Development of Stochastic Strength Model for Short Carbon Fiber Reinforced Thermoplastic

Abstract

Short carbon fiber reinforced thermoplastic (SCFRTP) is promising for use of the automobile members because of its high specific strength and moldability. An accurate prediction of its strength is necessary for practical use. Due to the uncertainty of carbon fiber distribution, it is difficult to obtain the stochastic property of strength of members without numerical investigation, to which complexities solved by high performance computing is associated. In this study, we develop a stochastic macro strength model to predict minimum strength of SCFRTP specimen in terms of fiber volume fraction and fiber direction tensor regarded as spatial stochastic variables. As a first step, we have obtained a three-dimensional full model of specimen, where carbon fiber and resin are clearly separated, via X-ray computed tomography. We divided this full model into small volumes, and fracture simulations for both models to obtain macro stress-strain curves. We investigate methodology to constitute macro strength model through the comparison of both simulations in regard to stochastic homogeneity.