Validation of DEM using macroscopic stress-strain behavior and microscopic particle motion in sheared granular assemblies

Abstract

Validation and/or calibration of distinct element method (DEM) models is usually performed by comparing element test simulation results with the corresponding stress-strain relationships observed in the laboratory [1]. However, such a validation procedure performed at the macroscopic level does not ensure capturing the microscopic particle-level motion [2]. Thus, the reliability of the DEM model may be limited to some stress paths and may not hold when the material response becomes non-uniform for example when shear bands develop. In this study, the validity of the DEM is assessed by comparing the numerical result with experimental data considering both particle-scale behavior (including particle rotations) and macroscopic stress-strain characteristics observed in shearing tests on granular media. Biaxial shearing tests were conducted on bi-disperse granular assemblies composed of around 2700 circular particles under different confining pressures. Particle-level motions were detected by a novel image analysis technique. Particle rotations are observed to be a key mechanism for the deformation of granular materials. The results from this study suggest that to properly calibrate DEM models able to capture the mechanical behavior in a more realistic way particle scale motions observed in laboratory experiments along with macroscopic response are necessary.