Abstract
Mesh size constitutes a critical parameter governing both computational accuracy and computational efficiency in fracture simulation of rock-like materials employing an advanced hybrid continuum-discrete element method integrated with a cohesive fracture model. However, determining an optimal mesh size remains nontrivial: conventional trial-and-error mesh refinement strategies inherently undermine computational efficiency and lack theoretical grounding. This study presents a systematic approach to mesh size identification via integrated theoretical analysis, experimental validation, and numerical parametric studies within the Cohesive Zone Model (CZM) framework via the combined finite-discrete element method (FDEM), which explicitly captures fracture propagation and fragmentation in simulation. Particular emphasis is placed on identifying a new general approach for optimal mesh size to enhance computational efficiency while preserving solution fidelity, thereby reducing overall computational costs constraints in engineering applications. First, mesh size estimates are derived from extensive fracture mechanics criteria and comprehensive statistical analysis of experimental KIC–σt data across multiple rock types. Second, the Modified Maximum Tangential Stress (MMTS) criterion that incorporates T-stress effects is introduced to correct the fidelity of fracture process zone (FPZ) length (lFPZ) estimation; predictions derived from the enhanced criterion are validated against independent experimental observations and analytical benchmarks. Third, quantitative evaluation of mesh size effects is conducted through FDEM simulations of uniaxial compressive strength (UCS) and Brazilian Disc (BD) tests. Integration of a meshable lFPZ from MMTS-based, back-correction inferred from KIC–σt relationships, and FDEM results reveals a robust convergence threshold: computational stability is attained when the FPZ is resolved by approximately 3–4 elements. Finally, heterogeneity effects on mesh size demonstrate that representing grain-scale features with single elements stabilizes rock fracture numerical stability, establishing the mesoscopic grain scale as the lower resolution bound for rock mesh sizing. The methodology and findings hold substantial implications for mesh design in rock fracture simulations and demonstrate the practical feasibility of mesoscale approaches in addressing complex geomechanical simulation challenges through hydro-based techniques.
扫码关注我们
求助内容:
应助结果提醒方式:
