Effect of Specimen Size and Boundaries on the Results of Direct Simple Shear Tests

Direct simple shear (DSS) tests are widely used to capture the shearing behavior of soils. A major concern with a DSS test is the non-uniform distributions of stress and strain on the horizontal platens of this device. A series of DSS simulations were carried out in this study using 3D discrete element analysis (DEM) to examine the effects of specimen size and boundary conditions. The gradation of a coarse sand was used to generate different sphere sizes, which were then uniformly spread throughout the model. While smooth lateral walls were used, the horizontal boundaries had a high coefficient of friction to minimize slippage. A constant-volume condition was imposed by controlling boundary displacements. Different specimen diameters were modeled to examine the effect of diameter-to-height ratio (D/H) on stress path, peak shear stress, and post peak behavior. Numerical simulations indicated that boundary effects became more prominent at higher consolidation stresses for a given D/H. Such boundary effects were, however, reduced by enlarging the sample diameter (i.e., raising D/H). Specimen boundary effects were minimized in specimens with a D/H ≥ 2.8, resulting in the same shearing behavior as those with larger D/H ratios.