[Objective] The particle size distribution of fine-grained loess is a key indicator for evaluating its engineering mechanical properties. The traditional hydrometer method is limited by theoretical assumptions, experimental procedures, and dispersion effects, leading to insufficient measurement accuracy and long testing periods, which makes it difficult to meet the requirements for precise determination of particle size distribution of fine-grained soil. To address these issues, this study takes fine-grained loess from Lanzhou, Gansu Province, as the research object, and proposes a wet sieving test method suitable for fine-grained soil by modifying standard test sieves and using nylon filter cloth as a fine sieve medium. The study systematically investigates the determination of key test parameters and micromorphology. [Methods] After modifying the standard test sieves, multiple comparative tests were conducted focusing on key indicators including soil mass, water volume, test duration, and soil mass loss rate. Scanning electron microscopy (SEM) was employed to observe the micromorphology of particles in different size groups after separation. On this basis, comparative tests between the wet sieving method and the hydrometer method were conducted under conditions of full gradation and single-particle-size groups. [Results] The test results of key indicators determined the optimal test parameters as 30 g of soil and 4 000 mL of water. The duration of a single test was approximately 12 h, with the soil mass loss rate stably controlled within 3%. In contrast, the hydrometer method required approximately 48 h, indicating a fourfold improvement in efficiency. SEM results showed that the wet sieving method could achieve physical separation of fine-grained loess particles. The short-axis dimension of particles was identified as the key parameter controlling the sieving process. A high proportion of particles in the corresponding size groups after separation verified the reliability of the method. Comparative test results showed that the contents of clay and colloidal particles measured by the wet sieving method were significantly higher than those obtained by the hydrometer method, indicating that the traditional hydrometer method markedly underestimated the content of particles smaller than 0.005 mm in fine-grained loess. Further tests on single-particle-size groups confirmed that the hydrometer method exhibited large deviations in particle size determination for fine fractions, whereas the proposed method could directly and accurately reflect the particle size distribution. [Conclusion] This study indicates that the modified wet sieving method based on nylon filter cloth can accurately determine the particle size distribution of fine-grained loess and effectively compensate for the limitations of the traditional hydrometer method in fine particle testing. This method provides reliable technical support and experimental basis for precise geotechnical testing, engineering disaster prevention, and structural safety assessment in loess regions.

[Objective] Among numerical simulation methods for blasting fragmentation, the continuous medium simulation method has high efficiency, but its mechanical mechanisms are not rigorous and errors are significant when dealing with discontinuous problems; the discontinuous deformation analysis (DDA) method performs well for discontinuous problems, but when the fragment size becomes too small, excessively long computation time and non-convergence are likely to occur. This study aims to propose a numerical simulation method for blasting fragmentation that considers both computational efficiency and mechanical rationality. [Methods] Field tests were conducted to reveal the formation characteristics of blasting fragmentation, and the necessity of selecting appropriate numerical simulation methods for different fragment size ranges of blasting fragmentation was clarified. A continuous-discontinuous numerical simulation method for blasting fragmentation based on LS-DYNA+DDA coupling was proposed. In the near-field region of the blast hole, a continuous medium numerical simulation based on LS-DYNA was used to improve the computational efficiency of the crushing zone. In the middle- and far-field regions, a discontinuous method based on DDA was used to achieve discontinuous characterization of blasting fragmentation. The accuracy of using stress and velocity components as coupling parameters was compared. Finally, the LS-DYNA+DDA coupling method was validated based on the mining and blasting practice of Zhoushan Green Petrochemical Mine. [Results] Through field experiments and numerical simulation, it was determined that small-sized fragments were mainly concentrated within a very small range near the blast hole. The continuous medium method could efficiently simulate the distribution of small-sized fragments while ensuring accuracy. It was more reasonable to use DDA method to simulate the fragmentation of medium- and large-sized fragments. Using peak velocity as the coupling parameter between different methods could reduce the pressure loss during computation transmission. [Conclusion] Based on the measured results, comparison and validation between existing numerical simulation methods and the proposed LS-DYNA+DDA coupling method show that the proposed method improves the accuracy of blasting fragmentation prediction and has advantages in balancing the mechanical rationality of fragmentation mechanisms and computational efficiency. However, this method is currently applied in limited engineering scenarios, and its prediction efficiency needs further summary and optimization for different lithologies and blasting parameters.

[Objective] Research on predicting the hydraulic conductivity of cohesive soils is relatively lacking. The classical Kozeny-Carman equation provides an effective method for estimating the hydraulic conductivity of coarse-grained soils, but it performs poorly in predicting the hydraulic conductivity of cohesive soils. This study aims to improve the Kozeny-Carman equation and establish a method for calculating the hydraulic conductivity of cohesive soils. [Methods] We first constructed a relationship between bound water content and liquid limit (LL) in cohesive soils using statistical methods based on their correlation analysis. With this relationship as a bridge, we established a correlation between the total void ratio and the effective void ratio of the soil. Accordingly, the Kozeny-Carman equation was modified to develop a method for calculating the hydraulic conductivity of cohesive soils. Considering that parameter C in the modified equation was difficult to obtain in engineering practice, we developed a calculation model for specific surface area of cohesive soils, incorporating bound water, free water, and soil particles, in order to establish an engineering-friendly equation. A semi-empirical equation relating the specific surface area (S_s) of soil particles to liquid limit was derived, leading to a formula that calculated parameter C based on specific surface area. Data of 105 cohesive soils from published literature were employed to calculate hydraulic conductivity using both the original and modified equations, and the results were compared with measured values. After predicting the saturated hydraulic conductivity of cohesive soils using the improved model, we further evaluated the model’s predictive performance using two error metrics: Mean Absolute Error (MAE) and Root Mean Square Error (RMSE). Subsequently, the sensitivity of each input parameter was analyzed using the cosine amplitude method. Finally, the influence of the main clay mineral types and the order of magnitude of the measured hydraulic conductivity values on the model’s predictive performance was analyzed. [Results] (1) As the water content increased in cohesive soils, the hydration of clay minerals proceeded sequentially through tightly bound water, loosely bound water, and free water phases. A strong linear correlation existed between the ineffective void ratio and the logarithm of liquid limit (lgLL), with a coefficient of determination (R²) of 0.98. A discernible linear correlation was observed between the reciprocal of specific surface area (1/S_s) and the reciprocal of liquid limit (1/LL), with R²=0.83. Parameter C in the Kozeny-Carman equation exhibited a power-law relationship with soil specific surface area, with R²=0.85. The prediction reliability of the classical Kozeny-Carman equation was 56.2%, while that of the improved equation achieved 81.9%, representing a 25.7% improvement in accuracy. However, predictions exhibited divergence, primarily due to the heterogeneity of the experimental data sources, the error propagation from the indirect estimation of specific surface area data, and the fact that the improved formula relied solely on void ratio and liquid limit, potentially neglecting factors like particle size distribution and pore channel tortuosity. (2) Sensitivity analysis revealed that both void ratio and liquid limit were the primary parameters affecting the prediction accuracy of hydraulic conductivity. The model’s performance metrics for the database were MAE=0.29 and RMSE=0.36. For kaolinite-dominated clay, prediction reliability reached 72.4% (MAE=0.29, RMSE=0.38); that of montmorillonite-dominated clay achieved 94.4% (MAE=0.30, RMSE=0.32); and that of illite-dominated clay showed 77.8% (MAE=0.35, RMSE=0.38). Overall, the type of clay mineral had little influence on model performance. When the measured hydraulic conductivity value was within the 10^-9 m/s order of magnitude, the prediction reliability was 88.2% (MAE=0.24, RMSE=0.29);when it was within the 10^-10 m/s order of magnitude,the prediction reliability was 93.3% (MAE=0.20,RMSE=0.25);when it was within the 10^-11 m/s order of magnitude, the prediction reliability was 65.9% (MAE=0.42,RMSE=0.47). [Conclusion] These results show that the prediction reliability of hydraulic conductivity at the 10^-11 m/s order of magnitude is significantly lower than at the 10^-9 and 10^-10 m/s order of magnitude, with the errors and divergence much higher for the 10^-11 m/s order of magnitude. Therefore, the magnitude of hydraulic conductivity has a great impact on model performance, and the model has better applicability for predictions within the 10^-9 to 10^-10 m/s order of magnitude. The modified Kozeny-Carman equation proposed in this study provides a reliable theoretical reference for estimating the hydraulic conductivity of cohesive soils in geotechnical engineering practice.

[Objective] This study aims to investigate the mechanical softening behavior of cataclastic granite under different in-situ stress conditions through true triaxial compression tests, thereby providing an experimental basis and theoretical references for the safe construction and long-term stability assessment of deep soft rock engineering. [Methods] Using the self-developed TAXW-5000 multi-field coupled true triaxial simulation system, true triaxial compression tests on cataclastic granite under different minimum principal stresses (σ₃) were conducted for the first time. The effects of the minimum principal stress on the mechanical response, post-peak softening behavior, and shear-dilation characteristics were systematically analyzed.Furthermore, in combination with 3D-CT scanning technology, the internal crack structures of the failed specimens were extracted and reconstructed, revealing the spatial distribution characteristics of the cracks. The damage evolution patterns of cataclastic granite under different in-situ stress environments were discussed. [Results] As the minimum principal stress (σ₃) increased from 1 MPa to 10 MPa, the peak stress of the cataclastic granite rose from 48.98 MPa to 80.42 MPa, and the residual stress increased from 27.17 MPa to 75.67 MPa. Meanwhile, the softening modulus decreased from 25.67 GPa to 4.81 GPa. During the softening stage, the shear-dilation coefficient decreased from 1.37 to 0.21 with increasing σ₃. In the residual stage, the shear-dilation coefficient first increased and then decreased with higher σ₃, reaching a maximum value of 1.21 at σ₃=5 MPa. CT scanning results indicated that a lower σ₃ led to a greater number and larger apertures of internal cracks in the failed specimens. The damage factor, calculated based on crack statistics, decreased from 0.29 to 0.12 as σ₃ increased from 1 MPa to 10 MPa. [Conclusion] With an increase in the minimum principal stress (σ₃), the softening modulus of cataclastic granite shows a negative correlation with σ₃. The shear-dilation coefficient in the residual stage is relatively high overall and exhibits a trend of first increasing and then decreasing with rising σ₃. In contrast, the shear-dilation coefficient in the softening stage decreases significantly, indicating a transition in the rock deformation mechanism from brittle to ductile failure. A higher σ₃ environment leads to more pronounced softening behavior in cataclastic granite, resulting in a lower degree of crack development and lower damage levels after failure, thereby enhancing the overall engineering stability.

[Objective] The non-filter membrane straw drainage body (NSD) offers significant advantages in vacuum preloading treatment of dredged sludge, such as eliminating the need for filter membranes and preventing clogging, making it a promising solution for practical applications. However, the permeability characteristics and pore structure of the straw filter layer are not yet well understood, which limits its widespread adoption. [Methods] Laboratory permeability tests were conducted to investigate the variation in the permeability coefficient of the non-filter membrane straw drainage body with vacuum preloading time, filter layer thickness, and the initial water content of the dredged sludge. CT scanning was also used to further explore the influence of pore structure evolution under different treatment conditions on the permeability characteristics. [Results] (1) Initially, the NSD contained vertically and horizontally interconnected fissure drainage channels, and these channels formed a continuous seepage network through the connection of pores, which endowed the NSD with superior permeability. The voids occupied 31.42% of the total volume of the NSD, and the volume of fissure structures was 14 times greater than that of the pore structures. (2) The permeability performance of the straw filter layer was superior to that of conventional geotextile filter membranes. The permeability coefficient of the NSD-Reverse (NSD-R) filtration system decreased rapidly and then stabilized with increasing vacuum preloading time, ultimately reaching a stable value on the order of 10^-5 cm/s after 30 minutes, which was one order of magnitude higher than that of the reverse filtration system using geotextile filter membranes for clay (around 10^-6 cm/s). An increase in the filter layer thickness led to a decrease in the permeability coefficient of the filtration system, while a higher initial water content of the dredged sludge corresponded to a larger permeability coefficient of the NSD-R filtration system. (3) As vacuum preloading time increased, the pore structure of the straw filter layer in the NSD-R filtration system evolved. With the increase in preloading time, the porosity of the NSD filter layer decreased rapidly, with the proportion of centimeter-scale fissure structures significantly reduced, effectively blocking the continuous migration of fine particles. Subsequently, the proportion of millimeter-scale pore structures increased, enhancing soil retention while ensuring water permeability, thereby achieving a balance between soil retention and water permeability. [Conclusion] These findings provide theoretical support for the engineering application of NSD in environmentally sustainable stabilization of dredged sludge.

[Objective] The soil water characteristic curve (SWCC) can effectively reflect the relationship between soil water suction (matric potential) and water content, yet conventional testing methods pose challenges for the calibration of in situ soil hydraulic parameters. A more convenient, economical and reliable in situ testing approach is required for the rapid evaluation of in situ soil water status and hydraulic characteristics. [Methods] This study integrated soil dielectric theory with a soil water characteristic curve model and, by considering clay mineral composition, bound water, and free water, derived and established an evaluation model for the soil equivalent dielectric constant-matric suction characteristic curve (SEDCC). For model verification, remolded red clay samples with an average dry density of 1.2 g/cm³ were prepared from six regions of Yunnan Province. The prediction accuracy of the SEDCC curves for red clay across different regions was then systematically analyzed, and the feasibility of indirectly evaluating the SWCC of red clay using dielectric theory parameters of electromagnetic waves was further discussed. [Results] The results showed that the water retention characteristics of red clay were influenced by formation conditions, mineral composition, pore structure, and climatic environment, resulting in regional differences in the fitting accuracy of the Fredlund model. Volumetric water content was the key factor driving changes in matric suction and the equivalent dielectric constant. Using volumetric water content as a bridge, the equivalent dielectric constant gradually decreased as matric suction increased. By fitting the correlation between the equivalent dielectric constant and matric suction using the SEDCC model, the fitting accuracy of the SEDCC model in all regions was greater than 95%. An approximately symmetric relationship was observed between the matric suction inversion curve of red clay and the dielectric theory prediction curve. However, this relationship was still influenced by regional differences. By comparing the predicted values from the dielectric theory inversion curve with the measured values obtained by the filter paper method, it was found that the change trends of the soil water characteristic curves derived from the two were basically consistent, with an average relative error of 6.42%. [Conclusion] Therefore, the SEDCC curve proposed based on dielectric theory can fully reflect the coupling relationship between the soil equivalent dielectric constant and matric suction, indicating that the equivalent dielectric constant is a feasible indirect indicator for evaluating matric suction in red clay. Compared with traditional methods, this model is efficient, convenient, and nondestructive, and it shows potential for indirectly inverting the soil water characteristic curve using dielectric theory, thereby providing a convenient approach as well as theoretical model support and reference for the rapid and accurate detection of soil hydraulic parameters using electromagnetic wave techniques.

[Objective] The integrated support structure of precast piles with dentiform curtains can effectively reduce the earth pressure acting on the support structure and enhance its overall performance. This study aims to quantitatively analyze the load-reduction characteristics of the dentiform walls in this structure, reveal the key mechanisms of load reduction, and provide a theoretical reference for the design and calculation of this novel composite support technology. [Methods] Based on the load-reduction mechanisms of the dentiform walls in an integrated support system of precast piles with curtains, a cycloidal slip surface conforming to the actual sliding failure mode of the soil behind the walls was introduced. A calculation method for the load-reduction effect of the dentiform walls under soil sliding conditions was developed using the horizontal thin-layer differential theory. Based on the validation through calculation examples, the impact of key parameters—including dentiform wall structural parameters (spacing, width, thickness), soil strength parameters (cohesion, internal friction angle), and soil-wall interface strength parameters (soil-wall cohesion, soil-wall external friction angle)—on the load-reduction effect of earth pressure was further analyzed. [Results] When the dentiform wall spacing decreased from 3 m to 1.2 m, the load-reduction ratio increased from 23.77% to 44.02%. When the dentiform wall width increased from 0.5 m to 3 m, the load-reduction ratio increased from 17.16% to 46.34%. When the dentiform wall thickness increased from 0.4 m to 1.0 m, the load-reduction ratio only increased from 32.66% to 36.66%. When soil cohesion (c) increased from 1 kPa to 16 kPa, the load-reduction ratio increased from 28.14% to 59.86%. When the soil internal friction angle increased from 16° to 31°, the load-reduction ratio increased from 20.43% to 42.68%. When the soil-wall interface cohesion (c₁) increased from 0.6 kPa to 9.6 kPa, the load-reduction ratio increased from 27.41% to 50.41%. When the soil-wall interface external friction angle increased from 9.6° to 18.6°, the load-reduction ratio increased from 30.33% to 36.21%, with the rate of increase gradually slowing down. [Conclusion] The developed calculation method effectively solves the problem that earth pressure is difficult to quantitatively analyze when the integrated support system of precast piles with curtains has dentiform walls. The dentiform walls exhibit a significant load-reduction effect, with the load-reduction ratio of earth pressure from the sliding soil behind the walls reaching over 50%. Decreasing the dentiform wall spacing and increasing its width can both significantly enhance the load-reduction effect. Since the walls have a certain width, the contact area between the back of the dentiform walls and the sliding soil mass is limited, and increasing the wall thickness does not lead to a notable improvement in the load-reduction effect. Increases in both the cohesion and internal friction angle of the soil behind the walls can significantly raise the load-reduction ratio of the dentiform walls. Therefore, on sites with better soil conditions, adding dentiform walls is more beneficial for improving the performance of integrated support of precast piles with curtains. The increase in the strength of the soil-wall interface also contributes to a higher load-reduction ratio of the dentiform walls, but the increase in soil-wall cohesion has a relatively more significant impact.

[Objective] To investigate the drag reduction effect of residual layers in periodic debris flows, we quantitatively revealed the amplification effect of the impact of subsequent waves due to the presence of residual layers. We propose evaluation indicators centered on the impact force ratio F^* (amplification of subsequent wave impact relative to the first wave) and the momentum ratio R (amplification of total momentum flux relative to the head momentum flux), aiming to elucidate the dynamic mechanisms responsible for the increased destructiveness of subsequent waves. [Methods] Using a mesoscale flume, we examined debris flows with three different solid contents (40%, 50%, and 60%). Sensors measured flow height, normal stress, shear stress, and pore water pressure. Considering that the movement of subsequent waves over residual layers can be approximated as quasi-steady hydraulic jumps, we calculated unit width impact forces from post-jump velocities and water depths based on momentum conservation. Dimensionless indicators F^* and R were constructed to quantify the amplification effects of multi-wave impacts and intra-wave momentum due to residual layers. [Results] (1) Higher solid content resulted in thicker residual layers, which tended toward a quasi-equilibrium state of erosion and deposition under multiple wave actions. (2) The presence of residual layers altered the flow regime of subsequent waves, with the initial wave typically exhibiting the highest Froude number (Fr). Subsequent waves showed an overall decrease in Fr, indicating a shift from inertia-dominated to gravity-dominated flow control. Increased solid content significantly reduced liquefaction and mobility, as indicated by decreased liquefaction degree λ with increasing solid content, suggesting enhanced effective stress and reduced fluidity. This change influenced the interaction intensity between subsequent waves and residual layers. (3) Residual layers significantly amplified the impact forces of subsequent waves, showing an “increase then stabilize” trend. Under all solid contents tested, impact forces generally exhibited a “rapid increase followed by stabilization”, consistent with the thickening and stabilization process of residual layers. Impact force ratios F^* were greater than 1 for subsequent waves, indicating amplified peak impact forces under identical release conditions. (4) Momentum analysis revealed that R stabilized in later waves, aligning with the trend of F^*. As R>1 indicated total momentum flux exceeding head momentum flux, the share of momentum carried by thicker residual layers drove stronger impacts of subsequent waves, especially pronounced under higher solid contents and thicker residual layers. [Conclusion] (1) In periodic debris flows, the formation and stabilization of residual layers constitute the primary processes leading to enhanced destructiveness of subsequent waves. Even if release conditions are identical for each wave, significant increases in impact loads can occur due to residual layer influences. (2) The indicator system F^* (for comparing external manifestations across waves) and R (characterizing internal momentum distribution within a single debris flow wave) provides a concise assessment framework with clear physical meanings: residual layers amplify impacts by contributing “hidden momentum”, thereby influencing destructiveness amplification. (3) Engineering practices focusing solely on the first wave’s impact for protective structure verification may underestimate the destructiveness amplification effects of multi-wave events. It is recommended to consider the influence of residual layers in designing check dam scales, dam heights, and safety factors.

[Objective] This study aims to propose a new grouting technology to effectively solve the anti-seepage problem of fully-strongly weathered granite strata by optimizing grouting technology and material mix proportion, and to investigate the anti-seepage effects of frequency-pressure grouting technology in such strata. [Methods] An anti-seepage project of fully-strongly weathered granite strata in the upper reservoir of a pumped-storage power station under construction was taken as the research object. An equilateral triangle hole layout was adopted, and five groups of grouting tests (A, B, C, D, E) were conducted sequentially in Ⅰ, Ⅱ, and Ⅲ holes. The influences of grouting technologies (constant-pressure grouting and frequency-pressure grouting), grouting hole spacing (60, 120, 180 cm), and grouting materials (pure cement slurry and cement-bentonite mixed slurry with five water-to-cement ratios of 5∶1, 3∶1, 2∶1, 1∶1, and 0.5∶1) on the anti-seepage effects of grouting in such strata were compared and analyzed. The grouting effects were further evaluated using permeability tests, single-hole ultrasonic tests, single-hole shear wave velocity tests, and borehole color television tests. [Results] Field grouting trials demonstrated that frequency-pressure grouting technology significantly optimized the grouting effect of fully-strongly weathered granite strata through dynamic pressure adaptation and precise flow control, and it was superior to traditional constant-pressure grouting in terms of permeability coefficient control, material cost efficiency, and fracture filling integrity. Cement-bentonite mixed slurry had better controllability than pure cement slurry, especially in addressing leakage during grouting, and it also greatly reduced the large grout consumption caused by grout leakage and the extended construction period caused by multiple waiting times for grout to set. For fully-strongly weathered granite strata with low strength and loose soil, water pressure tests could not be completed, and only water injection tests were conducted to assess permeability before and after grouting. Single-hole shear wave velocity tests could reflect the density of strata before and after grouting to some extent. Grouting spacing in the range of 60-180 cm had no significant effect on the grouting outcomes, whereas the use of frequency-pressure grouting technology effectively improved the compactness of strata and notably increased shear wave velocity. After frequency-pressure grouting treatment, the permeability coefficient of fully-strongly weathered granite strata decreased from the order of magnitude of 10^-3 to 10^-5, indicating a great improvement in anti-seepage capacity. This demonstrated the feasibility of applying frequency-pressure grouting technology for anti-seepage treatment of fully-strongly weathered granite strata, and it could partially or completely replace the traditional cut-off wall scheme, thereby simplifying anti-seepage treatment, reducing construction cost, and minimizing strata excavation. [Conclusion] Compared with traditional grouting methods, frequency-pressure grouting can smoothly adjust grouting pressure and inflow rate according to the characteristics of the grouted strata and real-time grouting feedback, which can avoid excessive fracturing of low-strength strata, uncontrolled grout diffusion, and excessive grout consumption. This technology can be applied to anti-seepage treatment of fully-strongly weathered granite strata. Combined with cement-bentonite mixed slurry, this technology can effectively solve problems such as grout leakage and excessive grout consumption encountered with traditional grouting methods.

[Objective] This study focuses on water content as the key controlling factor to clarify the time-dependent patterns of thixotropic strength recovery of Zhanjiang Formation structural clay under different initial water contents. The microscopic mechanism is interpreted through three pathways: pore structure evolution, particle reorganization, and water action. The findings are expected to provide experimental evidence and theoretical support for predicting strength recovery and evaluating the stability of thixotropic clay foundations. [Methods] Remolded Zhanjiang Formation structural clay specimens were prepared and subjected to a 150-day thixotropy test. Specimens at different thixotropic durations were investigated using macroscopic and microscopic tests. For macromechanical testing, unconfined compressive strength (UCS) tests were conducted on cylindrical specimens. Direct shear tests were conducted on ring-knife specimens to obtain UCS, cohesion (c), and internal friction angle (φ), which were used to evaluate thixotropic evolution. A thixotropic strength ratio was defined as A_t = m_t/m₀, and two indicators—A_t(q) (based on UCS) and A_t(τ) (based on cohesion)—were used to compare recovery characteristics among different strength parameters. For microstructure, fabric evolution was observed using an SEM. Pore parameters, including porosity (M) and abundance (C), were extracted to quantitatively analyze pore structure evolution. Particle parameters, namely probability entropy (H) and distribution fractal dimension (D), were used to quantitatively characterize particle orientation/orderliness and aggregation degree, respectively. [Results] (1) Stage-dependent recovery: Both UCS and cohesion (c) increased with thixotropic duration and showed two stages: a rapid and significant recovery phase during 0-30 d, followed by a slower, stable phase during 30-150 d. The increment during 100-150 d was small, indicating near-stabilization, after which the test was terminated. (2) Dual effect of water content: At the same thixotropic duration, UCS generally decreased with increasing water content, reflecting weakened particle contacts and bonding and thus reduced instantaneous strength. However, higher water content resulted in a faster strength recovery rate, especially at early stage, indicating that water promoted the kinetics of self-adaptive structural adjustment during thixotropic process. (3) Indicator-dependent differences: Cohesion exhibited a higher thixotropic strength ratio and faster recovery within 1 d, suggesting that shearing promoted directional particle alignment and optimized the friction-bonding interface, making c more sensitive to structural rebuilding than UCS. (4) Coordinated micro-parameter evolution: As thixotropic duration increased, M and c decreased continuously. Pores shifted from “large and numerous inter-aggregate pores” to “small and fewer intra-aggregate pores”, while the overall pore shapes remained mainly quasi-equant but became denser. Additionally, H and D decreased synchronously, indicating enhanced particle orientation/orderliness and increased aggregation. These changes were most significant within the first 30 d, consistent with the rapid macroscopic recovery stage. SEM observations revealed a transition from an “open flocculated-dispersed” fabric to a “closed flocculated-aggregated” fabric. Pores between and within aggregates decreased, while particle contacts and continuity of force-transfer paths improved, thereby supporting strength recovery. [Conclusion] The thixotropic strength recovery of Zhanjiang Formation structural clay exhibits distinct time-stage characteristics and strong sensitivity to water content. Recovery generally progresses through a rapid phase (0-30 d) and a stable phase (30-150 d). Higher water content reduces the strength level but significantly accelerates the strength recovery rate. Cohesion exhibits a higher thixotropic strength ratio than UCS because shear-induced particle orientation facilitates more effective structural reconstruction. Microscopically, synchronous decreases in M/C and H/D indicate pore reduction, particle ordering, and aggregation densification. Water enhances particle activity by altering relative particle positions and expanding migration pathways, thereby accelerating self-adaptive adjustment and strength recovery during thixotropic process. Innovations included: (1) parallel comparison of UCS and rapid direct shear parameters within a single thixotropic framework, revealing the cohesion recovery advantage caused by shear-induced particle orientation; (2) linking the macroscopic two-stage recovery pattern with the coordinated evolution of M, c, H, and D, forming an evidence chain of “structural rearrangement—aggregation densification—strength recovery”; and (3) demonstrating that higher water content, while reducing instantaneous strength, accelerates recovery by enhancing particle mobility/activity.

[Objective] The accurate prediction of pipe roof deformation is critical for ensuring construction safety in ultra-shallow buried tunnels. Existing analytical methods frequently oversimplify the complex interaction mechanisms between the pipe roof and surrounding soil, particularly neglecting the load transfer mechanism, stress release effects during excavation, disturbance-induced soil weakening, and the time-dependent behavior of initial support systems. This study aims to develop a comprehensive theoretical framework that integrates these multiple effects into a unified model. The primary objectives include: establishing a vertical load equation that incorporates both the soil arching effect at the tunnel crown and the circumferential micro-arching effect between pipes; utilizing the Pasternak elastic foundation beam theory to simulate soil-structure interaction more accurately; introducing variable subgrade coefficients and a load release coefficient to represent stress redistribution and excavation disturbances; and ultimately formulating a reliable method for predicting pipe roof deformation under realistic construction conditions. The proposed model seeks to provide a practical and theoretically sound tool for design optimization and risk mitigation in pipe-roofed tunnel projects. [Methods] The research methodology combined theoretical derivation, numerical discretization, and empirical validation. First, the vertical load acting on the pipe roof was calculated by considering dual arching effects: the soil arching above the tunnel crown, modeled based on Terzaghi’s trap-door theory with inclined slip surfaces, and the micro-arching between adjacent pipes, with the load distribution derived assuming a parabolic arch axis between pipe contact points. The pipe roof was then modeled as an Euler-Bernoulli beam resting on a Pasternak elastic foundation, accounting for shear interaction between adjacent soil springs, offering a significant improvement over traditional Winkler-based models. To capture the construction-phase effects, the longitudinal span of the pipe roof was divided into five distinct zones: a fully enclosed support zone, an unenclosed support zone, an unsupported zone, a plastically disturbed zone, and an elastically disturbed zone, each with specific definitions of subgrade modulus and stress release rate. The governing differential equation was discretized using the finite difference method, with virtual nodes introduced to handle boundary conditions, and solved programmatically using MATLAB. The model was calibrated and validated against field monitoring data from a real-world ultra-shallow tunnel project, with additional comparative analysis against existing analytical models to demonstrate its superior performance. A detailed parametric study was conducted to evaluate the influence of the subgrade coefficient in front of the face, the excavation advance length, and the length of the unenclosed support segment. [Results] Validation against field data showed excellent agreement, with the predicted maximum deflection of 22.1 mm differing by only 5% from the measured value of 23.2 mm, confirming the model’s accuracy. The deformation curve generated by the proposed method was wider and smoother than those from existing theories, more accurately reflecting the continuous beam behavior of the pipe roof and aligning closely with monitoring results. Parametric analysis revealed that increasing the subgrade coefficient (k₀) of the soil in front of the excavation face from 10 MPa/m to 90 MPa/m significantly reduced the maximum deformation from 33 mm to 17 mm, although the marginal benefit diminished beyond 90 MPa/m. In contrast, the excavation advance length (s) had an exponential impact on deformation. Increasing s from 1.0 m to 3.0 m caused the maximum deflection to approach 160 mm, far exceeding the typical control limit of 20 mm and severely threatening face stability. The length of the unenclosed support segment (b) was found to have a negligible effect on deformation. Furthermore, the load transfer capacity of the pipe roof was observed to be highly sensitive to all three parameters under high overburden ratios (H/B). Excessive increases in k₀, s, or b under these conditions led to a significant transfer of load onto the soil in front of the face, increasing the risk of face instability. [Conclusion] This study successfully develops a multi-effect coupled analytical method for predicting pipe roof deformation in ultra-shallow buried tunnels. The integration of the soil arching effect, micro-arching effect, stress release, excavation disturbance, and support delay into a single model provides a more realistic and accurate representation of mechanical behavior than previously available methods. It emphasizes the effectiveness of improving the soil modulus in front of the tunnel face and strictly controlling the excavation advance length to manage deformation, while indicating that minimizing the unenclosed support length has limited benefits. However, the current study does not consider shear forces between differential elements, soil stiffness hardening, or small-strain behavior. Future research should incorporate these aspects to further enhance the model’s comprehensiveness and accuracy for a wider range of geotechnical conditions.

[Objective] Existing grouting theories and engineering experience are mostly based on hydrostatic or weakly flowing water conditions, making it difficult to accurately describe the diffusion and evolution characteristics of grout under dynamic water-flow environments. It is necessary to systematically reveal the diffusion mechanisms of grout in water-flowing fractures so as to provide theoretical support for grouting design under complex water inrush conditions in water-sealed caverns. In response to engineering conditions involving stable flowing water in a single fracture, this study aims to: (1) reveal the influence mechanisms of fracture geometric characteristics and construction parameters on grout diffusion behavior; (2) quantitatively analyze the controlling effects of key factors on grout diffusion distance, diffusion time, and sealing efficiency; and (3) clarify the relative importance of different influencing factors in the grouting of water-flowing fractures, thereby providing a basis for optimization of grouting parameters and construction decision-making for water-sealed caverns. [Methods] Based on the grout-water two-phase flow theory, a numerical model of grouting in a single fracture with flowing water was established using the finite element method. Variations in water flow velocity within the fracture and the driving effect of grouting pressure were comprehensively considered, and the diffusion, advection, and deposition processes of grout within the fracture were simulated. Through parametric comparative analysis, the effects of fracture aperture, fracture inclination, grouting pressure, flowing water velocity, and fracture boundary extent on the evolution of grout diffusion were systematically investigated. On this basis, a sealing efficiency index was introduced to comprehensively evaluate the grouting performance under different working conditions. [Results] Under flowing water conditions, the grout diffusion pattern, stabilization time, and final sealing performance within fractures were jointly controlled by multiple coupled factors. (1) Fracture inclination had a significant inhibiting effect on grout diffusion. As the fracture inclination increased, the coupling between the gravitational component and the flowing water direction was enhanced, causing the grout to more easily deviate along the down-dip direction. As a result, the ability of grout to migrate against the water flow was weakened, and the diffusion range was markedly restricted. (2) The time required for grout diffusion to reach a stable state increased significantly with an increase in the extent of the fracture domain, because a larger boundary extent provided a greater seepage space for grout diffusion. In contrast, increasing grouting pressure effectively accelerated the advance of the grout diffusion front and shortened the stabilization time, exhibiting a pronounced accelerating effect. (3) In terms of diffusion distance, the effective diffusion distance of grout was inversely proportional to fracture boundary extent and flowing water velocity. Higher flowing water velocity resulted in stronger scouring and transport effects on the grout, thereby reducing its retention capacity within the fracture. Conversely, increases in fracture aperture and grouting pressure facilitated the grout in overcoming water flow resistance, enabling longer diffusion distances and more sufficient fracture filling. (4) Comparative analysis of the influence degrees of various factors indicated that fracture aperture had the most significant effect on grout sealing efficiency, followed in descending order by flowing water velocity, grouting pressure, and fracture boundary extent. This demonstrated that fracture geometric characteristics and hydrodynamic conditions were the key factors controlling the success or failure of grouting under flowing water conditions. [Conclusion] Overall, in strong flowing water environments, relying solely on increasing grouting pressure does not significantly improve grouting performance, and comprehensive design must be carried out by jointly considering fracture aperture characteristics and groundwater hydrodynamic conditions. For areas with larger fracture apertures and higher flowing water velocities, measures such as staged grouting or advance water reduction should be preferentially adopted to enhance grout retention and sealing capacity within fractures.

[Objective] This study aims to investigate the influence of super-large-diameter double-circular pipe jacking construction on the settlement deformation of underground pipelines. Taking the water quality assurance project of Tiegang-Shiyan Reservoir in Shenzhen as the background, and based on the modified Peck formula, this study uses a combination of numerical simulation and field monitoring to systematically analyze the effects of multiple factors such as pipeline burial depth, material, pipe diameter, pipe jacking spacing, and spatial position. [Methods] The variable normalization method was used to analyze the influence degree of each factor on the pipeline, and the safety performance of the pipeline was evaluated. [Results] When the jacking pipes vertically crossed under the pipeline, the induced settlement range was the smallest, indicating a relatively reasonable construction method. When the pipeline burial depth, material, or pipe diameter was changed, the stratum displacement field ultimately showed a “V”-shaped distribution. However, when the spacing between the two jacking pipes increased to 1.5 times the jacking pipe diameter (i.e., 6 m), the displacement field shape transformed into a “W”-shaped pattern, and the influence range of pipeline deformation significantly expanded. Sensitivity analysis showed that the spacing between the two jacking pipes was the most significant factor affecting pipeline settlement (sensitivity=0.54), while pipeline diameter had the least influence (sensitivity=0.06), and pipeline burial depth had a moderate influence (sensitivity=0.40). Furthermore, the safety state of the sewage pipeline was evaluated using the allowable joint rotation angle. The calculated joint rotation angle under field monitoring conditions was 0.54°, which was lower than the standard control value of 1.15°, indicating that the pipeline joints remained in a safe state during construction and did not suffer damage due to uneven settlement. [Conclusion] Currently, there is considerable research on settlement deformation of underground pipelines caused by single-line pipe jacking construction, but research on the influence of super-large-diameter double-line pipe jacking with shallow burial depth is limited. This study clarifies the influencing mechanisms of key construction parameters, providing theoretical basis and data support for engineering practices involving large-diameter pipe jacking undercrossing existing pipelines.

[Objective] In recent years, the increasing number of buildings constructed on soft ground has made the treatment of soft soil foundations particularly important. Investigating the strength variation characteristics of silt foundations under different types of stabilizing agents and curing ages,as well as exploring a preliminary method for identifying the strength level of non-standard soil samples obtained in the field,is of great practical significance for engineering applications. [Methods] Representative silty soil layers from the Zhongshan area were selected. Cement-only mixing tests were first conducted to optimize cement content (ratio of cement mass to the mass of treated wet soil). Subsequently,silt was stabilized using lignosulfonate acid (LA),triethanolamine (TEA),and alkali-activated sodium silicate (AS) as stabilizing agents,respectively. Scanning electron microscopy (SEM) tests were then carried out to analyze the microstructures and stabilization mechanisms of untreated and stabilized soils. Finally,ArcGIS was used to construct independent elevation models from the SEM images of untreated and stabilized soils and to process them into three-dimensional images. Scatter plots were plotted in double-logarithmic coordinates,and the soil porosity and particle fractal dimension were further calculated and analyzed. [Results] 1) Using the unconfined compressive strength of specimens as the evaluation index,under the same curing age, the cement content was positively correlated with the unconfined compressive strength. Before a curing age of 7 days, the unconfined compressive strength increased rapidly; during 7-14 days, the growth rate slowed down; and from 14 to 28 days, the unconfined compressive strength continued to increase. Considering economic cost and code requirements, the optimal cement content was 18% of the wet soil mass. 2) At certain mixing ratios, single incorporation of LA, TEA, and AS all enhanced the strength of cement-stabilized soil and could be used as stabilizing agents. Based on the optimal contents of the three stabilizers, ternary mixing stabilization tests were conducted. The results showed that single incorporation of AS exhibited better stabilization performance than the other single-additive groups and the ternary mixing group. When the cement content was 18% and the AS content was 0.9%, the unconfined compressive strength of the stabilized soil reached a maximum value of 2.39 MPa. 3) SEM tests indicated that the specimens of the 18S blank group failed to generate sufficient gel-like hydration products (C-S-H) and needle-like Aft crystals. As a result, limited cementitious material existed between soil particles, and numerous pores were observed. After stabilization with 18S-0.9AS, a large amount of gel-like C-S-H hydration products and needle-like Aft crystals were rapidly generated, which interwove to form a large-area spatial network structure and initially formed a skeleton. This process led to particle bonding and aggregation and filled the interparticle pores. Overall, the 18S-0.9AS group exhibited the best stabilization effect. 4) Three-dimensional SEM images of untreated and stabilized soils were constructed and processed using ArcGIS. Data calculation and analysis showed that the untreated soil had 97 663 226.34 pore pixels, accounting for 53.22% of the total image pixels (porosity), with a fractal dimension of 1.399 8. The 18S-0.9AS stabilized soil had 50 153 642.75 pore pixels, accounting for 27.27% of the total image pixels (porosity), decreased by 25.95 percentage points compared with the untreated soil. The fractal dimension of this group was 1.853 5, and the unconfined compressive strength reached the maximum value of 2.39 MPa. [Conclusion] 1) The larger the particle fractal dimension of a specimen, the more complex the particle structure, the higher the surface roughness, the lower the porosity, and the higher the compressive strength. 2) The porosity of specimens and the compressive strength exhibit a nonlinear decreasing relationship, whereas the particle fractal dimension and the compressive strength exhibit a nonlinear increasing relationship. 3) When standard specimens cannot be obtained at construction sites, the compressive strength of specimens can be preliminarily inferred by using the porosity and particle fractal dimension of non-standard specimens, based on the physical significance and correlations of fractal dimensions among different specimens.

[Objective] Microwave-assisted rock breaking technology shows promising application potential in hard rock tunneling. The effectiveness of microwave irradiation on rocks is significantly influenced by their microstructure, particularly the mineral particle size distribution. Existing studies mostly focus on single mineral components or simple binary combinations, whereas systematic investigations into how the heterogeneity of complex mineral particle size distributions in natural rocks affects the microwave rock breaking remain limited. This study aims to quantify the heterogeneity of mineral particle size distribution through experiments and numerical simulations, and to reveal its influence mechanisms on the thermal-physical response, damage evolution, and mechanical property degradation of rocks. [Methods] Seven groups of standard granite specimens (Φ50 mm × 100 mm) with different mineral particle size heterogeneity coefficients were selected and subjected to microwave irradiation tests. Surface temperature variations of the specimens during irradiation were monitored, longitudinal wave velocities before and after irradiation were measured, and the peak strengths were obtained through uniaxial compression tests. Image processing techniques were used to extract the surface mineral distributions of the specimens, and quantitative indicators characterizing the heterogeneity of particle size distributions were defined and calculated. Using COMSOL software, a multi-physics numerical model coupling electromagnetic fields, heat conduction, and solid mechanics was established. The model precisely reconstructed the real mineral distributions with different heterogeneity coefficients, and simulated and analyzed the dynamic evolution process of the temperature fields, stress fields, and plastic damage zones of the specimens under microwave irradiation. [Results] (1) Thermal response: under identical irradiation conditions, the heterogeneity of mineral particle size distribution significantly affected the thermal response of the specimens. With the increase of heterogeneity, the temperature rise of the rock became more pronounced. The specimen with the highest heterogeneity (H=0.78) exhibited a final temperature approximately 44 ℃ higher than that of the most homogeneous specimen (H=0.34).(2) Damage and weakening: with increasing heterogeneity coefficient, the number of microwave-induced surface microcracks increased significantly. The reduction in longitudinal wave velocity intensified, with a maximum difference reaching 30%. The uniaxial compressive strength loss rate increased from 11.2% to 29.6%, with a maximum difference of 18.4%. These results indicated that the more heterogeneous the mineral distribution was, the more severe the internal damage induced by microwaves and the more significant the weakening effect on macroscopic mechanical performance became.(3) Mechanism: stronger heterogeneity led to more intense temperature gradients and thermal stress concentrations at the interface between strong microwave-absorbing minerals (such as potassium feldspar) and weak microwave-absorbing minerals (such as quartz). This was because specimens with higher heterogeneity contained larger potassium feldspar particles, which had stronger microwave absorption capacity, resulting in a rapid local temperature rise. Both the area proportion of the critical tensile stress zones (>15 MPa) and the area proportion of the plastic zones increased monotonically with the increase of the heterogeneity coefficient. The plastic zones first appeared at the contact interface between potassium feldspar and quartz and expanded over time. [Conclusion] The heterogeneity of mineral particle size distribution is a key microstructural factor controlling the effectiveness of microwave-assisted rock breaking. The defined quantitative heterogeneity coefficient can effectively predict the outcomes of microwave irradiation: rocks with higher heterogeneity are more likely to experience uneven heat accumulation and large interfacial thermal stresses under microwave irradiation, thereby leading to more extensive microcrack initiation, more significant wave velocity reduction, and more significant strength loss. This study identifies the potassium feldspar-quartz interface as the preferential site for damage initiation.

[Objective] As a new type of support anchor, the Glass Fiber-Reinforced Plastic (GFRP) screw anchor is increasingly used in engineering practice due to its simple construction, light weight, high strength, corrosion resistance, and cost-effectiveness. Previous calculation formulas for uplift bearing capacity of screw anchors mostly based on artificially defined boundaries between shallow and deep embedment, resulting in large differences in critical embedment depth ratio H/D (where H is the embedment depth and D is the anchor plate diameter) given by different scholars. This study aims to propose a calculation formula for uplift bearing capacity based on the generalized unified failure surface morphology, providing theoretical support for the design of GFRP screw anchors in engineering practice. [Methods] Using finite element numerical simulation software ABAQUS, numerical simulations of the uplift performance of GFRP screw anchors under different embedment ratios were conducted. The maximum uplift bearing capacity of vertical GFRP screw anchors in soil and the evolution of the failure surface morphology of anchored soil were investigated. [Results and Conclusion] (1) The load-displacement curves of GFRP screw anchor during uplift in anchored soil could be divided into three stages: the elastic stage, the local plastic stage, and the penetration failure stage. In the elastic stage, the load and displacement showed linear relationship, with the slope of the curve representing the equivalent stiffness of the anchor plate-soil system. In the local plastic stage, plastic deformation occurred in the local soil zone, deformation of the anchored soil gradually increased, system stiffness decreased, load-displacement relationship became non-linear (indicated by a gradually decreasing curve slope), and the interaction between the GFRP screw anchor and the anchored soil began to exhibit non-linear characteristics. In the penetration failure stage, the resistance of the anchored soil reached its ultimate value, cracks in the anchored soil became interconnected, the soil failure surface formed, and the curve entered a stable phase where the load no longer changed with displacement increase. The uplift bearing capacity factor N_γ showed a trend of initial increase followed by gradual stabilization with embedment ratio, which matched well with experimental results from other scholars, thereby verifying the simulation’s validity. The turning point occurred near H/D=9, suggesting that an embedment ratio H/D=9 could be considered as the critical embedment ratio distinguishing shallow from deep embedment. (2) Numerical simulation showed that during uplift, the anchored soil could be divided into three zones based on stress state: conical active zone formed by compaction above the GFRP screw anchor, passive zone extending outward along screw anchor edge influenced by extrusion from soil within failure surface, and transition zone between active and passive zones in plastic state. The soil within the failure surface consistently extruded the soil outside failure surface. Except for the outermost region, which might be in a state of at-rest earth pressure, the soil on the failure surface was generally in a state of passive earth pressure. With increasing embedment ratio, the failure surface morphology of the anchored soil underwent a dynamic evolution from “trumpet” shape➝“goblet” shape➝“light bulb” shape. No distinct “critical embedment ratio” between shallow and deep embedment for GFRP screw anchors existed. (3) By integrating the maximum uplift bearing capacities obtained from numerical simulations and the evolution patterns of the failure morphology in the anchored soil, a generalized unified failure surface model that did not require distinguishing between shallow and deep embedment for the soil anchored by GFRP screw anchors was proposed. This model was derived by introducing a cubic term based on the “trumpet” shape (with inclined linear boundaries) failure surface. On this basis, a calculation formula for the uplift bearing capacity under different embedment depths of GFRP screw anchors was derived, and the formula was compared and validated with numerous experimental results from domestic and international studies. The results demonstrated that the proposed calculation formula could effectively predict the maximum uplift bearing capacity under different embedment depth ratios.

[Objective] To explore the impact of freeze-thaw environments on the strength of guar gum-modified loess, direct shear test, unconfined compression strength test, Brazilian splitting test, and bender element small-strain test are conducted to examine the strength degradation and deterioration characteristics of guar gum-modified loess before and after freeze-thaw cycles. [Methods] In this experiment, cationic guar gum was selected for specimen preparation. The specimens were subjected to freeze-thaw cycles in a sealed environment. In the bender element tests, a single-pulse sine wave was used as the input signal at a frequency of 1 kHz. The initial arrival method was adopted to determine the travel time, from which the wave velocity was calculated. [Results] After freeze-thaw cycles, the tensile strength, uniaxial compressive strength, and shear strength of guar gum-modified loess decreased, but all remained greater than that of untreated loess subjected to freeze-thaw cycles. The strength degradation of guar gum-modified loess after freeze-thaw cycles was influenced by the coupling effects of water content and guar gum content, with water content having a more pronounced effect than guar gum content. Microscopic analysis revealed that guar gum formed filamentous cement by bonding with water molecules, which linked soil particles and restrained their movement. Consequently, the strength of the modified loess was notably enhanced. However, during the freeze-thaw cycles, the simultaneous damage of the above-mentioned cement and loess particles was the main reason for the more pronounced strength degradation of guar gum-modified loess than that of untreated loess. The shear wave velocity of loess modified by guar gum before and after freeze-thaw cycles was measured using bender element tests. The functional relationships of shear wave velocity with compressive strength and tensile strength of loess were established, revealing a good correlation with uniaxial compressive strength. [Conclusion] The research findings systematically reveal the influence of freeze-thaw action on guar gum soil stabilization technology, and a novel method is proposed to evaluate the strength degradation of modified loess with wave velocity. This provides environmentally friendly materials and new insights for soil stabilization in the loess regions of northwestern China, while offering a theoretical basis for the further application of guar gum in geotechnical engineering.

[Objective] This study investigates how soluble salts in clay redistribute under an applied direct-current electric field during electro-osmotic drainage and how the redistribution affects dewatering and consolidation efficiency. The study quantifies the spatiotemporal evolution of salt content through bulk electrical conductivity, distinguishes the individual effects of salinity and water content on conductivity, and infers ion-migration trends and their implications for the combined dewatering and desalination performance of saline clays. [Methods] Laboratory conductivity calibrations were conducted on remolded clay across practical ranges of water content and soluble-salt concentration. Based on these data, an empirical relationship was established that linked soil bulk conductivity to pore-fluid salinity while explicitly incorporating water content, enabling the conversion of measured conductivities into estimates of salt content. Subsequently, a one-dimensional electro-osmotic consolidation test was conducted. Segmented voltage, local conductivity, cumulative drainage, and current were monitored and recorded. Using this calibration, time-lapse conductivity profiles were processed to reconstruct salt-content distributions and their evolution. This method could provide a framework to monitor and interpret coupled ionic transport and water removal during electro-osmosis. [Results] Calibration showed that conductivity increased with both salinity and water content. However, when water content was considered, salinity accounted for a larger share of the variance in bulk conductivity. Accordingly, conductivity served as a reliable in-situ indicator of salt content during electro-osmosis. The electro-osmotic test revealed a distinct zonation of salt content consistent with electromigration toward the cathode. At the anode, salt content declined rapidly during the first 2 hours and then stabilized at approximately 2.0 g/L until the end of the test. In the mid-section, salt content also decreased over the first 2 hours, showing the smallest reduction among the three regions, followed by an increase and subsequent decline. By 6 hours, it temporarily exceeded the initial salinity. This peak reflected the convergence in the middle zone of cation fluxes migrating from anode to cathode and anion fluxes moving in the opposite direction. After 6 hours, the mid-section salinity decreased progressively and, at the end of the test, fell below that of the anode region. The cathode experienced the most pronounced change, showing a continuous decline throughout energization. By 8 hours, the cathodic zone had nearly approached a salt-free state. During electro-osmosis, the soil potential field was strongly modulated by both water content and salinity, producing spatially differentiated potential distributions that evolved over time. Water content and drainage rate exhibited non-uniform dynamics among regions and ultimately formed a moisture gradient of Anode < Middle < Cathode. Salinity exerted a pronounced control on potential pathways and transmission efficiency. Therefore, its evolution should be incorporated explicitly in design to optimize treatment outcomes. The combined evidence demonstrated that electro-osmotic drainage in saline clay could achieve two outcomes simultaneously: accelerated consolidation and effective removal of soluble salts. The latter mitigated adverse effects of high salinity on subsequent construction, including corrosion risk and strength variability, thereby improving the suitability of the treated ground. [Conclusion] This study delineates the migration and distribution patterns of soluble salts in high-salinity clays under electro-osmotic drainage, offering a new perspective for treatment and practical guidance for engineering application. Operationally, a critical point is reached when salinity in the cathodic zone drops to a very low level. Continuing energization beyond this point leads to sharply diminished drainage efficiency and disproportionately increased energy consumption. At the design stage, measuring soil electrical conductivity and conducting pre-tests to characterize the salinity-moisture relationship are recommended, thereby informing the required energization time. In practice, continuous conductivity monitoring provides a comprehensive indicator of overall dewatering progress. Wider adoption of these insights is expected to facilitate broader and more effective application of electro-osmosis in geotechnical engineering.

[Objective] This study systematically investigates the influence of particle shape on the mechanical properties of soil-rock mixtures, with a particular focus on strength and deformation characteristics. It aims to address the current knowledge gap regarding the specific mechanisms through which particle shape affects the mechanical behavior of such materials. By establishing quantitative relationships between shape parameters and mechanical response, the study aims to provide a scientific basis for predicting and controlling the settlement of soil-rock mixture subgrades in engineering practice. [Methods] High-strength α-hemihydrate gypsum powder was utilized to fabricate rock-like particles with controlled shapes. The Brazilian shape parameter (Y) was employed to quantitatively characterize particle morphology. A series of consolidated-drained triaxial tests were conducted using a medium-pressure triaxial apparatus to systematically evaluate the mechanical properties of soil-rock mixtures containing particles with different shape coefficients. The testing program included comprehensive measurements of peak deviatoric stress, internal friction angle, cohesion, and other shear strength parameters under different confining pressures. Microstructural analysis was performed to observe particle breakage patterns and stress transmission mechanisms. [Results] The experimental results revealed significant shape-dependent mechanical behavior. As the particle shape coefficient Y increased, the peak deviatoric stress of the specimens initially increased and then tended to stabilize. With increasing Y, the internal friction angle and initial shear angle φ₀ gradually decreased, while the cohesion exhibited a corresponding increase. The increment of shear angle Δφ showed a non-monotonic trend, first decreasing and then increasing. The shape coefficient was found to alter the stress transmission path within the particle skeleton, leading to preferential breakage of large-sized particles. When Y increased from 0.63 to 0.73, the failure mode transitioned from edge damage to localized rupture, and ultimately to complete fragmentation. Ellipsoidal particles (Y≥0.69) exhibited stress concentration at the long-axis ends, resulting in localized fracture concentrated in the shear band region. In contrast, near-spherical particles demonstrated uniform stress distribution and exhibited surface spalling. [Conclusion] This study successfully establishes the quantitative relationship between particle shape and mechanical properties of soil-rock mixtures, revealing the underlying mechanisms of shape effects. The findings demonstrate that particle shape significantly influences the strength, deformation, and failure characteristics through its control on stress transmission and particle breakage patterns. The study provides a scientific basis for the prediction and design of soil-rock mixture subgrade settlement, offering practical guidance for engineering applications. Future research should focus on extending these findings to field-scale conditions and developing predictive models that incorporate shape effects for improved design accuracy.

[Objective] During the parameter inversion process of groundwater models, frequent calls to the forward model result in excessive computational demand and prolonged processing time, which severely limits their practical applicability. To address the high time consumption in groundwater model inversion, this study proposes a coupled inversion algorithm capable of rapidly identifying unsteady permeability coefficients. [Methods] The proposed algorithm coupled the parameter inversion process with reduced-order model training, where the reduced-order model was employed instead of the original model for parameter inversion calculation, thereby reducing total inversion time. During the iteration process, the sum of squared errors between the reduced-order model calculation values and observations was used as the objective function, and an improved differential evolution algorithm with strong global search capability was employed as the optimization method for parameter inversion. In each iteration, the parameters that best matched the observations were identified as optimal, and snapshots of these optimal parameters were calculated to train the reduced-order model, thereby enhancing the inversion accuracy in the next generation. To enable the reduced-order model to accurately capture the time-domain response characteristics of the original model, a uniform snapshot strategy was employed to collect time-step snapshots. Based on the characteristics of the coupled inversion algorithm, the relative error of the reduced-order solution corresponding to the optimal parameters was calculated at all nodes within the time domain, and iteration was terminated when the maximum error fell below a preset threshold. [Results] Taking a two-dimensional pumping well model as an example, the proposed method was compared with an inversion approach based on the original model. The results indicated that: (1) For unsteady seepage parameter inversion, compared with the optimal time snapshot strategy, using a uniform snapshot collection strategy to construct the reduced-order model could achieve higher computational accuracy, while the reduced-order model had a lower average order. (2) While maintaining inversion accuracy comparable to that of the original model, the proposed algorithm could reduce computational time by approximately 95.37%. (3) Near the optimal parameters, the reduced-order model obtained by the proposed method showed almost identical responses to the original model. However, the error increased significantly when moving away from the optimal solution. (4) The effects of observation error, mesh density, and inversion dimensionality on the inversion accuracy of the proposed algorithm were consistent with those of the original model, but the computational time of the proposed algorithm was less than 5% of that of the original model. (5) The proposed algorithm was less affected by the order of the original model, and the increase in the computational time was proportionally smaller than the increase in model order, indicating higher computational efficiency for high-order models than for low-order ones. (6) Compared with low-dimensional inversion problems, the proposed algorithm exhibited greater time-saving efficiency in handling high-dimensional cases, suggesting stronger robustness against the curse of dimensionality. (7) Under different convergence accuracies, the proposed algorithm could consistently reproduce the results of the original model without a significant increase in computational time even as accuracy improved. [Conclusion] The proposed coupled inversion algorithm in this study, as a deterministic finite element-based inversion framework, innovatively couples the training process of the reduced-order model with the parameter inversion process and significantly improves the computational efficiency of parameter inversion.Characterized by a simple structure,ease of implementation, and no need for posterior error calculation,the algorithm has significant engineering application value and promising prospects for broad application.