Prediction of fabrics mechanical behaviour in uni-axial tension starting from their technical parameters

Abstract

On the basis of a phenomenological study, which consists in determining the response of a fabric sample to the traction and the relaxation, we proposed a rheological model with 15 coefficients. This model of mechanical behaviour in uni-axial tension takes into account the elastic and the viscous components. To characterize the coefficients of this model, simple tension and relaxation tests were performed for 29 fabrics in the warp and the weft directions. An artificial neuronal network trained with a retro-propagation algorithm performs functional mapping between these mechanical coefficients and the technical parameters of fabric, which allowed us to predict the mechanical behaviour in tension and relaxation of fabrics starting only from their technical parameters.