To enable experimental simulation of roadway rockburst induced by fault slip, some experimental principles were distilled based on rockburst process analysis. A large-scale experimental system for simulating roadway rockburst induced by fault slip was developed. The key functions and performance of the system were verified, and physical experiments were conducted to reproduce the phenomena of roadway rockburst induced by fault slip, thereby validating the system?s feasibility in simulating such dynamic events. The results show that: (1) A large-scale experimental system for roadway rockburst induced by fault slip is developed, with model dimensions of 3.0 m×1.5 m×1.5 m. The system features three-directional and four-surface loading capabilities, a maximum loading capacity of 20 MPa, and a loading precision of 0.15% F.S. The system has the function of rapid shear slip for faults, applicable to fault dip angles of 45°–90°. The main reaction frame innovatively adopts a concave-convex interlocking structure connection, achieving high overall stiffness (with a deformation of less than 3 mm under full load). (2) A dense array loading technique suitable for ultra-large-scale models under ultra-high stress is proposed. A compact array loader is invented, employing an embedded cylinder chamber and a shared wall design, enabling efficient multi-point and small-area uniform loading within limited space. This significantly improves load transfer effectiveness across large models. (3) A differential top-bottom coordinated loading method is developed to induce fault activation and slip in the model?s movable zone, where the bottom energy storage-loading is used for rapid retraction, and the top energy-storage loading is used for active follow-up loading. The minimum fault slip time can reach 6 s (slip amount 50 mm). (4) A fault-slip rockburst testing method is proposed, in which critical loading of the roadway is followed by fault activation and slip. During the fault slip process, seismic wave propagation is monitored, and two large-scale burst failures occur in the roadway, reproducing the complete chain process of “fault slip to generate seismic wave-wave propagation-seismic wave inducing roadway rockburst”. This study provides a robust experimental platform for investigating the mechanisms and mitigation strategies of roadway rockburst induced by fault slip.

To elucidate the cooperative bearing mechanism and failure characteristics of surrounding rock-lining structures in underground energy storage caverns subjected to high internal pressure, a large-scale three-dimensional physical model test system was developed. This system comprises a dual-servo internal pressure loading subsystem, a true-triaxial multi-segment confining pressure loading subsystem, a hydraulic servo self-balanced axial sealing subsystem, and a distributed optical fiber monitoring subsystem. It is capable of achieving a continuously adjustable internal water pressure ranging from 0 to 12 MPa, controllable confining stiffness, and axial boundary constraints. Using a lined rock cavern for underground hydrogen storage as a case study, a scaled physical model test was designed and executed, yielding multi-field responses, including internal pressure, confining pressure, lining displacement, and circumferential strain. The test results indicate that the composite lining remains globally stable when the internal water pressure reaches 10 MPa, with only a few controllable microcracks, each less than 0.3 mm in width, appearing in localized zones. The cooperative bearing of the surrounding rock effectively restrains radial deformation and crack propagation in the lining. The distributed optical fiber monitoring system meticulously captures the evolution from the elastic stage to the cracking stage of the composite lining. The proposed test system effectively replicates the high-pressure loading and surrounding rock-lining interaction in underground energy storage caverns, providing a high-fidelity experimental platform and essential testing techniques for the design and safety assessment of underground energy storage projects.

 Existing physical model tests for tunnels often suffer from small model scales, simplified confining pressure loading, and inadequate long-term loading capacity. To address these limitations, a large-scale physical model system for tunnel creep testing was developed to investigate the time-dependent behavior of deeply buried tunnel rock masses. The system comprises a high-stiffness annular reaction structure, a circumferential multi-point radial loading subsystem with servo control, a replaceable similar material model chamber, and a multi-source long-term monitoring platform. It is capable of generating an axially symmetric triaxial stress path and maintaining a stable confining pressure environment in the laboratory. A circular deep-buried tunnel with a geometric similarity ratio of 1:5 was selected as the prototype. A cement-based similar material that satisfies grouped similarity requirements was designed, and a stepped long-term creep test was conducted under incrementally increased confining pressure. Digital image correlation and multi-point convergence monitoring were employed to capture full-field deformation, crack evolution, and tunnel convergence throughout the entire loading process. The results indicate that the confining pressure in all sixteen loading channels remains stable for several tens of days and the circumferential stress distribution is nearly uniform, which demonstrates reliable stress transfer and long-term constant loading performance of the system. The surface strain field of the surrounding rock evolves from scattered strain nuclei to banded and networked high-strain zones. The cracking process progresses through three distinct stages: initial superficial cracks, stable extension and ultimately, through-going failure. The convergence curves clearly exhibit the transition from primary to secondary and tertiary creep stages. These findings demonstrate that the developed large-scale tunnel creep testing system can realistically reproduce the elastoplastic zoning and time-dependent deformation of surrounding rock after excavation. The system provides a robust experimental platform and high-quality data for the long-term stability assessment of deep tunnels and for the optimization of tunnel support design.

Anchorage support is extensively utilized in the stability control of underground structures. The performance of such support varies with anchorage parameters, particularly influenced by the relative position between the anchorage section and the joints in the surrounding rock, which significantly affects the shear resistance of the bolts. To investigate the impact of anchorage parameters on the shear mechanical behavior of bolted joints, a series of shear tests were conducted with varying anchorage lengths and thicknesses. The results indicate that, for a given anchorage thickness, increasing the anchorage length enhances the average shear peak strength and the rod breaking strength by 3.15% and 2.56%, respectively, while simultaneously reducing the average deformation range and the breaking displacement of the bolts by 18.41% and 16.48%, respectively. For a constant anchorage length, an increase in anchorage thickness improves the average shear peak strength and breaking strength by 1.93% and 1.64%, respectively. Post-failure, the deformed bolts exhibit an “S”-shaped profile. As the anchorage length increases, the deformation becomes more symmetric, the normal displacement at peak failure decreases, the axial force reduces, and the shear force provided by the bolts increases. In the pre-peak stage, the ring-down count and energy account for 61.7% and 58.5% of the total, respectively, while the b-value reaches its lowest point at the shear peak. During the post-peak deformation stage, the b-value continuously rises, peaking at the breaking point. Through Gaussian Mixture Model (GMM) statistics of crack types, it is concluded that the shear failure of the bolted joint predominantly occurs in shear, with an average of 86.05% attributed to acoustic emission shear failure and only 13.95% to tensile failure.

 To investigate the influence of liquid CO2 freeze-thaw cycles on the deformation and damage of coal bodies, a self-developed experimental system for liquid CO2 freeze-thaw cycles was utilized. The deformation and damage characteristics of coal under varying conditions of these cycles were examined. An analysis of the strain variation of the coal body revealed the underlying mechanisms of deformation and damage caused by liquid CO2 freeze-thaw cycles. The results indicate that as the number of freeze-thaw cycles increases from 1 to 12, the axial strain, circumferential strain, and volumetric strain of the coal body exhibit a U-shaped trend, first decreasing and then increasing. Notably, the change in axial strain is slightly greater than the reduction observed in circumferential strain, while volumetric strain shows an overall increasing trend. The relationship between the number of freeze-thaw cycles and the absolute growth rate of the minimum strain is exponential, whereas it increases linearly with the cumulative residual strain of the coal body. Greater numbers of freeze-thaw cycles correlate with higher residual strain, increased expansion and contraction rates, and more pronounced effects on the deformation and damage of the coal body. Furthermore, the permeability of freeze-thaw affected coal is positively correlated with the number of liquid CO2 freeze-thaw cycles. These findings provide valuable insights for optimizing gas extraction technologies using liquid CO2 cyclic freeze-thaw processes and for determining the relevant process parameters for liquid CO2 freeze-thaw cracking.

High temperatures in geothermal reservoirs increase the complexity of rock mechanical properties, constraining the effectiveness of hydraulic fracturing stimulation for enhanced permeability in hot dry rock (HDR). Research on the high-temperature rock mechanics relevant to HDR fracturing has primarily focused on two thermal mechanisms: intrinsic high-temperature environments and fluid-rock thermal shocks. Studies have found that, in general, heat treatment at ≥200 ℃ and real-time high temperatures of ≥100 ℃ significantly reduce the strength parameters of granite under compression, and alter the initial and critical states of the yield surface as well as its evolutionary process between them. This indicates that temperature alters the plastic mechanical properties of rock. Other experiments observed that increasing real-time temperature initially increases and then decreases the rock compressive strength. Thermal shock from lower temperatures leads to an approximately linear decrease in both compressive and tensile strength. Temperature increase causes the rock fracture toughness to first increase and then decrease, with the critical transition temperature generally lying between 100 ℃ and 200 ℃. The fracture process zone (FPZ) at the rock crack tip, characterized by plastic softening, notably expands with increasing temperature above 150 ℃, indicating enhanced rock thermoplasticity at high temperatures. Changes in rock microstructure with temperature primarily manifest as three competing mechanisms: the release of different types of bound water, the development of microcracks, and thermal expansion-induced mutual compaction of minerals. These competing mechanisms cause the observed trend of initial increase followed by decrease in rock strength and fracture toughness with rising temperature. Physical and numerical modeling of HDR fracturing indicates that thermal shock from cooler fracturing fluid reduces fracture initiation pressure, induces earlier initiation and delayed propagation of the hydraulic fracture, activates weak planes, alters the stress field around the fracture, and promotes the formation of branch fractures. A temperature-dependent microcrack zone exists around the hydraulic fracture, suggesting that the propagation of hydraulic fracture at high temperatures exhibits thermoplastic mechanical characteristics. Based on the above high-temperature mechanical properties, a theory of rock thermoplasticity has been developed. Focusing on the three thermoplastic zones around the hydraulic fracture—comprising the fracture process zone, the microcrack zone, and the compressive-shear plastic zone—a framework for thermoplastic constitutive modeling has been established. This includes fully coupled stress-strain-temperature constitutive relations and temperature-dependent yield/hardening/unloading criteria that distinguish between hydrostatic pressure and deviatoric stress. By linking the thermoplastic softening in the FPZ, a thermoplastic fracture model integrating plastic and fracture mechanics was developed, characterized by seven parameters that comprehensively describe high-temperature rock fracture. Three thermoplastic response factors were proposed to characterize the microcrack zone around a high-temperature hydraulic fracture, to evaluate the permeability enhancement effectiveness through fracturing. Future research should focus on the dependence of rock mechanical properties on thermal stress under boundary displacement constraints and should distinguish, both experimentally and in modeling, between the effects of heat treatment and real-time high temperature on rock mechanical behavior. It is also necessary to enhance the integration of thermoplastic theory with fracturing models within a coupled Thermo-Hydro-Mechanical-Chemical (THMC) framework and to further elucidate the mechanisms governing the formation and propagation of multiple fractures during HDR fracturing.

 Silica-sol grouting in fractured rock under dynamic water conditions is primarily influenced by the liquid-solid phase transition of the grout, which subsequently governs diffusion patterns and sealing performance. This study employed a 2D transparent fracture model to conduct an orthogonal test program considering variables such as fracture aperture, fracture inclination, and hydraulic head. The evolution of grout diffusion morphology and the corresponding seepage pressure response were analyzed. The results indicate that fracture inclination is the predominant factor affecting sealing efficiency, followed by fracture aperture and hydraulic head; however, significant interaction effects exist among these factors. Under phase-transition control, silica-sol diffusion undergoes a three-stage evolution: from circular spreading to fractal propagation, and finally to a U-shaped pattern. This evolution is influenced by injection pressure, hydrodynamic erosion, fracture geometry, and gelation kinetics. Fractal propagation is more pronounced at higher hydraulic heads and larger apertures, accompanied by increased fragmentation during the U-shaped stage. Mechanistically, the fractal pattern arises from interfacial instability driven by phase-transition kinetics. Progressive gelation results in a gel-shell/liquid-core structure at the advancing front, where a low-strength gel shell forms externally while a more mobile liquid core remains internally. Sustained injection causes radial pushing from the liquid core, alongside tensile stresses induced by dehydration condensation and external hydraulic erosion, promoting shell cracking and preferential channelization. This transformation alters the diffusion front from a smooth circular boundary to a fractal one. The seepage pressure response curve reflects the staged characteristics of grout diffusion and solidification, with its fluctuation amplitude closely correlating to erosion resistance, thus providing a quantitative indicator for assessing the effectiveness of silica-sol grouting under dynamic water conditions.

The mechanism of hydraulic fracture propagation and fracture network formation in tight sandstone is essential for deep energy development. To investigate the influence of injection rate on hydraulic fracture behavior in tight sandstone containing natural fractures, this study conducted large-scale true triaxial fracturing tests on specimens measuring 400 mm × 400 mm × 400 mm from the tight sandstone in the Puguang area of Sichuan. By integrating three-dimensional acoustic emission monitoring, isotope tracing, and fractal analysis of fractures, we systematically analyzed the fracture propagation process under three injection rates (100, 200 and 300 mL/min). The results indicate that: (1) The high quartz content (40.5%) and high brittleness index (64) of the reservoir provide a geological basis for the formation of complex fracture networks. (2) As the injection rate increases, the hydraulic fracture growth rate escalates from 19.95% at 100 mL/min to 67.01% at 300 mL/min, while the fractal dimension increases from 1.102 2 to 1.267 5. The fracture morphology transitions from a single dominant direction to a complex radial network. (3) Increasing the injection rate raises the maximum borehole pressure from 68 MPa to 100 MPa, with cumulative acoustic emission energy surging by 415.1% when the rate increases from 100 mL/min to 200 mL/min, while the increase slows to 6.1% when the rate rises from 200 mL/min to 300 mL/min, indicating that energy release tends to saturate at high injection rates. (4) Acoustic emission events are concentrated along the direction of maximum principal stress (at angles ≤ 45° to the wellbore). The time-series curve of the fractal dimension exhibits a three-stage characteristic: decline–rise–fluctuating decline, revealing the evolution of damage from disordered initiation to ordered connectivity. (5) Theoretical analysis indicates that breakdown pressure is influenced by the competing effects of injection rate and fluid infiltration. At low injection rates, fluid infiltration predominates, resulting in lower breakdown pressure, while at high injection rates, the injection rate dominates, leading to higher breakdown pressure. This study elucidates the influence of injection rate on hydraulic fracture propagation in tight sandstone and provides a crucial theoretical basis for optimizing parameters in deep volumetric fracturing.

Coal pillar instability arises from the interaction between the coal seam and the surrounding roof and floor strata. The stiffness ratio of the roof/floor to the coal seam directly influences the failure mode of the coal seam; thus, the roof, floor and coal seam should be studied as an integrated system. To investigate the effect of the stiffness ratio on the mechanical behavior of coal-rock combined specimens, uniaxial compression experiments were conducted on specimens with varying stiffness ratios. The deformation and failure behaviors during loading were quantitatively characterized based on the elastic modulus and failure intensity. Utilizing acoustic emission technology, the micro-fracture mechanisms and their evolutionary characteristics in combined specimens with different stiffness ratios were analyzed. The controlling mechanism of the stiffness ratio on the deformation and failure of these specimens is revealed from the perspective of micro-crack development and propagation. The research findings are as follows: (1) Variations in the stiffness ratio of composite specimens have minimal impact on the uniaxial compressive strength of the specimens, the strain of the coal samples, or pre-peak acoustic emission (AE) characteristic parameters (e.g., average ringing count, duration, amplitude, and energy). (2) The influence of stiffness ratio on the elastic modulus of composite specimens arises from differences in pre-peak deformation of the rock component. As the stiffness ratio increases, rock strain decreases, leading to reduced axial deformation of the composite specimens and an increase in their elastic modulus. (3) As the stiffness ratio of composite specimens decreases, the average particle size of failed coal samples diminishes while the fractal dimension increases, indicating an intensified degree of coal fragmentation with a decreasing stiffness ratio. (4) A decrease in the stiffness ratio of composite specimens results in a higher post-peak AE event rate, more severe signal blockage in acquisition channels, increased waveform overlap of AE events, and a higher proportion of high-amplitude and high-energy AE signals. Concurrently, the AE b-value and RA value decrease, indicating an increase in large-scale fractures. (5) The failure intensity of coal samples in composite specimens exhibits a negative correlation with the stiffness ratio. From the perspective of macroscopic energy storage and dissipation, a lower stiffness ratio leads to greater pre-peak axial deformation of the rock, allowing it to store more elastic strain energy. At peak strength, the elastic energy released by rock rebound does work on the coal sample, exacerbating coal failure. From the perspective of micro-fracture development and propagation, a decreasing stiffness ratio increases the proportion of large-scale fractures, which break the coal sample into smaller units and enhance coal fragmentation within the composite specimen. Additionally, the AF/RA ratio increases and becomes more concentrated, indicating a higher proportion of tensile fractures. Reduced energy dissipation during the failure process increases the residual energy available for conversion into kinetic energy, thereby intensifying failure.

To investigate the effects of liquid nitrogen freezing on the mechanical properties and meso-damage of coal at various temperatures, uniaxial compression tests and three-dimensional full-field strain measurements were conducted on coal samples subjected to liquid nitrogen freezing across a temperature range of 25 ℃ to 65 ℃. Numerical simulations of coal undergoing liquid nitrogen freezing at different temperatures were performed using a thermal-mechanical-damage coupling model to explore the meso-damage of coal. The results indicate that liquid nitrogen freezing significantly deteriorates the compressive strength and elastic modulus of coal, with the degree of deterioration progressively increasing with higher initial temperatures. Most failures in coal samples are tensile in nature, with the locations of surface cracks aligning with zones of stress concentration. As the temperature of the coal increases, the energy consumed during the deformation and failure of the coal samples decreases, and the stress required at each stage gradually declines. When the surface of the coal sample is exposed to liquid nitrogen freezing for 0.5 minutes, the contact surface temperature drops sharply, creating a low-temperature region around the coal wall and generating high thermal stress on the surface. After 5 minutes of liquid nitrogen freezing, the surface temperature of the coal sample closely matches that of the liquid nitrogen, leading to the development of a distinct temperature gradient within the coal, a gradual reduction in generated peak thermal stress, and an expansion of the thermal stress range. Following 20 minutes of freezing, the temperature at the center of the coal sample begins to decrease, the internal temperature difference narrows, and thermal stress further diminishes while the region affected by thermal stress expands and distributes uniformly within the coal sample. The variation in the mechanical strength of coal during low-temperature freezing with liquid nitrogen can be classified into three phases: a sharp decrease phase, a slow decrease phase, and a stable phase. As the coal temperature rises, the thermal stress induced by liquid nitrogen freezing increases, resulting in more pronounced coal damage and reduced mechanical strength. The compressive strength of coal samples at 65 ℃ decreases by 48.28% after liquid nitrogen freezing, which is significantly higher than the 11.08% reduction observed in coal samples at 25 ℃. This study provides a theoretical foundation for the engineering application of liquid nitrogen fracturing and permeability enhancement technology in deep low-permeability coal seams.

In deep true triaxial stress environments, the interaction between rock masses and support structures is intricate. Traditional constant normal load (CNL) tests, which neglect the effect of support stiffness, face challenges in characterizing the stress evolution paths and the genuine shear response governed by rock-support coupling. This study systematically compares the influence of constant normal stress (CNS) and CNL boundaries on the mechanical response of granite through true triaxial shear tests. The research focuses on analyzing the effects of varying lateral stresses (0–50 MPa) and normal stiffnesses (0–70 GPa/m) on true triaxial shear deformation, shear strength, shear failure modes, fracture surface morphology, and energy dissipation characteristics during shear fracture and friction sliding. The experimental results indicate that lateral stress primarily constrains lateral deformation, while its enhancement of normal deformation is an indirect effect. The CNS boundary effectively suppresses the normal dilatancy of granite. Under CNL conditions, increasing lateral stress exhibits a negative correlation with the rate of peak shear strength enhancement; specifically, the strength enhancement rate was 41.9% when lateral stress increased from 0 MPa to 10 MPa, but decreased to 11.36% when lateral stress further increased from 10 MPa to 50 MPa. The CNS boundary condition continuously suppresses normal dilation and enhances both peak and residual shear strength, with a maximum increase of 52.17%. Furthermore, under CNL conditions, lateral stress tends to induce feather-like secondary wing cracks and step-like fracture surfaces. Conversely, the coupling of dynamic normal stress and lateral confinement under the CNS boundary alters the energy dissipation mechanisms during fracture and friction sliding, resulting in a relatively smoother fracture surface morphology. This demonstrates the significant influence of boundary conditions on the failure mode. The study concludes that considering the CNS boundary provides a more realistic representation of the mechanical response of deep rock masses, offering crucial theoretical insights for support design and stability control.

To investigate the dynamic response characteristics of anchored coal in deep roadways, a split Hopkinson test system was utilized to perform impact tests on anchored coal samples subjected to varying load intensities. A numerical model of pre-stressed anchored coal was developed by incorporating a cohesive unit to simulate the anchorage interface, and the results of the model were compared with the experimental data. The findings demonstrate that the application of pre-stress enhances the dynamic compressive strength of anchored coal through the pre-tightening force of the bolt. Furthermore, as the impact intensity increases, both the peak stress and the dynamic elastic modulus of the samples show a positive correlation, while the expansion rate of the strain field within the anchored coal body accelerates. The end face of the sample first reaches the ultimate strain of the coal body, after which non-uniform strain rapidly develops along the axial direction, resulting in a strain concentration zone in the middle area of the anchorage. Subsequently, the anchored coal body exhibits outward diffusion from the anchorage center in the direction of the impact. The macroscopic failure mode evolves from circumferential shear failure to tensile failure. When the impact load is below 0.4 MPa, the pre-tightening force of the bolt provides some resistance to the propagation of the dynamic stress wave. However, when the impact load is excessively high, large-scale separation occurs between the bolt, the anchoring agent, and the coal body, leading to severe fragmentation of the anchored coal. Under low load strength, crack propagation within the sample is primarily governed by tensile-shear fractures, whereas at high load strength, the failure morphology is predominantly characterized by axial splitting. During stress wave loading, the incident compressive wave traverses the coal medium, crosses the anchorage interface, and propagates into the anchoring agent and bolt, where it undergoes transmission and reflection. The amplitude of the stress wave at the anchorage interface exhibits a linear increase with increasing load intensity, with the peak amplitude increasing by approximately 23%. As the incident compressive wave reaches the first anchorage interface, the stress wave amplitude decreases to 18%, while the peak residual stress wave in the bolt peaks at 25%. Under dynamic loading, the non-cooperative deformation of the anchored coal medium induces interface strain mismatch, and discontinuous deformations, such as anchorage failure and coal particle breakage, contribute to the gradual weakening and dispersion of the compressive stress field. This process exacerbates large-scale deformation and failure of the coal body. These findings enhance the understanding of dynamic stress wave propagation at the anchorage interface and provide an experimental basis for improving stability control strategies in deep roadways subjected to dynamic pressure.

The large-area hanging hard roof is prone to concentrated energy release, making it crucial to control the reasonable breaking step of the roof through subsection hydraulic fracturing to prevent rock bursts. This paper employs theoretical analysis, mechanical testing, numerical calculations, and field tests to examine the characteristics of mining stress influenced by various roof breaking steps. A method for determining the broken chain size during segmented fracturing under mining conditions is established, alongside a reasonable determination of segment length and its field application. The results indicate that the peak mining stress increases with mining activity under the same periodic breaking step of the roof at the working face. Furthermore, the peak mining stress rises with increasing periodic breaking distance across different breaking steps. As the roof breaking step increases, the loading and unloading amplitude of coal samples subjected to the mining stress path also increases, leading to a rise in cumulative acoustic emission energy, which facilitates the instability and failure of coal samples. By analyzing the failure of coal samples and the loading-unloading response ratio in relation to periodic breaking step mining, it is determined that coal samples subjected to a mining stress path with a 32.0 m periodic breaking step can be classified as impact coal samples. The numerical calculation suggests that a segment length of 30 m effectively mitigates the interference between adjacent cracks, resulting in a flatter crack propagation form, while maintaining a relatively small remaining interval between adjacent cracks. Field tests have effectively validated that the segmented length enables controllable energy release. Notably, the single-day microseismic frequency after fracturing increased by 52.0% compared to pre-fracturing levels, while the proportion of high-energy events significantly decreased, with high-energy events exceeding 104 J being effectively eliminated. These research findings provide a reference for the prevention and control of rock bursts in coal mines under similar geological conditions.

The stability of surrounding rock in deep underground engineering is governed by the interaction mechanism between the regional principal stress field and the structure of the layered rock mass. This study demonstrates through true triaxial tests that the mechanical behavior of layered rock is influenced by the synergistic effects of principal stress vectors and structural planes. Based on these findings, an analytical model for roadway stability that comprehensively accounts for principal stress vectors and layered structures is established. By introducing direction cosines, a transformation control equation for regional principal stress paths, derived from azimuth and inclination angles, is formulated. On this basis, a hierarchical extraction algorithm for the mechanical parameters of surrounding rock in a three-dimensional stress field is constructed, and a predictive method for the plastic zone of surrounding rock under the combined influence of principal stress deflection and layered structures is proposed. The failure characteristics of roadways, influenced by the magnitude, direction of principal stress, and layered structures, are systematically analyzed. The theoretical analysis indicates that the failure of roadway surrounding rock arises from the coupled interaction of stress magnitude, direction, and layered structures. Stress amplitude governs the spatial extent of the plastic zone, while the spatial relationship between the direction of principal stress and layered structures determines the predominant direction of failure development. Their synergistic effects result in asymmetric and heterogeneous characteristics within the surrounding rock. The validity of the theoretical analysis is corroborated through engineering case studies. Furthermore, strategies for controlling the stability of roadway surrounding rock are proposed, focusing on three aspects: optimization of support position spatial layout, stabilization of anchoring nodes, and enhancement of the mechanical properties of support materials. This establishes a theoretical foundation for the stability assessment and differentiated support design of deep layered rock mass roadways under complex stress paths.

The primary mineral component of bentonite is montmorillonite. The layered structure of montmorillonite causes the long axes of bentonite particles to align directionally during uniaxial static compaction into blocks, resulting in compaction-induced mechanical anisotropy. This paper systematically reviews the research progress on anisotropy in relation to swelling pressure, swelling/shrinkage deformation, compressibility, and strength characteristics, using Gaomiaozi bentonite as a case study. The findings indicate that swelling anisotropy is linked to the oriented swelling of montmorillonite crystal layers and the rearrangement of montmorillonite particles. An increase in dry density enhances the former, while higher salt concentrations suppress both mechanisms. Additionally, increased moisture content promotes structural homogenization of compacted bentonite. During loading, specimens with vertical compaction planes are more prone to particle slippage compared to those with horizontal planes, which results in reduced yield stress. Under high stress levels, initial anisotropy may diminish, while secondary anisotropy may emerge. Tensile strength is greater when tensile force is applied parallel to the crystal layer plane, and specimens with vertical compaction planes exhibit higher shear strength than those with horizontal planes. Future research should focus on the orientation of soil particles and other microstructural features, the development of predictive models for anisotropy under multi-factor coupling, in-situ testing in underground laboratories, and the behavior and evolutionary mechanisms of bentonite barriers under coupled thermos-hydro-mechanical-chemical (THMC) processes.

Under the influence of global warming and increased humidity, the frequency and intensity of extreme summer climate events on the Qinghai-Tibet Plateau have significantly impacted the stability of permafrost slopes. However, the synergistic mechanisms of combined high temperatures and heavy rainfall on permafrost slope stability remain inadequately studied. To elucidate the triggering mechanisms of summer warm-wet compound extreme climates on permafrost slope collapse disasters, this study conducted comparative experiments using a self-developed dual-controlled (bottom plate-atmosphere) environmental simulation system for cold regions. Three climatic scenarios were examined: normal climate mode, single extreme rainfall, and compound warming and humidifying mode. The hydrothermal-deformation response characteristics of permafrost slopes under these climatic conditions were quantitatively analyzed. The results indicate that summer rainfall induces short-term cooling in the shallow layers of permafrost slopes, while convective heat transfer from infiltrating rainwater accelerates warming in deeper soil layers. Under conditions of single extreme rainfall, thermal erosion of ground ice intensifies, resulting in an increase in toe displacement and collapse area by 22.6% and 100%, respectively, compared to normal climate conditions. Under compound extreme climate conditions, elevated temperatures further accelerate permafrost warming, leading to ground ice melting and the formation of deep, water-rich softening zones. The cumulative toe displacement (148 mm) and collapse area (0.8 m2) increased by 11% and 33%, respectively, compared to single extreme rainfall conditions. Across different climate modes, the severity of slope hydrothermal disasters exhibited a spatial pattern of toe＞middle slope＞crest. The development of slope cracks and the scale of failure followed the order: compound warming and humidifying mode＞single extreme humidifying mode＞normal condition mode, confirming the amplifying effect of compound extreme climates on slope disasters. This study provides a scientific basis for understanding the formation mechanisms of thaw-induced landslide disasters in permafrost slopes on the Qinghai-Tibet Plateau under extreme climate change and supports decision-making for disaster prevention and mitigation.

The coral sand foundations in dredged and reclaimed foundations of South China Sea islands exhibit complex gradation characteristics, with localized strata presenting a sand-gravel mixture. This study investigates the particle breakage and shear characteristics of uniformly graded and gap-graded coral sand-gravel mixtures through a series of high-pressure drained triaxial tests. Experiments were conducted under varying confining pressures, gravel contents, and relative densities. Results demonstrated that gravel content and density significantly influenced the shear strength. Under low confining pressures, the stress-strain curves showed a steeper initial slope, which gradually moderated as the confining pressure increased. The peak principal stress ratio decreased with increasing effective confining pressure. Volumetric strains under high pressures exhibited contractive behavior, intensifying with higher confining pressures. The internal friction angle increased with relative density, peaking at 25% gravel content, while the corresponding apparent cohesion was minimized. Particle breakage intensified with rising confining pressure, and the evolution of particle size distribution followed a pattern characterized by an initially rapid transition followed by gradual stabilization. Furthermore, the peak principal stress ratio decreased progressively with an increase in the relative breakage index. A power function relationship is established between the peak mobilized friction angle and the relative breakage index. Finally, analysis of the stress-dilatancy relationship under high pressure revealed distinct mechanisms of dilatancy behavior under high- versus low-pressure conditions, highlighting the significant impact of gravel content. These findings provide theoretical support for the design and construction of dredged foundation engineering in island and reef environments.

Banana fibers were extracted from discarded banana pseudostems by removing impurities from the leaf sheath surfaces. These fibers were processed through carding and cutting operations to yield structured banana fiber reinforcements. The fibers, prepared at predefined lengths and mixing ratios, were incorporated into loess to create banana fiber-reinforced loess composites. A series of tests—including unconfined compressive strength tests, triaxial shear tests, drying shrinkage cracking tests, disintegration tests, and scanning electron microscopy (SEM) examinations—were conducted to systematically evaluate the effects of banana fiber reinforcement on the strength-deformation characteristics, crack resistance, and water stability of loess from both macro and microscopic perspectives. The experimental results indicated that both the unconfined compressive strength and shear stress of the fiber-reinforced loess significantly increased with longer fiber lengths and higher fiber content. The incorporation of fibers effectively mitigated shrinkage deformation and restricted crack development during drying. Optimal crack prevention was achieved with a fiber length of 15 mm and a fiber content of 0.7%, at which the crack area ratio and average crack width reached their minimum values. Moreover, the inclusion of banana fibers notably enhanced the water stability of loess, as evidenced by reduced average disintegration rates and amounts, as well as prolonged disintegration time. Under varying fiber contents, the stress-strain curves obtained from consolidated drained triaxial shear tests exhibited a consistent strain-hardening behavior, aligning well with the hyperbolic hypothesis of the Duncan-Chang model. By analyzing the correlations between reinforced and unreinforced loess model parameters, empirical equations accounting for fiber content effects were established. Based on these relationships, a modified Duncan-Chang E-B constitutive model incorporating the influence of fiber content was proposed. Comparisons between predicted and experimental curves confirmed the validity of the model.

Deep-water super-large caisson foundations have been extensively utilized in cross-river and cross-sea bridge engineering. The sinking process entails the coupled interaction of significant soil deformation and structural dynamic response, making accurate simulation throughout the entire construction sequence challenging using traditional numerical methods. This paper presents a numerical simulation approach for the complete dynamic sinking process of deep-water super-large caissons, employing the Coupled Eulerian-Lagrangian (CEL) method. A unified model was developed to simulate the entire process, encompassing controlled touchdown, segment addition construction, and sinking through soil excavation. A user-modified Mohr-Coulomb constitutive model was proposed to characterize the distinct mechanical behavior of soil under excavation disturbance (fluid-like response) and structural compaction (solid-like response). The method was validated with the multi-dredging-well stepped-type caisson of the main pier of the Changtai Yangtze River Bridge as a case study. Results indicate that the model accurately reproduced the trajectory of changes in the caisson bottom elevation, with an average relative error of only 2.1% (1.0% error in the final sinking stage) between calculated and measured values, and a coefficient of determination reaching 0.993. This study revealed for the first time the dynamic coupling mechanisms of local soil upheaval induced by cutting edge penetration, the soil backflow phenomenon caused by air-lift soil excavation, and the re-accumulation of migrated soil. The overall kinetic energy to internal energy ratio of the model remained below 1%, confirming the quasi-static nature of the caisson-soil system during the construction process. The friction coefficient between the caisson sidewall and the soil was found to be negatively correlated with settlement, though parameter sensitivity was low. This method provides a crucial basis for ensuring construction safety in caissons subjected to complex soil-structure interactions.

Due to their excellent anti-sliding performance and efficient construction advantages, Portal-type Double-Row Anti-Slide Piles are extensively utilized in large-scale soil landslide control projects. However, existing analytical methods inadequately account for the distribution mechanism of landslide thrust exerted by the horizontal soil arch between piles in the loaded section. To address this issue, a simplified calculation method for landslide thrust distribution in the region behind the arch formed by horizontal soil arches between piles was proposed, based on the limit equilibrium method. Next, considering the slip surface as the boundary, the portal-type double-row anti-slide piles are divided into a load-bearing segment and an embedded segment. A calculation model for the reinforcement of soil landslides by portal-type double-row anti-slide piles is established, which takes into account the distribution effect of the inter-pile horizontal soil arch. By incorporating rigid and semi-rigid connection conditions between the pile top and the crown beam, the force method was adopted to solve the model and derive a corresponding calculation method for the internal forces and deformations of the pile shaft. Utilizing the proposed method, a case study was conducted to verify its accuracy through comparison with numerical simulation results, alongside a sensitivity analysis of key parameters. The study yields several conclusions: there is a significant difference in the pile top bending moment between the front and rear row piles, the rear row piles exhibit a sheltering effect on the front row piles. As the stiffness ratio between the crown beam and the pile body increases, the pile top horizontal displacement and the maximum bending moment of the pile shaft for both the front and rear row piles decrease non-linearly, and increasing the horizontal subgrade reaction coefficient of the embedded segment can effectively reduce the pile top horizontal displacement without significantly affecting the bending moment of the pile shaft.