Tekstil ve Konfeksiyon, cilt.27, sa.3, ss.289-299, 2017 (SCI-Expanded)
It is proposed that response surface designs with feasable region is an effective tool in prediction of a specific property of a product from the known properties of raw material, with a yarn irregularity exercise, for the first time in textile literature in this paper. The data used in this research is obtained from a cotton yarn spinning mill, being real production data. Response surface designs with feasible region are used to obtain the relationship between the response variable (yarn irregularity) and effecting factors (fiber properties) in 3D graphs and contour lines both for yarn in bobbin and cop form separately. It is concluded that this novel method of response surface designs with feasible region provides valuable results, is effective in prediction, is benificial in textile quality control, and can also be widely used in other industry branches when it is incorporated into a statistical quality control computer program.