Research on a Peridynamic Computational Method for Deep Underground Engineering under Realistic Stress Conditions

Geothermal energy is increasingly recognized as a clean, stable, and highly sustainable renewable energy source crucial for the global transition toward low-carbon economies. Reservoir rocks within Enhanced Geothermal Systems (EGS) are generally subjected to highly complex true triaxial stress conditions in deep underground environments. Consequently, their mechanical failure behavior strongly affects the efficiency of reservoir stimulation and the overall long-term engineering stability of the geothermal site. In this comprehensive study, an advanced true triaxial compression simulation method for predicting rock failure is developed based on the non-local peridynamic theory, which inherently excels at modeling discontinuous deformation. Within this novel computational framework, axial loading is systematically applied through designated virtual layers, while lateral principal stresses are accurately imposed by adjusting the body force density on the boundary layers. To verify the proposed approach, a detailed 50×50×100 mm granite numerical model is established and rigorously validated against corresponding laboratory experimental results. The comprehensive findings demonstrate that the proposed peridynamic method can effectively and accurately reproduce the entire progressive failure process, including crack initiation, propagation, coalescence, and ultimate instability failure of granite under true triaxial compression. Furthermore, the simulated complex crack paths and macroscopic failure patterns agree exceptionally well with empirical experimental observations. Ultimately, this study provides a highly useful and robust numerical approach for analyzing deep rock fracture mechanisms, optimizing EGS reservoir stimulation strategies, and predicting complex fracture network evolution in deep underground engineering projects.