In order to study the feasibility of upward mining in a certain coal mine, the mechanical parameters of the coal rock mass were obtained based on on-site investigation and rock mechanics experiments. A numerical model that conforms to the on-site mining layout was established using numerical simulation methods. The stress, displacement, and plastic zone distribution characteristics of the upper coal rock mass at different positions along the X, Y and Z directions after the initial mining of layered coal were analyzed. The critical thickness for tensile failure of the roof during upward mining was calculated using the thick plate theory. A ground penetrating radar was used on site to conduct three-dimensional detection of the fragmentation characteristics of the overlying roof rock. The results show that after the mining of 5# coal, the stress of the overlying roof surrounding rock in the goaf exhibits a dynamic change process of rising peak falling. As the vertical height from the working face increases, the stress in the center of the goaf and around the working face gradually decreases; the displacement distribution extends outwards from the center of the working face, forming a sinking basin with an upward opening in the direction of inclination and direction of the working face. The maximum displacement value is at the bottom of the basin; there is a tensile stress zone above the working face, and the plastic zone has the largest range at the center of the working face direction and inclination. The plastic zone has not evolved to coal 4# and coal 3#; the maximum critical thickness for tensile failure of the roof is 55.4 m, which is much smaller than the actual thickness of 99.82 m on site, and upward mining can be adopted.

To address severe mining-roadway deformation induced by roof stress concentration resulting from delayed fracture and caving of deep-buried thick and hard roofs, the ZC1103 working face of Yadian Coal Mine was selected as the engineering background. A fracture criterion based on the crack-tip stress intensity factor and rock fracture toughness was introduced to analyze fracture propagation and arrest along the shaped-charge direction. Bidirectional shaped-charge blasting models with different blasthole spacings were established using ANSYS and LS-DYNA, and multi-hole synergistic fracturing was evaluated from fracture coalescence, inter-hole pressure response, and overlap of damaged zones, followed by field validation. Results show that the bidirectional shaped-charge structure significantly alters the spatial distribution of blast-induced pressure around the blasthole. At 100 and 200 mm from the blasthole center, peak pressures in the shaped-charge direction increased by 48.9% and 35.1%, respectively, relative to conventional blasting, whereas those in the non-shaped-charge direction decreased by 28.0% and 34.9%, respectively. Field tests showed that, at a blasthole spacing of 1.0 m, distinct fracture traces were observed on the borehole wall 0.5 m from the adjacent blasthole, while the borehole walls remained largely intact. Compared with the non-pressure-relief zone, the cumulative deformation of the solid-coal rib, gob-side rib, and roof in the roof-cutting pressure-relief zone was lower by 41.7%, 51.8%, and 38.9%, respectively. These results provide a reference for designing bidirectional shaped-charge blasting spacing under similar deep-buried thick-and-hard-roof conditions based on fracture propagation and inter-hole coalescence criteria.

To address the challenge of intense pressure manifestation induced by remaining coal pillar from gently inclined multi-coal seam mining, a physical similarity simulation experiment was conducted using microseismic, thermal infrared, and borehole television monitoring. The overlying strata movement process, energy release characteristics, and fracture fractal evolution patterns were systematically investigated. The 3DEC numerical simulation was employed to evaluate the effectiveness of roof cutting measures. The results indicate that the overburden movement can be divided into three stages: formation of remaining coal pillar, suspension and fracturing of interlayer rocks, and subsidence and migration of overburden. The suspension of rock strata and the load transfer from uncollapsed strata are the key factors inducing dynamic disasters. Furthermore, the occurrence of disaster-causing energy events exhibited concentration and suddenness; only two IV high-energy events were recorded, accounting for 3.62% of the total energy, corresponding to the sudden fracturing of rock strata and the collapse of the entire overburden. A threshold of 250 J can serve as a criterion for identifying high-energy disaster events in similar models. Combined with the fractal dimension evolution process of fractures, it followed a pattern of growth, stabilization and renewed growth. After the upper-seam and W1121 working face were mined, the fractal dimension increased to 1.324. It remained within 1.372–1.377 during interburden suspension and reached 1.417 after interburden failure. These changes indicate that repeated multi-seam mining promotes upward fracture-zone propagation, potentially affecting the height of the water-conducting fracture zone. Finally, roof cutting was simulated in potentially hazardous coal-rock zones to evaluate pressure relief. Following pressure relief, suspension and hinging of fractured rock beams were largely eliminated. Overburden subsidence became more uniform, while lateral compression on one side of the panel coal pillar decreased markedly. Compaction within the goaf also became more uniform. These findings provide a theoretical basis for safe and efficient multi-seam mining and dynamic hazard control in mines with similar conditions.