to 4. This is so because in the first case maximum rigidity (as a function of weave) is quite close. In order for the fabrics with a low weave coefficient, corresponding to an average float of 4, to increase
Fabric design
229
Figure 1. The influence of the wool composition (La) and of the inverse average float (IAmed) on seam slippage
IJCST 17,3/4
230
Figure 2. The influence of the average fibre length and yarn count (in tex) on seam slippage
its consistency to an adequate level (good wear performance), it is necessary to increase the overall cover factor as a mean of compensation. This is why the seam slippage decreases as the average float increases, having in mind the above two values for this variable. 3.3.3 Discussion of equation for the worsted fabrics. The analysis of the equation that explains the property studied lead to the conclusion that the seam slippage (DC) decreases with the increase of cover factor (FaC). Seam slippage is at its maximum when the yarn count (Tex) is near of 32 tex. Seam slippage is at its minimum when the values of the average fibre length (Co) are near 61.5 mm and the number of yarn plies (NC) near 1.85 (Miguel, 2000). As explained in Section 3.3.1, the decrease in cover factor leads to the increase of seam slippage and vice versa. The fabrics with an average fibre length between 58 and 62 mm are 100 per cent wool fabrics. In these, the increase of the average fibre length corresponds to the decrease of twist coefficient and the change of 2-1 yarns plies. Likewise, there is a decrease of the average float (increase of the weave coefficient) and the fabric opacity also rises. Thus, the conditions are met that allow for a decrease of seam slippage (Figure 2). The fabrics with an average fibre length between 62 and 69 mm comprise 100 per cent wool and polyester/wool blends. For these fabrics, the increase of the average fibre length corresponds to the increase of the polyester content; the average fibre fineness and yarn count in tex. There is a group of polyester/wool worsted yarns with the average count of tex 25 £ 2; made of medium/coarse fibres (about 23 m), as the result of being blended with polyester, that have generally longer fibres than wool. This usually leads to lower woven structures (higher average float), lower fabric opacity, allowing for an increase of seam slippage. The fabrics made of yarns of the count in the range from 26 to 32 tex are the fabrics of thin yarns, with the finish type “washed only”. In this class of fabrics, the weave has a stronger impact on seam slippage, since there is no change of the finish type. An increase of the yarn count (in tex) corresponds to the increase of the average float (decrease of weave coefficient), resulting in lower seam slippage (Figure 2).
The fabrics having yarns of the count between 32 and 69 tex are the fabrics with several types of finish. An increase of the yarn count in tex corresponds to the increase of the fabric weight/m2 and, in general, to the increase of the finish type and fabric thickness. The consequences are clear, and cause the specific appearance of a large amount of the fibres per unit surface area of the fabric, and higher fibre entanglement resulting from higher internal cohesion and friction, which are the conditions reducing seam slippage. 4. Conclusions For most fabrics, the conventional variables that impact seam slippage most seriously are opacity, polyamide content, finish type and cover factor. Since this research does not deal with the variable of yarn crimp in fabrics, it is the cover factor that plays the central role, as the property determining seam slippage. The yarns with lower cover factor (less crimped) have more difficulty to slip between the perpendicular yarns (more crimped) and vice versa. Based on the equations given, and changing the most relevant variables concerning the explanation of the fabric seam slippage property, the fabric properties can be optimised for specific end-uses. References Miguel, R.A.L. (2000), “Modelling the influence of structural characteristics on wear properties of wool and blended fabrics”, PhD thesis, University of Beira Interior, Covilha˜,. Miguel, R.A.L. and Lucas, J.M. (2002), “Computer simulation of statistical evaluation and optimisation of wool and blended fabrics quality”, paper presented at The Textile Institute 82nd World Conference, Cairo.
Fabric design
231
The Emerald Research Register for this journal is available at www.emeraldinsight.com/researchregister
IJCST 17,3/4
232
The current issue and full text archive of this journal is available at www.emeraldinsight.com/0955-6222.htm
An experimental simulation of human body behaviour during sweat production measured at textile electrodes Georgios Priniotakis, Philippe Westbroek, Lieva Van Langenhove and Paul Kiekens Department of Textiles, Faculty of Applied Sciences, University of Gent, Gent, Belgium Abstract Purpose – In this paper an electrochemical cell is developed to test and follow up the quality of electrodes made of knitted, woven and non-woven conductive textile material. Design/methodology/approach – This cell is constructed of two electrodes planarly positioned against each other using the support of a PVC tube and two PVC plates. Between the electrodes and the electrolyte special membranes are placed that simulate the human skin. Findings – This research is a preliminary start of a study to investigate and understand the behaviour of textile electrodes and to gain insight in the inter-phases electrode-electrolyte and electrode-skin-electrolyte in order to be able to model the system and to use it for detection of parameters and body conditions. Research limitations/implications – As pointed out earlier, a lot of work still needs to be done but the preliminary work shows that promising possibilities can be offered. Originality/value – Simulation of human body behaviour during sweat production measured by textile electrodes. Keywords Textile testing, Electrochemical devices, Impedance voltage, Fibre testing, Electrodes, Simulation Paper type Research paper
International Journal of Clothing Science and Technology Vol. 17 No. 3/4, 2005 pp. 232-241 q Emerald Group Publishing Limited 0955-6222 DOI 10.1108/09556220510590939
1. Introduction The research, development and application activities in the field of intelligent textiles have grown tremendously over the past ten years. As a result an important amount of intelligent textile products were developed, such as respibelt, smart shirt, life shirt, wearable mother boards, cardiogram detection, smart bra, socks and gloves, MP3 jacket and intelligent protective clothing (Holcombe and Wallace, 2002; Rogale and Dragcˇevic´, 2001; Rogale et al., 1999; Van Langenhove and Hertleer, 2003). Despite the useful applications of these developed products, they are mainly used for qualitative measurement of parameters (e.g. cardiogram, respiration rate) and often show limited reproducibility, accuracy and selectivity. This is mainly caused by a trial and error approach instead of systematic, well-planned and managed research. Another reason for the limited reproducibility of intelligent textile sensing systems is their use in a very complicated environment, such as the human body surface. A lot of parameters can possibly interfere at the measured signal (e.g. a measured potential is not only dependent of heart beat but also of neuron reactions in the body, humidity
and oxygen concentration of the surrounding air, humidity of the skin in contact with the intelligent textile electrodes, contact surface between textile electrodes and skin surface, . . .) making it almost impossible to get reproducible results for the parameter of interest. In addition, the higher mentioned intelligent textile electrode products are mainly developed and used immediately at the human body (Pacelli et al., 2001; Van Langenhove et al., 2003; Schedukat et al., 2003). In these cases it is impossible to get reproducible results and/or understand and gain insight in the working mechanism of the sensor system because modelling of all parameters that contribute to the signal is too complicated. It becomes clear that, despite the promising possibilities for applications of intelligent textiles, still a lot of fundamental work needs to be done to understand the behaviour of the electrodes and their behaviour in contact with the human skin (Priniotakis et al., 2004a; Westbroek et al., n.d.; Westbroek et al., 2004; Priniotakis et al., 2004b; Kiekens et al., 2004). The results of this paper show a first attempt to model the interphase between textile electrodes and the human skin by using an electrochemical cell. This cell is described in the experimental section and mimics the system textile electrode-skin-electrolyte-skin-textile electrode as it is the case if two textile electrodes are positioned planarly at the human body, e.g. at an arm or a leg.
Simulation of human body behaviour 233
2. Experimental The electrochemical cell consists basically of two PVC plates, as shown schematically in Figure 1. The electrodes are positioned at the inner side of the plates (1) with rubber fittings (2), having a hole of a specified diameter. The distance between the electrodes is determined by the length of the tube (3) positioned between the plates, this tube is filled with electrolyte solution (4). The complete structure is kept together with screws (5). The metal sheets used to characterise the electrochemical cell consisted of 99,995 per cent pure Pd metal obtained from Goodfellow. The palladium surface was polished on emery paper 1,200 grid to obtain a fresh surface and further polished with Al2O3 polishing powder on polishing cloth. To smoothen the surface consecutively Al2O3
Figure 1. Scheme of the electrochemical cell consisting of: (1) PVC plates; (2) rubber ring fittings; (3) PVC tube; (4) electrolyte solution; (5) screws; and (6) textile electrodes
IJCST 17,3/4
234
powder of 1; 0.3 and 0.05 mm was used, respectively, for 5, 10 and 20 min. Finally, the surface was subjected to ultrasonic cleaning for 1 min. The textile electrodes were obtained from Bekintex and were composed of stainless steel fibres. Knitted, woven and non-woven structures were tested. NaCl was purchased from FLUKA, while doubly deionised water was used to prepare the solutions. For the experiments done with artificial sweat as electrolyte a solution containing 20 g l2 1 NaCl, 1 g l2 1 urea, 500 mg l2 1 of other salts and NaOH or HCl was used to maintain the pH of the sweat at a value of 5.8. Teflonw membranes were obtained from goodfellow. These membranes have a thickness of 0.175 mm, a pore density of 35per cent and the average pore diameter was respectively 5.0 and 0.45 mm. For the experiments, a potentiostat PGSTAT20 of ECO chemie was used, extended with a frequency response analyser (FRA) module in order to be able to perform impedance measurements. The frequency of the alternating potential that was applied was varied from 1 mHz to 1 MHz with maximum amplitude of 10 mV. 3. Results and discussion 3.1 The characterisation of the electrochemical cell First the electrochemical cell, described in the experimental section was characterized by using three different types of intelligent textile structures, a knitted, woven and non-woven structure, made of stainless steel fibers. Characterisation of the cell was done by applying an alternating potential between the electrodes positioned in the electrochemical cell and by measurement of the resulting alternating current at different frequencies in a range from 0.1 Hz to 1 MHz. From the applied potential and the measured current it is possible to obtain the impedance, which is composed of the real (R) and imaginary impedance ( jvRC) resulting in a Nyquist plot. Z¼
R 1 þ jvRC
ð1Þ
An example of such a plot is given in Figure 2(a), recorded in the electrochemical cell with knitted textile electrodes of 314 mm2, a distance of 112 mm between the electrodes and varying electrolyte concentration from 0.1 to 102 3 mol l2 1. For a complete characterization of the electrochemical cell the knitted textile electrode surface area exposed to the electrolyte solution was varied from 19.6 to 491 mm2 and the distance between the electrodes from 27 to 112 mm. In Figure 2(b) the real impedance and the phase angle shift between applied potential and measured current is given as a function of frequency of the applied potential and was obtained from the same experiment as performed for the data in Figure 2(a). The curves in Figure 2 are explained as follows: the used electrochemical cell is of the type of a typical conductivity cell and the equivalent electrical circuit corresponding to this cell is given in Figure 3(b), which is an extended form of the equivalent circuit in Figure 3(a). In Figure 3(a) the equivalent circuit for one electrode and its interface with the electrolyte is shown. This situation can be obtained by using a so-called three electrode setup, which will not be further described in detail. Rct is the charge transfer resistance, which is mainly the resistance that should be overcome to transfer electrons from the
Simulation of human body behaviour 235
Figure 2. (a) Nyquist; and (b) Bode plots recorded at palladium electrodes positioned in the electrochemical cell with a distance between the electrodes of 112 mm, and electrode surface area of 314 mm2 and electrolyte concentrations of: (1) 101; (2) 102 2; (3) 102 3 mol l2 1 NaCl
IJCST 17,3/4
236 Figure 3. Equivalent electrical circuit corresponding to the system in the electrochemical cell
liquid electrolyte to the solid electrode substrate. Zw1 expresses the Warburg impedance and corresponds to the impedance caused by diffusion of charged species in solution. Finally, when applying a potential over the electrodes an electrical double layer is established at the interfase of the electrodes with the electrolyte solution, hence the presence of a capacitor Cdl in the equivalent circuit. However, in this investigation it is not possible to use the eq uivalent circuit shown in Figure 3(a) because two identical electrodes are used instead of a three electrode setup. The equivalent circuit corresponding to this system is given in Figure 3(b), it can be seen that two elements are added to the EC of Figure 3(a), which are R1, the resistance of the electrolyte and C1 a geometrical capacitance, a parameter that is dependent of the positioning of the electrodes. From this circuit it can be seen that in the high frequency range the current flows exclusively through the capacitive part because its charging and discharging is so fast that it acts as a conductor. In Figure 2(a) this can be seen as an impedance that consists only of an imaginary part, with the real impedance near to zero. In Figure 2(b) this results in a phase angle shift of 908, a typical capacitive behaviour. However, with decreasing frequency the charging and discharging rate becomes slower, resulting in an increasing resistive behaviour of the capacitor, while the resistive behaviour of the resistive part R1 in the circuit remains constant. Therefore an increasing current fraction will flow through the resistive part R1 (the other fraction still flows through Cdl) resulting in a decreasing phase angle shift (Figure 2(b)) and the formation of a semi circle (Figure 2(a)). At the frequency where the semi-circle is completed, the electrical current flows only through the resistive part R1. Indeed at this point the imaginary impedance is zero, hence, the total impedance is equal to the real impedance. For the frequencies corresponding to the completion of the semi-circle a phase angle shift of zero is observed, which corresponds to pure resistive behaviour. Under these conditions the real impedance is equal to the resistance in direct current mode. It is this resistance that is of our concern and that reflects the resistance of the electrochemical cell. If the frequency is decreased further, diffusion of ions from the electrolyte becomes important, which is expressed by the Warburg Impedance (ZW). However, this type of impedance is not relevant for this investigation and not further discussed. It should be noted that in principle a second semi-circle should be obtained determined by Rct but in
this case it is not observed because the Warburg impedance interferes seriously at lower applied frequencies. For each experiment performed at different electrolyte concentrations (102 5, 102 4, 23 10 , 102 2 and 102 1 mol l2 1), different electrode surface areas (19.6, 78.5, 176, 314 and 491 mm2) and different distances between the electrodes (27, 91 and 112 mm) the real impedance at a phase angle shift of zero was determined and plotted logarithmically. Linear relationships are obtained between log R and log c (with c the concentration of the electrolyte solution) up to concentrations of 102 4 mol l2 1. The slope of these curves is near 2 1 ^ 0.05 which means that the obtained resistance of the cell is linear with c 2 1 as expected from theoretical considerations. It can also be seen that the slope of the relationships does not depend on the surface area and the distance between the electrodes. The curves are shifted towards less resistive behaviour (smaller R) for larger values of the electrode area and for decreasing distance between the electrodes. This is shown also in Figure 4 where log R is given as a function of electrode surface area A, for d ¼ 91 mm and for distance between the electrodes d, with A ¼ 491 mm2 : A slope of 2 0.65 ^ 0.03 is obtained for log R vs log A and 1 ^ 0.04 for log R vs d. The latter means that the resistance increases linearly with the distance between the electrodes. Similar results were obtained for the other values of A and d investigated in this work. Also for the other types of electrodes similar results were obtained. Again linear relationships were observed, which allows to quantify and characterize the electrochemical cell by the following equation: R¼k
d 0:98 cðAÞ0:62
ð2Þ
with R the resistance in V, k a constant being 22.5^ 1.0 at 298.0 K, d the distance between the electrodes in mm and A the surface area of the electrodes in mm2. This section of the investigation shows that quantifiable and reproducible results can be obtained for parameters such as electrolyte concentration, distance between the planarly positioned electrodes and surface of the electrodes. 3.2 Implementation of membranes in the electrochemical cell In this section of the investigation membranes were inserted in the electrochemical cells. These membranes are positioned between the textile electrodes and the electrolyte and mimic the skin of the human body. Before describing results it should be pointed out very clear that inserting these membranes makes the system much more complicated. Therefore it was at this moment only possible to obtain qualitative results. Quantitative analysis requires much more experiments with different types of membranes, cell configurations and variation of relevant parameters. This investigation will start in the near future. For the first set of data two different membranes were used, both made from Teflon (PTFE). The first membrane has pore sizes of 5 mm, a diameter that allows electrolyte solution to penetrate through the membrane. The second membrane has pore sizes of 0.45 mm, which is a membrane that does not allow electrolyte to penetrate easily through the membrane. The purpose of this type of experiments is to investigate the electrochemical impedance behaviour of the systems, which in fact mimic the human body in rest (no sweat production) and during sweating (electrolyte penetrating through the pores of the membrane or the skin).
Simulation of human body behaviour 237
IJCST 17,3/4
238
Figure 4. Calibration plots of: (a) log R vs log A; (b) log R vs log d (b) recorded at palladium electrodes in the electrochemical cell for electrolyte concentrations of 102 1 (1), 102 2 (2), 102 3 (3) and 102 4 mol l2 1 NaCl (4) A ¼ 491 mm2 in (b) and d ¼ 91 mm
The equivalent circuit shown in Figure 3(b) is extended with an additional RC element that expresses the presence of a membrane (Figure 5). Rm shows the resistance within the pores of the membrane, a resistance that is different from R1 because the microstructure in the pores results in different properties compared to the bulk of the electrolyte. Cm is the capacitive contribution of the interfase electrolyte – membrane and depends strongly on the contact surface of this inter-phase, thus this capacity depends on the pore size and the pore density. Experimental curves obtained in a cell with an electrolyte solution concentration of 102 2 mol l2 1 are shown in Figure 6. A clear difference can be observed, for the membranes with pore size of 5 mm two well-defined semi circles are detected, while for the membranes with pore size of 0.45 mm only one well-defined semi circle was detected. These data can be explained as follows for the left (5 mm pores) and the right (0.45 mm pores) figure in Figure 6, which are measured as a function of time. For the left not much difference is observed because equilibrium is obtained very fast, for the right one it takes some time (about 1 day) to obtain equilibrium due to the small pore sizes. For both figures a relatively small semi circle is obtained at high frequencies with a resistive value of about 8,000 V in the left figure, an unmeasurable resistance in the right one. This is in fact the electrolyte solution resistance R1, which is the only resistance to be measured at high frequencies. For lower frequencies, which means that Cm starts to behave as a resistive element, the current flows through R1 and Rpo resulting in a second semi circle with R ¼ 18:000 V: Probably a third semi circle can be detected but its shape us to rough to decide if this is a semi circle or not. More detailed information should be obtained first. In the right figure only one semi circle is observed and it takes quite a lot of time to obtain a steady shape of the semi circle. In the first curves it is not possible to obtain a semi circle because at that moment there is no contact between the electrolyte and the electrode surfaces due to the membrane acting as a barrier. In this case no charge can be transferred over this non-existing inter-phase that determines the total shape of the curve. After a few hours some electrolyte has penetrated in the pores and makes contact with the electrode, but still the actual contact surface is very small, which explains the high resistive behaviour (typical resistances of MV). The data above 3 MV are unreliable because for those high impedances the impedance of the equipment and the connecting wires starts to contribute to the overall impedance.
Simulation of human body behaviour 239
Figure 5. Extended equivalent circuit for presence of membrane
IJCST 17,3/4
240
Figure 6. Impendance plots as a function of time recorded in 102 3 mol l2 1 NaCl using a membrane with pore size of 5 mm (left) and 0.45 mm (right)
However, based on these results a first qualitative detection can be done by differentiating between a body that is sweating, e.g. during labour, sports, fever, and a body that is not sweating (during rest, sleep). 4. Conclusion This research is a preliminary start of a study to investigate and understand the behaviour of textile electrodes and to gain insight in the inter-phases electrodeelectrolyte and electrode-skin-electrolyte in order to be able to model the system and to use it for detection of parameters and body conditions. As pointed out earlier still a lot of work needs to be done but the preliminary work shows that promising possibilities can be offered. References Holcombe, B. and Wallace, G. (2002), “The brave new world of wearable intelligence”, Wool Technol Sheep Breed, Vol. 50, pp. 312-8. Kiekens, P. et al. (2004), “Electrochemistry as an analytical tool to study intelligent textile electrodes”, paper presented at ICCE – Eleventh Int. Conference on Composites or Nano Engineering, August 2004, Hilton-head Island, SC. Pacelli, M. et al. (2001), “Sensing threads and fabrics for monitoring body kinematic and vital signs”, Fibres Text Future Seminar, p. 55. Priniotakis, G. et al. (2004a), “Electrochemical impedance spectroscopy for quality control testing of textile structure electrodes”, Tekstil, Vol. 53 No. 11, p. 543. Priniotakis, G. et al. (2004b), “An electrochemical cell as simulator for the human body to study the behavior and properties of textile electrodes”, paper presented at 1st International Conference – Informatics and Quality, June, Athens. Rogale, D. and Dragcˇevic´, Z. (2001), “Intelligent clothing – a challenge for textile technology”, Tekstil, Vol. 50, pp. 107-21. Rogale, D. et al. (1999), “Intelligent garment manufacture and sales”, Tekstil, Vol. 48 No. 1, p. 17. Schedukat, N. et al. (2003), “Processing of high conductive yarns for signal transmission in smart textiles”, Proceedings Techtextil Symposium, p. 25. Van Langenhove, L. and Hertleer, C. (2003), “Smart Textiles, an overview”, Proceedings IIIth AUTEX Conference, p. 15. Van Langenhove, L. et al. (2003), “The use of textile electrodes in a hospital environment”, Proceedings IIIth AUTEX Conference, p. 286. Westbroek, P. et al. (2004), “New method for characterizing intelligent textiles by means of electrochemical impedance spectroscopy”, AUTEX 2004 Conference Proceedings, June, Roubaix. Westbroek, P. et al. (n.d.), “Method for quality control of textile electrodes used in intelligent textiles by means of electrochemical impedance spectroscopy”, Textile Research Journal (in press).
Simulation of human body behaviour 241
The Emerald Research Register for this journal is available at www.emeraldinsight.com/researchregister
The current issue and full text archive of this journal is available at www.emeraldinsight.com/0955-6222.htm
IJCST 17,3/4
Protective clothing – test methods and criteria of tear resistance assessment
242
Beata Witkowska The Institute of Textile Material Engineering, Lodz, Poland, and
Iwona Frydrych Technical University of Lodz, Lodz, Poland Abstract Purpose – The state-of-the-art of existing methods of tear resistance (static and dynamic) of clothing has been described, also presented are the parameters of static tear resistance for protective and work clothing depending on its application. Design/methodology/approach – For chosen group of fabrics the introduction of a new parameter of dynamic tear resistance was proposed. For research, five static tear test methods and the dynamic one were chosen. In order to find the relationship between the results of mean tear forces for the six described methods Kendall’s agreement coefficient was calculated. The comparative measurements for results of static tear resistance and dynamic tear resistance for protective and work clothing were carried out. On the basis of this, the value of tear dynamic force for these fabrics was established. Findings – When establishing the criteria for the tear strength for protective and work clothing, the most significant was fabric end-use and the minimal value of tear strength associated. Practical implications – The value of dynamic tear resistance can be the criterion for assessment of fabrics with regard to textiles exposed to tearing during application. It was the first comparative analysis of the measurement of tear resistance methods. Originality/value – Investigating test methods for the assessment of clear resistance. Keywords Textiles, Fabric testing, Protective clothing, Statics, Dynamics Paper type Research paper
International Journal of Clothing Science and Technology Vol. 17 No. 3/4, 2005 pp. 242-252 q Emerald Group Publishing Limited 0955-6222 DOI 10.1108/09556220510590948
1. Introduction The directive of EU No 89/686/EWG from 21 December 1989 deals with the unification of laws and normative acts determining general requirements of individual protective means. Analysing the requirements concerning the protective clothing shows, they depended on the purpose the clothing should fulfil. Many of them are determined by appropriate standard requirements. Owing to the health safety or human life saving, a lot of these demands include high mechanical tensile strength and abrasion resistance. Besides the individual protective means, the strength properties are the most important criteria of assessment for the fabrics intended for the special (high-tech) goods, technical or automobile textiles or textiles used as building interior materials. The paper presented describes the state-of-the-art of existing methods of tear resistance (static and dynamic) evaluation for clothing. The analysis of measurement methods and linear correlation relationships between results obtained by different measurement methods are presented. The parameters of static tear resistance for protective and work clothing, depending on its purpose are also presented. For the chosen group of fabrics, the introduction of a new parameter of dynamic tear resistance
is proposed, which can be a criterion of assessment of fabrics aimed for the textiles exposed to tearing during application, as a result of contact with sharp elements for example. The measurements were performed in the Institute of Textile Material Engineering in Lodz (Poland), in accredited laboratories, which in a complex way assessed the ability of using the fabrics examined for the protective clothing.
Protective clothing – test methods
2. Methodology Fabric tear resistance is the property determining the material strength under the action of static force (static tear test), or kinetic force (dynamic tear test). Different methods of tear test procedures are reflected in different standards. They are characterized by different ways of sample preparation, their shape and size, the way of clamping, the length of torn fabric distance, as well as the manner of reading and calculating the tear force (Z˙urek and Kopias, 1977). Five static tear test methods and a dynamic one are described below. All the methods (apart from the method No 1, dealing with the knitted fabrics as well, and the method No 5 is used for fabrics coated with rubber and polymers) were used for woven fabrics. In the course of the measurements, the method No 5 was taken into consideration as well. Although it concerns coated fabrics, it can be applied for testing the fireguard clothing. The methods are presented in the following standards: No 1 – PN-P-04640:1976 “Fabric measurement methods. Woven and knitted fabrics. Determination of tear strength”. No 2 – PN-EN ISO 13937-2:2002 “Textiles – Tear properties of fabrics – Part 2: Determination of tear force of trouser – shaped test specimens (single tear method)” (ISO 13937-2:2000). No 3 – PN-EN ISO 13937-3:2002 “Textiles – Tear properties of fabrics – Part 3: Determination of tear force of wing-shaped test specimens (single tear method)” (ISO 13937-3:2000). No 4 – PN – EN ISO 13937-4:2002 “Textiles. Tear properties of fabrics – Part 4: Determination of tear force of lounge – shaped test specimens (double tear test)” (ISO 13937-4:2000). No 5 – PN-P-04966:1993/A21:2002 “Rubber or plastics coated fabrics. Determination of tear force”. No 6 – PN-EN ISO 13937:2002 “Textiles. Tear properties of fabrics – Part 1. Determination of tear force using ballistic pendulum method (Elmendorf)” (ISO 13937-1:2000).
243
2.1 Static tear methods Static methods (No 1 4 5) differ among each other by the way of sample preparation and of clamping, tearing direction in relation to the acting force, distance between jaws etc. The static tear methods one described in Table I. The sample is torn at constant speed till the end of the measurement distance with all the methods applied. 2.2 Dynamic tear test The main element of the device for measuring dynamic tear strength is a ballistic pendulum, by means of which the force is suddenly applied to the appropriately
Table I. Description of static tear methods
5
4
3
1 2
PN-P-04640:1976 PN-EN ISO 13937-3:2002 PN-EN ISO 13937-3:2002 PN-EN ISO 13937-4:2002 PN-P-04966:1993/ AZ1:2002 Single
Double
Single
Single Single
Single or double tearing
75 ^ 1 75 ^ 1 145 ^ 1
k k
40 ^ 1 75 ^ 1
Tearing distance (mm)
’
’ k
Tearing direction: ’ or k to the acting force
100
100
100
100 100
Measurement rate (mm/min)
70
100
100
50 100
Distance between jaws (mm)
244
Method no Standard
IJCST 17,3/4
prepared sample. After it, the sample is mounted between two jaws, one – fixed and the other – moving, mounted onto the device body. The moving jaw is connected with the pendulum, which falls down due to the gravimetric force and the whole sample is torn by the displacement of the fixed jaw. Kinetic energy needed for the sample tear test along the initially cut distance is obtained as the result of the measurement. It is determined by the measurement of the work done during the sample tear test at the tearing distance. The advantage of the dynamic tear test is the possibility of obtaining the results of tear test (readings directly on the device) rather quickly. 3. Criteria of tear strength for protective and work clothing for the chosen groups Clothing used at the workplace can be divided into two groups: (1) Workwear, which protects or replaces one’s own clothing in the conditions not harmful to man. It is the clothing adjusted to the work (manipulations) and requirements of the technological process (definition according to PN-P-84525:1998 “Workwear. Work suits”). (2) Protective clothing, which covers or replaces the personal clothing, protecting against one or many dangers (definition according to PN-EN 340:1996 “Protective clothing. General requirements”), and its required at a particular workplace. The standards of the Polish Standardization Committee contain about 70 standards, dealing with the protective and work clothing. These standards contain very detailed requirement for fabrics to be used for the protective and work clothing, as well as the measurement methods concerning tear strength for protective and work clothing of selected job groups, and the designation of the chosen fabrics, qualifying for these groups, including the measurement methods. The requirements result from: the standards PN-EN 469:1998/Ap1:2003 “Protective clothing for fire-fighters. Requirements test methods for protective clothing for fire fighting”; PN-P-84625:1998 “Workwear. Work suits”; PN-P-82008:1990 “Cotton and cottonlike woven fabrics for work clothing”; the regulations between producers of protective and work clothing and clients: Headquarter of Fire Guard (Poland) – for protective clothing for groups of seekers, Military Research Centre of Uniformed Servants (Poland) – concerns the uniforms of this group. Requirements are presented in Table II. 4. The fabrics examined Cotton fabrics and their blends with viscose, PA and PES fibres, or other synthetic fibres are most often used in the manufacture of modern workwear. Fabric can be finished in a special way, for example, by coating. Typical blends used are: PES 65 percent/CO 30 percent; PES 50 percent/CO 50 percent; PES 35 percent/CO 65 percent. Pure cotton fabrics are also used for workwear (Polipowski, 2002). There are also some manufactured fabrics, with the threads of special properties like reinforcing, conductive, shining etc. These are introduced in one or both directions of the fabric. For the measurement No 7, the fabrics intended for protective clothing and 7 – for workwear, were chosen. The fabrics are sorted into groups presented in Table II. In the group No 1, protective clothing for fireguard (for action), there are four fabrics. They
Protective clothing – test methods 245
6
5
4
3
2
Work clothing (uniform for workers of meat mills), Fabric M Work clothing (depending on mass per square meter and end-use), Fabrics F, N
Protective clothing for fire fighting, Fabrics H, I, J, K Protective clothing (uniform of military pilot), Fabric L Protective and work clothing for fire fighting (for groups of life savers), Fabrics A, B, C, D, E Work clothing (for military barracks), Fabric G
1
Table II. Set of requirements concerning the tear strength of the protective and work clothing for the chosen job groups
Assortment
33 N 30 N
25 N
25 N
PN-P-04966/ AZ1:2002 PN-EN ISO 13937-2:2002 PN-P-04966/ AZ1:2002
PN-EN ISO 13937-2:2002
PN-EN ISO 13937-2:2002 PN-EN ISO 13937-2:2002
PN-EN 469:1998/Ap1:2003 Conditions established by Military Research Centre of Uniformed Servants Conditions established by the Headquarter of Fire Guard Conditions established by Military Research Centre of Uniformed Servants PN-P-82008:1990 PN-P-84525:1998
from 10 N to 30 N
25 N
Standard
Requirement
The value of parameter (not less than)
246
Lp
IJCST 17,3/4
are characterized by the high mechanical strength, durability, good thermal resistance and resistance to flame. They have barrier properties to such media as water, chemical agents, oils, etc. and maintain high physiological comfort. These properties qualify them for the outer layer of the protective clothing for fireguard in action. The group No 2 is represented by a fabric intended for the pilot protective clothing. It is characterized by a high strength, constant thermal resistance and resistance to flame, durability, as well as high physiological comfort. In the group No 3 there are five fabrics intended for the protective and work clothing for fireguards and for union suit for lifesavers. They are characterized by high strength and durability. The important feature is easy-care and no need for maintenance after washing. The group No 4 is represented by the fabric aimed for military barracks clothing. It is characterized by high strength, durability and physiological comfort. In the groups No 5 and 6 there are fabrics for general workwear. They have high strength, durability, easy-care and do not change after washing. Fabrics for uniforms for worker of meat mills have an aperture preventing the liquid transfer. It should be pointed out that the protective and work clothing for long usage, irrespective to the particular function, should fulfil common utility requirements, i.e. it should have determined durability, resistance and easy care properties, colour as well as chlorine resistance. The fabrics tested are characterized by mass per square meter: 190 to 425 g/m2; twill weave; warp density:280-589 threads per dm; weft density: 182-321 threads per dm; a raw material – the typical blends CO/PES, 100 percent CO, and for especially strength fabrics blends of meta-aramide with para-aramide and viscose fibres and minimum content of antistatic fibres. 5. Testing 5.1 Comparative measurements Full comparative measurements were done for the fabrics A-E (Table III), intended for protective and work clothing (Witkowska and Frydrych, 2003a, b; Witkowska and Frydrych, 2004). The Instron tensile tester was used for static measurements. For dynamic measurements, the Elmatear device of the firm J.H. Heal was used. The measurements were carried out in normal climate conditions, on the samples conditioned according to PN-EN 20139:1993. For the purpose of tear test, five static and one dynamic method were employed. For each direction (warp-weft), six specimens were used. As a result, the mean value or median of tear forces in [N], and values of random error at the significance test level a ¼ 0:05 were calculated. The results obtained for tear strength, depending on the method used, are presented in Table III. 5.1.1 Discussion. Analysing the values of the results obtained for tear strength employing five different static methods and one dynamic method, it should be noted that the results obtained for the mean tear force (or medium) for six methods are on different levels. Therefore, when starting the measurements, the appropriate measurement method should be chosen, taking into account the fabric end-use, the kind of tear performance that can occur in the conditions of end-use, as well as the parameters and criteria, which are in the form of normative documents for many fabrics. In order to establish the correlation between the results of the mean tear force (or median in the method No 5) for the six methods described, two Kendall’s agreement coefficients were calculated for five static tear methods and for the total of the six
Protective clothing – test methods 247
Table III. Tear strength results
21.8 32.5 41.9 31.4 32.5
a 41.0 76.9 77.0 49.5 65.7
A B C D E
A B C D E
a 0.5 0.5 3.8 0.8 1.8
U a 24.9 36.8 37.7 27.5 33.7
a
Weft
U a 1.7 5.5 1.0 0.7 2.0 a 44.0 74.0 68.4 45.8 73.5
PH-EN ISO 13937-4:2002 Method 4
Warp
U a 2.7 2.4 1.6 3.6 1.8
1.3 2.2 1.4 3.9 0.9
U a
Warp 0.6 3.5 4.1 1.2 0.8
U a
med. 27.5 68.8 49.8 40.0 47.0
med. 32.0 51.0 45.0 37.5 53.0
PN-P-04966:1993 Method 5
24.0 44.8 43.6 27.8 36.3
a
a 23.9 49.6 52.3 25.3 36.4
26.0 39.6 38.4 24.5 36.2
a 1.9 0.6 1.1 1.2 1.0
U a
18.0 28.3 37.5 28.4 27.6
U a 0.5 3.7 0.7 0.9 2.6
0.5 1.3 2.9 1.7 0.7
22.0 34.0 34.3 22.4 32.2
0.6 1.3 2.4 1.4 0.8
PN-EN ISO 13937-3:2002 Method 3 Static Warp Weft a U a a U a
PH-EN ISO 13937-1:2002 Method 6 Dynamic U a a 1.4 22.4 5.1 44.3 1.7 37.8 1.2 19.1 2.2 35.8
Weft
PN-EN ISO 13937-2:2002 Method 2
248
Fabric
PN-P-04640:1976 Method 1
IJCST 17,3/4
methods (static and dynamic taken together). The idea was to check whether the tear results obtained by different methods were coherent. The rank 1 indicates the lowest mean value (or median) of tear strength and the rank 10 the highest one (Udny Yule and Kendall, 1966; Volk, 1973). The following values of Kendall’s coefficients of agreement were obtained: for five static tear methods – W 1 ¼ 0:868; for six tear methods (static and dynamic) – W 2 ¼ 0:869: Taking into account that the value of agreement coefficient can be changed in the interval (0;1), and the values close to þ1 mean a high degree of agreement, it was clear that the obtained values confirmed quite a high degree of agreement of tear strength results. Furthermore, in order to confirm the relationship among particular tear methods, the linear correlation coefficients were calculated. The correlation coefficients can be seen in Table IV. The margin value of correlation coefficient at random degree n 2 2 ¼ 8 and the significance level a ¼ 0:05 above which the correlation exists was 0.632. Accordingly, there was no linear correlation between the methods 3 and 5 (Volk, 1973). The high values of correlation coefficients obtained, those above 0.8 (except in three cases), confirmed a strong linear correlation among the tear methods. Some rules should be noted: (1) The strongest linear correlation coefficient exist for the methods No 1 and No 3; No 2 and No 4; No 2 and No 6; No 4 and No 6. All of these methods have the same direction of tear force, in relation to the tearing direction. For the methods No 1 and No 3, the tear test is carried out in the direction perpendicular to the direction of the force applied, and for the methods No 2 and No 4 the tearing is carried out in the direction parallel to the force applied. Interesting situation exists in the methods No 2 and No 6, as well as in No 4 and No 6. The method No 6 concerns the dynamic tear test, so its procedure differs from the static tear tests described in the methods No 2 and No 4. Nevertheless, provided certain assumptions are made, similarities can be observed: . the way of sample mounting; the clamping line in both methods is parallel to the torn sample threads; Methods 1 1 1 1 1 2 2 2 2 3 3 3 4 4 5
and 2 and 3 and 4 and 5 and 6 and 3 and 4 and 5 and 6 and 4 and 5 and 6 and 5 and 6 and 6
Protective clothing – test methods 249
Correlation coefficient 0.839 0.984 0.861 0.637 0.814 0.818 0.972 0.883 0.979 0.861 0.626 0.788 0.884 0.937 0.843
Table IV. The correlation coefficients among the results of different tear methods
IJCST 17,3/4
250
.
acting tear force; in the static method a successive force increase causes the breaking of threads in the tearing direction; in the dynamic method there is a sudden force action on the similarly mounted threads and the torn threads are broken simultaneously.
Having all this in mind, it can be assumed that in the case of the methods No 2 and No 6, as well as for No 4 and No 6, tearing is carried out in the direction parallel to the direction of the force applied. It is confirmed by high values of correlation coefficients: 0.979 and 0.937. In the case of the pairs No 1 and No 3, No 2 and No 4, No 2 and No 6, No 4 and No 6, high value of linear correlation coefficient is observed for the data in both measurement directions (warp and weft), as well as for each direction separately. (2) The lowest linear correlation exists between the methods No 1 and No 5 as well as between No 3 and No 5. In these methods, the directions of force application differ. In the methods No 1 and No 3, it is perpendicular and in the method No 5 parallel to the force applied.
5.2 Establishing the value of tear dynamic force for the fabrics intended for the protective and work clothing For all the fabrics investigated A-N (Table III) the tear strength was measured using the method 2 (PN-EN ISO 13937:2), the method 5 (PN-P-04966) and the method 6 (PN-EN ISO 13937:1). The method 2 was chosen as it is widely applied in the laboratory tear measurements. The measurements were also done using the method 5, as it is also used for tear strength analysis in the case of the fabrics intended for the protective and work clothing for fireguards, while the results obtained from it are related to the criteria of tear strength. The comparison of the results obtained by the method 6 (dynamic tear test) and those obtained by the methods 2 and 5 showed: Step 1 – The method 2 was taken as a base, and the results obtained by it were taken as 100 percent, whereas the results obtained from the method 6 were presented as a percentage. Step 2 – The method 5 was taken as a base, and the results obtained by it were taken as 100 percent, whereas the results obtained from the method 6 were presented as a percentage. Six specimens were used for each direction (warp-weft). The mean value or median was calculated for the tear forces in [N], while the values of random error at the significance test level a ¼ 0:05: The results obtained for tear strength depended on the method used, and are presented in Table V. 5.2.1 Discussion. The following resulted from the analysis of the shear values obtained employing the method 6 (dynamic tear test): . For the method 2 – for 28 pairs of results (14 fabrics £ 2 directions) 16 pairs yielded results higher than by the method No 2 (100 4 148) percent, and 12 pairs of results lower than by the method No 2 (78 4 99) percent. As an average, they were on the level 106 percent of the base result.
Fabric A B C D E F G H I J K L M N Mean
PN-EN ISO 13937-2 PN-P-04966 Method 2 Method 5 Tear strength [N] Warp Weft Warp Weft a U a Median a U a 24.0 44.8 43.6 27.8 36.3 46.9 28.8 82.9 78.8 58.6 62.1 61.0 33.0 34.1
0.6 3.5 4.1 1.2 0.8 2.0 1.4 6.3 4.5 2.7 4.6 9.4 1.0 2.1
26.0 39.6 38.4 24.5 36.2 53.3 36.0 76.5 70.6 60.0 62.3 61.5 25.0 32.1
1.9 0.6 1.1 1.2 1.0 4.1 1.1 2.3 3.4 1.2 2.6 3.4 0.8 2.1
PN-EN ISO 13937-1 Method 6 Tear strength [N] A B C D E F G H I J K L M N Mean
.
23.9 49.6 52.3 25.3 36.4 45.2 41.6 73.2 74.3 65.6 67.1 62.5 37.9 50.6
1.4 5.1 1.7 1.2 2.2 2.0 5.0 1.9 2.2 3.0 2.0 2.1 4.0 4.0
22.4 44.3 37.8 19.1 35.8 53.8 46.2 66.4 66.9 64.5 65.7 63.0 24.2 47.3
0.5 3.7 0.7 0.9 2.6 3.0 3.0 3.0 2.0 3.5 2.0 2.0 2.0 3.0
27.5 69.0 50.0 40.0 47.0 54.0 36.0 123 98.0 68.0 75.0 85.0 40.0 45.0
32.0 51.0 45.0 37.5 53.0 58.0 46.0 76.0 78.0 68.0 76.0 66.0 33.0 42.0
Method 5 100 percent Method 2 Warp
Protective clothing – test methods
Weft
99 min 86 111 112 120 98 91 78 100 99 96 101 144 128 88 87 94 95 112 108 108 105 102 102 115 97 max 148 147 106 percent
Method 6 100 percent Method 2 87 70 72 87 105 84 63 min 51 77 68 84 93 115 100 60 87 76 87 96 95 89 86 74 95 95 73 max 121 112 86 percent
For the method No 5 – for 28 pairs of results five pairs yielded results higher than in the method No 2 (100 4 121) percent, and 13 lower than by the method 5 (51 4 96) percent. The results were at 86 percent of the base result.
Significant deviation of results in the method No 6, as compared to the methods 2 and 5 could be caused by different raw material content, different yarn linear density and different fabric density. These parameters influence significantly the results of tear strength. Nevertheless, when establishing the criteria for the tear strength for protective and work clothing, the most significant was fabric end-use and the minimal value of tear strength associated.
251
Table V. Set of tear strength results and share values obtained dependable on the method, for the fabrics intended for the protective and work clothing
IJCST 17,3/4
252
6. Conclusions The results obtained indicate the following: (1) The obtained values of the rank coefficients show a coherency among all the static tear test methods ðW 1 ¼ 0:868Þ; as well as between the static methods and the dynamic one ðW 2 ¼ 0:869Þ: (2) The high value of linear correlation for the methods using the same direction of force applied as related to the tearing direction ðr ¼ 0:937 4 0:984Þ; are observed. (3) There are certain similarities among the static tear methods applied when the direction of the force applied is parallel to the tearing direction, and the dynamic tear method, which is confirmed by the high correlation coefficient between these methods (r ¼ 0:937 and r ¼ 0:979). (4) Taking into account the percentage of the results obtained by the method No 6 (dynamic tear test) in relation to the results obtained by the methods No 2 and No 5 (static tear tests), the following values of dynamic tear strength parameter are proposed: . For the fabrics intended for the protective and work clothing (the criterion of static tear resistance determined according to PN-EN ISO 13937 Part 2), – dynamic tear resistance – not less than 27 N (static tear resistance – not less than 25 N), – dynamic tear resistance – not less than 324 35 N (static tear resistance – not less than 30 4 33 N), .
For the fabrics intended for the protective and work clothing (the criteria of static tear resistance determined according to PN-P-04966:1933/AZ1:2002) – dynamic tear test – not less than 22 N (static tear resistance – not less than 25 N).
References Polipowski, M. (2002), “Work clothing for a long usage”, Atest 9, pp. 19-21. Udny Yule, G. and Kendall, M.G. (1966), Introduction to the Statistics Theory, PWN, Warsaw. Volk, W. (1973), Applied Statistics for Engineers, WNT, Warsaw. Witkowska, B. and Frydrych, I. (2003a), “Clothing fabrics – static tear strength methods”, Przegle˛d Wło´kienniczy, No. 6. Witkowska, B. and Frydrych, I. (2003b), “Dynamic tear strength method and its correlation with the static methods”, Przegle˛d Wło´kienniczy, No. 8. Witkowska, B. and Frydrych, I. (2004), “A comparative analysis of tear strength methods”, Fabrics and Textiles in Eastern Europe, Vol. 12 No. 2(46), pp. 42-7. Z˙urek, W. and Kopias, K. (1977), Structure of Flat Textile Products (in Polish), WNT, Warsaw.
The Emerald Research Register for this journal is available at www.emeraldinsight.com/researchregister
The current issue and full text archive of this journal is available at www.emeraldinsight.com/0955-6222.htm
Uniformity and differentiation in fashion
Uniformity and differentiation in fashion
Ann Priest London College of Fashion, University of the Arts London, London, UK
253
Abstract Purpose – The paper offers a general outline of four broad consumer categories or groups, in part distilled from the range of detailed and colourfully named descriptors used to differentiate fashion and clothing consumer groups, identify and recognize trends. The paper will offer the opportunity to look at the long-term impact of external forces on fashion and clothing purchasing decisions. Design/methodology/approach – The method of research was diverse, but largely drawn from observation; media analysis and industry intelligence. In the course of the work, it was possible to draw on a varied range of sources to categorize purchasing decisions (illustrating consumer categories) into four main drivers. Findings – Thus to highlight some of the major forces that might drive the consumer. Practical implications – In reducing the detailed forecasts usually prepared for fashion and related products, the paper might be of interest to those considering the long term impacts of society, culture and politics on the purchasing decisions of fashion and clothing customers. Originality/value – An insight into the medium term future of fashion taking into consideration the consumer. Keywords United Kingdom, Fashion, Consumers, Society, Markets, Forecasting Paper type Research paper
1. Introduction Amongst the functions of fashion is to create uniformity amongst equals whilst at the same time differentiating status and background, signposting preferences and commitments. Reflecting the resulting market complexity, fashion forecasters have developed a range of detailed and colourfully named descriptors to differentiate consumer groups, identify, and recognise trends. In response, clothing retailers have grown and proliferated, chasing the increasing £’s spent on the sector as consumerism has raged across the UK and private consumption has become one of the prime movers in the British economy. With large profits at stake, achieving the right mix of product and price has been crucial and retailers have relied on the forecasters and interpretation of the same group of indicators to give them detailed direction. However, in consulting the same forecasters, watching the same international shows and receiving the same multiple media messages, it is no surprise that much of what has been on offer is similar in style, and available locally, regionally, and internationally. To describe and illustrate the UK clothing market in 2004, I have reduced the considerable range of ideas offered by forecasters to four broad consumer categories or groups. I will set out briefly some of the issues and contexts that are behind the formation of these groups. My premise and conclusion is that the over-arching factors that drive fashion offer longer-term indicators for fashion businesses; that fashion remains a matter of recognition and belonging, and the events and circumstances that affect our taste or purchase decisions are global and regional. The fashion market in
International Journal of Clothing Science and Technology Vol. 17 No. 3/4, 2005 pp. 253-263 q Emerald Group Publishing Limited 0955-6222 DOI 10.1108/09556220510590957
IJCST 17,3/4
254
the UK caters for sophisticated, well travelled and media hungry consumers – it will be no surprise then that most trends are to be found worldwide, and will be recognisable in most consumer societies. The four categories overlap, influence and relate to one another, and this is illustrated by the visual images that accompany the paper. Of the four, the final category is where much influence lies, highlighting that despite largely global factors: the value pre-occupation, domination in manufacturing, international news sharing, interdependent politics, rapid communication and shared multi media, acceptance is determined locally. The rising importance of the individual in society suggests regionalism and differentiation are not defunct. The challenge for clothing and fashion industries will be recognise broad trends whilst delivering to such customised and niche micro-markets. 2. UK clothing and fashion On the face of it, the UK clothing and fashion market remains attractive because of its size, some £32 billion in 2002. According to the UK Office of National Statistics; the five-year growth rate to 2000 was 18.5 per cent. In the year ending July 2001, sales of clothing and footwear rose by 12 per cent. For comparison, the European clothing market was worth e 272 billion in 2000 with the UK amongst the leading markets. Since 1999 however, growth has been fastest in the smaller, less mature markets such as Poland, Ireland, Hungary, Greece and Portugal, with some of the historically leader markets enjoying slowest growth. In the UK in 2002, women’s wear only grew by 4.1 per cent, in 2003 women’s outerwear grew by 2.5 per cent and in 2004 expectations are of just 2 per cent growth (Drapers, 2003; Retail Week, 2003). As elsewhere in Europe, specialist fashion retailers dominate the UK clothing market, although M & S is still the number one retailer and maintains a strong lead over its nearest women’s wear competitors. The star performer 01-03 continues to be next, with other specialists, New Look, H & M and Oasis gaining market share. Department stores are increasingly strong in the middle market, design led arena, due to clever retail styling. The value sector has developed and “democratised” and major grocery multiples Asda and Tesco have increased their market share considerably. Last year, internet shopping was failing to live up to the hype, now there is evidence that women are gaining in confidence in shopping on line, helped by their experience with known fashion retailers such as next. As anticipated, the availability of trusted brands through the internet has accelerated acceptance. Published research tells us that the increase in the value of the clothing market is offset against the considerable value for money offered by clothing over time. Consumers expect low prices and expect the downward trend for costs to continue. In guiding us to understand the extremely unpredictable and highly concentrated UK market, Jones and Hayes suggest that increasing incomes, particularly amongst women, are the key to increased economic activity and purchase determinants in the clothing market. With the result, that clothing and fashion purchase decisions are based on want rather than need (Jones and Hayes, 2002). The point is that clothing in the UK has changed from being a necessity to a luxury, i.e. “Luxury as standard”. The purchasing decision is therefore influenced by aspiration, the sense of what is deserved and how the customer wants to be perceived.
Downward pressure on prices continues to be significant. The value of the US dollar has had a tremendous impact for those sourcing in the Far East and has funded even lower prices and protected margins. Between the years 1996 and 2000, clothing prices in the UK fell in four out of five years, and the continued shifting of production to China is anticipated to fuel an intense price war at “value” level. Competition is fierce and the market has been characterised by excessive discounting. As an indicator of price development, the working time needed to purchase a selection of garments fell by two to three times between 1974 and 1988. A study of the metal working sector found that between 1982 and 1995, the time a motor vehicle worker had to work to buy a suit fell from 25 to 18 h (IMF, 1998). Anecdotally, there are interesting questions being raised in regard to the speed and totality of the shift of production to China: fashion oriented retailers have pressure for speed over margin and there is increasing concern in regard to ethical and other localised sourcing matters. This could be the harbinger of a new scenario for clothing and textile manufacturing in Europe. Another portent of the requirement for caution, in May 2002 report, clothing spend was reported as having “under performed” against spending on all consumer goods. It recognised that the consumer had broadened their aspiration and demand for “fashion, the new and exciting” to technologies, telecoms, and household items. Growth in the value market highlights the determination of consumers not to pay over the odds. At every level, there is evidence of bargain seeking and questioning “why pay more”? Bargain hunting has been democratised. The increase in retirees is significant, as are attitudes for example, toward spending on children’s clothing, once again underlining the shift from clothing as a necessity to a luxury. Most worryingly, the significant and influential 20-24 years old group are revealing some more negative attitudes to buying fashion as political and social attitudes, as well as student debt and inflation in the housing sector have kicked in. Overall, 2004 has brought signs of a slow-down in consumerism, and this reversal of the trend for increased spending is the cause of some medium term UK concern for the future. In Europe, the fastest growing markets are reported to be Ireland, Hungary, Greece, the Czech Republic and Switzerland. 3. The four categories 3.1 McFashion What has been termed “McFashion” has become disposable, quick fix international fashion, a mix of the bland and the predictable, and trendy, throwaway chic (Lee, 2003). 3.2 International superbrands The luxury brand continues to be a phenomenon; Fashion Designers or Designer Super-Brands with global recognition and multimillion dollar turnovers, supported by fragrance, cosmetics and widely acknowledged to have interdependency with the International Fashion Media. 3.3 UK (London) style Perhaps for a long time the herald of regional fashion is the acknowledged style and wit of UK, particularly London based, fashion; from time to time, heralded through such hype as “cool Britannia” and universally recognised as underpinning the
Uniformity and differentiation in fashion 255
IJCST 17,3/4
256
International fashion houses and super brands. Despite the link to international brands, at its basis are also micro retail outlets supporting students, localised traders, aspirational super branders or recent graduates. This is the feeding ground of the international labels and “suppliers to the UK high street”. Cult status and sales are fuelled through one off magazines and pamphlets emerging through the young “cognoscenti”, stylists, journalists and the music world. 3.4 Micro markets The final category is the most interesting and important in terms of the future market for clothing and fashion. There is significant and growing evidence for the emergence of the individual as a niche market. Recognition of this group is heralded by developing interest in spirituality, anti-consumerist movements, books, pamphlets and news media, proclaiming a new age from politics to work practices. This is not “grunge based” anti-fashion, and has no single or recognisable style trait. It is a societal cultural swell and its recognition requires that future “fashion” will take into account changing life stages, cultural cross over, political and ethical motivations and individualism as an aspiration. However, whilst identifying this category through these “marketing” terms, an empathy with creativity, the product and its values are also what is emerging here. The four categories or trends I have identified are of course overlapping and not exclusive. There are sub categories and cross group style trends. Recognising trends and forecasting demand have never been easy. The 1980s and 1990s were times when the fashion forecasting world was turned upside down in recognition of multiple styling levels, market niches, consumer types and the take over of fashion by marketing. There is evidence that what is now required could more accurately be called “intelligence”, “Style Vision” call it “mood consumption”. Many column inches in newspapers are dedicated to forecasting growth or decline in consumer spending. Merchants, economists and governments are all exercised by the guessing game and have whole armies of advisors adding to the focus group advice and feedback. Magazines such as Viewpoint have emerged as signposts to new movements and demographics simply by gathering the “intelligence” and analysing its meaning (Viewpoint, 2004). In 2003, internet based Style-Vision (www.style-vision.com), in a circular advertising a round table event, offered the following theme (Style-vision, 2004): “. . .your customer just evolved! Did you notice?” “A challenging landscape: . . .the luxury market is trading down, the middle segments are trading up, mobile phone companies are hiring fashion designers, fragrance companies are looking at furniture fairs, the automotive industry is becoming trendy, seniors are acting like teenagers, men are shopping like women, consumers are becoming more demanding. . . as such, the traditional borders between consumer segments are disappearing . . . ”
The UK market has many drivers; it is sensitive and remains as multi level and eccentric as has been increasingly the case since around 1975, recognised as the emergence of the modern market. Between 1975 and 1990, the market grew from 40 per cent of its 1999 size to 70 per cent (Jones and Hayes, 2002). Even with the current medium term jitters, the UK clothing market is a very attractive prospect. 3.5 McFashion “McFashion” is the term Michelle Lee coined in her 2003 book to describe the first trend, enthusiastically reported through UK news media. This provocative work argues that our high streets are full of cheap chic, the clothing version of fast food. She suggests that just as McDonalds has spread all over the globe offering uniformity and predictability, so mass market clothing retailers are succeeding in offering predictable, safe, recognisable styling. It can be bland and stylistically down market, or simply a fast food version of a “star trend”, but (it) enables “buy in” to social groups and fashion trends. These are products working to formulas. Those styles not “inspired” directly by what is on the catwalk, or worn by celebrities, are basic products churned out for the masses with a “style detail” or “twist” to ensure the consumer buys in safely to the latest trend. Not only does Lee argue that this is a tendency to uniformity and consistency, it is also throwaway, cheap and reliant on our greed and anxiety to be recognisably “cool”. Less provocatively, it highlights the concern with value for money for everyday purchases that helped to polarise the UK clothing market in the late 1990’s. If we take Lee’s argument and look for evidence, in the UK, there is our propensity to accumulate “fashion” items then destined for the bin or charity shops. In themselves, a major phenomenon. Perhaps giving a link to the fourth category and recognising that whilst we will buy into the McFashion world, we acknowledged the need for re-cycling as a sop for our greed and subsequent guilt. Big high street players nominated in the category of McFashion include GAP; H & M; ZARA; Marks and Spencer; Arcadia Group – and may be also Ralph Lauren and other “classic” designerwear. Also included: UK supermarkets; Asda, Tesco, Sainsbury, all of whom have expanded their product offer to include budget or value for money clothing as the proportion of incomes spent on food items continues to fall. They offer convenience shopping and a “buy in” for consumers seeking to buy in to fashion trend at a throw away price. Taking the supermarket clothing idea further, it is a logical extension of this business to expect internet purchases of basic items to be added to the grocery order and delivered to the door. Asda’s clothing business was last year quoted as generating sales of approximately £1 billion annually (Drapers, 2003; Retail Week, 2003). The potential of the sector is illustrated by changes in the denim market. Four years ago total consumer expenditure on branded adult jeans (Levi, Lee Wrangler, Pepe) in the UK was 51 per cent. By January 2003, it was 31 per cent. Retailers own label denim has reversed the position and is now accounting for 49 per cent from 31 per cent (Khabi, 2003). This extraordinary development is attributed to “thinking like brands rather than retailers”. That is to design, shape, fabric innovation and quality; add to that speed to market and price. Asda’ a basic five-pocket jean is retailing at £6, and even Moto’s fashion jean is £35. Thus a range of new markets for denim have opened up, leaving the traditional brands scrambling to develop value and high fashion lines, whilst the new denim producers look where next, rather than compete in a fragmented, competitive denim market.
Uniformity and differentiation in fashion 257
IJCST 17,3/4
258
McFashion may be characterised on the one hand by the bland and safe styling, but it extends to the fad and trend, quick throwaway fashion that relies on unerring recognition of the up coming trend, speed to market, a reliable supply chain, and in proven fashion, value for money. This is a competitive and challenging sector. It is a trend replicated in clothing and fashion markets globally, and increasingly, in the UK, McFashion is delivered by non-UK based companies such as GAP, Zara, Matalan and Uniglo. 3.6 International superbrands At the opposite end of the polarised UK market, are International Super-Brands. Whilst this diverse sector is a study in itself, for the purposes of this paper, it is useful to merge international brands or designer label merchandise into a single tendency. Led by the notion and reputation of “couture”, and as much an intangible as a tangible product, the message of these internationally branded fashion products is that the label on what you buy is critical. This is an international message led by the notion of super-luxury, rarity and quality, demonstrated (anecdotally) by the significant take up of products in this category by former eastern bloc countries as funds began to flow in to the economy. Influenced by class, money, film stars, music, Hollywood, Bollywood, sports and almost any glamour item in the media, this is big business and the stakes are high. International superbrands, the couture and ready to wear shows, and the media circus around them, are the drivers behind the aspirational element of fashion. Most of the International houses report couture customer numbers of barely up to 200 worldwide, and the majority of sales are acknowledged to be wedding dresses adapted from catwalk looks. The couture shows are acknowledged to be a fantastically effective marketing and selling tool, high impact advertising to maintain the excitement around the products that sell: the fragrances, bags, make up and the pared down ready to wear. In the UK, the total spend on men’s and women’s designer wear in 2002 was estimated to be worth £1.4 billion, a rise of 19 per cent compared to 1997 (Mintel, 2002). Once inflation is taken into consideration, the value of designer wear actually rose by 40 per cent over the same period. (Mintel, 2002) Since designer wear products are priced at a premium. Consumers are more likely to buy such items when levels of discretionary income are high. A rise in the number of working women – from £12.04 mn in 1997 to £12.74 mn in 2001, a rise of almost 6 per cent – has contributed significantly to the growth in this sector. Women’s designer wear has consistently accounted for the largest proportion of this market, accounting for 57 per cent in 2002, however the men’s designer wear market is increasingly influenced by the media and role models such as David Beckham, and therefore has the greatest potential for growth. In 2001 mens’ designer wear expenditure was recorded as 43 per cent of the market. This compares with the 1997 figure of 39 per cent Mintel forecast growth to 2006 of £747 mn or 23 per cent at current prices. Sufficient is the worldwide importance of the luxury goods sub-sector, and its determination to retain its market share globally, that two major conferences have been held in Paris, hosted by the International Herald Tribune, in order to discuss the issues of fashion and luxury brands. London College of Fashion’s Tim Jackson, prepared a review of the 2001 conference that was published in the Journal of Fashion Marketing and Management (Jackson, 2002). He identified the following notable quotes from speakers and published papers:
Tom Ford: Gucci; YSL – remarked: “globalisation is inevitable . . . ” and: “. . .the world has been united stylistically as it never was in the past. The entire world is watching the same films, listening to the same music and eating the same foods all at the same time. Our cultures are blending. Soon we will be one global culture”. Jean Marc Simon from Comite Colbert and Daniel Tribouillard, President of the Leonard Group, agreed that the market for such luxury goods was Europe, Asia and the USA; however, M. Tribouillard made clear that he regarded the future for luxury goods as China, Korea and Taiwan. “The growth of luxury brands . . .has been fuelled by the accelerated creation of new wealth. . .and its consumerist, status-motivated nature”. Mintel attribute a proportion of UK designer wear market growth to “a market partly fuelled by traditional mid-market consumers trading up to products sold at a premium price’. The purpose of these conferences is to share ideas and promote strategies to ensure continuing growth in the sector. The concern being that in attracting an aspirational customer with low loyalty and perhaps superficial perception of the brand essence, the brand in itself would be damaged. Returning to growth in the market, the conference’s concern (and Mintel’s forecast for the UK) might be borne out by research in the USA. In an article in the Harvard Business Review, Michael J Silverstein and Neil Fiske identified a new luxury goods market for the trading up middle market (Silverstein and Fiske, 2003). They describe a new customer (in the USA), a middle market customer prepared to pay a premium for well designed, well engineered and well crafted goods. They make the point that alongside the “old” luxury goods markets there are profits to be made by creators of “new luxury goods” by those who understand the aspiration of this customer and the factors that have shaped them. They point to “higher levels of taste, education and worldliness” attributed to being “better educated, more sophisticated, better travelled, more adventurous and more discerning than ever before”. Returning to the “old luxury brands” Tim Jackson’s parting shot in his review gives the clue for the Luxury Goods dilemma: “In a world of increasing homogeneity a big question for fashion business, luxury or otherwise, is how to reflect significant trends while being differentiated from competitors. For a luxury brand there is the additional problem of how to balance global availability on a sufficient scale to be profitable, with the requirement to be exclusive. The emergence of the multi brand, luxury group, such as LVMH or the Gucci Group NV, is one method used to increase group sales whilst maintaining individual brand exclusivity. Star designers, powerful imagery and “must have” limited editions have their place, but a key message of this conference was that a luxury brand must have integrity”. 3.7 UK (London) style The extraordinarily rich creative mix that has been recognised within the UK is characterised by the London fashion scene. Since the 1960s, there has been “Swinging London”, “Cool Brittania” and many other well worn (albeit sometimes slightly embarrassing) advertising and marketing slogans. Internationally, it is acknowledged that British design graduates have unique creative abilities. The UK educates around 3,000 graduates in fashion every year. Not all are fashion designers; they are design technologists, product developers, fashion buyers, stylists, merchandisers, pattern cutters, fashion journalists, “P.R.’s” and find work across the breadth of Fashion within the UK, and Internationally to great acclaim.
Uniformity and differentiation in fashion 259
IJCST 17,3/4
260
However, in the spirit of the new and creative, many start their own fashion business, or work on a freelance basis. The sheer numbers that embark on this path contribute to a unique and ever flourishing market that is both created by and kept lively by, young creative energy. A proportion goes on to build long term, established Designer businesses, and in a sense they then move to our previous group and become international. When this happens, they are replaced over again, with new young businesses at the cutting edge of fashion. This sector of the industry is hugely important to the UK economy, and scrutiny of published figures show considerable growth. In 1991, a survey of the UK fashion designer Industry was undertaken by Kurt Salmon Associates, which concluded that in 1990 the worldwide turnover of UK based “designers” was £75 mn. Using the same methods, Malcolm Newbury, working for the UK DTI in 2002, estimated the figure to be £700 mn (Malcolm Newbury Consulting, 2003). Labels such as Katharine Hamnett; John Rocha; Jasper Conran; Betty Jackson; Paul Smith, are examples of “Brit Fashion” gone International, although they are not in the Superbrand league. Some have gone into partnership with established retailers and multiples in design led partnerships that fund their own labels and presumably keep the wolf from the door. John Galliano; Stella McCartney; Alexander McQueen; Julian McDonald moved to Paris and took over creative direction of the houses of Givency, Dior and Chloe and have become the driving force behind the labels. Several have now established their own labels under the corporate wings of the Luxury Goods “stables” such as Gucci. Hussein Chalyan; Tristan Weber; Markus Lupfer, Rafael Lopez, Emma Cook are tipped to be exciting, if volatile, labels to watch. This category, however, to be truly a regional example, is characterised by those who have started small, or remained small, often working with teams of friends, tiny retailers and to be found in the “creative” or “trendy” areas of cities across the UK. Difficult to quantify and distinguished by its uncertainty and bravery, this is a sector that rejects the commoditisation of fashion. It has grown since the late 1970’s and now is established as a feeder for the international fashion world. The sector drives and is driven by unique creative communities of fashion designers, journalists, stylists, graphic designers and photographers that are responsible for the new fashion media. Titles such as Tank and Purple, and lately Marmalade have followed Dazed ad Confused and I-D as style bibles for the cognoscenti, the crowd that drive UK fashion worlds and ensure that fashion is pervasive and cutting edge. The energy, bravery, idealised concepts and rawness of this group have developed considerable recognition in the past forty years. Crucially, through the global media, there is recognition that this long recognised UK fashion strength is being chased in Australia, Hong Kong, India and other developing and design aware markets as their indigenous creative strengths are recognised and applauded. 3.8 Micro markets Truly the most directional of the tendencies, the most complex and the most intriguing in terms of how we view the future of clothing and fashion. This grouping has been driven by the past four decades of increasingly rapid developments in culture and society. It is epitomised by the fact that the children of the sixties are now over 50, and the trendy young things of the decade can be 15 or over 60. They are the hedonistic, self
focused baby boomers, and they have forsaken the trappings of retirement for clubbing, hiking and pleasure seeking – and incidentally, they want to be sold to in a way that reflects young markets. . .This is a hyper design and culture conscious culture. This is not however, “just” the group known as the “Rainbow Youth”, although they are a major factor (Viewpoint). This is a multi-dimensional, multi style group that ranges from those who demonstrate a carefree, yet studied casualness in regard to clothing and style, to those who collect international designer fashion and mix it with Top Shop chic, or antique textiles, and might have a recognisable personal style or mix and change with the weather. Many are reacting against brands and branding, whilst nonetheless being influenced by brand values and the day to day availability of information, advice and outrage. It is a growing tendency to please oneself in relation to what one knows and interprets according to experience, aspiration, position in society or conscience; a preference to be one’s own personal stylist because life, education and society have encouraged UK consumers to have personal views. International magazines have been criticised by the media for manipulating editorial and advertising, the one dependent on the other. The result, that the fashion pages are now still full of designer clothing, but styled and camouflaged to create a specific look or describe a narrative or story simply fuels this fourth category and its individual nature. Influences include Naomi Klein’s cult book “No Logo”, with its assertion that the most successful brands do not make anything (Klein, 2000). “Companies are switching from producing products to marketing aspirations, images and lifestyles. They are shedding physical assets by shifting production from their own factories in the first world to other people’s in the third” (Author unattributed, Economist. Com, 2001). Whilst rejecting outward and obvious brands, they are however influenced by values – will the recent political events affect direction? Most certainly, and not so slowly. This category will accept brands, but on their own terms. This sector is knowledgeable and opinionated. They have become accustomed to the bombardment of advertising and propaganda. They share the aspiration to achieve “value” with the first trend. However, the perception of value is not limited to cost, but extends to other areas of personal gratification, such as style recognition, ethics and crucially, individuality. The issues and preferences that help us recognise this tendency have been gathering for some time: Firstly, there are the demographic changes that place 15-24 year olds and more importantly, 55-64 year olds as sectors of greatest growth (þ 6 per cent). These are both major players in the new influence society. The spending capacity of childless, but informed consumers, including those referred to as the pink pound or gay society. In the political and macro sense, there is growing interest, although not so recently evident participation in the growing anti-globalisation movements, with their anti-capitalist stance and increasingly action – oriented proposals. The under performing clothing market, the falling spend on clothing as a proportion of income, which suggests growing interest in broader aspects of society and (other products) or aspects of life. The increase in interest in health, well being and spiritual aspects of life; knowledge and recognition of craftsmanship and “added value”, as a fall out from international travel and education as well as technological developments that enable body fit; technical fabrics that support the need for special performance.
Uniformity and differentiation in fashion 261
IJCST 17,3/4
262
The individual, as a phenomenon, “non-places” suggests, is making a comeback (Auge, 1995). Customised and made to measure clothing and shoes are available through the internet (www.landsend.com), designer wear (Gucci; Brooks Brothers), and is the subject of research and commercial development at UK national level. The fashion media reference the phenomenon by highlighting opportunities for customisation, personalisation and personal stylists, to unique products (Edelkoort 02). Style Vision’s latest report is Global Consumer Moods: Life attitudes to 2006 (Vol 2) (Anon, n.d.). They report “Consumption has entered a third phase. Mood consumption is rapidly replacing Mass Consumption and Lifestyle Consumption (Global Consumer Moods). Industries will have to re-think their branding strategies. Industries will have to re-think their product creation process and market research. In short: Industries will have to re-think their approach to consumption and consumers”. Added to this is the “non-places” suggestion that there is now the opportunity for the anthropology of individuals rather than groups. It seems that we may look forward to some long term changes to our clothing markets. 4. Conclusion The UK market for clothing and fashion continues to be excited and exciting, although it is clear that real growth may lie elsewhere in Europe. The mass market and recognition drive uniformity, but differentiation and individualism is growing. Changes in the spending patterns of (western) consumers may have an increasing impact on clothing and fashion if the final, influential, category lead the way to changed spending patterns for the future. That McFashion is global and continues to grow as the consumer society influences the developing world. Teenagers are as likely to be wearing McFashion as are 40, 50 and 60 some times. That International Super brands will protect their exclusivity whilst using aspiration and consumerism to attract a broader, multi level consumer market. That the UK regional attitude to ideas, eccentricity and creativity is the driving creative force for fashion within the UK, and to UK (London) style, and its influence on the International fashion industry. This category is increasingly replicated elsewhere and may be the major regional growth driver for fashion. The fourth category is a major driver for change. It size and influence is far reaching and is driven by the values and aspirations of an informed, opinionated group of individuals. The character and preferences of this group may hold the key to not just clothing and fashion within the UK, but to global consumption patterns. References Anon (n.d.), Global Consumer Moods: Life attitudes to 2006, just-style.com 2, AROG Ltd Auge, M. (1995), Non-places. Introduction to an Anthropoliogy of Supermodernity, Verso, London. Drapers (2003), 10 May. IMF (1998), The Purchasing Power of Working Time, International Metalworkers Federation, Geneva. Jackson, T. (2002), Journal of Fashion Marketing and Management, Vol. 6 No. 4, pp. 408-16. Jones, R.M. and Hayes, S. (2002), “The economic determinants of clothing consumption in the UK 1987-2000”, Journal of Fashion Marketing and Management, Vol. 6 No. 4. Khabi, M. (2003), “Denim dual”, Drapers, 5 April, p. 25.
Klein, N. (2000), No Logo, Flamingo, London. Lee, M.M. (2003), Fashion Victim: Our Love-Hate Relationship with Dressing, Shopping, and the Cost of Style, Lee Broadway Books, New York, NY. Malcolm Newbury Consulting (2003), Study of the UK Designer Fashion Sector, (DTI: BFC), UK. Mintel (2002), Market Size and Trends. Designer Wear – UK – July 2002, Mintel International Group, UK. Retail Week (2003), 13 June. Silverstein, M.J. and Fiske, N. (2003), “Luxury for the masses”, Harvard Business Review, April. Further reading Viewpoint (n.d.), Metropolitan Publishing BV, Amsterdam. Viewpoint (n.d.), Rainbow Youth, 13th ed., Metropolitan Publishing BV, Amsterdam.
Uniformity and differentiation in fashion 263
The Emerald Research Register for this journal is available at www.emeraldinsight.com/researchregister
The current issue and full text archive of this journal is available at www.emeraldinsight.com/0955-6222.htm
IJCST 17,3/4
The analysis of the seam strength characteristics of the PES-PTFE air-jet-textured sewing threads
264
Vaida Jonaitiene˙ and Sigitas Stanys Faculty of Design and Technologies, Department of Textile Technology, Kaunas University of Technology, Kaunas, Lithuania Abstract Purpose – The goal of the research presented is to analyse seam strength properties of polyester and polyester-polytetrafluoroethylene air-jet textured sewing threads. Design/methodology/approach – These threads are designed for sewing various garments and are manufactured by the Department of Textile Technology at Kaunas University of Technology. Manufacturing parameters are varied during air-jet-texturing, which includes air pressure, effect and core yarns overfeed. Tensile tests of sewing threads and seams strength tests are performed. Findings – They indicate that the strength of seams depends on the properties of sewing threads. Originality/value – Analysis of the seam strength of PES-PTFE air-jet-textured sewing threads. Keywords Polyesters, Thread, Lithuania, Research Paper type Research paper
International Journal of Clothing Science and Technology Vol. 17 No. 3/4, 2005 pp. 264-271 q Emerald Group Publishing Limited 0955-6222 DOI 10.1108/09556220510590966
1. Introduction The range of sewing threads available on the market has recently increased significantly. There are many reasons for this – the development of new fibres, the development and improvement of new thread manufacture processes, as well as the continuously increasing demand from the industry to get various sewing threads designed for sewing a wide assortment of articles (Kalaoglu, 2001). About 70-80 per cent of the sewing threads produced are used by the clothing industry and they are required to have good satiability in laundering and ironing. Sewing threads of cotton and polyester (PES) are the most commonly used in the manufacture of garments (Kalaoglu, 2001). Although they represent only a very small fraction of the cost of the garment, poor sewing threads can greatly increase production costs, as they cause frequent stoppages of sewing machines (Buzov et al., 1978). Textured PES threads are the most economical threads today and their introduction into clothing industry is one of the best methods to reduce production cost. During the process of sewing, the thread is rubbed against the guiding elements or the needle eye as the same point of the thread runs dozens of times through the eye. Textured threads can be used in most stitches except in the needles of lockstitches. They are also suited to surging and cover stitch machines and are widely used in knitted goods and action wear because of their high elongation and low costs. They are also used extensively as surging threads in woven goods (Miller, 1996). Air-jet-texturing is one of the most promising ways for manufacturing sewing threads of high quality and excellent mechanical properties (Buzov et al., 1978; Miller, 1996; Sengupta et al., 1996). The control of air-jet-texturing parameters enables the manufacture of threads possessing desirable features. Similarly other threads, the types of fibres used in the manufacture process,
composition of threads and finishing impact the quality of sewing threads. In the process of air-jet-texturing, various processing parameters significantly impact the composition of threads, i.e. air pressure, the ratio of speeds for feeding core-effect threads, etc. PES sewing threads have one disadvantage – at high sewing speeds and high friction with metal parts of the sewing machine these threads heat and start melting at , 2508C. Texturing is the process in which filaments are entangled by various methods, imparting softness and bulk to the product (Sengupta et al., 1996). Air-jet-textured threads are produced by feeding PES yarns through a turbulent region of compressed air. The yarn is opened, and loops are formed and closed. The loops are locked inside and on the surface by applying heat under tension. Such threads are often referred to as entangled ones. Air-jet-texturing offers sufficiently thick sewing threads, which ensures the stability of the seams. Due to circular cross-sections and smooth surface, PES multifilament sewing threads are more lustrous than those from natural yarns. PES is the best fibre for most sewing thread applications, being cheap, possessing high strength, good chemical properties, favourable elastic characteristics and good dye fastness (Hearle et al., 2001). 2. Experimental 2.1 Materials PES and polytetrafluoroethylene (PTFE) sewing threads, manufactured by the Department of Textile Technology at Kaunas University of Technology are analysed in the course of the research. Threads are manufactured on the Eltex air-jet-texturing machine. In the process of manufacture, the following three process parameters impacting the quality of a final product are varied: effect thread overfeed, core thread overfeed and the pressure of the air feed in the air-jet-texturing nozzle. Two PES multifilament threads are fed to the core and one PTFE multifilament thread to the effect. On the whole, three filament threads are used. Core threads are dampened in all cases. Effect threads overfeed (X3) vary from 13.20 to 46.80 per cent. Furthermore, the pressure of the air feed (X1) in the air-jet-textured nozzle is varied from 58:86 £ 104 to 117:72 £ 104 Pa and the core thread (X2) from 6.60 to 23.40 per cent. Mechanical indicators of the threads are analysed applying an experimental planning method and Box plan, the number of levels 2 2, the number of factors 2 3. After processing, the experiment data regression equations of the second-order were obtained. Dependence of the threads with structural effects on the three technological manufacturing parameters (overfeeds (X3, X2) and air pressure (X1) in the nozzle) for the analysed air-jet-textured threads, when thermo-setting was fixed, is presented graphically on a two-dimensional plane. It becomes a designing plane, whereas designing results become the coordinates of a two-dimensional plane defined by three variable factors. The general relation between the response Y (tensile characteristics) and manufacturing parameters X1, X2 and X3 (Krasovskyi and Filaretob, 1982) is as follows: Y ¼ B0 þ B1 X 1 þ B2 X 2 þ B3 X 3 þ B12 X 1 X 2 þ B13 X 1 X 3 þ B23 X 2 X 3 þ B11 X 21 þ B22 X 22 þ B33 X 23 The regression coefficients of the equations are given in Table I.
ð1Þ
The analysis of seam strength
265
IJCST 17,3/4
HemaJetw T321 air-jet-textured nozzle is used in the manufacture of the air-jet-textured thread. “Torlen FY HT” PES threads of increased strength (Table II) are used as the raw material. Furthermore, as the PTFE component is resistant to high temperatures, heterogeneous air-jet-textured sewing threads are suitable for working clothes that are exposed to high temperatures.
266 2.2 Methods The samples are sewn using the PES-PTFE threads on a universal lockstitch sewing machine, by straight superimposed seams. Stitch density is 4 cm2 1. Tensile tests are implemented according to the DIN EN ISO 13950, 1999 standard (ISO 2062, 1995). Tensile tests are performed on the CRE-type testing machine ZWICK/Z005 at the rate of extension 100 mm/min, and gauge length 200 mm. The number of tests per package is 5. Threads are enervated by both flexing and abrading. In addition, the developed test conditions are analogous to the impact of a sewing machine. Test conditions are as follows: load of threads 1.5, 2.0, 2.5 N, the number of circles 20; 50; 100, “Shmertz” needle of 90 Nm, needle-moving speed of 120 1/m, the number of tests: 10, the number of needles: 5. Tensile tests are implemented according to the DIN EN ISO 2062, 05/1995 standard (ISO 13935-1, 1999). Tensile tests are performed on the CRE-type testing machine ZWICK/Z005 at the rate of extension 500 mm/min, gauge length 500 mm, pretension 0.5 cN/tex, the number of tests per one package 20, stretching until the thread break. The samples are used for testing after the storage of at least 72 h in the conditioned laboratory (65 ^ 2 per cent RH, 20 ^ 28C). 3. Results and discussion Tensile characteristics of the PES-PTFE sewing threads are given in Table III. Air-jet-textured threads break differently than plain threads. Mechanical properties are determined by the core component, which is particularly important for stretching. Core component is the axis of an air-jet-textured thread, while the other components make a particular angle with this axis, defined by the overfeed quantity. Therefore, when
Table I. The regression coefficients of the PES-PTFE threads
Table II. The main parameters of the PES and PTFE threads used as raw material
Component
B0
B1
B2
B3
B12
B13
B23
B11
B22
B33
Dependencies of the seams breaking force 254.81 3.04 2 5.16 4.26 4.16 2 3.86 (0.76) 27.83 212.70 2 12.19
Parameter
PES value
PTFE value
Linear density of a single thread Number of filaments Specific breaking force Elongation at break Filament cross-section profile
133 dtex 32 54 cN/tex 16 per cent Circular
133 dtex 32 – – Circular
No. of PES-PTFE sewing threads 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
Resultant linear density, tex
Breaking force, N
Elongation at break (per cent)
Breaking tenacity, cN/tex
46.70 43.10 43.10 48.20 45.30 48.20 39.40 45.50 43.90 45.80 44.60 44.90 40.80 46.70 46.50
11.24 ^ 0.4 10.48 ^ 0.4 12.86 ^ 0.5 10.81 ^ 0.5 11.31 ^ 0.5 10.50 ^ 0.5 12.45 ^ 0.6 11.62 ^ 0.4 13.05 ^ 0.6 11.78 ^ 0.6 12.60 ^ 0.4 12.34 ^ 0.4 13.69 ^ 0.5 9.98 ^ 0.6 12.43 ^ 0.4
17.98 ^ 0.3 15.52 ^ 0.3 12.91 ^ 0.4 12.11 ^ 0.2 16.98 ^ 0.3 13.86 ^ 0.2 12.78 ^ 0.1 11.90 ^ 0.3 11.80 ^ 0.3 14.18 ^ 0.3 11.01 ^ 0.2 11.01 ^ 0.3 10.64 ^ 0.2 16.23 ^ 0.1 13.71 ^ 0.3
23.87 24.20 29.74 22.38 25.38 21.99 31.44 25.48 29.67 25.59 28.20 27.37 21.91 21.82 26.67
stretching an air-jet-textured thread, the core component breaks first, after which the effect component starts to deform and break. Besides, the character of the break is defined by the fact that the effect component is usually arranged in irregular filaments and in the process of stretching this component affects the core component by radial pressure. The indicator of specific break force of the air-jet-textured threads is not exact, since it is computed dividing the break force by linear density of the air-jet-textured threads. The thread breaks at the weakest point, and when computing specific break force total linear density of the threads with complex structure is considered. Specific elongation at break of the air-jet-textured threads depends greatly on the elongation of the core component. The effect component determines the strength when high elongation is reached. Results of testing the enervated PES-PTFE sewing threads are shown in Table IV. In the process of sewing threads are subjected to cycle tension and flexing loads, they are abraded in the contact with the fabric and the surfaces of the needle eye. Therefore, the strength of the threads decreases, whereas fibres in the filaments are damaged mechanically. In order to assess the influence of such an impact on the changes of the thread strength, cyclic flexing-abrasion tests are performed. Testing the enervated threads demonstrates that in many cases after subjecting to the enervation load of 1.5 N their breaking force increase after 20 cycles. This may be explained by the fact that filaments of the enervated threads are straightened, while physically they are not impacted to such a degree that could decrease their breaking force. The PES threads use feature-increased resistance; traditional PES multifilament threads have lower specific break force, i.e. ca. 30-40 cN/tex. PES multifilament threads with increased resistance are selected in order to upgrade the mechanical properties of the ready-made product. The PTFE threads are used to improve friction qualities of the sewing thread in the process of sewing. After the regression equations solving, the data obtained are processed and shown in Figures 1-3. The diagrams demonstrate the dependence of the properties of the seam
The analysis of seam strength
267
Table III. Tensile characteristics of the PES-PTFE sewing threads
Table IV. Cycle tension and flexing loads characteristics
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
12.83 10.34 14.13 11.74 9.76 12.93 11.29 10.13 12.25 12.55 13.93 13.57 11.14 11.50 13.33
1.5 N 20 cycles 12.33 13.09 12.23 11.71 10.99 9.51 13.34 – 11.69 – 6.44 9382 12.27 13.29 8.85
1.5 N 50 cycles 12.73 10.97 10.52 9.09 11.18 – – – 10.46 – – – – – –
1.5 N 100 cycles
2.0 N 50 cycles
Breaking force (n) 13.14 12.32 11.69 9.77 13.56 – 11.69 – 13.40 – 10.03 – 8.48 – 4.38 – 9.49 – 9.28 – 9.70 – 10.87 – 9.14 – 11.19 – 11.08 –
2.0 N 20 cycles 7.31 9.89 – – – – – – – – – – – – –
2.0 N 100 cycles
12.66 – 11.71 – 11.55 13.07 – 6.50 – – 11.03 – – – 11.83
2.5 N 20 cycles
268
Characteristics
9.78 – – – – – – – – – – – – – –
2.5 N 50 cycles
IJCST 17,3/4
breaking strength of the PES-PTFE threads as dependent upon the manufacturing parameters. Air-jet-textured threads break differently than simple plain threads. It is well known that mechanical properties are determined by the core component, and it is particularly important for stretching (Krasovskyi and Filaretob, 1982). The core component is the axis of the air-jet-textured thread, while the other components make a particular angle with this axis, defined by the overfeed quantity. Therefore, when stretching an air-jet-textured
The analysis of seam strength
269
Figure 1. Correlation between the seam breaking force and the manufacturing parameters of the PES-PTFE threads, when changing air pressure and effect threads overfeed
Figure 2. Correlation between the seam breaking force and the manufacturing parameters of the PES-PTFE threads, when changing air pressure and core threads overfeed
Figure 3. Correlation between the seam breaking force and the manufacturing parameters of the PES-PTFE threads, when changing effect and core threads overfeed
IJCST 17,3/4
270
thread, the core component breaks first and only then the effect component starts to deform and break. Furthermore, the character of the break is affected by the fact that the effect component is usually arranged in irregular filaments and in the process of stretching this component affects the core component by radial pressure. As the data in the figures show, the seam breaking force for the PES-PTFE threads depends directly on the manufacturing parameters. Maximum breaking force is achieved with mean values of the effect and core threads overfeed and air pressure as well. 4. Conclusions Air-jet-textured threads are manufactured in the Department of Textile Technology at Kaunas University of Technology using Eltex, state-of-the-art laboratory texturing machine with Heberlein air-jet-texturing nozzle and a dampening system. Different manufacturing parameters are varied during air-jet-texturing; including air pressure, effect and core yarn overfeeds. Tensile tests of the enervated sewing threads and seam strength tests are performed. Tensile tests of thread stretching indicate the following: . the usage of higher pressure in the process of air-jet-texturing allows decreasing elongation at break of the thread manufactured; . in individual cases it may be assumed that the force required to untwist elementary filaments is higher than their break force; and . the decrease of the air pressure and effect thread overfeed rises the specific break force of the thread manufactured. Summarising the results of the tests performed during this research indicates that by changing the parameters of the manufacturing process the properties of the thread, i.e. elongation at break, relative break force and absolute break force, may vary within certain limits. In industry, both types of threads may be used, due to their good properties and the possibility to preserve these properties in the process of sewing. Testing the enervated threads shows that, in many cases, after subjecting to the enervation load of 1.5 N the breaking force of the threads increased after 20 cycles. This may be explained by the fact that the filaments of the enervated threads straighten, while physically not impacted to such a degree that could decrease their breaking force. As the data in the figures show, the seam breaking force for the PES-PTFE threads depends directly on the manufacturing parameters. Maximum breaking force is achieved using mean values of effect and core threads overfeed and mean air pressure. It is evident that the strength of the seams depends on the properties of the sewing threads. Summarising of the results of the tests performed during this research indicates that by changing the manufacturing parameters the properties of the thread, i.e. elongation at break, relative break force, absolute break, tensile tests of enervated sewing threads and seams strength may vary within certain limits. In industry, the PES-PTFE threads may be used due to their good properties and because these properties can be preserved in the process of sewing. References Buzov, J.A., Modestova, T.A. and Alymenkova, N.D. (1978), Fabric Development in the Garment Industry, Lyogkaya Industria, Moscow, p. 480. Hearle, J.S.W., Hollick, L. and Wilson, D.K. (2001), Yarn Texturing Technology, Woohead Publishing Ltd, Abinto.
ISO 13935-1 (1999), “Textiles – seam tensile properties of fabrics and made-up textiles articles. Part 1: determinations of maximum force to seam rupture using the strip method”, p. 8. ISO 2062 (1995), “Textiles – yarns from packages – determination at break”, p. 8. Kalaoglu, F. (2001), “Holding it together”, TM, International Textile Magazine, No. 1, pp. 35-8. Krasovskyi, G.I. and Filaretob, G.F. (1982), Experimental Planning Method (in Russian), Legprombytizdat, Minsk. Miller, A. (1996), “Air-jet-texturing: a review”, Textile Technology Digital, No. 8, pp. 27-8. Sengupta, A.K., Kothari, V.K. and Sensarma, J.K. (1996), “Effects of filament modulus and linear density on the properties of air-jet textured yarns”, Textile Research Journal, Vol. 66 No. 7, pp. 452-5.
The analysis of seam strength
271
The Emerald Research Register for this journal is available at www.emeraldinsight.com/researchregister
The current issue and full text archive of this journal is available at www.emeraldinsight.com/0955-6222.htm
Objective measurement of pilling by image processing technique
Objective measurement of pilling
B.K. Behera and T.E. Madan Mohan Department of Textile Technology, Indian Institute of Technology, New Delhi, India
279 Received February 2005 Accepted February 2005
Abstract Purpose – This paper aims to report on a new pilling measurement system that has been developed using image processing technique. Design/methodology/approach – A pilling assessment cabinet is designed and developed which captures images and a software is developed to process and analyze the image of a pilled fabric to find out the various pilling parameters such as total number of pills, total area of the pills, mean area and number of pills per unit area. The image processing software processed image data of both the existing subjective assessment standards and pilled fabrics and assign suitable grades for comparison. Findings – The grades assigned by the machine correlates well with that of the experts grades and the results are reliably reproducible. The system can count the number of pills, find their total area, and their mean area. The results of EMPA-W2 and EMPA-W3 standards behave almost similar. The ASTM standards also gives somewhat the same results as the EMPA standards in number of pills but has a wide variation in the pilled area and mean pill area. The IWS standards produced an entirely different result from the other two standards, which leads us to the conclusion that all these standards are not objectively comparable to each other. Practical implications – The machine grade becomes a suitable methodology to compare the different grading systems. Originality/value – Traditional pilling tests are subjective by nature. Moreover, standards set by different organizations are not comparable with each other. This method presents a more universal and objective approach to describe the nature of the pilling. Keywords Image processing, Optical measurement, Fabric testing Paper type Research paper
1. Introduction Pilling is a long-standing problem in relation to staple fiber fabrics. The development of pills on a fabric surface, in addition to resulting in an unsightly appearance, can cause premature wear. Moreover, the advent of the synthetic fibers has accentuated the problem of pilling. Various properties of the synthetic fibers have been held responsible for their greater pilling propensity. Fabric resistance to pilling is commonly tested in the laboratory by specific machines by generating pilling on the fabric by simulating a wear. A sample of the original fabric is fixed in them in which wear is simulated by the action of abrasive materials. Generally, the machines are supplied with a standard reference consisting of photographs of samples with different degrees of pilling. The abraded fabric is then compared with the standard photographs, which are developed by the standard institutions like ASTM, AATCC, IWS, BIS, JIS, etc. and assign a degree of pilling accordingly. All the above said standards follow by and large the same principle to characterize the pilling severity of the fabrics. The standard photographs available for a particular
International Journal of Clothing Science and Technology Vol. 17 No. 5, 2005 pp. 279-291 q Emerald Group Publishing Limited 0955-6222 DOI 10.1108/09556220510616156
IJCST 17,5
280
fabric are confined to certain factors like the type of fabric, weave of the fabric, etc. Therefore, photograph of a woven fabric cannot be compared with a knitted fabric sample and similarly the photograph of a plain-woven fabric cannot be compared with a twill fabric sample, etc. Any of such subjective assessments are only accurate when performed by seasoned experts, and such experts may not be available at one place. Moreover, those standards set by different organizations are not comparable with each other. So there arises the need for a universal standard, which would be adopted worldwide that would best describe the nature of the pilling. If the piling severity is described by the number of pills, area of pills, their distribution over the sample, etc., then this would better give a clear idea about the quality of the fabric. 2. Pilling measurement In order to see the pilling tendency of a fabric the wear is simulated in laboratory. Generally, two main categories of machines are used to simulate this wear. In the first category, a flat abrasion tester like Martindale abrasion tester is used, which could be used with very low top loads. In the second category, probably more realistic from the point of view of wear behavior, a specially designed abrading system is used. Basically, the system consists of a container, whose internal surface is covered with a moderately abrasive layer. The specimens prepared in various manners are inserted in the container and kept there for a long time, during which they are subjected to an action of shaking and rubbing against each other and against the walls for pill formation. The pilled samples are generally assessed by the following three methods, (1) subjective assessment; (2) objective assessment; and (3) descriptive assessment. 2.1 Subjective assessment In subjective assessment, experts do the pilling evaluation of fabrics by visual inspection of the sample. A series of tested specimens of a specific fabric, which shows degree of pilling are stored under standard conditions i.e. under the conditions that will preserve their original form and appearance, were used for the comparison. Other than this type of real specimens, a set of five photographic standards of 80 £ 80 mm numbered 1-5 illustrating varying degrees of pilling from “very severe pilling” to “ no pilling”, are used for the visual comparison technique. The appearance of the face of the fabric samples are rated using the rating standards and with the scale shown in Table I. When the appearance of a test specimen falls between any two rating standards, then it is assigned the average value, for example, 3.5 or 2.5, etc. (Annual Book of ASTM Standards, 2000). The comparison could be made with a wide variety of standard set of photographs like ASTM, IWS, EMPA, JIS, etc. Generally, these types of grading are not accurate unless
Table I.
5 4 3 2 1
No pilling Slight pilling Moderate pilling Severe pilling Very severe pilling
experts who have a prolonged experience in grading the samples perform it. A study conducted by Neilly (1990) shows that the grades given by inexperienced observers vary widely. Moreover, the grades of one standard is not comparable with the other, i.e. EMPA standards characterize the size of pill as small, medium, and large pilling and then assign a grade in that category, whereas, IWS standards does not have such variations. So unless, if the customer has the specified standard photograph he cannot perceive what the test report says. Every country uses their own standards and they vary largely from one another. So there arises the necessity to standardize these standards and to quantify the pilling quantitatively by eliminating the standard photographs. 2.2 Objective assessment In order to get precise results, and to study about the behavior of fabrics in normal wear, researchers devised a simple objective method. Using a magnifying glass they simply counted the number of pills. Then the pills are sheared off from the fabric surface by a sharp razor blade, and weighed in a torsion balance. This method gives the real picture, of the behavior of a fabric in normal wear, i.e. the pilling propensity, weight loss of the fabric. Richards (1975) and Williams (1985) found the mean pill mass Mm and mean pill number Nm. From these values, they found the total pill weight from the product Nm and Mm. Richards found a good correlation between the weight of pills and the subjective visual assessment using standard photographs. Though this method gives a clear picture about the pilling tendency of a fabric, it is not followed in the industry because the method is quite laborious and need delicate skills to handle the seared off pills. Konda et al. (1988a, b) used an image analysis system to evaluate the pilling grade using a video camera. The image of the sample is captured with a video camera with a resolution of 240 £ 256: The image was then transformed to gray scale and converted to binary transformation. Similarly, they captured the images of the JIS standard photographs and converted to binary form. Both the images are later compared and then assigned a suitable grade. This method is particularly suited for solid shade fabrics where both the warp and weft are of same color. As the gray image is directly converted to binary image the speckles formed due to weave interlacement could not have been eliminated. However, as they did the same procedure to the standard photographs also, it could have resulted in unwanted speckles in that image also, where both the disturbances would have stood cancelled. Amirbayat and Alagha (1994) utilized the laser triangulation method for the grading of fabric pilling. 2.3 Descriptive assessment Neilly (n.d.) proposed a descriptive method of grading the surface pilling on knitted and woven fabrics. The purpose was to compare the conventional photographic method of grading the pilling to a descriptive method of grading which is simply based on five main descriptive grades. The correlation between each method and the consistency between individual observers for each fabric using each method was analyzed. The five grades used were: no change, fuzziness, pills beginning to form, pills formed and severe pilling. 3. Proposed method for pilling measurement The main objective of this work is to develop a pilling measurement system, which would calculate the important parameters that would give a clear picture of the pilling
Objective measurement of pilling 281
IJCST 17,5
282
severity. The system eliminates the subjective assessment method where an ambiguity always stays. A pilling assessment cabinet is designed and developed which captures images and a software is developed to process and analyze the image of a pilled fabric to find out the various pilling parameters such as total number of pills, total area of the pills, mean area, number of pills per unit area. The complete measurement process comprises two broad steps: imaging and image processing. For acquiring pilled images, an imaging system is designed and the same is interfaced with computer for storing image data. The image processing software is developed to process image data of both existing subjective assessment standards and pilled fabrics for comparison. 3.1 Imaging system For grabbing image of pilled fabric by digital camera, a closed chamber with controlled illumination system is designed. The chamber has a sample platform, camera holder and camera-computer interfacing arrangement so as to acquire the image and transfer the same to computer for processing, as shown in Figures 1 and 2. 3.2 Image processing In order to count the number of pills and to measure their area, the pills formed on the fabric surface has to be identified against the background fabric. The fabric may have a variety of weaves starting from plain to jacquard designs. And the yarn color effects may also present to produce various design from simple stripes to checks. The pill formed may be adhering to any part of the fabric surface. Now the main requirement is to segregate the pills from the inherent design of the fabric and no design or artifact of the fabric shall be misinterpreted as a pill. So the first step is to interpret the design of the fabric, and separate the pill from the fabric surface. The design identification of the fabric can be done by many methods; a general procedure, which was employed to search for the repeating patterns inside the image. But this search was carried out in spatial domain, i.e. directly in the image, which would take a lot of time for the calculations. Instead of doing all the calculations in spatial domain, the image was transformed into frequency domain where identification of the repeating frequencies
Figure 1. Image acquisition system
Objective measurement of pilling 283
Figure 2. Image acquisition chamber
were quite easy and faster than the operations done in the spatial domain. The frequency domain operation is carried out using the fast Fourier transformation, where the image is represented in the frequency domain and the image processing operations were carried out. 3.3 Fourier transforms The Fourier transform is an important image-processing tool, which is used to decompose an image into its sine and cosine components. The output of the transformation represents the image in the Fourier or frequency domain, while the input image is the spatial domain equivalent. In the Fourier domain image, each point represents a particular frequency contained in the spatial domain image.
IJCST 17,5
The Fourier transform theorem can be stated in one or multiple dimensions, depending on the number of independent variables used in the transformed function. If a function f(x) in the time (or spatial) domain is known, its Fourier transform F(u) is defined as, Z 1 f ðxÞe2j2pux dx FðuÞ ¼ 21
284
pffiffiffiffiffiffiffi where u is the variable frequency and j ¼ 21: For digital computation of the Fourier transform, a discrete version of the integrals is needed. Suppose M columns and N rows of data points are sampled from a two-dimensional function f(x, y) at evenly spaced intervals in both directions. The sampled image f(m, n) (m ¼ 0 to M 2 1 and n ¼ 0 to N 2 1) is repeated to form an endless sequence. The discrete Fourier transform (DFT) pair of this M £ N data array is, Fðu; vÞ ¼
21 M 21 X 1 NX f ðm; nÞe2j2Pðum=M þvn=NÞ MN n¼0 m¼0
f ðm; nÞ ¼
21 M 21 X 1 NX Fðu; vÞe 2j2Pðum=M þvn=N Þ MN v¼0 u¼0
and
Usually, M is selected to be identical to N. Unfortunately, the DFT is an extremely slow process, especially when the data set is large. So, in order to increase the efficiency of the process a newer algorithm has been introduced. This algorithm and other variants are now known as the fast Fourier transform (FFT) and are widely used in various applications. The basic idea of the FFT is that a DFT of an N data set is the sum of the DFTs of two subsets, even-numbered points and odd-numbered points. The data set can be recursively split into even and odd until the length equals one. Since the DFT of one point is equal to itself, the DFT of the data set becomes a series of simple operations. The FFT algorithm can reduce N2 operations in the DFT to N log2 N operations in the FFT. If N ¼ 1; 000; the FFT is about 100 times faster than the DFT. To use the FFT, N has to be a power of two. Since the resulting transform of a function is a complex function, the magnitude and phase have to be displayed separately. For an image f(m, n), the square of the magnitude M(u, v) is called the power spectrum P(u, v). The power spectrum is often displayed against frequency to show the contributions of each frequency to the function. According to the property of conjugate symmetry of the Fourier transform we can exchange the first quadrant of the spectrum with the third one, and the second quadrant with the fourth one to display a centered power spectrum. The power at he origin P(0, 0) indicates the so-called direct component (DC), reflecting the average brightness of the image, and the power spectrum at the other locations shows the different frequency terms present in the original image. Pixels on the same radius in a region near the origin represent low frequency terms that provide the overall structure of the image, while pixels in a region far from the origin represent high frequency terms such as the edges of the objects. Since frequency terms can be readily isolated in a power spectrum, the spectrum is particularly useful for identifying the periodic structure and filtering noise. The phase
information is seldom displayed, because it is difficult or impossible to visually interpret a phase display nevertheless it is needed for conducting the inverse FFT, which restores the image, based on the power spectrum and the phase. Applying the above said techniques the frequencies of the unwanted data are removed and the image was reconstructed from the modified spectrum. The camera using the computer interface first grabs the image of the sample. The camera freezes the frame in RGB, which contains all the data including the color information. The RGB image is then converted to gray image (Figure 3(a)), by removing the color information. Then the gray image is converted to frequency spectrum by the fast Fourier transformation (Figure 3(b) and (c)). The FFT spectrum shows the high frequency data by the spectacular peaks located according to the position of those repeating structures present in the spatial domain. These peaks are searched and found by a suitable search, and replaced by zeroes, if it is found to be greater than a specified threshold (Figure 3(d) and (e)). Replacing the peak at the center by zeroes, which is the representative of the high frequency repeat unit, entirely removes those image information, which belong to the weave pattern and designs formed due to the colors of the warp and the weft. The filtered FFT spectrum is then reconstructed and converted to binary by thresholding. This ON pixels, i.e. the white part of this binary image represents the pill which is actually located in the fabric. As the pills have been separated from the original image now it is easier to count them and find the area, their distribution, etc.
Objective measurement of pilling 285
Figure 3. Sample no. 1
IJCST 17,5
286
Figure 4.
3.4 Measurement algorithm The basic algorithm used for image processing of the pilled fabric is shown in Figure 4. Threshold becomes the key operation in identifying the pills present. A global threshold value is obtained for every image using the Otsu’s algorithm, which is known for computational efficiency and efficacy. Thus the threshold value is obtained and the image is converted to binary scale with 0s and 1s.
4. Pilling parameters The outputs obtained from the image are always in terms of pixels. For calculating all the pilling parameters the output obtained from the processed image has to be translated to the readable scale, i.e. in inches or centimeters. The conversion is done using the resolution of the original image and its size. Let the image resolution be M £ N pixels. And its size shall be P £ Q square inches, i.e. P represents the sample length and Q its width both in inches. Then, Total number of pixels ¼ Total area of the sample; i.e. M £ N ¼ P £ Q: Then, Area per pixel ðaÞ ¼
P £Q square inches M £N
4.1 Number of pills The term can be defined as the total number of pills present in the entire sample. Total number of pills ðN Þ ¼ Number of pills present in the area ðSÞ where (S) is the area of the sample in square inches. 4.2 Total area of the pills It is the total area of all the pills present in the sample. In order to find the area of the pills, the “ON” pixels, i.e. the pixels having the value “one” has to be counted in the final binary image. The product of total number of white pixels and the area of a pixel, would be the area of pills. Let the total number of white pixels be represented by T Total area ðAÞ ¼ a £ T where T is the number of white pixels and a the area of a pixel in square inches. 4.3 Mean area It is the average area of all the pills. ¼ A=N Mean area ðAÞ where A is the total area of the pills and N the number of pills. 4.4 Pills per unit area It can be defined as the number of pills present in a unit area. As we know the total number of pills (N ) are known, and so also total area of the sample (S), the pills per unit area shall be found out as follows,
Objective measurement of pilling 287
IJCST 17,5
Pills per unit area ¼
N S
These four parameters are directly transferred and stored in a report file from where print outs could be taken and used for further reference and analysis.
288
5. Results and discussion The standard photographs were scanned and the digital image was first processed in the newly built software. The pilling parameters which were discussed in the earlier chapters were obtained from the output of the software, i.e. the total number of pills, their area, mean pill area and the number of pills present in a square inch was obtained. The EMPA standard has two series of photographs. The K-series and the W-series, where the K-series stands for knitted fabrics, W1 stands for nonwoven fabrics and W2, W3 for woven fabrics. The standards W2 and W3 were processed in the system and the results are shown in Tables II and III, respectively. The ASTM standards’ image has given the results shown in Table IV and that of IWS in Table V.
Table II. Details about the pilling parameters of standards obtained from the system for EMPA standards (W2)
Grade
Table III. Details about the pilling parameters of standards obtained from the system for EMPA standards (W3)
Grade
1-2 2-3 3-4 4-5
1-2 2-3 3-4 4-5
Grade Table IV. Details about the pilling parameters of standards obtained from the system for ASTM standards
1 2 3 4 5
Grade Table V. Details about the pilling parameters of standards obtained from the system for IWS standards
1 2 3 4 5
Pill count
Area (square inches)
Mean pill area (square inches)
Pills per square inch
67 53 31 5
1.5924 0.9602 0.6640 0.0023
0.0238 0.0181 0.0214 0.0004
2.8089 2.2212 1.2992 0.2095
Pill count
Area (square inches)
Mean pill area (square inches)
Pills per square inch
81 40 17 0
1.8995 1.0151 0.5978 0.0000
0.0235 0.0254 0.0352 0.0000
3.39 1.67 0.71 0.00
Pill count
Area (square inches)
Mean pill area (square inches)
Pills per square inch
65 38 15 6 0
0.2702 0.0787 0.0262 0.0129 0.0000
0.0042 0.0021 0.0017 0.0022 0.0000
3.81 2.23 0.88 0.35 0.00
Pill count
Area (square inches)
Mean pill area (square inches)
Pills per square inch
16 12 7 2 0
0.5191 0.4449 0.1461 0.0548 0.0000
0.0324 0.0371 0.0208 0.0274 0.0000
9.102 6.826 3.980 1.137 0.000
The pilling parameters of the samples are compared with the pilling parameters obtained for the standard fabric images of ASTM and the equivalent grade is deduced. For example, sample no. 2 has the pill count as 37, pill area as 0.062 square inches and the mean pill area as 0.00186 (Figure 5). Comparing the values with the values obtained for the standard fabric images shows that the equivalent grade is “2” since the total number of pills, pill area and the mean pill area are 38, 0.0787 and 0.0021 which are in close approximation for the results of sample no. 2. Considering sample no. 3 (Figure 6), the equivalent ASTM grade becomes “4” since their parameters have close approximation, which could be observed from Tables VI and IV.
Objective measurement of pilling 289
Figure 5. Sample no. 2
IJCST 17,5
290
Figure 6. Sample no. 3
Table VI. Details about the pilling parameters of the samples obtained from the system
Sample no. 1 2 3
Pill count
Pill area (square inches)
Mean pill area (square inches)
35 37 3
0.054 0.062 0.0083
0.00157 0.00186 0.00143
Pills per square Equivalent ASTM inch grade 3.88 4.11 0.33
2 2 4
The following are the resultant images of fabrics tested using the software and the machine grade, on the basis of ASTM standards, of the samples are tabulated in Table VI. 6. Conclusions A pilling measurement system has been developed which could count the number of pills, find their total area, and their mean area. The results of EMPA-W2 and EMPA-W3 standards behave almost similar. The ASTM standards also gives somewhat the same results as the EMPA standards in number of pills but has a wide variation in the pilled area and mean pill area. The IWS standards produced an entirely different result from the other two standards, which leads us to the conclusion that all these standards are not objectively comparable to each other. The machine grade becomes a suitable methodology to compare the different grading systems. References Amirbayat, J. and Alagha, M.J. (1994), “The objective assessment of fabric pilling. Part-II. Experimental work”, Journal of Textile Institute, Vol. 85 No. 3, pp. 397-401. Annual Book of ASTM Standards vol. 07.01, 07.02, ASTM D 3514-99 (33-36), ASTM D 3511-99a (18-21), ASTM D 3512-99a (23-27), ASTM D 4970-99 (656-659) (2000). Konda, A., Xin Ling, C., Takadera, M., Okoshi, Y. and Toriumi, K. (1988a), “Members of the textile machinery society of Japan”, Transaction, Vol. 41 No. 7, pp. T113-123. Konda, A., Xin Ling, C., Takadera, M., Okoshi, Y. and Toriumi, K. (1988b), “Members of the textile machinery society of Japan”, Transaction, Vol. 41 No. 11, T152-T168 pp. 96-107. Neilly, D.G. (1990), “Grades of pilling”, Textile Asia, pp. 68-70. Richards, N. (1975), Journal of Textile Institute, Vol. 66 No. 73. Williams, V.A. (1985), Textile Research Journal, Vol. 55 No. 312.
Objective measurement of pilling 291
The Emerald Research Register for this journal is available at www.emeraldinsight.com/researchregister
IJCST 17,5
3D apparel creation based on computer mannequin model Part I: system kernel and formulas
292 Received June 2004 Accepted February 2004
The current issue and full text archive of this journal is available at www.emeraldinsight.com/0955-6222.htm
Jing-Jing Fang and Chang-Kai Liao Department of Mechanical Engineering, National Cheng Kung University, Tainan, Taiwan, Republic of China Abstract Purpose – This ongoing research revolute the conventional clothing design process by garment constructions in truly three dimensions rather than in two dimensions by ways of pattern design. The aim of the research is to develop a computer-assisted clothing design tool in complete three dimensions. It would provide the garment designers the capabilities of 3D basal garment creation, restyling, and static fitting analysis when wearing on a digital mannequin. The kernel of the design environment and the mathematical formulas used in garment creation are described, results and implementations will be presented later in part II of this paper. Design/methodology/approach – In this paper, a mannequin-based garment design and restyling tools in three dimensions is proposed. The tools are based on mathematical formulas which provide an intuitive way of computer-aided garment design. Findings – Free style creation on clothes is performed by the provided tools and its formulas behind. Feature-based mannequin model is initially constructed by its features interpolation. The crucial girths on garment, for instances, collar girth and sleeve girth are generated from the neck girth and the armhole girth, respectively. Based on the feature girths on the mannequin, garment surface is “radial grown” from the digital mannequin. B-spline surface, loft surface, and sweep surface are used to build blouse, sleeve, and collar for creation and restyling. Research limitations/implications – Basal garment is initially “grown” from the computer mannequin model, which means, size grading no longer becomes extra work. 3D restyling tool is then invoked to conduct versatile designs by exhibiting designers’ imagination space. Static fitting analysis is easily performed by the corresponding features on the mannequin. Originality/value – In this paper, a new three dimensions method in clothing design for clothes style creation and restyling in three dimensions on a digital mannequin model is proposed. In this research, a truly 3D garment design tool is developed in order to break through currently paper draft design concept. Keywords Garment industry, Clothing, Design for assembly Paper type Research paper
International Journal of Clothing Science and Technology Vol. 17 No. 5, 2005 pp. 292-306 q Emerald Group Publishing Limited 0955-6222 DOI 10.1108/09556220510616165
1. Introduction On account of automatic technique development, the conventional labor-based production line gradually lost its power of economic competition. There is no exception in clothing industry. In order to overcome time consuming and labor intensive problems, the technique of how to invoke automation in garment industry is crucial. In general, most of the garment design system uses computers to play an auxiliary role. This kind of system provides the functions of two-dimensional pattern design, modification and local scaling. It relies completely on the knowledge of experienced pattern designers in order to produce suitable patterns in various sizes. However, the
designs based on 2D pattern drafts cannot develop multiple styles fit on the mannequin without try-and-error. For the purpose of creating personal dressing style, we develop the clothing design software to conduct the modality of 3D visualization and modification. Not only provides garment styles creation but also its restyling. The technique consists of parameterized geometric modeling and undeveloped surface generation. B-spline geometric surfaces are utilized in this study for mapping the pivotal features from the mannequin (Tsai and Fang, 2003) on the designed garment. In terms of, the designed garment is “grown” from its “base”, which means, the mannequin. Using the interactive kernel interface provided by the software, clothing style is freely created by moving the control vertices on the garment geometric models. The designer is able to preview his or her own design concept of creativity in 3D. The aim of part I of the research is to develop a fundamental kernel environment for garment design in three dimensions, also to provide handy tools for 3D modification on the clothing models. The mathematical formulas used in clothing models are also described here. 2. Relevant study investigation In recent year, massive progress in computer hardware and software triggers many novel sciences, which may not be accomplished decades ago. Computer assisted drawing, design, engineering, manufacturing, or even tele-operation surgery, adopt the research powers in computer geometric modeling. With the ability of 3D computer animation in real-time, relevant studies and applications are widely expanded in apparel industry. Here, we are going to briefly survey the relevant academic researches, classify the different approaches in clothing techniques between their and our works. There are a few research teams working on the fields of human modeling and clothing simulation. Founded in 1989, Professor Magnenat-Thalmann from University of Geneva, leads a team working on virtual human dressing simulation. Their work (Hadap et al., 1999; Volino and Magnenat-Thalmann, 1997) focused on virtual garment draping effect on a virtual human in mixed realities. The effect soon be applied to real time animation (Cordier and Magnenat-Thalmann, 2002), and then presented on internet online clothing store (Cordier et al., 2003). The 2D patterns are constructed and imported in their garment simulator around a generic body. In a similar work done by Vassilev (2000) and Vassilev et al. (2001), they proposed a fast technique for cloth animation by using local B-spline patches to form human manikin and clothing. A similar research was simultaneously conducted in Hong Kong University of Science and Technology. Professor Yuen leads a group of researchers working on the similar subject of virtual human modeling for garment industry. Mannequin and manikin dummies are laser-scanned to form geometric templates (Au and Yuen, 1999; Wang et al., 2003a, b). Simple dress and skirt are constructed by employing variational subdivision approach for mesh interpolating from the given curves on mannequin profile. Based on the development of pattern design script language (Kang and Kim, 1999a) in Seoul National University, South Korea, Kang and Kim (1999b, 2003) present an apparel CAD system for the purpose of garment pattern generation. Garment model was aligned to scan body by convex hull technique to prevent major garment/body intrusion (Kim and Kang, 2003). In this paper, we propose a new 3D method in clothing design for clothes style creation and restyling in three dimension on a digital mannequin model. Feature-based
3D apparel creation. Part I
293
IJCST 17,5
294
mannequin model is initially constructed by its features interpolation. The features for clothing creation, such as shoulder points, neck shoulder points, armpits, bust points, crotch point, shoulder lines, bust girth, armscye, waist girth, hip girth, sidelines, front and back centerlines, front and back princess lines, neckline, etc., are first found by a generic algorithm (Tsai and Fang, 2003). Based on body measurements in anthropometry for apparel (ASTM, 1999; ISO, 1999), these crucial features points and girths on the specific mannequin are automatically extracted by mathematical definitions and implementations. In this research, we develop a truly 3D garment design tool in order to break through current paper draft design concept. Basal garment is initially “grown” from the computer mannequin model, which means, size grading no longer becomes extra work. Three dimension restyling tool is then invoked to conduct versatile designs by exhibiting designers’ imagination space. Static fitting analysis is easily performed by the corresponding features on the mannequin. 3. Infrastructure of computer-assisted 3D apparel design In this paper, we present an infrastructure for clothing design and restyling by way of computer-assisted 3D clothing creation in real-time. To achieve the goal of above, B-spline formulas and rule surface method are employed for surface reconstruction and style creation. Based on the frame of microsoft foundation classes (MFC) and three-dimensional drawing library (OpenGL), the system integrates object-oriented methodology as software developing infrastructure and its functional expansions. In the following, we describe the relationship between two modules of the software, the interface and the kernel. Messages flow between the kernel and the interface is divided into three types of command including design, draw, and restyle. Basic garment is constructed in the design object by ways of parameterized input, whereas, restyle creation from basic design is manipulated in the restyle object by ways of drag-and-drop controlled by input devices. Exhibition object, draw, demonstrates the geometric outcomes from both design and restyle objects. Figure 1 shows the relationships between the two modules. The kernel transmits messages through CDesign3dView. Once CDesign3dView
Figure 1. Relationship between the interface and the kernel module
receives commands from the user, a dialog window is triggered to demand data from the user. Geometric objects are then constructed and illustrated accordingly. Drawing modalities, such as point, line, mesh, and render, one of them is chosen and then transmitted to the kernel before canvas drawing. In kernel, every part of garment is associated to the “Garment” object (see Figure 2). The designed basal garment combines the parameterized blouse, collar, sleeve, dress, skirt, trouser, and others. Based on B-spline creation, refine procedure for garment style creation is simply achieved by dragging the movable control points distributed on garment modeling. Through interface module, the users indicate one or a group of control points on the designed clothes modeling for reshaping, the associated shape on surface is instantly changed according to the users manipulation (for example, see Figure 3). For the purpose of manipulation three-dimensional movement by ways of two-dimensional mouse or digitizer, at least seven different constraints are provided for single point or a group of cross section points. Each constrain provides movement of the control points along one of horizontal, sagittal, frontal, or radial plane; or in either direction of forward/backward, left/right, or upward/downward. In terms of this the movement of the control points would change the shape of the surfaces to perform restyling.
3D apparel creation. Part I
295
4. Kernel theory The kernel technique of the 3D garment design tools involved geometric modeling and computer graphics theorem, provide the clothing designer the capabilities of draping on a mannequin or modifying a pre-designed dress in 3D. The creative garment is then exported for further automatic pattern generation in the next stage, which is not described in this paper. Garment parts are constructed according to the flowchart in Figure 4.
Figure 2. The kernel structure
Figure 3. Restyle interface with: (a) control points distributed above clothes surface; and (b) a group of sectional points moving induce the changing on dress shape
IJCST 17,5
296
Figure 4. Part construction procedure
The pivotal dimensions are required including mannequin model, and its crucial features for blouse construction. These crucial features include but not limit to bust girth, waist girth, mid waist girth, hip girth, princess lines, shoulder lines, centerlines, etc. are necessary for basal garment construction. Accordingly, sleeve girth and collar girth on the blouse are, respectively, generated from armhole girth or neck girth by mathematic formulas. B-spline surfaces are using in blouse, dress, and sleeve parts, whereas, rule surfaces are using in collar part. The following sections describe how each garment part creates and its mathematic formulas behind. 4.1 Blouse Blouse is the part that covers human body from shoulder to hip which involving a couple of boundary girths, such as collar girth, sleeve girth, shoulder lines, centerlines, and sidelines. B-spline surfaces (Piegl and Tiller, 1997) are generated from boundary girths and other features on the mannequin to cover the mannequin. Based on ISO 8559 (1999) and ASTM D521999 (1999), the computerized mannequin (Tsai and Fang, 2003) implicitly stores significant feature points, feature girths and dimensions for garment making. It comprises of 5,416 points including every feature points and feature girths on it, which are clearly enough to represent a general mannequin for garment construction. Figure 5 shows the associated formulas involving in sequence. According to the blouse type the user select, fundamental collar girth and sleeve girth are generated from the neck girth and the armhole girth on the mannequin with respected associative formulas. The simple convex hull method is employed to create a dress-like garment around the bust portion. Refer to feature lines on the formed 2D pattern (Figure 6) and the inquired dimensions input, we create 3D reference points and used them to generate the geometric surfaces of front blouse and back blouse. However, by way of blouse dimensions input can only scale the specific lengths of garment portion, it may not meet the users requirements of design diversity. To enforce the variety of garment style, restyle function is provided in this paper. B-spline methodology is employed in collar girth generation, sleeve girth generation, and garment surface construction, also as a fundamental formula of restyling. In the following sections, we are going to describe them in detail.
3D apparel creation. Part I
297
Figure 5. Blouse construction
Figure 6. Blouse pattern features
IJCST 17,5
298
4.2 Collar girth generation Neck girth is one of the crucial feature girths on mannequin. According to the feature-based anthropometry, bottom neck girth does not lie on a planar in physical. To construct a basic collar girth for either convertible collar or shirt collar, practical formula is given by experienced clothing designers to meet the correct length of the collar girth. Here we briefly describe how to generate 3D collar girth from the structuralized neck girth points. Assume that B-spline curves C(u) is represented in terms of blending functions Ni,k(u). X N i;k ðuÞV i ð1Þ CðuÞ ¼ i
Vi represents the control points shown in dots in Figure 7. Suppose Qi represents the structuralized neck girth points, then Qj ¼ Cðuj Þ ¼
n X
N i;k ðui ÞV i
ð2Þ
i¼0
Equation (2) is rewritten in matrix form, then control points Vi are gained. C ¼ NV ) V ¼ N 21 C
ð3Þ
Figure 8 shows a top view diagram of both convertible and shirt collar girth generated by the associative predefined extending ratios from a neck girth. Suppose Ai is one of the data points in neck girth, Pc is the centroid of the girth, and Bi is the corresponding data point in collar girth radial stretched from Ai. Pfn notates the front neck point of the mannequin. Hence, ui ; /P fn P c Ai : Define distance dðui Þ ¼ jAi B~i j;
i ¼ 0; 1; 2; . . . ; n
ð4Þ
For each ui,ui is variable that depends on associated Ai obtained from equation (3). The modifiable collar girth generated here would allow the user to create various shape of collar girth. In addition, based on the designed collar girth, we are able to
Figure 7. Top views of shirt (left) and convertible (right) collar girths generated from a neck girth
3D apparel creation. Part I
299 Figure 8. Sleeve girth generation
generate associated collar band and its connected collar wing. The detailed method is illustrated in Fang (2003). In general, the size of a ready-made shirt is determined by the collar girth of the shirt. Therefore, the circumference of collar girth in our design is initially defined about 1 cm longer than the person’s (mannequin’s) neck circumference. 4.3 Sleeve girth generation Since our selected mannequin does not carry armhole feature, therefore, the armhole girth is approximated near the shoulder portion of the digital mannequin. Based on the bending value method (Wang et al., 1995), we first search the maximum bending value on the mannequin between the shoulder girth and the bust girth. The computerized 21 girths from shoulder to bust on the digital mannequin are shown in isometric view in Figure 8(a). The top view of the three pivoting girths, shoulder girth, bust girth, and sleeve girth is shown in Figure 8(b). The convex hull method is used to mimic the physical tape-measurement of the mannequin for the purpose of bust girth generation. The girth described in this paper is in a form of polygon which connects a group of scanning points. Therefore, armhole girth is the polygon in the ways of connecting every bending point of the 21 girths by maximum bending value method. Define P i P iþk ¼ xifd ; yifd ;k [ N ; ’k; V ¼ {P i j;i [ N } bd bd P i P i2k ¼ xi ; yi where 8 fd > x ¼ xiþk 2 xi > > i > > > < y fd ¼ yiþk 2 yi i > xbd > i ¼ xi2k 2 xi > > > > ybd ¼ x 2 x : i2k i i Let
8 > > xci ¼ xifd þ xibd > > < ) > > fd c bd > > : yi ¼ yi þ yi
IJCST 17,5
300
bending value BV ðPÞ ¼ Max xci ; yci
ð5Þ
and the P is the selected bending point. Notice that the calculated bending points of the armhole girth is not coplanar. Therefore, we use the three feature points: shoulder point, front and back armpits to create a virtual plane (see Figure 8(a)). Consequently, the top half of the sleeve girth is developed by projecting the armpit girth to the virtual plane. On the virtual plane, we mirror the preset half sleeve girth by the line through front and back armpits in order to obtain the other half of the sleeve girth. The outcome of the obtained sleeve girth is shown in Figure 8. 4.4 Blouse surface B-spline surfaces are widely used in many different fields of design. General CAD/CAM software benefit from its flexibility in advanced design ability. In this research, we adopted one of its derivative methodologies, skinned surface (Piegl and Tiller, 1997), to develop the geometric surface of the blouse. The original form of the B-spline skinned surface Qk;l ¼ Sðuk ; v l Þ ¼
n X m X
N i;p ðuk ÞN j;q ðvl ÞP i;j
ð6Þ
i¼0 j¼0
was rewritten into Qk;l ¼ Sðuk ; v l Þ ¼
n X m X
N i;p ðuk ÞðN j;q ðvl ÞP i;j Þ ¼
i¼0 j¼0
n X
N i;p ðuk ÞRi;l ;
ð7Þ
i¼0
whereas Ri;l ¼
m X
N j;q ðvl ÞP i;j
j¼0
Qk,l denotes the known reference points; u k ; v l represent the corresponding data to two parametric variables u, and v. Using chord length distribution to obtain ukl0 with its numbers of r £ sðr ¼ n þ p þ 1; s ¼ m þ q þ 1Þ, then u k ¼
s X
u0kl =s:
l¼0
Similar way to obtain the parameter v i . In order to retrieve those control points uniformly distributed on the garment surface, we reorganized the positions by equal chord length method. Therefore, movable control points are distributed well proportioned for further use in shape restyling. Figure 9 shows these control points distribution from original to uniform positions.
3D apparel creation. Part I
301
Figure 9. Original (left) and uniform (right) control points on a half digital mannequin
4.5 Collar surface Collar girth is a spatial curve that the collar sweep along in order to generate a collar surface. However, the appeared segmentations (Figure 10(b)) are not always consistent. They vary relying on orientation and position beyond collar girth. Shapes of general shirt collar and convertible collar are defined according to wearing figure in practical. Mathematical formula is given to build its dimensions and segmentations. In general, variation of convertible collar frequently appears in garment design. Three parameters are given from the garment designer as shown in Figure 11. l1 notates the rear collar height, l2 represents rear collar width, and variable angle w is the wide-open angle from the sigittal plane to the front edge of collar face.
Figure 10. Half convertible collar: (a) pattern; and (b) segmentation
Figure 11. Parameterized convertible collar
IJCST 17,5
It was initially defined that
a1 ðui Þ ¼ ða1 ðun Þ 2 a1 ðu0
302
ÞÞ* sin
ui 2 w p þ a1 ðu0 Þ; p2w* 2
i ¼ 1; 2; · · ·; n 2 1
ð8Þ
ui 2 w p þ a2 ðu0 Þ; p2w* 2
i ¼ 1; 2; · · ·; n 2 1
ð9Þ
and
a2 ðui Þ ¼ ða2 ðun Þ 2 a2 ðu0 ÞÞ* sin
L1 ðui Þ ¼
3 X
aj uij ;
i ¼ 0; . . . ; n
ð10Þ
j¼0
Rewrite equation (12) in matrix form L1 ¼ QA ) A ¼ Q21 L1
ð11Þ
and L 2 ð ui Þ ¼
3 X
bj uij ;
i ¼ 0; . . . ; n
ð12Þ
j¼0
To be a fundamental convertible collar, b0 =constant, and b1 ¼ b2 ¼ b3 ¼ 0. Figure 12 shows the screen displays of the user-defined lengths as l1 ¼ 50 mm, l2 ¼ 100 mm, and w ¼ 9. The major difference between convertible collar and shirt collar is, the shirt collar consists of two pieces of pattern instead of one in convertible collar. The prolonged collar band length provides space for button and buttonhole. Therefore, in 3D profile, the collar band of shirt collar is closer to human neck than convertible band for the sake of necktie applying. Figure 13 shows the notations of user-defined parameters, such as rear collar band height l1, rear collar wing width l2, front collar width l3, and wide-open angle w. The preset formulas are given collar band height L1(ui), and collar wing width L3(ui) 2 L2(ui).
Figure 12. Different point of views of a convertible collar
Define L1(ui) and L3(ui) as 8 L 1 ð ui Þ ¼ l1 ; p=3 # ui # p > > > < 3 X L 1 ð ui Þ ¼ aj uij ; 2p=6 # ui # p=3 > > > : j¼0 L3 ðui Þ ¼
3D apparel creation. Part I ð13Þ
303
ui 2 ul ðl3 2 ðl2 þ l4 ÞÞ þ l2 þ l4 ; w 2 ul *
C 1 ðL3 ; ui Þ ¼ ðl2 þ l4 ; ul Þ
i ¼ 0; · · ·; ul
ð14Þ
C 2 ðL3 ; ui Þ ¼ ðl3 ; 0Þ
Curves C1 and C2 are defined by L3(ui) and ui, which determine the expansion of the collar wing area near tip of collar wing. It is given that l1 ¼ 30 mm, l3 ¼ 40 mm, l4 ¼ 60 mm, then four-view outcomes are shown in Figure. 14. 4.6 Sleeve surface Owing to un-presence of the mannequin arm, sleeve construction mimics the 3D sleeve figure. Based on the pivotal lengths in 3D projection view (Figure 15) and sleeve pattern, a composite B-spline loft surface is created to mimic a three-dimensional sleeve. General sleeve is generated from the sleeve girth to the cuff. Figure 15 demonstrates a sleeve diagram. The sleeve girth developed in Section 4.3 is located on the section plane SP0, which is the virtual plane, passes through shoulder point, front armpit, and back armpit. The B-spline loft surface is passing through sectional planes SP1, SP2, and SP3. Sleeve orientation as shown in Figure 16 demonstrates the lift and extension angles between mannequin body and virtual mannequin arm. The length of cap height in Figure 16 is determined by the lift angle b between the virtual arm and the z-axis. The shorter the cap height is, the greater the lift angle.
Figure 13. Half shirt collar definitions
Figure 14. Different point of views of a shirt collar
IJCST 17,5
304
Figure 15. Sleeve diagram
Figure 16. Sleeve orientation
Associate to sleeve pattern in Figure 17, length of sleeve cap is equal to the curve length from Cen0, Cen1, to Cen2 in Figure 15. The cuff width is equal to the intersection curve length between SP2 and the composite B-spline loft surface. The ratio g between the length of front biceps and back biceps would set-up the extension angle a between the virtual arm and the y-axis. If ratio g ¼ 1, then a ¼ 0 the sleeve is perpendicular to the xz plane. If ratio g is greater than 1, then sleeve extension is over the transversal plane of the mannequin. In general, ratio g is greater than 0 and less than 1 in order to fit body kinematics in steady state. 5. Conclusion In this paper, we propose a mannequin-based garment design and restyling tools in 3D. The tools are based on mathematical formulas, which provide an intuitive way of computer-aided garment design. Free style creation on clothes is performed by the provided tools and its formulas behind. The crucial girths on garment, for instances, collar girth and sleeve girth are generated from the neck girth and the armhole girth, respectively. Based on the feature girths on the mannequin, garment surface is “radial grown” from the digital mannequin. B-spline surface, loft surface, and sweep surface are used to build blouse, sleeve, and collar for creation and restyling.
3D apparel creation. Part I
305
Figure 17. Sleeve pattern
References ASTM (American Society for Testing and Materials), Standard Terminology (1999), “Relating to body dimensions for apparel sizing”, D5219-99 , ASTM, Philadelphia, PA. Au, C.K. and Yuen, M.M.F. (1999), “Feature-based reverse engineering of mannequin for garment design”, Computer-Aided Design, Vol. 31 No. 12, pp. 751-9. Cordier, F. and Magnenat-Thalmann, N. (2002), “Real-time animation of dressed virtual humans”, Proceedings of EUROGRPHICS, Vol. 21 No. 3. Cordier, F., Seo, H. and Magnenat-Thalmann, N. (2003), “Made-to-measure technologies for an online clothing store”, IEEE Computer Graphics and Applications, Vol. 23 No. 1, pp. 38-48. Fang, J.J. (2003), “3D collar design creation”, International Journal of Clothing Science and Technology, Vol. 15 No. 2, pp. 88-106. Hadap, S., Bangarter, E., Volino, P. and Magnenat-Thalmann, N. (1999), “Animating wrinkles on clothes”, paper presented at the IEEE Visualization ’99, IEEE Computer Society Press, San Francisco, pp. 175-82. ISO (1999), “Garment construction and anthropometric surveys – body dimensions”, IOS 8559:1999(E), International Standards Organization, Geneva. Kang, T.J. and Kim, S.M. (1999a), “Development of three-dimensional apparel CAD system part I: flat garment pattern drafting system”, International Journal of Clothing Science and Technology, Vol. 12 No. 1, pp. 26-38. Kang, T.J. and Kim, S.M. (1999b), “Development of three-dimensional apparel CAD system part II: prediction of garment drape shape”, International Journal of Clothing Science and Technology, Vol. 12 No. 1, pp. 39-49. Kim, S.M. and Kang, T.J. (2003), “Garment pattern generation from body scan data”, Computer-Aided Design, Vol. 35 No. 7, pp. 611-18.
IJCST 17,5
306
Piegl, L. and Tiller, W. (1997), The NURBS Book, 2nd ed., Springer-Verlag, New York, NY. Tsai, M.J. and Fang, J.J. (2003), “A feature based data structure for computer manikin”, Taiwan patent number 04083- 09220535030, 2003; USA Patent pending number 10/699,640. Vassilev, T.I. (2000), “Dressing virtual people”, paper presented at the SCI’2000 Conference, Orlando , pp. 23-6. Vassilev, T.I., Spanlang, B. and Chrysanthou, Y. (2001), “Fast cloth animation on walking avatars”, Proceeding of 2001 Eurographics. Volino, P. and Magnenat-Thalmann, N. (1997), “Developing simulation techniques for an interactive clothing system”, Proc. VSMM’97, Geneva, pp. 109-18. Wang, M.J., Wu, W.Y., Huang, L.K. and Wang, D.M. (1995), “Corner detection using bending value”, Pattern Recognition Letters, Vol. 16, pp. 575-83. Wang, C.C.L., Chang, T.K.K. and Yuen, M.M.F. (2003a), “From laser-scanned data to feature human model: a system based on fuzzy logic concept”, Computer-Aided Design, Vol. 35 No. 3, pp. 241-53. Wang, C.C.L., Wang, Y. and Yuen, M.M.F. (2003b), “Feature based 3D garment design through 2D sketches”, Computer-Aided Design, Vol. 35 No. 7, pp. 659-72.
The Emerald Research Register for this journal is available at www.emeraldinsight.com/researchregister
The current issue and full text archive of this journal is available at www.emeraldinsight.com/0955-6222.htm
3D apparel creation based on computer mannequin model Part II: implementations Jing-Jing Fang and Chang-Kai Liao Department of Mechanical Engineering, National Cheng Kung University, Tainan, Taiwan, Republic of China
3D apparel creation. Part II
307 Received June 2004 Revised February 2005 Accepted February 2005
Abstract Purpose – A significant garment restyling tool was developed to perform garment design in three dimensions. It provides the professional designers the abilities of 3D garment creation, restyling, omni-angle visualization, and fitting evaluation on a digital mannequin model. According to the body tape-measurements defined in ISO 8559:1999(E) and ASTM D5219-99 (1999), the extracted feature lines on computer mannequin dominate the shape of the apparel and also its associated fitting results. In this paper, the garment creation by the provided interfaces and its outcomes based on the developed system kernel and its formulas described in part I of the paper is demonstrated. Design/methodology/approach – In part II of this paper, a three-dimensional garment creation and restyling software based on the kernel infrastructure and formulas is implemented. Findings – Currently, three fundamental dresses, two basic collars, and sleeve are successfully implemented in the creation of free style mannequin-made apparels. Fitting results in static status are easily performed by detecting the allowances along the body feature lines and its near by. The base of an intuitive 3D computer-aided garment design and manufacture is gradually formed starting from here. Originality/value – In this paper, it is proved that garment creation and restyling can be achieved in three dimensions. This work provides a solution to how to manipulate a modifiable geometry in three dimensions and provide a friendly tool for reshaping. Keywords Garment industry, Clothing, Design for assembly, Computer aided design Paper type Research paper
1. Introduction Although automation technique is successfully used in computer-aided apparel manufacturing industry, the garment designers are still using pen-drawing on a piece of paper to express their creativity on making fashions. As long as the primitive drawing used in the garment design process, computer-aided apparel design cannot be promoted from 2D to 3D. In a broad sense, computer-aided apparel design should include not only the 2D pattern drawing but also 3D style creation and 2D pattern generation. However, most of the current apparel CAD is narrowly focused on the area of 2D pattern construction. In which, it may still follow the specific prototypes such as Mode et Mode, Bunka prototype, and Dress Making from Japan. The authors would like to thank the long-term collaboration partners, Professional Wu L.J., and Lecturer Huang S.C. from Department of Fashion Design, Tainan Woman’s College on Arts and Technology.
International Journal of Clothing Science and Technology Vol. 17 No. 5, 2005 pp. 307-319 q Emerald Group Publishing Limited 0955-6222 DOI 10.1108/09556220510616174
IJCST 17,5
308
The ultimate target of apparel industry automation is to fulfill the individual requirements of made-to-measure, subjects of uniqueness and creation of personal style. However, human body is rather hard to be unified; therefore, many different sizing mannequins are statistically evaluated in order to produce ready-made garment. The automation system based on sizing mannequins would provide the garment designer an intuitive and handy creation in fully three dimensions. Hence, the basic requirements of clothing design are framed on the mannequin’s geometry. Through the self-developed body scanner, mannequin geometric features are automatically extracted by mathematical analysis. Cheng (2001) and Tsai and Fang (2003) introduced the feature-based methodologies to reconstruct the surface of digital mannequin as shown in Figure 1 by forms of points, lines, polygons, and rendering. According to anthropometry and body measurements for garment manufacturing (ASTM (American Society for Testing and Materials), Standard Terminology, 1999), crucial features are extracted and regenerated by a limit number of structuralized points. Phenomenally, full body points below 25,000 are capable to rebuild the geometric shape of the mannequin trunk in detail. In addition, the tape measurements for garment construction are automatically obtained from the computer model. Based on the computer model, drafted garment is directly generated from the computer mannequin. The associated points, girths, lengths, widths, heights, and even shapes gradient on the designed garment is fidelity stored. The aim of this 3D garment design software provides the tools of visualization, modification, and restyling in three dimensions. Creativity is carried out from three-dimensional garment creation, which are generated and built from the computer model by given associate parameters. Through the given interface of clothing design tool, customers and designers can easily communicate through the window navigation, in terms, a computer screen. Complex and laborious design procedure of made-to-measure garment becomes simple and reasonable. Compare to conventional 2D sketch on piece of papers, it provides a friendly communication interface between customers and designer. In addition, it allows designers and customers to modify and refine the fashion outcomes in 3D before it goes into the stage of manufacturing. 2. Relevant software investigation In recent year, progress in computer hardware technique promotes high performance software development among general personal computers. Computer assisted drawing,
Figure 1. Computer mannequin model shown in (a) points, (b) vertical lines, (c) lattices, and (d) rendering
design, engineering, manufacturing, or even surgery, adopting the techniques of computer graphics and geometric modeling are in its florescence. With the ability of 3D computer animation in real-time, relevant developments are comprehensively applied in apparel industry. Below, we are going to briefly survey commercial apparel CAD products, classify the different approaches among these tools, and then compare them with our ongoing work. Currently, the commercially available apparel design software such as the products from PatternMaker Inc. (1992), Geoffrey E. Macpherson Ltd (1938), and DoCAD Ltd (2001), developed a series of pattern CAD tools. It highly relies on the tailors’ skills and experiences to transfer a fantastic design of drawing into realistic wearable clothes. Providing automatic marking, grading, and manufacturing in Gerber Technology Company (1969), they have successfully applied patterns to apparel sewing. However, they focused on reducing the tedious works of try and error on 2D pattern developing for pattern makers. Rather than truly design in three dimensions, the appearance of 3D garment in Pad system Inc. (1988) and Lectra Ltd (1973) provided advanced functions for 3D visualization and sizing on both the specific dress and the manikin, but not for designs variety. Most of the commercial software for apparel CAD are based on two-dimensional pattern maker, and thus, sewing up a three-dimensional garment for visualization. Since 2D pattern pieces are forced to be virtually sewed together on its associated sewing edges without draping over a mannequin, on which forms an unnatural shape as a “gas tank”. In addition, it lacks fitting survey and also draping effects. The 3D approached design software for sark, named 3D Design Concept by Computer Design Inc. (1992), allows the users drawing the brassiere profile on a computer mannequin in order to flatten into patterns. Their projective method from screen drawing a ready-made-bra on a grading mannequin top is allowed to produce tightened underwear. In additions, it is anticipated that the design would not adapt to women’s bra that usually be discovered incompletely symmetry. To deal with such imperfectly symmetric human body in bra manufacturing is one of our research goals in the next stage. Different from the known apparel CAD software described above, our research goal is to develop a complete 3D design tool rather than a 2D pattern maker. Our work is not going to replace the current industrial successful 2D pattern maker, but try to promote the traditional pen-drawing fantastic fashion into a visible computerized fashion. The work is two stages ahead the flow of apparel design/manufacturing automation. Our anticipation of apparel design/manufacturing automation is shown in Figure 2. A feature-based computer mannequin provides the crucial positions and lengths for garment generation. By given associated dimensions on the 3D apparel fashion software, a selected garment style is “grown” above the mannequin’s top. In order to produce various shapes of clothes, tools for modifiable design and effects of draping are provided. Combine with the pattern makers’ expertise, draft pattern including pleats are directed flatten from the designed clothes. The draft patterns would then import to the general commercial software in common DXF format for transferring into current apparel production line. Fitting analysis on body top is easily performed by simply mapping corresponding garment features to its original mannequin model.
3D apparel creation. Part II
309
IJCST 17,5
310
Figure 2. Automation flow for apparel design and manufacturing
3. Interfaces and implementations Based on the kernel interface and geometric methodologies described in the part I of the paper, the provided 3D design interfaces and results are illustrated here. It reveals that the 3D apparel CAD modeling does not involve regular surface to form various shapes of clothes. The ongoing work initially displayed the designed contents of dress, sleeve, collar, and their combinations. Flowchart in Figure 3 shows the design process for garment part construction.
Figure 3. The design process for garment part construction
A digital mannequin is initially loaded into the design system. The system provides the users to select the part of the garment to be constructed, such as “dress”, “sleeve”, or “collar”. Refer to the feature lengths of the mannequin, garment parameters are given by the users for basal garment surface construction. Fashion is designed by mouse dragging the control points distributed on above the garment surface in specific directions, such as moving along either one of selected horizontal, sagittal, frontal, or radial plane; or in either direction of forward/backward, left/right, or upward/downward of the mannequin. The interactive icons for surface modification of moving constraints are shown in Figure 4. In the following sections, we are going to demonstrate a few fashion samples constructed from the garment restyling software.
3D apparel creation. Part II
311
3.1 Dress Currently, three types of dresses, blouse, sheath dress, and shift dress, are available in our system. Their tape-measurements of bust girth, waist girth, and hip girth are automatically calculated and appeared in read-only edit boxes shown in the mid column of Figure 5. Basal lengths of the blouse are predefined by increasing associated weighting from the girth lengths; certainly, it is changeable by the users as displayed in the right column of Figure 5. By given associated lengths for blouse, sheath dress, and shift dress, Figures 6-8 show original designs with different view angles.
Figure 4. Moving constraints
Figure 5. Dress dialog box
IJCST 17,5
312
Figure 6. Basal blouse
Figure 7. Sheath dress
At the process of shape modification, we demonstrate the reshaped sheath dress (Figure 8(c)) from its original design in Figure 8(a). For the purpose of changing the shape of the dress, a series of control points distributed above the dress surface are provided for surface dragging by ways of mouse handle (Figure 9). More advance restyling is shown in Figure 10, new style is created by moving the control points on the collar girth and skirt hem of a dress.
3D apparel creation. Part II
313
Figure 8. Shift dress
Figure 9. Restyling from original sheath dress
3.2 Collar Two fundamental collars, the convertible collar and the shirt collar, are provided for collar creation in the system. For example, a convertible collar (Figure 11), collar geometry is constructed by the given lengths of collar stand height l1, collar height l2, and the open angle of collar wing w. Here, we demonstrate three styles of convertible collar with respective parameters l1, l2, w to 3, 8, 5; 4, 9, 15; and 5, 13, 20, from left to right in Figure 12. As a shirt collar (Figure 13), collar surface construction is based on the given lengths of collar stand height l1, neck height l2, neck point height l3, and its open angle of collar wing w. By setting the construction dimensions l1, l2, l3, w, with
IJCST 17,5
314
Figure 10. Collar girth and skirt hem restyling from original sheath dress
Figure 11. Convertible collar setting
Figure 12. Parameterized convertible collars
3D apparel creation. Part II
315
Figure 13. Shirt collar setting
respect to 3, 4, 6, 4; 3, 5, 8, 5; and 2, 3, 8, 5 would illustrate shirt collar in three dimensions from left to right shown in Figure 14. A top with restyled shirt collar is shown in Figure 15. 3.3 Sleeve Sleeve surface is generated from the defined 3D sleeve girth, and its given lengths of cap height, front biceps, back biceps, sleeve length, cuff width, and cuff depth. Figure 16 shows a long-sleeve top generated from its setup of cap height 5.5 cm, front
Figure 14. Parameterized shirt collars
Figure 15. A top with restyled shirt collar
IJCST 17,5
316
Figure 16. A top with long sleeve
biceps 15 cm, back biceps 16 cm, sleeve length 40 cm, cuff width 10 cm, and cuff depth 5 cm. Here, we also demonstrate a half-length sleeve dress in the left of Figure 17, which was modified to be a dress with lantern sleeve in which cuff width and cuff depth remain unchanged. In addition, accessory would always promote the value of garment creation. Figure 18 shows a shirt creation by adding buttons and pocket (on the left chest) on it. 4. Fitting analysis Based on the computer mannequin model and its generated garment, the users are able to restyle the designed garment by provided kernel interface. According to the crucial feature lines on a mannequin, body-fitting diagram reveals detailed allowances
Figure 17. A dress with half sleeve and restyled lantern sleeve
3D apparel creation. Part II
317
Figure 18. Shirt with buttons
between the mannequin and the designed garment. In this section, we are going to show the static fitting results of the designed garment on a mannequin. A fitness diagram is shown in Figure 19 which displays the allowances between half of the mannequin body and the designed dress worn on it. In the longitudinal axis, number 0, 10, 20, 30, 40 represent the front center line, front princess line, side line, back princess line, and back center line on the mannequin, respectively. Curves with different colors denote respective allowances along its specific girths and the mannequin body. Several girths used in garment manufacturing are provided to comprise the fitting diagram. By simply clicking on the check box besides, these allowance curves of shoulder girth, bust girth, under bust girth, waist girth, middle
Figure 19. Fitting diagram
IJCST 17,5
318
waist girth, hip girth, and crotch girth, are in either visible or invisible status. Value of each allowance curve respective to every longitudinal index is also listed in fitting table (Figure 20). Allowances of the feature lines near by would be found by simple interpolation. 5. Conclusion In part II of this paper, we implement a three-dimensional garment creation and restyling software based on the kernel infrastructure and formulas. Currently, three fundamental dresses, two basic collars, and sleeve are successfully implemented in the creation of free style mannequin-made apparels. Fitting results in static status are easily performed by detecting the allowances along the body feature lines and its near by. The base of an intuitive 3D computer-aided garment design and manufacture is gradually formed starting from here. 6. Discussion In this paper, we have proved that garment creation and restyling is possible to be achieved in three dimensions. Why the 3D apparel CAD software has not been conducted in the apparel industry? From the viewpoint of manufacturing, we have picked up a few reasons or problems. How to manipulate a modifiable geometry in three dimensions and provide a friendly tool for reshaping? How to transfer the designed geometries (usually are undeveloped surfaces) into a workable 2D pattern? How to handle the appearances of distorted graph patterns printing on the garment surface? How to involve draping to promote the shape of dressing? Concern to fitted garment analysis, how to address the fabrics mechanical properties and its material properties in achieving the final shape is a complex problem. The first problem could be solved in this paper. Our ongoing studies are concurrently working on creating various apparel styles by cutting; adding more
Figure 20. Fitting table
garment pieces to the system such as trousers, skirt; accessory design and its database construction. Besides that, intelligent pattern generation from the undeveloped surface (usually B-spline surface) and patterns layout from the 3D designed dress is also one of our successive works. References ASTM (American Society for Testing and Materials), Standard Terminology (1999), “Relating to body dimensions for apparel sizing”, D5219-99 , ASTM, Philadelphia, PA. Cheng, K.H. (2001), “Feature extraction from 3D human body’s data point and construction of computer mannequin”, Master thesis, Department of Mechanical Engineering, National Cheng Kung University, Taiwan. Computer Design Inc. (1992-1995), 3D Design Concept Reference Manual, vols 1/2. DoCAD Ltd (2001), DoCAD Drawing Software Reference Manual, DoCAD, Taiwan. Geoffrey E. Macpherson Ltd (1938), available at: www.macphersons.co.uk Gerber Technology Company (1969), available at: www.gerbertechnology.com/ Lectra Ltd (1973), available at: www.lectra.com/ Pad System Inc. (1988), available at: www.padsystem.com PatternMaker Inc. (1992), available at: www.patternmaker.com Tsai, M.J. and Fang, J.J. (2003), “A feature based data structure for computer manikin”, Taiwan patent number 04083- 09220535030, 2003; USA Patent pending number 10/699,640. Further reading ISO (1999), Garment Construction and Anthropometric Surveys – Body Dimensions, IOS 8559:1999(E), International Standards Organization, Geneva.
3D apparel creation. Part II
319
The Emerald Research Register for this journal is available at www.emeraldinsight.com/researchregister
The current issue and full text archive of this journal is available at www.emeraldinsight.com/0955-6222.htm
IJCST 17,5
An experimental study of the factors affecting the indentation of fabric under a pinch gripper
320
H. Lin and P.M. Taylor
Received February 2004 Revised March 2005 Accepted March 2005
School of Mechanical and Systems Engineering, The University of Newcastle upon Tyne, Newcastle upon Tyne, UK, and
S.J. Bull School of Chemical and Advanced Materials, The University of Newcastle upon Tyne, Newcastle upon Tyne, UK Abstract Purpose – This paper presents an experimental study of the influence of variables such as strain rate, the number of fabric plies, the type of fabric, the kinds of fibre and the shape of indenter on the indentation of fabric under differently shaped pinch gripper. Design/methodology/approach – This experimental study will be approached from three different angles. It will look into an indenter pressing a sample with a much larger size, which is important in practice in the world of grasping by a pinch gripper. It will research a flat indenter, but also an indenter with a curved surface and will investigate fabric compression particularly with regard to the differences between single-layer and multi-layer stacks. Findings – The type of fabric architecture and the kind of fibre have been proven to be important for the indentation. Even more important is the indenter geometry. Evidence collected to date suggests that the grasping action is more sensitive to indenter geometry. This leads to three possible approaches: close regulation of the materials and processes, handling processes to change in the material properties, and thirdly, intelligent systems which can learn from and adapt to each situation. Research limitations/implications – This study suggests that a picking up operation should change in the material properties, that is, the operation should be controlled by using fabric characteristics as the control information in an intelligent environment. Originality/value – Previous work on compression has been concentrated on an indenter with a size identical to a specimen, this study will look into an indenter pressing a sample with a much larger size. On compression, previous work has focused on single-layer fabric compression by a flat indenter, but this research will not only research a flat indenter and single layers, but also an indenter with a curved surface, and multi-layer stacks. Keywords Clothing, Garment industry, Automation, Materials handling Paper type Research paper
International Journal of Clothing Science and Technology Vol. 17 No. 5, 2005 pp. 320-334 q Emerald Group Publishing Limited 0955-6222 DOI 10.1108/09556220510616183
Introduction The automation of garment manufacture is an important field of study in the textile industry. In this field, a major focus has been placed on materials handling, because it accounts for 80 per cent of the time needed to manufacture apparel (McWaters and Clapp, 1994). In an automated materials handling line, one of the most frequently occurring processes is the picking and placing of fabric panels. This can be carried out using pinch grippers, which comprise two pegs that push down on the top of the fabric.
The pegs are then brought together so that the fabric buckles up and is secured between them (Taylor et al., 1996). The implemention of intelligent of picking and placing of fabric in garment automation depends on a thorough understanding of the operation. For instance, (1) What is relationship between the performance of indenter, external load, deformation and the properties of material being handled? (2) What is happening within the material when it is being gripped? (3) How significant are the roles of indenter geometry and the material properties in the overall picking up action? (4) How much force should be applied to a fabric to achieve reliable gripping without damage? (5) How do different materials have different responses to the action? (6) How are the effects of variables such as the number of fabric layer, compression rate, etc. on this operation? (7) How to decide/adjust manipulation and control strategies for the action according to the behaviour of material in an intelligent environment of picking and placing of fabric panels? The first four questions can be answered by performing a mathematical analysis and studying the resulting internal stress and strain distributions, which has been completed (Lin et al., 2005a). The questions 2-5 can be answered by developing a numerical model for providing/identifying deformation profiles and stress fields, which also has been completed (Lin, 2002; Lin et al., 2005b). This study provides an experimental assessment of experimental variables effects on the operation. The last issue will be dealt with based on the results of questions 1-6. These efforts will be directed toward developing a theoretical foundation to the intelligent picking-up of fabric panels by a robotic gripper and the fundamental requirements of gripper design. This experimental study will be approached from three different angles to the literature. Previous work on compression has been concentrated on an indenter with a size identical to a specimen (Saunders et al., 1997, 1999a, b; Lekakou et al., 1996), this study will look into an indenter pressing a sample with a much larger size, which is important in practice in the world of grasping by a pinch gripper. On the other hand, previous work on compression has focused on fabric compression by a flat indenter (Saunders et al., 1997, 1999a, b; Lekakou et al., 1996), this research will not only research a flat indenter, but also an indenter with a curved surface. Furthermore, most of the previous work on compression has been done on single-layer fabrics (Matsudaira and Qin, 1995; de Jong et al., 1986; Taylor and Pollet, 2000), this research will investigate fabric compression particularly with regard to the differences between single-layer and multi-layer stacks, the latter is important in garment manufacture. Materials, indenters and experimental procedures Materials and indenters Three different types of weave: plain, twill and satin, and three different kinds of fibre: cotton, polyester and wool were used in the present study. Their specifications are
An experimental study
321
IJCST 17,5
322
given in Table I. All the samples have a size of 60 £ 40 mm: During the lay-up stage for multi-layer fabric testing, fabric samples were laid up by aligning warp-to-warp and weft-to-weft matched as best as possible from one layer to the next (Figure 1). All the fabrics used for the study are commercially produced and widely used in garments. A fresh sample was used for each test and all the samples were conditioned at 20 ^ 18C temperature and 65 ^ 3 per cent relative humidity for 48 h at least before testing. Two types of indenter were used for this study, i.e. a rectangular indenter with a size 60 £ 15 £ 10 mm and a cylindrical indenter 60 £ 15 mm (length, diameter). Experimental procedures Indentation tests were carried out on Instron Series IX Automated Materials Testing System 1.04. The indenter was attached to the cross-head. The pressure foot, i.e. the rectangular indenter is brought down to apply a pressure of 0.5 cN/cm2 on a fabric sample for 20 s, and the initial thickness of sample is measured under the initial pressure. The value of the thickness is also as an initial thickness for the cylindrical indenter testing. Then, the indenter is displaced downwards at a constant speed
Table I. Specifications of sample
Figure 1. A schematic of fabric compression by the rectangular indenter
Fabric code
Material
C1 C2 P1 P2 P3 W2
100 100 100 100 100 100
per per per per per per
cent cent cent cent cent cent
Structure cotton cotton polyester polyester polyester wool
Plain Twill Plain Twill Satin Twill
Area density (g/m2)
Thickness at 0.5 cN/cm2(mm)
Fabric count warp/weft (thread/cm)
173 517 59 199 130 208
0.63 1.034 0.11 0.68 0.38 0.52
16/13 19/17 15/15 12/11 26/22 14/13
0.5 mm/min (excluding from the testing the effect of the compression rate on the fabric behaviour) until the target load of 200 N is achieved. Results and analysis The influence of strain rate The first stage was to investigate the viscoelastic behaviour of fabric, for each of the examined three types of weave and three kinds of fibre, under different compression speeds in the range of 0.05-1 mm/min. Figure 2 shows the effects on assemblies of ten plies each of cotton twill, polyester twill and wool twill weave. The conclusion is that as the compression rate is reduced the compression curves are shifted to lower pressures and higher strains, which could be attributed to fibre slippage and frictional effects occurring during slow compression. It shows that the behaviour of fabric is time dependent and suggests that the influence of compression speed on multi-layer fabric compression strain should not be neglected, which is in contrast to the compression studies of Saunders et al. (1997, 1999a). Their experiment results indicated that changes in compression speed within the range of 0.05-1 mm/min show no significant effects on the compression data for assemblies of ten plies of cotton plain cloth. This inconsistency could be at least in part due to the differences in the experimental apparatus or methods. Comparing Figure 2(a) and (c) with (b), it can be found that the polyester fabric behaves in a more viscous manner than cotton fabric and wool fabric as expected. Elastic effects are more important than viscous effects in the compression of cotton and wool fabric assemblies. Relatively speaking, the compression speed has no significant effect on the compression of single-layer fabric in the both cases of the cylindrical indenter and the rectangular indenter, indicating elastic deformation is dominant in single-layer fabric compression (Figures 3 and 4). The difference of the influence of compression speed on the behaviour of fabric deformation between the single-layer and multi-layer fabrics is due to different deformation mechanisms between them. Nesting is the dominant factor in multi-layer fabric compression (Saunders et al., 1997) and this leads to more viscous deformation (Saunders et al., 1999b). Hence, multi-layer fabric behaves in more viscous manner. The favourable mode of compression in single-layer fabric is fibre and yarn deformation which includes a more significant elastic contribution, compared to nesting (Figure 5). The influence of number of plies Figure 6 shows the effect of the number of cloth layer on the compression of examined plain, twill and satin weaves by the rectangular indenter at a strain rate of 0.5 mm/min. No significant effects are observed for the number of plies greater than 2, which is in agreement with the compression studies of Saunders et al. (1999a). However, comparing the single-layer fabric with the multi-layer fabric compression, the compression curve of single-layer fabric moves to higher strain, indicating the difference of compression mechanisms between the single-layer and the multi-layer fabrics. The reason for higher compression load as the number of layer increases may be by the nesting between adjacent cloths. As the number of fabric in an assembly is raised higher than two nesting is dominant, which would give less strain for multi-layers.
An experimental study
323
IJCST 17,5
324
Figure 2. Effect of compression rate on the compression of ten plies twill fabric by the rectangular indenter: (a) cotton; (b) polyester; and (c) wool
An experimental study
325
Figure 3. Effect of compression speed on the compression of single-layer twill fabric by the rectangular indenter: (a) cotton; (b) polyester; and (c) wool
IJCST 17,5
326
Figure 4. Effect of compression rate on the compression of single-layer twill fabric by the cylindrical indenter: (a) cotton; (b) polyester; and (c) wool
An experimental study
327
Figure 5. A stack fabric compression by a cylindrical indenter illustration
In the case of the cylindrical indenter, the effects are present for n from 1 to 10 in Figure 7. As the number of layers is increased from 1 to 10, the compression curves move consecutively to lower strain, which is somewhat different from the case of rectangular indenter, where single-layer compression curve is separate from the multi-layer curves. There are two possible explanations for this difference. One could be the contact area of the indenter/sample changes with external load. Another one could be that the number of compressed fabric layers varies through the compressed region at a given load, such that more layers of fabric are compressed underneath the middle of the indenter than towards its the edges. Figure 8 shows this concept. In this figure, there are three layers of fabric are compressed at x ¼ 0; but only one layer of fabric is compressed at x ¼ 2: The influence of architecture of fabric During the compression of woven fabrics, various changes take place within their structure. From the viewpoint of fibres, bending of individual fibres, changes in the free distance between two contact points, and fibre-to-fibre slippage are the most important changes taking place during compression. From the viewpoint of structure, closer nesting, deformation of the yarn waveform, and compression of the yarn cross-section are the most important features. These changes depend on the properties of the fibre and the structure of the fabric. Figures 8 and 9 show the effect of the architecture of fabric on compressibility of cloths. In all the cases, i.e. single layer and multi-layer, the rectangular indenter and the cylindrical indenter, the conclusions are in agreement, which is also in agreement with the compression studies of Saunders et al. (1997). That is, the twill fabric shows very low compressibility, plain fabric has very high compressibility and the satin fabric shows intermediate compressional characteristics. In the case of the twill fabric, there is little chance of fibre-to-fibre slippage during compression owing to a relatively compact structure (Table I), resulting in low compressibility. Moreover, the twill has a non-symmetric and non-flat structure, and hence it and its assembly is the most difficult to compress. The plain weave examined is easily compressed at low
IJCST 17,5
328
Figure 6. Effect of the number of the plies on the compression of the fabric by the rectangular indenter: (a) polyester satin; and (b) wool twill
and intermediate pressures due to its high degree of crimp and the high possibility for nesting between layers for multi-layer fabric. Five harness satin weave has a flat structure and hence its assemblies can be compressed more easily than assemblies of twill but, on the other hand, a high number of ends and picks (Table I), and the lack of crimp and non symmetry minimize perfection of nest-fitting between layers, resulting in single- and multi-layer satin fabrics being more difficult to compress than plain weave cloths. However, in multi-layer fabric compression (Figure 9), the effect of the architecture of the fabric on the compressibility decreases as the pressure increases, because the amount of nesting between layers determined by the structure of the fabric is only dominant at low and intermediate pressure. When the assembly of cloths is indented, the fibre yarns are deformed by decreasing the amplitude of the yarn waveform and the fibres are compressed and deformed, so the properties of fibres and yarns become
An experimental study
329
Figure 7. Effect of the number of the plies on the compression of fabric by the cylindrical indenter: (a) polyester satin; and (b) wool twill
significant at high pressure. This result suggests that in fabric compression the interlacing of the yarns in the fabric is more important than the actual yarns from which the fabric was made. The influence of type of fibre Figure 10 gives the thickness-compression curves for the same fabric weave but for different types of fibres. These curves indicate the difference in compressibility between the different fibres for the same fabric architecture. The woollen fabric shows the maximum compressibility in the examined three kinds of fibres, followed by the polyester fabric and the cotton is the most difficult to compress as expected. Note the difference in the steepness of both loading and unloading curves between the different fibres. Cotton, c2, shows a steep compression curve and can, therefore, be regarded as a less compressible or harder material whereas, for example, the polyester, p2, and woollen fabric, w2, are soft and show a slowly increasing compression.
IJCST 17,5
330
Figure 8. A comparison between different types of polyester fabrics, namely plain, twill and 5 harness satin weave: (a) rectangular indenter; and (b) cylindrical indenter (number of plies ¼ 1)
Figure 9. A comparison between different types of weaves, namely plain, twill and 5 harness satin weave for polyester fabric 10 layers
An experimental study
331
Figure 10. A comparison between different types of fibre, namely cotton, polyester and wool (weave: twill, (a) 1 layer, (b) 10 layer, indenter: rectangular)
This is determined by the difference in their Young’s Modulus. The Young’s Modulus of cotton, polyester and wool fibres are 8.8, 7.3 and 3.0 N/tex, respectively (Morton and Hearle, 1975). The influence of shape of indenter Figure 11 shows that maximum compressive strain under the cylindrical indenter is much greater than that of the rectangular indenter for the same testing conditions, because the maximum pressure p0 of the cylindrical indenter is 3/2 times the mean pressure of rectangular indenter (Johnson, 1987). Furthermore, the contact region of the cylindrical indenter with the sample is smaller in the same test conditions (Figure 12). The influence of compression direction In the both cases of the rectangular indenter and the cylindrical indenter, the results of experiment show that the fabric has higher compressibility in the weft direction. This behaviour is determined by the basic structure of woven fabric, i.e. that weft yarn
IJCST 17,5
332 Figure 11. Polyester twill fabric 1 layer compression by the rectangular and the cylindrical indenter
Figure 12. The effect of the direction of indenter on fabric compression: (a) polyester twill 10 layers, rectangular indenter; and (b) polyester satin 10 layers, cylindrical indenter
usually has less twist and the fabric count in the weft direction is smaller than that in the warp direction (Table I). In other words, the woven fabric is usually loose in the weft direction, leading to easier compression in that direction. Conclusion Experimental research has been performed on single-layer and multi-layer fabrics compression using a rectangular indenter and a cylindrical indenter. When considering the main action of picking up a fabric panel in the interaction at the fabric/indenter interface, the properties of the fabric have been proven to be important for the operation. In general, the lower the Young’s Modulus of the fibre the higher its compressibility will be. If we, on the other hand, compare indenter orientation then it is noted that, for all kinds of samples used, alignment with the weft direction gives higher compressibility since fabric and yarns are normally looser in that direction. An even greater effect has been found when the type of fabric architecture is changed. For all fibres and the two types of indenter, the compressibility of plain fabric is highest, twill is lowest and intermediate for satin weave. Evidence for the effect of viscous slippage was found when the compression speed was changed, although more work is required to determine if this is significant at the real speed of the picking-up action. Tests have shown that in the range of 0.05-1.0 mm/min, the maximum difference in strain is 0.02 for multi-layer fabric. This variation in compressibility does not seem to occur for a single-layer fabric, and so it is postulated that effects are occurring in the nesting mechanism. The number of fabric plies in a multi-layer stack shows no significant effect on this operation. However, a different phenomenon is observed in load-strain curves from the two types of indenter. Much attention has been given to the influence of the shape of the indenter on the indentation action. For all fabrics, the compressibility by the cylindrical indenter is much higher in comparison to the rectangular indenter. The reason for this is thought to be the concentration of stress in surfaces compressed by a curved surface indenter. The effect of the indenter shape on the grasping action undoubtedly needs more attention, not only for fabric systems but for all other kinds of materials too. A study, which determines the shape influence on the grasping operation, would benefit both the garment automation and the gripper design world. Evidence collected to date suggests that the grasping action is very sensitive to the indenter geometry. More research on compression by a finger gripper (rounded and compliant) should hopefully give a better understanding of grasping mechanism of human hand. This study suggests that a picking up operation should change in the material properties, that is, the operation should be controlled by using fabric characteristics as the control information in an intelligent environment. References de Jong, S., Snaith, J.W. and Michie, N.A. (1986), “A mechanical model for the lateral compression of woven fabrics”, Textile Research Journal, Vol. 56 No. 12, pp. 759-67. Johnson, K.L. (1987), Contact Mechanics, Cambridge University Press, Cambridge, p. 100. Lekakou, C., Johari, M.A.K.B. and Bader, M.G. (1996), “Compressibility and flow permeability”, Polymer Composites, Vol. 17 No. 5, pp. 665-72. Lin, H. (2002), “Modelling the indentation of non-rigid materials by a pinch gripper”, PhD thesis, The University of Newcastle upon Tyne, Newcastle upon Tyne.
An experimental study
333
IJCST 17,5
334
Lin, H., Bull, S.J. and Taylor, P.M. (2005b), “Simulation of the deformation and stress distribution within a flexible material pressed by a pinch gripper”, Journal of Materials Processing Technology (in press). Lin, H., Taylor, P.M. and Bull, S.J. (2005a), “A mathematical model for grasping analysis of flexible materials”, Modelling and Simulation in Materials Science and Engineering, Vol. 13, pp. 185-201. McWaters, S.D. and Clapp, T.G. (1994), “Automated apparel processing ‘computer simulation of fabric deformation for the design of equipment’”, International Journal of Clothing Science and Technology, Vol. 6 No. 5, pp. 30-8. Matsudaira, M. and Qin, H. (1995), “Features and mechanical parameters of a fabric’s compressional property”, Journal of the Textile Institute, Vol. 86 No. 3, pp. 476-86. Morton, W.E. and Hearle, S. (1975), Physical Properties of Textile Fibres, 2nd ed., The Textile Institute, Manchester. Saunders, R.A., Lekakou, C. and Bader, M.G. (1997), “Compression and microstructure of fibre plain woven cloths in the processing of polymer composites”, Composites Part A, Vol. 29A, pp. 443-54. Saunders, R.A., Lekakou, C. and Bader, M.G. (1999a), “Compression in the processing of polymer composites 1. A mechanical and microstructural study for different glass fabric and resins”, Composites Science and Technology, Vol. 59, pp. 983-93. Saunders, R.A., Lekakou, C. and Bader, M.G. (1999b), “Compression in the processing of polymer composites 2. Modelling of the viscoelastic compression of resin-impregnated fibre networks”, Composites Science and Technology, Vol. 59, pp. 1483-94. Taylor, P.M. and Pollet, D.M. (2000), “A preliminary study of the low-load lateral impact compression of fabric”, International Journal of Clothing Science and Technology, Vol. 12 No. 1, pp. 12-25. Taylor, P.M., Pollet, D.M. and Grießer, M.T. (1996), “Pinching secure handling of fabric panels”, Assembly Automaton, Vol. 16 No. 3, pp. 16-21.
The Emerald Research Register for this journal is available at www.emeraldinsight.com/researchregister
The current issue and full text archive of this journal is available at www.emeraldinsight.com/0955-6222.htm
Learning-based fuzzy colour prediction system for more effective apparel design Chi-Leung Hui, Tak-Wah Lau and Sau-Fun Ng
Fuzzy colour prediction system 335
Institute of Textiles and Clothing, The Hong Kong Polytechnic University, HungHom, Hong Kong, People’s Republic of China, and
Chun-Chung Chan Department of Computing, The Hong Kong Polytechnic University, HungHom, Hong Kong, People’s Republic of China Abstract Purpose – This paper aims to design and develop a learning-based fuzzy colour prediction system for providing more effective apparel design in computer-aided design system. Design/methodology/approach – In this study, we propose using a fuzzy system integrated with preliminary knowledge of colour prediction for facilitating apparel design. The performance of the proposed system is evaluated in terms of its computational efficiency and robustness. In addition, the proposed system is evaluated by target group of customers. Findings – It was found that the performance of the proposed system is better than the traditional approach. Research limitations/implications – Although the proposed system has some limitations, the outcome of this study could be used to produce a future breakthrough in providing an intelligent computer-aided design system for apparel product. Originality/value – Using such an approach, an apparel designer could predict the favourite colours of garment for a target group of customers. The system uses preliminary knowledge about the customers’ profiles and evaluations. Such fuzzy approach for colour prediction is established, which is not used in a traditional way in apparel design. Keywords Computer aided design, Clothing, Control systems, Predictive process Paper type Research paper
1. Introduction In apparel design, the first element to grab the customer’s eye is the colour of a garment. Colour is the most fundamental fashion element. It is usually the first decision that a designer makes each season (Johnson and Moore, 2001). An apparel designer predicts colour trend for a target group of customers based on the general characteristics of the customers’ group such as gender, age, height, and skin colour. In a traditional process of colour prediction, an apparel designer relies on his/her sensations on the characteristics of a target group of customers. It is not an effective way because such prediction of colour trend for a target group of customers may produce a great error and it is very time-consuming. In order to improve the deficiency in such colour prediction of garment for a target group of customers, it is necessary to develop a learned-based intelligent system that helps an apparel designer to predict the colour in each season more effectively. This
International Journal of Clothing Science and Technology Vol. 17 No. 5, 2005 pp. 335-348 q Emerald Group Publishing Limited 0955-6222 DOI 10.1108/09556220510616192
IJCST 17,5
336
paper attempts to develop a learned-based intelligent system for such colour prediction based on customers’ profile so as to provide an effective way for apparel design. To perform learned-based colour prediction, a fuzzy intelligent system is used in this study. There are two main reasons for using a fuzzy system instead of another type of intelligent system, such as neural networks. First, colour perception cannot be clearly defined by the customer. No person can give the exact degree of preference for a particular colour. These preferences are usually expressed in general linguistic terms, for example, “like” or “dislike”. However, fuzzy mathematics has strong tools to convert linguistic variables into numerical variables. Secondly, most of the knowledge in colour prediction can be represented as input-output pairs. Fuzzy systems can also be integrated with optimisation tools, such as gradient-descent-based training (Wang, 1995), to match the input-output pairs. There are two obvious constraints that apply in using traditional fuzzy system. First, the computation preformed at the server should not be too complex, so that it can support as many users as possible. Secondly, to maintain sufficient accuracy, the number of iterations during training should not be too many. Normally, a fuzzy system is designed to be evolved from a blank structure using Gaussian membership functions. A long training time and complex computation are required. In this paper, an effective learning methodology, which uses supervised training on the preliminary knowledge of the customers’ profiles, is proposed. This methodology is used to speed up the processing performance on the server and to reduce the number of iterations required during training. The rest of this paper is presented as follows. Section 2 reviews the current literature on design of fuzzy system. Section 3 outlines the research methodology. Section 4 presents and discusses the results. Finally, Section 5 concludes this paper by describing the limitations of the current study and by providing several suggestions for future research in this area. 2. Literature review 2.1 Ways of designing fuzzy systems Fuzzy architecture describes a general technology. There is no systematic procedure to design fuzzy systems that has been met with wide acceptance and usage. Traditionally, design of a fuzzy system, like a fuzzy controller design for a rice cooker, is carried out using a time-consuming trial-and-error process (Baldwin, 1981). This process aims to incorporate expert knowledge into a fuzzy model. Some general approaches try to formulate an optimal fuzzy structure by using well-known mathematical techniques, such as gradient descent (Wang, 1995) and Kalman filtering (Simon, n.d.). Although different approaches are used to construct a fuzzy system, the evaluation of fuzzy systems is the same. The performance of a fuzzy system strongly depends on its robustness and computational efficiency in a changing environment. 2.2 Standard fuzzy system The structure of a standard fuzzy system (Wang, 1997) is shown in Figure 1. The central part of the fuzzy system is the fuzzy rule base. The fuzzy rules for colour prediction are collected and stored in the fuzzy rule base. The input to the system is a customer’s profile that we need to analyse. The profile is converted to a fuzzy set through the fuzzifier. The fuzzy input and the fuzzy rules are
compared in the fuzzy inference engine. If the differences between them are small, then the output follows the rules. Finally, the fuzzy output is then defuzzified using the defuzzifier, and a favourite colour for the customer is proposed. In our study, since the inputs and outputs are discrete in nature, we can use a common approach (Wang and Mendal, 1992) using a singleton fuzzifier, product inference engine and a maximum defuzzifier. Under this configuration, the input-output relationship can be written in the following simplified form. Suppose there are M possible colours for selection and N sections of user profiles to analyse. The predicted favourite colour is f ðxÞ ¼ y l* , where l * [ {1; 2; . . . ; M } in such a way that it has the largest membership value compared with the others. Or equivalently, N Y i¼1
mAl * ðxi Þ $ i
N Y i¼1
mAl ðxi Þ; i
;l [ {1; 2; . . . ; M };
Fuzzy colour prediction system 337
l * [ {1; 2; . . . ; M }
where mAl ðxÞ is the rule related to section i and the lth colour. Variable xi is the input i from the customer’s profile. 2.3 Gradient descent optimization In the development of the fuzzy system, the key issue is how to construct the rules in the fuzzy rule base. Some systems create the rules based on common sense. However, these rules are not optimal and the performance cannot be guaranteed. This has led to research on training the fuzzy rules. Gradient-descent optimisation (Wang, 1995) is one of the popular training algorithms and is used in this study because the computation is more simple compared with other existing algorithms. Gradient descent (Yang and Amari, 1998) is a function optimisation method that uses the first derivative of the function and the idea of steepest descent. The first derivative of a function is simply the gradient of the function at that point. Suppose that the gradient of a function is known, the value of the function will be decreasing in the negative direction of the gradient (i.e. 27FðxÞ). Gradient descent is an iterative method. The method used is described as follows: (1) Compute the derivative of the function with respect to its independent variables. The derivative is denoted as F 0 ðxÞ; where F is the function to be minimized, and x is the vector of independent variables. (2) Change the value of x as follows: x nþ1 ¼ x n 2 hF 0 ðx n Þ; where the subscript n refers to the iteration number, and h is a step size (which must be chosen so that it is neither too big nor too small). Too large a value for h will result in overshooting the function minimum, and too small a value will result in a long convergence time.
Figure 1. The structure of a standard fuzzy system
IJCST 17,5
338
(3) Repeat the above two steps until the procedure converges on the minimum of the function F(x). 2.4 Fuzzy system using gradient descent optimisation The standard set-up of fuzzy system using gradient descent optimisation is shown in Figure 2. Fuzzy system F0 is constructed using an arbitrary structure, for example, using a Gaussian membership function with height, attenuation and centre parameters. A predicted colour is presented in the output of F 0 . The error function, measuring the difference between the actual favourite colour and the estimated one, is computed from the evaluation process. Gradient descent training is applied to train the parameters of the Gaussian membership function to match the actual favourite colour. 3. Research methodology In this section, we describe the research methodology used in this study. It is divided into three stages. The first stage develops the preliminary knowledge of colour prediction. The next stage integrates the preliminary knowledge of colour prediction with the fuzzy system. The last stage involves carrying out performance analysis on the proposed system. 3.1 Development of the preliminary knowledge of colour prediction In the gradient descent fuzzy algorithm, the system is evolved from a blank initial structure. Knowledge is gained during the training process. The performance of the fuzzy system strongly depends on the initial parameter settings. If the wrong initial parameters are chosen, a poor fuzzy system can be developed. Looking at an analogy in the human world, this makes sense because a human also cannot do a good job if they learn badly. Learning from scratch seems to be inefficient in the view of system’s development constraints. Preliminary knowledge should be integrated in the system before it starts to learn. Again, looking at a human analogy, babies can cry and breathe without any learning. This knowledge is embedded in the baby’s brain before they start to learn. With the help of preliminary knowledge, a human can learn much faster and avoid unnecessary learning paths. Similarly, with the fuzzy system, training time can be reduced and some bad initial parameters can be avoided. This makes the system converge faster and makes the solution found more likely be the global minimum. To define preliminary knowledge on colour prediction, the factors significantly influencing the customers’ favourite colours are analysed in the following sub-sections.
Figure 2. Standard set-up of fuzzy system using gradient descent optimisation
3.1.1 Cultural factor. Cultural variations play an important role in influencing human colour preference. For example, British and Chinese have their own perceptions of colour. Lin et al. (2001) pointed out that these cultures have a close agreement in their two languages in terms of colour categories, but there is a large difference in the meaning associated with colours due to cultural differences. For example, white can create two extreme feelings in eastern and western countries. In Britain, white is commonly associated with new and pure, whereas in China, white is commonly associated with death and unhappiness. Because of cultural differences, it is impossible to define a unique preliminary knowledge model. However, we can define a unique model for people from the same cultural background. Therefore, the scope of this study is restricted to the Chinese culture. 3.1.2 Other factors influencing the colour preference. The outcome of specific research (Lam, 1987; Wan, 2000; Chu, 1992; Bohdanowicz, 1997) on the features of colour preference within the Chinese culture is summarized in Table I. There are some other factors, such as emotional (Lam, 1987), social relationship (Bohdanowicz, 1997) and habits (Bohdanowicz, 1997), which also influencing the favourite colour. After interviewing 125 Chinese people (75 males and 50 females) randomly chosen at five railway stations located in different districts, we found that six out of nine relevant factors influencing colour preference were statistically significant at 0.01 level (Table I).
Fuzzy colour prediction system 339
3.2 Integration of fuzzy system with preliminary knowledge Based on the significant factors analysed in Table I, the relationship between the significant factors and their colour preferences can be modelled by membership functions. These preliminary membership functions categorize the initial structure of the fuzzy system, and it is applicable for general customers in Chinese. For individuals, this structure is trained by using the evaluation process. A customer’s profile is combined with the preliminary knowledge in a fuzzy processing core to make a prediction. The system’s architecture, to perform fuzzy colour prediction, is shown in Figure 3. The dimension of the system is reduced because only feedback information is considered. Furthermore, gradient descent optimisation (Wang, 1995) is used in the training process to improve performance. As a result, the computational complexity can be reduced and acceptable accuracy can be achieved.
Factors
References cited
Seasonal Age Gender Height Skin colour Fashion Emotional Social relationship Habits
Lam (1987) Lam (1987) Wan (2000) Wan (2000) Chu (1992) Bohdanowicz (1997) Lam (1987) Bohdanowicz (1997) Bohdanowicz (1997)
Significant at 0.01 level p , 0.001 p , 0.001 p , 0.001 p , 0.01 p , 0.01 p , 0.001 p . 0.01 p . 0.01 p . 0.01
Table I. Analysis of the factors influencing the colour preference in Chinese culture
IJCST 17,5
340 Figure 3. Architecture of the fuzzy colour prediction system
3.2.1 Preliminary membership functions design. There are five steps to construct the preliminary membership functions based upon the results of interviewing 125 Chinese people (75 males and 50 females). (1) Define the centre and the height of the membership function for each of the areas of preliminary knowledge. As an example, the rule related to “Age” and “Red” colour preference is shown in Figure 4. Three fuzzy sets (children, young and old) are defined for the “Age” factor. The centre of the membership function is defined as the mean age of each fuzzy set. The mean is found using the following procedure. Suppose that there are M statistical results between certain ages and the relative degree for a group can be found: ½agei ; degðyoungÞi ;
;degð†Þ [ ½0; 1;
i ¼ 1; . . . ; M
The mean for each category (k people) is defined as the weighted average: M X
avageðyoungÞ ¼
agek degðyoungÞk
k¼1 M X k¼1
with the variance,
Figure 4. The fuzzy relation between the “Age” and “Red” colour preference
degðyoungÞk
M X
sageðyoungÞ ¼
jagek 2 avageðyoungÞ jdegðyoungÞk
k¼1 M X
Fuzzy colour prediction system
degðyoungÞk
k¼1
Similarly, the height of the membership function can be found by the above procedure. We collect the following samples using the results of interviews: ½colouri ; degðyoungÞij ;
;degð†Þ [ ½0; 1;
341
i ¼ 1; . . . ; N c ; j ¼ 1; . . . ; N s
The corresponding height is calculated as: Ns X
H colour_iðyoungÞ ¼
degðyoungÞj
j¼1
Ns
;
where Ns is number of samples for each colour, and Nc is the number of colours for prediction. The height represents the relative importance of that colour for each fuzzy group. The relative importance deg(†) is measured by five levels of colour preferences, from “dislike” to “like”, as shown in Figure 5. (2) Interpolate the samples by using triangular membership function. The heights calculated in step 1 are meaningful around the centre of their corresponding fuzzy sets. This relationship can be realized by interpolating the centre with a triangular membership function. The triangular membership function has a parameter yc denoting the centre and s denoting the spread. A smaller spread represents a smaller fuzzy variation. 8 < s2jy2yc j ; if y [ ½yc 2 s; yc þ s s uð y; yc ; sÞ ¼ : 0; if y [ ½yc 2 s; yc þ s In the example of the rule related to “Age” and “Red”: yc ¼ avageðyoungÞ ; s ¼ sageðyoungÞ : The interpolated result is shown in Figure 6. Notice that our design should cover all the input space. If there are some empty regions in the input space, we can define more triangular membership functions to enhance the resolution of the output space. For example, we define five fuzzy sets instead of defining three fuzzy sets for the “Age” factor. Other methods, such as extrapolating the rules from neighbourhoods (Wang and Mendal, 1992), are also possible if the number of statistical results is limited.
Figure 5. Five levels of colour preferences
IJCST 17,5
342
Figure 6. The interpolated fuzzy relation between the age and colour preference for the red colour
Figure 7. The preference output when input age is X *
(3) Construct the fuzzy system using preliminary knowledge. When the age of the customers is entered into the system, such as X*[1], the fuzzy system will compare the input with the fuzzy rules. The output is obtained by calculating the membership value. Figure 7 shows that the preference output is 0.5 when the input age is X *. (4) Repeat steps 1-3 for different colours. After step 4, the membership value for each colour has been calculated. In this study, the 216 colour indexes (216 Color Chart, n.d.) are chosen as the reference colours. These colours appear the same in different computer systems. An example of the membership function for colour preference for age X* using the 216 colour indexes is shown in Figure 8. (5) Construct preliminary membership function for a customer. After colour preferences for the six significant factors influencing the colour preference in Chinese (Table I) are computed for the six fuzzy systems, six sets of colour preferences are combined into one set of outputs in the fuzzy combination engine. The fuzzy intersection T-norm, the product operation in particular, is used to perform in this combination process. In Figure 9, the signal flow graph shows how the system obtains the preliminary membership function from a user profile. A sample preliminary membership function for a customer is shown in Figure 10. The membership function shown in Figure 10 is treated as the preliminary knowledge of a customer. Each customer has their own preliminary membership
function to summarize his or her favourite colours. The height of each colour component represents the degree of preference for a particular colour. With this membership function, we can simply apply the gradient-descent training algorithm to optimise the predicted output that can predict the customer’s favourite colour. The gradient descent algorithm is described in the next section. 3.2.2 Fuzzy learning through gradient descent optimisation. The structure of the proposed fuzzy prediction system is defined by the preliminary membership function. To have more robust control, feedback training is necessary to optimise the performance of the fuzzy system. Feedback knowledge is collected from evaluating the predicted colour. Gradient descent optimisation is used as the training strategy. During the evaluation procedure, the colour index with the largest membership value is proposed to the customer. Five degrees of preference evaluation from a scale of
Fuzzy colour prediction system 343
Figure 8. Preliminary membership function between colour preferences for age X * and 216 colour indexes
Figure 9. Signal flow graph of the preliminary membership function
Figure 10. Preliminary membership function for customer A
IJCST 17,5
“dislike” to “like” (Figure 5) are evaluated. The evaluation stops when the customer indicates a colour they “strongly like”. For each evaluation, the prediction error is calculated as follows. E colour_i ¼ absðevaluated valuecolour_i 2 membership valuecolour_i Þ
344
where i ¼ 1; 2; . . . ; 216: To minimize Ecolour_i, the gradient descent algorithm is used. Because the preliminary membership function is discrete in nature, differentiation is replaced by a difference equation that is equal to Ecolour_i. The training equation is obtained as follows. ðmembership valuecolour_i Þkþ1 ¼ ðmembership valuecolour_i Þk 2 hðmembership valuecolour_i Þk 2 ðevaluated resultcolour_i Þk where k ¼ people. The parameter h is the rate of the gradient descent. A larger rate will result in overshooting the local minimum. A smaller rate will result in a long convergence time. 3.3 Performance analysis A prototype of the system has been implemented based on the proposed approach. The system was implemented using Matlab 6.0. The colour chart (www.unisys.com/hw/ servers/clearpath/mcp/hlx6100spec.asp) was selected as the testing samples. An experiment was conducted to evaluate the proposed approach. The experimental set-up is described below. A popular server, the LX6100 series from the Unisys Company (www.unisys.com/ hw/servers/clearpath/mcp/hlx6100spec.asp), is taken as the reference platform in measuring the computational efficiency. The server specification is listed in Table II. One thousand five hundred and fifty Chinese people (800 males and 750 females) were randomly chosen at a railway station for this experiment. Each participant was asked to evaluate each predicted single colour in front of the same LCD monitor. This set-up can eliminate the effect of colour differences between different monitors. Participants were asked to rate the predicted single colour using a five-point scale, from “strongly dislike” to “strongly like”. The training is stopped when participants indicated a predicted colour is rated as “strongly like”.
Items Processor Memory
Table II. Server specification for the control experiment
OS supported Monitor
LX6100 server specification Two Intel Pentium III XEON 550 MHz with 2 MB Cache 256 MB minimum and up to 8 GB ECC PC100 SDRAM using sets of four 64, 128, 256, or 512 MB DIMMs. 16 Sockets Windows NT 2000 Advanced Server operating system Dell 1500 FP 15’ LCD resolution: 1024 £ 768
The processing time required in each prediction was counted and the number of training samples required for each participant was recorded. This experiment was repeated for three different training step sizes. The reason that we recorded the processing time is to ensure that our system is capable of working on a general server. The number of users supported at latency less than 3 s is calculated to measure the system’s capacity. The number of training samples required for each participant was recorded to measure the robustness of the system. By monitoring the number of training samples over time, we could analyse the learning capability of the fuzzy prediction system. The step size, h can be chosen as part of the system design. It is related to the prediction performance directly. In this experiment, we used three different step sizes, for h ¼ 0:1; 0.2 and 0.4, and the changes in performance were monitored.
Fuzzy colour prediction system 345
4. Experimental results The run-time performance of the fuzzy system mainly depends on the performance in the training phase. In this section, we focus on analysing the computational efficiency and robustness of the training phase. The average results are summarized in Table III. 4.1 Computational efficiency analysis Under the Matlab environment, the amount of computation time required is analysed in Table IV. As shown in Table IV, the processing time in the Matlab environment is reasonable. The time spent on accessing the customer database is longer than the time spent on a single iteration. However, for each login, the database is only accessed once. Therefore, the time spent on the algorithms iterations become the most critical area in the computational analysis. This encouraging result shows that our prediction scheme is capable of working on a general server. In the next step of this work, we optimised the code using Visual Cþ þ language. This language offers control in forming the data structures and compile-time
Step size (h) Average processing time to get and update users’ information Average processing time on each iterations Average number of iterations needed in first access Average number of iterations needed in second access Average number of iterations needed in third access
Average processing time to get and update user database Average processing time on each iterations Number of clock cycle required Number of users supported with latency ,3 s
0.1
0.2 0.4 s
0.4
30.41 18.47
0.01 s 20.54 14.55
27.58 21.85
12.24
13.47
17.55
Table III. Averaged results from the controlled experiment
0.010 s 0.01 * 550 M ¼ 5.5 M 300
Table IV. Computational analysis of the experiment
0.4 s
IJCST 17,5
346
optimisation that Matlab does not support. We estimate that the processing time can be reduced by a factor of 10-50. 4.2 Robustness The relationship between the number of training samples required and number of database accesses for h ¼ 0:1; 0.2 and 0.4 is shown in Figure 11. Using a small step size, for example 0.1, a large number of iterations are needed in the first two iterations. However, the reduction rate is much faster when using a larger step size. This is because the model is more accurate when using a fine step size. In contrast, a large step size, for example 0.4, gives poor performance in the number of iterations required for each access, since the local minimum is overshot frequently and it requires a long time to converge to make it stable. We conclude that the optimal step size is around 0.2. The number of training samples required is reducing through the knowledge accumulation. This demonstrates the learning capability of the proposed fuzzy colour prediction system. 4.3 Evaluation of the proposed system Five apparel designers were invited to test the prototype of proposed system. They were asked to input the characteristics of target group of customers in terms of six significant factors (Table I). Based upon the characteristics of target customers, 100 target customers were randomly selected at a railway station and, in average, 85 per cent of respondents rated the predicted single colour as “strongly like”. It showed that the proposed system is able to predict the customers’ favourite colours for apparel design. 5. Conclusion In this paper, we proposed a fuzzy approach for colour prediction in apparel design. Using such an approach, an apparel designer could predict the favourite colours of garment for a target group of customers. The system uses preliminary knowledge
Figure 11. Relationship between the number of trainings samples required and number of database accesses for h ¼ 0:1; 0.2 and 0.4
about the customers’ profiles and evaluations. Such fuzzy approach for colour prediction is established, which is not used in a traditional way in apparel design. A prototype system has been developed based on the proposed approach. A controlled experiment was conducted using six significant factors (analysed Table I). Results show that the proposed system is able to predict the customers’ favourite colours for apparel design efficiently. The proposed system only applies to single-colour prediction case. There is also a limitation on the colour inconsistency for different cathode ray tubes (CRT) displays (Cui et al., 2000). This causes the evaluation result with a lot of variation. To tackle this problem, Finlayson et al. (1999) proposed a methodology to unify the colours displayed in CRT using feedback. Song and Luo (2000) also derived a colour difference formula for CRT displays. Our future work includes developing a better preliminary knowledge representation and more sophisticated algorithms for learning and applying the knowledge to multi-colour prediction case. Minimum involvement is required from the customers to train the system. For a wide range of applications, colour inconsistency is another issue that deserves future study. Note 1. It might make more sense to use a numerical value in this example, rather than use a mathematical symbol. References Baldwin, J.F. (1981), “Fuzzy logic and fuzzy reasoning”, in Mamdani, E.H. and Gaines, B.R. (Eds), Fuzzy Reasoning and Its Applications, Academic Press, London. Bohdanowicz, J. (1997), Fashion Selling, Ng Lam. Chu, L.Y. (1992), Let You Get More Attractive Colour, Kai Shing. 216 Color Chart (n.d.), available at: www.exoticblades.com/tamara/colorchart.html (accessed 14 March 2002). Cui, G., Luo, M.R. and Li, W. (2000), “Colour-difference evaluation using CRT colours”, paper presented at Colour and Visual Scales Conference, National Physical Laboratory, 3-5 April. Finlayson, G.D., Hordley, S. and Hubel, P.M. (1999), “Unifying colour constancy”, paper presented at The 7th Color Imaging Conference, IS&T and SID , pp. 120-6. Johnson, M.J. and Moore, E.C. (2001), Apparel Product Development, 2nd ed., Prentice-Hall, Upper Saddle River, NJ. Lam, M.C. (1987), Plan for Colour, Arts Publisher. Lin, H., Luo, M.R., MacDonald, L.W. and Tarrant, A.W.S. (2001), “A cross-cultural colour-naming study. Part I: using an unconstrained method”, Color Research and Application, Vol. 26 No. 1, pp. 40-60. Simon, D. (n.d.), “Kalman filtering for fuzzy discrete time dynamic systems”, submitted for publication. Song, T. and Luo, M.R. (2000), “Testing colour difference formulae on complex images using a CRT monitor”, The 8th Color Imaging Conference Proceedings, The Society for Imaging Science and Technology & Society for Information Display, 7-10 November , pp. 44-8. Wan, S. (2000), Colour Image Chart, Lung Tai Publisher.
Fuzzy colour prediction system 347
IJCST 17,5
348
Wang, L.X. (1995), “Analysis and design of fuzzy identifiers of nonlinear dynamic systems”, IEEE Transactions on Automatic Control, Vol. 40 No. 1, pp. 11-23. Wang, L-X. (1997), A Cause in Fuzzy System and Control, Prentice-Hall, Englewoods Cliffs, NJ. Wang, L.X. and Mendal, J.M. (1992), “Generate fuzzy rules by learning from examples”, IEEE Trans. on Systems, Man, and Cybern., Vol. 22 No. 6, pp. 1414-27. Yang, H.H. and Amari, S. (1998), “The efficiency and the robustness of natural gradient descent learning rule”, in Jordan, M.J., Kearns, M. and Solla, S.A. (Eds), Advances in Neural Information Processing Systems, MIT Press, Cambridge, MA, pp. 385-91.
The Emerald Research Register for this journal is available at www.emeraldinsight.com/researchregister
The current issue and full text archive of this journal is available at www.emeraldinsight.com/0955-6222.htm
COMMUNICATIONS
A new way of understanding the customer for fibre manufacturers Vicky Lofthouse and Tracy Bhamra
Fibre manufacturers
349
Department of Design and Technology, Loughborough University, Leicestershire, UK, and
Tom Burrow Tencel Ltd, Spondon, Derby, UK Abstract Purpose – This paper describes the novel approach taken in a collaborative research project that aimed to investigate new ways of understanding the customer, for Derby-based fibre manufacturer, Tencel Limited. The overall aim of the research described in this paper, was to help identify and establish a significant retail programme with a major UK store group for Tencel limited. Design/methodology/approach – In an iterative process, the target customer for the focus groups was identified, the main aims of the process were discussed, the test garments were identified and the empathic design tools were adapted. The team developed a programme of activities that would capture customer focused information on these critical issues. Findings – Using the Grove techniques helped to make the project transparent and inclusive, and enabled the whole team to be involved in the decision-making process. Using these techniques have provided Tencel with a non-scientific way of understanding how their end customer perceives their fibre, providing unequivocal evidence of the customer true feelings recorded as “raw” video based evidence. Practical implications – The Tencel/IF project has also led to a number of additional advantages for the company, such as the development of new relationships within the supply chain, the development of new relationships within Tencel and the enhancement of multi-disciplinary team working and concurrent engineering. Originality/value – This paper has presented the novel approach that the combined Tencel/IF team took to develop a better understanding of the end customer and illustrates how techniques which were developed for one industry can be successfully adapted and applied to a quite different industry with excellent results. Keywords Customer satisfaction, Research work, Textile industry, United Kingdom Paper type Case study
1. Introduction This paper describes the novel approach taken in a collaborative research project that aimed to investigate new ways of understanding the customer, for Derby-based fibre manufacturer, Tencel Limited. The project was carried out between September 2001 and October 2002 as part of the Textile & Clothing Industry Forum. This initiative The authors would like to extend their thanks to Jenny Cater, Denise Lofthouse and Chris Dowie for their help in setting up the focus groups in Essex, Leicester and York. They would also like to thank their colleagues at Tencel; Linda Mears, Patricia Manning, Jim Taylor, Anne Whineray, Karen Cooper and Caroline Lavery, for their commitment to the project, and their unwavering enthusiasm to succeed.
International Journal of Clothing Science and Technology Vol. 17 No. 5, 2005 pp. 349-360 q Emerald Group Publishing Limited 0955-6222 DOI 10.1108/09556220510616200
IJCST 17,5
350
funded by the UK Department of Trade & Industry (DTI) and companies from the textile and clothing industry, aimed to demonstrate and facilitate the delivery of profitable supply chains in partnership with UK manufacturers and retailers, and to develop the processes for sustaining competitive advantage (Fashion Industry Forum, 2002). The remit of the Industry Forum (IF) is to organise “live” projects, which are carried out within the workplace with participating retailers and suppliers. One of the specific aims of IF is to take technologies and approaches which have a proven track record in other industries (such as automotive and electronic product manufacturing) and adapt them for use within the textiles and clothing industry (Bhamra et al., 1998). The overall aim of the research described in this paper, was to help identify and establish a significant retail programme with a major UK store group for Tencel limited. In 1978, Courtaulds Fibres began research into solvent spun cellulosic fibre technology in response to predicted market changes. The resulting fibre TENCELw, a registered trademark of Tencel Limited for Lyocell (Tencel Ltd, 2002), was launched in 1992 claiming to be the first new fibre to be introduced on the market for 30 years (Curtis, 1997). Tencel Limited is currently the world’s leading supplier of Lyocell, with an 80 per cent share of the market and an annual turnover of approximately 100 million Euros. Over the last 10 years a new fibre variant has been developed – Tencel A100, ideally suited for use in knitted applications. The new fibre has proven to be very successful in jersey fabrics but has only achieved success in limited geographical markets for knitwear, and developments in western markets particularly have been slow. Initial research through IF suggested that problems with the perception of the Tencel fibre existed at several levels within the supply chain. Within processing companies, it was perceived that there would be technical difficulties regarding processing the fibre into yarn and this was a valid reason for them not to “push” A100 through the supply chain. At the retail level of the supply chain there was a perception that garments manufactured from Tencel would be more expensive to produce than products made from equivalent yarns (Lofthouse and Thomas, 2002). In order to move beyond the problems outlined above, Tencel Limited and the IF team set out to gain an understanding of the end customer’s (i.e. the consumer) perception of A100, based on the belief that this would provide a convincing package of evidence to illustrate to buyers that customers wanted garments knitted in this fibre variant. It was felt that strong support from the end-user would help to “pull” the fibre through the supply chain. This paper describes the approach adopted in order to gain an understanding of the end customers’ perception of Tencel’s product and how this knowledge was used to create a package of evidence that could be used to demonstrate the positive features of their product. 2. Moving beyond the traditional In deciding to “speak” to the end customer, Tencel were making a radical move and breaking away from the traditional structure of the apparel supply chain. Figure 1 shows this change, showing the traditional approach to the textiles supply chain which relied on yarn being “pushed” through via the fibre manufacturers. This is compared to the new approach adopted in this project which set out to empower Tencel to utilise their customer understanding to “pull” the yarn along the supply chain. It was believed that this would allow them to gain confidence in their downstream customers and demonstrate that they have a product that satisfies an end-user need.
Setting out to better understand the end customer was an unaccustomed approach for the technical team of a fibre manufacturer that traditionally focused on science and technology rather than the “touchy feely” side of human emotions. In the traditional set-up a fibre supplier like Tencel would only speak to the manufacturer of the textile product, and these sorts of conversations would have a very technical/scientific basis. To be able to effectively communicate with the end customer (non-technical shoppers) the Tencel team would have to learn a whole new vocabulary and develop new ways of talking about knitwear features. In order to support this investigation, tools and techniques which were developed to investigate emotional responses to design features in the automotive industry, were adapted for use within the textiles industry (Burns et al., 1999, 2000; Burns and Evans 2000, 2002. These tools are described in more detail in Section 4. In addition to these data collection techniques, an approach developed by The Grove Consultants International (2003) was adapted to help with the management of this complex and intricate project. The Grove Consultants International is a process consulting firm that has pioneered the use of visual approaches to help facilitate strategic visioning in collaborative work environments. Their techniques aim to stimulate participation, encourage teams to think about the “big picture” and enhance group learning and memory (The Grove Consultants International, 2001). This paper outlines the tools which were used throughout the Tencel/IF collaboration and illustrate how they benefited the project team and helped Tencel to generate their first insight into the perceptions that their end customers have of their product offering.
Fibre manufacturers
351
3. Managing the project The Tencel/IF project was carried out by a multidisciplinary team, located at three different sites in the UK. It consisted of two members of the IF team, a project leader who managed the project and a research assistant with comprehensive experience of the textiles industry and of running focus group activities. These team members were based at Cranfield University (now at Loughborough University). The multidisciplinary Tencel team consisted of the fabric development manager, the knit specialist, the retail account manager, the dyeing and finishing development manager and the support unit manager. The majority of the Tencel team were based at the Spondon site in Derbyshire, with the exception of the retail account manager who was based in their London offices.
Figure 1. Supply chain diagram
IJCST 17,5
352
Figure 2. The Tencel “Journey Diagram”
The Tencel/IF project set out to provide the Tencel team with the ability to be able to carry out similar projects in the future, rather than provide them with a project consultant. As a result of this the whole team made decisions on every aspect of the project, and were engaged in the development of the focus group tools. Although the IF team members often presented proposals and ideas for activities, these were always discussed and manipulated by the whole team to ensure that the best possible result was obtained. For this reason each member of the team had a degree of responsibility and consequently a stake in the success of the project. In order to achieve a high degree of team working, manage the project effectively across the different geographical locations and make effective use of the different expertise available, the first of the “new” tools used within the project was introduced at the start of the project. A visual communication tool, adapted from a template entitled “Your Company’s Gameplan” designed by The Grove Consultants International (2003) was used to plan the project and record the progress of the team. The adapted template was known as the “Journey Diagram” and consisted of four main areas. A section on the left hand side of the diagram entitled “Resources” recorded the team members names along with any external resources that could be drawn upon. The main section, which was illustrated by a large blue arrow that indicated the direction of travel, recorded the deliverables and tasks which needed to be completed. A large yellow target on the right hand side of the diagram reminded the team of the ultimate aim of the project and helped to reduce deviation from the correct goal. At the bottom of the diagram green hills provided the team with an area in which to record the challenges that faced them during the project. A photograph of Tencel’s “Journey Diagram” being used as a work in progress is shown in Figure 2. The “Journey Diagram” was used to record all of the activities that the team identified were necessary to progress the project towards the final target. Tasks were discussed at regular team meetings, recorded on individual Post-it Notes and stuck onto the plan. The flexibility of this approach meant that activities could be moved around and reordered until the team were happy. At each meeting, the Journey Diagram was fixed to the wall and used for reference and further planning. Activities which should have been completed were reported on and ticked off when they were finished with. Future activities were identified and assigned an owner and a completion date.
Throughout the project, the template was used during meetings and then photographed and distributed to the team shortly after each meeting, along with comprehensive minutes which recorded all major decisions and the reasoning behind them. The “Journey Diagram” provided many benefits which helped in generating the successful outcome of the project. In the early stages of the project, using the template helped to draw the new team members together and provide them with a common focus. As time progressed, it helped to ensure each team member was aware of the activities that needed to be carried out, when they needed to be done and who was responsible for them. At one stage during the project a new team member joined the group and it was an easy task to brief her on the progress that had been made, the direction in which the project was headed and the reasoning behind the decisions which had been made. The template provided an effective way of making everyone accountable for their tasks in a non-confrontational and supportive way. This approach made the project very transparent and helped to ensure there were no miss-understandings within the team, which might delay its progress and threaten the tight deadline, it also enabled each team member to have a high level of project ownership, which was rewarded by commitment and enthusiasm throughout the project. In addition to this, using the template allowed the team to visually map the project over the year timescale, enabling them to record an end date and work backwards, hence ensuring there was enough time available to complete the required activities. This was crucial to the success of the project, due to the short timescales in which we were trying to fit the later activities of the project. Finally, the visual nature of the template meant that the team could easily see the project progressing and celebrate success when the milestones were reached. It was within this context and structure that the project was carried out. 4. Understanding the customer As was outlined in the introduction, collecting customer focused data was a new experience for the fibre manufacturer. In order to support this activity, the team drew from the experiences of the researchers at Cranfield University (Evans et al., 2002) who had developed empathic design tools in order to better facilitate the collection of customer focused information for Nissan. The resulting tools are “participatory, in-depth and qualitative in nature” (Burns and Evans, 2002). Many of the tools aim to provide designers with the opportunity to “walk in the shoes of those they are designing for . . . to equip designers with the previously uncaptured customer information that they need to stimulate innovation, to differentiate their product and delight their customers” (Burns and Evans, 2002). In order to support this textiles project, these tools were either adapted or used as a basis for the development of new techniques, suitable for garments. 4.1 Setting up the focus groups In an iterative process, the target customer for the focus groups was identified, the main aims of the process were discussed, the test garments were identified and the empathic design tools were adapted. The pilot study demonstrated that it was important to identify a specific customer base to focus on, to ensure that the tools which were to be developed were appropriate to that sector. After some initial investigations the Tencel/IF team selected a female
Fibre manufacturers
353
IJCST 17,5
354
customer group as their target audience. A profile of this female customer was drawn up by the retail account manager and used as a basis for consideration during the development of the tools (Table I). In order to ensure that the activities in the focus groups generated information on the most important and appropriate areas for Tencel, the team spent some time identifying the critical areas of interest, these included: . customer perceptions of the Tencel garment on first sight; . customer perceptions of the Tencel garments in comparison with cotton; . customer perceptions of the comfort for the garments – short-and medium-term; and . customer perceptions of the wash performance of the Tencel garments. Through a series of discussions it was decided that the focus groups would be used to test 100 per cent Tencel A100 fibre variant against 100 per cent cotton. It was felt that if the results showed that A100 was as good as cotton with similar qualities then this would provide a very strong message for Tencel. Forty test garments in 100 per cent cotton and forty test garments in 100 per cent A100 were produced by a local knitwear company. Building from the findings from a pilot study, a plain black, round neck garment was developed as the test piece. 4.2 The focus group activities and adapted tools Based on the interests identified above, the team developed a programme of activities that would capture customer focused information on these critical issues. Two sets of focus group activities were developed, that would be carried out with three different sets of female’ customers of the type described above, within a 2-week period. Each focus group had an agenda of activities that the participants were aware of (Figure 3) into which adaptations of empathic design tools and techniques were “woven” to achieve the desired result. The focus groups were held in the evenings and were designed around a novel “Cocktails and Cake” theme, with the intention of them feeling very feminine, luxurious and fun. The meeting rooms were decorated with flowers and the participants were treated like ladies and provided with a selection of wines and beautiful cakes during the evening, to maintain a “special” feeling. This theme evolved in response to profile data that was provided on the female customer group. Although there were some structured activities the majority of the activities were designed to flow and encourage conversation, rather than force it. The focus groups had a relaxed and “chatty” air and often felt more like a party than a work group.
Table I. The characteristics of the type of customer selected for the Tencel trials
Female Average age: 30-60 years Middle/upper class Professional/middle management Family oriented Travels
Gardens Into cooking Self assured Traditional/loyal Classic look in her approach to fashion Environmentally aware
Fibre manufacturers
355
Figure 3. Outline of the programme of activities involved in the Tencel project
To ensure that all opinions and feelings were captured, recording equipment was used throughout the focus groups. Although the ladies were made aware of the camera and could see it through out the focus group, they often forgot about the camera and talked candidly to each other about their true feelings. This had the added benefit of providing Tencel with hard hitting evidence of exactly what the customer thought of the product – “warts and all”. It has been seen in the past (Burns et al., 1999) and again in this project that giving this type of evidence to a project team has a very powerful effect as it shows the team exactly what the customers think, rather than a diluted version. 4.2.1 The Kano questionnaire. At the beginning of the first focus group each of the participants were given a short questionnaire to fill in. The questionnaire which aimed to gather more information about the customer experience, was based on the Kano model (Matzler et al., 1996; Matzler and Hinterhuber, 1998) an approach which provides a way of thinking about products in terms of customer needs. Through the questionnaire we asked participants to imagine that they had recently bought a knitwear garment and to consider how happy they were with the purchase, we asked them questions about various aspects of the whole life cycle of the garment in a positive way – if it had the feature, and in a negative way – if the feature was missing. The aim of this approach is to pinpoint the ladies’ exact attitude to the aspect concerned. This formal activity was placed at the beginning of the evening so that it would not break up the flow of activities later on. 4.2.2 A gift wrapped present. After the questionnaire the flow and pace of the evening changed completely and all participants were given their Tencel garment wrapped up in a beautiful package (Plate 1). This activity was intended to provide Tencel with information on the customers “first impressions” and was designed to replicate the participants receiving a present. How did they feel about their gift?
IJCST 17,5
356
Plate 1. Opening the gift wrapped Tencel garments – “First impressions count”
The initial reaction was considered to be very important to capture, as this is what sales are based on in the retail environment. The approach used at this stage in the focus group drew from an empathic design technique called “Scenario of Use” which aims to provide insights into how the customer thinks, feels and talks about products, by replicating specific situations that put the customer in the environment that is to be discussed. During this activity another technique known as “Murmur of the Customer” was used. The “Murmur of the Customer” is a technique which captures what the customers actually think about products rather than what they tell market researchers or sales staff. In order to capture this “murmur” at various stages during the activities, for example, after the present had been unwrapped, the organisers left the room under the guise of collecting coffee, leaving the participants to chat about the garments on their own. At this stage the conversation always become more animated and candid, and the video equipment ensured that these valuable insights into their needs, wants and opinions were captured. This process ensured that comments that would not otherwise have been mentioned due to the participants “being too polite” were captured and could subsequently be used to improve and develop the product more accurately. 4.2.3 Comparing Tencel A100 and cotton. The next activity gave Tencel the opportunity to see how their garment compared with an identical garment knitted in cotton. The participants were given the cotton garment and asked to compare the two. They were told that each garment was the same style, the same colour and that both garments were knitted out of the same weight fibre, and dyed with the same quantity of dye. Initially during this activity the participants focused on the colour comparisons between the two. After a while they were asked to close their eyes and then talk out the
difference in the garments. This was to encourage them to think about senses other than sight. Again the “Murmur of the Customer” technique was used and the participants were left to talk in “private”. The whole activity was recorded for analysis later on. 4.2.4 Washing and wearing. At the end of the first focus group the participants were introduced to the wash wear activity. The aim was that between the focus group sessions the participants would take away their new garments for two weeks and treat them as they would normally. They were asked to wash and wear them as they wanted, but to ensure that each garment was worn and washed at least once. This activity again drew on the empathic design technique of “Scenario of Use”. It aimed to gain an insight into the customer’s experiences of owning the product. The value of using this type of technique is that it enables the product developer to gain an understanding of how the actually customer feels about the product and how it actually performs, rather than making assumptions. This approach to assessing the wash/wear ability of the garment was very different to the traditional scientific approach which was usually taken, which involves systematically and repeatedly washing the garment using a specific washing powder and then drying it under scientific conditions. During the two week period the participants kept a diary of their experiences, they recorded how comfortable the garments were to wear, how they wore, how they felt after washing, how they dried and any other thoughts and feelings. “Diary Keeping” is a type of self-administered questionnaire again used within empathic design. This approach enabled substantial amounts of data to be generated with minimal amount of effort on the part of the organisers. At the second focus group, the same participants were encouraged to talk about their experiences of owning the different products. They were encouraged to talk in groups and were subtly led through a series of specific areas to concentrate on. Again the activities were video recorded and the “Murmur of the Customer” technique was used to ensure that candid opinions were captured. At the second set of focus group the results of traditional scientific wash/wear trial, that had been run in parallel with the consumer trials, were presented to the ladies. Again their responses were recorded. This provided a very useful insight into the issues that captured the imagination of the ladies. 4.3 Results analysis A huge range of findings in a variety of different forms (e.g. questionnaires, video data, diaries) were generated as a result of the Tencel/IF project. In order to identify the main findings from the project and present them to the rest of the research and development team at Tencel, all the findings were analysed using a coding and clustering methodology common for qualitative data (Miles and Huberman, 1994; Silverman, 2000; Yin, 1993). Utilising and adapting the empathic design tools and applying them to the focus group situation enabled a huge range of rich qualitative data to be collected from all participants. They provided detailed descriptions of their first opinions of the garments, using descriptive language to express how they felt about the feel and texture of the A100 garment, such as “lovely soft feel”, “lustrous”, “it feels like you are putting on a cuddle”. They also provided detailed opinions about the garment’s drape,
Fibre manufacturers
357
IJCST 17,5
358
colour, how little it creased and the overall appearance of the garment that they were given. This level of detail and use of descriptive language was repeated for the cotton garment. For the wash/wear trial the participants provided the team with detailed information on how they washed and dried their garments and whether they referred to care labels. They also provided an insight into the performance of both garments during and after washing, in terms of texture/feel, colour performance, shrinking, stretching and creasing. Finally they provided the Tencel team with a view of the “owning experience”, providing detail on their first impressions, fit and ease of wear on both first and subsequent wearing. The main findings resulting from the analysis of the focus groups were collated into a PowerPoint presentation, and presented to an audience of representatives from the Tencel product development team which included the Head of Tencel’s Research and Development. Throughout the presentation, critical issues were illustrated with extracts of film, thus allowing the R&D team to hear the true “Voice of the Customer”. Being able to highlight the points that were made by illustrating them with video clips from the workshops, provided a very valuable way of demonstrating to the team exactly what the ladies opinions were, both positive and negative. This provide to be an extremely important exercise for the team, providing them with a real “pat on the back” when positive comments were made and honestly highlighting where improvements were needed “straight from the horses mouth”. The results from the project have been extremely valuable to Tencel, and are currently being fed into their research and development process. 5. Conclusions This paper has presented the novel approach that the combined Tencel/IF team took to develop a better understanding of the end customer and illustrates how techniques which were developed for one industry can be successfully adapted and applied to a quite different industry with excellent results. Using the Grove techniques helped to make the project transparent and inclusive, and enabled the whole team to be involved in the decision-making process. As a result of this approach, the Tencel team have a pack of tools and activities that they had been involved in developing, and the confidence to carry out future projects on their own. Drawing on the empathic design research (Evans et al., 2002) has added both interest and value to the data collection work, made it fun for the participants and enabled rich/genuine information to be collected. Using these techniques have provided Tencel with a non-scientific way of understanding how their end customer perceives their fibre, providing unequivocal evidence of the customer true feelings recorded as “raw” video based evidence. This powerful evidence means that Tencel can further strengthen their research and development programme based on an understanding derived from the “Voice of the Customer” rather than on preconceived, scientific ideas. The Tencel/IF project has also led to a number of additional advantages for the company, such as the development of new relationships within the supply chain, the development of new relationships within Tencel and the enhancement of multi-disciplinary team working and concurrent engineering. Tencel have recognised that whilst their primary focus is to develop and manufacture fibre, there is still incredible value to be gained from talking to the end customer, and because customers
think and behave very differently to scientists and engineers it is imperative to understand what is important to them from their perspective. This project has shown that the ability to identify customer delighters not only provides an excellent and powerful marketing message but also provides sound development support for R&D. The powerful video evidence has provided the Tencel team with a mechanism for communication which can be used to make the senior management sit up and take notice of the need to understand customers. As a consequence, a continuation project was started to work with the manufacturer who had produced the garments used with the focus groups. The target was to develop a Tencel knitwear product which could be launched on the market. Unfortunately, the manufacturer was experiencing financial difficulties and was closed by its owners shortly after the completion of the project. Hence it is not possible to point to a success that is a direct result of the project. But the methodology, the results from the focus groups and the message about the value of listening to the ultimate consumer have become engrained in the Tencel business and have been used with other customers, both retailers and knitwear manufacturers, around the world. As a footnote, Tencel has recently been taken over by its main competitor, Lenzing. The new enlarged Lenzing business will have access to the results of this work to help develop the market for Tencel fibre. References Bhamra, T., Heeley, J. and Tyler, D. (1998), “A cross sectoral approach to new product development”, The Design Journal, Vol. 1 No. 3, pp. 2-16. Burns, A. and Evans, S. (2000), “Insights into customer delight”, in Scrivener, S.A.R., Ball, L.J. and Woodcock, A. (Eds), Collaborative Design, Chapter 29, Springer, London. Burns, A. and Evans, S. (2002), “Empathic design: a new approach for understanding & delighting customers”, International Journal of New Product Development and Innovation Management, Vol. 3 No. 4. Burns, A., Barrett, R., Evans, S. and Johansson, C. (1999), “Delighting customers through empathic design”, paper presented at the 6th International Product Development Management Conference, 5-6 July, Cambridge. Burns, A., Evans, S., Johansson, C. and Barrett, R. (2000), “An investigation of customer delight during product evaluation”, paper presented at the 7th International Product Development Management Conference, 29-30 May, Leuven. Curtis, H. (1997), “Implementation of environmentally conscious design in courtaulds fibres: tencel”, MSc Research Project, Textiles/Fashion. 1997, Manchester Metropolitan University, Manchester, p. 160. Evans, S., Burns, A. and Barrett, R. (2002) in Lofthouse, V.A. (Ed.), Empathic Design Tutor, Cranfield University, Cranfield. Fashion Industry Forum (2002), “Website of the fashion Industry Forum”, available at: www. industryforum.net (accessed September 2002). Lofthouse, V.A. and Thomas, M. (2002), Understanding Customer Perceptions in Knitwear, Cranfield University, Cranfield, pp. 1-61. Matzler, K. and Hinterhuber, H.H. (1998), “How to make product development projects more successful by integrating Kano’s model of customer satisfaction into quality function deployment”, Technovation, Vol. 18 No. 1, pp. 25-38.
Fibre manufacturers
359
IJCST 17,5
360
Matzler, K., Hinterhuber, H.H, Bailom, F. and Sauerwein (1996), “How to delight your customers”, Journal of Product & Brand Management, Vol. 5 No. 2, pp. 6-18. Miles, M.B. and Huberman, A.M. (1994), An Expanded Sourcebook – Qualitative Data Analysis, 2nd ed., Sage, Thousand oaks, CA. Silverman, D. (2000), Doing Qualitative Research: A Practical Handbook, Sage, London. Tencel Ltd (2002), Experience More. Experience TENCELw, Tencel Ltd, available at: www.tencel. com/ (accessed 25th June 2003). The Grove Consultants International (2001), “A graphic facilitation retrospective: charting what we learned”, available at: www.grove.com/new/new_gfretro.html (accessed June 2001). The Grove Consultants International (2003), The Grove Consultants International, available at: www.grove.com (accessed 18 June 2003). Yin, R.K. (1993), “Applications of case study research”, Applied Social Research Methods Series, Sage, London.
International Journal of Clothing Science and Technology
ISSN 0955-6222 Volume 17 Number 6 2005
International textile and clothing research register Editor-in-Chief Professor George K. Stylios
Access this journal online __________________________
2
Editorial advisory board ___________________________
3
Editorial __________________________________________
4
Research register __________________________________
6
Research index by institution_______________________
65
Research index by country _________________________
66
Research index by subject__________________________
67
Research index by principal investigator ____________
68
Access this journal electronically The current and past volumes of this journal are available at:
www.emeraldinsight.com/0955-6222.htm You can also search over 100 additional Emerald journals in Emerald Fulltext at:
www.emeraldinsight.com/ft See page following contents for full details of what your access includes.
CONTENTS
www.emeraldinsight.com/ijcst.htm As a subscriber to this journal, you can benefit from instant, electronic access to this title via Emerald Fulltext and Emerald Management Xtra. Your access includes a variety of features that increase the value of your journal subscription.
Additional complementary services available
How to access this journal electronically
E-mail alert services These services allow you to be kept up to date with the latest additions to the journal via e-mail, as soon as new material enters the database. Further information about the services available can be found at www.emeraldinsight.com/alerts
To benefit from electronic access to this journal you first need to register via the Internet. Registration is simple and full instructions are available online at www.emeraldinsight.com/admin Once registration is completed, your institution will have instant access to all articles through the journal’s Table of Contents page at www.emeraldinsight.com/0955-6222.htm More information about the journal is also available at www.emeraldinsight.com/ ijcst.htm Our liberal institution-wide licence allows everyone within your institution to access your journal electronically, making your subscription more cost effective. Our Web site has been designed to provide you with a comprehensive, simple system that needs only minimum administration. Access is available via IP authentication or username and password.
Your access includes a variety of features that add to the functionality and value of your journal subscription:
Research register A Web-based research forum that provides insider information on research activity world-wide located at www.emeraldinsight.com/researchregister You can also register your research activity here. User services Comprehensive librarian and user toolkits have been created to help you get the most from your journal subscription. For further information about what is available visit www.emeraldinsight.com/usagetoolkit
Key features of Emerald electronic journals Automatic permission to make up to 25 copies of individual articles This facility can be used for training purposes, course notes, seminars etc. This only applies to articles of which Emerald owns copyright. For further details visit www.emeraldinsight.com/ copyright Online publishing and archiving As well as current volumes of the journal, you can also gain access to past volumes on the Internet via Emerald Fulltext and Emerald Management Xtra. You can browse or search these databases for relevant articles. Non-article content Material in our journals such as product information, industry trends, company news, conferences, etc. is available online and can be accessed by users. Reference linking Direct links from the journal article references to abstracts of the most influential articles cited. Where possible, this link is to the full text of the article. E-mail an article Allows users to e-mail links to relevant and interesting articles to another computer for later use, reference or printing purposes. Emerald structured abstracts New for 2005, Emerald structured abstracts provide consistent, clear and informative summaries of the content of the articles, allowing faster evaluation of papers.
Choice of access Electronic access to this journal is available via a number of channels. Our Web site www.emeraldinsight.com is the recommended means of electronic access, as it provides fully searchable and value added access to the complete content of the journal. However, you can also access and search the article content of this journal through the following journal delivery services: EBSCOHost Electronic Journals Service ejournals.ebsco.com Huber E-Journals e-journals.hanshuber.com/english/index.htm Informatics J-Gate www.j-gate.informindia.co.in Ingenta www.ingenta.com Minerva Electronic Online Services www.minerva.at OCLC FirstSearch www.oclc.org/firstsearch SilverLinker www.ovid.com SwetsWise www.swetswise.com TDnet www.tdnet.com
Emerald Customer Support For customer support and technical help contact: E-mail
[email protected] Web www.emeraldinsight.com/customercharter Tel +44 (0) 1274 785278 Fax +44 (0) 1274 785204
EDITORIAL ADVISORY BOARD Professor Mario De Araujo Minho University, Portugal Professor H.J. Barndt Philadelphia College of Textiles & Science, Philadelphia, USA Professor Dexiu Fan China Textile University, Shanghai, China Professor B. Knez University of Zagreb, Croatia
Professor Masako Niwa Nara Women’s University, Nara, Japan Professor Issac Porat School of Textiles, UMIST, UK Professor Ron Postle The University of New South Wales, Australia
Editorial advisory board
3
Professor Rosham Shishoo Swedish Institute for Fibre and Polymer Research, Mo¨lndal, Sweden
Professor Carl A. Lawrence University of Leeds, Leeds, UK
Professor Paul Taylor University of Newcastle, Newcastle upon Tyne, UK
Professor Trevor J. Little North Carolina State University, USA
Professor Witold Zurek Ło´dz´ Technical University, Poland
Professor David Lloyd University of Bradford, Bradford, UK
International Journal of Clothing Science and Technology Vol. 17 No. 6, 2005 p. 3 # Emerald Group Publishing Limited 0955-6222
IJCST 17,6
4
International Journal of Clothing Science and Technology Vol. 17 No. 6, 2005 pp. 4-5 Emerald Group Publishing Limited 0955-6222
Editorial Championing the research efforts of the community The International Textile and Clothing Research Register (ITCRR) is in its 11th year of publishing the research efforts of our community. It provides a breadth of activity in the field of textile and clothing research and it encourages participation and dissemination to those working in this discipline and further a field. Again as you will see in this new edition, textile and clothing research and practice is increasing in volume, in quality and in diversity, all good news for all of us involved in it. Research, development and innovation can, without doubt, give us more wisdom, enable our industries to become more competitive, and contribute to our quality of life. I believe that registering research projects will provide the due credit to originators of the research and contribute much more to the future development of this field. Groups of expertise can be identified in this manner, repetition and reinvention can be avoided, leading to best utilisation of time and funding for faster and better directed research in the international arena, since globalisation is on everybody’s agenda. The ITCRR has been set up with all these things in mind. Textiles and clothing originate from the physiological need to protect ourselves from the environment. This has made necessary the art of hand knitting and weaving, and cut and sew processes, which have been evolving for many centuries. Although the original need for clothing has somewhat changed, mechanisation of this process started after the industrial revolution and has continued this century, with automation developments on a massive scale. Considering the upstream part of the whole textile and clothing production chain, yarn-making is the most highly automated area, followed by fabric, with high speed knitting and weaving machines. The downstream part of garment making, however, still remains probably the less developed connection in this chain; one that no doubt many of us have our eyes on as the candidate for development into the new century. With massive computerisation over the last 20 years, logistics and sales have also changed dramatically from pen and paper to electronic data interchange and electronic point of sale. New challenges are already upon us with nanotextiles; nanofibres nanocoatings, with multifunctional and smart textiles and clothing, and with wearable electronics. This year in particular we have welcome contributions from textile and clothing aesthetics, design and fashion. It is our view that in our field design and technology go hand to hand, and we predict that more exciting projects will come from this synergy. Consistent and extensive research and development in textile and clothing design and science and technology underpin all these developments by the international research community, whether in educational establishments, in research trade organisations, or in companies. IJCST has been set up 16 years ago, as a platform for the promotion of scientific and technical research at an international level. With the statement that the manufacture of clothing in particular needs to change to more technologically advanced forms of manufacturing and retailing, IJCST continues to support the community in these and other efforts. The journal continues with its authoritative style to accredit original technical research. The refereeing process will continue and will try to reduce waiting time during the refereeing process as much as q Professor George K. Stylios
possible. IJCST will be instrumental in supporting conferences and meetings from around the world in its effort to promote the science and technology of clothing. I finally praise the enthusiasm of our research community and those authors that have made IJCST an invaluable resource to all involved with textiles and clothing. I thank our editorial board for their continuous support and our colleagues who have acted in a refereeing capacity, with commitment to progress our research efforts. I take the liberty to list some of those names below (apologies in advance if anyone has accidentally been omitted from this list): . Professor Paul Taylor, University of Newcastle; . Professor Haruki Imaoka, Nara Women’s University; . Professor Mario De Araujo, University of Minho; . Professor H.J. Barndt, Philadelphia College of Textiles and Science; . Professor Masako Niwa, Nara Women’s University; . Professor Jachym Novak, Vysoka Skola Sronjni a Tectilffi; . Professor Isaac Porat, UMIST; . Professor Roy R. Leitch, Heriot-Watt University; . Professor Ron Postle, The University of New South Wales; . Professor Gordon Wray, Loughborough University of Technology; . Dr Taoruan Wan, University of Bradford; . Professor David Lloyd, University of Bradford; . Professor G.A.V. Leaf, Heriot-Watt University; . Dr David Brook, University of Leeds; . Dr C Iype, University of Leeds; . Dr Jaffer Amirbayat, UMIST; . Dr Norman Powell, Leeds Metropolitan University; . Dr David Tyler, Manchester Metropolitan University; . Dr Jintu Fan, Hong Kong Polytechnic University; . Dr Lubas Hes, University of Minho; . Dr Jelka Gersak, University of Maribor; . Dr Han Fan, Heriot Watt University; . Dr Hua Lin, Heriot Watt University; and . Dr Sharon_po_tang, Heriot Watt University. The Editor-in-Chief thank all subscribers, authors, editorial board members, referees, publishing team, Ms Anita Hill, colleagues and students for their support. Address for correspondence: Heriot-Watt University, School of Textiles, Netherdale, Galashiels, Selkirkshire, TD1 3HF, Scotland, UK. E-mail:
[email protected]. George K. Stylios Editor-in-Chief
Editorial
5
Research register
IJCST 17,6
Arad, Romania 6
AUREL VLAICU University, Bd. Revolutiei 77, 310130 Arad, Romania, Tel: 040257283010; Fax: 040257280070; E-mail:
[email protected] Principal Investigator(s):Prof. Michaela Dina Stanescu PhD, Department of Chemical and Biological Sciences Research Staff:Lect. Silvia Mihuta PhD, PhD Student Andreia Stanislav, PhD Student Mihaela Dochia
Biotechnology for textile waste waters Other Partners: Maribor University Slovenia Ass.Prof. Vania Kokol
Academic
Industrial Textile factories from the region
None None Project started: September 2005 Project ends: September 2007 Grant value: 20000 Euro Source of support: Ministry of Education from Romania and Slovenia Publications Not Available.
Athens, Greece National Technical University of Athens, Iroon Polytechniou 9, Zografos Campus, GR-15773 Athens, Greece, Tel:+302107721520; Fax: +302107722347; E-mail:
[email protected] Principal Investigator(s): Prof. Christopher Provatidis, Lect. Savvas Vassiliadis, School of Mechanical Engineering Research Staff: Dipl. Eng. Dimitris Venetsanos, Lect. Kleanthis Prekas, Dipl. Eng. Ioannis Koukoulis
Numerical modelling and simulation of textile materials Other partners: Academic University of Maribor, Faculty of Mechanical Engineering
Industrial None
Project started: 1 July 2003 Project ends: 31 October 2005 Grant value: 11.740 Euro Source of support: General Secretariat of Research and Technology-GR Keywords: Objective measurements, Mechanical model, Textile fabrics The textile materials have a complex structure. The structural complexity of the fabrics imposes relevant complexity in the study of their mechanical properties. The proposed project deals with the low-stress analysis and simulation of the mechanical behavior of the fabrics. The special loading conditions refer to the characteristic “handle” of the fabrics, as well as to other properties like sewability of fabrics etc. In order to create the suitable micromechanical models of the fabrics, it is necessary to analyze the structural characteristics of the fabrics and to study their microstructure from the geometrical point of view. Based on that, the parametric micromechanical computational models will be compared and crosscorrelated. A series of special tensile, shearing, compression, bending, geometrical roughness etc laboratory tests will be performed in both standard Kawabata Evaluation System for Fabrics as well as in the automated version of it. The obtained measurements, will be compared and they will be used for the evaluation of the performance of the computational models. Finally the model correction actions are foreseen, so that the maximum possible convergence between the experimental and the computational results will be achieved.
Project aims and objectives Textile fabrics will be tested using the Kawabata Evaluation System for Fabrics in both Standard and Automated version. The results will be compared and correlated. The mechanical data will be used for the evaluation of the performance of numerical models of fabrics created using Finite Elements.
Research deliverables (academic and industrial) Correlation of the measurements of the Standard and Automated version of the KES-F. Evaluation of FEM models of textile fabrics. Publication “Contact mechanics in the two-dimensional finite element modelling of fabrics”, IJCST
Athens, Greece Technological Education Institute of Piraeus, P. Ralli & Thivon 250, GR-12244, Athens, Greece, Tel: +302105381224; Fax: +302105450965; E-mail:
[email protected] Principal Investigator(s): Prof. Maria Rangoussi, Prof. Christopher Provatidis, Lect. Savvas Vassiliadis, Department of Electronics Research Staff: Lect. Kleanthis Prekas, Prof. Thanos Peppas, Ass. Prof. Anastasios Delopoulos
Research register
7
IJCST 17,6
8
Electrically conductive textiles Other partners: Academic
Industrial
National Technical University of Athens Aristoteles University of Thessaloniki Project started: 1 January 2004 Project ends: 30 June 2006 Grant value: 60.000 Euro Source of support: Ministry of Education GR Keyword: Electrically conductive textiles The project focuses on the properties of the textile electrically conductive materials. In principle the textile fibres made from natural or man-made polymers are electrical isolators. The electrically conductive fibres are made through adding metallic materials in the man-made polymer structures. They combine finally the character of the textile fibres with the specific electrical conductivity. The electrical and mechanical properties of textile yarns made of man-made electrically conductive fibres will be studied taking in account the influence of the mixing and spinning parameters. The electrically conductive yarns will be used for the production of textile fabrics with special characteristics in terms of Joule effects, as well as of their shielding behaviour in electromagnetic fields. The deformation of the yarns in the fabric structure, which is expected to change the initial electrical properties of the yarns, will be studied and it will be modeled. In parallel the weaving parameters will be correlated to the final electrical characteristics of the fabrics. Electrical and mechanical tests are planned in order to prove the suitability of the end products in various applications.
Project aims and objectives The project aim is to define the parameters for the production of electrically conductive fabrics dedicated for specific uses under controlled conditions to meet specific needs and standards. The objectives are the thorough study of the textile and electrical properties of the fibres, yarns and fabrics taking in account the changes of the behaviour of the materials occurring when complex structures are developed.
Research deliverables (academic and industrial) A thorough study of the influence of the textile structures and the deformations on the electrical properties of the conductive textiles. Production of sample fabrics for the investigation of the suitability in specific enduses. Publications Not available.
Athens, Greece Technological Education Institute of Piraeus, P. Ralli & Thivon 250, GR-12244, Athens, Greece. Tel: +302105381224; Fax: +302105450965; E-mail:
[email protected] Principal Investigator(s): Savvas Vassiliadis, Thanos Peppas, Anthony Primentas, Department of Electronics, Department of Textiles
Strengthening university-industry links in Uzbekistan Other partners: Academic
Industrial
TecMinho Associacao Universidade Tashkent Institute of Textile Empresa para o Disenvolvimento and Light Industry Namangan Engineering – Economic Institute Logotech S.A. Project started: 2005 Project ends: 2008 Grant value: 400.000 e Source of support: European Commission, Directorate General Education and Culture Keywords: Textiles, Industrial liaison office, Distance learning One of the main objectives of the project is the establishment and the operation of the Industrial Liaison Office. It deals with the technology transfer between universities and industries mainly in the textile sector, one of the strongest industrial sectors of Uzbekistan. Another objective is the organization of distance learning courses. They will support the open and life-long courses to the personnel of the industries and to the general public as well. The project will support the design and the realization of the distance learning structure.
Project aims and objectives Establishment of the Industrial Liaison Offices and the enhancement of the links between industry and universities. Organization of the distance learning courses according to the needs and the future of the textile sector in Uzbekistan.
Research deliverables (academic and industrial) A network of industries and universities for the technology transfer between them. A distance learning organization supporting the life long learning needs in Uzbekistan. Publications Not available.
Research register
9
IJCST 17,6
10
Bornova I˙zmir, Turkey Faculty of Engineering, Textile Engineering Department, Ege University, TR-35100 Bornova I˙zmir, Turkey. Tel: +902323399222; Fax: +902323399222; E-mail:
[email protected] Principal Investigator(s): Prof. Dr Kerim Duran, Ege University, Faculty of Engineering, Textile Engineering Department, Sabanci University, Faculty of Engineering and Natural Sciences Research Staff: Assoc. Prof. Dr Ays¸ egu¨l Ekmekc¸˙I Ko¨rlu¨, PhD. O.Ug˘ur Sezerman, M. ˙Ibrahim Bahtiyari, Gu¨nseli Bayram
Preventing pilling problem of cellulosic fabrics with different types of enzymes and comparison of their effects Other partners: Academic
Industrial
Sabanci University, Faculty of Engineering and Natural Sciences Ege University, Faculty of Engineering, Textile Engineering Department Project started: 1 August 2003 Project ends: 1 June 2005 Grant value: USA 52 500 $ Source of support: TU¨BI˙TAK TAM Keywords: Viscose, Cellulase, Biopolishing, Enzyme modification In the first part of the project the effect of commercial cellulases on to the bio-polishing of viscose based fabrics was investigated as reported in previous registration form (2004). However, despite the loss of tensile strength, the sufficient pilling values could not be obtained with the use of these commercial cellulases in bio-polishing of viscose fabrics and different commercial cellulases in different enzyme components results slight differences in pilling values after bio-polishing. So, in order to minimize tensile strength loss and improve pilling values after bio-polishing, enlarging the cellulases by a modification method was carried out. For this purpose, cellulase molecules were crosslinked chemically by using a bifunctional linker like glutaraldehyde. Then the effect of crosslinked enzyme was determined with evaluating the pilling tendency and bursting strength of fabrics. The project was completed successfully.
Project aims and objectives Despite the loss of tensile strength, the sufficient pilling values could not be obtained with the use of commercial cellulases in bio-polishing of viscose fabrics. The aim of this project was to overcome this problem with modified cellulase enzymes or with the use of feasible blend of cellulase components.
Research deliverables (academic and industrial) Crosslinked cellulase enzymes cause less bursting strength losses ( , 10 per cent) with 1 point better pilling values than those obtained with the use of commercial cellulase. Publications “Treatment of viscose with crosslinked commercial cellulases.” paper presented at the 3rd International Conference on Textile Biotechnology, Graz, 13-16 June. “Use of cellulase-Jeffamine networks in place of the free enzyme may alleviate the problem of lost fiber strength during textile processing” paper presented at the 229th ACS National Meeting, San Diego, CA, 13-17 March. “Effect of crosslinked and commercial cellulases on bio-polishing of viscose based knitted fabrics” paper presented at the 5th AUTEX World Textile Conference, Portoroz, 27-29 June.
Bucharest, Romania The Research-Development National Institute for Textile and Leather, Lucretiu Patrascanu Street, no. 16, sector 3, Bucharest, 030508, Romania, Tel: (0040) 21-340.49.28, (0040) 21-340.42.00; Fax: (0040) 21-340.55.15, E-mail:
[email protected] Principal Investigator(s):Romanian Research and Education Ministry, The Department of Laboratory Apparatus and Equipment Meant for the Textile Industry Research Staff:Ion Mituleasa, Engineer, Emilia Visileanu, Doctor Engineer, Daniela Isar, Senior Researcher
Functional model stand – for determining the filter parameters and testing the quality of the textile filter cartridge Other partners: Academic
Industrial
Polytechnic University Bucharest None Project started : 14 October 2002 Project ended: 30 May 2004 Grant value: 21.100 EUR Source of support: Governmental Budget: 74 percent, The Research-Development National Institute for Textile and Leather: 26 percent Keywords: Textile filter cartridge, Filtration, Quality, Stand The characteristic parameters of the filters which define the particle retaining capacity in a fluid, are made evident by: the absolute rating, representing the diameter of the largest particle passing through the filter, expressed in mm, the nominal filter rating, an abstract value representing the retaining percentage in the contaminants mass; the mean filter rating, a measure of the average size of the filtering element pore; the filter
Research register
11
IJCST 17,6
12
rating; beta(b) ratio which states the relation between the number of particles upstream, larger than a specific size and the number of particles downstream, which is bigger than a specified size. The functional test stand allows testing the textile filter cartridge having a retaining capacity of the impurities between 5 and 100 mm, making possible to determine the drop pressure and the maximum capacity that may penetrate the filter for a certain preset drop pressure. The filtering system may be: galvanizing liquors, electrolytic solutions, organic solvents, various industrial emulsions, mineral and vegetal oils, photographic solutions, detergents, lacquers and dyes, water cosmetics, solutions used in the food or in the medicine industry, etc.
Project aims and objectives The project is aiming to designing, performing and experimenting of a functional StandModel for testing textile filter cartridges, in the purpose of quality verifying and determining the filter parameters
Research deliverables (academic and industrial) .
Project and stand functional model for verifying the textile filter cartridges quality.
.
Testing methods for determining pressure fall and filter rating.
Publication Textile Industry Magazine, No. 1, 2005, p. 24
Bucharest, Romania The Research Development National Institute for Textile and Leather (INCDTP), Lucretiu Patrascanu Street, 16, Bucharest, Romania, Tel: 004 021 3404200, Fax: 004 021 3405515, E-mail:
[email protected] Principal Investigator(s): Dr eng. Emilia Visileanu, Dr eng. Iuliana Dumitrescu, Department for Products testing, Controlling and Notifying Research Staff:Eng. Maranda Erdes, Eng. Alina Popescu, Phys. Marilena Niculescu
Development of methodology for determining heavy metals on textile materials and leather articles in view of harmonization with the EU regulated frame Other Partners: Academic UPB INCDTP Production Project Stared: 14 October 2002
Industrial SC. Interclan Prod SRL SC. Antilopa S A SC. Flaros S A Project ended: 30 March 2004
Grant value: 625.000.000 lei Source of support: Ministry of Education and Research Keywords: Heavy metals, Detection limits, Procedures, Validation The process of Romania integration in EU, imposes, on the other hand, the harmonization of Romania legislation with the European one in various fields. The most of the companies worldwide forbids the marketing of textile and leather materials that contain, over a certain limit, pesticides, fungicides, insecticides, heavy metals, azo and allergen dyes that by reduction, they generate cancerigen, formaldehyde, phthalates. Due to the complex problems implicated in the identifying of these compounds, there is no standard recognized worldwide for their determining on textile and leather materials, each company specifying a certain methodology, equipment, depending on the destination of the final product. The international projects in the last years aimed to establish faster and more sensitive methods for determining heavy metals, especially from water, air, soil by means of immunologic reagents. The absence of some international standards is determined by the fact that the tests results are strongly influenced by numerous factors among which the most important are methods of compounds extraction and purification and sensitiveness of detection apparatus. The methods will be validated on various textile and leather materials, finished with various chemical auxiliaries, supplied by co-financing partners interested in the existence in the country of a lab capable to release ecological bulletins necessary for exporting the goods.
Project aims and objectives The aim of the research work is the elaboration of methodologies for identifying heavy metals existing on textile and leather materials in view of protecting human health and environment. The major objectives are: a) obtaining information on the chemical-physical characteristics, the processes by which they reach the textile materials, interactions with the other existing components, methods for qualitative and quantitative Identifying, effects on human body; b) development of analytic procedures; c) validation of analytic procedure.
Research deliverables (academic and industrial) The research results were: .
creation of operational procedures for determining heavy materials on textile materials
.
development of specification material for making analyses for spectrophotometry of atomic absorption there were compared the parameters achieved for the established method with the requirements of EU Directive 178 and the German standard DIN 53314 the work method was validated
.
.
Publication “Methods for determining heavy metals on textile and leather materials”, Industria Textila, Vol. 55, No. 3, 2004.
Research register
13
IJCST 17,6
14
Bucharest, Romania The Research Development National Institute for Textile and Leather, 16 Lucretiu Patrascanu Street, cod 030508, Sector 3, Bucharest, Romania, Tel: 0040-21-3404200 ext. 115, Fax: 0040-21-3405515, E-mail:
[email protected] Principal Investigator(s):Eng. Alina Popescu, Candidate for a Doctor’s Degree, Textile Chemical Processing and Environmental Protection Department Research Staff:Dr Eng. Emilia Visileanu, Prof. Dr Eng. Micaela Dina Stanescu, Eng. Georgeta Vulpe, Candidate for a Doctor’s Degree, Dr Eng. Magdalena Fogorasi
Development of biotechnologies for processing of textile materials made from natural fibers in order to reduce the environmental pollution Other Partners: Academic
Industrial
Aurel Vlaicu University, Arad SC CARPATEX SA, Brasov Project started: September 2002 Project ended: June 2004 Grant value: Source of support: Ministry of Education and Research, Romania Keywords: Enzymes, Pectinase, Amylase, Protease, Lipase, Cotton bio-scouring, Anti-felting of wool, Raw wool bio-scouring Within the present project, in conformity with the project activity schedule, a detailed documentary study regarding the present state in the field of textile materials enzymatic treating technologies was presented with reference to: enzymes specific structure and action, enzymes defining characteristics and factors of influence over their action, mechanisms of the enzyme-textile substrate interaction. There were elaborated and experimented biotechnologies for treating cotton and wool textile materials, namely: .
Enzymatic procedure for cotton fabrics bio-scouring – an alternative to the classic procedure of alkaline boiling.
.
Enzymatic procedure for cotton fabrics de-sizing – bio-scouring in simultaneous phase.
.
Procedure for cotton fabrics bio-scouring – bleaching in simultaneous phase.
.
Enzymatic procedure for anti-felting treatment of wool fabrics.
.
Enzymatic procedure for raw wool scouring.
Achieved biotechnologies were presented and demonstrated at the co-financing partners of the project.
Results obtained within the project were disseminated through articles published in the specialty literature (2 articles), papers and posters presented with the occasion of scientific national and international events (11 presentations). Information linked to this project are available on the web site of The ResearchDevelopment National Institute for Textile and Leather – Bucharest, at the following address www.certex.ro
Project aims and objectives .
Properties study and highlight for various enzyme types with applicability in finishing natural fibers textile materials.
.
Study through modern spectroscopy and physical-chemical methods for the interaction mechanisms of various enzyme types with natural fibers textile materials.
.
Elaboration and realization of bio-technologies for the treatment of textile materials with natural fibers content, in view of obtaining multiple finishing effects.
.
Settling of optimal application conditions for the bio-preparations with different support specificity.
.
Experimentation and quantification (through physical-chemical, physicalmechanical, scanning microscopy methods) of obtained technical performances and of effects induced to treated textile materials, for each enzyme type product/application field. Identification of novel elements for the realized bio-technologies and the elaboration of invention patents documents. Delivery of project results in national symposiums, round tables organized within Colorists and Chemists Society from Romania, meetings of workgroups organized within COST Action 847, COST Action 628 and their publishing in specialty magazines.
.
.
. .
Design and realization of WEB page. Demonstration of functionality for elaborated bio-technologies at co-financing partners.
Research deliverables (academic and industrial) Eco-friendly procedure for preliminary preparation with enzyme products of the cotton textile materials, is accomplished through exhaustion or padding-storing, with an aqueous solution containing 0.025-3 g/l alkaline pectinase, 1-5 ml/l non-ionic tensionactive product poly-glycol ether-based derived from greasy alcohol, with or without adding 0.15-0.2 g/l EDTA complexation product, with or without addition of 0.5-2 g/l a-amilase from non-pathogenic micro-organisms, pH ¼ 7-9, incubation temperature 55-65 C, for 30-60 min Advantages: .
Decrease of oxygen chemical consumption (COD) from waste waters by over 40 percent, of oxygen bio-chemical consumption (BOD) with over 40 percent and of total sediment with over 80 percent.
Research register
15
IJCST 17,6
.
Reduction of energy consumption by reducing reaction temperature and processing time.
.
Reduction of water consumption with over 25 percent.
.
Reduction of waste waters cleaning costs by eliminating NaOH from the process.
.
If processing cotton blends with other fibers like polyester or wool, degradation of the other components is not generated due to the action specificity of pectinolythic enzyme and to mild treatment conditions.
16
Eco-friendly procedure for raw wool scouring with alkaline lipase encapsulated in polyethylene-glycol, is carried out in basins containing aqueous solutions as follows: . basin 1: 0.5 g/l tension-active product based on greasy alcohol, 2 g/l Na2CO3, pH ¼ 9.5-10, T ¼ 45 C .
basin 2: 0-2.5 g/l tension-active product based on greasy alcohol, 0-2 g/l Na2CO3, 0-4 g/l lipase, pH ¼ 9.5-10, T ¼ 50-52 C
.
basin 3: 4 g/l lipase, pH ¼ 9.5, T ¼ 50-52 C
.
basin 4: 0-4 g/l lipase, pH ¼ 8-9.5, T ¼ 45-50 C
.
basin 5: water at the temperature of 35-45 C
.
basin 6: water at the temperature of 30-35 C
Advantages .
Efficient and gently removal of greasy substances during scouring, confirmed by low values of greasy substances content, of solubility in alkali, and through Pauly test.
.
Decrease of chemical oxygen consumption (COD) and bio-chemical oxygen consumption (BOD) existent in the wastewater with over 40 percent, of the content in substances extractible in petroleum ether with over 60 percent.
.
Increase of added value for the textile materials obtained out of wool fibers enzymatically scoured.
Eco-friendly procedure for anti-felting of wool is accomplished through exhaustion in jet type apparatus with an aqueous solution containing 0.5-2 g/l proteolitic enzyme, 0.5-1 g/l non-ionic tension-active product, pH: 8-9, incubation temperature 70 C, for 30-60 min. Advantages .
felting capacity reduction for 100 percent wool textile materials; reduction of pilling formation; improving of tinctorial affinity; soft handle;
.
mantaining of natural character of wool;
.
decrease of bio-chemical oxygen demand (BOD) with approx. 30 percent, and the concentration of total organic substances existent in the wastewater (COD) with approx. 21 percent, in comparison with wastewaters coming from the classical anti-felting treatment (oxidative pre-treatment, reducing treatment, followed by silicone polymer application)
.
decrease of the content of inorganic substances (solid residue) with approx. 68 percent against the classical treatment;
Publications Popescu, A., Iorga, I. and Vulpe, G. (2003) “Biotechnologies for the treatment of textile materials made of protein fibers. Part I. Treatments for the reduction of wool felting capacity through the usage of protheoitic enzymes”. Industria textila˘, Vol. 55, pp. 103-107 Sta˘nescu, M.D. (2003) “Textile bio-technologies – present stage and perspectives”. Industria textila˘, Vol. 54, pp. 105-106 Sta˘nescu, M.D. and Fogorasi, M. (2002) “Enzymes usage in the textile field”, Annals of AUREL VLAICU University, Textile Series, pp. 69-73 Sta˘nescu, M.D., Bucur, M.S., Pustianu, M., Mihuta and S., Raileanu M. (2003) “Modifying wool dyeing properties with immobilized enzymes”. DWI Reports, Vol. 126, pp. 465-468
Budapest, Hungary Budapest University of Technology and Economics, Budapest XI. Mu˝egyetem rkp. 3, Postal A., H-1521 Budapest, Hungary. Tel: +36-1-463-1376; Fax: +36-1-463-1376; E-mail:
[email protected] Principal Investigator: Judit Borsa, Department of Plastics and Rubber Technology, Department of Physical Chemistry, Department of Chemical Technology
Modification of cellulose fiber for extension of its application Other partners: Academic
Industrial
Johan Be´la National Center for None Epidemiology, Inst. for Isotops and Surface Chemistry of the Hungarian Academy of Sciences, Johannes Kepler University, Linz, Austria, Dr Habil. Ildiko Tanczos Project started: 1 January 2005 Project ends: 31 December 2008 Finance/support: N/A Source of support: Hungarian National Research Fund (OTKA), Governmental Fund (GVOP) Keywords: Cellulose, Cotton, Hemp, Swelling, Chemical modification, Carboxymethylcellulose, Functional textile, Antimicrobial textile, Textile for hospital use
Research register
17
IJCST 17,6
Cellulosic fibers are modified by physical and chemical methods: .
Interaction of cotton cellulose with quaternary ammonium hydroxide (tetramethylammonium hydroxide) is studied in comparison with sodium hydroxide.
.
Cotton fiber is modified by slight carboxymethylation. Effect of technology parameters on the properties of the modified fiber, theoretical aspects of modification and some possible application of modified fiber (e.g. antibacterial textile for hospital use) are investigated. Delignification and refinement of various kinds of Hungarian hemp are studied.
18 .
Research deliverables (academic and industrial) None Publication Borsa, J., La´za´r, K. and La´szlo´, K. (2005), “Antibacterial effect of carboxymethylated cotton fiber” Paper presented at The Fiber Society Spring Conference, St Gallen, May 2005.
Dharwad, Karnataka, India University of Agricultural Sciences, Dharwad, Karnataka, India, Sr Scientist, College of Rural Home Science, University of Agricultural Sciences, Dharwad 580005, Karnataka, India. Tel: (0836) 2743190; Fax: 091-0836-2448349; E-mail:
[email protected] Principal investigator(s): Dr Geeta Mahale, Laboratory
Womens’ wear: development of basic pattern catalogue of different sizes for commercial production Other partners: Academic None
Industrial Magnum Solutions Pvt Ltd, New Delhi Training on Lectra Software Project ends: 30 March 2006
Project started: 4 April 2004 Grant value: RS 134,05,140 Source of support: Indian Council of Gricultural Research, New Delhi Keywords: Sizes and figure types, Blouse, Anthropometrical measurements, Standardization, Size groups, Pattern catalogue ISO size designation standards (ISO 36370) considered adult as the period between 25 and 35 years. Adult is the period of growth of pelvic bones and biacromial circumerferance in which a woman attains complete adult but changes in body
proportions go on throughout the life. The present adult group being very conscious and particular about the fitting and comfort always search for quality fitted outfits. In India, majority of women wear saree blouse. Blouse is a common tight fitting upper garment adorns with the national dress saree. Saree blouse being a tight fitted garment, women mostly depend on the tailor for its stitching. The available drafting systems of blouses do not answer the needs of all sizes and figure types in respect of precision and goodfit. Fitting of the blouse is highly dependent on pattern sizing and figure type of the women. Sizing of the blouse is mainly based on anthropometrical measurements, bust size and figure type are the variables, which differs from women to women. Hence there is a need to take up studies on standardization of the various anthropometrical measurements of women; categorization of measurements into different size groups and development of saree blouse pattern catalogue for different size groups.
Project aims and objectives .
To study the anthropometric (body) measurements of adult women.
.
To categorize these body measurements into small, medium and large sizes. To prepare standard basic bodice block for different size groups by flat pattern method. To develop paper patterns using basic bodice block to fit different sizes and to asses the fitting characteristics by confirmative trials. To create a few saree blouse patterns by adaptation method.
.
.
. . .
To assess the fitting satisfaction by conducting wear trials. To release the saree blouse pattern of different size groups for commercial production
Research deliverables (academic and industrial) Phase I – Data collection Anthropometric measurements of 1200 adult women between age group of 25 and 40 years is collected. Phase II – Categorization of size groups is done based on the mode and standard deviation values of the round chest measurements. Development of basic bodice block for different size categories. Construction of saree blouse patterns for different size groups. Phase III – Conducting wear trials for fitting. Assessing the level of fitting satisfaction through scoring/ranking. Introducing and releasing the saree blouse pattern catalogue For commercial production through intervention packages for functioning trials. Publications Mahale, G. (1991), Anthropometric analysis of rural women for sizing blouse, Karnataka Journal of Agricultural Sciences, Vol. 4 Nos. 3&4, pp. 185-186 Mahale, G., Jyoti, V., Sunanda, R. K. (1999), Anthropometric analysis of adolescent girls, New Cloth Market, Vol. 13 No. 7, pp. 19-20 Varughese, G. and Mahale, G. (1989), “Development of basic blouse size for rural women”, The Indian Textile Journal, Vol. 99 No. 11, pp. 136-137
Research register
19
IJCST 17,6
20
Dharwad, Karnataka, India University of Agricultural Sciences Dharwad Karnataka, Dr Geeta Mahale, Sr Scientist, Dept Textiles and apperal Designing, College of rural home science. UAS, Dharwad, Karnataka 580005. Tel: 091-0836-2743190; Fax: 091-0836-2448349; E-mail:
[email protected] Principal Investigator(s): DR Geeta Mahale, All India Co-ordinated Research Project on Clothing and Textiles Research Staff: Mrs Vanishree, S., and Mrs Iramma, G.
Value addition for agro and animal based fibres – (on-going project) Other partners: Academic
Industrial
None None Project started: 26th September 1996 Project ends: Continued project Grant value: Rupees 11 lakhs per year Source of support: Indian Council of Agricultural Research, New Delhi. Keywords: Value addition, Agro and Animal based Fibres, Non-farming Activities, Mestha, Agave, Hemp and Pina Fibres, Income generation Agro an animal based fibres are of several types such as cotton, wool and silk have developed as major organized sectors but not developed to that extent in decentralized sector. On the other hand the other minor agro and animal based fibres like rabbit wool, wild silk, mestha.camel’s hair, hemp, agave, pina etc. are yet to develop at the levels of organized and decentralized sectors. Some of these fibres are put to limited use at family level. Farming activities are seasonal, which also include cultivation/rearing of fibre yielding crops and animals. Processing of fibres from these sources take place at farm level to some extent, for example deseeding of cotton, retting of jute, agave and hemp. Shearing of wool, reeling of silk etc. These may give employment to the community to a limited extent. These fibres are further procured at industry level giving greater employment outside rural areas. It is generally observed that most of the farming families look for the additional income before the onset of farming activities as expand gains they have made through agriculture. If technologies were developed for cottage level processing of the fibres and for product diversification then value addition would augment income generation to the farming community even during lean season. Therefore there is a need for exploring the possibilities of income generation through non-farming activities using the local natural resources. This would also help the nonfarming families to earn their livelihood to certain extent.
Project aims and objectives .
To identify the resources – agro and animal based fibres, indigenous natural dyes etc. and related by-products and their present utilization.
.
To evolve the national profile of availability and trend of agro, animal based fibres and dyes.
.
To introduce intervention for improvement in the existing practices in processing fibres and dyes.
.
To develop new technologies through different techniques for cottage level adoption. To assess the economic viability of the developed technologies.
.
Research deliverables (academic and industrial) Phase I (1996-98) – Identification of resources and utilization of agro and animal based fibres were studied Phase II (1998-99) – Optimization of Dyeing conditions using Arecanut and Marigold flowers. Dyeing Silk with Arecanut and Marigold Phase III (1999-2001) – Optimization of Dyeing conditions using Arecanut and Marigold flowers. Dyeing cotton with Arecanut and Marigold Phase IV (2001-2002) – Optimization of Dyeing conditions using Acalypha and Teak leaves. Dyeing cotton and silk with Acalypha and Teak. Preparation of value added products viz, saree.cushion covers, files, purses, macrame´ wall hangings.pouches, bags etc. using natural dye sources. Phase V (2002-2003) – Dyeing wool with Arecanut, Marigold, Acalypha and Teak dye extract. Optimization of dyeing conditions using aforesaid natural dye sources with different eco-friendly mordant combinations. Phase VI (2003-2004) – Optimization of Printing conditions and Printing with natural dye sources, Optimization of dyeing conditions using Red sander bark and dyeing cotton, silk and wool with Red sander bark. Preparation of value added products using natural dye sources with printing techniques viz – Table cloth.deewan set. Dress material, pouches, greeting cards, cushion covers, files and handkerchief. Phase VII (2004-2005) – Dyeing cotton, silk and wool with mahaguny leaves, Optimization of dyeing conditions using mahaguny leaves extract and dyeing cotton, silk and wool with mahaguny leaves extract. Preparation of value added products using natural dye sources with dyeing and printing techniques. Publications Mahale, G., Sakshi and Sunanda, R.K. (2004), “An Eco-friendly dye for silk-Teak leaves” Manmade Textiles in India, Vol. XLVII No. 4, pp.130-134. Mahale, G., Sakshi and Sunanda, R.K. (2004), “Acalypha leaves, an Eco dye for Wool”, Textile Asia, Vol. 35 No. 4, pp. 39-43 Mahale, G., Sakshi and Sunanda, R.K. (2003), “Teak leaves, a dye source for cotton”, Textile Asia, Vol. XXXIV No. 9. pp.52-6 Mahale, G., Vanishree, S. and Iramma, G. (2005), “Natural Colourant for Silk”, Asia Textile and Apparel Journal, Vol. 15 No. 6, pp. 42-43, Hong Kong. Mahale, G., Sunanda, R.K. and Sakshi (2004), “Hedge plant – a natural colourant for wool” Textile India, September pp. 31-32.
Research register
21
IJCST 17,6
22
Mahale, G., Sunanda, R.K. and Sakshi (2004), “Acalypha dyed wool-dyeing conditions”, Manmade Textiles in India, Vol. XLVII No. 10, pp. 386-9. Mahale, G., Vanishree, S. and Sunanda, R.K. (2004), “Diversification of natural waste into dyestuff for textile material”, Indian Silk, Vol. 43 No. 3, pp. 29 Mahale, G., Sunanda, R.K. and Sakshi. (2004), “Eco-dyed cotton with Marigold” International Dyer, Vol. 188 No. 4, pp. 46-8. Mahale, G., Sunanda, R.K. and Sakshi (2003), “Value addition – acalypha leaves extract”. paper presented at the seminar of “Natural seminar on Impact of New Economic policies on Rural Industrialization” National Institute of Rural Development, Hyderabad 8 to 10 September. Mahale, G., Sakshi. and Sunanda, R.K. (2003), “Silk dyed with Acalypha (Acalypha wilkesiana) and its fastness”, Indian Journal of Fibre and Textile Research, Vol. 228 No.3, pp.86-9 Mahale, G., Sakshi. and Sunanda, R.K. (2003), “Arecanut – A natural colourant for silk”, Manmade Textiles in India, Vol. 46 No. 4, pp. 136-41 Archana A., Mahale G. and Sakshi (2003), “Colour fastness of Parthenium dyed Silk”, The Textile Industry and Trade Journal, Vol. 41 No. 1-2, pp.49-50. Mahale, G., Sakshi. and Sunanda, R.K. (2002), “Printing with natural dyes – an enterprise. NIRD souvenir on Strategies for rural industrialization through decentralized planning”, 24-25 October, pp. 1-8. Mahale, G., Sunanda, R.K., and Sakshi. (2002), “Colour fastness of Eco dyed cotton with Marigold”, Textile Trends, Vol. 44 No. 10, pp. 35-9 Mahale, G., Sakshi. and Sunanda, R.K. (2002), “Fastness properties of Acalypha on cotton”, International Dyer, Vol. 187 No. 9, pp.39-41 Mudugal S. and Mahale G. (2002), “Fountain flower dyed UAS sheep breed wool – its colourfastness in acidic media”, Manmade Textiles in India, Vol. 45 No. 4, pp. 140-44 Mahale, G., Sunanda, R.K. and Sakshi. (2002), “Eco-dyeing – Diversification of Teak leaves”. Proceedings of the International Conference on Eco-balance and Life Cycle Assessment in India, 13-15 February, pp. 76-79. Mahale, G., Sunanda, R.K. and Sakshi. (2001), “Natural dyeing of silk with Teak leaves and its fastness”. Proceedings of Convention on natural Dyes, IIT, New Delhi: pp. 111-16 Mahale, G., Sunanda, R.K. and Sakshi., (2001), “Eco-dyeing of cotton with teak and its fastness”, The Textile Industry and Trade Journal, Vol. 39 Nos 9-10, pp.33-6. Mahale G., Sakshi. and Sunanda R.K. (2001), “Colour fastness of Arecanut dyed cotton”. ManMade Textiles in India, Vol. 44 No. 6, pp. 243-46. Mahale, G., Sunanda, R.K., Bhavani, K. and Sakshi (2000), “Colourfastness of Arecanut dye in acidic pH”, The Textile Industry and Trade Journal, Vol. 38 Nos 11-12, pp. 159-63. Neelima, G., Mahale, G. and Mulla, J. (2000), “Effect of reactive dyes on yarn properties of UAS sheep breed wool”, TextileTrend, Vol. 13 No. 8, pp. 39-40 Neelima G. and Mahale G. (2000), “Colourfastness of Acid dyes-UAS sheep breed wool to washing, sunlight and hot pressing” Manmade Textiles in India, Vol. 43 No. 7, pp. 310-12. Neelima G. and Mahale G. (2000), “Colourfastness of UAS sheep breed wool to crocking”, Textile Industry of India, Vol. 39 No.4, pp.17-20. Neelima G. and Mahale G. (2000), “Effect of acid dyes on fibre properties of UAS sheep breed wool”, The Textile Industry and Trade Journal, Vol. 37 No. 3-4, 61. Mahle G., Sunanda R.K., Bhavani K. and Sakshi. (2000), “Optimization of dyeing condition for arecanut dye”, paper presented at the National seminar on Indian Natural Colouring Agent Beyond 2000 AD, 11 to 13 February, 2000:01 Neelima, G. and Mahale, G. (2000), “Effect of Reactive dyes on fibre properties of UAS sheep breed wool”, Manmade Textiles in India, Vol. 18 No. 2, pp.73-4. Neelima, G. and Mahale, G. (1999), “Effect of acid dyes on yarn properties of UAS sheep breed wool”, New Cloth Market, Vol. 13 No. 12, pp. 59-60
Mahale, G., Bhavani, K., Sunanda, K. and Sakshi. (1999), “Colour fastness properties of Areca Catechu in alkaline pH”. Indian Silk, Vol. 38 No.7, pp.18-21. Mahale, G., Bhavani, K., Sunanda, K. and Sakshi. (1999), “Standardizing dyeing conditions for African Marigold”, Man-Made Textiles in India, Vol.42 No. 11, pp. 453-58. Mahale, G., K., Bhavani, K., Sunanda and Sakshi. (1999), “Marigold – A natural colouring agent: assessment of its colourfastness”, Textile Industry of India, Vol. 38 No. 11, pp. 7-13. Neelima, G. and Mahale, G. (1999), “Dyeing of UAS sheep breed wool with reactive dyes”, Textile Industry of India, Vol. 38 No. 10, pp. 10-11 Mahale, G., Sunanda, K., Bhavani, K. and Sakshi. (1999), “Tagetes Erecta: Its colorfastness in acidic media”, Textile Trends, Vol. 42 No. 7, 223-26 Mahale, G., Sunanda, R.K., Bhavani, K. and Pratibha, B.R. (1999), “Natural Dyeing-Silk with Arecanut extract”, Textile Industry of India, Vol. 38 No. 7, pp. 20-2 Mahale, G. et al. (1999), “Karantaka-Woollen Blanket weavers”, Indian Journal of Small Ruminants, Vol. 5 No. 1, pp. 39-42. Neelima, G. and Mahale, G. (1999), “Shrink proofing of Wool”, Indian Textile Journal, Vol. 109 No. 7, pp.56-8. Rashidabanu, U. and Mahale, G. (1999), “Pineapple leaf fibre (PALF)”, The Textile Industry & Trade Journal, Vol. 37 Nos. 1-2, pp. 27-9. Mahale G., Sunanda R.K. and Bhavani K. (1998), “Value addition – cotton yarns”, The Textile Industry & Trade Journal, Vol. 36 Nos. 11-12, pp. 75-9. Mahale, G., Bhavani, K. and Sunanda, R.K. (1998), Bamboo (Bambusa arundinaca): “Immense possibilities”. Textile Industry of India. 12-4 November. Mahale, G. and Jayashree, Y. (1995), “Colour fastness of napthol dyed cotton fabric”, Textile dyer and printer, Vol. 28 No. 12, pp. 23-7
Edinburgh, UK Heriot-Watt University, Riccarton, Edinburgh, EH14 4AS, Tel: 0131 451 3034; Fax: 0131 449 5542; E-mail:
[email protected] Principal investigator(s): Professor John J.I.B. Wilson, Physics, EPS
Flexible solar cells Other partners: Academic R.R. Mather Project started: – Grant value: Source of support: Keywords: Solar cells, Coatings, Plasma
Industrial Project ends: –
Plasma chemical vapour deposition of silicon films on textile substrates for solar cells (under development). This project has received some EPSRC support for a feasibility study – we have shown that the silicon coating process may be applied to polyester substrates, and that
Research register
23
IJCST 17,6
the addition of a plasma polymer film can protect against atmospheric degradation of the silicon; further work is underway to develop an efficient solar cell structure to provide flexible devices.
Project aims and objectives An effective, low cost, flexible solar cells on textiles.
24
Research deliverables (academic and industrial) Demonstrate photovoltaic response. Identify durability problems (e.g. due to flexure). Prototype solar cell device. Publication www.technical-textiles.net (December 2002)
Ghent-Zwijnaarde, Belgium Department of Textiles, Ghent University, Technologiepark 907, 9052 Ghent-Zwijnaarde, Belgium. Tel: +32-9-2645735; Fax: +32-9-2645846; E-mail:
[email protected] Principal Investigator(s): Prof. Gustaaf Schoukens, Department of Textiles Research Staff: Prof. Paul Kiekens, Prof. Gustaaf Schoukens
Development of high performance artificial grass for football applications Other Partners: Academic CMSE Centre for Material Science and Engineering Project started: March 2004 Grant value: e435.704 Keyword: Artifical turf
Industrial Desso DLW Sports Systems Project ends: February 2007 Source of support: IWT
This project wants to contribute to a breakthrough in the acceptance of artificial turf for football, having constant playing properties during the whole season.
Project aims and objectives The most important aims of the project are: .
complete acceptance of artificial turf by sports men
.
guaranteed quality as to playing properties (sliding, ball roll, ball bounce,. . .)
Research deliverables (academic and industrial): In order to meet these aims, the project involves: . research into an optimal stalk of artificial grass based on fixed required characteristics .
research into and development of an alternative construction based on experimentally determined criteria and numerical simulation
.
research into and development of testing methods to translate intuitive aspects (e.g. sliding, ball contact,. . .) into objective criteria.
Publications Not available.
Ghent-Zwijnaarde, Belgium Department of Textiles, Ghent University, Technologiepark 907, 9052 Ghent-Zwijnaarde, Belgium. Tel: +32-9-2645735; Fax: +32-9-2645846; E-mail:
[email protected] Principal Investigator(s): Dr Philippe Westbroek, Department of Textiles Research Staff: Dr Philippe Westbroek
Study of electrochemically deposited Bi-Te layers on gold, platinum and stainless steel and their (semi-conductor) properties using electrochemical and spectroscopic methods Other partners: Academic None Project started: January 2005 Grant value: e 170.700 Source of support: FWO
Industrial None Project ends: December 2007
Publications Not available.
Ghent-Zwijnaarde, Belgium Department of Textiles, Ghent University, Technologiepark 907, 9052 Ghent-Zwijnaarde, Belgium. Tel: +32-9-2645735; Fax: + 32-9-2645846; E-mail:
[email protected] Principal Investigator(s): Carla Hertleer, Department of Textiles Research Staff: Prof. Lieva Van Langenhove
Research register
25
IJCST 17,6
26
Application of smart textiles in clothing Other partners: Academic IMECCentexbel Project started: September 2004 Grant value: e 65.120 Source of support: IWT TIS Keyword: Smart textiles
Industrial Project ends: August 2006
The application of smart textiles in clothing provides on the one hand an answer to realistic needs relating to health, protection, communication, comfort, and safety of various users. On the other hand, this application creates new and economically interesting opportunities for the innovative textile entrepreneur. Moreover, smart clothes aim at a whole range of users, such as sportsmen, intervention teams, older people, disabled people, . . . The field of smart textiles practically has unlimited possibilities and what is more, it is to a large extent virgin territory. The time has come to instigate Flemish companies in the “clothing” value chain to take this innovative course. At the moment, too few textile companies in Flanders are familiar with the possibilities of the existing intelligent materials or with new technologies or textile materials enabling further product innovation. Through a dynamic networking between the players involved, the real possibilities with regard to the application and the maintenance of smart textiles will be made public, demonstrable and negotiable. This action aims at encouraging Flemish companies to start company specific innovation projects, whether or not in co-operation with specialised technology providers and/or knowledge centres.
Project aims and objectives This project aims at promoting networking between all parties involved in the “clothing” value chain, the specialized knowledge centres and the technology providers in the framework of using smart textiles in clothing. Publications Not available.
Ghent-Zwijnaarde, Belgium Department of Textiles, Ghent University, Technologiepark 907, 9052 Ghent-Zwijnaarde, Belgium. Tel: +32-9-2645735; Fax: + 32-9-2645846; E-mail:
[email protected] Principal Investigator(s): Prof. Lieva Van Langenhove, Department of Textiles Research Staff: Prof. Lieva Van Langenhove
Data transmission and wireless communication for smart textiles Other partners: Academic None Project started: January 2005 Grant value: e 123.600 Source of support: FWO
Industrial None Project ends: December 2008
Publications Not available.
Ghent-Zwijnaarde, Belgium Department of Textiles, Ghent University, Technologiepark 907, 9052 Ghent-Zwijnaarde, Belgium. Tel: +32-9-2645735; Fax: +32-9-2645846; E-mail:
[email protected] Principal Investigator(s): Dr Philippe Westbroek, Department of Textiles Research Staff: Dr Philippe Westbroek
Flexible Peltier elements based on textile structures Other partners: Academic None Project started: March 2005 Grant value: e 140.079 Source of support: IWT
Research register
Industrial None Project ends: February 2007
Peltier elements are currently very successful given the possibility to use them as a cooling or heating system when applying an electrical current. Depending on the current direction, cooling or heating is obtained; when inverting the current direction, the cooling and heating effect will be inverted as well. This means that heating and cooling can be obtained at the same plane of the element by inverting the current which is sent through that element. However, Peltier elements have a rigid structure, which reduces their application possibilities. For that reason, the idea was put forward to develop flexible Peltier elements on the basis of textile structures. This means that fibres have to be available possessing the appropriate semi-conductor properties (such as Bi-Te) for an optimal Peltier effect. Alternative semi-conductor fibre structures have already been described, but they are less suitable for application in Peltier elements.
27
IJCST 17,6
28
Project aims and objectives A first aim in this research project is to take a first step in that direction by conducting fundamental research into the electrochemical deposition of Bi-Te layers on different electrode materials and under varying circumstances and to study the properties of these layers with electrochemical and surface techniques. Gaining insight in the deposition and the properties of the deposited layer in view of the production of suitable semi-conductor properties is a second aim in this project. In a second project comprising more application-directed research, the aim is to deposit the semi-conductor layers with the most suitable properties on conductive metal wires, which are to be processed into a textile structure, suitable for the study of the Peltier effect. Publications Not available.
Ghent-Zwijnaarde, Belgium Department of Textiles, Ghent University, Technologiepark 907, 9052 Ghent-Zwijnaarde, Belgium. Tel: +32-9-2645735; Fax: +32-9-2645846; E-mail: johanna.louwagie @ugent.be Principal Investigator(s): Ing. Johanna Louwagie, Department of Textiles Research Staff: Prof. Paul Kiekens
Bilateral networking and quality training in textiles (NeQuaTex) Other partners: Academic Technical University “Gh. Asachi” Iasi Romania
Industrial Center for continuing education and training – Cetex Romania, S.C Filbac S.A Tg.Lapus ROMANIA Project ends: January 2007
Project started: February 2005 Grant value: e 90.962 Source of support: Ministry Flemish Community
In their (attempts for) mutual cooperation, companies in the textile and clothing sector in Flanders and Romania encounter many practical problems. This project wants to deal with the most important ones: understanding and implementation of European quality standards, knowledge of EC regulations for CE marking, access to accredited laboratories, quality certification and management, mentality differences, efficiency, implementation of the latest technological developments. . . Other problems will be diagnosed in the initial phase of the project and incorporated in the work programme.
Research deliverables (academic and industrial) In both countries, training sessions, workshops and exchanges will be organised, revealing the basic principles of modern quality management in a free European market environment and new technological developments. Publications Not available.
29
Ghent-Zwijnaarde, Belgium Department of Textiles, Ghent University, Technologiepark 907, 9052 Ghent-Zwijnaarde, BELGIUM. Tel: +32-9-2645735; Fax: +32-9-2645846; E-mail:johanna.louwagie @ugent.be Principal Investigator(s): Ing. Johanna Louwagie, Department of textiles Research Staff: Prof. Paul Kiekens
T3-SQM – training in textile technology, standardisation and quality management Other partners: Academic Technical University of Sofia Project started: January 2005 Grant value: e 99.988 Source of support: IWT
Research register
Industrial State Agency of Standardisation and Metrology (SATM) Project ends: December 2006
Over the last few years, the Bulgarian textile industry has made some progress in adapting itself to the free market economy. However there is still a long way to go. In order to come to a successful adaptation, the local textile companies – especially SMEs – and laboratories need assistance and training in the field of textile technology, standardisation and quality management. As a consequence, this project mainly focuses on the introduction and the implementation of standardisation and quality management systems according to EN and ISO standards in SMEs and testing laboratories. It will also help the local industrials and laboratories to get actively involved in the development of new testing methods and the implementation of EN and ISO standards.
Research deliverables (academic and industrial) In both countries training sessions for personnel will be organised, revealing the basic principles of modern quality management in a free market environment, standardisation and textile technology. In order to get a clear view of the local situation, at the project kick-off, a preparatory visit will be paid to the partners in Bulgaria. In consultation with the partners, training sessions will be developed in different modules (Product Quality, Standardisation, Quality management,. . .). At the beginning, these modules will be taught in English by UGENT-Tex to trainers from Bulgaria (training of trainers). The courses will be
IJCST 17,6
30
summarised into the local language. Subsequently, the training sessions will be repeated for a broader public, by a mixed team of trainers (training of trainees). Employees from the partner organisations and the local industry will also be invited to participate in short-term traineeships and workshops at the Department of Textiles of Ghent University, and possibly other laboratories and/or companies. This will give them the opportunity to experience how a modern and functional quality system in a laboratory and/or company works. Apart from basic subjects, this project also foresees in thematic training sessions that will reveal the latest developments in the field of technological quality management and textile testing. . At the end of the project, a training and consulting centre will be established at the Department of Textiles, Technical University of Sofia (Centre for Technological Quality Management). This centre will see to the further dissemination of the acquired knowledge long after the end of the project. Publications Not available.
Ghent-Zwijnaarde, Belgium Department of Textiles, Ghent University, Technologiepark 907, 9052 Ghent-Zwijnaarde, Belgium. Tel: +32-9-2645735; Fax: +32-9-2645846; E-mail:
[email protected] Principal Investigator(s): Dr Philippe Westbroek, Department of Textiles Research Staff: Prof. Paul Kiekens, Dr Philippe Westbroek
Development of an electrochemical sensor for the determination of benzene(chlorine)phenols and PCBs Other partners: Academic
Industrial
Project started: – Project ends: – Grant value: e Source of support: Ghent University – BOF The aim of the project is to develop an electrochemical sensor system that is able to detect (chloro)phenols, PCBs and dioxins in the ppb concentration range in aqueous (waste water) and non-aqueous (oil, fat, food) medium. These compounds are very toxic and difficult to decompose, a property that explains their danger and difficulty to be detected. However, these compounds can be oxidised electrochemically at the surface of an electrode if a catalyst is used.
Research deliverables (academic and industrial) In this project, tetrasulfonated metallophthalocyanines (MTSPcs) are used as catalyst and are immobilized at the surface of an electrode, with M ¼ Co(II); Fe(II) or Ni(II). At this stage
of the project, Co(II)TSPc and Fe(II)TSPC were immobilized at the surface of a gold electrode and the deposited layer is fully characterized by electrochemical methods and XPS. Oxidation of chlorophenols at this electrode in aqueous alkaline solution resulted in poisoning of the electrode but for this effect solutions are sought and will be found in the near future. At present, the electrode is able to detect chlorophenols in a concentration range from 100 ppb to 500 ppm in a reproducible way (standard deviation of less than 1 per cent). Solutions will be found to avoid or circumvent poisoning of the electrode, ultramicro electrodes will be investigated for their suitability in order to improve on the detection limit (goal is 20 ppb) and to be able to measure in non-aqueous solutions. Finally, the developed electrode will also be tested for its usability in detection of PCBs and dioxins. Publications Not available.
Ghent-Zwijnaarde, Belgium Department of Textiles, Ghent University, Technologiepark 907, 9052 Ghent-Zwijnaarde, Belgium. Tel: +32-9-2645735; Fax: +32-9-2645846; E-mail:
[email protected] Principal Investigator(s): Dr Philippe Westbroek, Department of Textiles Research Staff: Prof. Paul Kiekens, Dr Philippe Westbroek
Remedent: voltammetric determination of hydrogen peroxide in gels for bleaching teeth Other partners: Academic None Project started: January 2005 Grant value: e 27.500 Source of support: IWT
Industrial Remedent Project ended: June 2005
Publications Not available.
Ghent-Zwijnaarde, Belgium Department of Textiles, Ghent University, Technologiepark 907, 9052 Ghent-Zwijnaarde, Belgium. Tel: +32-9-2645735; Fax: +32-9-2645846; E-mail:
[email protected] Principal Investigator(s): Prof. Lieva Van Langenhove, ing. Marian Ledoux, Department of Textiles Research Staff: Prof. Lieva Van Langenhove
Research register
31
IJCST 17,6
Self-learning machine speed for air-jet looms AUTOSPEED II Other partners: Academic
32
None Project started: March 2005 Grant value: e 370.591 Source of support: IWT
Industrial Picanol Project ends: August 2006
Publications Not available.
Ghent-Zwijnaarde, Belgium Department of Textiles, Ghent University, Technologiepark 907, 9052 Ghent-Zwijnaarde, Belgium. Tel: +32-9-2645735; Fax: +32-9-2645846; E-mail:
[email protected] Principal investigator(s): Dr Philippe Westbroek, Department of Textiles Research staff: Dr Philippe Westbroek
Development of nano-structured metallophthalocyaninebased surfaces for use in analysis of environmentally, medically and biologically important molecules Other partners: Academic
Industrial
South-Africa, Rhodes University, Department of Chemistry Project started: February 2005 Project ends: January 2007 Grant value: e 65.000 Source of support: Ghent University / VLIR There has been a growing interest in thiol-derivatized metallophthalocyanines (MPcs) in recent years for the fabrication of self-assembled monolayers (SAMs). SAMs are an incredibly versatile means of extending the functions of an electrode and have been known to offer greater advantages over films produced by other methods such as spin coating. However, our joint work on SAMs has shown limitations on the usable range for electrodes modified with SAMs. Thus in this proposal thiol-derivatized MPc complexes will be synthesized and deposited as nano and micro thin films by electrospinning. Electrospinning produces nano-fibres or sprays of thiol-derivatized MPc on conductive surfaces. The thiol-derivatized MPc complexes will also be deposited on surfaces modified by silane deposition. The catalytic activity as well as the stability and reproducibility of surfaces modified by the various methods (electrospinning, SAM and
binding to silanes) towards the analysis of environmentally, medically and biologically important molecules will be evaluated and the results obtained compared. This work will be an important step towards the development of electrochemical sensors for analysis of these molecules. The analytes chosen include environmentally important molecules (such as thiols and organohalides), biologically important molecules (such as cysteine and cysteinecontaining proteins) and medically important molecules (such as the neurotransmitters; serotonin and dopamine). Publications Not available.
Istanbul, Turkey Faculty of Civil Engineering, Department of Environmental Engineering, Istanbul Technical University, 34469 Maslak, Istanbul, Turkey. Tel: 0090 212 285 37 86; Fax: 0090 212 285 65 870; E-mail:
[email protected]. Principal investigator(s): Idil Arslan-Alaton (ITU), Laboratory of the Environmental Engineering Department Research staff: Idil Arslan-Alaton (ITU), Fatos Germirli Babuna (ITU), Gulen Eremektar (ITU)
Environmental pollution effects and best available integrated technologies for textile industry preparation, dyeing and finishing process chemicals Other partners: The project has been accepted within the scope of EU COST Action 628: Life Cycle Assessment of textile products, eco-efficiency, and definition of BAT of textile processing
Academic
Industrial
Project started: 1 February 2004 Project ended: 31.07.2005 Grant value: e 43125 Source of support: Turkish Technical and Scientific Research Council (TUBITAK) Keywords: Textile dyeing and finishing industry, Textile auxiliary chemicals, Advanced oxidation, Biological treatment, Acute toxicity, Biodegradability The textile sector remains one of the most serious environmental polluters and major industries in Turkey as well as many other countries from both the economical and strategic point of view. The future of this sector directly and strongly depends on environmental problems associated with it and “source based pollution reduction” as well as “sustainable development” driven pollution prevention and control strategies. With the above mentioned facts in mind, the goal of the proposed project is to develop a reliable and quantitative methodology as to identify the ecological effects and indicators
Research register
33
IJCST 17,6
34
of problematic process chemicals commonly used in the textile preparation, dyeing and finishing industry, and to assess the efficiency of alternative, innovative, integrated endof-pipe treatment technologies. Within the scope of the ongoing COST ACTION 628 ending in 2005, the source-based environmental impacts of potentially toxic and/or refractory textile process chemicals will be evaluated via collective, critical pollution parameters (TOC, COD, BOD, as well as fate of acute toxicity towards the water flea Daphnia magna, and AOX, i.e. adsorbable organic xenobiotics or halogens), accompanied with biological treatability and best available integrated chemical + biochemical oxidative treatment technology assessment (ozonation, catalyzed ozonation and Fenton’s reaction preceding biotreatment). Besides, biological degradability studies based on respirometric activity analysis of biosludge and quantitative COD fractionation will be implemented to establish a quantitative, reliable, empirical pollution control and most appropriate treatment profile for the textile preparation, dyeing and finishing sector.
Project aims and objectives In the present study, the treatability of commercially important textile auxiliaries by ozonation at varying pH (at pH ¼ 3.5, i.e. the natural application pH of textile tannins, and at pH ¼ 7) and ozone doses (i.e. 500, 750, 900, 1100 mg/h) was investigated. The major purpose of the experimental study was to improve the biocompatibility of the studied chemicals via partial oxidation, i.e. ozonation at relatively low doses to minimize treatment costs. The efficiency of ozonation in degrading the parent compounds and their oxidation intermediates was comparatively assessed in terms of the collective environmental parameters COD, TOC and UV optical density (absorbance at 254 nm and 280 nm; UV254 and UV280). Changes in biodegradability and acute toxicity of the selected textile auxiliaries after ozonation under optimized reaction conditions was assessed by BOD5 measurements, a modified, activated sludge inhibition test (ISO 8192) using mixed cultures resembling sewage sludge, the water flea Daphnia magna and the marine algae Phaedactilium tricornutum.
Research deliverables (academic and industrial) Two progress and one final reports submitted to TUBITAK Publications Arslan-Alaton, ˙I., Eremektar, G., Germirli-Babuna, F., Selc¸uk, H. and Orhon, D. (2004), “Chemical pretreatment of textile dye carriers with ozone: effects on acute toxicity and activated sludge inhibition”, Fresen. Environ. Bull. Vol. 13 No. 10, pp. 1040-44. ¨ zerkan, B. and Selcuk, H. Arslan-Alaton, ˙I., Eremektar, G., Germirli-Babuna, F., ˙Insel, G., Teksoy, S., O (2004), “Advanced oxidation of commercial textile biocides in aqueous solution”. IWA 4th World Water Congress and Exhibition, Marrakech, 19-24 September, CDROM, paper Nr. 85867. Arslan Alaton, I., Insel, G., Eremektar, G., Germirli Babuna, F. and Orhon, D. (n.d.), “Effect of textile auxiliaries on the biodegradation of dyehouse effluent in activated sludge”, paper presented at the Chemosphere 2005 (in press). Arslan Alaton, I., Insel, G., Eremektar, G., Germirli- Babuna, F. and Orhon, D. (2005), “The fate of textile biocides and dye carriers in biological activated sludge systems”, Paper presented at the 2nd International Conference “Textile Processing State of the Art & Future Developments” Cairo, 11-13 April p. 58. Koyunluog˘lu, S¸., Arslan-Alaton, ˙I., Eremektar, G. and Germirli Babuna, F. (2005), “The effects of ozonation on the biodegradability and toxicity of tannins used in the textile dyeing & finishing
industry”, Paper presented at the 3rd International Conference on Ecological Protection of the Planet Earth, 10-11 June, Istanbul, Session: “New Technologies, Innovations, Alternative Energy Resources in Water”
Izmir, Turkey
35
Textile Research Centre (Tubitak-TAM), Tubitak Bornova 35100 Izmir, TURKEY. Tel: +90 232 3421410; Fax: +90 232 3421410; E-mail:
[email protected] Principal investigator(s): Ass. Prof. Arif Taner Ozguney, Tubitak Textile Research Centre Research Staff: Ass. Prof. Cankut Taskin, Ass. Prof. Serap Donmez Kretzschmar, Ass. Arzu Ozerdem, Ass. Pelin Gurkan, Ass. Gonca Ozcelik, Tex.Eng. Burak Baykaldi
Comparison of the properties of woven and knitted fabrics produced by conventional ring spun and compact yarns before and after dyeing and printing processes Other partners: Academic
Research register
Industrial Go¨khan Tekstil San. ve Tic. A.S¸./Denizli Tokullar Tekstil A.S¸./Izmir Project ends: 05.2005
Ege University, Faculty of Engineering Textile Department Project started: 04.2003 Grant value: Source of support: TU¨BI¨TAK Keywords: Compact yarn, Conventional ring-spun yarn, Woven fabrics, Knitted fabrics, Printing-dyeing, Pilling Compact and conventional ring spun yarns in two different yarn counts and twists were produced for simple woven and knitted fabrics by using the same cotton fiber blend. The woven fabrics produced with compact yarns have better pilling values. With respect to air permeability values, some variations exist based on yarn count, due to the effects of hairiness and bulkiness of yarns. It has not been determined any significant statistical differences between abrasion resistance, colour efficiency and rubbing fastness values of fabrics produced with compact and ring spun yarns. The effects of yarn production type on singing and mercerizing processes were also investigated and it was found that according to the end-use of the fabrics woven with compact yarns, these two processes can be eliminated. When the test results of knitted fabrics produced from compact and ring-spun yarns are compared, both grey and processed knitted fabrics produced from compact yarns have better pilling and bursting strength values than fabrics knitted from ring-spun yarns. There are not any statistically significant differences between the abrasion resistance, colour efficiency and rubbing fastness values.
IJCST 17,6
Handle properties of woven and knitted fabrics are evaluated by subjective and objective methods. It is found that the handle properties of the fabrics produced with compact and ring spun yarns are changeable according to the finishing processes.
Project aims and objectives .
To examine the differences between the physical and handle properties of woven and knitted fabrics in various finishing processes, produced with conventional ring spun and compact yarns in detail.
.
To examine whether low twisted compact spun yarns can be used instead of high twisted ring spun yarns.
.
To investigate the possibility of elimination of the singing and mercerizing processes by using compact yarns in woven fabrics.
.
To investigate the possibility of elimination of the enzymatic treatment by using compact yarns in knitted fabrics.
36
Research deliverables (academic and industrial) Low twisted compact spun yarns can be used instead of high twisted ring spun yarns. By this way production can be increased. It is found that hairiness of compact yarn is lower, whereas tensile strength and elongation values are higher than ring spun yarns. As the yarn becomes finer, the difference with reference to tensile strength in warp direction increases in favour of woven fabrics produced with compact yarns. According to the end use of fabrics, singing and mercerizing process can be eliminated when compact spun yarns are used. As compact yarns have the advantages of less hairiness and smooth surface, there is obvious difference in the pilling properties of the fabrics produced from compact and ring spun yarns. According to the results of abrasion resistance, air permeability, colour efficiency, rubbing fastness, there is no difference between the fabrics produced from compact and ring spun yarns. Publication ¨ zerdem A. (2004), “Kompakt ve Tas ¸ kin C., O¨zgu¨ney A., Do¨nmez S., Gu¨rkan P., O¨zc¸elik G. and O Konvansiyonel Ring ˙Ipliklerinin Fiziksel O¨zelliklerinin Bobinleme ˙I¸slemi Sonrası Kars¸ılas ¸ tırılması”, Tekstil Teknik, Mayıs, p. 260
Izmir, Turkey Textile Research Centre (Tubitak-TAM), Tubitak Bornova 35100 Izmir, Turkey Tel: +90 232 3421410; Fax: +90 232 3421410; E-mail:
[email protected] Principal investigator(s): Dr O¨zlenen ERDEM ˙IS¸MAL, Tubitak Textile Research Centre Research staff: Ass. Prof. Arif Taner Ozguney, Tex.Eng. Arzu Arabaci
Influence of the enzymatic scouring on printing of cotton with pigment and reactive dyes Other partners: Academic
Research register
Industrial
Taner Triko Sanayi ve Ticaret Ege University, Textile and apparel A.S¸ C¸orlu-Tekirdag˘ research application center Project started: July 2004 Project ends: January 2006 Grant value: Source of support: TU¨BI¨TAK Keywords: Cotton scouring, Pectinase, Cellulase, Pigment and reactive printing At the first part of the study, 100 per cent cotton knitted fabric was treated with the alkaline and acidic pectinase, acidic and neutral cellulase, lipase, laccase, ligninperoxidase and manganaseperoxidase in one and two step processes by exhaust method under the laboratory conditions. Appropriate absorbancy and whiteness values could be achieved by enzymatic scouring for dyeing. However whiteness needs to be improved for printing. Therefore small amounts of oxidative agents and activators were used in combination with the enzymes. After the evaulation of the laboratory scale experiments, selected trials were applied under the mill conditions. At the second part of the study, pigment and reactive printings and reactive dyeings were applied to the enzymatically and conventionally pretreated fabrics in different concentrations. Fabrics were tested for water absorbancy, whiteness, bursting strength, colour depth, DE, DL, pilling and fastness properties.
Project aims and objectives The goal of this ongoing project is to research the possibility and the effectiveness of enzymatic pretreatments on printing and dyeing properties of cotton, to discuss the optimum process conditions by comparing bioscouring with the conventional pretreatments.
Research deliverables (academic and industrial) Enzymatic scouring is an ecofriendly alternative and has many advantages over alkaline scouring. It seems that more energy, cost and time saving processes can be obtained with the enzymatic treatments. Also its combinations with the oxidative agents are more economical than the conventional pretreatments in terms of energy, time and cost. Publications Not available.
Izmir, Turkey Textile Research Centre (Tubitak-TAM), Tubitak Bornova 35100 Izmir, Turkey Tel: +90 232 3421410; Fax: +90 232 3421410; E-mail:
[email protected]
37
IJCST 17,6
Principal Investigator(s): Prof. Abbas YURDAKUL, Tubitak Textile Research Centre Research Staff: Ass. Rıza ATAV, Tex.Eng. Arzu ARABACI
38
Comparison of the performances in reactive dyeing of natriumchloride and natriumsulphate salts that are used in textile dyeing Other partners: Academic Ege University, Faculty of Engineering Textile Department
Industrial
Sodas¸ A.S¸* . /Denizli, Otuzbir Kimya Ve SanayiTu¨rk Ltd. S¸ti. */Istanbul-Denizli, *Natriumsulphate producer Project ends: August 2004
Project started: January 2004 Grant value: Source of support: TU¨BI¨TAK Keywords: Natriumchloride, Natriumsulphate, Reactive dyeing, Color yield, Brightness, Dyeuptake speed In the first step of this research which is made with the aim of comparing the performances of natriumchloride and natriumsulphate salts in reactive dyeing, effects of these salts on color yield when they are used in same amounts and also in amounts that they give stoichiometrically equal natrium cations were investigated. Besides color yield, the effects of these salts on dyeuptake speed, levelling and brightness of dyeings, ease of removal of these salts in washing treatments were also examined. According to the experimental results, it is observed that the salt type used in dyeing has different effects on color yields. In this research it is found that salt type has no important effect on dyeings carried out with heterocyclic ring type reactive dyes, but it can be told that dyeings made by using natriumchloride have per cent 3-5 higher color yields compared to natriumsulphate. In vınylsulphone type reactive dyes it is observed that dyeings made by using natriumsulphate have per cent 3-15 higher color yields compared to natriumchloride. It is determined that salt type doesn’t cause any important differences on dyeuptake speed, levelling and brightness of dyeings, ease of removal of salts in washing treatments for both types of reactive dyes. As it is known that for some dyes solubility decreases markedly in the presence of natriumchloride. By using natriumsulphate salt instead of natriumchloride in dyeing the precipitetion problem is eleminated. Beside that natriumsulphate salt has the advantage of being less corrosive for textile dyeing machineries.
Project aims and objectives .
To compare the performances of natriumchloride and natrium sulphate salts in reactive dyeing and determine the advantages and disadvantages of these salts.
Research deliverables (academic and industrial) Project was printed as booklet and distributed to industry and in first 6 months it is reported by producers that there was a per cent 20-25 increase in natriumsulphate consumption in textile factories. Publication Yurdakul A., Atav R., Arabacı A. (2005), “Tekstil Boyacılıg˘ında Kullanılan Sodyumkloru¨r ve Sodyumsu¨lfat tuzlarının Performanslarının Kars¸ılas ¸ tırılması”, Tekstil Teknolojisi ve Kimyasındaki Son Gelis ¸ meler Sempozyumu X-Bursa, Haziran 2005, (Presentation), pp. 59-74.
Ksar-Hellal, Tunisie Institut Supe´rieur des Etudes Technologiques de Ksar-Hellal, b.p.68, ISET de Ksar-Hellal, 5070 Ksar-Hellal, Tunisie, 00216 73, Tel: 475900, Fax: 475163, E-mail:
[email protected] Enseignant coope´rant bulgare
Principal Investigator(s):Dr Ivelin Rahnev, Personal investigations, Department of Textiles Research Staff: None
Mechanics of the textile threads General:Mechanical Module of Spinning CAD Stage-2005:Torsion Simulation Module
Other Partners: Academic
Industrial
None None Project started: 05 May 1998 Project ends: None Grant value: None Source of support: None Keywords: Vectorial analysis, Tensor computation, Elastic potential fibrous structure, Thread’s morphology, Initial elastic properties synthesized thread’s shape The torsion of the fibrous beams forms the linear textile products and this ancient tradition to obtain the threads is well studied, described and controlled. In the known technologies, the manufacturing of the yarns and the twisted threads benefits from the perpetual tests. In general, the set of the contemporary spinning mill does not depend so much on the diversity of the materials or of the installed machinery capacity, but on the technological level and on the qualification of the personnel. That leads to a narrow specialization of the production without using the performances of the machines. The contradiction between the fibrous variety, the powerful machines and the impossibility for fast change in the set comes from the lack of universal model of the twisted thread.
Research register
39
IJCST 17,6
40
The simulation of the torsion of the threads requires formalization and succession during the application of the mathematical apparatus of the analytical mechanics. The universal description of the structure and the mechanical parameters builds the image and directs the technological design of the textile linear products. The theoretical models of the torsion envisage the general character and the probable physical behaviour of textile thread. The known methods are based on the analytical, numerical or empirical principles with the common purpose to guide the experimental optimization of the design. In the present analysis we applied the universal theorems of the mechanics to the diversity of the real fibrous matters. The deductions correspond to the textile practice and the technological knowledge concerning the behaviour of the textile linear products. The analytical mechanics in the continuous mediums of textile fibres solves simultaneously the structure and the properties of the thread.
Project aims and objectives Our objective is the analytical model of the structural modification and the tenacity of the fibrous beam during torsion. In this work we would like to display the process of torsion, the deformation of the fibres and; the morphology and the awaited properties of the thread. This work is intended for the needs of textile engineers training as well as for the textile products design. The algorithm contains a group of modules of the spinning CAD.
Research deliverables (academic and industrial) The methodological result consists in the prognostication of the thread properties and its visualization could be helpful for the field engineers practice. We think that the one of the main results is the synthetic visualization of the fibrous structure and of the thread morphology shape. The computed values of the structural parameters and the constraints of the preloaded initial state are essential for the consecutive simulation of the strained thread. This work is intended for the needs of textile engineers training as well as for textile products design. The algorithm contains a group of modules of the spinning CAD. Publications Rahnev I., “The industrial establishment and the principal algorithm of the spinning CAD”, (in Bulgarian), Textile and Garment, pp. 7-12, 5/2004. Rahnev I., “Virtual mechanics of the textile torsion”, paper presented at Fifth Autex Conference, Portorozˇ, Slovenia, 27-29 June 2005, pp. 333-8.
Lodz, Poland Technical University of Łodz, Faculty of Engineering and Marketing of Textiles, Zeromskiego 116, 90-245 Lodz, Poland. Tel: 48-42-6313337; Fax: 48-42-6313343; E-mail: cybulska@wipos. p.lodz.pl
Principal Investigator(s): Maria Cybulska, Department of Architecture of Textiles Research Staff: Maria Cybulska, Tomasz Florczak, Jerzy Maik
Methods of identification and reconstruction of textiles for special applications Other Partners: Academic
Industrial
Institute of Archaeology and Ethnology, None Polish Academy of Science Grant value: 100000 PNZ Project started: November 2004 Project ends: March 2006 Source of support: The Minister of Science and Information Society Technologies Keywords: Textiles, 3D graphics, Image analysis, Identification of structure, Raw material and colour, 3D reconstruction Problems related to identification and reconstruction of textiles can be found in different areas referred not only to textile science and engineering. .
Science – analysing the structure-to-mechanical behaviour relationships.
.
Industry – on-line quality control.
.
Designing – visualisation of the end product.
.
Criminology – sample identification.
.
Archaeology – determination of the origin and structure of finds, virtual reconstruction.
While first two problems can be solved using image processing and analysis methods, the next three problems need additionally the implementation of 3D graphics methods to simulate – virtually construct or re-construct – the product. The research includes the following tasks: (1) Application of some well known and development of the new methods allowing full identification of the properties of product represented by the sample. It means the raw material, colour and structure. (2) Development of 3D graphics tools allowing simulation of any kind of textile products – from fibre and yarn to fabric and cloth (3) Virtual reconstruction of product in its original form on the basis of obtained results of sample analysis Publications None
Research register
41
IJCST 17,6
42
Maribor, Slovenia Faculty of Mechanical Engineering, University of Maribor, Smetanova 17, 2000 Maribor, Slovenia, Tel: +386 2 220 7910, Fax: +386 2 220 7990, E-mail:
[email protected] Principal Investigator(s):Prof. Dr Alenka Majcen Le Marechal, Institute of textiles, Laborytory for chemistry, dyes and polymers Research Staff: Synthesis and modification of dyes and polyfunctional reagents, synthesis of active organic substances, Application of polyfunctional reagents in textile engineering, Introduction of nanoencapsulation in medical and hygienic textiles, Decoloration of textile waste water (H2O2/UV, H2O2/O3, H2O2/Fenton, H2O2/ultrasound), Development of sensors for monitoring of ecological parameters, Analytical methods in textile industry, Development of alternative methods for the determination of free formaldehyde
Advanced oxidation processes and biotreatments of water recycling in the textile industry (sixth framework programme) Other Partners: Academic
Industrial
None Industrial Robert Blondel SA, TSP, Helios Intalquartz, DAMA, OBEM Project started: 01.01.2005 Project ends: 31.12.2006 Grant value: 363.471 EUR Source of support: EC Keywords: Advance Oxidation Processes, UV-activated, Hydrogen peroxide, Thermal activated oxidation process, Decolouration, Bioflotation, ANN-based process control software The AdOPBio project aims to develop a decolouring and recycling treatment of the wastewaters in the textile finishing industry, based on two alternative methods: Advance Oxidation Processes (UV-activated photolysis of hydrogen peroxide and thermal activated oxidation process) for the decolouration of the spent bath, combined with a bioflotation process for the destruction of the residual organic load. The combination of these wastewater treatments is expected to achieve a complete decolourisation of the process waters for every type of wet process (finishing, bleaching, dyeing, etc.). The project will also develop and implement a process-control software based on artificial neural network and systems dynamics. Research centres in collaboration with textile finishing companies and suppliers of dyeing machines and wastewater treatment equipment will develop a prototype that will be tested and validated by the end-user companies (textile finishing companies) in order to accumulate experiences and improve the capability of the plant to match a wide range of industrial needs.
The project includes all the steps in developing a wastewater treatment unit such as: . .
modelling and laboratory investigations of AOP and bioflotation processes design and manufacture of AOP and bioflotation reactors
.
design and manufacture of a dyeing machine, interfaced with both AOP reactors
.
implementation of an ANN-based process control software interfacing the dyeing machine with the bioflotation treatment plant tests of the plant in and industrial validation of the decolouring and recycling process
.
Project aims and objectives Today more than 4.000 compounds are used in the textile finishing process, which complicates design and setting up of a single cleaning up and recycling technology. The equipment used in decolouring and cleaning up processes is very hard to set and tune with the continuous variation in load and composition. Moreover, the pollutant charge can overload the capability of the cleaning plant, therefore failures are common and operational costs are prohibitive for SME companies. One of the purposes of the project will be to investigate different dyeing and finishing processes, drawing guidelines for the convenience of recycling water by this system. ADOPBIO will focus on a decolouring and cleaning up treatment for textile finishing wastewaters based on an UV-activated photolysis of the hydrogen peroxide (an targetted Advanced Oxidation Process, AOP) combined with a bioflotation treatment. The combination of these treatments can achieve a complete decolourization and recycling of the process waters for every type of wet process (finishing, bleaching, dyeing, etc.). ADOPBIO will also focus on the development and implementation of process control software, based on artificial neural network and systems dynamics. The textile finishing wastewater treatment is expected to achieve the following characteristics: Quality of the treated process water: .
full decolorization (. 99% for interfering dyes; . 90% for other colour substances)
.
reduction of surfactants $ (99%) and toxic compounds (COD reduction $ 95%) if not recyclable
.
recycling of at least 75% of the wastewaters
Research deliverables (academic and industrial) Technical report on the end users wet processes Economic impact report on the end users wet processes End users specifications AOPs kinetic models Optimized laboratory AOP reactors AOPs and bioflotation scheme guidelines AdOPBio web-space (diagrams, run-time models, simulations) UV-activated AOP reactor scheme Thermal activated AOP reactor scheme Bioflotation reactor scheme Dyeing machine scheme
Research register
43
IJCST 17,6
44
ANN-based process control software Learning and operational simulation tools AOP simulated control panel UV- and thermal-activated AOP pilot reactors Bioflotation pilot reactor Dyeing machine prototype Integrated AdOPBio system of water treatment Industrial validation report Runnable models (predictive, sensitivity tests) Management tools (project calendar, reporting, cost follow up) Project quality indicator and milestone report Mid-term review Plan for using and disseminating the kowledge Publications Brodnjak-Voncˇina, D. and Majcen Le Marechal, A. (2003) “Reactive dye decolorization using combined ultrasound/H2O2”. Dyes Pigm., Vol. 59, No. 2, pp. 173-9 Kurbus, T., Slokar, Y.M., Majcen Le Marechal, A. and Brodnjak-Voncˇina, D. (2003) “The use of experimental design for the evaluation of the influence of variables on the H2O2/UV treatment of model textile waste water”. Dyes Pigm., Vol. 58, pp. 171-8 Kurbus, T., Majcen Le Marechal, A. and Brodnjak-Voncˇina, D. (2003) “Comparison of H2O2/UV, H2O2/O3 and H2O2/Fe2+ processes for the decolorisation of vinylsulphone reactive dyes”. Dyes Pigm., Vol. 58, No. 3, pp. 245-52 Kurbus, T., Slokar, Y.M. and Majcen Le Marechal, A. (2002) “The study of the effects of the variables on H2O2/UV decoloration of vinylsulphone dye”. Part II. Dyes Pigm., Vol. 54, pp. 67-78 Voncˇina, B., Bezek, D. and Majcen Le Marechal, A. (2002) “Eco-friendly durable press finishing of textile interlinings”. Fibres Text. East. Eur., Vol. 10, No. 3, pp. 68-71 Voncˇina, B., Majcen, N., Majcen Le Marechal, A., Brodnjak-Voncˇina, D. and Bezek, D. (2004) “Free formaldehyde determination using HPLC”. Mater. Sci. Forum, Vols. 455-456, pp. 801-04 Voncˇina, B. and Majcen Le Marechal, A. (2005) “Grafting of cotton with [beta]-cyclodextrin via poly(carboxylic acid)”. J. Appl. Polym. Sci., Vol. 96, No. 4, pp. 1323-8 Vajnhandl, S. and Majcen Le Marechal, A. (2005) “Ultrasound in textile dyeing an the decolouration/mineralization of textile dyes”. Dyes Pigm., Vol. 65, No. 2, pp. 89-101
Maribor, Slovenia Faculty of Mechanical Engineering, University of Maribor, Smetanova ulica 17, SI-2000 Maribor, Slovenia, Tel: +386 2 220-7960, Fax: +386 2 220-7990, E-mail:
[email protected] Principal Investigator(s):Prof. Dr Sc. Jelka Gersˇak, Department of Textiles, Laboratory for Clothing Engineering, Physiology and Construction of Garments Research Staff:Research Unit Clothing Engineering, Research Unit Textile Technology
Clothing engineering and textile materials Other Partners: Academic
Industrial
None None Project started: 2004 Project ends: 2008 Grant value: 17.936.050 SIT or 74.890 ECU for 2004 Source of support: Ministry of Higher Education, Science and Technology Keywords: Clothing, Fabric, Fabric mechanics, Behaviour Comfort, Prediction The research programme Clothing engineering and textile materials is based on complex research studies of fabric mechanics, structural properties of textile materials and their thermo-physiological comfort. The programme comprehends tree associated thematic parts: a) basic research on fabric mechanics with the emphasis on fabrics as complex textile structures, b) behaviour modelling of complex textile structures and c) characterisation of the parameters of thermo-physiological comfort. The important results of the research can be given in the form of the following achievements: .
It was established on the basis of the hysteresis behaviour of textile fabrics that such specific behaviour is caused by fabric’s structural parameters as well as by warp and weft friction properties. Fabric mechanical and physical properties, as well as relaxation are directly influenced by structural parameters of the fabric.
.
Textile fabrics must be treated as non-homogenous and anisotropic structures, and that is the reason why their mechanical characteristics such as tensile, shear, bending and compression properties are non-linear. That indicates that individual yarn and fibre movements within the fabric structure during deformation is rather complex. This is the reason that fabric’s mechanical properties should be regarded as a structural body instead of continuum.
.
Elastic potential of textile fabrics is directly influenced by parameters of mechanical properties. Some fabrics show specific values of elastic potential. Responsible for that is raw material composition, type of weave and fabric’s structural parameters. Tensile elastic potential usually can increase with the increase in the work of deformation and is directly dependant on the ability of particular fabric to be relaxed as well as on its structural parameters and friction properties of warp and weft. Multicomponents mechanical models can be stated as most appropriate models for proper description of relaxation phenomena in fabrics containing elastic components. Generally, according to the achieved results, it can be stated that with the increase of the share of elastic component, also relaxation time increases. That means that in fabrics with higher share of elastic component, because of relaxation of the tension at a constant deformation, the tension is slowly decreasing. After one or 24 hours relaxation it reaches higher values compared with the fabrics containing less elastic component.
.
Research register
45
IJCST 17,6
.
The direct connection between the mechanical and physical properties of fabrics integrated into the garment, and finished garment appearance quality, which is defined on the basis of five criterions (garment fall, appropriate 3D shape of the garment, quality of garment fit, quality of the finished seams: seam puckering and flotation of seams, garment appearance as a whole) has been established. The results of the research work showed that different number of mechanical and physical fabric properties influence the garment appearance quality.
.
Human body thermoregulation as a system of heat exchange between the human body and garment and/or other textile product and environment is influenced by the garment type and its construction, as well as by fabric’s matter and thermal properties.
46
Project aims and objectives The main objectives of this project are: .
definition of relationships between fabric mechanical properties and quality of a produced garment
.
design of a model for prediction of garment appearance quality
.
set-up of a model of fabric behaviour as shell
.
numerical simulation of heat exchange between the human body and garment system or other textiles products and environment.
Research deliverables (academic and industrial) Having realized the proposed research programme and gaining the knowledge about deformation and relaxation processes in fabrics, as well as by gained theoretical cognition and definition of interactions of mechanical and physical properties of fabrics and degree of garment appearance quality of the clothes to be made, the new knowledge has been developed. This new cognitions will be used for prediction of the non linear mechanical properties of fabrics in low loading area and simulation of fabric behaviour. Furthermore, the theoretical cognition from this study area of interactions between matter properties of textile materials, heat exchange and humane thermoregulation will enable the development of the numerical simulation of heat exchange between the human body and garments or other textiles products and environment. Publications Gersˇak, J. (2003) “Investigation of the impact of fabric mechanical properties on garment appearance” (Istrazˇivanje utjecaja mehanicˇkih svojstava tkanina na izgled odjec´e), Tekstil, Vol. 52, No. 8, pp. 368-78 Gersˇak, J. (2004) “Study of relationship between fabric elastic potential and garment appearance quality”, International Journal of Clothing Science and Technology, Vol. 16, Nos 1/2, pp. 238-51 Jevsˇnik, S. and Gersˇak, J. (2004) “Modelling the fused panel for a numerical simulation of drape”, Fibres Text. East. Eur., Vol. 12, No. 1, pp. 47-52 Zavec Pavlinic´, D. and Gersˇak, J. (2004) “Design of the system for prediction of fabric behaviour in garment manufacturing processes”, International Journal of Clothing Science and Technology, Vol. 16, Nos 1/2, pp. 252-61 Zavec Pavlinic´, D. and Gersˇak, J. (2004), “Vrednovanje kakvoc´e izgleda odjec´e”. Tekstil, Vol. 53, No. 10, pp. 497-509
Urbanija, V. and Gersˇak, J. (2004) “Impact of the mechanical properties of nappa clothing leather on the characteristics of its use“, J. Soc. Leather Technol. Chem., Vol. 88, No. 5, pp. 181-90 Kocik, M., Zurek, W., Krucinska, I., Gersˇak, J. and Jakubczyk, J. (2005) “Evaluating the bending rigidity of flat textiles with the use of an Instron tensile tester”, Fibres Text. East. Eur., Vol. 13, No. 2(50), pp. 31-34. (available at: http://www.fibtex.lodz. pl/50_09_31.pdf) Jevsˇnik, S., Gersˇak, J. and Gubensˇek, I. (2005) “The advance engineering methods to plan the behaviour of fused panel”, Int. J. Cloth. Sci. Technol., Vol. 17, Nos 3/4, pp. 161-70
Maribor, Slovenia Faculty of Mechanical Engineering, University of Maribor, Smetanova ulica 17, SI-2000 Maribor, Slovenia, Tel: +386 2 220-7960, Fax: +386 2 220-7990, E-mail:
[email protected] Principal Investigator(s):Prof. Dr Sc. Jelka GERSˇAK, Department of Textiles, Laboratory for Clothing Engineering, Physiology and Construction of Garments Research Staff: Dr Sc. Jelka GERSˇAK, Dolores TRUCˇL, Dipl.Eng., MSc. Andreja RUDOLF
Introduce the new lubrication technology stretched PES filament sewing threads Other Partners: Academic None
Industrial TSP d.d. TOVARNA SUKANCEV IN TRAKOV, Maribor Project ended: September 2004
Project started: April 2003 Grant value: 8.379.335 SIT or. 34.990 ECU Source of support: Ministry of the Economy Keywords: Sewing thread, PES filament, Hot extension, Lubrication, Viscoelastic properties
The project was conducted in five phases. In the first phase processing and specific demands of the high-quality PES filament threads according to the automobile industry demands were defined. In the second phase were researched specific demands of the filament yarns for achieving high-quality threads and methods of the threads surface treatment were analysed. For this purpose the influence of the constructional parameters of the PES filament threads (twist direction, number of yarn and thread turns) on the mechanical properties was researched. Results of the analysis show, that the number of thread turns influence the significantly the thread breaking tenacity and breaking extension. Mathematical models for prediction the thread breaking tenacity and breaking extension as function of the number of thread turns and filament yarn turns were formed. In the third project phase different methods of surface treatment, respectively lubrication of the PES filament thread at simultaneously drawing in warm medium with the aim of improving the mechanical and surface properties of the existent
Research register
47
IJCST 17,6
48
PES threads were analysed. The aim of the research was to study the new technology of cold jet lubrication and simultaneous drawing of the thread in warm medium. It was found, that the wanted quality and smoothness of the thread influences the drawing parameters of PES filament thread in warm medium, as well as the correct position of the jet for cold lubrication and velocity of the lubricating agent. In the fourth project phase the influence of the amount and type of the lubricating agent, as well as the drawing parameters on smoothness of the jet lubricated filament threads was been analysed. The research results show that the drawing ratio has an important role. The increasing drawing ratio causes linear increase of the thread breaking tenacity and decrease of the breaking extension. In the fifth phase of the project the thread sewability on the basis of the thread mechanical and structural properties before and after the sewing process was analysed. The structural properties of the drawn threads were analysed with differential scanning calorimetry (DSC), dynamic mechanical analysis (DMA) and with determination of the birefringence. In the frame of the mechanical properties the behaviour of the PES filament threads at tensile loading was analysed. The mechanical properties before and after the sewing process, tensile load of the thread during the sewing process, the thread dynamic tension and the alteration of the mechanical properties of the threads and separated filament fibres from threads after the sewing process were analysed. It was found, that the degree of crystallinity influences the fibre mechanical and thermal properties. This is reflected in the breaking tenacity and extension, elasticity modulus, glass transition temperature and melting temperature. Results of the DMA analysis show that the drawing ratio has an influence on dynamic mechanical properties, namely, first of all in the region of b-relaxation transition, which gives us information about the fibre crystalline and amorphous phase order. The thread drawing process at increased temperature influences the relaxation properties of the fibres. With raising the drawing ratio the crystallinity degree increases, which decreases the intensity of b-relaxation. It was further found, that with the increasing of the drawing ratio the average factor of orientation increases. This is the condition for the increase of the breaking tenacity, modulus elasticity and decrease of breaking extension with increasing drawing ratio. After the sewing process the degree of crystallinity and average orientation decrease, while amorphous orientation increases for lower drawing ratios and decreases for higher drawing ratio. This is the effect of greater tensional and thermal loadings of the threads during the sewing process, which influences on the re-forming of the fibres’ structure properties. This is confirmed by the results of the thread’s mechanical properties before and after the sewing process.
Project aims and objectives The main objectives of this project are: .
To investigate the effect of the lubricant agent type and amount as well as the effect of lubrication technology on assuring reliability of the sewing process at minimal dynamic tension and alteration of the thread mechanical properties
.
To develop and introduce the new lubrication technology stretched PES filament sewing threads.
Basic purpose of the introduction of the new lubrication technology of drawn PES filament threads is to the improve processing properties and resistance of the drawn PES filament threads against the dynamic and thermal loadings in the sewing process at minimal mechanical deformations, which is the condition for the process stability and assurance of the aimed seam strength.
Research deliverables (academic and industrial) Important contribution regarding the academic as well as industrial point of view can be seen as new cognitions of the influence of thread’s constructional parameters, hot extension and lubrication technology on change of mechanical and structural properties of the thread and twisted filament yarns. Achieved cognitions can be considered as starting points for projecting the sewability of filament threads and their applied properties. The new lubrication technology stretched PES filament sewing threads was adopted by industry. The new technology of lubrication of drawn PES filament threads assure the desired processing properties and greater resistance of the drawn PES filament threads against dynamic and thermal loadings in the sewing process at minimal alteration of the mechanical properties, which is the condition for the process stability and assurance of the aimed seam strength. Publications Rudolf, A. and Gersˇak, J. (2002) “Influence of twist on the mechanical properties of sewing thread”, Proceedings of the 1st International Textile, Clothing & Design Conference ITC&DC, Dubrovnik, Faculty of Textile Technology, University of Zagreb, Zagreb, pp. 395-400. Rudolf, A. and Gersˇak, J. (2003) “The planning of thread mechanical properties”, Proceedings of the 4th International Conference Innovation and Modeling of Clothing Engineering Processes IMCEP 2003, Faculty of Mechanical Engineering, Institute for Textile and Garment Manufacture Processes, Maribor, Slovenia, pp. 252-8. Rudolf, A., Nagy, P., Gersˇak, J. and Sfiligoj-Smole, M. (2005), “Dynamic mechanical analysis of drawn thread”. Proceedings of the 5th World Textile Conference AUTEX 2005, Portorozˇ, 27-29 June. Book 2. Faculty of Mechanical Engineering, Department of Textiles, Maribor, pp. 841-6. Rudolf, A. and Gersˇak, J. (2005), “The influence of thread twist on alteration of fiber mechanical properies”, Textile Research Journal (in press)
Maribor, Slovenia Faculty of Mechanical Engineering, University of Maribor, Smetanova ulica 17, SI-2000 Maribor, Slovenia, Tel: +386 2 220-7960, Fax: +386 2 220-7990, E-mail:
[email protected] Principal Investigator(s):Prof. Dr Sc. Jelka Gersˇak, Department of Textiles, Laboratory for Clothing Engineering, Physiology and Construction of Garments Research Staff: Dr Sc. Jelka GERSˇAK, Dr Vili BUKOSˇEK, Dr Dunja SˇAJN, MSc. Rozalija BLEKACˇ
Research register
49
IJCST 17,6
50
Study of the relationship between deformation and relaxation of fabrics containing elastane yarns in the spreading process Other Partners: Academic
Industrial ELKROJ, modna oblacˇila d.d., NAZARJE
Faculty of Natural Sciences and Engineering, Department of Textiles, University of Ljubljana Project start Date: July 2004 Project end Date: June 2007 Grant value: 4.051.470 SIT or 16.890 ECU for 2004 Source of support: Ministry of Higher Education, Science and Technology Keywords: Fabrics containing elastane yarns, Relaxation phenomena, Stress relaxation, Deformation, Spreading process The aim of the research project is to study the elastic behaviour of fabrics with elastane yarns taking into account the elastic potential, elastic deformations and relaxation ability based on the study of fabric mechanics, with the purpose to gain new cognition about the relationship between elastic deformations and relaxation phenomena in such fabric type. The main part of the research is focused on the study of elastic deformations and relaxation phenomena in fabrics with elastane yarns, where the influence of the parameters of tensional-elastic properties of such fabrics on elastic deformations and relaxation level will be investigated taking into account loading intensity. Special attention is given to the study of time dependence of relaxation phenomena in fabrics with elastane yarns and special relaxation conditions of such fabrics during spreading and laying into fabric layers, i.e., influence of the length, surface friction between fabric layers and number of layers. For this purpose the stress relaxation during witholding under constant deformations and the relaxation of fabrics after manually unwinding and spreading were studied. The relaxation of fabrics represents as elastic recovery and response of fabric as the result of loading during winding and spreading to form by lay. Based on given results of relaxation of fabric as the consequence of winding process was found that with higher percentage of containing elastane in the yarn grows the degree of deformation.
Project aims and objectives The main goal of the research is to find the correlation between the elastic deformation and relaxation level in fabrics with elastane yarns and to set up the model for predicting the relaxation of fabrics taking into account the parameters of fabric’s tensional-elastic properties, parameters of the spreading process, as well as the length and number of fabric layers. Publications Gersˇak, J., Sˇajn, D. and Bukosˇek, V. (2005) “A study of the relaxation phenomena in the fabrics containing elastane yarns”. Int. J. Cloth. Sci. technol., Vol. 17, Nos 3/4, pp. 188-99.
Sˇajn, D. and Gersˇak, J. (2004) “ITMA 2003 – merilne naprave za merjenje nateznih obremenitev tekstilij ¼ ITMA 2003”, “Instruments for tension measuring of textile products during production process”. Tekstilec, Vol. 47, Nos 3/4, pp. 136-44. Sˇajn, D., Gersˇak, J. and Bukosˇek, V. (2004) “A study of the relaxation phenomena of fabrics containing elastane yarns”. Book of proceedings of 2nd International Textile, Clothing & Design Conference Faculty of Textile Technology, University of Zagreb, Dubrovnik, Croatia, Zagreb, pp. 605-10. Sˇajn, D., Gersˇak, J. and Bukosˇek, V. (2005) “Correlation between the relaxation and deformation of fabrics containing elastane yarn”. Proceedings of 5th World Textile Conference AUTEX 2005, Faculty of Mechanical Engineering, Department of Textiles, Maribor, Portorozˇ, Slovenia, pp. 406-11
Mulhouse, France Ecole Nationale Supe´rieure des Industries Textiles de Mulhouse, Universite´ de Haute-Alsace, 11 rue Alfred Werner, F 68093 Mulhouse, France, Tel: +33389336320; Fax: +33389336339; E-mail:
[email protected] Principal Investigator(s): L. Schacher (ENSITM France), E. Strazdiene (KTU – Lithuania), Laboratoire de Physique et Me´canique Textiles FRE CNRS 2636 Research Staff: ensitm L. Schacher – D. Adolphe M. ISSA – S. BenSaid – JY. Dre´an KTU E. Strazdiene – M. Gutauskas – L. Naujokaityle – L.Valatkiene
Handle and tactile evaluation of fabric by sensory evaluation and instrumental measurements Other partners: Academic Department of Clothing and Polymer Products Technology (APGTK)
Industrial Faculty of Design and Technologies, Kaunas University of Technology (KTU), Donelaicˇio g. 73, LT 51424 KAUNAS – LITHUANIA
Project started: January 2005 Project ends: December 2006 Source of support: Exchange programs “Gilibert” (PAI) between French and Lithuanian research laboratories Keywords: Sensory analysis, Griff Tester, Tactile feeling Historically textile and clothing industries are producing products in appropriate qualities to satisfy consumer needs. In this regard handle is the most fundamental attribute that determines weather or not a particular fabric is suitable for a given enduse, ant that it often determines the commercial success or failure of textile manufacturing processes or products. Textiles can differently be evaluated and its sensory perception can be described by its mechanical properties. In this sense ENSITM (France) has a great experience in
Research register
51
IJCST 17,6
52
fabrics sensory (tactile) evaluation, while KTU (Lithuania) in creating testing devices which are much closer to real fabric loading conditions than most of the existing equipments. The aim of this research is to create reliable basis of textile hand evaluation and to lay foundation for its further standardisation. For this the following will be done: .
sensory evaluation of fabrics belonging to different categories
.
investigation of some parameters effect upon the variation of fabrics sensory properties;
.
investigation by the original devices of KTU,
.
setting of relationships between measurements and sensory parameters,
.
evaluation of sensitiveness threshold and saturation in the terms of touch.
It is evident that joint research will produce a scientific base for the control of fabric quality and performance as a result of new process and product development. This study will be valuable in the sense of economy, because it would allow to define the best fabric that the industries can use to get a specific effect in order to diminish the costs.
Project aims and objectives The aim of this research is to create reliable basis of textile hand evaluation by the joint research of two participating institutions and to lay foundation for its further standardisation. For this the following must be done: .
subjective evaluation of fabrics belonging to different categories of finishing, e.g. easy-care, anti-crease, etc.;
.
investigation of constructions parameters or of finishing treatment effect upon the variation of fabrics sensory properties;
.
investigation of fabrics of different categories of finishing by the original devices of biaxial loading i.e. by punching device and by KTU-Griff-Tester, which operates on the principle of specimen pulling through a rounded hole;
.
setting of relationships between fabrics objective characteristics obtained by KTU-Griff-Tester and between its subjective parameters obtained by sensory evaluation, e.g. between complex criterion and finishing;
.
evaluation of sensitiveness threshold and saturation in the terms of touch.
Research deliverables (academic and industrial) .
new knowledge of textiles hand and its subjective evaluation that will be included in study programmes of master students;
.
it will provide new research themes for master and doctoral students of Universities;
.
it will extend the existing testing base of Universities laboratory by new testing methods and the interpretation of the obtained results;
.
new knowledge of subjective evaluation will be disseminated for textile and clothing industries of Lithuania providing them new possibilities in their competitiveness, because the laboratory will be able to give qualified consultation services.
Besides all mentioned above, the results of joint research will be spread not only by the publications in the journals with the referee, but also in such well known scientific forums as European conferences organised by AUTEX (Association of Universities for Textiles) and Fibre Society conferences organised in US Publications Philippe, F., Schacher, L., Adolphe, D. and Dacremont, C. (2003), “The sensory panel applied to textile goods: a new marketing tool”, Journal of Fashion Marketing and Management, Vol. 17, No. 3, pp. 235-48. Philippe, F., Abreu, M.J., Schacher, L., Adolphe, D. and Silva, M.E. (2003), “Influence of the sterilisation process on the tactile feeling of surgical gowns”, International Journal of Clothing Science and Technology, Vol. 15, Nos 3/4, pp. 268-75. Philippe, F., Schacher, L., Adolphe, D. and Dacremont, C. (2004), “Tactile feeling: sensory analysis applied to textile goods”, Textile Research Journal, Vol. 74, No. 12, pp.1066-72. Philippe, F., Schacher, L., Dacremont, C. and Adolphe, D. (2001), “Sensory analysis: State of Art. application to textile evaluation” International Textile Congress Fibres, yarns, fabrics, finishing, management, innovations, Terrassa (Espagne), Actes, 18-20 June, pp. 216-22. Philippe, F. Schacher, L. and Adolphe, D. (2003), “Characterisation of different finishing treatments using tactile sensory analysis” paper presented at The Fiber Society Spring 2003 Symposium, Loughborough University, Actes 30 June-2 July, pp.11-12. Philippe, F. (2001), “Contribution a` l’e´valuation tactile des produits textiles par analyse sensorielle”, the`se de doctorat en Sciences pour l’Inge´nieur, Universite´ de Haute Alsace, Mulhouse (France), January. Chollakup, R. (2003), “Me´langes soie-coton en filature fibres courtes: caracte´ristiques des fils et analyse sensorielle des tricots”. The`se de doctorat en Sciences pour l’Inge´nieur, Universite´ de Haute Alsace, Mulhouse (France), January. Strazdiene˙ E. and Gutauskas, M. (2003), “Behaviour of stretchable textiles with spatial loading”, Textile Research Journal, Vol. 73, No. 6, pp. 530-4 Daukantiene˙, V., Papreckiene˙, L. and Gutauskas M. (2003), “Simulation and application of the behaviour of a textile fabric while pulling through a round hole”, Fibres and Textiles in Eastern Europe,, Vol. 11 No. 2, pp. 38-42 Strazdiene˙, E., Daukantiene˙, V. and Gutauskas, M. (2003), “Bagging of thin polymer materials: geometry, resistance and application”, Materials Science (Medz˘iagotyra), Vol. 9 No. 3, pp. 262-5 Strazdiene˙, E., Domskiene˙, J. and Gutauskas, M. (2003), “New method for the evaluation of coated textiles performance properties”, Proceedings of the World Textile Conference, (Third AUTEX Conference). Gdansk, pp. 118-21
Philadelphia, USA Philadelphia University, Computer Information Systems, Philadelphia PA, Tel: 215-951-5356; E-mail:
[email protected] Principal investigator(s): Prof L Sztandera, Computer information systems Research staff:
Research register
53
IJCST 17,6
54
Genetic algorithms in molecular design of novel fibers Other partners: Academic
Industrial
Prof Chih-Chung Chu, Biomedical Engineering, Cornell University Dr Hugh Cartwright, Chemistry Department, University of Oxford Project started: 1 May 2005 Project ends: 30 April 2006 Grant value: $ 155,000 Source of support: National Textile Centre Keywords: Polymer, Genetic Algorithm, Artificial Intelligence, Neural network, Inverse problem, Commercial value To design novel polymers computationally we are using two techniques from within the field of Artificial Intelligence. With a Neural Network, we aim to develop a tool that is able to predict the properties of a given copolymer or homopolymer from its structure, thereby solving what is known as the forward problem. The second technique is the Genetic Algorithm, which solves the inverse problem by forming the heart of the search procedure whose role is to find the optimum formulation. These two methods are combined together in a collaborative manner, to form a paired algorithm in which the neural network is used in the calculation of the genetic algorithms’ fitness function; the tool created in this way is known as a Hybrid Intelligent System. Object-oriented design was chosen for the development of the software tool, rather than a functional approach, because it is considered to produce more maintainable and easily understood system architecture and code. Initial work has focused on the second of the two tools required for the Hybrid System, the Genetic Algorithm (GA). The present formulation of the GA comprises two sub-systems: a GA engine with five fitness functions and a graphical user interface (GUI) front end which provides the user with an easy route into the functionality of the GA engine. The design methodologies used for the production of the GA are analogous to the evolutionary incremental software engineering model. Starting from a basic genetic algorithm, input and output functionalities have been added to yield a checked test-bed model which will be used as the basis of the polymer modeling system. We are developing an extensive structure-property database to aid in designing fibers with improved properties, such as stretch, strength, bulk, comfort and dyeability.
Project aims and objectives Our goal is twofold: (i) to establish an extensive structure-property correlation database library; (ii) to develop computational techniques that can draw upon data within this library to predict the formulation for a range of novel polymers and fibers with desirable characteristics.
Research deliverables (academic and industrial) Computer software capable of predicting the properties of previously unknown polymers, allied to a supporting database.
Publications NTC FY (2004) (New Project Proposal (accessible through the NTC web site, www.ntcresearch.org/) “Genetic Algorithms in molecular design of novel fibers”. NTC Research Project C04-PH02s Report June 2005. (accessible through the NTC web site, www.ntcresearch.org/)
Research register
55
Selkirkshire, UK School of Textiles and Design, Heriot-Watt University, Netherdale, Galashiels, Selkirkshire TD1 3HF. Tel: 01896 89 2135; Fax: 01896 75 8965; E-mail:
[email protected] Principal investigator(s): Prof. George K. Stylios and Dr R.M. Christie, Research Institute for Flexible Materials (RIFleX) Research staff: Dr Taoyu Wan, Dr Manzoor Arain, Dr Xiaoming Zhao
Engineering the performance and functional properties of technical textiles: the effect of manufacturing parameters on the performance of technical textiles Other partners: Academic
Industrial
University of Manchester University of Leeds
IQ Textiles TechniTex Faraday Ltd
Project started: December 2002 Project ends: November 2005 Grant value: £358,000 Source of support: Engineering and Physical Sciences Research Council (EPSRC) Keywords: Coating, Optimization, Variables The functionality and performance of end products can be achieved by establishing the precise machine settings and optimisation of process variables during mechanical and chemical processing. Consequently, manufacturing efficiency and product quality is dependent upon knowledge of the interaction between manufacturing conditions, material properties and end product requirements. A key feature in the manufacture of technical textiles is functional setting. This comprises the processes for add-on new functionality to fabrics. These include coating and laminating processes for functionality such as liquid and stain resistance, anti microbial properties and biocompatibility, and for introducing appropriate mechanical performance, comfort and aesthetic requirements to finished fabrics. This is a part of a £1,000,000 EPSRC funded project “Engineering the performance and functional properties of technical textiles”, with the University of Leeds and Manchester under the TechniTex Faraday Partnership in Technical Textiles.
IJCST 17,6
56
Coating is the focus activity of this research, which integrates with the research done at Leeds and Manchester in the overall research programme. The research will concentrate on the investigation of the coating processes, taking into account their relationship with fabric, i.e. investigating the fabric/coating process interface. Coating effectiveness depends upon the structural, physical and chemical properties of the material. This is the focus of this part of the research, which will determine the interaction of physical and chemical properties of a given fabric on coating in one hand and the optimum coating variables for the effective performance treatment of fabrics in the other. Coating is selected for this investigation because, being the final stage of manufacture, it has a major effect on the performance attributes of a technical textile product and it provides the important added value for their end use.
Project aims and objectives .
Establish coating parameters for end users.
.
Establish functional criteria based on coating and how it is measured.
.
Establish fabric/coating process interactions. Optimisation of mechanical and physical properties of textile fabrics.
. .
Provide a manufacturing system for the effective coating of textile products.
.
Establish manufacturing specifications.
.
Linkage of modelling and measurement with coating treatment Verified interactions with exemples
. .
Optimise coating requirements
A new coating methodology specifically focusing on micro/nano porous coating is also investigated, based on phase separation, which is to yield controlled pores which can be used as payload delivery systems.
Research deliverables (academic and industrial) .
Measurement methods for fabric properties
.
Generic sensory tools
.
Optimum fabric coating conditions Modelled performance/properties/machine conditions
. .
Relationship between performance and manufacture
.
Constructed fuzzy logic model Micro/nano porous coating
. .
Manufacturing interface and predictive right first time manufacturing system
.
Final report
Publications Wan, T. (2005), “Porous polymer membranes from microphase separation”, IMACS 2005 World Congress, Paris, 11-15 July.
Stylios, G.K. (2005), “Optimisation of commercial coatings for technical textiles”, Technical Textiles International, pp. 19-22. Stylios, G.K. (2005), “Engineering the performance and functional properties of technical textiles”, International Seminar on Functional Clothing and Intimate Apparel, Hong Kong, May. Stylios, G.K. (2004), “Technitex core research”, Presentation at the House of Lords, London October. Wan, T., Stylios, G.K., Han, F. (2004), “Functionality development in technical fabric coating for medical application”, Proceedings of the Textile Institute 83rd World Conference, Shanghai, 23-27 May.
Selkirkshire, UK School of Textiles and Design, Heriot-Watt University, Netherdale, Galashiels, Selkirkshire TD1 3HF, Tel: 01896 89 2135, Fax: 01896 75 8965, E-mail:
[email protected] Principal Investigator(s):Prof. George K. Stylios, Research Institute for Flexible Materials (RIFleX) Research Staff: None
Interactive wireless and smart fabrics for textiles and clothing Other Partners: Academic
Industrial
None None Project started: September 2002 Project ends: September 2006 Grant value: None Source of support: Worshipful Company of Weavers Keywords: Smart, Interactive, Textiles, Garment, Clothing, Sensors, Wireless The last few years have witnessed an increased interest in wearable technologies, smart fabrics and interactive garments. This has come about by certain technological innovations in the areas of sensor-based fabrics, micro devices, wire and wireless networks. In terms of textiles, most of current developments are towards the fashion markets and have resulted in glorifying garments as gimmicky gadgets. However, some efforts are also being directed in using the technology for improving the quality of life, or even for life saving purposes. Examples of such uses can be found in the military, healthcare, fire fighting, etc. This research project investigates new interdisciplinary technologies in fabrics, sensors and wireless computing, for the development of a prototype interactive garment for monitoring various functions of the wearer.
Research register
57
IJCST 17,6
58
Project aims and objectives The general aim of the project is to develop technologies for use in interactive garments, which can provide monitoring functions for various applications such as the clinical or healthcare sector. More specifically, objectives are: .
.
Develop suitable wireless sensors for various measurements, including ECG, temperature, breathing, skin conductivity, mobility and movement, humidity, positioning, etc. Develop a Personal Area Network and a Wireless Communication Centre
.
Optimise suitable wireless technologies such as Bluetooth to enable communication between sensors and a central processing unit
.
Conceptualise a smart multilayer fabric Integrate technologies
.
Research deliverables (academic and industrial) . .
.
Wireless sensors for physiological and other measurements Wireless communication centre for relaying information between sensors, wearers, central processing unit and Internet Conceptual multilayer fabric suitable for interactive garments
Publications Stylios, G.K. and Luo, L. (2003) “Investigating an interactive wireless textile system for SMART clothing”, paper presented at 1st International Textile Design and Engineering Conference (INTEDEC 2003), Fibrous Assemblies at the Design and Engineering Interface, Edinburgh, 22-24 September. Stylios, G.K. and Luo, L. (2003) “The concept of interactive, wireless, smart fabrics for textiles and clothing”, paper presented at 4th International Conference, Innovation and Modelling of Clothing Engineering Processes – IMCEP 2003, Maribor, Slovenia, 9-11 October. Stylios, G.K. and Luo, L. (2004) “A SMART wireless vest system for patient rehabilitation”, Wearable Electronic and Smart Textiles Seminar, Leeds, 11 June. Stylios, G.K., Luo, L., Chan, Y.Y.F. and Lam Po Tang, S. (2005) “The concept of smart textiles at the design/technology interface”, paper presented at 5th International Istanbul Textile Conference, Recent Advances and Innovations in Textile and Clothing, Istanbul, 19-21 May.
Selkirkshire, Scotland Heriot-Watt University, Netherdale, Galashiels TD1 3HF, Selkirkshire, Scotland. Tel: 01896 892136; Fax: 01896 758965; E-mail:
[email protected] Principal Investigator(s): Professor G.K. Stylios, Research Institute for Flexible Materials Research Staff: Dr Fan Han, Dr Hua Lin, Mr Liang Luo
HOMETEX: a virtual trading centre for textiles Other partners: Academic None
Research register
Industrial OCF Ltd, Silicon Graphics Inc., Scottish Enterprise Borders, Scottish Textiles Manufacturers Association, Borders Textile Forum Project ends: 28 February 2006
Project started: 1 September 2001 Grant value: £1,000,000 Source of support: EU ERDF Objective 2 Keywords: Drape, Augmented reality, Virtual trading, Home shopping, 3D simulation, Dynamic draping In recent years we have witnessed a revolution in networking of information on a global scale via the Internet. Many companies have capitalized on this provision and have used it in many diverse ways, from electronic mailing to marketing, selling and trading of products and services. Marketing and selling of limp products such as textiles and, particularly, garments using new multimedia techniques would be extremely beneficial to the industry, since it would enable companies to reduce product to market, to enhance product development through 3D visualization and to trade directly without the intervention of retailers. But selling of garments is not as easy as selling other commodities; garments are made of limp materials which take up the configuration of the wearer; customers would in most cases like to wear the garment, or, in the case of buyers, see the garment worn by a model. For the effective exploitation of these possibilities, we should, therefore, develop a multimedia environment to enable the simulation of drape behavior of garment designs on virtual models that may resemble real customers. The textile industry chain, being a traditional industry, is very conservative in the use of multimedia for manufacture, advertising and/or sales. The reason is firstly because the industry consists of small companies, which do not have the resources to use multimedia technologies effectively without training, and secondly there is no technological infrastructure available to realistically visualize new products from home. This project aims to pilot such possibilities, which have other benefits to this industry in terms of “just-in-time” manufacture, 3D visualization of new products and better communication with their customers.
Project aims and objectives The main objectives of this scheme are as follows: .
To develop a virtual Home Trading Centre for the textile, clothing and retailing industries: “HomeTex”.
.
To enable the production of virtual fashion shows for buyers through CD-ROM and Internet presentations.
59
IJCST 17,6
.
.
60
.
To network companies with customers (directly) via the technology and to regularly upgrade and manage company trade data, for piloting the technology. To establish and provide through “HomeTex other trade data, real-time electronic mail, Tele Trading and, possibly, banking”. To enable textile and clothing companies to interface with this technology so that new products can be made much faster and to minimize energy, raw materials and other resources.
Research deliverables (academic and industrial) A Virtual Trading Centre in Textiles operating from the Borders of Scotland. Publications Stylios, G.K., Han, F. and Wan, T. (2004), “The integration and development of new computer vision technologies for global retailing”, Proceedings of the 83rd Textile Institute World Conference, Shanghai, 23-27 May. Stylios, G.K., Han, F. and Wan, T. (2003), “3D modeling, simulation and visualization of ‘Virtual Wearer Trials’ with digitally cloned humans”, Proceedings of the 4th International Conference Innovation and Modelling of Clothing Engineering Processes – IMCEP 2003, Maribor, Slovenia, 9-11 October. Stylios, G.K., Han, F. and Wan, T. (2003), “Modelling real humans for virtual garments”, Proceedings of the 1st International Textile Design and Engineering Conference (INTEDEC 2003), Fibrous Assemblies at the Design and Engineering Interface, Edinburgh, 22-24 September. Stylios, G.K. and Wan, T.R. (1998), “A new collision detection algorithm for garment animation”, International Journal of Clothing Science and Technology, Vol. 10 No. 1 pp. 38-49. Stylios, G.K. and Zhu, R. (1998), “The characterization of static and dynamic drape of fabrics”, Journal of the Textile Institute, Vol. 88 No. 4, pp. 465-75. Stylios, G.K., Wan T.R. and Powell, N.J. (1995), “Modeling the dynamic drape of fabrics on synthetic humans: a physical limped parameter model”, International Journal of Clothing Science and Technology, Vol. 7 No. 5. pp. 10-25. Stylios, G.K., Wan, T.R. and Powell, J.J. (1996), “Modeling the dynamic drape of garments in synthetic humans in a virtual fashion show”, The International Journal of Clothing Science and Technology, Vol. 8 No. 3, pp. 44-55.
Selkirkshire, UK School of Textiles and Design, Heriot-Watt University, Netherdale, Galashiels, Selkirkshire TD1 3HF, Tel: 01896 89 2135, Fax: 01896 75 8965, E-mail:
[email protected] Principal Investigator(s):Prof. George K. Stylios Research Institute for Flexible Materials (RIFleX) Research Staff:Mr Mohammad Mahfuzur Rahman Chowdhury
Investigating nano fibre production by the electrospinning process Other Partners: Academic
Industrial
None None Project started: July 2004 Project ends: July 2007 Grant value: Source of support: Keywords: Electrospinning, Electrospinning process, Parameters, Polymer, Nanofibre application Electrospinning is a unique way to produce novel polymer nanofibres with diameter typically in the range of 10 nm to 500 nm. Using this process, the polymer nanofibres can be made from a variety of polymer solutions or melt to produce fibres for a wide range of applications. Electrospinning occurs when the electrical force at the surface of a polymer solution or melt overcomes the surface tension and causes an electrically charged jet to be ejected. When the jet dries or solidifies, an electrically charged fibre remains. This charged fibre can be directed or collected or accelerated by electrical forces, then collected in sheets or other geometrical forms. This research project is an investigation of the electrospinning process and the effect of process variables on orientation, crystallinity, microstructure and mechanical properties of the nanofibres produced. Some of the polymeric parameters investigated are polymer type, solvent type, molecular weight, solution properties, viscosity, conductivity and surface tension. In the case of process parameters, the electric potential, flow rate, concentration, distance between capillary and collection screen, ambient parameters are important.
Project aims and objectives .
.
To investigate process-structure-property relationships in polymer fibres with nanosize diameters produced by electrospinning To systematically investigate the process in stages: jet initiation, jet elongation (bending instability) and nanofibre formation
.
To investigate the morphology and properties of the polymer nanofibres
.
To produce fibres of uniform diameters
Research deliverables (academic and industrial) .
Nanofibres of uniform diameter
.
Defined mechanical and physical properties Process-structure-property relationships
. .
Research register
Detailed understanding of the electrospinning process
61
IJCST 17,6
.
Nanofibres suitable for applications such as air filtration, protective clothing, fibre reinforced support, and other technical fabrics
Publications None
62
Shropshire, UK Rapra Technology Ltd, Shawbury, Shrewsbury, Shropshire, SY4 4NR, Tel: +44 (0)1939250383, Fax: +44 (0)1939 251118, E-mail:
[email protected] Principal Investigator(s):Chris Hare, Large Projects Research Staff: R. Venables, S. Wallace
BIOCOMP Other Partners: Academic Upper Austria Research, Lulea Univesrity of Technology, Wroclaw University of Technology
Industrial Fraunhofer ICT, Risoe, Gaiker, Celabor, VTT, APC Composites, BAFA, Transfurans chemicals, PJH, Chemont, Tecnaro, MERL, Net composites, Tehnos, Griffner, MEDOP, Haidlmair, Fiedler, Ekotex, National Institute of wood
Project started: 1 April 2005 Project ends: 1 October 2008 Grant value: e6.5 million Source of support: EU, Integrated Project, FP6 Project title: New Classes of Composite Engineering Materials from Renewable Sources Project acronym: BIOCOMP Project summary: Future product design requires sustainable processes and eco-innovation in material development for engineering applications. The innovative approaches use new engineering materials – biocomposites and their development has to be knowledge-based, whereas predominant issues are resource saving, variability in properties and functionality, light weight, low costs and eco-efficiency in all stages of the product life cycle. The main objective of this project is to obtain a breakthrough for SMEs on the development and use of engineering thermoplastic and thermosetting materials mainly from natural resources, like lignin from the paper industry and from the High Presure Hydrothermolyses (HPH) process, other biopolymers (here referred as biopolymers: e.g. Polylactide, Polyhydroxy-butyrate, Starch), furan resins, woven and non-woven cellulose fibres and fibre mats to final model products. The technical work programme will comprise the complete technical path from the input of natural raw materials (fibres, polymers and natural additives) to the output of final top quality engineering composite materials and model products (e.g. housings for
electronic equipment, car front end interiors, glass frames etc.) with an environmentally friendly life cycle. In parallel, there are activities concerning standardisation of characterisation and test procedures and quality control. Demonstration by the model products supporting the dissemination and the exploitation of results will exhibit the benefits of the materials and deliver a first input to material databases. An integrated concept of sustained skill and education of staff and students will provide routines and access to the material data. It includes the most interesting approaches of all current developments for engineering biocomposites. Innovative additives will provide flame retardancy and colouring.
Project aims and objectives To produce a range of “hi tech” composite panels using a variety of natural fibres and natural resins.
Research deliverables (academic and industrial) Many deliverables in the project, some of which are: Reports on: raw material characterization, data and tolerances. Compounding of materials Test data Safety and emissions Environmental benefits Economic evaluation Sample tools and parts Demonstrator parts Training, exploitation and dissemination. Publication None
Tehran, Iran Amirkabir University of Technology, Hafez Ave., Tehran. Iran, Tel: +98216406673; Fax: +98216406673; E-mail:
[email protected] Textile Engineering Department Research Staff: Dr F. Dadashian, Dr M. A. S. Monfared, Dr H. Riahi
Design of a new quality assessment for textile product using fuzzy Tauguchi functions Other partners: Academic None Project started: 2000 Finance/support: 17000 $
Industrial Project ended: 2002
Research register
63
IJCST 17,6
64
Source of support: Iranian Standard and Industrial Research Organization Keywords: Textile quality assessment, Multiple criteria decision making (MCDM), Fuzzy function, Taguchi loss function, AHP This project studies on the development of a new multiple criteria decision by a making (MCDM) methodology to assess textile products using experimental data gathered from Standard Organization of Iran. A quality attribute is characterized fuzzified taguchi quality function and integrated into a hierarchical structure to enable a true assessment of the complex concept of product quality. The theoretical aspects of the method are discussed and experiments are conducted to validate the method using real data.
Project aims and objectives The aim of this project is related to present a new methodology for quality assessment of textile products.
Research deliverables (academic and industrial) None Publications “Design of a new quality assessment system using fuzzy Tauguchi functions”, paper presented at 17th IMACS World Congress, Paris, 11-15 July. “Numerical quality assessment of textile products”, Published in the Proceeding of the 5th Iranian National Textile Engineering Conference 2004, Tehran. “An investigation of the effect of different strategic programs of Iranian standard and industrial research organization on the product quality”, Published in the Proceeding of the 6th Congress on Government, University and Industry Cooperation For National Development, 17-18 December 2002, Tehran.
Research index by institution
Index by institution
Institution
Page
Amirkabir University of Technology, Iran
63-64
Aurel Vlaicu University, Romania Budapest University of Technology and Economics, Hugary Ege University, Turkey
6 17-18 10-11, 35-36
Ghent University, Belgium,
24-33
Heriot-Watt University, UK
23-24, 55-61
Institut Supe´rieur des Etudes Technologiques de Ksar-Hellal, Tunisia
39-40
National Technical University of Athens, Greece
6-7, 9
Philadelphia University, UK,
53-55
Rapra Technology Limited, UK
62-63
Technical University of Lodz, Poland
40-41
The Research-Development National Institute forTextile and Leather, Romania
11-17
Universite´ de Haute-Alsace, France
51-53
University of Agricultural Sciences, India
18-24
University of Maribor, Slovenia
42-50
65
Research index by country
IJCST 17,6
66
Country
Page
Belgium
24-33
France
51-53
Greece
6-9
Hungary
17-18
India
18-24
Iran
63-64
Poland
40-41
Romania
6,11-17
Slovenia
42-50
Tunisia
39-40
UK
10-11, 33-39
USA
62-63
Research index by subject Subject
Page
Nonofibres, electrospinning
3D graphics, reconstruction, textile simulation, fabric drape, home shopping
41, 59
Objective measurement, sensory evaluation handle, tactile properties, mechanical modelling, textile fabrics, fabric properties, mechanics
Anthropometrical measurements, sizing Distance learning, networking and training, quality management
18 9, 28, 29
6-10, Dyeing and finishing, biological 14, 33, treatment, biodegradability, 37, 38, biotreatment, water recycling, 42 bioflotation, textile dyeing, reactive dyes scouring, cotton, pigment and reactive printing biotechnology, biopolishing, enzyme modifications bioscouring, wool-antifelting, cotton EU regulation, heavy metal detection textile waste, composite materials, renewable sources, electrochemical sensor, chlorophenols, dioxins
6, 12, 30, 62
Fibres, animal based fibres; metha, agave, hemp, pina fibres, molecular design, genetic algorithms
20, 54
Quality assessment, Tauguchi function Sewing threads, mechanics, fabric, clothing, comfort Spreading, lubrication, viscoeleastic properties, elastane yarn, relaxation Technical textiles, conductive textiles solar cells, coatings, plasma SMART textiles, clothing, fabrics, wireless cooling and healing, textiles peltier elements functional textiles, cotton, chemical modification, artificial turf, fabric coating, optimisation, technical textiles, filtration
Index by subject
62 7, 35, 38, 39, 45, 47, 50, 51 63 39, 45, 47 50
8, 11, 17, 23, 24, 26, 55, 57, 61, 62
Weaving, air-jet looms, self-learning
32
Yarn spinning; ring-spun yarns, woven, knitted fabrics properties
35
67
Research index by principal investigator
IJCST 17,6
68
Principal investigator
Page
Thanos Peppas
9 9
Idil Arslan-Alaton
33
Anthony Primentas
Judit Borsa
17
Christopher Provatidis
6-7
R.M. Christie
55
Ivelin Rahnev
39
Maria Cybulska
41
Maria Rangoussi
7
Luliana Dumitrescu
12
L. Schacher
51
Kerim Duran
10
Gustaaf Schoukens
24
Ozlenen Erdemismal
36
Micaela Dina Stanescu
6,14
44,47-49
E. Strazdiene
51
Chris Hare
62
G.K. Stylios
55, 57, 58, 60
Carla Hertleer
25
Lieva Van Langnehove
26
Alenka Majcen Le Marechal
42
Savvas Vassiliadis
6-7
Jelka Gersak
Johanna Louwagie
28-29
Emilia Visileanu
Geeta Mahale
18-20
Philippe Westbroek
12,14 25, 27, 30-32
Ion Mituleasa
11
John Wilson
23
Arif Taner Ozguney
35
Abbas Yurdakul
38
