Effect of Cotton/Polyester Blend Ratio, Loom Speed, and Air Pressure on Yarn Twist Loss and Yarn Strength Loss: The Case of Air-Jet Loom

ABSTRACT Production of high-quality woven fabric with uniform fabric properties is an essential requirement for manufacturers. But in air-jet weaving, some quality parameters of the fabric are affected by twist loss of weft yarn during weft insertion. The aim of this work is to show the effect of polyester/cotton (PC) blend ratio, loom speed, and air pressure on weft yarn twist loss and tensile properties of the fabrics produced by air-jet weaving machines. Box-Behnken design has been used to design and analyze the experiment. 27 fabric samples were on an air-jet loom with a combination of four factors each at three levels (loom speed (A) 300rpm, 425rpm, and 550rpm), left side relay nozzle pressure (LSRNP) (B) (2.5 bar,3.25 bar, and 4 bar), Right side relay nozzle pressure (RSRNP)(C) (3 bar, 4.75 bar, and 6.5 bar), cotton/polyester blend (D) 100%/0%, 75%/25%, and 50%/50%,). The results of the experiment revealed that PC blend ratio, loom speed, and air pressure affect the twist level and the strength of the yarn significantly. It is found that relay nozzle air pressure is directly proportional to the twist losses and weft yarn strength loss. But the PC blend ratio has a more significant impact on weft yarn twist loss and mechanical properties of the yarn.


Introduction
Woven fabrics made from blended yarns have superior properties than fabric made from a single fiber. The most common types of weft yarn insertion methods for woven fabric are shuttle, rapier, water jet, air-jet, and projectile insertion. However, a modern air-jet weft insertion method is preferred due to the high speed of weft insertion for lightweight to medium-weight fabrics.
In air-jet weaving, the weft yarn insertion process is complex because the motion of the yarn is determined by a complex interaction between the surface of the yarn and the air (Brun, Corti, and Pozzetti 2008). The pressurized air is inserted through a nozzle to carry the weft yarn from one end of the loom to the other end across the width of the fabric (Gong, Chen, and Zhou 2018). During picking, the weft yarn is moved by the frictional drag between the air stream and the yarn surface. The driving force that moves the weft yarn form one side to the other side of the loom in air-jet insertion is determined by friction between the yarn surface and the air which is defined by the formula indicated in Equation (1).
Where: ρ -air density; [kgm− 3 ], C f -skin friction coefficient, X -is the direction of weft yarn movement, D -yarn diameter [m], L -Length of the weft along the reed width [m], V -weft yarn velocity [ms− 1 ], U -air velocity [ms− 1 ], and F f -friction force (Szabó, Patkó, and Oroszlány 2010). The yarn is pulled by the air at the tip rather than pushed from behind through the insertion to minimize buckling (Adanur 2020). As a result, the air pressure untwists the yarn which leads to loss of yarn strength after weaving as well as the quality of the fabric. The amount of twist determines the yarn behavior by providing lateral forces that prevent the fibers in the yarn from slipping over one another, which affect fibers' closeness together.
The level of twist loss varies depending on relay nozzle air pressure, the width of air-jet looms (Mohamed, Barghash, and Barker 1987), type of relay nozzles (Parekh et al.), weft yarn tension (Adanur 2020), and loom speed (Adanur and Mohamed 1988). "In general, 5 to 10% of twist losses were recorded due to untwisting of free yarn end. But the level of twist loss between open-end and ring-spun yarns shows differences" (Mohamed, Barghash, and Barker 1987).
Numerous researchers have examined the complex relationships between fiber types, yarn properties, and fabric structural parameters on weft yarn twist loss and fabric properties. Regarding the yarn properties, the linear density of the yarn (tex) increases as the percentage of polyester in the PC blend increases (Canoglu and Tanir 2009). It can be observed from Umair et al. (2017) that, the twist loss increases as the fineness of the yarn increases. The reason is finer yarns store more potential energy due to the higher twist insertion in the finer yarn and as a result of this the compressed air not only untwisted the yarn but also accelerated the twist loss process.
Moreover, the amount of twist level governs the strength and structure of the yarn such as, elongation, dye absorbency, appearance, drape, and other fabric characteristics (Parekh et al.). The twist loss affects the strength of the weft yarn (Brun, Corti, and Pozzetti 2008) and makes a difference in the properties of the fabric. The strength of the yarn woven by air-jet weaving is significantly lower than that of shuttle loom and other shuttle-less looms like a rapier, projectile, and water jet looms. These lead to poor tensile and tear strength of the fabric (Dhamija and Chopra 2007).
This study was carried out to figure out the effect of the cotton/polyester blend ratio on the twist loss of weft yarn during air-jet weaving. The twist loss of the weft yarn with different proportions of PC blend ratio was tested based on different weaving speeds and relay nozzle air pressure. Finally, the twist loss of PC blend weft yarn as well as the tensile properties of fabrics were measured with the contrast experiment.
The ANOVA is used to determine whether the factors have a significant influence on the response variable or not. The first hypothesis is that there is no difference in mean values of the weft yarn twist level as well as in yarn tensile strength losses when PC blend ratio, air pressure, and loom speed are changed. The alternate hypothesis is that there is a mean difference between the values of weft yarn twist level and yarn strength lose when there is a change in PC blend ratio, speed, and nozzle air pressure.
To determine whether any of the differences between the mean values are statistically significant, 95% confidence interval has been used. A significance level of 0.05 indicates a 5% risk of assuming that a difference exists when there is no actual difference. If the p-value is less than 0.05 (p ≤ .05), we can reject the null hypothesis and conclude that not all population means are equal. However, if p > .05, the difference between the means is not statistically significant.

Materials
Two different cotton-polyester blend weft yarns and one 100% cotton yarn produced by a ring spinning machine (Reiter G35 model) with 20 Ne count was used. The weft yarns were obtained from Combolcha Textile Share Company, Combolcha, Ethiopia, with the following specification (Table 1).

Yarn property testing
The following properties of weft yarn before and after weaving (unraveled from the fabric) were tested.

Yarn twist losses.
Yarn twist is defined as the number of turns in a given length of yarn. The average twist of yarns can be determined using the conventional untwisting method, which involves untwisting the yarn until it has no twist and counting the number of turns. To determine the number of turns lost during the weaving process, the number of twists was measured by unraveling the weft yarn from the body of the fabric according to ISO 2016:2015 standard with a gauge length of 500 mm using a quadrant type tester with untwist-retwist methods. As recommended by (Wang, Liu, and Hurren Deakin 2008), the 500 mm length of yarn was taken 150 mm away from the selvage on both sides of the fabric, as indicated in Figure 1. Because the fabric within 150 mm of the selvedge can change yarn properties and they are no longer representative of the bulk. Ten replicates have been taken for each experiment and an average value have been taken for analysis.  Yarn tensile strength. The tensile strength properties of unraveled yarns from the fabric were tested by using a Universal tensile strength tester (STATIMAT ME+ tensile tester) according to ISO 2062(ISO 2062-2013. Ten replicates have been taken for each test, and the average values were taken for analysis.

Designing the experiments and analysis
The main factors or independent variables of this research are PC blend ratio, loom speed, and two zones of relay nozzle air pressure (left and right). The responses are yarn twist loss and weft yarn tensile strength.
Design expert software with Box-Behnken Design has been employed to design the experiment and analyze the results of the experiment. Three level values are taken for all factors as indicated in Table 2. These levels are called high, medium, and low, respectively. 27 fabric samples have been produced depending on the design of the experiment as illustrated in Table 3.
The values of PC blend weft yarn twist loss and its effect on yarn properties after weaving were analyzed. The analysis includes regression analysis, ANOVA, effect of each factor on responses and a 3D response surface plot. Regression analysis is used to realize the empirical relationship between the factors and responses.

Result and discussion
In this study, twenty-seven fabric samples were produced on air-jet weaving machine based on the predetermined combination of PC blend ratio, loom speed, and air pressure (as shown in Table 4). The yarn and fabric properties such as weft yarn twist loss and weft yarn tensile strength losses were measured in the laboratory and the results of the measured values of each yarn and fabric properties were analyzed with design expert software (Box-behnken design).

Effect of loom speed, relay nozzles air-pressure and PC blend ratio on twist loss of weft yarn
Yarn twist is the number of turns within a certain length of yarn, and twist holds the fibers together and gives strength to the yarn (Rosiak and Przybyl 2003). Twist loss of weft yarn is determined by the twist difference between weft yarn from the supply package (cone) and the twist of unraveled weft yarn from the fabric. In this study, the effects of PC blend ratio, loom speed, and air pressure on the yarn twist loss are discussed as follows.
The ANOVA Table 5 with p-value <.0001 implies that the quadratic model is significant. In this study, loom speed, left-side relay nozzle air pressure, right-side relay nozzle air pressure, PC blend ratio, and its interaction effects (AC, AD, BC & CD) as well as quadratic terms (A 2 , C 2 and D 2 ) are statistically significant with p-values less than 0.05. This shows that the average twist of weft yarn before weaving has been found different from the average twist of weft yarn after weaving, and the twist loss is statistically significant. As a result, PC blend ratio, right side relay nozzle pressure, and loom speed have a great impact on weft yarn twist loss.
Depending on the above result, the alternate hypothesis was accepted. Because there is a mean difference between the values of weft yarn after weft insertion twist level when there is a change in PC blend ratio, loom speed, and nozzle air pressure during the air-jet fabric manufacturing process. A regression model equation has been developed to predict the level of twist loss at various values of PC blend ratio, loom speed, and left and right relay nozzle air pressure. The equation is used to make predictions about the response at the given levels of each factor. The quadratic model of regression after removing non-significant terms is given in Equation (2). The regression analysis shows loom speed, PC blend ratio (when the amount of cotton increases in the blend), and interaction effect (AC and CD) have a negative correlation with twist loss of weft yarn. While left-side relay nozzle air pressure, right-side relay nozzle air pressure, interaction effects (AD and BC), and quadratic terms (A 2 , C 2 , and D 2 ) have a negative correlation with weft yarn twist loss during air-jet weaving.
Twist loss ¼ 60:67 À 8:92A þ 4:54B þ 14:58C À 15:21D À 4:25AC þ 5:5AD þ 4:5BC À 4CD þ 4:6A 2 þ 7:6C 2 þ 3:67D 2 The positive relationship between the factor and the response indicates that twist loss of weft yarn increases when the value of independent variables has increased. On the other hand, the negative coefficients show that when the value of input variables is increased, the twist loss of weft yarn is reduced. The best fitted model for twist loss of weft yarn is a quadratic model, with coefficient of determination (R 2 ) value of 0.9791, as shown in Table 6. This suggests that the analyzed factor explains 97.91% of the twist loss for 20 Ne weft yarn. Furthermore, the predicted R 2 of 0.8856 is in reasonable agreement with the adjusted R 2 of 0.9547 as the difference is less than 0.2 (i.e., 0.069 < 0.2).
In Figure 2, the scattering graphs of the actual vs. predicted plot tell us how the model is performing. The Y-axis displays the predicted values from the model, and the X-axis displays the actual values from the experiment. The diagonal line in the middle of the plot is the estimated regression line. The results of twenty-seven fabric samples are closest to the fitted line, which indicates the model is well designed. Therefore, a strong correlation is achieved between predicted and actual values for twist loss of different proportions of PC blend weft yarns. Actual vs predicted values of twist losss (a), and yarn tensile strength loss (b)

The effect of loom speed on weft yarn twist loss
In air-jet weaving, loom speed is a critical factor which affect the property of the weft yarn. When the insertion speed increases, weft yarn twist loss reduces due to the yarn moving rapidly inside the warp shade, as shown in Figure 3(a). Furthermore, the pick insertion time is shorter than the time needed to complete redistribution of the twist throughout the entire weft length. When loom speed increases from 300 to 550 rpm, twist loss reduces rapidly which is also in agreement with agreement with (Umair et al. 2017; Zegan and Ayele 2022).

The effect of air pressure on yarn twist loss
The twist loss of weft yarn is directly proportional to air pressure. When the amount of air pressure increases, yarn twist loss increases due to the frictional drag between the air stream and the yarn surface, which leads to untwisting the yarn during the insertion time. Figure 3(b,c) shows the weft yarn twist loss increasing slowly when the left-side relay nozzle air pressure increases from 2.5 to 4 bar. But, the level of twist loss is very high at the right-side relay nozzles. Because, the yarn on the right side traveled higher distance than the yarn on the picking side. Furthermore, the right-side relay nozzles apply more pressure to the yarn, leading to increasing the yarn and air interaction, causing more twist loss on the receiving side. This is also in agreement with Yao-Qi (1984) and Parekh, and Raichurkar. Table 6 shows that PC blend ratio, right-side relay nozzles, and loom speed have a great impact on weft yarn twist loss. But the PC blend ratio has more influence compared to other input factors. Figure 3(d) indicates that the twist loss percentage of PC blend yarn is higher than cotton yarns. The higher rigidity of polyester fiber is the possible reason for the high twist loss of weft yarn. Because of the more flexural rigidity of polyester fiber, PC blend weft yarns have a tendency to untwist more during the weft insertion process of the air-jet loom compared to cotton yarn at the same level of twist. As a result, polyester fibers or their blends in yarn have a higher tendency to untwist during the weft insertion process. As a result of this research, we can conclude that lower twist loss is obtained at 100% cotton yarns due to better twist retention behavior and the highest twist loses is observed at 50/50 pc blend due to the reason mentioned earlier and this is also supported by (Umair et al. 2017).

3D Surface graph for factor interaction effect on yarn twist loss
The 3D response plot in Figure 4(a) shows that the minimum twist loss is observed (49 turns/m) at 550 rpm of the loom and 2.5 bar left-side relay nozzle air pressure, and the maximum of 82 turns/m is obtained at 300 rpm and 4 bar left-side relay nozzle air pressure at a constant PC blend ratio of 25%/ 75% and 4.75 bar right-side nozzle air pressure. At a constant left side relay nozzle pressure and PC blend ratio, the minimum (55 turns/m) twist loses is observed at 550 rpm and 3 bar right side relay nozzle pressure and maximum (101 turns/m) twist losses is observed at 300 rpm and 6.5 bar right side relay nozzle pressure, as shown in Figure 4(b). Figure 4(c) shows that twist loss of weft yarn is reduced with increasing loom speed and by increasing the percentage of cotton in the blend. It can be observed from the surface graph that the maximum losses (94 turns/m) occurred at 300 rpm speed of the loom and 50/50% PC blend ratio and the minimum losses (56 turns/m) occurred at 550 rpm speed of the loom and 100% cotton yarns at a constant left and right-side relay nozzle air pressure. The result shows that PC blend ratio has more significant effect compared to the working speed of the air-jet loom. The interaction effects of left side relay nozzle pressure and right-side relay nozzle pressure in Figure 4(d) show that the minimum (52 turns/m) twist loss occurred at 4 bar left side relay nozzle pressure and 3 bar right side relay nozzle pressure. Whereas the maximum (94turns/m) loss of twist was obtained at 6.5 bar right side relay nozzle pressure and 4 bar left side relay nozzle pressure.
The 3D response graph in Figure 4(e) shows that the minimum twist loss of 44.5 turns/m was obtained at 100% cotton weft yarn and 2.5 bar left side relay nozzle pressure, and the maximum twist loss of 87 turns/m was obtained at 50%/50% PC blend ratio and 4 bar left side relay nozzle pressure at a constant loom speed and right-side relay nozzle pressure air pressure. In Figure 5(f), the minimum (45 turns/m) twist loss occurred at 100% cotton yarn and 3 bar right side relay nozzle pressure, and the maximum (104 turns/m) loss was obtained at 6.5 bar right side relay nozzle pressure and a 50/50% PC blend ratio at a constant speed of loom and left side relay nozzle air pressure. The interaction between the right-side relay nozzle air pressure and the PC blend ratio has a significant effect on weft yarn twist loss.

Effect of loom speed, relay nozzles air-pressure, and PC blend ratio on strength loss of weft yarn
The single yarn strength obtained by unraveling from the body of the fabric is always lower than that of the single yarn strength from the package. Because the yarn strength while in the fabric is affected by friction due to interlacement and cohesive forces between the adjacent yarns, in addition to fabric construction parameters. In general, the main factors that affect the yarn strength loss are the type of fiber and their properties, yarn structure, and machine setting during the fabric construction process (Fiori, Brown, and Sands 1954). In this section, the effect of machine settings such as loom speed and nozzle air pressure as well as the effect of PC blend ratio on the yarn strength loss were analyzed. The best fitted model for yarn strength loss is a quadratic model, with R 2 value of 0.9849, as shown in Table 6. This suggests that the analyzed factor explains 98.49% of the strength loss of the yarn.  Therefore, a strong correlation is achieved between predicted and actual values for strength loss of different proportions of PC blend weft yarns, as indicated in Figure 3(b).  The ANOVA results in Table 7 show that the model is significant with p-value <.0001. In addition, all factors, CD interaction and quadratic terms (A 2 , C 2 and D 2 ), are also significant. This indicates that the average strength loss of weft yarn before weaving has been found different from the average strength loss of weft yarn after fabric formation, and that the yarn strength loss is statistically significant with a 0.05 significance level.
Depending on the above result, the alternate hypothesis was accepted for all significant factors. Because there is a mean difference between the values of weft yarn strength before weaving (from the package) and after weaving (unraveled from the fabric) when there is a change in PC blend ratio, loom speed, and nozzle air pressure during the air-jet fabric manufacturing process.
The regression analysis after removing non-significant terms is given in Equation (3). Depending on the equation, the loom speed and PC blend ratio have a negative correlation with the strength loss of weft yarn. Whereas, left-side nozzle air pressure, right-side nozzle air pressure, interaction effect (CD), and quadratic terms (A 2 , C 2 , and D 2 ) have a positive correlation with weft yarn strength loss. The equation can be used to make predictions about the strength loss of weft yarn at given levels of each factor, and the factor coefficients are used to identify the relative impact of each factor.
Yarn strength loss ¼ 0:65 À 0:325A þ 0:101B þ 0:288C À 0:546D þ 0:108CD þ 0:1A 2 þ 0:19C 2 þ 0:196D 2 (3) The effect of loom speed on weft yarn strength loss ANOVA Table 7 shows that loom speed has a significant impact on weft yarn strength loss. The strength loss of weft yarn is reduced when the insertion speed increases during air-jet weaving. High insertion speed increases the yarn feeding speed and decreases the total cycle time. As a result, the shortest weft picking time leads to the lowest possibility of changing the properties of the yarn structure. So, it can be concluded that the higher the loom speed, the lower the strength loss, which is in agreement with Zegan and Ayele (2022). Figure 5(a) indicates that when the loom speed increases from 300 to 550 rpm, the yarn strength loss reduces from 1.2 cN/tex to 0.46 cN/tex.

The effect of air pressure on weft yarn strength loss
The compressed air pressure in the air-jet loom affects the strength of the yarn in particular and the quality of the yarn in general. Table 7 shows that both left-side and right-side relay nozzle air pressure have a significant impact on weft yarn strength loss. But the level of strength loss between left-side and right-side relay nozzle air pressure has shown the differences, as shown in Figure 5(b,c). That means right side relay nozzle pressure has more impact on strength loss due to the fact that yarn on the right side is exposed for a longer period of time compared with the left-side. This is in agreement with Umair et al. (2017). When the left-side relay nozzle pressure increases from 2.5 bar to 4 bar, the yarn strength loss increases slightly, as shown in Figure 5(b). Whereas, as shown in Figure 5(c), as the rightside relay nozzle air pressure increases from 3 bar to 6.5 bar, the amount of strength loss increases rapidly.

The effect of PC blend ratio on weft yarn strength loss
The lowest single yarn strength loss was found in 100% cotton weft yarns, while the maximum occurred in 50/50% PC blend weft yarns. This is due to the higher TPI loss occurring in the high amount of polyester fiber in the blend, as shown in Figure 4(d). Polyester fiber makes the yarn more compact and smoother. As a result, it reduces the yarn velocity and increases the insertion time (Colak and Kodaloglu 2004). This leads to an increase in the reduction of yarn strength due to high TPI loss (Umair et al. 2017).

3D Surface plot for factor interaction effect on yarn strength loss
As indicated in Figure 6, the minimum (0.12 cN/tex) strength loss occurred at 100% cotton weft yarn and 3 bar right-side nozzle air pressure. But the maximum (1.72 cN/tex) was obtained at 50% polyester in the blend and 6.5 bar RSRNP at a constant speed of loom and left side relay nozzle air pressure. The interaction between both factors has a significant effect.

Conclusion
The research confirmed that there is a significant level of twist loss in the weft yarn because of the presence of the free leading end. With an increase in the loom speed, the percentage of twist loss is lower due to the shortest time weft yarn spent in the shade. When the amount of air pressure increases during the insertion period, weft yarn twist loss increases due to the higher frictional drag between the air stream and the yarn surface. Polyester fibers or their blends with cotton in the yarn have a higher tendency to untwist compared to cotton yarns during the pick insertion process. Because of the more flexural rigidity of polyester fiber, PC blend weft yarns have more of a tendency to untwist during the weft insertion process than cotton yarn at the same level of twist. In addition, a higher percentage of polyester in the PC blend lowers the hairiness value. Furthermore, reducing loom speed increases the weft yarn strength loss significantly because the weft yarn was exposed to air for a longer period of time. Relay nozzle air pressures are directly proportional with and weft yarn strength lose and inversely related with PC blend ratio (when the amount of cotton in the blend increases). In general, the results of the current research indicated that the maximum loss of twist and weft yarn strength loss were obtained at 425 rpm, 3.25 bar left side relay nozzle pressure, 6.5 bar right side relay nozzle air pressure, and a 50/50% PC blend ratio.