A Multiscale Bridging Material Parameter and Damage Inversion Algorithm from Macroscale to Mesoscale Based on Ant Colony Optimization

Abstract

The study of material failure phenomena with coupled multiple scales is a need and a challenge. The fundamental difficulty in these problems is due to multiscale bridging material parameters and damage, which ensure that different scales of analysis should be concordant. To address this issue, an inversion algorithm based on ant colony optimization was developed to establish a bridge from macroscale to mesoscale. By using the developed algorithm, mesoscopic parameters including elasticity modulus and damage can be predicted by relying solely on the macroscopic mechanical response, which can be easily obtained from experiments. The developed algorithm can promote the development of multiscale mechanics, which can support an effective method to determine harmonious material parameters and damage in different scales. As a case study, the multiscale bridging inversion results, including elasticity modulus, stress and strain, and damage from macroscale to mesoscale of a concrete specimen, are given and compared with the corresponding experiment. The ability of the developed inversion algorithm is verified.