Complex network and fabric driven non-affine kinematics of aggregate-rubber mixtures

The ensemble behavior of aggregate-rubber (AR) mixtures during shear is governed by the particle level rearrangements and their interaction especially due to rubber particle deformation. This study presents a quantitative framework that connects meso-scale network topology to macroscopic mechanical behavior, emphasizing the stabilizing influence offered by rubber particles on strain localization and the formation of force chains. This study conducts 2D numerical simulations with a coupled Discrete-Element Free Galerkin Method (DEM-EFG) to accurately model the deformation of individual rubber particles within granular assemblies. AR mixtures with different proportions of rubber content are subjected to simple shear to investigate non-affine kinematics at large strains. Complex network graphs are constructed to represent particle centers and their connections with neighboring particles, enabling the analysis of weighted and unweighted strong sub-networks. The network metrics reveal that the addition of rubber particles increases the density of triangular motifs and restricts the formation of larger cycle memberships, thereby reducing sliding contacts and particle rotation. The collapse dynamics of force chains are analyzed using surrogate network-based measures of kinetic energy, capturing the reduced temporal percolation of strain localization with addition of rubber particles. A key finding of this study is the slow destruction of small-sized cycles in the AR meso-structures, along with weak fabric anisotropy of AR contacts. This behavior is attributed to the reduced couple stress and the rapid decay of spatial correlation in non-affine displacement within AR mixtures with high rubber content.