Application of the four-dimensional lattice spring model in direct shear testing of intact rock

Abstract

Rock shear strength and its influencing parameters, such as cohesion and friction angle, are critical for both surface and underground project rock engineering designs. The application of direct shear testing is more commonly employed for rock joints than for intact rock due to several limitations, such as the nonuniform normal stress distribution. Hence, triaxial tests are more frequently used to determine the cohesion and friction angle of intact rock specimens. However, compared to the basic tools of the direct shear test, these tests require more complex experimental equipment, making the triaxial testing less accessible to laboratories in general. Therefore, it is important to develop a method for making intact rock direct shear test results more accurate for engineering applications. To evaluate and enhance the accuracy of intact direct shear tests, laboratory experiments on intact cubic marble specimens and consequent numerical simulation were performed using the four-dimensional lattice spring model (4D-LSM). The 4D-LSM is arguably more efficient than the conventional discrete element method since it uses direct calibration of microscale parameters from macroscale parameters while maintaining the ability to reproduce material fracture. In this study, a two-stage simulation was introduced in the 4D-LSM to represent the stages when the normal and shear loads are applied. To determine the failure state of the specimen, the Mohr– Coulomb failure criterion was implemented into the 4D-LSM. A bilinear stiffness model was used to reproduce the shear force-displacement curve from the experiments. Our results show that this method was able to reproduce numerical results that are similar to those obtained in experimental direct shear tests.