Orateur
Description
Understanding the micromechanics of granular flows is critical for modeling natural hazards and industrial processes, yet the role of statistical particle shape variability remains unclear. Indeed, in most reported studies all particles have the same shape while in natural granular materials there is generally a strong variability around an ‘average’ or ‘reference’ shape. Using 3D Discrete Element Method (DEM) simulations of assemblies of icosahedral particles under triaxial compression, we investigate how a geometrical perturbation around the reference icosahedral shape influences the transition from a dense static state to continuous critical-state flow. We find that at the stress peak state, macroscopic shear resistance is sensitive to shape variability, driven by transient geometric interlocking and spatial polarization of the strong-contact network. However, as the assembly tends towards a steady flow state, the effect of volume dilation reduces the effect of interlocking and thereby all flowing assemblies converge into a universal, shape-independent critical state. In this regime, macroscopic equilibrium is sustained exclusively by a residual frictional network where tangential forces play a uniquely dominant role compared to classical spherical models. Our results suggest that statistical geometric disorder can be employed as a parameter to control internal force-transmission routes from fabric-organized to friction-mobilized in complex granular flows.