An extended grain-based model accounting for microstructures in rock deformation

Abstract

Reliable prediction of the rock fracturing process is a challenging issue in exploitation of deep earth resources in which artificial creation of complex fracture networks is employed. The grain‐based modeling (GBM) approach is a promising numerical technique for its unique capability to simulate the fracturing behavior of crystalline rocks. An extended grain‐based model is developed to improve the traditional Voronoi GBM from two aspects. First, digital image processing technique is presented to incorporate actual rock microstructures into the numerical model. Second, the effect of initial microcracks is considered by integrating a statistical discrete fracture network model into GBM. By simulating semicircular bending tests on 16 extended GBMs and 3 Voronoi GBMs, the effects of rock microstructures and initial microcracks on microcracking behavior and mechanical properties are analyzed. Cracking patterns are classified into four types for the first time with respect to fracture toughness and crack initiation threshold. The results indicate that the use of a statistical structure or a purely deterministic GBM without consideration of initial microcracks cannot realistically describe grain‐scale discontinuities, which likely leads to biased evaluations of the rock failure process.