Thermoplastic Polyurethane Based on the 3d Printing Fashion Clothing-Conceptual Model of The Fashion Industry

Many researchers in the past have conducted studies on the use of 3D printing in the fashion industry. Vanderploeg et al., (2017) [2] conducted research on different applications of 3D printing in the fashion industry, Sun & Zhao, (2017) [3] presented a model on the impact and challenges of the use of 3D printing in the fashion industry, Chun, (2017) [4] studied the use of 3D printing technology to cultivate new designs, D. Sun & Valtasa, (2019) [5] conducted research on the integration of 3D printing technology into fashion clothing production processes, McCormick et al., (2020) [6] studied the Chinese fashion market for use of 3D printing, Spahiu et al., (2020) [7] researched 3D printing techniques.

Both in the fashion industry and in 3D printing, the material used plays a very important role. In this direction, one material that has started to grab the attention of the researchers, producers, and customers is thermoplastic polyurethane (TPU). It is known for its outstanding material properties like tear resistance, resistance to chemicals, abrasion resistance, skin compatibility, large elongation break, resistance to ozone, low-temperature resistance, resistance to oxygen, low long-term deformation rate, resistance to oil and fuels, antimicrobial properties, resistant to fungus, soft elasticity, and excellent bending and tension strength [9,10]

TPU can be coated onto a piece of fabric to deliver protection, comfort, and aesthetic values to the fabric. Owing to these properties and properties enlisted in Figure 1, TPU is considered to be a good choice for making garments and textiles. Owing to these aspects, many researchers have studied the use of TPU material for 3D printing fashion products. Kim et al., (2019) [1] and Han & Kim, (2018) [11] studied the use of TPU-based material for 3D printing fashion products, Jin & Li, (2020) [12] researched the development of 3D printing clothing using TPU material, Jeong et al., (2021) [13] used TPU material in 3D printing for developing parametric design in fashion products. Although TPU has certain limitations such as its rough feel and limitations in creating diverse designs such as those with severe curves, its biggest advantage is that it is sustainable, which makes it worth adopting.

Fig. 1

While many have studied 3D printing technology using TPU material in the fashion industry from different perspectives, fewer researchers have addressed the actual adoption of TPU-based 3D printing in the fashion industry. Thus, the present research has been focused to propose an adoption model for TPU-based 3D-printed fashion clothes.

The Chinese fashion industry is growing rapidly. This can be attributed to the fact that the purchasing power of the Chinese population is at its peak. Now, fashion has become an important for the Chinese population, and clothing is no longer a mere necessity. Moreover, fashion is now one of China's most lucrative industries. Not only local, but also foreign companies are looking for ways to get a strong foothold in the fashion market of the world's most populous country. Today, the Chinese fashion market is the second-largest apparel market in the world with an estimated value of USD 2 billion [14]. This scenario of the Chinese fashion industry makes it vital to study TPU-based 3D printed fashion clothes because TPU has the potential to be the future of fashion and the Chinese fashion market, which holds the maximum potential in the world as it has the biggest consumer market).

The items considered under Benefits of TPU-Based 3D printed fashion clothes have been enlisted in Table 1.

Table 2 presents KMO-Bartlett test. A high value of KMO (0.925) indicates that the factor analysis may be useful with the data collected for the Benefits of TPU Based 3D Printed Fashion Clothes. Since Bartlett's Test reported a statistically significant value of 0.000 (<0.05) further stating factor analysis may be useful with the data collected for benefits of TPU-Based 3D printed fashion clothes

Since the study reported appropriate values for KMO-Bartlett test, the researchers moved ahead with Factor Analysis as follows.

Here, high extraction values (>0.5) were obtained which indicate that the factors thus received are good as the proportion of each variable's variance that can be explained by the retained factors are high.

The total number of factors retained in the total variance explained test is two. Furthermore, these two factors account for 77.85% of the variance, which is a respectable number pointing towards the credibility of the retained factors. The number of factors retained can further be observed in the Scree Plot presented in Figure 1(a). It can be observed in Figure 1(a) that factor 1 and factor 2 possess the highest Eigenvalues, thereby contributing to the highest variance. Post the second factor, the line is almost flat, meaning each successive factor is accounting for smaller and smaller amounts of the variance.

Table 3 presents the items segregated under two factors. It can be seen B2 and B4 lie under factor 2 while all other items lie in factor 1.

On analyzing the items under factor 1 and factor 2 (Table 1), it was found that the items under factor 1 were Customer Related Benefits, while those under factor 2 were strength related. Thus, factor 1 was named Customer-Related Benefits, while factor was called Strength-Related Benefits

Items considered under Challenges of TPU-Based 3D Printed Fashion Clothes have been enlisted in Table 1.

Table 4 presents KMO-Bartlett Test. A high value of KMO (0.899) indicates that the factor analysis may be useful with the data collected for challenges of TPU based 3D printed fashion clothes. Since Bartlett's Test reported a statistically significant value of 0.000 (<0.05) further stating factor analysis may be useful with the data collected for challenges of TPU based 3D printed fashion clothes

Since the study reported appropriate values for KMO-Bartlett Test, the researchers moved ahead with Factor Analysis as follows

Here, high extraction values (>0.5) were obtained, which indicates that the factors thus received are good as the proportion of each variable's variance, which can be explained by the retained factors being high except for item C8 (0.202).

The total number of factors retained in the total variance explained test is two. Also, these two factors account for 74.121% of the variance, which is a respectable number pointing towards the credibility of the retained factors. The number of factors retained can further be observed in the Scree Plot (Figure 1 b). It can be observed in Figure 1 (b) that factor 1 and factor 2 possess the highest Eigenvalues, thereby contributing to the highest variance. Post the second factor, the line is almost flat, meaning each successive factor is accounting for smaller and smaller amounts of the variance.

Table 5 presents the items segregated under two factors. It can be seen C3 lies under factor 2 while all other items lie in factor 1. Further, the correlation value of C8 was extremely small, which is less than 0.5 owing to which it has not been categorized under any factor.

Re-running the test after the removal of item C8. Post removal of C8, the value of KMO increased from 0.889 to 0.904.

Here, high extraction values (>0.5) were obtained which indicate that the factors thus received are good as the proportion of each variable's variance that can be explained by the retained factors are high.

Post-removal of item C8, the total number of factors retained is still two. Further, the variance accounted for by these two factors post-removal of C8 increased from 74.121% of variance to 82.925% thereby enhancing the credibility of the retained factors. This depicts an improvement in the quality of the retained factors.

It can be observed in Figure 1(c) that factor 1 and factor 2 possess the highest Eigenvalues, thereby contributing to the highest variance. Thus, not much difference was observed from the Scree plot with C8 (Figure 1c).

Table 7 presents the items segregated under two factors. It can be seen C3 lies under factor 2 while all other items lie in factor 1 which is the same as was the case with a rotated component matrix with item C8 (Table 10). Further, the correlation value of C3 can be observed to be increased from 0.898 to 0.996 post-removal of C8. On analysing the items under factor 1 and factor 2 (Table 1), it was found that the items under factor 1 were TPU material related challenges while those under factor 2 were printer related. Thus, factor 1 was named TPU Material Related Challenges while factor was called Printer Related Challenge.

The items considered under Motivation to adopt TPU-Based 3D Printed Fashion Clothes for Customers have been enlisted in Table 1.

Table 8 presents KMO-Bartlett Test. A high value of KMO (0.897) indicates that the factor analysis may be useful with the data collected for Motivation to adopt TPU-based 3D printed fashion clothes for customers. Since Bartlett's Test reported a statistically significant value of 0.000 (<0.05) further stating that factor analysis may be useful with the data collected.

Since the study reported appropriate values for KMO-Bartlett Test, the researcher moved ahead with Factor Analysis as follows. Here, high extraction values (>0.5) were obtained, which indicate that the factors thus received are good as the proportion of each variable's variance that can be explained by the retained factors are high.

The total number of factors retained in the total variance explained test is two. Furthermore, these two factors account for 98.985% of the variance, which is a respectable number pointing towards the credibility of the retained factors.

The number of factors retained can further be observed in the Scree Plot (Figure 1(d))

It can be observed in Figure 1(d) that factor 1 and factor 2 possess the highest Eigenvalues, thereby contributing to the highest variance. Post the second factor, the line is almost flat, meaning each successive factor is accounting for smaller and smaller amounts of the variance.

Table 9 presents the items segregated under two factors. Some cross-loadings can be observed. Thus, the factor with a higher correlation was selected to keep the item. Herein M1 to M6 has been allotted to Factor 2 while M7 to M15 are under factor 1.

On analyzing the items under factor 1 and factor 2 (Table 1), it was found that the items under factor 1 revolved around customers, while those under factor 2 were focused on brands. Thus, factor 1 was named Brand Specific Motivation, while factor 2 was called Customer-Centric Motivation.

Table 10 presents KMO Bartlett Test. A high value of KMO (0.795) indicates that the factor analysis may be useful with the data collected for Attitude towards the adoption of TPU-Based 3D Printed Fashion Clothes for Customers. Since Bartlett's Test reported a statistically significant value of 0.000 (<0.05) further stating factor analysis may be useful with the data collected.

Since the study reported appropriate values for KMO-Bartlett Test, the researcher moved ahead with Factor Analysis as follows. Here, high extraction values (>0.5) were obtained which indicate that the factors thus received are good as the proportion of each variable's variance that can be explained by the retained factors are high except.

The total number of factors retained is four. Furthermore, these four factors account for 78.828% of the variance, which is a respectable number pointing towards the credibility of the retained factors. The number of factors retained can further be observed in the Scree Plot (Figure 1(e).

It can be observed in Figure 1 (e) that factors 1, 2, 3, and 4 possess the highest Eigenvalues thereby contributing to the highest variance. Post the fourth factor, the line is almost flat, meaning each successive factor is accounting for smaller and smaller amounts of the variance.

Table 11 presents the items segregated under 4 factors. Some cross-loadings can be observed. Thus, the factor with a higher correlation was selected to keep the item. Herein CK1 to CK3 and PI1 to PI3 are placed under factor 1, SN1 to SN3 under factor 2, PN1 to PN3 under factor 3, and finally, EC1 to EC3 under factor 4. The allotment of all the factors is the same as was originally selected by the researcher except for “purchase intention” (PI1 to PI3, Table 24) which was considered as a separate factor by the researchers but after factor analysis, it was found to be grouped with “customer knowledge” (CK1 to CK3, Table 24) that is under factor 1. Thus, factor 1 was named Customer Knowledge and Intention.

Figure 1 presents the Scree Plot of all the above factors.

To assess the association of the above factors on “Adoption of TPU-Based 3D Printed Fashion Clothes” it is imperative to determine the factor structure (model). For doing so, exploratory factor analysis (EFA) has been conducted. EFA will facilitate variable reduction thereby identifying the number of latent constructs and the underlying factor structure of a set of variables.

As per the perception of the researcher, all the above factors might lead to the adoption of TPU-based 3D printed fashion clothes. Thus EFA has been conducted to establish a model for the same based on the factors extracted via factor analysis conducted in previous sections.

The model thus obtained post exploratory factor analysis has been presented in Figure 2. The analysis of model fit has been conducted as follows.

Fig. 2

Minimum discrepancy per degree of freedom (CMIN/DF)< 3 indicates an acceptable fit between the hypothetical model and sample data [24]. Thus, the proposed model of the present research has an acceptable fit based on CMIN/DF (2.621, Table 12).

NFI value between 0.8 and 0.9 is considered to be acceptable [25]. Thus, the NFI value (0.80, Table 12) in the present research makes the proposed model acceptable. Further, a TLI value of more than 0.9 is considered to be a good fit [25] but in the present research, the TLI value is lower (Table 12). Thus, the research has considered other matrices to base the results on. Finally, a CFI value between 0.8 and 0.9 is considered acceptable which is the case with present research as the CFI value obtained is 0.856.

For FMIN Model Fit Indices, the more the value is close to 0, the better the model fit [26]. Since the values of FMIN Model Fit Indices in the present research are close to zero, it points towards a good model fit.

According to MacCallum et al., (1996) [27], the RMSEA value of less than 0.08 is an acceptable fit, thereby making the proposed model in the present research acceptable.

Since all the model fit indices are within acceptable limits, the model can be stated to be a good fit. Thus, the proposed model has been presented in Figure 2.