Understanding rock damage evolution is essential for deep resource extraction and geological hazard prevention. Sandstone and mudstone are sedimentary rocks exhibiting multiscale damage behaviors critical to assessing engineering rock mass stability. We investigated their damage evolution under uniaxial loading using in situ real-time computed tomography (CT) scanning and mineral-based discrete element simulations. The porosity evolution coefficient ( K pec ), pore structure coefficient ( K pnsc ), and damage evolution factor ( D k ) were derived from CT data. Fractal theory was used to determine the fractal dimension of throats ( D t ). These parameters were utilized to evaluate mesoscopic damage statistically and spatially. Threshold segmentation was used to extract minerals, and a grain-based model considering CT results (CT-GBM) was constructed to reveal the macroscopic mechanical response and the evolution of microcracks and contact force chains. Results showed that sandstone, with abundant initial pores and unobstructed throat channels, had higher K pec and D t than mudstone, reflecting rapid compression due to the cementation of hard particles. The K pec and D t of mudstone, which has high viscosity and low porosity, increased and decreased during loading. The K pnsc value indicated that sandstone had a more complex pore network and higher stability, whereas mudstone had a simpler structure and lower mechanical stability. K pnsc was positively correlated with the degree of damage. Different mineral distributions influenced model performance. Due to the distinct mineral boundaries in sandstone, the CT-GBM simulated the crack evolution and anisotropic failure with high accuracy. Conversely, the simplified GBM model was more suitable for describing the brittle failure of mudstone due to its fine and uniform particles. The failure mechanism of sandstone was primarily intracrystalline cracks, with the contact force chains exhibiting dispersion, compaction, and shearing. In contrast, the mudstone displayed uniform brittle failure and rapid crack propagation. This study offers methodological insights for predicting rock mass safety.
Liu et al. (Sun,) studied this question.