The Relationship Between Weave Structure And Wrinkle Resistance Properties Of Non-Finished And Finished Cotton Fabrics

The research on the relationship between cotton fabric wrinkle property and weave structure is rare in China at present. In other countries, the research of poplin-like fabrics is popular whereas plain cloth-like fabrics and the finishing agent is not the low-formaldehyde agent. So in this paper, 88 fabrics of different weave, different density are designed and tested. The relationship between fabric wrinkle property and weave structure is discussed and simulated by artificial neural network.Research shows that weave is the important factor in influencing wrinkle resistant property and tensile strength before and after finishing. Different weave has different properties. Weave factor C is in direct relation to wrinkle recovery angle (WRA) and in inverse relation to tensile property (TS). This trend also exists after wrinkle resistant finishing. Among the weaves studied in this paper, 3/1 left-twill fabric is one special weave. Excellent WRA and higher TS property can be obtained after finishing.Research also shows that fabric threads influences the WRA and TS before and after finishing. Fabric threads is in direct relation to (TS) and in inverse relation to (WRA). Higher WRA can be obtained when the parameters of warp threads is similar as that of the weft threads (Defined as Even structure). 3/1 left-twill fabric with Even structure can obtain best wrinkle recovery angle (WRA) and higher tensile strength retain (TSR).New Tightness coefficient (T) is developed in this paper. Research shows that T effects fabric WRA adversely for non-finished fabric and finished fabric. T is in direct relation to tensile strength and formaldehyde release. Weft tensile strength is much effected by T than Warp tensile strength. No distinct relationship can be found between T and TS loss.The relationship between weave structure and wrinkle resistant property is complex. In this paper, an artificial neural network has been applied for the estimation of WRA and TS. It is a high non-linear problem between the influencing factors and wrinkle resistant property, etc. On the basis of modified BP algorithm, training with a sequence of past time course dataof WRA and TS, a good neural network prediction model was obtained. The weight coefficients of each of the network' s interconnection are stored. The well trained network is estimated accurately and rapidly, both the experimental results and predicated activity values are very agreed.