For the purpose of improving computer simulation of granular materials, a three-dimensional discrete element model using random arrays of elastic ellipsoidal particles is developed. Major technical implementations relating to the development of this new numerical model (computer code ELLIPSE3D) include the contact detection algorithm based on a geometric potential concept, elastic contact laws based on Hertzian theory and Mindlin's solution, geometric representation and kinetic computation for ellipsoidal particles, and the explicit finite difference scheme for 3-D rotational kinematic equations of ellipsoids. Numerical results produced by ELLIPSE3D show the important effects of particle shape on shear strength, deformation behaviour and fabric statistics of particle arrays. These results demonstrate that using nonspherical particles in a discrete element model improves the simulations of granular materials and creates the possibility to study the effect of particle shape on engineering properties systematically.
A series of drained tests were performed numerically on specimens with ellipsoids of two different shapes. The drained tests include axial extension, constant mean stress compression, lateral compression, lateral extension, plane strain, triaxial compression, triaxial extension and six constant-b tests. The six constant-b tests were conducted by varying the intermediate principal stress with b = (σ2 − σ3)/(σ1 − σ3) = 0, 0·2, 0·4, 0·6, 0·8 and 1. These tests were performed using the discrete element method (DEM) with pressure boundaries. Two specimens composed of 1170 ellipsoids with two different aspect ratios were created. The two specimens were consolidated isotropically before being loaded along different stress paths. The effects of intermediate principal stress, mean stress, particle shape and stress path were investigated. The results show that the stress–strain behaviour of constant-b tests was affected by mean stress. Similar behaviours were observed for the drained tests with b = 0 regardless of the stress path. The friction angles of these tests were then compared with four failure criteria that have been proposed in the literature. Overall, the Lade and Ogawa models are better than the models proposed by Matsuoka & Nakai and Satake. The Ogawa model has a slightly better performance than the Lade model.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.