Internal voids and other defects in coal-rock readily trigger local or global failure, causing dynamic hazards such as roadway collapse and rock bursts. To quantify the effects of voids on mechanical behavior and failure, displacement-controlled uniaxial compression simulations were performed on coal specimens with five void counts and eight distribution patterns. Key findings are: (1) Compressive strength, Elastic modulus, Pre-peak strain energy decreases monotonically with increasing number of voids and is approximately linear. Void distribution also affects strength, with a maximum reduction of 14.7 %. (2) The impact energy index decreases with increasing void count. Void distribution has a pronounced effect: the index spans 6.40–11.09 across the eight distribution patterns, indicating strong impact propensity in all cases. (3) As the number of voids increases, the crack-initiation stress first decreases and then increases, while the cumulative impact count decreases monotonically. The vertical three-void layout yields the highest cumulative impact count. (4) As void count increases, cracking progressively localizes toward the specimen's lower region. (5) A damage constitutive model for voided coal was developed by coupling statistical distribution theory with the Drucker-Prager (D-P) criterion and subsequently validated. The model achieves greater than 90 % accuracy and reliably reproduces the stress-strain response of voided coal. (6) An energy-migration model for instability in voided coal was established, and energy accumulation, redistribution, and release under loading were analyzed to elucidate the associated instability mechanism. These findings offer theoretical guidance for roadway support design, stability assessment in mining engineering, and the prevention and control of coal-rock dynamic hazards.

A new methodology for coupling Smoothed Particle Hydrodynamics (SPH) with a Finite Volume (FV) solver for multiphase flow is presented. The approach follows a patched-domain strategy in which the two domains are disjointly separated by a discrete coupled boundary. Consequently, the coupled SPH–FV method is able to benefit from both methods’ strengths in their respective domains. The SPH solver employs a custom scheme, whereas a preexisting FV solver employing a Volume-of-Fluid (VoF) multiphase representation is adapted with minor modifications. Key features of the coupling include the determination of the volume fraction at the coupled boundary, the preservation of the fluid interface, and surface tension modeling in the vicinity of the coupled boundary. The method is validated against a series of benchmark cases, showing very good agreement. To our knowledge, this is the first coupled SPH–FV method that allows interface advection across the coupled boundary for multiphase flow with surface tension.