Development of multifunctional warming fabric part II: Multi-quality process parameter optimization of hygroscopic heating and antistatic function yarn

Abstract

This series study develops a multifunctional poly(ethylene terephthalate) (PET) fabric with far-infrared function and moisture absorption heating properties for warming clothing. In Part I, the far-infrared function yarn was made by a PET and nano-germanium composite to improve the properties of far-infrared emissivity and far-infrared heating. The moisture absorption heating PET yarn includes hygroscopic heating and antistatic properties. In Part II, this moisture absorption heating PET yarn is made by the hydrophilic PET, and the process parameters optimization for melt spinning is presented. First, in order to improve the hydrophilicity of PET, the modified PET is polymerized by terephthalic acid, ethylene glycol, polyethylene glycol oligomers, and sodium-5-sulfo-bis-(hydroxyethyl)-isophthalate. Then, the hydrophilic PET is made into a moisture absorption heating PET yarn, specified to a 75d/72f fully drawn yarn by melt spinning. The properties of this yarn, including hygroscopic heating, surface resistance, tensile strength, elongation at break and yarn count, are presented. In order to improve the properties of this yarn, the Taguchi method combined with grey relational analysis is used to obtain the optimal multi-quality process parameters for melt spinning. According to experiment results, this modified yarn has excellent hygroscopic heating, surface resistance and elongation at break. At the same time, it is shown that the warming fabric forms from Part I and this modified yarn can achieve far-infrared emission of 89%, far-infrared heating of 6.3 ℃, hygroscopic heating of 1.8 ℃ and surface resistance of 10⁷Ω.

Keywords

hygroscopic heating moisture absorption heating PET Taguchi method ANOVA grey relational analysis

The textile industry currently plays an important role in responding to the needs of the human body in relation to environment and climate. As the demand for better, more advanced textiles increases, the industry has turned its attention towards the development of functional fabrics.¹ The application of functional fabrics in winter clothing focuses mainly on hygroscopic heating and antistatic properties. The hygroscopic heating function uses the exothermic process of the human body. This process, commonly known as sweating, involves the evaporation of liquid sweat at the surface of the skin, which is capable of absorbing a large amount of heat. As the sweat evaporates, body temperature can drop significantly. When the sweat vapor condenses in vitro, a great deal of the heat of condensation is released. The hygroscopic heating fiber releases a large amount of heat as the enthalpy decreases in the hygroscopic process, increasing the temperature inside the clothing for thermal insulation.² The electrostatic phenomenon refers to the contact separation electrification of fabric, leading to nonuniform surface charge distributions. Positive charges or negative charges accumulate gradually, forming potential differences to the ambient environment. When charges transfer along different potentials, we refer to this as electrostatic discharge (ESD). In cold, dry environments, the electrostatic phenomenon can be a safety hazard, as it has the potential to cause fire or explosion.³ Shang and Zhang⁴ presented the heating mechanism of hygroscopic heating, demonstrating it can absorb the moisture given off by skin and transform it into heat energy. As moisture absorption increases, so does the heating effect. Pinar, Oleksiewicz and Wróbel⁵ illustrated that surface resistance will decrease when the relative humidity or hygroscopicity of a textile is enhanced. According to references, it is proven that the textile hygroscopic heating and antistatic properties will increase simultaneously when the hygroscopicity of the textile material increases.

Research on hydrophilic modification of PET material includes a variety of modification methods including: hydrophilic coating layer,^6–7 surface radical graft polymerization^8–9 and monomer polymerization.^10–12 Hsiao et al.¹⁰ presented the physics and kinetics of alkaline hydrolysis of cationic dyeable poly(ethylene terephthalate) (CDPET) and polyethylene glycol (PEG) modified CDPET polymers. The crystallinities of hygroscopic PET and PEG-modified PET polymers declined as the dimethyl 5-sulfoisophthalate sodium salt (SIPM) and PEG contents increased. The rate constant of the alkaline hydrolysis of hygroscopic PET and PEG-modified PET polymers increased with SIPM/PEG content and the molecular weight of the PEG increased. In the study of Miranda, Santos and Soares,¹¹ PET fabric modification was provided by using PEG and N,N′-dimethylol-4,5- dihydroxyethylene urea chemically modified resin, and the hydrophilicity of modified PET fabric was increased. Zhao et al.¹² presented the preparation and characterization of PET copolyesters modified with sodium-5-sulfo-bis-(hydroxyethyl)-isophthalate (SIPE) and PEG. The incorporation of PEG units with lower molecular weight led to more ether bonds and hydroxyl end-groups in molecular chains, showing PET copolyesters had better hydrophilicity with decreasing molecular weight of PEG units.

In this study, SIPE and PEG was copolymerized with terephthalic acid (TPA) and ethylene glycol (EG) to form a hydrophilic PET copolymer. The reaction mechanism is shown in Figure 1. In the copolymer, the SIPE segment provides an -O-Na functional group that enhances the hydrophilicity of the copolymer. However, as the content of the rigid benzene in the copolymer increases, the viscosity of the copolymer in the molten state increases, making the process of the copolymer more difficult.^13–14 PEG provides ether bonds and hydroxyl value to reduce the rigidity of the molecular chain and increase the process properties of the copolymer.¹² Therefore this form of modified PET is suitable for making an antistatic and hygroscopic heating yarn. In addition, the lower molecular weight of PEG oligomer results in higher hydrophilicity,^11–12 making it unlikely to hydrolyze. Therefore, the PET polymerization monomer is mixed with molecular weight-400 PEG oligomer and SIPE, so as to create a hydrophilic PET material. The clothing industrial specification 75d/72f fully drawn yarn is made by melt spinning and melt drafting, thus making a yarn with hygroscopic heating and antistatic function.

Figure 1.

The reaction mechanism of PET modification.

Melt spinning process

The melt spinning uses a screw to squeeze the molten polymer through a spinneret with many orifices at a constant rate, so as to produce continuous filaments. Typical melt spinning equipment includes a feeder, a melting device (extrusion zone and heating zone), a gear pump for quantification, a spinneret, a cooling thread eye and a fiber take-up system, as shown in Figure 2.^15–16

Figure 2.

The process of melt spinning.

During melt spinning, the polymer is fed into an extruder consisting of a screw used for melting and mixing. The polymer melt is then pumped through a spinneret. The extruded polymer is quenched with cold air, allowing the molten mass to solidify into filaments. Since the as-spun filaments lack adequate strength for industrial applications, the melt spinning is followed by the mechanical drawing of the extruded filament, resulting in alignment of molecular orientations along the filament axis. This will result in improved physical and mechanical characteristics.¹⁷ The process parameters of melt temperature, mold temperature, nozzle temperature, gear pump speed and take-up speed will influence the yarn qualities and the yarn forming in melt spinning.¹⁸

Process optimization

The process parameters of melt temperature, mold temperature, nozzle temperature, gear pump speed and take-up speed will influence yarn qualities and yarn forming during melt spinning. The process parameters design is important for the melt spinning process. The Taguchi method is a readily available and effective process parameter tool for providing a singular maximized quality. In comparison to traditional experimental methods, the Taguchi method reduces the number of experiments and confirms the effect of process parameters.^19–21 For the optimization of multiple product qualities, the Taguchi method can be combined with other theories for evaluation.^22–23 Pervez Mozumder and Mourad²⁴ presented an investigation on the optimization of the injection molding parameters of high-density PET/TiO₂ nanocomposites using grey relational analysis (GRA) along with the Taguchi method. Nine experimental runs were carried out based on the Taguchi L₉ orthogonal array. The data was processed according to grey relational steps. The optimal process parameters were found based on the average responses of the grey relational grades. Najarian et al.²⁵ illustrated the experiments about the optimization of converting process parameters based on the Taguchi orthogonal array design. The multi-objective optimization methods were employed to determine the optimum control factors. The result shows that GRA optimization had the minimum mean error percentage. Kuo et al.²⁶ applied the Taguchi method in planning a research experiment using the orthogonal array combined with the GRA to obtain the optimal parameters of the melt spinning process. It can be observed that GRA can sequence the effects of factors on quality with a grey relational coefficient. In collaboration with the Taguchi method, optimal process parameters for different product qualities can be found. Therefore, the Taguchi method is combined with GRA in this study to optimize melt spinning process parameters. The optimization design process is shown in Figure 3.

Figure 3.

Optimization design process.

This study uses the Taguchi L₁₈ orthogonal array to design the experiment, as shown in Tables 1 and 2. The full factorial experiment requires 216 (3⁶) times to determine the six parameters and three levels of the experimental design; while the Taguchi L₁₈ orthogonal array only requires 18 experiments, as it makes the orthogonality for each parameter and level. Every parameter level is discussed in the same times to present their quality effects. In comparison to the full factorial experiment, the orthogonal array can implement analysis with fewer experiments.^27–28

Table 1.

The parameters and levels design for experiment

Parameters		Levels
Parameters		1	2	3
A	PEG content (PEG/TPA weight ratio)	3/100	5/100	7/100
B	SIPE content (SIPE/TPA molar ratio)	2/100	3/100	4/100
C	Melt temperature (℃)	275	280	285
D	Mold temperature (℃)	285	290	295
E	Nozzle temperature (℃)	290	295	300
F	Gear pump speed (m/min)	1600	1920	2240
G	Take-up speed (m/min)	2700	3300	3900

PEG: polyethylene glycol; SIPE: sodium-5-sulfo-bis-(hydroxyethyl)-isophthalate; TPA: terephthalic acid.

Table 2.

Experimental parameters configuration in Taguchi orthogonal array

Exp. No.	Factors
	A	B	C	D	E	F	G
	PEG content (PEG/TPA weight ratio)	SIPE content (SIPE/TPA molar ratio)	Melt temperature (℃)	Mold temperature (℃)	Nozzle temperature (℃)	Gear pump speed (m/min)	Take-up speed (m/min)
1	3/100	2/100	275	285	290	1600	2700
2	3/100	3/100	280	290	295	1920	3300
3	3/100	4/100	285	295	300	2240	3900
4	5/100	2/100	275	290	295	2240	3900
5	5/100	3/100	280	295	300	1600	2700
6	5/100	4/100	285	285	290	1920	3300
7	7/100	2/100	280	285	300	1920	3900
8	7/100	3/100	285	290	290	2240	2700
9	7/100	4/100	275	295	295	1600	3300
10	3/100	2/100	285	295	295	1920	2700
11	3/100	3/100	275	285	300	2240	3300
12	3/100	4/100	280	290	290	1600	3900
13	5/100	2/100	280	295	290	2240	3300
14	5/100	3/100	285	285	295	1600	3900
15	5/100	4/100	275	290	300	1920	2700
16	7/100	2/100	285	290	300	1600	3300
17	7/100	3/100	275	295	290	1920	3900
18	7/100	4/100	280	285	295	2240	2700

PEG: polyethylene glycol; TPA: terephthalic acid.

In Table 1 and 2, the seven parameters and three levels are discussed by the Taguchi L₁₈ orthogonal array. The parameter factors selected are PEG content, SIPE content, melt temperature, mold temperature, nozzle temperature, gear pump speed and take-up speed. Hygroscopic heating, surface resistance, tensile strength, elongation at break, and yarn count are selected as the target qualities according to the required properties of the moisture absorption heating yarn. However, the detection of surface resistance can only be expressed in a hierarchical manner instead of continuous data, and non-continuous data can't be analyzed by the Taguchi method. As stated in the references, ^10–12 the fabric hydrophilicity enhances the hygroscopic heating and reduces surface resistance of fabric simultaneously, so the qualitative surface resistance will be validated in the finished yarn of multi-quality optimization.

This study uses the Taguchi method for single quality parameter optimization. The analysis of variance (ANOVA) is used to calculate the factor contribution of a selected quality and its respective weights to the other various desired qualities, combined with GRA for parameter optimization of multiple qualities. The method is described as follows.

Taguchi method

This study uses an orthogonal array to design the experimental parameters, and uses the S/N ratio to judge the experimental optimal parameters and significant experimental factors. The method is described below:

1. S/N ratio calculation. According to the quality of product, its S/N ratio could be defined as:

A. The large the better: hygroscopic heating, tensile strength and elongation.

S / N_{LTB} = - 10 log [\frac{1}{n} \sum_{i = 1}^{n} \frac{1}{{y_{i}}^{2}}]

(1)

where y_i is the test data for each experiment, n is the number of experiments.

B. Nominal the best: yarn count.

S / N_{NB} = 10 log \frac{\bar{y} 2}{s^{2}}

(2)

where

\bar{y}

is the mean of response values, and s is standard deviation.

2. Main effect analysis.

The main effect analysis combines the S/N ratio of the experimental results with the orthogonal array, to carry out the response of each level of each quality, so as to find out the combination of optimal control factors. Response of effect factor A represents the average value in relation to experimental results with A_i (i=1,…,18). According to the results of main effect analysis, the response of each level of each quality is made into a response table.

Analysis of variance (ANOVA)

ANOVA splits the variation of each factor data to determine the effect of the factor on quality, so as to decide the significant factor and analyze the error of each factor. The factors with insufficient significance are classified as pooled error. ANOVA can analyze the influence of each factor on quality characteristics in the melt spinning process.

ANOVA calculates the following items:

Degree of freedom (DOF)

DOF = Thenumberoflevels - 1

(3)

DO F_{T} = Thenumberofexperiments - 1

(4)

Total sum of squares (SS_T)

S S_{T} = \sum_{j = 1}^{n} {(y_{i} - \bar{y})}^{2}

(5)

where y_i are the response value of L₁₈ orthogonal arrays, and

\bar{y}

is the mean of response value.

Sum of square (SS) of factors

S S_{A} = \frac{1}{n} {(A}_{1}^{2} {+ A}_{2}^{2} {+ A}_{3}^{2} + \dots A_{n}^{2}) - CF

(6)

where A_i are the factor's response in the main effect analysis.

CF = \frac{\sum {y_{i}}^{2}}{n}

(7)

where n is the number of experiments.

Sum of square of error (SS_E)

S S_{E} = S S_{T} - \sum S S_{A}

(8)

Mean sum of square (MS)/Mean sum of square of error (MSE)

MS = \frac{SS}{DOF}

(9)

MSE = \frac{S S_{E}}{DO F_{E}}

(10)

DO F_{E} = DO F_{T} - DOF

(11)

where DOF_E is the degree of freedom of error.

F ratio.

Fratio = \frac{MS}{MSE}

(12)

ANOVA uses F ratio to represent the effect relation of factors. The larger the F ratio is, the greater is the effect of the factor on the system. Therefore, the F ratio can be applied to order the importance of factors and can analyze the effect of experimental parameters on quality. This is normally a weakness within the Taguchi experiment.

Partial sum of square (SS') of factors

{SS}_{A}^{'} = S S_{A} - DOF \times MSE

(13)

Contribution

ρ = \frac{S {S^{'}}_{A}}{{SS}_{T}}

(14)

Contribution is the influence of each factor on quality characteristics. The higher the contribution, the greater its influence.

Grey relational analysis

The GRA seeks the numerical relation between each factor and quality in the experiment by analysis. If the factors have a consistent data plot variation trend versus quality, the correlation grade is higher. Therefore, the GRA provides quantized indexes for an experimental developmental variation trend.²⁹ The GRA process is described below.

1. Target determination The target values of various quality characteristics are defined. Hygroscopic heating, tensile strength, elongation at break, and yarn count characteristics should be maximized in this study. The maximum value of experimental results is defined as the target value.

2. Grey relation sequence generation According to the qualities (the larger the better or nominal the best), the grey relation sequence is defined as follows:²⁵

A. The larger the better.

x_{i} (k) = \frac{x_{i} (k) - min x_{i} (k)}{max x_{i} (k) - min x_{i} (k)}

(15)

where i is the number of experiment in each quality, k is the total number of desired qualities.

B. Nominal the best.

x_{i} (k) = \frac{min x_{i} (k) - x_{i} (k)}{max {x_{i} (k) - OB, OB - min x_{i} (k)}}

(16)

where i is the number of experiment in each quality, k is the number of qualities, OB is the target value.

3. Grey relational coefficient calculation

ξ_{i} (k) = \frac{Δ min + ξ Δ max}{Δ oi (k) + ξ Δ max}

(17)

where Δmin is the minimal difference between

x_{i} (k)

and target value, Δmax is the maximal difference between

x_{i} (k)

and target value,

ξ

is the distinguishing coefficient which is usually set at 0.5, ¹⁷ Δoi(k) is the difference between

x_{i} (k)

and target value.

4. Grey relational grade calculation

r_{i} = \frac{1}{n} \sum_{k = 1}^{n} w (k) \times ξ_{i} (k)

(18)

where w(k) is the weight of each quality, it is set as the contribution in each quality. The grey relational grade represents the correlation grade of various groups of experimental parameters to the ideal values of quality characteristics. The larger the value, the closer the corresponding process parameter is to the ideal value.

Result and discussion

This study dealt with single quality and multi-quality process parameter optimizations simultaneously for the moisture absorption heating PET yarn melt spinning process. The Taguchi method is used for single quality process parameter optimization to convert the experimental value of L₁₈ orthogonal array into an S/N ratio. The main effect analysis is applied to calculate the optimum process parameters of each quality, and the factors influence for each quality can be determined by ANOVA. The GRA is employed to deal with multi-quality optimization process parameters. The process parameter optimizations for single quality and multi-quality are as follows.

Hygroscopic heating in single quality optimization analysis

The hygroscopic heating of yarn is tested by the standard FTTS-FA-023. The mean value of the hygroscopic heating property is calculated by five observations in each experiment, and the results of the experiments are shown in Table 3.

Table 3.

Experimental data of hygroscopic heating

Exp. No.	Factors							Hygroscopic heating
Exp. No.	A (PEG/TPA weight ratio)	B (SIPE/TPA molar ratio)	C (℃)	D (℃)	E (℃)	F (m/min)	G (m/min)	Mean (^oC)	Standard deviation (^oC)
1	3/100	2/100	275	285	290	1600	2700	1.51	0.0070
2	3/100	3/100	280	290	295	1920	3300	1.56	0.0068
3	3/100	4/100	285	295	300	2240	3900	1.61	0.0097
4	5/100	2/100	275	290	295	2240	3900	1.62	0.0092
5	5/100	3/100	280	295	300	1600	2700	1.65	0.0068
6	5/100	4/100	285	285	290	1920	3300	1.68	0.0052
7	7/100	2/100	280	285	300	1920	3900	1.62	0.0182
8	7/100	3/100	285	290	290	2240	2700	1.70	0.0134
9	7/100	4/100	275	295	295	1600	3300	1.80	0.0076
10	3/100	2/100	285	295	295	1920	2700	1.50	0.0042
11	3/100	3/100	275	285	300	2240	3300	1.56	0.0100
12	3/100	4/100	280	290	290	1600	3900	1.62	0.0096
13	5/100	2/100	280	295	290	2240	3300	1.61	0.0121
14	5/100	3/100	285	285	295	1600	3900	1.60	0.0185
15	5/100	4/100	275	290	300	1920	2700	1.68	0.0144
16	7/100	2/100	285	290	300	1600	3300	1.61	0.0080
17	7/100	3/100	275	295	290	1920	3900	1.70	0.0131
18	7/100	4/100	280	285	295	2240	2700	1.80	0.0111

PEG: polyethylene glycol; SIPE: sodium-5-sulfo-bis-(hydroxyethyl)-isophthalate; TPA: terephthalic acid.

The S/N ratio is calculated by the hygroscopic heating result of the L₁₈ experiments. The effect of various factors is calculated by main effect analysis. The response table is shown in Table 4 and the response figure is shown in Figure 4.

Figure 4.

Response figure of hygroscopic heating.

Table 4.

Response table of hygroscopic heating (Unit: dB)

	A	B	C	D	E	F	G
Level 1	3.858	3.967	4.308	4.233	4.270	4.242	4.284
Level 2	4.303	4.232	4.306	4.251	4.312	4.201	4.277
Level 3	4.629	4.590	4.176	4.305	4.207	4.347	4.229
Effect	0.771	0.623	0.132	0.071	0.105	0.146	0.055
Ranking	1	2	4	6	5	3	7

According to Table 4 and Figure 4, the optimum factors of hygroscopic heating are A₃, B₃, C₁, D₃, E₂, F₃ and G₁. Factor A (PEG content) has the greatest effect on hygroscopic heating, and followed by B (SIPE content), F (gear pump speed), C (melt temperature), E (nozzle temperature), D (mold temperature) and G (take-up speed).

Table 5 shows the ANOVA for hygroscopic heating, factor D and G are incorporated into the error term because the F ratio is less than 2. As can be seen, the PEG content and SIPE content have the greatest effect on hygroscopic heating, followed by melt temperature, gear pump speed and nozzle temperature. This means that the addition of PEG and SIPE enables the polymerized PET to have hydrophilic functional groups, so that the modified PET fiber can have higher hygroscopic heating properties.

Table 5.

ANOVA of hygroscopic heating

Factor	DOF	SS	MS	F ratio	SS'	Contribution
A	2	1.796	0.898	120.884	1.781	55.82%
B	2	1.175	0.587	79.052	1.161	36.37%
C	2	0.069	0.034	4.638	0.054	1.70%
D	2	0.016	0.008	1.109	–	–
E	2	0.034	0.017	2.267	0.019	0.59 %
F	2	0.068	0.034	4.604	0.054	1.68%
G	2	0.011	0.005	0.716	–	–
Error	3	0.022	0.007	–	–	–
Combined Error	7	0.049	0.021	–	–	3.84%
Total	17	3.191	–	–	3.191	100%

ANOVA: analysis of variance.

Tensile strength in single quality optimization analysis

The tensile strength of yarn is tested by the standard ASTM D3822. The mean value of the tensile strength property is calculated by five observations in each experiment, and the results of the experiments are shown in Table 6.

Table 6.

Experimental data of tensile strength

Exp. No.	Factors							Tensile strength
Exp. No.	A (PEG/TPA weight ratio)	B (SIPE/TPA molar ratio)	C (℃)	D (℃)	E (℃)	F (m/min)	G (m/min)	Mean (g/d)	Standard deviation (g/d)
1	3/100	2/100	275	285	290	1600	2700	3.44	0.0397
2	3/100	3/100	280	290	295	1920	3300	3.44	0.0270
3	3/100	4/100	285	295	300	2240	3900	3.53	0.0628
4	5/100	2/100	275	290	295	2240	3900	3.45	0.0305
5	5/100	3/100	280	295	300	1600	2700	3.59	0.0487
6	5/100	4/100	285	285	290	1920	3300	3.61	0.0576
7	7/100	2/100	280	285	300	1920	3900	3.51	0.0950
8	7/100	3/100	285	290	290	2240	2700	3.34	0.0488
9	7/100	4/100	275	295	295	1600	3300	3.58	0.0367
10	3/100	2/100	285	295	295	1920	2700	3.38	0.0654
11	3/100	3/100	275	285	300	2240	3300	3.45	0.0539
12	3/100	4/100	280	290	290	1600	3900	4.28	0.1332
13	5/100	2/100	280	295	290	2240	3300	3.34	0.1053
14	5/100	3/100	285	285	295	1600	3900	3.45	0.0586
15	5/100	4/100	275	290	300	1920	2700	3.50	0.1516
16	7/100	2/100	285	290	300	1600	3300	4.32	0.2946
17	7/100	3/100	275	295	290	1920	3900	3.73	0.0696
18	7/100	4/100	280	285	295	2240	2700	3.44	0.0604

PEG: polyethylene glycol; SIPE: sodium-5-sulfo-bis-(hydroxyethyl)-isophthalate; TPA: terephthalic acid.

The S/N ratio is calculated by the tensile strength result of L₁₈ experiments, the effect of various factors is calculated by main effect analysis. The response table is shown in Table 7 and the response figure is shown in Figure 5.

Figure 5.

Response figure of tensile strength.

Table 7.

Response table of tensile strength (Unit: dB)

	A	B	C	D	E	F	G
Level 1	11.063	11.019	10.937	10.832	11.143	11.489	10.745
Level 2	10.851	10.873	11.096	11.355	10.773	10.946	11.142
Level 3	11.211	11.233	11.091	10.937	11.208	10.690	11.237
Effect	0.360	0.361	0.160	0.523	0.435	0.799	0.492
Ranking	6	5	7	2	4	1	3

According to Table 7 and Figure 5, the optimum factors of tensile strength are A₃, B₃, C₂, D₂, E₃, F₁ and G₃. Factor F (gear pump speed) has the greatest effect on tensile strength, and followed by D (mold temperature), G (take-up speed), E (nozzle temperature), B (SIPE content), A (PEG content) and C (melt temperature).

Table 8 shows the ANOVA for tensile strength. Factor C is incorporated into the error term because the F ratio is less than 2. As can be seen, the gear pump speed has the greatest effect on tensile strength, meaning that it has the greatest influence on the tensile strength quality, followed by mold temperature, take-up speed, nozzle temperature, SIPE content and PEG content. The results show that under the drafting force provided by the gear pump and take-up system, molecular orientation and crystallinity increase, which increases the tensile strength of the yarn, resulting in values consistent with the literature.³⁰

Table 8.

ANOVA of tensile strength

Factor	DOF	SS	MS	F ratio	SS'	Contribution
A	2	0.593	0.296	3.099	0.401	5.93%
B	2	0.595	0.297	3.110	0.404	5.96%
C	2	0.099	0.050	0.518	–	–
D	2	1.119	0.560	5.852	0.928	13.71%
E	2	0.861	0.430	4.500	0.669	9.89%
F	2	2.200	1.100	11.501	2.008	29.67%
G	2	1.016	0.508	5.312	0.825	12.18%
Error	3	0.287	0.096	–	–	–
Combined Error	5	0.386	0.145	–	–	22.66%
Total	17	6.769	–	–	6.769	100%

ANOVA: analysis of variance.

Elongation at break in single quality optimization analysis

The elongation at break of yarn is tested by the standard ASTM D3822. The mean value of the elongation at break property is calculated by five observations in each experiment, and the results of the experiments are shown in Table 9.

Table 9.

Experimental data of elongation at break

Exp. No.	Factors							Elongation at break
Exp. No.	A (PEG/TPA weight ratio)	B (SIPE/TPA molar ratio)	C (℃)	D (℃)	E (℃)	F (m/min)	G (m/min)	Mean (%)	Standard deviation (%)
1	3/100	2/100	275	285	290	1600	2700	39.63	0.7292
2	3/100	3/100	280	290	295	1920	3300	42.47	0.9226
3	3/100	4/100	285	295	300	2240	3900	45.22	0.6051
4	5/100	2/100	275	290	295	2240	3900	41.93	1.5860
5	5/100	3/100	280	295	300	1600	2700	42.68	0.3026
6	5/100	4/100	285	285	290	1920	3300	45.39	0.5166
7	7/100	2/100	280	285	300	1920	3900	44.89	0.8464
8	7/100	3/100	285	290	290	2240	2700	44.94	0.6191
9	7/100	4/100	275	295	295	1600	3300	43.80	0.9024
10	3/100	2/100	285	295	295	1920	2700	42.12	0.3756
11	3/100	3/100	275	285	300	2240	3300	42.57	0.5643
12	3/100	4/100	280	290	290	1600	3900	41.16	0.8201
13	5/100	2/100	280	295	290	2240	3300	43.05	0.6209
14	5/100	3/100	285	285	295	1600	3900	43.23	0.6167
15	5/100	4/100	275	290	300	1920	2700	40.49	0.3878
16	7/100	2/100	285	290	300	1600	3300	44.46	0.5305
17	7/100	3/100	275	295	290	1920	3900	43.36	0.4865
18	7/100	4/100	280	285	295	2240	2700	41.77	0.3768

PEG: polyethylene glycol; SIPE: sodium-5-sulfo-bis-(hydroxyethyl)-isophthalate; TPA: terephthalic acid.

The S/N ratio is calculated by the elongation at break result of L₁₈ experiments, and the effect of various factors is calculated by main effect analysis. The response table is shown in Table 10 and the response figure is shown in Figure 6.

Figure 6.

Response figure of elongation at break.

Table 10.

Response table of elongation at break (Unit: dB)

	A	B	C	D	E	F	G
Level 1	32.495	32.592	32.447	32.641	32.641	32.557	32.443
Level 2	32.619	32.707	32.596	32.571	32.572	32.685	32.789
Level 3	32.838	32.652	32.909	32.740	32.738	32.710	32.719
Effect	0.342	0.116	0.462	0.168	0.166	0.153	0.346
Ranking	3	7	1	4	5	6	2

According to Table 10 and Figure 6, the optimum factors of elongation at break are A₃, B₂, C₃, D₃, E₃, F₃ and G₂. Factor C (melt temperature) has the greatest effect on elongation at break, and followed by G (take-up speed), A (PEG content), D (mold temperature), E (nozzle temperature), F (gear pump speed) and B (SIPE content).

Table 11 shows the ANOVA for elongation at break. Factors B, D, E and F are incorporated into the error term because the F ratio is less than 2. As can be seen, the melt temperature has the greatest effect on elongation at break. In the melt spinning process, relatively high temperatures can easily reduce the viscosity or cause thermal cracking in the material. The entanglements of the molecular chains will be reduced, and the molecular arrangements will be loosened to increase the fiber defects.³⁰

Table 11.

ANOVA of elongation at break

Factor	DOF	SS	MS	F ratio	SS'	Contribution
A	2	0.360	0.180	2.901	0.236	12.40%
B	2	0.040	0.020	0.322	–	–
C	2	0.666	0.333	5.366	0.542	28.48%
D	2	0.086	0.043	0.689	–	–
E	2	0.084	0.042	0.673	–	–
F	2	0.081	0.040	0.651	–	–
G	2	0.401	0.200	3.228	0.277	14.53%
Error	3	0.186	0.062	–	–	–
Combined Error	5	0.853	0.395	–	–	44.59%
Total	17	1.904	–	–	1.904	100%

ANOVA: analysis of variance.

Yarn count in single quality optimization analysis

The yarn count of yarn is tested by the standard ASTM D1577-07. The mean value of the yarn count property is calculated by five observations in each experiment, and the results of the experiments are shown in Table 12.

Table 12.

Experimental data of yarn count

Exp. No.	Factors							Yarn count
Exp. No.	A (PEG/TPA weight ratio)	B (SIPE/TPA molar ratio)	C (℃)	D (℃)	E (℃)	F (m/min)	G (m/min)	Mean (d/72f)	Standard deviation (d/72f)
1	3/100	2/100	275	285	290	1600	2700	75.32	1.5559
2	3/100	3/100	280	290	295	1920	3300	80.61	2.4899
3	3/100	4/100	285	295	300	2240	3900	79.87	2.7414
4	5/100	2/100	275	290	295	2240	3900	76.67	3.5924
5	5/100	3/100	280	295	300	1600	2700	78.08	5.7593
6	5/100	4/100	285	285	290	1920	3300	78.92	3.6823
7	7/100	2/100	280	285	300	1920	3900	77.32	1.1683
8	7/100	3/100	285	290	290	2240	2700	76.08	4.1587
9	7/100	4/100	275	295	295	1600	3300	79.81	4.6985
10	3/100	2/100	285	295	295	1920	2700	77.74	1.4581
11	3/100	3/100	275	285	300	2240	3300	73.36	3.6227
12	3/100	4/100	280	290	290	1600	3900	79.84	4.2784
13	5/100	2/100	280	295	290	2240	3300	76.28	1.9257
14	5/100	3/100	285	285	295	1600	3900	78.32	2.6251
15	5/100	4/100	275	290	300	1920	2700	77.38	1.5018
16	7/100	2/100	285	290	300	1600	3300	78.23	1.7938
17	7/100	3/100	275	295	290	1920	3900	79.45	2.2204
18	7/100	4/100	280	285	295	2240	2700	74.67	3.3625

PEG: polyethylene glycol; SIPE: sodium-5-sulfo-bis-(hydroxyethyl)-isophthalate; TPA: terephthalic acid

The S/N ratio is calculated by the yarn count result of L₁₈ experiments, the effect of various factors is calculated by main effect analysis. The response table is shown in Table 13 and the response figure is shown in Figure 7.

Figure 7.

Response figure of yarn count.

Table 13.

Response table of yarn count (Unit: dB)

	A	B	C	D	E	F	G
Level 1	37.793	37.707	37.703	37.634	37.777	37.840	37.657
Level 2	37.777	37.774	37.801	37.838	37.824	37.895	37.805
Level 3	37.782	37.870	37.847	37.880	37.751	37.617	37.890
Effect	0.017	0.163	0.144	0.245	0.074	0.279	0.234
Ranking	7	4	5	2	6	1	3

According to Table 13 and Figure 7, the optimum factors of yarn count are A₁, B₃, C₃, D₃, E₂, F₂ and G₃. Factor F (gear pump speed) has the greatest effect on yarn count, and then D (mold temperature), G (take-up speed), B (SIPE content), C (melt temperature), E (nozzle temperature) and A (PEG content).

Table 14 shows the ANOVA for yarn count. Factors A, C and E are incorporated into the error term because the F ratio is less than 2. The gear pump speed has the greatest effect on yarn count quality, that is, the geometry of fibers cross-sectional profile, followed by mold temperature, take-up speed and SIPE content.

Table 14.

ANOVA of yarn count

Factor	DOF	SS	MS	F ratio	SS'	Contribution
A	2	0.001	0.000	0.024	–	–
B	2	0.080	0.040	2.325	0.046	5.37%
C	2	0.065	0.032	1.886	–	–
D	2	0.207	0.103	5.997	0.172	20.24%
E	2	0.017	0.008	0.486	–	–
F	2	0.262	0.131	7.593	0.227	26.71%
G	2	0.168	0.084	4.875	0.134	15.69%
Error	3	0.052	0.017	–	–	–
Combined Error	5	0.117	0.050	–	–	31.99%
Total	17	0.850		–	0.850	100%

Multi-quality optimization analysis

The Taguchi method is combined with GRA to obtain multi-quality optimization parameters. The grey relational grade is obtained by Eqs. (15)–(18) as shown in Table 15. The total factor's contribution of each quality in ANOVA analysis is used to determine the weights of Eq. (18). The weights of hygroscopic heating, tensile strength, elongation at break and yarn count are 0.9616, 0.7734, 0.5541 and 0.6801, respectively.

Table 15.

The grey relational grade of each experiment

Experiment No.	$r_{i}$	Experiment No.	$r_{i}$	Experiment No.	$r_{i}$
1	0.1968	7	0.3547	13	0.3001
2	0.1355	8	0.4236	14	0.2551
3	0.2796	9	0.4072	15	0.2918
4	0.2919	10	0.1534	16	0.4700
5	0.3180	11	0.2609	17	0.3558
6	0.3876	12	0.3369	18	0.4819

The effect of various factors is calculated by main effect analysis according to Table 16. The response table is shown in Table 13 and the response figure is shown in Figure 8.

Figure 8.

Response figure of grey relational grade

Table 16.

Response table of grey relational grade

	A	B	C	D	E	F	G
Level 1	0.227	0.294	0.301	0.323	0.333	0.331	0.311
Level 2	0.307	0.291	0.321	0.325	0.287	0.280	0.327
Level 3	0.416	0.364	0.328	0.302	0.329	0.340	0.312
Effect	0.188	0.073	0.027	0.023	0.046	0.060	0.016
Ranking	1	2	5	6	4	3	7

According to Table 16 and Figure 8, the optimum factors of multi-quality optimization are A₃, B₃, C₃, D₂, E₁, F₃ and G₂. It means that PEG content 7/100 (PEG/TPA weight ratio), SIPE content 4/100 (SIPE/TPA molar ratio), melt temperature 285 ℃, mold temperature 285 ℃, nozzle temperature 290 ℃, gear pump speed 2,240 m/min and take-up speed 3,300 m/min. Factor A (PEG content) has the greatest effect, and followed by B (SIPE content), F (gear pump speed), E (nozzle temperature), C (melt temperature), D (mold temperature) and G (take-up speed).

Confirmation experiment

Modification verification

The properties of the moisture absorption heating PET yarn and general PET yarn are compared in Table 17.

Table 17.

Comparison of general PET yarn and the moisture absorption heating PET yarn in this research

Quality materials	Hygroscopic heating (℃)	Surface resistance (Ω)	Tensile strength (g/d)	Elongation at break (%)	Yarn count (d/72f)
General PET yarn	1.3	10⁹	4.61	38.44%	73.55
Moisture absorption heating PET yarn	1.8	10⁷	4.16	44.89%	75.02

According to Table 17, the hygroscopic heating of the moisture absorption heating PET yarn is 1.8 ^oC, higher than the general PET yarn 1.3 ^oC of 0.5 ^oC. The surface resistance of the moisture absorption heating PET yarn is 10⁷Ω which is classified as “good,” and the surface resistance of the general PET yarn is 10⁹ which is classified as “fair.”³¹ The moisture absorption heating PET yarn achieved higher elongation, but less tensile strength than the general PET yarn. However, the tensile strength of 4.16g/d and elongation at break of 44.89% are suitable for weaving fabrics.^32–33 The yarn count of the moisture absorption heating PET yarn is 75.02d/72f. It complies with the clothing industrial specification 75d/72f.

The results show that the modified moisture absorption heating PET yarn can improve the hygroscopic heating and surface resistance of the yarn, but will reduce the tensile strength of the yarn. Hsiao et al.¹⁰ also reported that the mechanical properties of the copolymer gradually declined as the SIPE content increased.

Quality optimization verification

In order to validate the single quality and multi-quality optimum parameters of this study, a sample was fabricated using the following optimized parameters:

Hygroscopic heating optimized yarn: PEG content 7/100 (PEG/TPA weight ratio), SIPE content 4/100 (SIPE/TPA molar ratio), melt temperature 275 ℃, mold temperature 295 ℃, nozzle temperature 295 ℃, gear pump speed 2240 m/min and take-up speed 2700 m/min.

Tensile strength optimized yarn: PEG content 7/100 (PEG/TPA weight ratio), SIPE content 4/100 (SIPE/TPA molar ratio), melt temperature 280 ℃, mold temperature 290 ℃, nozzle temperature 300 ℃, gear pump speed 1600 m/min and take-up speed 3900 m/min.

Elongation at break optimized yarn: PEG content 7/100 (PEG/TPA weight ratio), SIPE content 3/100 (SIPE/TPA molar ratio), melt temperature 285 ℃, mold temperature 295 ℃, nozzle temperature 300 ℃, gear pump speed 2240 m/min and take-up speed 3300 m/min.

Yarn count optimization yarn: PEG content 3/100 (PEG/TPA weight ratio), SIPE content 4/100 (SIPE/TPA molar ratio), melt temperature 285 ℃, mold temperature 295 ℃, nozzle temperature 295 ℃, gear pump speed 1920 m/min and take-up speed 3900 m/min.

Multi-quality optimized yarn: PEG content 7/100 (PEG/TPA weight ratio), SIPE content 4/100 (SIPE/TPA molar ratio), melt temperature 285 ℃, mold temperature 285 ℃, nozzle temperature 290 ℃, gear pump speed 2240 m/min and take-up speed 3300 m/min.

The properties of the single quality and multi-quality optimized yarn are shown in Table 18.

Table 18.

The properties of the single quality and multi‐quality optimized yarn

Properties Type of yarn	Hygroscopic heating (^oC)	Tensile strength (g/d)	Elongation at break (%)	Yarn count (d/72f)	Grey relational grade
Hygroscopic heating optimized yarn	1.84	3.71	42.66	76.42	0.5419
Tensile strength optimized yarn	1.65	4.48	44.23	77.31	0.5513
Elongation at break optimized yarn	1.63	4.33	46.62	73.58	0.5500
Yarn count optimized yarn	1.72	3.77	41.02	75.28	0.4548
Multi-quality optimized yarn	1.80	4.16	44.89	75.02	0.6967

In Table 18, the results of the single quality parameters optimization included hygroscopic heating, tensile strength, elongation at break and yarn count are compared with the experimental data in Table 3, Table 6, Table 9 and Table 12. The optimized qualities hygroscopic heating, tensile strength, elongation at break, and yarn count are superior to the yarns form orthogonal arrays experiment. It is verified that the single quality parameters optimization of melt spinning using the Taguchi method is effective in this study.

For multi-quality optimization verification, the grey relational grade provides a comprehensive indicator of multi-quality optimized yarn in the setting weights. The results of multi-quality parameters optimization are compared with the experimental data in Table 15. The multi-quality optimized yarn improves properties comprehensively and presents the highest grey relational grade in experimental yarns. The results confirm that the Taguchi method combined with the GRA optimization theory can indeed optimize the single quality and multi-quality properties of the yarn.

Composite fabric

In order to validate the properties of far-infrared function PET yarn and moisture absorption heating PET yarn composite fabric, the yarns from Part I and Part II are woven together into a fabric, the properties of the composite fabric are: far-infrared emission 89%, far-infrared heating 6.3 ℃, hygroscopic heating 1.8 ℃ and surface resistance 10⁷. The properties of the composite fabric are the same as the yarns of Part I and Part II. It is shown that the composite fabric doesn't affect the properties of the original yarns. Combining the far-infrared heating and the hygroscopic heating, the heating property of the composite fabric can reach 8.1 ℃.

Conclusion

A moisture absorption heating PET yarn, which includes both a hygroscopic heating and antistatic function, is developed in this study. The TPA, EG, PEG oligomer, and SIPE are used in hydrophilic PET modification. The hydrophilic PET is made into a moisture absorption heating PET yarn to a75d/72f specified, fully drawn yarn, by melt spinning and melt drafting. The Taguchi method is combined with GRA to optimize the melt spinning process parameters for multi-quality optimization. The properties of this yarn, including hygroscopic heating, tensile strength, elongation at break and yarn count, are discussed in the quality optimization. Based on the results, the conclusions are as follows:

According to the comparison between unmodified PET yarn and the moisture absorption heating PET yarn in this research, the modified moisture absorption heating PET can improve the hygroscopic heating and surface resistance of the yarn simultaneously, but will reduce the tensile strength of the yarn. It was confirmed in this study that using TPA, EG, PEG oligomer and SIPE polymerization to modify the hydrophilic PET was successful.

For the yarn application, in this research the moisture absorption heating PET yarn got higher elongation but less tensile strength than the general PET yarn. It achieved a tensile strength 4.16g/d and an elongation at break 44.89%, which is suitable for weaving fabrics.

According to the comparison between the single-quality optimized yarn and the experimental yarn, the optimized qualities as hygroscopic heating, tensile strength, elongation at break and yarn count are superior to the experimental yarn. It is verified that the single quality parameter optimization of melt spinning using the Taguchi method is effective in this study.

The multi-quality optimized yarn improves properties comprehensively and presents the highest grey relational grade in experimental yarns. The results confirm that the Taguchi method combined with GRA optimization theory can indeed optimize the single quality and multi-quality properties of the yarn.

The properties of composite fabric formed from the yarns of Part I and Part II are far-infrared emission 89%, far-infrared heating 6.3 ℃, hygroscopic heating 1.8 ℃ and surface resistance 10⁷. Combining the far-infrared and hygroscopic heating, the heating property of the composite fabric can reach 8.1 ℃.

Footnotes

Acknowledgement

The research was supported by the Ministry of Science and Technology of the Republic of China under grant number 106-2221-E-011-137-MY2.

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

References

Pan

Hsieh

Lou

, et al. Elastic knits with different structures composed by using wrapped yarns: Function and comfort evaluations. Fiber Polym 2017; 18(9): 1816–1824.

Niedermann

Rossi

. Objective and subjective evaluation of the human thermal sensation of wet fabrics. Text Res J 2012; 82(4): 374–384.

Greenwald

. Electrical hazards and accidents: Their cause and prevention, New York: Van Nostrand Reinhold Publications, 1991.

Shang

Zhang

. A study of wear abilities of softwarm heating warm fiber fabric. Adv Mater Res 2013; 709: 233–237.

Pinar

Oleksiewicz

Wróbel

. Assessment of the electrostatic properties of polyester knitted fabrics containing carbon fibres after enzymatic modification for the improving of hygroscopic properties. Fibres Text East Eur 2014; 22: 84–90.

Zhou

Fang

Wang

, et al. Improving bonding strength between a hydrophilic coating layer and poly (ethylene terephthalate) braid for preparing mechanically stable braid-reinforced hollow fiber membranes. J Appl Polym Sci 2018; 135(14): 46104.

Chetouane

Fafet

Barbet

, et al. Hydrophilic–impermeable modified polyethylene terephthalate for selective endothelialization. IOP Conference Series: Mater Sci Eng 2017; 254(6): 062003.

Tamizifar

Sun

. Surface modification of poly(ethylene terephthalate) fibers via controlled radical graft polymerization. J Appl Polym Sci 2018; 135(11): 45990.

Takke

Behary

Perwuelz

, et al. Surface and adhesion properties of poly (ethylene glycol) on polyester (polyethylene terephthalate) fabric surface: Effect of air-atmospheric plasma treatment. J Appl Polym Sci 2011; 122(4): 2621–2629.

10.

Hsiao

Kuo

Tang

, et al. Physics and kinetics of alkaline hydrolysis of cationic dyeable poly (ethylene terephthalate)(CDPET) and polyethylene glycol (PEG)-modified CDPET polymers: Effects of dimethyl 5-sulfoisophthalate sodium salt/PEG content and the number-average molecular weight of the PEG. J Appl Polym Sci 2005; 98(2): 550–556.

11.

Miranda

TMR

Santos

Soares

GMB

. Soil-release behaviour of polyester fabrics after chemical modification with polyethylene glycol. IOP Conference Series: Mater Sci Eng 2017; 254(3): 032005.

12.

Zhao

, et al. Preparation and characterization of poly (ethylene terephthalate) copolyesters modified with sodium-5-sulfo-bis-(hydroxyethyl)-isophthalate and poly (ethylene glycol). J Appl Polym Sci 2014; 131(3): 39823.

13.

Eisenberg

Hird

Moore

. A new multiplet-cluster model for the morphology of random ionomers. Macromolecules 1990; 23(18): 4098–4107.

14.

Pal

Gandhi

Kothari

. Effect of comonomer on structure and properties of textured cationic dyeable polyester. J Appl Polym Sci 1996; 61(3): 401–406.

15.

Kuo

C-FJ

Lan

Chen

, et al. Property modification and process parameter optimization design of polylactic acid composite materials. Part I: polylactic acid toughening and photo-degradation modification and optimized parameter design. Text Res J 2015; 85(1): 13–25.

16.

Kuo

C-FJ

Lan

Chen

, et al. Property Modification and Process Parameter Optimization Design of Polylactic Acid Composite Materials Part II: Application of Response Surface Methodology and Multiobjective. Text Res J 2015; 85(7): 687–700.

17.

Rawal

Mukhopadhyay

. Advances in filament yarn spinning of textiles and polymers: Chapter 4 - Melt spinning of synthetic polymeric filaments, United Kingdom: Woodhead Publications, 2014.

18.

Kuo

C–FJ

Fan

, et al. Nano composite fiber process optimization for polypropylene with antibacterial and far–infrared ray emission properties. Text Res J 2016; 86(16): 1677–1687.

19.

Park

Cho

Park

. Effects of material and processing conditions on powder-binder separation using the Taguchi method. Powder Technol 2017; 321: 369–379.

20.

Moayyedian

Abhary

Marian

. Optimization of injection molding process based on fuzzy quality evaluation and Taguchi experimental design. CIRP Manuf Sci Techn 2018; 21: 150–160.

21.

Chauhan

Singh

Kumar

, et al. Experimental investigation and optimization of impinging jet solar thermal collector by Taguchi method. Appl Therm Eng 2017; 116: 100–109.

22.

Kuo

C–FJ

Lin

. A study of multi-quality processing parameter optimization for sueded fabric. Text Res J 2017; 87(4): 389–398.

23.

Kuo

C–FJ

Tzeng

Lan

, et al. A study on blending polyethylene terephthalate with titanium dioxide particles in melt spinning process parameter optimization. Text Res J 2013; 83(8): 813–826.

24.

Pervez

Mozumder

Mourad

AHI

. Optimization of injection molding parameters for HDPE/TiO₂ nanocomposites fabrication with multiple performance characteristics using the Taguchi method and grey relational analysis. Materials 2016; 9(8): 710.

25.

Najarian

Alipour

Rad

, et al. Multi-objective optimization of converting process of auxetic foam using three different statistical methods. Measurement 2018; 119: 108–116.

26.

Kuo

C–FJ

Syu

Lin

, et al. The application of principal component analysis and gray relational method in the optimization of the melt spinning process using the cooling air system. Text Res J 2013; 83(4): 371–380.

27.

Tarng

Juang

Chang

. The use of grey-based Taguchi methods to determine submerged arc welding process parameters in hardfacing. J Mater Process Tech 2002; 128(1–3): 1–6.

28.

Mohammad

Chin

Talebian

, et al. Electrospinning of polymethyl methacrylate nanofibers: optimization of processing parameters using the Taguchi design of experiments. Text Res J 2015; 85(4): 356–368.

29.

Bezuglov

Comert

. Short–term freeway traffic parameter prediction: Application of grey system theory models. Expert Syst Appl 2016; 62: 284–292.

30.

Zhang

Dong

. Advances in filament yarn spinning of textiles and polymers, United Kingdom: Woodhead Publishing Limited in association with the Textile Institute, 2014.

31.

The committee for conformity assessment of accreditation and certification on functional and technical textiles, standard of FTTS-FA-009, Specified requirements of far infrared textiles, 2005. http://www.ftts.org.tw/images/fa009E.pdf (accessed 25 April 2018).

32.

Pramanik

Patil

. Physical characteristics of cotton/polyester core spun yarn made using ring and air-jet systems. AUTEX Res J 2009; 9(1): 14–19.

33.

Crow, Rita M., and Malcolm M. Dewar. Examination of the Load–Elongation Properties of Fabrics, Yarns and Fibres at 20 and 40 C. No. Dreo-923. Defence Research Establishment Ottawa (ONTARIO), 1985.