Dynamically Propagating Crack in Graded Particle Dispersed Composite Modeled by Neuro-Fuzzy Inference

摘要

A computational method for analyzing a particle dispersed material is presented. The potential difficulty of developing a constitutive law for the material especially in the nonlinear regime is circumvented by developing a computational constitutive modeling that employs a FEM mesomechanics analysis of a grain model, neural network, and fuzzy inference techniques. Dynamic crack propagation in the functionally graded material(FGM) subjected to a tensile load or a thermal shock is demonstrated to find the grading effect of the material on crack severity and plasticity as the crack propagates in FGM.