Prediction of fabric pilling tendency based on blended worsted fabric construction

Abstract

This article investigates a comparison between three different models to predict pilling rates of blended worsted fabrics using the ICI box testing method. Regression analysis was used to relate between weft yarn counts, picks per inch, fabric weave structure and pilling rates. Based on three different selected materials of weft yarns, a coding system was established to differentiate between single and plied weft yarn counts. The results showed an independent equation for every model, whereas the R-square values were acceptable for all models. Furthermore, fabric weave structure was found as the closest variable affecting pilling rates in two models when considering the value of coefficient of correlation.