Experimental study on sandstone permeability under cyclic loading and unloading of temperature and confining pressure

This study investigates the coupled effects of cyclic temperature and confining pressure variations on the permeability evolution of deep sandstone through a series of multi-physical field experiments conducted using the Rock Top multi-field coupling test system. Cyclic loading-unloading seepage tests were performed at real-time temperatures of 25, 50, 75, and 90 ℃ to systematically analyze the thermo-mechanical interactions governing permeability characteristics, alongside microscopic structural analysis to elucidate underlying mechanisms. Results show that under constant temperature conditions, sandstone permeability follows a consistent trend with confining pressure changes: permeability decreases as confining pressure increases and rises as confining pressure decreases. Moreover, within the same confining pressure loading-unloading cycle, permeability during loading is consistently higher than during unloading, with permeability and confining pressure exhibiting an exponential relationship. Under constant confining pressure, permeability initially decreases then increases with rising temperature, displaying a two-stage characteristic, and also follows an exponential function with temperature. Permeability loss diminishes with increasing confining pressure and reduces progressively with cycle number. Under varying confining pressures, permeability loss first decreases and then increases as temperature rises, while also decreasing over successive cycles. Additionally, regardless of temperature or confining pressure, permeability loss is mainly concentrated in the initial loading-unloading stage, accounting for approximately 60% of total loss. Under combined temperature and confining pressure variations, sandstone permeability loss steadily decreases with increasing loading-unloading cycles. The most significant reduction occurs during the first loading phase, driven by rapid pore compaction and microcrack closure, while subsequent cycles show diminishing permeability changes as the rock structure stabilizes. The permeability response reveals distinct sensitivities: during initial loading, confining pressure dominates permeability reduction through mechanical compression; in later cycles, temperature becomes predominant—with thermal expansion and microcrack propagation (notably above 75 ℃) mitigating compaction effects, leading to a transition from permeability reduction to potential recovery at 90 ℃ where interconnected crack networks form. Microstructural analysis via SEM and CT indicates that lower temperatures (50–75 ℃) mainly promote pore densification with limited new cracking, resulting in sustained permeability decline. Conversely, 90 ℃ induces extensive intergranular cracking and fracture coalescence, fundamentally altering flow pathways and increasing permeability despite cyclic loading. Further analysis under constant temperature confirms an exponential relationship between permeability and confining pressure. Due to hysteresis, permeability during loading consistently exceeds that during unloading at equal pressures. Under constant pressure, permeability exhibits a two-stage exponential response to temperature—initially decreasing (25–75 ℃) as thermal expansion closes microcracks, then increasing (≥90 ℃) as thermal cracking dominates. Quantitative data reveal permeability loss decreases with both increasing pressure and cycle number, with approximately 60% of total loss occurring in the first cycle. Pearson correlation coefficients verify the transition of controlling factors from confining pressure (early cycles) to temperature (later cycles). Microscopically, 50–75 ℃ reduces crack apertures to less than 1 μm via grain rearrangement, while 90 ℃ generates pervasive crack networks exceeding 10 μm in width, explaining permeability rebound. This study establishes quantitative relationships between thermo-mechanical conditions and permeability evolution, providing direct implications for stability evaluation in deep mining engineering involving temperature-pressure cycling—especially for ventilation design, water intrusion prevention, and support system optimization. The experimental methodology and identified temperature-pressure-permeability thresholds offer valuable references for similar sedimentary rocks in geoengineering applications. Future research should consider mineral composition effects and undertake larger-scale validation.