Abstract
The surface roughness of fabric is one of the fabric properties that is used to evaluate the sensorial comfort of clothes. However, its objective evaluation requires sophisticated and expensive testing instruments and skilled testing expertise. Developing a predictive model is therefore an alternative approach to overcome such limitations. This study investigated a regression model to predict the surface roughness of 3/1 twill fabric using weft yarn count and weft thread density. Nine samples were produced by varying the weft yarn count and weft thread density at three different levels, while their surface roughness was determined using a Kawabata instrument (KES-FB4) under standard testing conditions. A two-factor predictive model equation was developed using design expert software. Based on the results and findings, the effects of count and density on the roughness of 3/1 twill fabric were found to be statistically significant for the developed model at a confidence interval of 95%. The model was tested by correlating the measured and predicted surface roughness values of 100% cotton 3/1 twill fabric. The results of the model test indicate a significant correlation (R2 = 0.9644) between the measured and predicted surface roughness values of 3/1 twill fabric, with a 95% confidence interval.
Model validation was performed, and the study showed that the measured and predicted values of the surface roughness of 3/1 twill fabric have a 0.828 coefficient of determination (R2). This indicates that the surface roughness of 3/1 twill fabric can be predicted well by weft yarn count and weft thread density. This model can be thus used in textile industries and by research institutes for predicting the surface roughness of 3/1 twill fabrics in the new product development process.