To address the critical strategic need for sustainable and permanent habitat construction in global lunar base development, this study adopted the concept of in-situ resource utilization on the Moon. Using a simulant with a composition similar to that of Chang’e-6 lunar regolith as the foundational raw material, lunar regolith-based geopolymers (LRG) were synthesized via a solid sodium silicate–activated “one-part” process. A systematic investigation was conducted into the evolution of mechanical performance under the Moon’s extreme surface temperatures (–178 to 113 ℃). The study focused on how key elemental ratios in the regolith (Ca/Si and Al/Si) affected the compressive strength of LRG. Furthermore, leveraging the advanced analytical techniques such as scanning electron microscopy coupled with energy-dispersive X-ray spectroscopy (SEM-EDS), X-ray diffraction (XRD), thermogravimetric analysis (TGA), and nitrogen adsorption–desorption (NAD) tests, this study elucidated the microstructural evolution mechanisms of LRG under the combined effects of thermal cycling and elemental composition. The results showed that increasing Ca/Si (0.26–0.35, Al/Si=0.34) enhanced the compressive strength of LRG but concurrently heightened its susceptibility to cryogenic conditions; increasing Al/Si (0.34–0.58, Ca/Si=0.26) led to a rise-then-fall trend in strength, while cryogenic sensitivity initially decreased and subsequently increased. Notably, when Ca/Si was set at 0.35 and Al/Si at 0.58, the compressive strength of LRG reached 52.25 MPa and remained relatively robust even after cryogenic exposure. Further analysis indicated that strength degradation under cryogenic exposure arose from a combination of internal and external factors: internally, pore-structure deterioration due to collapse of the gel network; externally, pore-water freezing-induced cryo-suction and gel incompatibility collectively contributing to pore expansion. Degradation in high-calcium systems was attributed to the synergistic effects of internal and external factors, whereas in low-calcium systems it was predominantly governed by external factors. Moreover, this study proposed that the optimal construction window for lunar building operations entailed completing mixing and casting within approximately 100 hours prior to the onset of the high-temperature phase of the lunar daytime. This work not only clarified the regulation mechanism governing LRG mechanical performance via the synergistic interplay of Ca/Al/Si but also unveiled the cryogenic degradation patterns across diverse elemental compositions, thereby providing an important theoretical basis and technical support for the efficient utilization of lunar in-situ resources and economically sustainable extraterrestrial construction.

To explore the rock fragmentation characteristics and crack propagation mechanism of tunnel boring machine (TBM) double-disc cutters, scaled penetration tests and numerical simulations were carried out on red sandstone, employing double-disc cutters with two distinct cutter-edge profiles at varying cutter spacings. The mechanical response, failure characteristics, and fragmentation efficiency during rock fragmentation induced by double-disc cutter penetration were subsequently analyzed. The penetration failure process and crack propagation mechanism were revealed using the continuous-discontinuous element method (CDEM). The results show that the rock fragmentation characteristics and mechanical response of the two blade-types of double-disc cutters exhibit significant differences. Compared with the double-disc wedge-shaped cutter, the double-disc cutters with constant cross-section generate a more regular crushing pit profile, characterized by reduced undulation at the crushing pit base and fewer macroscopic cracks on the rock profile. When double disc cutters work synergistically, their superimposed stress fields interact. Both the peak penetration force and the specific energy for rock fragmentation exhibit a trend of first decreasing and then increasing with increasing cutter spacing. Both parameters reach their minimum values at a cutter spacing of s=16 mm. At lower penetration depths, wedge-shaped cutters preferentially generate tensile cracks and cause localized failure, with a significantly higher proportion of tensile cracks compared to constant cross-section cutters. Tensile cracks between constant cross-section cutters can coalesce faster. As penetration depth increases, shear cracks in the compacted core zone continuously propagate, and the proportion of tensile cracks for two types of cutters converges. Double-disc constant cross-section cutters consistently produce larger damaged zones and severely damaged zones, indicating that their energy is concentratedly dissipated in damage zone expansion and shear crack propagation. This study can provide a reference for the design and selection of cutter spacing and cutter-edge profiles for TBM double-disc cutters.

The combined effect of high temperature and alternating rain and heat significantly accelerates the disintegration and degradation of red-bed soft rocks, frequently triggering geological disasters such as slope cracking and deformation, collapse and landslide, as well as water and mud inrushes. This poses a severe challenge to the disaster prevention system in red-bed areas. Investigating the evolution patterns of rock fractures under hydro-thermal cycles plays a pivotal role in preventing and controlling disasters in red-bed regions. To quantitatively analyze the propagation of red-bed soft rock fractures under hydro-thermal cycles, digital image processing technology was adopted alongside a fracture aperture calculation method based on axial transformation. The evolution patterns of fracture apertures and the morphological changes in contour lines of red-layer soft rock fractures were systematically analyzed across a temperature range from room temperature to 100 ℃ were systematically analyzed. The results show: (1) Under hydro-thermal cycle action, red-bed soft rock fractures exhibit significant expansion and evolution characteristics, and the increase in temperature significantly intensifies the fracture expansion effect. At 100 ℃, the fracture apertures increases by 1.96 to 10.54 times compared to the initial value, markedly higher than the 2.25 to 7.93 times observed at room temperature. Further research reveals that fractures with smaller initial apertures are more sensitive to hydro-thermal cycles, and this sensitivity becomes more pronounced with increasing temperature. (2) Spatially and temporally, the analysis of fractures aperture evolution reveals that aperture differences between adjacent measurement points increase significantly with cycling, indicating non-uniform fractures propagation. Higher temperatures significantly promote soft rock crack propagation. (3) The sand and gravel debris at fracture edges, being highly resistant to hydrothermal erosion, develops pronounced resistance differences compared to surrounding clay minerals undergoing continuous erosion. This leads to progressively larger aperture differences between adjacent points. Once clay matrix erosion reaches a critical level, the debris undergoes structural detachment. Synchronous observations reveal that despite local fluctuations in fracture contour roughness, it exhibits an overall significant decline during cycling, indicating sustained abrasion of fracture walls due to hydro-thermal coupling. (4) Hydrothermal cycles exacerbate soft rock damage through the combined effects of thermal expansion/contraction stresses and pore water phase transitions. Temperature variations promote microcrack propagation and enhance permeability while inducing capillary tension via evaporation-condensation cycles. Repeated hydration-dehydration cycles in clay minerals further induce tensile fracturing, resulting in significantly amplified cumulative damage.

With the gradual commencement of construction of key projects in central and western regions of China, the control of large deformation in the surrounding rock of deep buried tunnels has become an engineering technical challenge that needs to be solved in China. The steel pipe shrinkable energy-absorbing cable, characterized by its stable resistance, high controllability and excellent energy absorption capacity, holds promise for addressing the issue of the large deformation in deep engineering construction. Given the commonly observed shear damage characteristics of the support structures in the jointed rock mass, a full-scale double shear test on the cable was conducted to investigate yielding deformation characteristics of the steel pipe shrinkable energy-absorbing cables under rock mass shear action. The test results reveal a constant resistance stage in the load-bearing curve of the energy-absorbing cable. This stage effectively prevents cable fracture through the yielding deformation of the constant resistor, which accommodates rock shear deformation while reducing shear stiffness during the shearing process. Concurrently, the axial load experienced by the energy-absorbing cable remains constant after reaching the designed constant yielding resistance value, enabling it to provide constant resistance to yielding deformation that aligns with the deformation of surrounding rock while maintaining a constant support resistance. Compared with traditional cables, the maximum shear displacements of the 160 kN and 350 kN energy-absorbing cables increase by 89% and 76% respectively, with the absorbed energy rising by 29% and 108% respectively. In the process of rock shear deformation, the energy-absorbing cable can effectively accommodate large surrounding rock deformation while significantly enhancing its energy absorption capacity, thereby providing robust technical support for the prevention and control of surrounding rock deformation disasters in underground engineering construction.

To mitigate the frequent storm surge disasters in the South China Sea, this study proposes the construction of a gabion-armored and geosynthetic-reinforced coral sand sea dike utilizing locally abundant coral sands within the marine area. This approach offers several advantages, including environmental sustainability, low carbon footprint, ease of construction, high stability, and cost-effectiveness, demonstrating broad application prospects. Large-scale wave flume experiments were conducted to investigate the dynamic response characteristics and stability of gabion-armored and geosynthetic-reinforced coral sand sea dike under 50-year return period wave conditions in the South China Sea. The results indicate that the sea dike exhibits excellent stability under tested wave conditions, with the maximum vertical deformation at the crest accounting for only 23.2% of the allowable value in the relevant code. The positive wave pressure on the slope decreases from the toe to the crest, while the negative pressure first decreases and then increases along the same direction. The gabion armors enhance wave breaking, resulting in a reduction of both positive and negative wave pressures on the upper slope. Alternating positive and negative pore pressure oscillations are observed on the offshore side within the dike, while the amplitude of pore pressure oscillations decreases on the far offshore side, albeit with a small amount of accumulated pore pressure. The maximum pore pressure remains significantly lower than the effective stress at the corresponding depth, indicating that liquefaction does not occur in the sea dike. The geogrid undergoes periodic tensile and compressive states, with an average cycle of approximately 24 s, 7.6 times the wave period. The maximum tensile strain experienced by the geogrid is 0.21%, which is below the typical strain range for geogrids in land-based reinforced structures under working stress conditions.

The stabilization treatment of expansive soil and the resource utilization of industrial solid waste are two prominent research areas in geotechnical engineering. The utilization of industrial solid waste to solidify expansive soil offers both economic and environmental benefits. This study employs ground granulated blast furnace slag, fly ash, and waste tire textile fibers to synergistically expansive soil. Through swelling strain tests, unconfined compressive strength tests, California bearing ratio tests, X-ray diffraction analysis, and scanning electron microscopy analysis, the systematic investigation was conducted on the effects of binder dosage, ground granulated blast furnace slag/fly ash ratio, and waste tire textile fiber content on the performance of expansive soil. The results show that when the binder dosage is 15%, the ground granulated blast furnace slag/fly ash ratio is 8:2, and the waste tire textile fiber content is 0.7%, the proposed solidified expansive soil achieves optimal mechanical properties. Compared to untreated expansive soil, the unconfined compressive strength at curing ages of 7, 14, and 28 days increases by more than fivefold. The California bearing ratio of the stabilized soil reaches a maximum value of 25.22%, representing a substantial improvement over the untreated expansive soil, with a California bearing ratio of only 3.16%. X-ray diffraction and scanning electron microscopy analyses confirmed the reinforcing effects of the binder and waste tire textile fibers in the stabilized soil. The proposed method for stabilizing expansive soil not only enhances its engineering performance but also facilitates the resource utilization of waste tire textile fibers—a type of solid waste with extremely low utilization rates—thereby demonstrating substantial potential for widespread adoption and application.

Discontinuity traces constitute linear features generated at the intersection of rock mass surfaces and rock discontinuity, serving as a fundamental and critical parameter for characterizing the strength of rock masses. Quick and accurate identification of the traces of rock mass discontinuities is of great significance for the evaluation of rock mass stability. This paper proposes a novel method for automatically identifying discontinuity traces from 3D point cloud data. Firstly, the potential feature points of the trace are identified based on the differences in the roughness features of the point clouds. Then, the L1 median algorithm is employed to refine the potential feature points and extract the trace skeleton points. This process eliminates the need for point cloud triangulation and can directly process the raw data. Finally, an improved skeleton line growth algorithm is used to connect the discrete skeleton points to generate a continuous trace line. The effectiveness of the new method is validated via four point cloud datasets, and the recognition results are compared with those obtained using CloudCompare and existing methods. The results show that the new method performs exceptionally well in trace line extraction with a high degree of automation, thereby providing effective data support for rock mass quality assessment and stability analysis.

Coral sand (CSD) was mixed with fine soil in varying proportions within real reef ground, exhibiting intricate dynamic behaviors under cyclic loadings, such as earthquakes. Undrained cyclic triaxial tests under diverse confining pressures and cyclic loading stress ratios (CLSR) with various fines content (FC) were conducted to systematically investigate the liquefaction characteristics and excess pore water pressure (EPWP) development. The coarse-grained CSD and fine-grained CSD displayed distinct failure modes, with the fine-grained CSD exhibiting a greater propensity for liquefaction. The liquefaction resistance of CRR₂₀ (cyclic resistance ratio) showed a trend of first decreasing and then increasing with the increase of FC, and reached the minimum value near the critical FC (FC_th). As FC and confining pressure increased, the axial deformation transformed from a gradual accumulation mode to a slow-rapid development mode. Particle breakage decreased with increase of FC but increased with confining pressure and CLSR. The plastic exhibited a trend of initial decrease followed by subsequent increase as FC rose, attaining its minimum value around FC_th. EPWP development curve changed from Type A to Type B as FC and confining pressure increased. EPWP exhibited differential responses during the third stage of Type B development pattern, when FC was relatively low (coarse-grained CSD), the growth rate of EPWP slowed down; while when FC was relatively high (fine-grained CSD), it continued increasing. The EPWP development of CSD mixtures exhibited notable discrepancy from that of quartz sand, thus a new unified EPWP model was proposed. The new findings obtained in this paper provided valuable insights for seismic design of infrastructure on coral island.

To tackle the problems of poor dehydration effect and unqualified removal of heavy metals in the traditional vacuum preloading for contaminated sludge treatment, we propose a vacuum preloading-radiation electroosmosis technology that integrates both dehydration and heavy metal removal capabilities. Through the theoretical derivation of the reinforcement mechanism of the radiation electrode and the removal mechanism of heavy metal ions, we discovered that, compared with the traditional vertical electrode, the radiation electrode substantially improves the drainage volume per unit time and shortens the consolidation time via two mechanisms: superimposing the transverse electric field to augment the overall electric field intensity, expanding the effective action area. Its consolidation and drainage performance surpasses that of the traditional vertical electrode. Moreover, the migration duration of heavy metal ions is longer in electroosmosis with traditional vertical electrodes than with radiation electrodes, proving that heavy metal ions at the identical position are subjected to a more intense electric field from the radiation electrode, leading to accelerated ion migration. To assess the dehydration and heavy metal removal efficacy of the vacuum preloading-radiation electroosmosis technology in treating contaminated silt, we conducted six sets of laboratory model tests, encompassing traditional vacuum preloading (VP), vacuum preloading + electro-osmosis (VP+EO), vacuum preloading + electro-osmosis with cathode-anode radiant electrode (VP+EOCAR), vacuum preloading + electro-osmosis with sustained power (VP+EOSP), vacuum preloading + electro-osmosis with intermittent power (VP+EOIP), and vacuum preloading + electro-osmosis with graded power (VP+EOGP). We analyzed indicators such as drainage volume, settlement, vane shear strength, water content, degree of consolidation, and Cu2+ content. The findings reveal that VP+EOGP yields the optimal dehydration results, with a 38.19% increase in drainage volume, a 26.63% rise in settlement, a 39.51% to 73.4% enhancement in vane shear strength, and a 3.41% to 5.36% reduction in water content compared to VP. For heavy metal removal, VP+EOSP performs best (with a Cu²⁺ removal of 224.04 mg/kg), followed by gradual power-on VP+EOGP (with a Cu²⁺ removal of 220.1 mg/kg). This finding offer a theoretical foundation for the improvement of vacuum preloading technology.

Loess exhibits distinctive properties, including notable macroporosity, water sensitivity, and structural characteristics, with its mechanical behavior being significantly affected by saturation levels. This study focuses on the constitutive modeling of the coupled hydro-mechanical elastoplastic behavior in structured loess, employing Bishop's effective stress and saturation as the primary driving variables within the model framework, while considering the corresponding soil skeleton strain and suction as conjugate variables. Accounting for the variation in structural properties with saturation, we introduce saturation-dependent yield stress structural parameters and frictional strength structural parameters, and formulate an evolution equation for these structural parameters in relation to plastic strain and saturation. Utilizing the yield equation for remolded saturated loess as a foundation, we construct the LC (loading-collapse) yield equation for unsaturated structured loess within the effective stress-saturation space, based on the correlation between saturation and yield stress. A volumetric strain-dependent soil-water characteristic curve is adopted to delineate the hydraulic behavior of loess. By integrating volumetric strain with its mechanical response, we establish a coupled hydro-mechanical elastoplastic constitutive model for unsaturated loess. The model encompasses 14 parameters, which can be calibrated through standard unsaturated triaxial tests and isotropic consolidation tests. The model's validity is confirmed by predicting the outcomes of compression and shear tests on loess under conditions of constant suction and constant moisture content, and comparing these predictions with established experimental data. The findings demonstrate that the model accurately captures the yield strength of loess across varying saturation levels, forecasts the strain softening behavior of unsaturated loess under low confining pressures, and adequately describes the coupled hydro-mechanical behavior of structured loess, thereby offering essential theoretical underpinnings for deformation and stability analyses in loess engineering.

The safe disposal of high-level radioactive waste (HLW) is a prerequisite for the sustainable development of nuclear power. Due to construction-induced disturbances during early excavation and support operations, coupled with the persistent heat generated from radionuclide decay in the subsequent stages, the surrounding rock of the HLW repository undergoes a sequence of excavation, support installation, heating, and cooling. Consequently, fractures within the surrounding rock are subjected to cyclic shearing under thermal conditions, potentially compromising the sealing efficacy of the repository. This study focuses on Beishan granite, the designated host rock for China’s HLW geological disposal repository, as the research subject. To investigate the shear deformation-seepage behavior of rock fractures under disposal-relevant temperatures, cyclic shear-seepage tests were conducted under constant normal stiffness (CNS) and controlled temperature conditions using a novel thermo-hydro-mechanical (THM) coupled testing system. The mechanism governing the evolution of shear deformation-seepage in granite fractures under CNS, with consideration of temperature effects, was elucidated. The findings are as follows: (1) During the first forward shear process, the fracture deformation-seepage evolution is governed by the dominant roughness structure. The hydraulic aperture experiences a rapid increase concomitant with shear deformation, subsequently stabilizing gradually at an elevated level. In subsequent shear cycles, the hydraulic aperture exhibits no significant variation, while normal deformation continuously demonstrates shear contraction during forward shearing and shear dilation during reverse shearing. (2) Under a low normal boundary stiffness (k_n=0.8 and 2 GPa/m), the hydraulic aperture of granite fractures is highly sensitive, exhibiting a sudden increase by up to 6.5 times when the shear behavior switches from contraction to dilation. Conversely, under a higher normal boundary stiffness (k_n=4 GPa/m), the residual shear contraction reaches up to 4.12% after each shear cycle, and the hydraulic aperture decreases by up to 9.84% relative to the preceding loading cycle. (3) The impact of rising temperature on the deformation and seepage characteristics of granite fractures is notably pronounced during the shear process. Throughout the initial shear cycle, an increase in temperature leads to a diminished fluctuation in hydraulic aperture. In subsequent shear cycles, at 90 ℃, a pronounced residual shear contraction of up to 26.31% is observed following each shear cycle. Elevated temperatures induce a pronounced reduction in the permeability of granite fractures. Specifically, the hydraulic aperture at 60 ℃ and 90 ℃ decreases to 0.386 and 0.081 times that measured at 25 ℃, respectively. (4) The fracture with higher roughness (JRC=15.53) exhibits a significant reduction in permeability during cyclic shearing compared to its counterpart with lower roughness (JRC=10.31), with the hydraulic aperture decreasing by up to 40.10%. The research results provide important references for the long-term safety evaluation of HLW geological disposal projects. The observed reduction in the permeability of granite fractures induced by heating and cyclic shearing is positive for the tightness of underground engineering projects, such as HLW geological disposal and compressed air energy storage.

Energy piles are an innovative building energy-saving technology that serves the dual purposes of structural load-bearing and geothermal energy exploitation. The impact of different operational modes and long-term service on the heat transfer performance of the pile and the thermal response of the surrounding soil cannot be overlooked. However, there is a relative scarcity of systematic research on the evolution patterns of thermal responses between piles and soil under long-term service across various operational modes. Addressing this gap, model tests of energy piles in saturated soft soil conditions were carried out under different operational modes. A hydrothermal numerical model for energy piles was developed and validated against experimental findings. The study primarily focused on examining the effects of long-term service and atmospheric temperature on the thermal response of energy piles and the extent of thermal influence on the soil under different operational modes. The results reveal that in short-cycle tests, a longer thermal recovery period corresponds to a higher heat transfer power upon completion of the same thermal cycle. The pile temperature demonstrates a pattern where the central section heats up sharply, while the ends experience a milder temperature rise. Atmospheric temperature exerts a notable influence on the temperature of the upper soil layer, and the thermal influence range of the pile foundation progressively expands in a spindle-like manner. During long-cycle simulations, the proportional disparity in heat exchange powers among various operational modes progressively widens over time, peaking at up to 60%. The maximum temperature differential within the pile foundation consistently hovers around 3 ℃, with relatively gentle fluctuations in the surrounding soil temperature. In the initial phase of the cycle, the thermal influence radius increases linearly with depth until it reaches a certain value and then stabilizes; in subsequent phases, it continues to expand linearly with depth. A longer thermal recovery period within the cycle correlates with a relatively smaller thermal influence radius. Nevertheless, over extended operational periods, the discrepancies in thermal influence radii among different modes gradually diminish.

During the excavation of extremely soft coal roadways, advanced support techniques are commonly employed to ensure construction safety. To investigate the stability of the excavation face under such advanced support conditions, this study develops a combined failure model comprising an inclined prismatic body and a logarithmic spiral surface. This model is developed based on the Hoek-Brown failure criterion and accounts for the effects of advanced support mechanisms. By integrating the upper-bound limit analysis method, strength reduction approach, and geometric equivalence techniques, an objective function for stability safety factor applicable to non-circular roadway faces is derived. A computational program is then implemented to evaluate the stability safety factor, and the results are compared with those from existing studies to validate the rationality and applicability of the proposed method. Furthermore, the influence of various factors on excavation face stability is systematically analyzed. The findings reveal that the proposed method provides a more accurate representation of the nonlinear failure behavior of surrounding rock across different stress zones. Among the influencing parameters, the internal friction angle of the surrounding rock exerts the most significant effect on enhancing excavation face stability. Additionally, increasing the cross-sectional area of advanced pipes and reducing their ring spacing both contribute to improved face stability. The installation of face bolts proves to be a critical support measure for maintaining excavation face stability. Notably, different geometric equivalence methods yield substantially varied results in computing the stability safety factor for non-circular roadway faces.

Landslide-debris flows are characterized by high speed, long travel distance, and high energy, making them highly destructive. Their movement process and deposition patterns are significantly influenced by terrain undulations, but existing studies still lack sufficient understanding of how these factors regulate the kinetic mechanisms. This study systematically conducted 28 sets of debris flow experiments with different volumes (6.76×10³−2.50×10⁴ cm3) and particle sizes (0.4−3.0 cm) using a chute physical model, setting three types of terrain obstacle heights: 8 cm, 15 cm, and 25 cm. These experiments provides a theoretical basis for analyzing debris flow disaster mechanisms in complex terrain. The experiments show that low terrain undulations of 8 cm can trigger particle ejection and secondary acceleration, but with significant fluctuations, while high terrain undulations of 25 cm retain particles, increasing the energy dissipation rate to 91%−96% and significantly inhibiting motion. As particle volume increases, the initial kinetic energy is enhanced, but increased internal friction reduces particle mobility. Larger particle sizes reduce frictional energy loss and enhance fluidity, extending the flow range, and due to their specific flow characteristics, large particles are more influenced by terrain undulations. Analysis of Froude numbers (F_r<0.8) and centroid displacement indicates that the debris flow is in a slow flow state, and terrain undulations dominate energy distribution through the interplay of inertial force and resistance. This study quantitatively reveals the synergistic regulation mechanism of terrain–particle–volume, providing a theoretical basis for risk prediction of debris flow disasters in high mountain canyon areas.

Rainfall infiltration is a critical factor triggering the instability of cobble-boulder mixed soil slopes in Southwest China. To reveal the infiltration characteristics and wetting front migration of such soils under rainfall conditions, in-situ artificial rainfall experiments were conducted on slopes in the Zhala hydropower station reservoir area in Xizang. The dynamic responses of volumetric water content and pore air pressure were monitored under varying rainfall intensities and durations. In combination with two-phase flow numerical simulations, inverse modeling and validation of unsaturated parameters were performed, and the spatiotemporal evolution of wetting fronts under rainfall–post-rainfall conditions was further revealed. A predictive formula for wetting front migration was then established. Results indicate that water content variations are predominantly governed by rainfall intensity, duration, and burial depth, with shallow layers responding sensitively and deep layers exhibiting lag and attenuation. The slope of the water content curve characterizes the relative magnitude of rainfall intensity and depth, surface saturation rarely occurs, and pore air pressure has negligible retardation effects on infiltration. Wetting front migration consists of a rapid advancement during rainfall phase and a slow stabilization phase after rainfall, with both the rate and depth strongly dependent on rainfall intensity and duration. The proposed predictive formula shows high accuracy across different rainfall scenarios. These findings provide theoretical support and parameter references for rainfall-induced seepage analysis and engineering design of cobble-boulder mixed soil slopes.

Landslide disasters frequently occur in the mountainous regions of Southwest China. In recent years, the extensive construction of transportation infrastructure has exacerbated risks associated with landslides. Excavation spoil accumulation and rainfall are primary triggers; however, the instability mechanism of accumulation landslides under rainfall and surcharge loading remains inadequately understood. Taking the Weicheng large-scale accumulation landslide along the Wu-Chang Expressway in Guiyang city as the prototype, this study conducted laboratory physical model tests with a geometric scale of 1:150 to investigate the water content, soil pressure, pore water pressure and displacement of the landslide under rainfall and surcharge loading, thereby revealing its deformation and instability mechanism. The results indicate that graded loading induces stress concentration beneath the loading zone, with the maximum increase in soil pressure reaching approximately 55.9%, and settlements exhibiting a stepwise incremental pattern. In the rainfall phase, soil pressure initially undergoes a slight increase followed by a significant decrease, with the maximum reduction recorded at 93.7%. Meanwhile displacement exhibits a growth trend that is initially sluggish but subsequently accelerates. Additionally, the water content and pore water pressure in shallow soils demonstrate fluctuations in response to rainfall cycles. The accumulation landslide exhibits a failure mode characterized by rearward pushing and front traction. Its instability process can be categorized into three distinct stages: initial stability, toe erosion, and progressive failure. The failure mechanism involves the following sequence: initially, loading induces stress concentration at the rear; subsequently, rainfall infiltration leads to an increase in pore water pressure and a reduction in effective stress; shear strength degradation at the slope toe then results in localized collapse; stress redistribution drives progressive crack propagation and coalescence, ultimately forming multi-level slip surfaces and triggering the progressive failure of the landslide. These findings offer a scientific foundation for hazard assessment, engineering optimization, and risk mitigation of loose accumulation landslides induced by the combined effects of rainfall and surcharge loading.

The southwestern region of China features complex geological conditions and a unique climatic environment, making stability issues of red-bed accumulation slopes unavoidable during infrastructure construction. This study focuses on the Ya'an area and investigates the dynamic response characteristics and deformation failure mechanisms of red-bed accumulation slopes subjected to seismic activity alone as well as the combined rainfall-earthquake effects through dual-model parallel shaking table tests. The research integrates multi-dimensional approaches, including macroscopic deformation analysis, moisture content and pore water pressure monitoring, and dynamic response evaluation. Results show significant differences in deformation and failure patterns between the two conditions. Under seismic action alone, deformation primarily concentrates at the slope crest, manifesting as a “seismic-initiated push-type failure”. In contrast, under combined rainfall-earthquake action, the failure zone expands to both the slope crest and toe, exhibiting a “progressive sliding failure” pattern. In terms of dynamic response, seismic action alone leads to responses concentrated at the slope crest, forming a vertical response zone, whereas under combined rainfall-earthquake action, the dynamic response extends to the upper and middle parts of the slope, propagating from the slope shoulder and surface inward. Analysis indicates that rainfall infiltration increases soil moisture content, reduces effective stress, and raises the damping ratio, which constitutes the primary mechanism for these differences. Moreover, as the number of vibrations increases, both conditions demonstrate an enhanced elevation amplification effect, with the cumulative effect at the slope crest being particularly pronounced under combined rainfall-earthquake action. This highlights the synergistic influence of rainfall and elevation as the core mechanism intensifying the dynamic response of slopes.

Ignoring the liquid-phase inertia might be one of the important reasons for the deviation between the pore water pressure calculated based on one-dimensional consolidation theory and its observed value. To investigate the cause of this deviation, the dynamic consolidation coefficient was introduced to reflect the influence of the liquid-phase inertial force. The Biot dynamic consolidation equation was degraded to a one-dimensional case and applied to the foundation settlement analysis, and its numerical solution was obtained by using the finite volume method. The validity of the proposed numerical solution method was verified by comparing it with the examples solved by the finite difference method in the literature. Then the influences of porosity, dynamic consolidation coefficient and different boundary conditions on the consolidation process were investigated. The numerical results illustrated that both porosity and dynamic consolidation coefficient significantly affect the lag effect of pore water pressure. Among them, a larger porosity corresponds to a shorter duration required for pore water pressure to attain its peak value. Furthermore, there exists a critical porosity, at which the influence of the inertial force of the liquid phase becomes negligible. Porosity exerts a certain degree of influence on the average consolidation degree defined by pore water pressure in the early stage of consolidation, but it will gradually weaken in the middle and later stages. A larger dynamic consolidation coefficient results in a more pronounced lag effect of pore pressure in the early stage of consolidation will be. In addition, when considering the dynamic effect, the distribution of pore water pressure along depth under two-way drainage condition will gradually tend to be symmetrical from the initial asymmetry, and the peak value of pore water pressure exceeds the overburden load value.

The unloading-rebound and recompression deformation characteristics of soft clay significantly affects the design and structural safety of deep foundations, as well as embankments and culvert passages constructed on preloaded soft subgrades subjected to reverse excavation. To investigate the underlying mechanisms, this study conducted a systematic experimental investigation into the unloaded rebound and recompression behaviors of soft clay from Anxiang, Dongting Lake. The theoretical findings were applied to engineering practice subsequently. The results reveal a clear three-stage evolution of rebound deformation with increasing unloading ratio R: when R≤0.2, the rebound rate δ is extremely low, corresponding to a rebound incubation period; when 0.2<R≤0.8, δ increases nonlinearly with R, entering a stable rebound stage; and when R>0.8, δ rises sharply, indicating an accelerated development stage. The critical unloading ratio R_cr of the soil is 0.2, and the ultimate unloading ratio R_u is 0.8. The rebound rate δ is significantly influenced by the preloading history. Under the same unloading ratio, a higher preloading load leads to a higher δ. The maximum rebound rate δ_max after unloading shows a strong linear correlation with the preloading load p_max. The relationship curves between the rebound modulus E_c and the unloading ratio R (E_c-R and lgE_c-R) exhibit two linear segments. The inflection points of these curves, with horizontal coordinates slightly greater than R_cr and close to R_u, respectively, effectively define the stable deformation zone and the significant rebound development zone. The recompression ratio r and the reloading ratio R′ show a nonlinear increasing relationship, which can also be divided into three stages and well fitted by a cubic polynomial without a constant term. A formula for calculating the rebound and recompression deformation of deep foundation pits was established based on the r-R′ relationship. Using R_u and R_cr as boundaries, the soil at the bottom of a foundation pit was categorized into strong, moderate, and non-rebound zones. The critical depth of the rebound zone (i.e., the neutral point depth of piles) was determined, and a formula for calculating the ultimate uplift load of foundation piles at the pit bottom was proposed. These findings providing a quantitative basis for the accurate calculation and risk control of rebound and recompression deformation of soft clay foundation pits, and has been applied in engineering projects such as highway culverts and passages in the Dongting Lake area.

To address the stability issues of stope roof in deep metal mines under high-stress conditions, this study establishes a coupled roof deflection model incorporating lateral confining pressure (surrounding rock pressure) based on elastic theory, using a mining area in China as the engineering context. Subsequently, the analytical solutions for roof deflection equations and maximum deflection are derived. Through systematic orthogonal experiments and parametric sensitivity analysis, the deformation behavior of stope roof is unveiled under the coupled effects of multiple factors, including roof span, thickness, elastic modulus, and lateral pressure coefficient. The results indicate that the additional amplification effect of lateral confining pressure on roof deflection via the P-δ effect (where P denotes the axial pressure on the roof, and  δ denotes the lateral deflection of roof) is relatively limited (just a 0.74% deflection increase in the case study). However, but its impact escalates with the degradation of roof stiffness. Range analysis reveals that roof span (ranging from 146.6 to 294.8 mm) and thickness (ranging from 129.0 to 260.1 mm) are the dominant control factors, with markedly higher sensitivity than elastic modulus and lateral pressure coefficient. Based on parameter sensitivity, an engineering optimization design criterion is proposed: when the burial depth is between 1 000 and 2 000 m, priority should be given to controlling stope span (B≤32 m) and roof thickness (a ≥ 5 m), and reinforcement support measures should be taken for low elastic modulus rock strata (E <100 GPa). The research outcomes provide theoretical foundation and methodological support for safety design in deep metal mine stopes.

To investigate the mechanical response and damage characteristics of conglomerate under blasting conditions, a theoretical computational model for conglomerate blasting zoning was proposed, grounded in the principles of multi-media wave dynamics and tailored to the structural properties of conglomerate. A mesoscopic numerical model for conglomerate blasting was developed, incorporating the matrix, gravel, and their cementation interfaces. Dynamic impact tests on conglomerate specimens were conducted using an electromagnetic split Hopkinson pressure bar (SHPB) system to validate the accuracy of the numerical model material parameters. Subsequently, single-hole blasting simulations were performed on conglomerate with varying explosive types and blast hole diameters, and the area of the crushed zone and crack distribution were quantitatively characterized. The results show that the effective stress in gravel exceeds that in the matrix under blasting loading, and the matrix around gravel is more prone to damage. With the increase in blast hole diameter and explosive performance, both the area of the crushed zone and the fractal dimension of cracks increase. However, when using low-performance explosives, the growth rate of the crushed zone area slows down as the diameter of the borehole increases, while the growth rate of the fractal dimension of cracks accelerates. These findings can provide reference for gravel blasting engineering.

To investigate the influence of temperature effects on the crack evolution mechanism and failure modes of rock, a numerical model for simulating the crack evolution process in rock under thermo-mechanical coupling was proposed based on the ordinary state-based peridynamics method. The model incorporated the maximum principal stress and the Mohr-Coulomb strength criterion to determine crack propagation types. The reliability of the model in simulating material fracture behavior under thermo-mechanical coupling was verified by comparison with numerical solutions from coupled thermo-mechanical finite element method, analytical solutions for stress concentration around an orifice, and rock uniaxial compression fracture test results. Additionally, the mechanisms of temperature effect, ambient temperature, and loading rate on crack initiation and propagation in rocks were explored. The research results indicate that following pre-cooling treatment, the initiation of rock cracks occurs earlier, the propagation rate accelerates, and the peak strength diminishes by approximately 13.1%−17.2%, exhibiting a tensile-dominated failure characteristics. As the ambient temperature increases, crack propagation paths become more concentrated, peak strength improves by 3.2%−18.6%, the failure process is delayed, the development of tensile cracks is suppressed, and the failure mode gradually shifts to shear-dominated crack propagation. When the loading rate rises, the crack evolution duration is shortend. Comparing to room temperature conditions, the difference in the proportion of through-going cracks at the moment of failure under pre-cooling treatment decreases from 22.8% to only 2.95%, suggesting that higher loading rate markedly reduces the sensitivity of temperature effects to propagation.

Accurate identification of crack types during rock fracturing is of great significance for revealing rock mass failure mechanisms and preventing engineering disasters. Shear tests and Brazilian splitting acoustic emission (AE) tests were conducted on sandstone. The Pearson correlation coefficient and random forest methods were employed to optimize AE parameters. The SMOTE (synthetic minority over-sampling technique) method was used to balance the tensile-shear crack dataset, thereby constructing an integrated AE dataset that incorporates tensile-shear crack characteristics. By considering various machine learning model performance evaluation metrics, the convolutional neural network (CNN) was determined as the foundational architecture for the identification model. Subsequently, three deep learning models were constructed: CNN-LSTM (long short term memory)-multi head attention, CNN-GRU (gated recurrent unit), and BO (Bayesian optimization)-CNN-LSTM. It was found that the BO-CNN-LSTM model is most suitable for the real-time identification of rock crack types. The eight optimal identification parameters are average frequency, duration, initial frequency, rise time, peak frequency, centroid frequency, amplitude, and root mean square (RMS) voltage, achieving a comprehensive accuracy of 97.89%. Based on sandstone uniaxial compression tests, the generalization performance of the three deep learning models was compared. The difference in crack classification results between the optimal real-time identification model BO-CNN-LSTM and the conventional Gaussian mixture model-support vector machine (GMM-SVM) clustering model was less than 5%. The newly developed real-time crack type identification model possesses high recognition accuracy and generalization capability, providing a reliable basis for the real-time identification of rock crack types.

To improve the precision and robustness of P-wave picking in microseismic monitoring, this study develops an automatic picking method based on a dual-branch fusion and lightweight deep regression network. The model employs parallel time- and frequency-domain branches to extract temporal dependencies and spectral features, respectively. Specifically, the time-domain branch leverages convolutional neural network–bidirectional long short-term memory network（CNN-BiLSTM）to model waveform dynamics, while the frequency-domain branch employs Mel-frequency cepstral coefficients (MFCC) with two-dimensional convolutional neural network (2D CNN) for robust spectral representation. Both branches integrate squeeze-and-excitation (SE) channel attention channel attention and are adaptively fused through a gated module. The fused features are fed into a lightweight multilayer perceptron (MLP) regression layer to produce a probability curve for precise P-wave onset localization. Experiments on 1 028 real records demonstrate that the proposed method attains a precision of 79.9% within a 5 ms tolerance range, outperforming conventional AIC and STA/LTA methods and maintaining stable performance under low signal-to-noise ratio (SNR) conditions. This method integrates a lightweight design with high precision, thereby offering a reliable tool for phase identification and event localization in microseismic monitoring.

In cold regions, fractured rock masses experience frost heave forces due to the phase change expansion of fracture water, leading to fracture propagation and even rock mass failure. Although existing research has developed models for calculating frost heave pressures, the problem of frost cracking in fractured rocks, which involves the coupled temperature-seepage-stress fields due to moisture migration, phase change heat transfer, and fracture propagation, remains a challenging aspect of current research. Drawing upon the developmental process and mechanism of water-ice phase transition, this study proposes an equivalent water expansion method to simulate phase transition expansion. A finite element model incorporating multi-field coupling is established using COMSOL software to couple the stress field, temperature field and seepage field. Based on experiments on frost heave forces in frozen rock masses, a corresponding multi-field coupled model is established, and the proposed multi-field coupled numerical simulation method is validated from two aspects: frost heave pressure and fracture propagation. The stress, damage and deformation of the fractured rock mass during the freezing process are analyzed. Finally, the evolution patterns of seepage and pore water pressure in the fractured rock are analyzed under varying freezing directions. The simulation results show that the frost heave pressure is mainly composed of the pore water pressure, and the frost heave pressure can be neglected in the region of freezing regions; the range of frost heave pressure action shrinks as the freezing front advances; the fracture propagation starts from the stress concentration area at the fracture tip, with damage extending to the periphery; the freezing direction significantly influences the seepage-hydraulic pressure response: the uniform freezing forms a closed freezing shell to limit the seepage, whereas the unidirectional freezing facilitates the seepage of fracture water out of the fracture, resulting in a reduction of the frost heave pressure. The proposed method offers a theoretical framework for elucidating the freeze-thaw damage mechanism in fractured rock masses and advancing geotechnical engineering research in cold regions.

To accurately invert the dynamic resilient modulus of in-service subgrades from pavement vibration responses under impact loading, overcoming the limitations of existing methods in simulating the dynamic characteristics of layered pavement systems during testing, an inversion method integrating dynamic modeling and artificial intelligence is proposed. First, fully considering the layered nature of pavement systems and the unsaturated characteristics of subgrades, a three-layer pavement dynamics model is established. The solution for the model’s dynamic response is derived using the Laplace-Hankel double transform and its inverse transform. A comprehensive database correlating pavement surface vibration responses with the dynamic resilient moduli of layered pavement systems is constructed through extensive computations. Subsequently, an artificial neural network (ANN) model is developed based on this database. Bayesian optimization is employed to adjust its hyperparameter combinations, yielding an intelligent rapid inversion framework for subgrade dynamic resilient modulus. To validate the method, measured data from 19 typical test sections in the U.S. long-term pavement performance program (LTPP) database are utilized for comparative verification. Results demonstrate that the dynamic resilient moduli obtained via the intelligent inversion method exhibit strong correlation with laboratory-measured values (R²=0.815 0), significantly outperforming conventional inversion methods relying on static mechanics- based models.

Rapid prediction of soft soil foundation consolidation is essential for embankment stability assessment and post- construction settlement evaluation. Existing data-driven models often rely on large-scale labeled samples and neglect the governing physical mechanisms, which limits their accuracy and generalization under complex drainage boundaries and stratigraphic conditions. To address this issue, a physics-informed deep operator model, termed TDC-PI-DeepONet (time-dependent consolidation- physics-informed DeepONet), is proposed by incorporating Terzaghi’s consolidation theory into the DeepONet (deep operator network) framework. In the proposed model, the branch network encodes the consolidation coefficient field, while the trunk network takes the time-depth coordinates as input. Their feature representations are combined to learn the nonlinear operator mapping from the consolidation coefficient field to the spatiotemporal evolution of excess pore water pressure. Considering the spatial variability of soil properties, lognormal Gaussian random fields are used to generate consolidation coefficient samples, and the governing equation, initial condition, and boundary conditions are embedded into the loss function as physical residuals. The model is validated using single-layer and double-layer consolidation cases under different drainage boundaries and is compared with analytical solutions, PINNs (physics-informed neural networks), and standard DeepONet. The results indicate that, compared to the PINNs, the TDC-PI-DeepONet model exhibits more accurate prediction for pore pressure dissipation under various stratigraphic and drainage conditions. The error level is overall lower and the distribution is more concentrated. Based on the statistical results of the test set EL₂, the average error of the model is reduced by approximately 26%, and the deduction time for new cases is reduced from hundreds of seconds to tens of milliseconds. When compared to the DeepONet model, the TDC-PI-DeepONet model achieves lower errors and more stable predictions in multilayer heterogeneous cases. The TDC-PI-DeepONet model provides an efficient and physically consistent approach for rapid consolidation prediction of complex soft soil foundations.

Internal deformation data of surrounding rock serve as a critical basis for evaluating the stability of surrounding rock in deep underground engineering. Based on the principles of ultra-weak fiber Bragg grating (UWFBG) and time division multiplexing sensing, an externally anchored UWFBG strain sensing optical cable was developed. With a sensing point spacing of 1 m, the proposed optical cable enables quasi-distributed measurement of surrounding rock deformation. Calibration tests were conducted to investigate the sensing characteristics, error indices, and long-term stability of the proposed strain sensing optical cable. Subsequently, the optical cable was embedded into cement mortar specimens with varying strength grades to carry out three-point bending tests. These tests were designed to explore the strain transfer effect of the optical cable within cement mortar media exhibiting different mechanical properties. On this basis, a correction coefficient for strain transfer efficiency that takes the uniaxial compressive strength of the host material into account was proposed. Furthermore, drawing upon the deep roadway engineering project at Pingmei No. 4 Coal Mine, field tests were conducted on the strain sensing optical cable to validate its reliability and accuracy under complex geological conditions. The research results demonstrate that the strain sensing optical cable exhibits excellent linearity, repeatability and long term stability, with a maximum measuring range of 30 000×10⁻⁶. After introducing the proposed correction coefficient, the measurement accuracy of the strain sensing optical cable is significantly improved.