Principal Component Analysis Of Processing Properties And Prediction Of Overfeed Sewing Shrinkage Rate For Light Wool Fabrics

Mechanical properties tested by FAST system that is very important controlling index in apparel process have an important impact on fabric’s handle and appearance quality of apparel overfeed sewing. Based on mechanical properties tested by FAST system, the objective evaluation method for wool fabric handle and sewing quality predicting model have established by experiment method with fabrics produced in 80s of last century. Since 90s, as the decreasing of wool fabric’s unit weight it needed to connect fabric property design with overfeed sewing design to guarantee apparel process quality. And nowadays the theory of relationship between fabric mechanical property and overfeed sewing is still not perfect. Light wool fabric production in our country is still in the development period, and has obvious differences compared with abroad light wool fabric. In order to analyze the main differences of light wool fabric processing properties in home and abroad and their influence on apparel overfeed sewing quality which could supply theory reference and data support to home light wool fabric production and apparel design. This thesis took home and abroad light wool fabrics as research object to research further theoretical relationship between mechanical property and overfeed sewing shrinkage and acquire predicting method of overfeed sewing shrinkage by means of finding processing property main differences. The thesis’s work included two parts: one was principal component analysis of processing properties of home and abroad light wool fabrics; the other predicted overfeeding sewing shrinkage of light wool fabrics.Principal component analysis of processing properties of light wool fabrics was illustrated respectively in chapter two, three and part one of chapter six. It selected thirty-five samples of home and abroad light wool fabrics (Foreign were twenty, home were fifteen). According to FAST controlling drawing, it selected fifteen processing property indexes as evaluation index by combined wool product uses. Through testing and calculating, it compressed fifteen indexes to seven comprehensive evaluation indexes. According to each index contribution rate, it defined respectively such as formability principle component, moisture expansion principal component, thick-weight principal component, relaxation shrinkage principal component, warp tensile elasticity principal component, shearing rigidity principal component, and appearance thickness principal component. Through calculating of principal component value, it could find the biggest difference of home and abroad light wool fabric was formability principal component, and then was fabric thick-weight property. Foreign fabric showed light and had bigger formability. It could take formability and thick-weight property as improved aim in order to enhance of home fabric’s quality. Five worsted fancy fabrics were developed by product design and process management. When testing, the woven fabric had reached expectation effect.Wool fabric overfeed sewing shrinkage was illustrated respectively in chapter four, five and part two of chapter six. Three parts were conducted in this research process. Firstly it built predicting math model of overfeed sewing shrinkage. The research made simplification premise to reduce problem that did‘t consider sewing thread apply force and friction apply force between fabrics, and ignored fabric weight, and took bending modulus as constant and so on. In base of applying big deformation mechanics equation, apparel design parameters that include fabric connection length ratio (overfeeding amount) and connection length (needle distance) were brought. It built math model of between fabric sewing shrinkage and bending–tensile modulus ratio (fabric bending modulus than tensile parameter), and fabric connection length, and fabric connection length ratio to link fabric property and apparel design. Through math analysis of math model, it could get the smaller of overfeeding sewing shrinkage when the fabric bending–tensile modulus ratio and the fabric connection length ratio were bigger, and the connection length was smaller. It derived the method for testing and calculating of fabric bending–tensile modulus ratio based on FAST testing system. It analyzed the relationship between bending–tensile modulus ratio and formability of FAST testing index which indicated that the bigger of formability, the bigger of bending–tensile modulus ratio. Because home fabric has lower formability and bending–tensile modulus, it could cause sewing shrinkage higher and produce crepe easily. The thesis provided the new sewing formability concept as a reference value to represent different fabric’s sewing shrinkage prediction in overfeeding sewing which generalized the application of formability. Secondly it further researched the change of fabric connection length ratio and connection length in overfeeding sewing process. Fabric was caused relaxation shrinkage and moisture expansion shrinkage under the moisture and hot affection. The thesis took water shrinkage as dimensional stability index to analyze its influence on fabric connection length and fabric connection length ratio, which revealed moisture expansion rate and relaxation shrinkage rate would influence overfeeding sewing shrinkage in overfeeding process.It deduced calculating method of oblique shrinkage based on warp and weft shrinkage which supplied the base for fast reaction. Lastly in order to revise error of model premise and raise predicting accuracy, it selected seven indexes such as bending–tensile modulus ratio, fabric connection length, fabric connection length ratio, moisture expansion rate & relaxation shrinkage rate of long and short fabrics, as input parameters of artificial neural network. By simulating arm overfeed sewing it founded predicting model of artificial neural network of overfeeding sewing, and compared with multiple linear regression model. The result showed artificial neural network model had an advantage over multiple linear regression model.Summary, principal component analysis of home and abroad light wool fabric sample made us know clearly the main differences of home light wool fabric which indicted the research direction and supplied data supporting for enhancing process quality of home light wool fabric. The founded predicting model of overfeeding sewing and method of error correction were not only to analyze the influence factors of overfeeding sewing quantitatively, but also could go on company quality management, supply guide for process reference’s optimization, and improve imperfection of predicting sewing shrinkage through experiment method at present. Meanwhile on theoretically it revealed an influence on apparel overfeeding sewing caused by process property differences of home and abroad fabrics.