Multi-scale Investigation into the mechanical properties and shear-slip failure processes of oil shale with varying bedding orientations

The efficient extraction of deep shale oil fundamentally relies on a precise understanding of the anisotropic shear-slip failure governed by natural bedding planes. However, conventional numerical simulations frequently overestimate the macroscopic failure strength because they neglect the authentic geometric heterogeneities of internal micro-defects. To address this limitation, this study establishes a multi-scale discrete element modeling (DEM) framework coupling Micro-CT scanning with macroscopic triaxial compression tests under a constant confining pressure. Within this framework, the spatial geometries of authentic microcracks are mapped to dynamically degrade the Smooth Joint Model (SJM) parameters via a reduction coefficient, successfully reproducing the synergistic weakening effect between native microscopic defects and structural planes. At the macroscopic scale, results demonstrate that both the compressive strength and elastic modulus exhibit an asymmetric U-shaped anisotropic trend alongside varying bedding dip angles. At the mesoscopic scale, single-factor sensitivity indices and the evolution of contact force anisotropy explicitly reveal the micro-mechanical drivers of this anisotropic behavior. Meanwhile, quantified by the displacement gradient tensor and micro-crack accumulation, the failure mode systematically shifts from matrix-dominated tensile splitting at extreme dip angles to highly localized, interface-driven shear slip at intermediate inclinations. Ultimately, this multi-scale approach provides a rigorous physical basis for anisotropic failure, offering reliable guidelines for wellbore stability and fracture optimization in deep shale formations under fixed stress environments.