Abstract

This paper focuses on a Finite Element (FE) based approach with micromechanical RVE (Representative Volume Element) method for predicting nonlinear material properties for three-dimensional (3D) composite materials with randomly distributed inclusions. Assuming the RVE is uniformly repeated over the domain of an entire structure, the effective constitutive properties of the RVE will characterize the entire domain as well. A 3D RVE generator is developed to facilitate automatically one or multiple types of material inclusions that are randomly distributed inside the RVE. The RVE geometry are symmetric on each pair of two opposite boundary faces - totally 3 pairs of a cubic RVE. To achieve fast virtual tests, an automatic meshing method is developed and employed to generate a complete periodically symmetric mesh for the RVE to serve a rigorous homogenization compute process for various configurations of the RVE with different inclusion distributions. This procedure can easily apply to RVEs with different randomly distributions of inclusions to generate meshes. Analysis procedures are defined with respect to loading cases based on far-field solution to obtain local solution fields from finite element analysis. Different cases of the interfaces between inclusions and matrix are investigated in this study: perfectly bonded and debonded using interfacial damage initiation and evolution with cohesive behaviors. In the computational study, a new Abaqus iterative solving technology is used to solve the large models with high performance than direct solving algorithm. Using a rubber material as an industrial application of this developed solution, the calibrated nonlinear material properties from RVE virtual test results are applied in a steady-state rolling analysis of a tire traveling at a ground velocity of 10.0 km/hr on a flat rigid road surface. The results of rubber material with hard inclusions predict differences in the free rolling equilibrium solutions and the contact pressure distribution between the tire and the road comparing with the results without inclusions.