Response Surface Regression Models for Prediction of Ring Spun Yarn Properties

Abstract

In this study, the response surface methodology is used to predict the mechanical properties of yarn, their unevenness and hairiness by using the high-volume instrument (HVI) properties of raw cotton and the parameters of the spinning process. Therefore, five different blends of cotton are processed and spun into ring yarns (Nm13, Nm19, Nm 21, Nm31 and Nm37). Each count is spun at five twist levels (450, 500, 650, 750 and 850 trs/m).

The models that are developed by using response surface regression with many iterations on a Minitab16 statistical software predict very well the different yarn properties since the R² values obtained are very important. In addition, these models show that metric number and twist have the highest effect on the four studied parameters

Citation

Ghanmi, H., Ghith, A. and Benameur, T. (2015), "Response Surface Regression Models for Prediction of Ring Spun Yarn Properties", Research Journal of Textile and Apparel, Vol. 19 No. 4, pp. 1-10. https://doi.org/10.1108/RJTA-19-04-2015-B001