Multiscale digital core fluid–solid damage coupling simulation for deep reservoirs based on free-flow and seepage coupling

Abstract The efficient exploitation of deep energy resources and secure CO 2 sequestration hinge on understanding fluid–solid damage coupling in reservoir rocks. However, existing pore-scale models often overlook the multiscale nature of pore systems and the evolution of rock damage. To address this, we develop a multiscale fluid–solid damage coupling simulation method based on digital core technology and continuum damage theory. By fusing high-resolution SEM images with large-field-of-view CT scans via machine learning, we construct a digital core that explicitly represents both micrometer-scale macropores (from CT) and four categories of sub-resolution matrix pores (from SEM). The Darcy-Brinkman-Stokes equation couples free flow and seepage, while the Span-Wagner equation captures the properties of supercritical CO 2 . Microstructural analysis shows total damage is most correlated with micropore fraction ( R 2 = 0.95), while apparent permeability strongly correlates with average micropore size ( R 2 = 0.89). The results reveal that confining pressure critically regulates the damage mechanism. When the confining pressure is below 60 MPa, total damage decreases and tensile failure dominates; when the confining pressure exceeds 60 MPa, total damage increases because shear failure rises significantly. Consequently, permeability loss can reach 63.7% under high confining stress. Micrometer-scale macropores dominate flow. Well-connected Category 1 matrix pores show flow velocities 298 times higher than the least connected matrix pore category, demonstrating the strong control of pore connectivity on microscale flow dynamics. Furthermore, increasing CO 2 injection pressure reduces effective stress, enhancing tensile damage and apparent permeability by up to 37%, but elevated fluid viscosity creates a competing resistance that can suppress flow velocities. This work provides a quantitative framework for predicting flow capacity and damage evolution in deep reservoirs under multi-field coupling conditions.