The Development Of New Sports Fabrics And Fabric Subjective Heat And Moisture Comfort Intelligent Prediction Model

Based on predecessors' research about the thermal-wet transmitting mechanism of fabrics, a series of fabrics with thermal-wet comfort properties were developed in this paper. The static thermal-wet transmitting indexes and the temperature and humidity changing curves in the process of dynamic thermal-wet transmitting of 36 kinds of knitting fabrics were tested and analyzed. The subjective thermal-wet comfort properties of these fabrics were assessed by wearing tests. The two prediction models of fabric subjective thermal-wet comfort properties based on BP neural network by taking static objective indexes as input parameters and by taking dynamic objective indexes as input parameters were founded. And the intelligent prediction software facing subjective thermal-wet comfort properties of fabrics based on BP neural network model with Matlab program language was developed. It included the following contents.For the present three-layer plating stitch has a long-standing problem that the plating effect is not good, and the ground yarn is easy to appear on the face side of fabric, a novel weft plating yarn carrier with three guide eyes suited for large cylinder diameter was designed. The three-layer elastic knitted fabrics with thermal-wet comfort properties were developed, by using cotton yarn outside, superfine filament or special-shaped section filament inside and lycra in the middle. Superfine filament or special-shaped section filament has good capillary effect, and cotton yarn has nice moisture absorbability. The cotton yarn outside could absorb the sweat transmitted by the filament inside and release the sweat to the atmosphere to keep the skin dry. The lycra in the middle offered the fabric favorable elasticity. At the same time, a serious of thermal-wet comfort knitted fabrics made of "U" shaped section filament were developed.36 kinds of knitted fabrics suited for making sports wears were selected. The 6 static thermal-wet transmitting objective indexes of air permeability rate, moisture transmission rate, wick height, moisture regain rate, drying rate and thermal resistance were tested and analyzed.For the existing microclimate heat and wet transmitting testing equipments were focused on sweating hot plate which could only test one-dimensional heat and wet transmitting of fabrics, a new microclimate testing equipment, sweating cylinder, was designed to test the thermal-wet transmitting properties of fabrics. This equipment could simulate human body shape to test two-dimensional hot and wet transmitting of fabrics, which was closer to the people real wearing state. Using this sweating cylinder, the temperature and humidity changing curves of the 36 kinds of fabrics in the microclimate were tested from sweating start to sweating end. At the same time, the 5 indexes of K_ts,T_equ,â–³T,K_Te,RH_equ, which most comprehensively reflected the dynamic thermal-wet comfort properties of fabrics, were selected. The temperature and humidity changing curves of different fabrics were contrasted and analyzed.The 4 subjective thermal-wet comfort indexes of mugginess feel, dampness feel, stickiness feel and cold feel after exercise of the 36 kinds of fabrics were assessed by wearing tests. Using specialist appraising method, the weight values of the 4 subjective feeling indexes to the comprehensive assessing of clothing thermal-wet comfort were gained, and then based on weight sum method, the comprehensive assessing values of clothing thermal-wet comfort were calculated. The results showed that the single eyelet fabric made of "U" shaped section filament with good wicking ability had the best thermal-wet comfort properties.Based on the results of static objective experiments, dynamic objective experiments and subjective wearing experiments, two prediction models of fabric subjective thermal-wet comfort properties based on BP neural network by taking static objective indexes as input parameters and by taking dynamic objective indexes as input parameters were founded. At the same time, the modeling values and the measured values of two models were contrasted. The results showed that the two models both had comparatively high predicting accuracy, but the predicting accuracy of the model by taking dynamic objective indexes as input parameters was higher. To contrast with the predicting accuracy of the models based on BP neural network, two prediction models of fabric subjective thermal-wet comfort properties based on multiple linear regression by taking static objective indexes as independent variables and by taking dynamic objective indexes as independent variables were founded. From the results of the modeling values and the measured values, it showed that the models based on BP neural network had higher predicting accuracy than the models based on multiple linear regression.The visualization technology was used in the fabric subjective thermal-wet comfort prediction. The intelligent prediction software facing subjective thermal-wet comfort properties of fabrics based on BP neural network with Matlab program language was developed. This software could directly predict the 4 subjective thermal-wet comfort indexes by inputting the 5 dynamic thermal-wet comfort objective indexes or by inputting the 6 static thermal-wet comfort objective indexes in the visual interface. The aim of fast predicting the subjective thermal-wet comfort properties of fabrics was realized.