A novel method for generation of particle packing model used in numerical simulation for the mechanical behavior of multi-component material

Abstract

This study presents a novel algorithm to establish a numerical model for the study of mechanical behavior of multi-component material such as particle reinforced metal matrix composites, concrete, particle filled polymers, etc. The algorithm consists of two steps, one is to divide the given region into sub-regions of a specified size distribution, and the other is to fill each sub-region with particles of specified shapes. During the filling process, the contact states between particles can be controlled by the number and positions of reference points. Taking the quasi-static uniaxial tensile simulation of concrete as a verification example, two concrete models with the same particle size distribution but different shapes have been established. Through the comparison of elastoplastic responses of different aggregate shape models in different tensile directions, it is concluded that the models meet the requirements of macroscopic uniformity, meso-scale randomness and statistical rationality. This verifies the feasibility of modeling method and its advantages in precise control of particle shape, distribution and other parameters.