Response surface methodology and artificial neural network modeling of work of adhesion on plasma-treated polyester-cotton-woven fabrics

This paper aimed to study the effects of plasma treatment parameters on polyester-cotton of woven fabric surfaces through the work of adhesion test using artificial neural network (ANN) and response surface methodology (RSM). This study used plasma treatment parameters, such as electrode distance, voltage, and plasma exposure time, as inputs for the models. We used surface tension as a function of contact angle (theta) to measure the work of adhesion (w(SL)), the model's output. Results showed that adhesion is closely related to the selected input variables. In addition, the development of artificial neural networks and response surface methodology could predict the experimental data with the coefficient of determination results were 0.902 and 0.719, and the root-mean-square error (RMSE) values were 2.135138 and 3.685359, respectively. Based on this research, compared with SRM, ANN has higher accuracy in calculating the work of adhesion. We concluded that ANN is expected to be a valuable quantitative method to predict and understand the adhesion effect of plasma treatment on the surface modification of polyester-cotton woven fabrics. The novelty of this study is that we used both ANN and RSM for the first time to predict the work of adhesion of polyester-cotton woven fabric treated with corona plasma. The use of the artificial neural network to simulate and predict the effect of plasma treatment on improving the work of adhesion is another novelty of this research.

Automated summary

This study used artificial neural network and response surface methodology to investigate the effects of plasma treatment parameters on the adhesion of polyester-cotton woven fabric surfaces. The results showed a close relationship between adhesion and the selected input variables, with higher accuracy observed in the artificial neural network compared to the response surface methodology.