Abstract
The present paper concerns the statistical analysis of the surface roughness evaluation of knitted fabrics by the Textile Surface Tester. The main objectives were, firstly, focused on investigating the effect of knitted fabrics structural factors and the test conditions on the surface absolute roughness, the total roughness and the standard deviation. Secondly, the relationship between sample characteristics (face, yarn count, loop length), the test conditions (the force and the slipping speed of the sensor feeler on the sample and signal sampling time), and the surface roughness parameters were analyzed and modeled through regression analysis. The combined effects of the input parameters and their two-way interactions on the test bench outputs were investigated using the analysis of variance (ANOVA). The percent contribution ratio was used to show the influence of inputs and their interactions on surface roughness parameters.
The results show how much surface roughness is mainly influenced by the knits structural factors. Also, it is underlined that the applied force by the sensor feeler on the fabric has an important effect on outputs. Finally, the sensor slipping speed on the sample and the signal sampling time have no important effects on outputs. Models were developed using experimental results from a full factorial experimental design. The adjusted coefficients of determination R2adj were found to be greater than 80%.