Coal burst is a sudden dynamic disaster that occurs under the superposition of dynamic and static loads, characterized by a violent failure process, concentrated energy release, and complex occurrence mechanism, posing a serious threat to mine safety. To reproduce the failure process of coal burst and reveal its dynamic essence in laboratory conditions, conventional physical simulation methods based on static similarity principles are inadequate to fully capture its dynamic characteristics. Therefore, with the dynamic disaster mechanism of coal burst as the research focus, a coal-rock dynamic similarity criterion system centered on the acceleration similarity ratio is proposed, aiming to establish a theoretical framework for physical simulation that can realistically reflect the dynamic response characteristics of coal burst. Through dimensional analysis, the key parameters involved in the coal burst process were systematically analyzed. Using the acceleration similarity ratio as the primary control parameter, a theoretical equation set of dynamic similarity criteria for coal burst was derived, and a system of dynamic similarity coefficients was constructed, including the stress similarity ratio, elastic modulus similarity ratio, time similarity ratio, and strain rate similarity ratio. By introducing the dynamic similarity coefficient, the coupling relationships and constraint conditions among various similarity ratios were analyzed, and a similarity criterion system describing the impact tendency of physical models was established. Furthermore, based on the derived theoretical relationships, the evolution laws of model strength curves under different geometric similarity ratios were summarized, and the applicable range and parameter intervals of the dynamic similarity criteria were clarified. The results show that the dynamic similarity coefficient is the key parameter governing dynamic similarity relationships, and its value directly determines the similarity of inertia response and energy transfer between the model and the prototype. The optimal geometric similarity ratio and wave velocity similarity ratio for the dynamic similarity criteria were determined. The proposed coal-rock dynamic similarity principle for coal burst breaks through the limitations of traditional static similarity theory, achieves quantitative characterization of inertia effects and energy release behavior, and provides an operable theoretical basis for dynamic physical simulation experiments of coal burst.

The dynamic damage mechanisms of granite with different grain sizes under multi-velocity and multi-angle penetration are crucial for optimizing the design of underground protective structures and evaluating the impact resistance of rock masses. Although previous studies have focused on the penetration mechanisms of rocks under individual variables, such as penetration velocity, biting angle, and grain size, they often overlook the coupled effects of these factors on crack propagation behavior and failure patterns during penetration. In this study, hypervelocity penetration tests were conducted on granite with a two-stage light gas gun (50/20 mm caliber) under four different impact pressures (14, 15, 16, and 17 MPa) and at four different biting angles (0 degrees, 15 degrees, 30 degrees, and 45 degrees). To quantify the resulting crater morphology parameters-such as the equivalent diameter, penetration depth, crater area, and crater volume-3D scanning and MATLAB point cloud processing techniques were combined with both quantitative and qualitative analyses of fragment splashing during penetration. The evolution of the crater parameters for two granite types (coarse-grained and fine-grained) at various penetration velocities and angles was analyzed in detail. A theoretical model for predicting penetration depth, which incorporates the coupled effects of grain size, penetration velocity, and biting angle, was proposed. The results indicate that fragment splashing velocities result in an axisymmetric distribution under normal penetration conditions. As the biting angle increases, high-speed fragments are shifted toward the impact-opposite side, whereas higher penetration velocities reduce the total number of splashed fragments. In coarse-grained granite, the crater parameters increase with increasing penetration velocity but decrease with increasing biting angle, primarily because of weak grain boundaries and multi-scale crack branching. The fine-grained granite exhibits a similar trend, with the crater parameters increasing with increasing velocity but the penetration depth decreasing with increasing biting angle. However, at high penetration velocities and intermediate biting angles (e.g., 30 degrees), the synergistic propagation of transgranular cracks results in a non-monotonic trend for the crater area and volume: they initially increase, but then decrease as the biting angle rises. At lower velocities, insufficient energy limits short-range crack propagation, resulting in monotonic decreases in the crater area and volume with increasing biting angle. The increasing rates of the crater parameters in coarse-grained granite decrease monotonically with increasing biting angle within a certain velocity range. In contrast, those in finegrained granite first increase but then decrease because of the synergistic effects of normal-tangential stresses at higher biting angles. Compared with fine-grained granite, coarse-grained granite has a higher strain rate sensitivity coefficient owing to its more energy-intensive crack propagation paths. The penetration depth prediction model, modified with energy conservation and grain size effects, effectively quantifies the three-factor coupling mechanisms, providing a theoretical tool for designing gradient protective structures.

To address the deviation between rigid confining pressure experiments and actual engineering conditions of deep backfill mining, where backfill near the working face has less confining pressure, while that in deep goaf areas is under high confining pressure, this study investigates the load-bearing characteristics of rock granular materials under flexible passive confining pressure. Customized PC molds with varying wall thicknesses and rigid steel molds were used to construct a gradient confining pressure environment. Compression tests were conducted, combined with the characterization of acoustic emission (AE) monitoring, strain measurement, particle sieving, and scanning electron microscopy (SEM) observation. The results show that flexible passive confining pressure divides the particle compression process into three stages that are different from those under traditional rigid constraints, namely the initial compaction stage, the crushing failure stage, and the lateral confinement-dominated stage. AE signals exhibit a bimodal energy distribution, and the time interval between the two can vary by more than 4 times with changes. The failure modes transition from shear to tension. Compared with intact materials, granular materials under lateral confinement maintain continuous volume contraction, and can even maintain a continuous volume contraction trend at least when the strain reaches 8%. And lateral confinement stiffness significantly enhances axial bearing capacity: when the axial strain reaches 30%, the axial stress in the rigid confinement group is nearly 5 times that in the flexible confinement group. Fractal dimension increases from 1.94 to 2.39 as the confinement stiffness rises. This study clarifies the influence mechanism of lateral confinement stiffness on granular mechanics, providing fundamental support for optimizing backfill design based on goaf locations and improving surrounding rock control in deep green mining. (c) 2026 China University of Mining & Technology. Publishing services by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

To address the limitations of traditional microseismic localization methods in layered heterogeneous media-including insufficient accuracy, dependence on initial values, and reliance on extensive labeled data-this paper proposes a simulation-data-driven Adaptive Residual Neural Network (ARNN) for microseismic event localization in coal mines. Based on actual geological data and Snell's law, a high-fidelity simulated dataset of P-wave first-arrival travel times was constructed using ray tracing, serving as a substitute for costly numerical simulation approaches. The lightweight ARNN model incorporates residual connections and the Adaptive Moment Estimation (Adam) optimizer to enhance training stability. Experimental results demonstrate that the proposed method achieves an average localization error of 11.8 m on 324 test samples and 7.9 m on 10 independent test events, significantly outperforming the Newton iteration method (minimum error of 73.01 m) and the grid search method (minimum error of 37.47 m). Moreover, even when a substantial amount of invalid samples is introduced into the training data, the localization error remains within 15.3-17.0 m, indicating strong robustness. This study provides a high-precision, highly adaptable intelligent solution for microseismic localization in complex geological settings with limited labeled data, offering significant engineering application value.

Bolt support represents a widely adopted technique for controlling strainbursts in underground excavations. Clarifying the mitigation mechanism of pre-tensioned bolts against strainburst is essential for rational support design. This study investigates the control mechanism of pre-tensioned bolts on strainburst through integrated theoretical, experimental, and numerical approaches. Results demonstrate that pre-tension force significantly increases both coal bursting threshold stress and fracture characteristic stress, with both parameters growing logarithmically with higher pre-tension. Pre-tension also significantly reduces ejected debris mass, surface damage ratio, seismicity density, and crack dimensions. Moment tensor inversion quantifies the regulatory role of pre-tension in suppressing tensile fracture propagation. Pre-tensioned bolt improves elastic strain energy density in the surrounding rock by expanding the spatial envelope and magnitude of the support stress field, creating an energy barrier effect. They lower elastic strain energy release rate (ESERR), thereby reducing coal bursting proneness and intensity. This energy regulation exhibits spatial heterogeneity: along the bolt axis from the free to anchorage section, ESERR remains consistently below 0.065 J m⁻3 s⁻1, reflecting minimal dynamic activity. Radially outward within the coal rib, ESERR increases sharply (peaking at 0.99 J m⁻3 s⁻1) due to shortened dynamic fracture duration as support stress attenuates, which locally intensifies strainburst severity. This study systematically clarifies the mitigation mechanism of pre-tensioned bolt against strainburst, contributing to the understanding of strainburst control in deep underground engineering.

Understanding the support effectiveness of pre-tensioned bolt under varying anchorage parameters is crucial for roadway support design. Previous studies focus on verifying and demonstrating bolt support effects, while lack a comprehensive examination of how anchorage parameters collectively govern both the support stress field and the fracture behavior of surrounding rock. Therefore, this study used experimental and numerical methods to investigate the stress/energy evolution and corresponding fracture transition behavior in bolt-reinforced coal ribs. Pre-tensioned bolts effectively mitigate coalburst through crack suppression, where the stress thresholds of crack initiation and crack damage are logarithmically increased with pre-tension force and decreased with anchorage length. Under end-anchoring conditions, increasing the pre-tension force from 250 N to 1250 N significantly suppresses coalburst intensity, manifesting as a 16.8% decrease in surface damage rate, a 4.0 g reduction in coalburst debris mass, an 8.5% increase in the acoustic emission b-value, a 15.2% reduction in tensile fracture proportion. Conversely, maintaining a constant pre-tension force of 250 N while increasing anchorage length from 18.0 mm to 60.0 mm aggravates coalburst intensity, characterized by a 9.9% increase in surface damage rate, a 6.8 g rise in coalburst debris mass, a 4.4% decrease in the b-value, a 7.3% increase in tensile fracture proportion. Discrete element model results elucidated distinct “two-compression-one-tension” distribution pattern of support stress field and anisotropic force chain network. Afterwards, quantitative correlations of energy storage capacity and kinetic energy conversion with anchorage parameters were determined as well. The control mechanism of anchorage parameters is conclusively revealed, in the manner of effective zone expansion of support stress field, enhancement of force chain anisotropy and contact fracture resistance, to increase minimum principal stress and bond strength.

Investigation of the damage effects of bursting layer materials represented by granite under hypervelocity oblique penetration is important for the design of underground protective engineering. Through hypervelocity oblique penetration tests, this study systematically reveals the damage mechanism and dynamic mechanical response of granite targets under different striking angles. Tests were conducted at striking angles of 0°, 15°, 30°, and 45° using a two-stage light gas gun, with velocities ranging from 1838 m/s to 2664 m/s, while high-speed photography and 3D scanning were employed for quantitative characterization of penetration damage. For striking angles of 0°, the differences in anti-penetration properties between medium-grained and fine-grained granite were compared. The experimental results indicate that striking angles greater than 30° significantly decrease damage parameters. A velocity threshold exists at 45°, where increasing velocity from 2343 m/s to 2436 m/s causes the penetration depth and the volume loss to plummet by 34.66% and 59.99%, respectively. Multivariate regression analysis revealed linear correlations between normal velocity (V0cosβ) and damage parameters. Furthermore, based on the Extended Dynamic Cavity Expansion Model (EDCEM), a penetration depth prediction model was modified to consider the striking angle. Among the four designed velocity gradients, the average penetration depth, crater area, volume loss and peak strain of fine-grained granite are all smaller than those of medium-grained granite. Their failure modes were observed microscopically, demonstrating superior anti-penetration properties of fine-grained granite. The experimental results address a gap relative to traditional vertical penetration studies and provide essential data for the performance assessment of granite subjected to oblique penetration.

Accurate seismic source location is crucial for assessing rock stability and mitigating dynamic hazards in underground mining. However, high-accuracy source location remains challenging in environments with sparse geophone networks, which are typical in mining operations. The framework strategically integrates three core components in a 'region-first, precision-later' workflow: (1) deep learning based on multi-scale attention mechanisms for coarse region location, (2) the regulated signature distance (RSD) method for physics-guided regional refinement, and (3) the sparrow search algorithm (SSA) for metaheuristic-optimized velocity model inversion and grid search. A boundary discrete network is utilized to optimize the initial velocity model structure. The method also integrates a double-difference location technique to reduce velocity model errors. Through comparative and ablation experiments involving both simulations and field tests in coal mines, the method demonstrates superior performance, achieving significantly lower epicenter and source errors compared to traditional approaches. The proposed method distinguishes itself by offering a comprehensive, integrated solution that addresses data scarcity, model inaccuracy, and computational efficiency simultaneously in sparse network scenarios, a challenge not fully tackled by existing hybrid approaches.

This study investigates dynamic fragmentation mechanism in defective brittle materials under impulsive loading through an analysis of fragment size distribution in rock Mott rings. A hybrid finite-discrete element method with embedded cohesive fracture modeling is employed to simulate fracture propagation, with rigorous validation of mesh dependency effects via convergence analysis. A novel energy-based dynamic fragmentation model is proposed, integrating pre-existing defects into an energy equilibrium framework that balances kinetic energy conversion and fracture dissipation across both inherent flaws and newly formed cracks. Comparative analysis with the classical Grady model establishes marked improvements in resolving multi-stage fracture progression, particularly through the model's capacity to capture defect-mediated strain-rate dependencies. Results reveal three distinct fragmentation phases, elastic expansion, defect-dominated primary fragmentation, and strain-rate-dominated secondary fragmentation, in defective systems as the strain rate increases. Cumulative fragment distributions in defect-embedded rings identify three distinct failure regimes: defect-driven fragmentation, defect-strain rate coupled fragmentation, and strain-rate-dominated fragmentation. Parametric analysis reveal that (1) increased defect fracture energy destabilizes phase boundaries between primary and secondary fragmentation stages, (2) higher defect strength accentuates three-stage fragmentation under constant loading rates, while (3) defect elastic modulus shows minimal impact on staging characteristics. Fragment size agrees well with predictions from the newly proposed theoretical model, demonstrating its effectiveness in capturing defect-mediated fragmentation dynamics.

Traditional uniform-diameter boreholes may cause either inadequate or excessive destressing effectiveness, potentially inducing coal burst or exacerbating roadway convergence and deformation. To address this hazard, a segmented variable-diameter destressing borehole design is proposed based on the zonal structure of surrounding rock and the stress-regulating mechanism of destressing boreholes. Quasi-static compression tests were subsequently conducted on rectangular roadway specimens containing different destressing boreholes. By integrating ejection velocity, three-dimensional bursting crater morphology, burst thresholds, and structure strength, the effectiveness of small-diameter, large-diameter, and segmented variable-diameter destressing boreholes was systematically evaluated. Through analysis of zonal fracture-induced acoustic emission responses, this study clarifies the mechanical-energy effects transfer from deep toward shallow surrounding rock during the catastrophic moment, thereby providing new insights into the triggering process and mitigating mechanism of coal strainburst. Research indicates that the instantaneous elastoplastic dilation of deep surrounding rock constitutes the core mechanism driving coal strainburst. Specifically, shallow tensile failure induces transient unloading, which triggers shear fracture in deeper zones. The resulting rapid fracture-released energy further accelerates shallow instability, causing a high energy release rate and ultimate burst occurrence. Energetically, the proposed borehole design attenuates the amplitude of fracture-released energy. Spatially, it drives shear-fracture sources deeper into the surrounding rock while preserving the bursting resistance and energy-attenuation capacity of the shallow zone as a protective barrier. Consequently, the design significantly reduces the surplus energy converted into debris kinetic energy, thereby effectively mitigating coal strainburst.

Rockburst real-time monitoring and early warning systems rely on stress perception using sensors installed in boreholes around roadways. Establishing a theoretical method for the monitoring threshold of these stress sensors, based on the quantitative interaction between the sensor and borehole, is the fundamental requirement for rockburst mitigation. In this study, the driving forces behind borehole convergence and rockburst occurrence are characterized, establishing a theoretical damage model for the interaction between stress sensors and boreholes within rockburst-prone roadways. Characteristic formulae and curves to describe the interaction between boreholes and stress sensors are derived and drawn, and the main control factors and their laws of interaction between boreholes and stress sensors are discussed. Based on these findings, design principles and monitoring threshold for rockburst control are proposed. The results indicate that, driven by the critical mining stress required for rockburst occurrence, the intersection between the borehole convergence curve and squeezing deformation curve of the stress sensor is key to quantitatively determine the monitoring threshold. Finally, we proposed a theoretical criterion and method to determine the warning threshold of stress sensors, which is directly related to rock strength, rock brittleness, mining stress, and borehole size. Through a case study from typical rockburst mines in China, the engineering applicability of the criterion is verified, with the determined error range between the theoretical warning threshold and actual warning threshold being 4.2%-6.5%.

This study investigates the coupled material–structural influence on coal bursts in deep coal roadways. Integrating theoretical analysis, numerical simulation, and physical model testing, the roles of coal bursting proneness, surrounding rock stiffness, and dynamic load disturbances in triggering coal bursts were systematically examined. An FDEM (finite-discrete element method) numerical model simulated the failure processes of coal-rock composites under varying roof stiffness conditions. The analysis revealed distinct energy conversion patterns governing different coal bursting proneness types subjected to roof pressure. Roadway failure simulations demonstrated lower coal fragment ejection velocities in non-burst-prone seams compared to highly burst-prone seams. Critical parameters for evaluating coal mass instability were determined based on energy conversion patterns within the roof–coal–floor system, confirming the influence of roof and floor stiffness changes on dynamic failure of coal. Combined static–dynamic loading tests, conducted via both numerical simulation and physical modeling, identified key instability evaluation parameters under dynamic conditions and revealed a distinct transition in roadway failure modes under dynamic disturbances. Finally, by analyzing the mechanisms of coal bursting proneness, surrounding rock stiffness, and external dynamic loading, a material–structural coupling mechanism for roadway failure was proposed. This mechanism provides a theoretical basis for coal burst prevention and control strategies in deep mining.

Water injection into coal seams and destressing boreholes are commonly used techniques for preventing coal burst disasters on roadways. However, the combined effect of these techniques on coal remain unclear, and there is a need to elucidate the strong bursting liability observed under dynamic loading. In this study, a Split Hopkinson Pressure Bar (SHPB) experimental system was employed to investigate the relationship between the mechanical properties of coal under impact loading with multi-parameter modifications, including water content, borehole diameter, and borehole layout. The strength characteristics, energy evolution, and geometric fractal behavior were systematically analyzed. The borehole layout had a minor influence, whereas the water injection and destressing boreholes had a synergistic weakening effect on the dynamic strength of coal. Energy analysis showed that the reflected energy decreased, whereas the transmitted and dissipated energies rose with increasing water content and borehole diameter (reaching 9.17% and 41.66%, respectively). The fractal dimensions of the fragments peaked at 2.68, indicating a shift in the fracture mode toward fine powderization and a substantial enhancement in energy dissipation efficiency. This study revealed the damage superposition mechanism of the combined modification, demonstrating its importance in synergistically regulating impact energy propagation by optimizing wave impedance matching and enhancing stress wave scattering. These findings provide an important experimental and theoretical basis for optimizing the parameters of combined destressing techniques and formulating multi-process prevention strategies against coal bursts in deep mines.

In a high-gas and rockburst combined coal mine, the effectiveness of long-borehole blasting in rock floors for rockburst elimination in ultra-thick coal seams is evaluated. The protective layer's failure in coal safety mining is analyzed, and a numerical model is developed to study the local accumulation of high-energy seismic events, considering the spatio-temporal relationship between the protective layer and the working faces in coal seam. A multi-round blasting scheme is designed using a gas-extraction roadway under the coal seam, with a blasting efficiency index (Se) proposed for assessment. The study compares blasting efficiency in the coal pillar area and solid coal area. Results show that the protective layer, compacted by long-term strong mining stress, loses its pressure relief function, leading to stress concentration and energy accumulation in the coal pillar, causing high-energy seismic events and damage to adjacent roadways. Floor blasting effectively prevents rockbursts by inducing high-energy seismic events that dissipate energy through coal and rock mass rupture. The initial local blasting reduces strain energy by 78 %, outperforming subsequent rounds; Se reaches 0.519-0.604 near thrust faults and stabilizes at 0.3-0.5 in solid coal areas. The local initial blasting outperforms subsequent rounds, the frequency of seismic events in the blasting treatment area increases, and the elastic energy accumulated in the coal and rock mass is released as low-energy seismic events. Se evaluates seismic induction, energy release, and pressure relief, reflecting the energy accumulation of coal and rock in different regions where Se is stable in areas with low mining disturbance, while areas with repeated high-energy seismic events show more complex energy re-accumulation processes. Multi-rounds for inducing seismic events and releasing strain energy should be adjusted based on monitoring results during mining.

Mining disturbances can induce fault unloading slip, posing major challenges in underground engineering. This study examines fault slip behavior under varying fault roughness and confining pressures using laboratory experiments and numerical simulations. Acoustic Emission (AE) and Digital Image Correlation (DIC) were employed to capture full-process slip characteristics, while Confocal Laser Scanning Microscopy (CLSM) revealed microstructural changes on fault surfaces. Discrete element modeling with the Smooth Joint Model (SJM) in PFC was used to explore microcrack evolution and localized stress-strain responses. Results show that fault unloading slip is governed by the evolution of normal and shear stresses on the fault plane and proceeds through four stages. In the linear steady stage, stress and deformation fields remain stable. When confining pressure reduces to 80 %-85 % of its initial value, the nonlinear steady stage emerges, with stable displacements. Further unloading leads to the meta-instability stage, characterized by 5 %-10 % stress fluctuations, localized displacements, and increased high-energy AE events. Finally, the instability stage exhibits strong stress oscillations and rapid macroscopic slip. Energy release cycles lengthen with increasing confining pressure and roughness. For a 180-grit fault, the cycle extends from 58 s at 10 MPa to 105 s at 30 MPa; at 30 MPa, roughness from 1000-grit to 60-grit increases the cycle from 68 s to 142 s. CLSM and simulations confirm that high-roughness surfaces concentrate stresses at asperities, enhancing damage, whereas smooth surfaces show uniform wear. With confining pressure >= 30 MPa and roughness <= 180-grit, deformation shifts from elastic to plastic and wear from abrasive to adhesive. Coupled AE b-value and entropy evolution-transitioning from high to low states-effectively indicate slip stages. These findings advance understanding of fault slip instability in deep coal mining and provide guidance for monitoring and early warning.

Understanding the influence of anchorage parameters on bolt support effectiveness is crucial for optimizing support designs. Although prior studies have recognized the compensatory effect of bolts, the underlying mechanisms by which anchorage parameters modulate the support stress field and fracture behavior poorly understood. This study integrates laboratory experiments with discrete element method (DEM) simulations to investigate the spatiotemporal stress/energy evolution and fracture transition in bolt-reinforced coal ribs under compression. Results show that pre-tensioned bolts effectively mitigate coalburst by suppressing fracturing. Specifically, the crack initiation and damage stress thresholds increase logarithmically with pre-tension force but decrease with anchorage length. Concurrently, with the increase in pre-tension, the surface damage rate, the mass of far- and near-field coal burst-spalling debris, and the crack dimensions all exhibit a progressive decrease. Conversely, extending the anchorage length yields the opposite trends. Furthermore, DEM simulations reveal a distinct “two-compression-one-tension” support stress field coupled with an anisotropic force chain network, highlighting the asymmetric nature of load transmission and localized stress partitioning mechanisms. Crucially, both the magnitude and spatiotemporal evolution of the loading-induced support stress field are fundamentally dictated by the initial stress state, thereby providing new insights into the support design. Driven by the stress compensation-reinforcement mechanisms, the energy storage capacity and bonding strength of the coal ribs is enhanced, while the kinetic energy conversion ratio is significantly minimized. This elevates the critical threshold for coalburst initiation and diminishes the potential energy source for debris ejection, thereby successfully mitigating coalburst hazards.

To investigate the dynamic response and damage mechanism of anchored structures under repeated impact loads,a combined approach of theoretical analysis,laboratory experiments,and numerical simulations is employed to con-duct the study of the dynamic mechanical behavior of end-anchored anchor bodies subjected to"multiple low-energy im-pacts followed by fewer high-energy impacts".Based on one-dimensional stress wave theory and the Kelvin elastic solu-tion,an analytical model for stress wave propagation and superposition in the rock bolt is established.Utilizing a self-de-veloped impact testing system for anchored structures,the dynamic mechanical behavior of the anchoring system under different energy-level cyclic impacts is quantitatively analyzed.The results show that a significant stress wave superposi-tion effect exists in the free segment of the rock bolt,with the peak stress exhibiting a segmented distribution characterist-ic.Within the anchored segment,the stress wave decays exponentially with propagation distance,Within the anchored segment,the stress wave decays exponentially with propagation distance,and the reflection coefficient at the anchored-free interface was measured to be 0.332,reflecting significant waveform reflection characteristics caused by impedance mismatch.Under repeated impacts,the anchoring interface exhibits a clear progressive failure characteristic,with rupture initiating at the anchored-free interface and gradually propagating deeper,forming a tripartite failure morphology of"com-posite failure zone-damage front zone-intact zone".The cumulative displacement of the anchored structure shows stage-wise nonlinear growth with the number of impacts,especially after exceeding the energy threshold,where the single max-imum displacement induced by a high-energy impact surges to 5 mm,this indicates that the interface has transitioned from micro-damage accumulation to a state of macroscopic plastic failure,Numerical simulation results further reveal the inter-action mechanism between damage accumulation and stress wave propagation,interface damage of the anchored structure initiates near the anchored-free interface and expands inward,causing the energy dissipation mechanism to shift from be-ing dominated by bond-slip(attenuation coefficient α=0.35)to friction dominance(α decreases to 0.12),Consequently,an abrupt change in stress wave attenuation characteristics is triggered.These findings reproduce the performance degrada-tion process of anchored structures under repeated impact loads and provide a theoretical basis for optimizing support designs and assessing dynamic stability in deep,high-stress roadways.

Similarity is a fundamental principle in physical analog modeling, guiding the selection of materials and experimental conditions. However, the similarity criteria for modeling the behavior of impact coal during coal burst remain underdeveloped. This study establishes a comprehensive set of similarity criteria by deriving critical parameters, such as impact stress, energy, time, and velocity, using mechanical models and similarity transformation methods. These criteria ensure consistency in the deformation and failure mechanisms of coal and rock under both static and dynamic conditions. The static similarity criterion is derived based on the Froude criterion, while the dynamic similarity criterion is investigated by varying the acceleration similarity ratio. Stress–time curves and acceleration ranges under dynamic conditions are obtained to validate the criteria. Numerical simulations further confirm the rationality of the proposed static and dynamic similarity criteria. The findings have significant practical implications. The proposed criteria enable the accurate design of physical models for simulating coal burst in laboratory conditions, facilitating a deeper understanding of coal burst mechanisms. This provides critical insights for predicting and mitigating coal burst risks in underground mining. Moreover, the methods and results can be extended to other geo-mechanical problems involving dynamic failure, such as coal burst prediction, tunnel stability assessment, and dynamic hazard control. By bridging the gap between experimental modeling and real-world applications, this study contributes to safer mining operations and improved hazard management strategies.

Mining-induced seismicity in strata has become one of the main dynamic phenomena in some coal mines.The theoretical identification of the main controlling rock stratum where mining-induced seismicity occurs is the basis for ac-curate prevention and control of mining-induced seismicity.In this paper,theoretical analysis,numerical simulation and on-site observation are used.The connotation and type of mining seismic events and mining-induced seismicity are sorted out,and the relationship between mining seismic events and mining-induced seismicity is clarified.The concept of the Key Strata of Mining-induced Seismicity(KSMIS)in overburden rocks is put forward,and the characteristics of the KSMIS are summarized,and the difference between the KSMIS and key strata is presented.The identification method of the KSMIS is proposed,the crack development and energy evolution during the fracture process of the KSMIS are analyzed,and the energy release mechanism for the fracture of the KSMIS is revealed.The results show that:① Mining seismic events can be classified into micro-seismic events,high-energy seismic events and mining-induced seismicity,and mining-induced seismicity can be classified into overburden rocks,faults,coal pillars,floor,folds and composite types;② The KSMIS in overburden rocks refers to a layer or group of layers in the roof that control the occurrence and distribution of the mining-induced seismicity,and is classified into two types of high-level and low-level according to the location of the KSMIS;③ Considering the failure criteria and energy characteristics of thick and hard rock layers,a method of identify-ing the KSMIS is proposed,and the method is verified by case study;④ In addition to the crack zones on both sides of the mined area,horizontal shear cracks appeared between the KSMIS and the roof in low position.The strain energy and shear dissipation energy of the interlayer joints were concentrated,and the shear slip between the rock layers occurred.There are strain energy and shear dissipation energy accumulation zones in some areas of the KSMIS,and layering damage exists within the KSMIS;⑤ When the actual maximum stress of the rock exceeds the strength limit of the rock layer or structur-al contact surface,the KSMIS will be broken or unstable,resulting in the formation of mining-induced seismicity in over-burden rocks.In the process,some of the elastic strain energy and gravitational potential energy is converted into mining-induced seismicity energy as well as various types of dissipation energy,etc.The results of the study can provide theoret-ical guidance for the prevention and control of mining-induced seismicity in overburden rocks.