Discrete element simulations of direct shear specimen scale effects

This paper presents a study of the micromechanics of granular materials as affected by the direct shear test scale using the discrete element method (DEM). Parametric studies were conducted to investigate the effects of specimen length and height scales (in relation to the particle size) on the bulk material shear strength and shear banding behaviour in the direct shear test. A mesh-free strain calculation method previously developed by the authors was used to capture and visualise the evolution of strain localisation inside the direct shear box. Simulation results show that the maximum shear strength measured at the model boundaries increases with decreasing specimen length scale and increasing specimen height scale. Micromechanics-based analysis indicates that the local and global aspects of fabric change and failure are the major mechanisms responsible for the specimen scale effect. Global failure along the primary shear band prevails when the specimen length scale and length to height aspect ratio are small, while progressive failure becomes more likely when the specimen length scale and aspect ratio become larger. Further quantifications of the specimen scale effects on the macroscopic behaviour of granular materials rely on the fundamental understanding of quantitative relationships between material fabric, anisotropy and bulk strength.