Building upon research into the effects of thermal disasters and their geological genesis, this study addresses the prevention of thermal disasters in tunnels located within geothermal anomaly zones. To overcome the limitations of conventional single-approach engineering solutions, a novel technical framework was developed, encompassing disaster source identification and control measures. This framework is guided by the methodology of “disaster-pregnant mode analysis, hazard-causing structure identification, process risk assessment, and control intervention.” First, the geological conditions, disaster dynamics, and precursor characteristics of tunnel thermal disasters were analyzed, leading to the identification of four disaster-pregnant modes: convective, conductive, diffuse, and diffuse-conductive composite. Second, recognizing that regional-scale geothermal features dictate the magnitude of hazards, regional-scale hazard structure identification techniques were established through regional geothermal field analysis and geological-geothermal coupling relationships, facilitating the optimal selection of low-temperature corridors. Third, acknowledging that tunnel-scale hazard structures determine disaster locations, collaborative geological-geothermal investigation techniques that integrate geochemical tracing, comprehensive geophysical exploration, and drilling were proposed for the survey-design phase. This approach enabled static risk assessments through the preliminary identification of thermal convergence features, including thermal recharge, high-temperature sections, and thermal phenomena. Fourth, to target potential high-temperature zones, construction-phase multiparameter tracking techniques were introduced to monitor borehole rock temperature, hydrochemical characteristics, and gas composition/concentration. These techniques supported dynamic risk assessments and enhanced the targeted identification of thermal convergence characteristics such as thermal intensity, scale, and severity in high-geothermal sections. Finally, multidimensional control technologies were formulated to address various thermal effects, encompassing ventilation, hygrothermal regulation, heat resistance measures, surrounding rock reinforcement, and concrete composite modification. This culminated in a tripartite synergistic prevention strategy that integrates heat source blocking, environmental control, and structural protection. Future research directions were also outlined. The established technical framework offers a novel system for thermal disaster prevention, providing guidance for safe tunnel construction.

The cases of tunnel damage observed during the 2022 Menyuan earthquake confirm that tunnels traversing active fault zones are simultaneously affected by both fault dislocation displacement and strong seismic forces. A correct understanding of the coupling effects of “strong ground motion and fault dislocation” on fault-crossing tunnels during earthquakes is crucial for ensuring the safe construction and operation of such tunnels in seismic zones. Consequently, a physical model test has been conducted for the first time to investigate the coupling effects of “strong ground motion and fault dislocation” on fault-crossing tunnels. A novel physical simulation system has been developed to simulate fault-crossing tunnels under coupled “strong shaking and fault dislocation” conditions. This system comprises a twin shaking table array, a non-uniform loading model box capable of applying combined “seismic motion and permanent displacement”, and a co-moving sliding support module. Additionally, an artificial ground motion synthesis method has been proposed to achieve synchronous and non-uniform inputs of “seismic motion and permanent displacement”, thereby establishing a hybrid input framework for the coupling effects of “strong ground motion and fault dislocation”. This framework facilitates the physical simulation of the mutual feedback response between the seismic wave field and the deformation displacement field on either side of the fault zone. Simultaneously, physical simulation experiments of the coupling effects of earthquakes and dislocation were conducted using the Daliang Tunnel as the background project. The effectiveness of the experimental device and method was analyzed based on the response characteristics of the shaking tables, the model test box, and the surrounding rock and lining structure of the fault-crossing tunnel. Furthermore, the damage phenomena observed in the Daliang Tunnel during the Menyuan earthquake under the coupling effects of “earthquake and dislocation” were successfully reproduced. This study provides a physical simulation methodology for subsequent in-depth research into the mechanisms of the “strong ground motion and fault dislocation” coupling effects in cross-fault tunnels and is expected to advance seismic design practices for fault-crossing tunnels.

 Accurate prediction of tunnel rock mass strength is essential for ensuring construction safety. This study introduces a Bayesian network (BN)-based method that leverages multidimensional data. First, an enhanced Swin-Transformer, integrated with UNets capability for small-scale feature extraction, achieves high-precision segmentation of apparent rock mass features (e.g., leakage and fractures), attaining an average accuracy of 89.3%, which surpasses conventional models. Subsequently, a multidimensional dataset is constructed by integrating apparent features, internal parameters, physical-mechanical properties, and design parameters. Test results indicate that the constructed Bayesian network achieves an accuracy of 89.6%, outperforming traditional machine learning, deep learning, and experience-driven Bayesian network methods. It maintains an accuracy of 82.3% on untrained samples, demonstrating exceptional generalization capabilities. Sensitivity analysis reveals that fracture and weathering parameters are the dominant factors influencing rock strength. The proposed framework facilitates interactive operation and allows for partial parameter input, thereby providing a practical tool for intelligent tunnel engineering.

To elucidate the seepage-deformation failure mechanism of subgrade slopes filled with completely decomposed granite (CDG), hypergravity model tests and finite element analyses were conducted to examine the coupled effects of groundwater and rainfall. In both hypergravity tests, seepage failure occurred at the trailing edge due to a rapid rise in the water level. The critical water level at the trailing edge was approximately 12 m, and the critical hydraulic gradient within the slope model was around 0.47 for both tests. The finite element model, validated by the hypergravity tests, indicates that the onset of osmosis failure is primarily attributed to a surge in the local hydraulic gradient resulting from the swift increase in water level at the trailing edge; subsequently, seepage leads to backward piping failure within the slope. Furthermore, as fine particles accumulate and permeability decreases at the slope toe, the stability of the slope diminishes with the rise in groundwater level. Conversely, when fine particles at the slope toe are eroded and permeability increases, the groundwater level decreases, resulting in enhanced slope stability. The hydraulic gradient gradually diminishes from the slope surface to the trailing edge, aligning with the development process of backward piping failure, until the piping channel is fully formed and the hydraulic gradients on both the front and back of the slope become consistent. These findings provide valuable insights for mitigating water-induced failures in similar subgrade slopes.

The seismic landslide hazard chain encompasses various physical processes, including landslide initiation, slope movement, river damming, and dam-break flooding. It is characterized by multifactorial influences, the coupling of multiple hazards, and an interconnected sequence of disaster events, highlighting the need for a comprehensive risk assessment model. This paper develops a risk assessment model for the seismic landslide hazard chain based on Bayesian theory. Initially, the primary nodes of the Bayesian network were identified through bibliometric mining combined with multi-source data. Subsequently, by integrating a database of typical seismic landslide hazard chain cases, an initial Bayesian network model was constructed using the Expectation-Maximization algorithm. A more refined coupled model with enhanced nodes was then developed by incorporating the cascade mechanisms among various hazards. Finally, using the Brier score criterion and taking the Mogangling landslide hazard chain as a case study, the predictive capabilities of both models were evaluated by calculating the mean squared error between predicted probabilities and actual hazard occurrences. The results indicate that: (1) Predictions of landslide volume, dammed lake capacity, and barrier dam stability from both the initial and coupled models are consistent with field observations. The selection of network nodes, determination of structure, and the use of posterior probabilities derived from parameter learning as prior probabilities for the predictive model effectively replace subjective expert experience, thereby enhancing the reliability of holistic risk assessments for seismic hazard chains. (2) The Brier score of the coupled model reached 0.146 7, outperforming the original model?s score of 0.242 7, demonstrating that the multi-hazard coupling method improves the predictive performance of the Bayesian network model. (3) The multi-hazard coupled model significantly reduces the Brier score by enhancing the efficiency of available data utilization, thereby addressing the issue of data scarcity in parameter learning across the entire seismic landslide hazard chain process. This study offers novel insights and decision support for hazard chain prediction and disaster mitigation through chain-breaking strategies.

When mountain railway bridges traverse debris-flow gullies characterized by “tributary stronger than main river” dynamics, they face not only direct impact risks during normal flow but also the potential for burial due to backwater sedimentation caused by main river blockage. Existing design standards for bridge heights rarely account for this latter factor. To quantitatively assess the real-time sediment depth at the bridge girder and its implications under blockage-induced backwater conditions, this study employs a case study of a railway bridge crossing debris-flow gully Z on a major trunk line in western China. Given certain height restrictions, a depth-integrated continuum mechanics model was utilized. Using Massflow, the entire dynamic process—including debris-flow outflow, river blockage, and backwater sedimentation—was simulated both before and after the construction of a check dam within the gully. From this, the critical storage capacity and optimal dam height were determined. The results indicate that: (1) Prior to dam construction, backwater sedimentation following blockage reached 10.93 m, exceeding the bridge clearance limit of 6.65 m. (2) Using the clearance limit as the post-regulation control target for sediment depth, the critical storage capacity of the check dam was calculated to be 6.08×104 m³ based on the necessary reduction in total outflow. The optimal dam height was subsequently established at 6 m according to the capacity-height relationship. (3) After regulation by the check dam, the maximum backwater sediment depth at the bridge site was reduced to 6.16 m, satisfying the clearance requirement. (4) To address the dual objectives of peak flow reduction and outflow volume regulation, a configuration of two beam-type slit dams was identified as the optimal spatial arrangement. Numerical simulations confirmed that their regulatory effectiveness corresponded with the design objectives. This research offers theoretical methods and design references for enhancing the disaster resilience of linear transportation infrastructure traversing debris-flow hazards in high-mountain gorge regions.

Deep underground excavation inevitably induces varying degrees of initial damage to the surrounding rock, significantly altering its load-bearing capacity and deformation behavior. However, conventional strength criteria are insufficient for quantitatively characterizing these damage effects. This study aims to elucidate the deterioration mechanisms of sandstone under initial damage and to develop an improved strength criterion suitable for engineering predictions. Sandstone specimens with varying levels of damage were prepared through uniaxial incremental cyclic loading-unloading tests, from which their peak strength, deformation modulus, and shear strength parameters were obtained via triaxial compression tests. Additionally, PFC2D was utilized to simulate crack initiation, propagation, and coalescence, providing microscopic insights into the damage-induced weakening of strength. Based on these findings, the staged influence of damage on the Hoek-Brown parameters m and s was quantified, leading to the establishment of a “damage-parameter” relationship incorporated into the generalized Hoek-Brown criterion, thereby enhancing the strength prediction method. The results indicate that initial damage causes exponential decay in peak strength and stiffness, accompanied by an increase in Poisson′s ratio, with the damage effect being strongly dependent on confining pressure. The reconstruction and weakening of the crack network were identified as fundamental causes of macroscopic deterioration, while high confining pressure effectively suppresses tensile cracking and promotes frictional slip, thereby delaying failure. The improved Hoek-Brown criterion accurately captures strength reduction at different damage levels. Engineering case verification further demonstrates that the proposed method yields higher prediction accuracy for surrounding rock deformation compared to conventional criteria. This study provides a quantitative approach for evaluating strength degradation in excavation-induced damaged rock and offers valuable guidance for stability assessment and support optimization in deep underground engineering.

The permeability differences between soft and hard coal seams resulting from tectonic evolution are significant. Investigating the dynamic response of coal and roof rock layers with varying strengths during blasting is crucial for optimizing blasting parameters and controlling the surrounding rock in underground roadways. To address the discrepancies in permeability enhancement due to blasting in coals of different strengths, this study employed a combined approach involving model testing, numerical simulation, and field application to examine the attenuation of blasting energy and the damage-fracture characteristics of coal and rock. To quantitatively characterize the damage evolution process within coal, the fractal dimension was introduced to quantify the complexity of the crack network induced by blasting. Based on Weibull theory and the damage probability density function, a quantitative relationship between damage variables and fractal dimension applicable to both hard and soft coal was established, leading to the construction of a blasting damage prediction model based on fractal dimension. The results indicate the following: (1) Blasting in hard coal results in extensive crack propagation, leading to effective permeability enhancement but significant roof damage. In contrast, blasting in soft coal creates a larger crushing zone, resulting in limited permeability enhancement and minimal roof disturbance. Analysis of strain and vibration acceleration data reveals that as the explosive stress wave travels from soft coal across the coal-rock interface into the rock layer, it experiences greater attenuation compared to its travel from hard coal. (2) Damage evolution analysis shows that damage in hard coal appears as striated patterns, while in soft coal, it manifests as regional damage centered around the borehole. Calculations using ImageJ software demonstrate that the fractal dimension of hard coal is greater than that of soft coal at the same time point. The developed damage prediction models for hard and soft coal exhibited high goodness-of-fit, with R² values of 0.96 and 0.94, respectively, confirming the reliability of using fractal dimension for quantitative damage assessment. (3) Based on experimental and numerical simulation results, the charging parameters for blasting permeability enhancement in hard coal seams at Zhaogu No. 2 Mine were designed. Field results indicated that three hours after blasting, the gas drainage concentration increased from 15.33% to 36.74%, and the pure drainage volume rose from 0.085 m³/min to 0.587 m³/min, with no significant damage observed in the roadway roof. These research findings provide a reference and theoretical basis for charge design in blasting permeability enhancement engineering for coals of varying strengths.

To investigate the influence of support paths on the stability characteristics of roadway surrounding rock reinforcement and to address the significant deformation challenges associated with traditional support methods that neglect the reinforcement path effect, we utilized our independently developed ZST–1500 microcomputer-controlled electro-hydraulic servo coal-rock dynamic-static combined adaptive coupling testing system to conduct experiments on the mechanical properties of coal under graded pressurization. We systematically analyzed the mechanical response mechanisms, failure characteristics, and energy evolution patterns of coal and rock subjected to graded pressurization paths. The research findings reveal that: (1) As the coal-rock mass transitions into the plastic zone and confining pressure increases, its load-bearing capacity significantly improves, with a lower initial confining pressure associated with a greater increase in compressive strength. (2) The implementation of stepwise pressure increments enhances the plastic deformation capacity of the coal-rock mass, with the ratio of plastic deformation initially rising before declining as the initial confining pressure increases. (3) Stepwise pressure increments lead to multi-stage damage accumulation within the coal-rock mass, evidenced by a decrease in the inclination angle of the primary crack and an increase in the number of secondary cracks. (4) Compared to conventional triaxial tests, the peak acoustic emission (AE) ringing count decreases under the stepwise pressure increment path, while the range of high ringing counts expands. (5) Increased confining pressure significantly boosts both the energy absorption and ultimate storage capacity of the rock. However, stepwise pressure increments result in a higher proportion of pre-peak dissipated energy density, thereby intensifying pre-peak strain hardening and ultimately enhancing the overall load-bearing capacity of the coal-rock mass. These findings provide a reliable theoretical foundation for the design of surrounding rock support and the optimization of reinforcement strategies in deep roadways.

To investigate how the length of liquid oxygen storage influences rock failure behavior and the evolution of the fracture process zone (FPZ) during phase-change expansion fracturing, four types of oxygen- storage charges were applied to cubic bluestone sandstone specimens. Ultra-high-speed imaging combined with Digital Image Correlation (DIC) technology was employed to capture full-field displacement and strain evolution throughout the loading process. Based on this data, crack propagation patterns, failure modes, strain-time responses, fractal dimensions, and energy dissipation characteristics were analyzed. The results indicate that under medium-to-low storage lengths (5–7 cm), crack growth is primarily governed by axial through-fractures, transitioning from a “multi-crack linkage” mode to a “single continuous penetration” mode. The strain-time response exhibits a characteristic three-stage behavior, comprising rapid increase, peak stabilization, and nonlinear attenuation. The crack initiation time decreases linearly with increasing storage length, while the fractal dimension rises from 1.700 to 1.749, accompanied by a shift in fragmentation from large blocks to more uniformly sized medium fragments. When the storage length exceeds approximately 7 cm, excess phase-change energy triggers competitive crack propagation, resulting in the formation of a composite network of axial and eccentric cracks. This leads to a sharp increase in the fractal dimension to 1.801 and a notable non-uniformity in fragmentation. Further analysis reveals that storage length plays a critical role in controlling displacement field patterns and FPZ evolution. Under low storage conditions, fragments rotate outward around the bottom contact point; as storage increases, symmetric horizontal ejection becomes dominant. Beyond a critical range, the competitive propagation of eccentric cracks inhibits the development of axial cracks, resulting in limited displacement of central fragments. Consequently, FPZ length shifts from a decreasing trend to an increasing one, indicating a fundamental change in the energy dissipation mechanism.

The deformation and fracture of hard roofs are among the primary causes of impact disasters at working faces, posing significant threats to the safety and efficiency of coal mining operations. This study analyzes the geological conditions of the working face beneath the hard roof of CG1302 in a mine located in Shandong Province, elucidating the processes of fracture and the destabilizing impacts of hard roofs in deep well caving mining. It is established that controlling impact risks associated with hard roofs hinges on minimizing bending deformation, reducing energy accumulation and stress concentration, and preventing the breaking and caving of the hard roof. The mechanisms of energy absorption and impact reduction in deep well filling mining are clarified: the filling material provides support to the hard roof through effective contact, thereby limiting deformation of both the hard roof and the coal-rock mass at the working face. Simultaneously, the filling material absorbs and dissipates a portion of the energy through compressive deformation and plastic damage. A mechanical model of the hard roof under deep well caving and filling mining is constructed, elucidating the bending deformation state, energy accumulation in the hard roof, and the distribution characteristics of compressive strain energy in the coal body. It is observed that the bending moment and energy of the hard roof, along with the energy of the coal body in front of the working face during filling mining, are reduced by 77.61% to 96.98% compared to unfilled mining. The areas at the end of the coal support zone (i.e., 0 to 5 m from the junction) and the middle of the goaf are identified as high-risk zones for bending deformation and energy accumulation in the hard roof. Factors influencing energy absorption and impact reduction in filling mining, such as overburden load, the elastic foundation coefficient of the filling material, the elastic modulus of the hard roof, and the elastic foundation coefficient of the coal seam, are discussed. Notably, the overburden load and the elastic foundation coefficient of the filling material significantly affect the bending deformation and energy accumulation of the hard roof, while the elastic modulus of the hard roof and the elastic foundation coefficient of the coal seam have a lesser effect. The effects of energy absorption and shock reduction at various filling rates were quantified using FLAC3D numerical simulations. As the filling rate increased from 0% to 90%, the maximum subsidence and peak strain energy density of the main roof, as well as the peak strain energy density of the coal body in front of the working face, were significantly reduced, showing decreases of 88.49%, 64.52% and 64.84%, respectively. Furthermore, the peak strain energy density and the range of high strain energy at the working face decreased markedly with increasing filling rates. Filling mining mitigates the bending deformation of hard roofs, weakens energy accumulation, and effectively reduces impact hazards. Combined with field practices and research findings, this study demonstrates that filling mining can significantly diminish the intensity of mine pressure at the working face, facilitating effective prevention and control of rock bursts, thereby providing a scientific basis for the safe and efficient green mining of deep coal resources.

Regarding the extraction of unconventional shale oil, in-depth investigations of two-phase flow patterns are crucial for describing residual oil content, displacement efficiency, and ultimate recovery. Using a pore-scale rock as a case study, a two-phase flow model has been developed based on phase-field theory and the Navier-Stokes equations to simulate the water-oil two-phase flow process, influenced by pore-throat size and structural wettability under specified flow velocities. The results indicate that variations in pore-throat size inversely correlate with the advancement of the leading-edge positions of the water-oil two-phase interfaces on both sides. It is observed that dominant flow channels form within the pores, with internal flow velocity being inversely proportional to pore size. Moreover, the pressure difference distribution in small-pore-throat structures is considerably greater, while the average pressure change at the inlet gradually diminishes and stabilizes. The stabilization times for both average inlet pressure and leading-edge position are consistent. As the pore-throat size decreases, displacement efficiency within the pores progressively increases, although it deviates from stable displacement. Additionally, variations in pore structure wettability result in the advancement speed of the leading-edge positions being directly proportional to contact angle size, and the central two-phase interface exhibits irregular capillary fingering characteristics; however, final displacement saturation is inversely proportional to contact angle size. Furthermore, the initial pore pressure distribution is uniform, and the pressure gradient at the two-phase interface increases as wettability weakens, particularly with the average pressure change at the inlet gradually diminishing and stabilizing. The advancement speed of the displacement front increases with larger contact angle sizes, and displacement within the pores is significantly influenced by capillary effects, especially for a wetting angle of 30°, leading to an extensive displacement range of the two-phase interface that affects fingering on both sides. These investigations elucidate the water-oil flow mechanism during unconventional shale oil reservoir extraction, providing a scientific basis for optimizing efficient development strategies.

Quasi-negative Poisson′s ratio (Quasi-NPR) bolts, a novel reinforcement component, have been increasingly utilized in slope and tunnel support engineering. To further investigate their anchoring characteristics in large-scale rock masses with muddy interlayers, direct shear tests were conducted on mud-filled joints anchored with Quasi-NPR bolts. The analysis evaluated the anchorage shear performance under varying filling degrees (0–2) and joint lengths (100–400 mm), with comparisons made to conventional positive Poisson′s ratio (PR) bolts (HRB400). The results indicate the following: (1) At the same filling degrees, the post-peak stress drop in HRB400-bolted filled joints becomes more pronounced with increasing joint length. In contrast, Quasi-NPR-bolted filled joints exhibit relatively stable performance, accompanied by a gradually diminishing decay in residual shear strength. This suggests that Quasi-NPR bolts offer anchoring advantages during the residual stage, primarily due to their ability to mitigate the size effect on residual shear strength. (2) Although Quasi-NPR bolts develop smaller plastic hinges in small-scale specimens, these hinges gradually expand with increasing joint length, ultimately exceeding those observed in HRB400 bolts. This highlights the superior deformation adaptability of Quasi-NPR bolts in large-scale jointed rock masses. (3) The anchorage resistance ratio of Quasi-NPR bolts decreases with increasing filling degree and joint length. When the filling degree exceeds 1.6 and the joint length surpasses 300 mm, the ratio tends to stabilize. These findings provide valuable guidance for determining key design parameters of Quasi-NPR bolts in the reinforcement of muddy interlayer rock masses.

The mechanical properties of rocks are critically important for the design and analysis of rock mass engineering. However, under specific conditions—such as core disking due to high in-situ stress, highly jointed rock masses, and extraterrestrial rock bodies—obtaining standard cylindrical rock specimens that meet conventional testing requirements can be challenging. To address this issue, this study develops an integrated prediction framework for estimating the macroscopic mechanical parameters of rocks based on nanoindentation tests of the constituent minerals. First, nanoindentation is employed to obtain meso-scale mechanical parameters of rock-forming minerals, including hardness, elastic modulus, and mode-I fracture toughness. The generalized means method is then applied to upscale the mineral elastic moduli to the macroscopic elastic modulus of the rock, demonstrating significantly higher prediction accuracy than traditional homogenization approaches. Second, leveraging the physical consistency between indentation hardness and uniaxial compressive strength, a multiple linear regression model incorporating the influence of minerals with varying strength grades is established to predict the uniaxial compressive strength of rocks, revealing the dominant role of low-strength minerals in controlling macroscopic strength. Furthermore, by integrating fracture mechanics theory with an equivalent weighting model, a predictive model for the tensile strength of rocks is formulated based on the mode-I fracture toughness of minerals. This framework provides an effective approach for the rapid estimation of rock mechanical parameters under sampling-constrained conditions and, through the introduction of the Hoek-Brown criterion, establishes a methodological foundation for achieving cross-scale predictions of mechanical parameters from “rock cuttings nanoindentation-rock core-rock mass”.

To investigate the fracture evolution of strain bursts in granite at varying burial depths, we employed a self-developed true triaxial rockburst experimental system that integrates high frame rate infrared thermal imaging and acoustic emission monitoring technologies. Rockburst experiments were conducted at five initial in-situ stress levels corresponding to burial depths ranging from 400 to 1 600 m. By analyzing the infrared radiation temperature field—including average temperature, spatio-temporal evolution of high-temperature points, and the differentiation rate of the temperature field—alongside acoustic emission characteristic parameters such as RA-AF values and energy concentration r values, we elucidated the macroscopic failure process of rockbursts, the infrared thermal image response to crack propagation, and the precursor behavior of acoustic emissions. The findings indicate that: (1) the macroscopic failure of rockbursts occurs in four stages: a calm period, external bulging of the rock plate, fracture, and debris ejection. Crack evolution exhibits a sequential morphology comprising shear, tension, and a combination of both. The intensity of rock explosions increases with greater burial depth, resulting in larger explosion pits. (2) Infrared thermal imaging effectively captures the entire process of crack initiation, propagation, and penetration. At greater burial depths, strip-shaped high-temperature zones dominated by tensile cracks serve as precursors to severe rockbursts. The average infrared temperature demonstrates a characteristic pattern of “stable-oscillating increase-sharp rise” prior to rockbursts, with high-temperature points aggregating and evolving along the rupture path. (3) Acoustic emission RA-AF analysis revealed that tensile cracks predominated the failure mode (70%–84.51%); the r value of energy concentration transitions from stable fluctuations to an accelerated decline before a rockburst, serving as a potential early warning indicator. (4) By assessing the temperature field differentiation rate, we can effectively identify shear (heating) and tension (cooling) fractures, with the appearance of their peaks and troughs occurring significantly earlier than the emergence of macroscopic cracks. This provides a novel observational basis for understanding the microscopic fracture mechanisms and predicting rockbursts. The research outcomes hold substantial theoretical value for identifying and mitigating rockburst risks in deep engineering contexts.

Understanding the mechanisms of rock fragmentation caused by blasting gas during explosive loading is crucial for optimizing blasting design and achieving precision blasting. In this study, we develop a theoretical model for rock fracturing that accounts for the combined effects of explosive stress waves and blasting gas. We propose an analytical method for determining gas volume and pressure based on crack opening displacement and the gas equation of state. This model is integrated into a finite-discrete element solver to analyze the rock-breaking process driven collaboratively by stress waves and blasting gas. The results indicate that stress waves and blasting gas play fundamentally different roles in rock fracturing. Enhanced stress-wave loading increases the crushed zone but has a limited impact on the fracture zone. In contrast, an increase in blasting gas pressure leads to longer primary cracks and significant expansion of the fracture zone, while the crushed zone remains nearly unchanged due to the restraining effect of the gas. Stress waves dominate the peak particle velocity (PPV), and increasing the charge weight amplifies the contribution of stress waves, resulting in higher PPV in the surrounding rock. The contribution of blasting gas to PPV is modest, ranging from approximately 1% to 15%, and this contribution diminishes further with increasing charge weight. By introducing a gas-pressure correction term into the Sadovskii formula, we establish a relationship among PPV, charge weight, scaled distance, and gas pressure, quantifying the respective contributions of stress waves and blasting gas to blasting-induced damage and vibration. These findings provide a valuable reference for controlling blasting-related dynamic hazards in engineering practice.

Under the dual influences of global warming and engineering disturbances, creep settlement in cold regions has become increasingly prominent. Currently, the research on the shear and volumetric deformation characteristics of frozen soil during the creep process remains insufficient, particularly concerning the constitutive models that describe the underlying mechanisms. Therefore, through experimental analysis, mechanism exploration, and theoretical modeling, triaxial creep tests were conducted on frozen clay under varying high, low, and highly negative temperature conditions. The shear and volumetric deformation characteristics during the creep process at negative temperatures and different shear stress levels were analyzed in detail, leading to the derivation of a creep constitutive model that quantitatively describes the physical mechanisms of ice-water phase transition and local bonding damage. The main research findings are as follows: (1) under conditions of negative temperature and stress level, as well as during temperature rise, the shear creep characteristics of frozen soil exhibit both attenuating and non-attenuating behaviors, with shear strain and shear stress showing a linear correlation. The volumetric deformation characteristics during the creep process reveal that high negative temperatures are associated with shear shrinkage, while low negative temperatures lead to shear expansion. Additionally, volumetric deformation initially increases and then decreases with rising average stress. Moreover, an increase in negative temperature promotes the development of creep deformation, and at high stress levels, the behavior can transition from attenuating to non-attenuating, ultimately leading to creep deformation failure. (2) Negative temperature and stress levels influence the internal ice-water phase transition and the evolution of local cementation damage, affecting changes in material volume fraction, creep mechanical properties, and interaction modes in frozen soil, thus controlling the overall trend of macro creep development. (3) A creep constitutive relationship that incorporates the ice-water phase transition (pressure melting equation) and the local shear and volumetric cementation failure mechanisms has been established, with its validity and applicability confirmed through predictions based on test data. The results of this study provide a theoretical foundation for the accurate prediction of long-term creep settlement and the evaluation of long-term service performance of roads in cold regions in the context of climate warming.

To investigate the crack evolution process—from initiation to propagation and stabilization—in soils with varying organic matter contents at the slopes of open-pit coal mine dump sites, and to assess the influence of organic matter content on soil water stability, this study focuses on the Yimin Open-Pit Coal Mine in the Inner Mongolia Autonomous Region. Considering the effects of drying-wetting and drying-wetting-freezing-thawing cycles, laboratory tests were conducted on soil samples with different organic matter contents. These tests included drying tests, drying-wetting cycles, drying-wetting-freezing-thawing cycles, and disintegration tests. Using image digital processing software (PCAS), we analyzed the crack initiation and development patterns of soil samples under various testing conditions, as well as the impact of organic matter content on water stability. The experimental results indicate that during the drying process, soil samples with higher organic matter content exhibit faster moisture evaporation rates under the same drying duration. As organic matter content increases, the total crack area, total crack length, and average crack width of the soil surface all show an upward trend, with the cracks developing in a more oriented manner. Following drying-wetting cycles, organic matter has an inhibitory effect on soil desiccation cracking, and after four cycles, the crack morphology tends to stabilize. For soil samples subjected to drying-wetting-freezing-thawing cycles, both the organic matter content and the number of cycles promote crack development. As organic matter content and the number of cycles increase, the total crack area, total crack length, and average crack width also increase correspondingly, with the crack morphology of the soil surface becoming essentially stable after three to four drying-wetting-freezing-thawing cycles.

The interaction between clay mineral particles and pore salt solutions significantly influences intergranular capillarity and adsorption effects, thereby altering the shear strength of unsaturated clay and impacting the stability of infrastructure in saline soil regions. To elucidate the influence of salt solutions on the physicochemical interactions of unsaturated clay, total saturation is categorized into capillary water saturation and adsorbed water saturation, based on generalized soil skeleton theory. Subsequently, an effective stress equation and a soil-water characteristic curve (SWCC) model that incorporate osmotic suction are proposed. The effective stress equation is derived by examining the relationship between the virtual work responsible for volume changes in the generalized three-phase system and energy changes in the capillary water state, which primarily comprises net stress, capillary stress, and osmotic stress. The SWCC model integrates the impact of osmotic suction on adsorbed water content, as well as the linear relationship between the air entry value and osmotic pressure, effectively characterizing the water-holding properties of unsaturated saline clay. Building on this foundation, a shear strength model for unsaturated saline clay is proposed, and its predictive capability is validated using experimental results from unsaturated clay subjected to varying saline conditions. The findings demonstrate that the proposed shear strength formula not only accurately captures variations in the shear strength of unsaturated saline clay under different matric suction and osmotic suction conditions but also reflects the physical and mechanical mechanisms by which salt solutions affect shear strength.

Accurate prediction of shear behavior in structured soils is crucial for geotechnical engineering design and construction. Existing research indicates that the initial structure during shearing often does not completely deteriorate, preventing the critical state line from returning to that of the reconstituted state in the void ratio-stress plane. To address this characteristic, a fractional non-orthogonal elastoplastic constitutive model for structured clay with interparticle bonding has been developed. By introducing parameters that describe critical state properties and combining them with the evolution law of structural state variables, a state-dependent yield surface evolution equation is established. Utilizing fractional calculus theory, the non-orthogonal gradient is derived by directly solving the fractional derivative of the yield surface, which determines the direction of plastic flow. The sub-loading surface theory is incorporated to account for the influence of stress history while facilitating smooth transitions in stress-strain relationships. The relationship between equivalent yield stress and structural state variables is established through isotropic compression behavior, thereby constructing a comprehensive constitutive framework. In this model, the yield surface shape, fractional order, and structural state variables are interconnected, allowing for dynamic adjustments of the plastic flow direction as structural degradation occurs. The model comprises ten parameters, all of which can be determined through conventional geotechnical tests. The model?s validity is verified through parametric sensitivity analysis and comparisons with test results from Corinth marl, Wenzhou marine clay, and Osaka clay. Results demonstrate that the model effectively captures the nonlinear compression characteristics, deviatoric stress peak strength, and volumetric deformation behavior of soils. Compared to the associated flow rule, the model exhibits superior applicability in predicting the mechanical response of structured clays, particularly showing advantages in peak strength and volumetric deformation predictions.