A DEM-DFN Method for Assessing Mining-Induced Continuous Fracturing and Fracture Connectivity in Rock Masses

Abstract Predicting continuous fracturing above mined-out areas is crucial for mining impact assessments and is typically achieved through numerical simulations. However, conventional modelling approaches often rely on indirect criteria to infer fracture propagation, leading to low-confidence approximations and random errors. This paper presents a novel method that incorporates the Discrete Fracture Network (DFN) technology into Discrete Element Method (DEM) modelling to enhance the representation and quantification of interconnected fracture networks. To implement it, an algorithm is developed with a core logic of traversing all physical contacts and using the DFN technology to generate fractures with position, length, orientation, and aperture attributes inherited from open contacts. A critical feature of the algorithm involves bridging gaps caused by the default corner rounding configuration in DEM models to ensure the continuity of linearly connected fractures. The algorithm is validated in a jointed rock mass model and then applied to a field-scale model based on a case study, demonstrating reliable computational functionality and high accuracy in connected fracture identification. In addition, the method enables a hierarchical representation of fracture connectivity relative to the goaf or the ground surface, depending on the reference object. Further application to sequential panel extraction scenarios reveals that fracture networks within individual panel footprints can interact through narrow pillars, leading to enhanced fracture propagation and potential connection between surface cracks and deeper fracture networks. Overall, this research provides an effective modelling method for assessing the impact of mining activities on rock masses and groundwater systems.