A Numerical Meso-Scale Elasto-Plastic Damage Model for Modeling the Deformation and Fracturing of Sandstone Under Cyclic Loading

Abstract

A sound understanding of the influence of cyclic loading on mechanical properties of rock is of great importance for the management and optimization of engineering rock mass structures such as nuclear waste repositories and geothermal resources. In the present manuscript, we first performed a series of conventional uniaxial compressive tests and fatigue failure tests (i.e. cyclic loading tests) on sandstone samples from Sichuan Province, China. The experimental results show that the fatigue failure of sandstone is achieved once the cumulative axial deformation under cyclic loading reaches the axial strain that corresponds to the post-peak phase of the stress–strain curve from a conventional uniaxial test. Based on our laboratory data, and considering that the fatigue process is characterized by local material degradation as well as irreversible deformation that accumulates during each loading cycle, we propose here a two-dimensional meso-scale elasto-plastic damage model to simulate the fatigue behavior of rock during cyclic loading. We then performed a suite of numerical simulations using this model to better understand the fatigue process in sandstone. The results of our numerical modeling show that the ratios of maximum stress and stress amplitude to the uniaxial compressive strength are the most important factors affecting the fatigue life of sandstone. A larger maximum stress and/or stress amplitude results in a shorter fatigue life for the numerical sandstone specimen. However, we find no significant correlation between sample homogeneity and fatigue life. Importantly, our modeled results show that specimen failure occurs once the cumulative axial deformation under cyclic loading reaches the axial strain that corresponds to the post-peak phase of the stress–strain curve from a conventional uniaxial test simulation, in accordance with our laboratory testing. Therefore, the model presented herein provides a new platform to investigate the fatigue failure of rocks.