Three-dimensional non-persistent joints containing random rock bridges were fabricated by adopting 3D printing technology. Direct shear tests were conducted on specimens with joint persistence of five different percentages under three levels of normal stress. On the basis of test data, the laws of variation of shear strength parameters of three-dimensional non-persistent joints with joint persistence were systematically analyzed. The results show that the peak shear strength of joints decreases obviously with the increase of joint persistence. There is a significant difference in the variation of cohesion and internal friction angle with the increase of joint persistence. Cohesion decreases rapidly in a nonlinear manner, while the internal friction angle remains relatively stable. Furthermore, based on the laboratory direct shear test results, rheological elements were introduced to characterize the shear mechanical behavior of specimens, and a shear damage evolution constitutive model was established for non-persistent jointed rock mass. It is found that the calculated results of the model are in good agreement with the experimental data.

Aiming at the reasonable safe width of pillars in cut-and-fill stoping of steeply inclined thick and large phosphorite ore bodies, a mechanical analysis model of stope pillars was established based on elastic mechanics theory. The average bearing load of pillars due to the backfill was calculated, and a stope structural layout consisting of roof pillars and stope pillars was adopted. Cusp catastrophe theory was used to analyze the critical instability conditions of pillars, and the formula for critical failure width of pillars was then deduced, with which the critical pillar width is calculated to be 3.591 m. On this basis, numerical simulation was carried out to investigate the deformation and failure characteristics of surrounding rock in a stope with width of 10 m and pillar widths of 2 m, 3 m, 4 m and 5 m respectively. The comparison results of displacement and plastic zone indicate that roof subsidence gradually decreases with the increase of pillar width. When the pillar width is 5 m, the maximum roof subsidence is 30.95 mm and the plastic zone is only distributed in local roof areas. Engineering practice has verified that with pillars 5 m in width, the maximum displacement and maximum stress of the pillars are 18.47 mm and 4.32 MPa respectively, and both the stope and backfill remain in a stable state, which meets the requirements of engineering safety and economic efficiency.

To improve the efficiency of structural plane investigation and the accuracy of failure mode identification for open-pit slopes, the open-pit slope of Nannihu Molybdenum Mine in Henan Province was taken for research. The open-pit slopes were divided into 6 sub-areas, and a joint and fissure investigation method based on point cloud recognition was adopted to obtain the occurrence data of rock mass structural planes in each sub-area, and the recognition accuracy was verified by comparing with the manual scanline measurements. Cluster analysis was performed for the recognized structural plane data to divide the dominant structural plane sets in each sub-area. The spatial combination relationship between the dominant structural planes and the slope was analyzed by adopting stereographic projection to determine the potential failure modes of each sub-area. The results show that the occurrence data of the structural plane obtained by point cloud recognition are in good agreement with the manual measurements, which can meet the requirements of slope structural plane investigation and analysis. A total of 14 sets of dominant structural planes were identified in the six sub-areas: two sets of dominant structural planes were developed in each of zones Ⅰ, Ⅱ, Ⅴ and Ⅵ, while three sets of dominant structural planes were developed in each of zones Ⅲ and Ⅳ. Based on stereographic projection analysis, there are certain differences in the potential failure modes of slopes in different sub-areas, mainly in planar sliding and wedge failure modes. It is found that a single failure mode occurs in some part of sub-areas, and a combination of two failure modes occur in others.

 To address the challenges of slope stability and ground pressure control during the transition from open-pit to underground mining in Jinchuan Longshou Mine, the stress evolution and plastic zone variation laws of the open-pit slope and orebody surrounding rock during underground stoping and backfill process were analyzed with numerical simulation. The results show that: Underground mining leads to the formation of an elliptical subsidence area on the surface of the open pit, where the subsidence displacement is positively correlated with the mining depth, and the maximum subsidence displacement reaches 1.53 m; underground stoping induces a significant redistribution of the secondary stress field, with the compressive stress in the bottom area of the open pit decreasing from 2.0 MPa to 0.1 MPa, forming a low-stress relaxation zone penetrating to the surface; the plastic zones are mainly distributed at the bottom of the open pit and the toe of the south slope, while the backfill body effectively restricts the plastic zone from extending deep into the slope, and there is no penetrating plastic zone inside the slope body, indicating the overall stability of the slope. In the subsequent deep mining process, it is recommended that microseismic monitoring and multi-point displacement observation should be strengthened in areas with developed plastic zones. Meanwhile, the backfill roof-contact rate should be strictly controlled to prevent local surrounding rock caving from inducing slope instability.

For rapid and automatic recognition of blast fragmentation in the Husab Mine in Namibia of Africa, a dynamic recognition model was proposed for blast fragmentation in open-pit mines. By acquiring stereo images through a binocular camera and constructing a depth map, in combination with an improved YOLOv8 instance segmentation model, the classification of gravel and fragmentation analysis were realized. Moreover, transfer learning, ROI extraction, and shadow data augmentation technologies were all adopted to optimize the model, enhancing its recognition accuracy in complex environments. The model training results show that the optimized model can achieve a precision of 0.375, a recall of 0.600, and an accuracy of 0.300. This model exhibits strong robustness in extremely complex scenes of open-pit mines, and is capable of successfully detecting most effective rock blocks from high-noise and high-interference backgrounds. Field measurement data indicate that the model can operate stably in environments with high dust and dynamic lighting, achieving a recognition accuracy of 85% for effective rocks on site, and the error of the fragmentation distribution curve is controlled within 8%. This model provides reliable technical support for the efficient evaluation and intelligent detection of blasting effects in the Husab Mine.

Based on PFC^2D, a numerical model was constructed to systematically analyze the effects of confining pressure and fracture dip angle on the crack propagation, failure modes, and energy evolution of rock mass with cross-fractures. The results show that with the increase of confining pressure, the failure mode transitions from tensile or mixed failure mode under low confining pressure to shear-dominated failure under high confining pressure. As the dip angle increases, the secondary fracture - dominated fracture is shifted to primary fracture - dominated fracture. The confining pressure significantly affects the propagation morphology and quantity of cracks, whereas the dip angle mainly influences the crack initiation location and propagation path. The confining pressure enhances the energy storage capacity of rock samples, promotes strain energy accumulation, and strengthens the overall toughness of the structure. The dip angle mainly affects the energy allocation path. When the dip angle of cross-fractures is 45°, the specimen exhibits an extremely high energy storage capacity and delayed failure behavior, indicating a strong load-bearing capacity but a high rockburst risk. When the dip angle of cross-fractures is 60°, the lower energy storage limit of the specimen leads to sudden burst of energy, demonstrating typical brittle failure characteristics.