Abstract
During deep coal mining, coal is often subjected to vertical cyclic loading and lateral unloading. Understanding the failure characteristics of coal under these combined mechanical conditions is essential for safe extraction. This study utilized the MTS816 system to conduct cyclic loading tests on unloaded coal samples under varying confining pressures. From an energy perspective and based on the obtained stress-strain curves, the energy evolution and damage characteristics of the coal samples were analyzed throughout the entire process from intact state to failure. A CT scanning system was used to examine the fracture distribution and failure characteristics of the post-test coal samples. Finally, based on energy dissipation principles, an energy-damage constitutive model was developed and validated with experimental data. The results show that, under cyclic loading, increases in stress level and confining pressure lead to greater input energy, enhancing the coal samples' capacity to store and dissipate energy. The input energy is primarily stored as elastic energy, and this storage capacity is independent of confining pressure. At a given confining pressure, the damage variable increases with higher stress levels. With increasing stress levels under different confining pressures, the damage curves gradually diverge, showing significant differences in the damage variable, which reaches its maximum at the highest cyclic stress level. Under varying confining pressures, the two-dimensional/three-dimensional fractal dimensions and fracture volumes of the coal samples first increase and then decrease as confining pressure increases. The internal fractures in the tested coal samples consist of through cracks and micro-crack planes, with the degree of failure quantitatively characterized by fractal dimensions. The experimental loading stage curves show strong consistency with the energy-damage constitutive model curves, accurately describing the deformation behavior of unloaded coal samples under cyclic loading. These findings offer a valuable reference for controlling coal stability under cyclic loading conditions induced by mining activities.