Addressing the limitations of current research on tunnel thermal disasters, which primarily focuses on managing high rock temperatures and ambient temperatures during construction while neglecting heat source mechanisms and accompanying disasters such as high-temperature water inrush and harmful gases, this study systematically deconstructs the formation mechanisms and disaster effects of tunnel thermal disasters by constructing a novel cognitive system termed “heat source-gas source-response-disaster.” Firstly, based on thermal phenomena observed in geothermal anomaly zones and extensive engineering practices related to thermal disasters, tunnel thermal disasters are redefined to collectively encompass high rock temperature, high-temperature water, and harmful gas inrush. Secondly, through a statistical analysis of 87 global cases of tunnel thermal disasters, thermally-induced disasters in geothermal anomaly zones are classified into five types: high rock temperature, hazardous gas, high rock temperature combined with high-temperature water inrush, high rock temperature combined with hazardous gas, and high rock temperature combined with both hazardous gas and high-temperature water inrush. This classification reveals a strong correlation with surface thermal anomalies and the formation lithology. Thirdly, the effects of tunnel thermal disasters are systematically categorized, illustrating their detrimental impacts on the tunnel’s temperature and humidity environment, exacerbation of surrounding rock deformation and failure, induction of cascading failures in support structures, and triggering of multi-dimensional gaseous disaster cascades. Building upon this, the Qinghai—Tibet Plateau is examined as a case study to investigate the origins, sources, and conduction mechanisms of heat and gas in typical geothermal anomaly zones. This establishes a theoretical paradigm for the synergistic genesis of thermal disasters, involving crustal, mantle, or crust-mantle heat and gas generation, coupled with tectonic heat and gas conduction. Finally, the reliability of the research findings is validated through an analysis of a typical tunnel thermal disaster case. This study provides a theoretical foundation for further exploration of disaster-prone patterns, identification of hazard structures, and formulation of prevention and control measures for tunnel thermal disasters.

The highly heterogeneous three-dimensional pore network structure of coral reef limestone results in significant anisotropy and size effects on its mechanical properties. This research developed digital rock core models of coral reef limestone using CT scanning and adaptive threshold segmentation techniques, enabling the conversion between these digital models and finite element models through spatial mapping relationships. The mechanical response, damage evolution, and changes in porosity of coral reef limestone specimens of varying sizes under quasi-static compression were systematically investigated. The results indicate that: (1) As the model size increases, peak strength nonlinearly rises from 0.25 MPa to 7.26 MPa, with material behavior transitioning from ductile to brittle, and the post-peak response shifting from gradual strain softening to a sudden stress drop; (2) As specimen size increases, stress distribution evolves from highly concentrated patterns to more uniform distributions, ultimately developing into complex discrete multipolar states, with dominant stress directions becoming increasingly pronounced; (3) Larger models exhibit more complex fractal characteristics and progressive pore reconstruction processes, with significant nonlinear relationships observed between fractal dimension and porosity changes. This study elucidates the interaction mechanism between multi-scale pore structures and macroscopic geometric dimensions, providing a theoretical foundation for stability evaluation in island reef engineering.

Deep rock masses are situated in a complex environment characterized by high ground stress and elevated water pressure. Under repeated blasting disturbances, these rock masses are prone to inducing water inrush disasters. However, research on the effects of water pressure on the strength degradation characteristics of rock remains limited, resulting in a lag of relevant theoretical studies behind engineering practices. In this context, a self-developed dynamic test system has been employed to conduct cyclic impact tests on red sandstone under hydro-mechanical coupling conditions. Based on the test results, the evolution of dynamic peak stress, dynamic degradation coefficient, dynamic intensity factor (DIF), and damage variable with cyclic impacts has been analyzed. Additionally, dynamic damage accumulation and fatigue life models for rocks subjected to hydro-mechanical coupling conditions have been established to explore the mechanisms by which hydro-mechanical coupling affects strength degradation. This research aims to elucidate the mechanisms behind water inrush disasters in deep rock mass engineering and to enhance the safety resilience of such projects. Key findings are as follows: both the dynamic peak stress and DIF of rock under hydro-mechanical coupling conditions exhibit accelerated decreasing trends, while the dynamic degradation coefficient and damage variable increase rapidly, demonstrating distinct strength degradation characteristics. When the number of cyclic impacts approaches 50% of its fatigue life, the growth rate of the dynamic degradation coefficient spikes, resulting in an “accelerated deterioration phenomenon”. Water pressure mitigates the rate of change of dynamic peak stress, DIF, dynamic degradation coefficient, and damage variable in rock, thereby attenuating the “accelerated deterioration phenomenon” and prolonging fatigue life. Conversely, axial static stress increases the rate of change of these parameters, intensifying the accelerated deterioration phenomenon and shortening the fatigue life of the rock. The developed fatigue life model for rocks demonstrates an error rate below 16%, accurately predicting the fatigue life of deep rock masses.

As a precursor signal of rock failure, magnetic fields offer distinct advantages such as slow attenuation, a wide propagation range, and strong anti-interference capability. Investigating the generation mechanisms and evolution patterns of magnetic fields during rock failure not only enhances our understanding of rock failure mechanics but also provides theoretical support for geological hazard prediction. To elucidate the effects of loading rate and water saturation on the magnetic field response mechanisms associated with rock failure, this study selected three typical rock types—coarse sandstone, limestone, and granite—and conducted uniaxial compression tests under controlled loading rates (0.1 mm/min, 0.3 mm/min and 0.5 mm/min) and water conditions (natural dry and vacuum-saturated states). Stress fields, magnetic field signals, and acoustic emission (AE) signals were synchronously recorded to analyze the relationships among magnetic field signals, loading, and AE throughout the entire rock failure process. The results indicate that: (1) Peak magnetic induction intensity increased to varying degrees across all three rock types with increasing loading rates, exhibiting enhancements of approximately 1.6 to 2.7 times; (2) Water saturation triggered a dual-path response: for saturated limestone and granite with low porosity and minimal water content, peak stress remained relatively unchanged compared to the natural state, whereas peak magnetic induction intensity increased substantially due to streaming potential effects; conversely, saturated coarse sandstone, significantly affected by water-induced degradation, exhibited reduced compressive strength and weakened AE activity, with magnetic field characteristics transitioning from abrupt jumps to accelerated growth after peak stress due to the absence of brittle fracture; (3) Loading tests on saturated specimens demonstrated that water reduced both peak stress and AE activity to varying extents while significantly enhancing magnetic field signals; (4) X-ray diffraction (XRD) analysis revealed that Si- and Fe-bearing minerals such as quartz and pyrite possess electrical and magnetic properties, serving as key material prerequisites for rock magnetic field response. Combined with scanning electron microscopy (SEM) imaging, the analysis confirmed that rocks with dense structures and intense fracturing are more prone to generating significant magnetic field signals. This study highlights the distinct differences in rock magnetic field evolution mechanisms under varying loading rates and water saturation conditions, providing important insights into the physical processes of rock failure and geological hazard prediction.

To address the empirical challenges and insufficient quantification in current rockburst identification methods for deep, high-stress hard rock tunnels, this study proposes an energy-based rockburst criterion grounded in distortional and volumetric energies, accompanied by a corresponding energy-release calculation method and a constitutive failure model. Laboratory rock mechanics tests, combined with acoustic emission (AE) monitoring, were conducted to systematically investigate the evolution of distortional and volumetric energies during compression loading and their intrinsic relationship with rockburst energy release behavior. The results indicate a sharp increase in AE activity as the ratio of distortional to volumetric energy approaches a critical threshold. Furthermore, higher confining pressures significantly suppress AE events, illustrating the inhibitory effect of volumetric energy on brittle failure under elevated confinement. Numerical simulations of tunnel excavation, based on the proposed constitutive model, further reveal the dynamic evolution of energy partitioning and the transitions in failure modes within the surrounding rock during excavation. These findings demonstrate that appropriate support measures can optimize the stress path of the surrounding rock, enhance the proportion of volumetric energy, and shift the failure mode from brittle to ductile, thereby effectively mitigating the risk of rockbursts.

Collapse controlled by stress and rock mass structure (hereafter referred to as stress-structure collapse) is a prevalent hazard in relatively intact rock masses of large deep underground caverns, significantly threatening construction safety and the stability of supporting structures. This study systematically investigates the characteristics and formation mechanisms of stress-structure controlled collapse in hard rock under strong unloading conditions, based on a collapse case database established from the roof excavation of the underground powerhouse at the Baihetan (BHT) Hydropower Station. The investigation employs field studies, multi-factor analysis of construction-stress-geology, rock mass unloading tests, and theoretical analysis. The research findings indicate the following: (1) Collapses predominantly occur in stress concentration zones of cavern sections near the working face, where substantial damage to the surrounding rock is evident, often accompanied by spalling failure. The depth of the collapse is generally less than 2 m. The rock mass in the collapse area is primarily classified as Class II or III, typically containing two or three groups of hard unfilled structural planes of Class IV, with a geological strength index (GSI) ranging from 45 to 70. (2) The stress threshold value inducing collapse formation is approximately 30% to 50% of the rock’s uniaxial compressive strength, with at least one group of dominant structural planes intersecting the tunnel axis and the excavation tunnel wall at a small angle (i.e., 0° to 40°). (3) Under the strong unloading effects of cavern excavation, the dominant structural planes in the shallow surrounding rock propagate along their tips due to the concentration of tangential stress, gradually extending in a direction parallel to the excavation surface and ultimately connecting with other structural planes and the cavern wall, leading to collapse. Scanning electron microscopy (SEM) and laboratory test results indicate that this cracking process is primarily governed by tensile failure. (4) Analysis of construction-induced factors demonstrates that stress-structure collapse can be effectively mitigated or redirected by optimizing excavation schemes, support parameters, and support timing. The cracking-restraint design method (CRD) can be employed to enhance prevention and control strategies for such collapses. The research outcomes provide guidance for understanding the mechanisms of stress-structure collapse and its prevention in deep rock engineering.

Carbon dioxide (CO2) capture, utilization, and storage (CCUS) represent an effective decarbonization pathway that can accelerate achievement of dual-carbon targets. However, during geological CO2 storage, the injection rate and fault roughness are critical factors influencing the risk of fault activation, yet the underlying mechanisms remain unclear. To address this, this study simulated the fault activation process through shear slip of fractures in a laboratory setting, systematically conducting triaxial shear-seepage tests on sandstone fractures under varying CO2 injection rates and roughness conditions. The morphology of the fracture surfaces before and after the tests was reconstructed and analyzed using a 3D scanner. The results indicate that: (1) The critical activation pressure of the fracture increases with the CO2 injection rate but decreases with increasing fracture roughness. (2) As either the CO2 injection rate or roughness increases, the time required for fracture activation during the injection-driven phase decreases. The deformation mechanism gradually shifts from being predominantly governed by shear slip to being co-controlled by shear slip and normal displacement, with normal displacement occurring prior to shear slip. (3) For smooth fractures, changes in surface morphology post-activation are negligible, and permeability increases only slightly. In contrast, for rough fractures, surface roughness decreases significantly after activation; however, permeability increases substantially due to the dilation effect. The findings of this research provide a theoretical reference for optimizing injection strategies and warning against fault instability risks in CO2 geological storage.

High-frequency acoustic emission (AE) signals generated during rock microcracking contain essential source parameters—such as moment magnitude, source radius, and stress drop—that characterize the underlying rupture process. To facilitate their quantitative inversion, this study develops and calibrates a broadband, high-fidelity AE sensor, AcouSeek, which exhibits a flat displacement response across the frequency range of 10 to 1000 kHz. Two standard sources—steel ball impact and capillary glass rupture—are employed to establish an absolute calibration methodology, grounded in Hertzian contact theory and a step-loading model, respectively. In a three-point bending test with controlled crack opening, 24 AcouSeek sensors are utilized to record AE signals. A total of 561 microcracking events are analyzed through displacement-based spectral fitting and physical source models, resulting in moment magnitudes ranging from –8.6 to –7.2, source radii between 1.6 and 2.9 mm, and stress drops between 30 and 82 kPa. The obtained parameters are consistent with those reported for mining-induced and natural earthquakes, demonstrating cross-scale compatibility. This study establishes a comprehensive workflow—from broadband sensing to absolute calibration and source parameter inversion—providing a standardized and extensible approach for integrating laboratory AE studies into the framework of quantitative seismology.

To investigate the scale effect in blasting crater tests, this study was conducted utilizing dimensional analysis and similarity theory. By considering the strength size effect, strain rate effect, and explosive energy distribution, the mechanical mechanisms underlying the size effect in blasting craters were elucidated. A modified charge calculation formula that incorporates the energy coupling coefficient was proposed. Coupled with experimental data, the applicable scale range of conventional blasting model tests was evaluated. The results indicate that, compared to medium-charge tests (measured in kilograms), small-charge tests (measured in grams) are significantly influenced by the strength size effect and strain rate effect, resulting in higher apparent material strength and greater energy consumption for material fragmentation. In contrast, in large-charge tests (measured in tons), a substantial portion of the explosive energy is expended in overcoming gravitational potential, which must be considered when analyzing energy distribution. These differences in energy partitioning give rise to the observed size effects in blasting crater experiments. Unlike traditional empirical formulas based solely on geometric similarity, the improved formula accounts for the energy coupling coefficient and the distribution characteristics of blasting energy across various charge scales. For instance, in small-charge iron ore blasting tests documented in the literature, the improved formula reduced the average relative error in crater radius prediction from 110.0% to 13.3%, demonstrating a significant enhancement in accuracy. Furthermore, the improved formula allows for reliable extrapolation from medium-charge tests to large-charge scenarios. Based on test data from desert alluvium, it is shown that this extrapolation remains valid within a geometric scale factor of less than 20. However, under the influence of the strength size effect and strain rate effect, the energy distribution in small-charge tests is markedly different from that in medium-charge tests, particularly in terms of fragmentation energy, which leads to considerable deviations when applying small-charge results to larger scales.

The national standard mandates that sensors be installed on the surface of natural soil when assessing the attenuation characteristics of ground vibrations caused by metro systems. However, urban environments often lack exposed natural soil surfaces conducive to ideal testing conditions. Consequently, most current measurements are conducted on hardened ground surfaces, which clearly contravenes regulatory requirements and raises concerns regarding the reliability of the findings. To investigate the effect of hardened ground on the attenuation of metro-induced ground vibrations, this study develops an analytical model of a semi-infinite layered foundation soil subjected to moving harmonic loads and compares the ground vibration responses before and after the installation of the hardened surface. The results from both measurement and theoretical analysis indicate that the influence of hardened ground on surface vibrations is significant. Due to the high wave impedance and waveguide effects of hardened surfaces, surface vibrations at different frequencies may either attenuate or amplify in far-field regions. The impact of hardened ground becomes more pronounced when the metro is buried at a shallow depth, when the foundation soil is soft, or when the hardened surface layer is thick. These findings provide a theoretical foundation for the measurement and evaluation of metro-induced environmental vibrations.

The bauxite deposits in Shanxi Province are located within the Carboniferous-Ordovician composite aquifer system. Understanding the mechanisms of ore body weakening and fracture propagation due to water-rock interactions is essential for ensuring the safe and efficient mining of bauxite. This study focuses on the Loufan bauxite in Shanxi, conducting Brazilian tests on specimens with varying water saturations (0%, 25%, 50%, 75%, and 100%) to investigate tensile strength, damage evolution, and fracture energy characteristics. A parametric program was developed to quantitatively characterize the fracture surface morphology and study the effect of water saturation on the splitting characteristics. Molecular dynamics simulations and CT scanning tests were employed to elucidate the micro-mechanisms of water-rock interactions that weaken bauxite strength. Additionally, ESEM images were analyzed to reveal the fracture mechanisms at different water saturations. The results indicate that: (1) Diaspore, the primary component of the studied bauxite, exhibits strong hydrophilicity. (2) As water saturation increases from dry to fully saturated, the tensile strength decreases from 14.851 MPa to 6.364 MPa, and the fracture energy drops from 10 426.63 J/m² to 6 772.35 J/m², indicating a progressive weakening of its resistance to failure. (3) The Joint Roughness Coefficient (JRC) increases from 7.30 to 17.02, and the fractal dimension rises from 1.923 00 to 1.948 45, suggesting that the fracture surface morphology becomes increasingly complex. (4) CT scan results indicate that mineral expansion and dissolution are not the primary mechanisms contributing to strength weakening. (5) Molecular dynamics simulations and ESEM analysis reveal that water molecules interact with the diaspore crystal through hydrogen bonding, which reduces inter-crystal friction and alters the fracture mechanism. The failure transitions from transgranular fracture (dominated by the cleavage of Al-O covalent bonds) to intergranular fracture (dominated by the rupture of hydrogen bonds (O?-H…O?)). Macroscopically, this results in rougher fracture surfaces and a gradual reduction in tensile strength. These findings provide experimental evidence for the mining of diaspore-type bauxite.

Affected by geological processes, the rock mass of the reservoir bank slope typically displays initial damage. During reservoir operation, the rock mass within the water-level fluctuating zone experiences prolonged ‘soaking-air drying’ dry-wet cycles, which may be further affected by tectonic earthquakes or reservoir-induced seismic disturbances. To elucidate the mechanisms of damage and deterioration of the reservoir bank slope rock mass under these complex conditions, typical sandstone from the reservoir bank slope was selected, and damaged rock samples were prepared through cyclic loading and unloading. Systematic dry-wet cycling and dynamic response tests were conducted. The results reveal that: (1) The compressive strength of both intact and damaged rock samples decays in a “fast-slow-stable” manner under dry-wet cycle conditions. Initially, the strength of damaged rock samples is slightly higher than that of intact samples due to pore compaction; however, significant deterioration occurs in the later stages under the influence of dry-wet cycles. (2) Under dry-wet cycles, the dynamic elastic modulus of both types of rock samples continues to decay, while the damping coefficient, damping ratio, and energy dissipation ratio increase synchronously. Notably, damaged rock samples exhibit a higher rate of deterioration and response sensitivity, with their dynamic parameter evolution displaying clear nonlinear and phased characteristics. (3) The product of porosity and pore throat median diameter serves as a comprehensive damage variable, effectively and quantitatively characterizing the microstructural evolution of rock samples with initial damage under dry-wet cycles. (4) The degradation of dynamic characteristics in damaged rock samples under dry-wet cycles is driven by both initial damage and cycling effects. While cyclic loading and unloading induce short-term densification through stress concentration and particle breakage, they also facilitate crack propagation. Initial defects enhance seepage pathways and weaken cementation, leading to structural loosening and continuous damage accumulation until stabilization occurs. Therefore, in long-term seismic performance analyses of the reservoir bank slope, it is essential to systematically consider the coupled effects of initial damage and dry-wet cycles on the deterioration of the rock mass’s dynamic characteristics.

To assess the suitability and safety of surrounding rocks in abandoned mines for Compressed-Air Energy Storage (CAES) projects, this study examines the abandoned iron ore mine located in the eastern part of the Nalengele River in Qinghai Province, China. It systematically analyzes the engineering geological features and static-dynamic mechanical behaviors of the rocks. Researchers employed digital image correlation (DIC), Split Hopkinson Pressure Bar (SHPB) tests, and ultra-high-speed photography to accurately characterize the dynamic responses and failure mechanisms of three key rock types: granite, marble, and skarn. The results reveal distinct differences in static properties; both granite and skarn exhibit high elastic moduli and a significant potential for rockburst. Under dynamic impact, all rock types follow a consistent pattern of strength enhancement, fracture evolution, and energy dissipation. Dynamic compressive strength and elastic modulus increase sharply with strain rate, and the dynamic increase factor (DIF) shows a linear correlation with the logarithm of the strain rate. As the strain rate rises, crack propagation accelerates, transitioning macrocrack forms from tensile to shear dominance. Specifically, granite demonstrates limited strength growth, characterized by rapid through-going linear fractures that primarily convert energy to crack expansion and debris, resulting in low energy absorption efficiency. Marble exhibits the most substantial increase (DIF ranging from 1.70 to 3.45), with restricted, winding cracks facilitating top energy dissipation. Skarn presents intermediate characteristics, featuring irregular cracks and plastic failure following ductile slip, with moderate energy dissipation. At an impact air pressure of 0.30 MPa, granite and skarn display the phenomenon of “failure strain saturation,” where the slope of the peak strain relative to the increasing strain rate is less than 0.05, indicating saturation, and the strain response ceases to increase significantly with rising air pressure. In contrast, marble continues to demonstrate excellent energy absorption and buffering capacity. These insights provide essential theoretical and practical guidance for site selection, air pressure thresholds, and rock stability in abandoned-mine CAES projects.

The morphological development of coal-rock fractures is a critical factor influencing their mechanical properties, and accurate identification of these fractures is essential for ensuring mine safety. However, existing detection methods often struggle to accurately identify fine fractures and fragmented zones in complex environments. To address this challenge, this study introduces an enhanced AttRes-UNet model based on the ResNet-UNet framework, which incorporates both channel attention (SE) and spatial attention (CBAM). The model utilizes residual structures to improve feature representation and employs attention mechanisms to enhance the extraction of key fracture regions, thereby significantly boosting fracture recognition in complex conditions. A high-quality annotated dataset of real coal-rock images was constructed, and the impacts of two residual backbone depths, ResNet18 and ResNet34, on model performance were systematically assessed. Experimental results reveal that the AttRes34-UNet model significantly outperforms traditional UNet and ResNet-UNet models in terms of accuracy, F1 score, and IoU, demonstrating superior stability and generalization in fine fracture detection and in delineating fragmented zone boundaries. The proposed AttRes-UNet model offers an efficient and precise technical approach for automatic recognition of coal-rock fractures, providing scientific support for safety monitoring and disaster prevention in coal mining operations.

Support optimization for hydraulic tunnels poses a significant technical challenge during dynamic adjustments in site investigation, design, and construction. Traditional analytical methods and conventional numerical software often fall short of meeting the on-site engineering demands for high accuracy and rapid response. Consequently, there is an urgent need for the development of automated and ultimately intelligent design solutions in underground engineering. To address challenges in hydraulic tunnel support design, such as low modeling efficiency and insufficient standardization of parameters, this study first utilizes the 3DEC platform interface to propose an integrated modeling framework. This framework combines a unified representation of geometric parameters with a quantitative representation of multi-source rock mass parameters, enabling high-fidelity reconstruction of tunnel models under realistic geological conditions. Next, based on rock mass quality and the spatial distribution characteristics of discontinuities, the study automatically generates an initial sample library of support schemes, facilitating the automatic matching of support design options to surrounding rock classes and cavern spans. Concurrently, by incorporating a rock bolt spatial positioning algorithm and a stiffness equivalence model, the study introduces a parametric design approach and a unified mechanical representation method for the support system. Following this, the study analyzes the tunnel stress-deformation response and the spatial distribution of block collapse by traversing candidate schemes. A multi-criteria matrix evaluation is then conducted to identify the optimal support design. Finally, an intelligent tunnel support design platform is developed, which integrates “parametric modeling, automatic scheme generation, iterative comparison and selection, and multi-criteria decision-making” into a cohesive workflow. Using a large-scale hydraulic tunnel project as a case study, this work combines orthogonal experimental design, range analysis, and multi-criteria matrix evaluation. The results demonstrate that, in practical engineering applications, the three categorical indices achieve prediction accuracies exceeding 80%, while the quantitative prediction indices attain coefficients of determination (R²) greater than 0.89, confirming the feasibility and validity of the developed platform. Thus, the proposed approach offers an efficient and portable computational tool along with technical support for optimizing support and enabling dynamic adjustments during hydraulic tunnel construction.

The macroscopic mechanical behavior of rock is fundamentally governed by its mesostructure. However, most existing mesoscopic DDA models rely on ideal geometric units and structural simplifications, making it challenging to accurately represent the actual mesoscopic characteristics of rock. To address this limitation, this study proposes a mesoscopic numerical modeling approach for rock that integrates Mask R-CNN and DDA, enabling the direct construction of high-fidelity numerical models from real mesoscopic images. The proposed method first enhances mineral boundary features in rock images through wavelet transform, and then utilizes Mask R-CNN for accurate identification and segmentation of minerals and pores. Based on this, a high-precision DDA model is established to systematically analyze the influence of mineral composition and pore characteristics on the mechanical properties of rock. The results indicate that quartz can inhibit the initiation and propagation of microcracks due to its high elastic modulus and strong interface contact strength, while the low contact strength of feldspar exacerbates rock stress concentration, leading to a linear positive correlation between the quartz-to-feldspar ratio (QFR) and uniaxial compressive strength (UCS). The geometric heterogeneity of pores and minerals induces local stress concentration in the rock, promoting an increase in microcracks in high-porosity areas and resulting in a nonlinear negative correlation between porosity and UCS. This method overcomes the limitations of traditional modeling associated with oversimplification, achieving a realistic reproduction of rock mesostructures and highlighting the impact of mesoscopic heterogeneity on macroscopic mechanical behavior. It provides a novel tool and insights for revealing the mesoscopic mechanisms underlying the macroscopic mechanical behavior of rock.

response to the challenge posed by the lack of coordinated linkage between the drying and solidification stages in the treatment process of existing flowable sludge, which results in excessively high costs, this study proposes a collaborative implementation method utilizing phosphogypsum-based cementitious materials. This approach seamlessly integrates drying and solidification through the development of phosphogypsum-based powders and aggregates characterized by rapid water absorption and slow-setting hydration. Consequently, a “drying-first, solidification-later” mechanism is achieved, supported by systematic evaluations of mechanical properties, durability, environmental impacts, and both economic and ecological benefits. The results indicate that when 15% phosphogypsum-based powder and 20% phosphogypsum-based aggregates are incorporated into flowable sludge (with a water content of 110%) and cured for 28 days, the unconfined compressive strength (UCS) reaches 4.84 MPa, the cohesion is 355.7 kPa, and the internal friction angle is 30.8°. After undergoing 10 cycles of wetting-drying or freezing-thawing, the UCS remains at 3.6 MPa and 3.3 MPa, with loss rates of 26.3% and 32.5%, respectively, thereby meeting the bearing capacity requirements for solidified soil foundations (≥1.0 MPa). Additionally, the concentrations of soluble phosphorus and soluble fluorine comply with Class II surface water standards, ensuring environmental safety. Mechanistic analysis reveals that phosphogypsum-based aggregates, containing 80% phosphogypsum, form an aggregate-like structure through rapid local water absorption and slow hydration, optimizing particle gradation and facilitating skeleton construction. Meanwhile, the phosphogypsum-based powder, containing 15% phosphogypsum, generates needle-rod-shaped ettringite (AFt) and network-shaped calcium silicate hydrate (C-S-H) gel via the same rapid water adsorption and slow hydration mechanism, playing a crucial role in cementation. The synergy between these two components achieves temporal separation and efficiency integration in the “drying-solidification” process, ultimately realizing the technical goal of “one-time implementation, two-stage completion.” Economic and environmental analyses reveal that the unit strength cost of this method is only 63% of that of traditional cement-based solidification, while carbon emissions are reduced by 94%. This innovative approach demonstrates significant economic and ecological advantages, making it highly valuable for widespread application.

The investigation of the water retention characteristics of Gaomiaozi (GMZ) bentonite is crucial for selecting buffer/backfill materials and designing engineering barrier systems for high-level radioactive waste (HLW) geological disposal. This study examined the water retention properties of GMZ bentonite with varying montmorillonite contents and dry densities. The Nuclear Magnetic Resonance (NMR) technique was utilized to analyze the microscopic characteristics and mechanisms of water distribution in GMZ bentonite under different conditions. The results indicated that the effect of montmorillonite content on the water retention properties of GMZ bentonite was dependent on the controlled suction during the tests. When the controlled suction was below 113.8 MPa, the volumetric change rate and water content of the samples increased gradually with higher montmorillonite content. However, no significant correlation was found between increases in montmorillonite content and volumetric change rate or water content at controlled suctions exceeding 113.8 MPa. Throughout the suction range of 2.7 MPa to 367.8 MPa investigated in this study, the dry density of GMZ bentonite did not significantly influence its water retention characteristics. The water retention curves of GMZ bentonite with varying montmorillonite contents and dry densities were well fitted by the Van Genuchten model. Finally, NMR tests were conducted on samples following the water retention tests, which confirmed the findings obtained from the water retention characteristic tests of GMZ bentonite.

To address the issue of uneven fiber dispersion during the preparation of fiber-reinforced soils, a fluid sedimentation method was proposed. Fiber-reinforced soil specimens were prepared using both the fluid sedimentation method (referred to as the “fluid method”) and the traditional compaction method (referred to as the “traditional method”), with variations in fiber types (polypropylene, PP/polylactic acid, PLA), fiber lengths (5 mm and 10 mm), and fiber contents (0.2% to 1.0%). Unconfined compressive strength tests, slaking tests, cracking tests, and scanning electron microscopy (SEM) analyses were conducted to investigate the effects of the sample preparation method and fiber parameters on the mechanical and water stability properties of fiber-reinforced soils. The test results indicate that the fluid method disperses fibers uniformly in a three-dimensional isotropic manner through slurry adsorption. The unconfined compressive strength of the fiber-reinforced soil prepared using the fluid method is significantly higher than that of the soil prepared by the traditional method. Additionally, the PLA fiber-reinforced soil produced by the fluid method demonstrates a markedly enhanced capacity for plastic deformation. In the slaking test, the specimens prepared using the fluid method exhibit a significantly lower disintegration rate, attributed to the strong bonding of soil particles and the network confinement provided by the fibers. Notably, the PLA fiber specimens prepared by the fluid method display a “disintegration without dispersion” phenomenon. The cracking test results reveal that the PLA fiber-reinforced soil prepared by the fluid method exhibits excellent crack inhibition performance, with optimal results observed at a fiber length of 5 mm and a fiber content of 1.0%. Microscopic analysis shows that the fluid method results in finer soil particles and smaller pore volumes, with both PP and PLA fibers evenly distributed within the soil matrix. Through comparative analysis, it is confirmed that the fluid sedimentation method offers significant advantages in terms of fiber dispersion and performance enhancement. These findings provide a novel and effective solution for the preparation and engineering application of fiber-reinforced soils.

Surface faults are commonly filled with fault gouge that contains crushed rock and clay, whose mechanical properties significantly influence fault stability. This type of fault gouge exhibits clay-like creep behavior under stress, while its mechanical strength is generally greater than that of pure clay due to the presence of rock fragments. To accurately describe the creep process of this fault gouge, in-situ samples from a hydropower station were selected as the research subject. The clay mineral composition was analyzed and determined through X-ray diffraction and scanning electron microscopy tests. Based on the Boltzmann linear superposition principle, creep shear tests were conducted on the fault gouge. The total strain was decomposed into three components: linear viscoelastic strain, linear viscoplastic strain, and nonlinear viscoplastic strain. Corresponding constitutive relationships were established for each component, leading to the development of a novel composite constitutive model based on a generalized viscoelastic body, an Abel dashpot, and a power-law empirical model. The model parameters were determined and validated using experimental data, and the accuracy of the model was compared with that of the traditional H-K model. The results indicate that: (1) the primary clay mineral component of the fault gouge in the study area is montmorillonite, and the creep process of the gravel-containing fault gouge exhibits typical nonlinear characteristics; (2) after establishing the composite model and determining all its parameters, comparative validation with the traditional H-K model demonstrates that the proposed model can more accurately capture the sharp increase in deformation rate during the accelerated creep stage, confirming its good applicability and superiority.