ABSTRACT
In laboratory-scale numerical simulations of rock damage, the grain structure is commonly approximated through an assembly of bonded Voronoi blocks, and a set of properties that represents the micro-mechanical properties of the grains and grain contacts is calibrated to numerically replicate the macro-mechanical behavior of the rock. Such assemblies provide reasonable approximations of actual grain structures. However, the random nature of Voronoi assemblies increases the uncertainty of the contact micro-properties obtained from the calibration process, potentially leading to incorrect estimations of the rock strength. This study evaluates how different representations of the grain structure, in terms of grain geometry and grain arrangement influence predictions of brittle rock mechanical behavior. Different 2D Bonded Block Models of granite specimens were generated and used as a basis for Uniaxial Compressive Strength test simulations. The quality of agreement between the strength of actual granite specimens and the strength predicted using Bonded Block Models was tested through a comparative analysis. The modeling results show a notable influence of grain shape and grain arrangement. The results also prove that it is possible to predict the strength of rock with reasonable accuracy (within 13% of variability) using reasonably simplified representations of the grain structure in Voronoi models and previously published micro-properties.
It is well known that the grain structure of a rock controls the micro-mechanical behavior of the grains, and consequently, the macro-mechanical behavior of a rock (Gao et al., 2016; Wang and Cai, 2018). The mechanical behavior and strength of rocks are usually characterized using laboratory tests such as uniaxial compression, triaxial compression, and tensile tests. Laboratory studies show how individual rock specimens can exhibit different fracturing behavior as well as different strength even when they are the same rock type (Liu et al., 2018). The reason for such differences is the heterogeneous nature of rocks. At the grain-scale, rocks are composed of diverse minerals (i.e., mineral grains) in different shapes and sizes and are also affected by different defects (e.g., microcracks). That micro-structural heterogeneity governs a complex micro-mechanical behavior that generates localized stress concentrations within a rock specimen and ultimately results in fracture development (Gao et al., 2016; Wang and Cai, 2018).
