This study investigates the degradation mechanism of shear strength in overlying loess within the sliding zones of loess-red bed landslides governed by base-cover interfaces in the plateau regions of Northwest China. The Chigou, Luoyugou, and Shuiyanzhai landslides in Gansu Province were selected as representative cases. A series of seepage-corrosion tests, direct shear tests, X-ray diffraction (XRD) analysis, ion chromatography (IC), and scanning electron microscopy (SEM) were conducted to examine the evolution of mineral composition, chemical composition, and microstructure during seepage-induced strength deterioration. The main findings are as follows: (1) Both the internal friction angle and cohesion decrease progressively during seepage, with cohesion exhibiting a more pronounced reduction. (2) Intense chemical reactions-primarily hydrolysis, acid-base exchange, and ion exchange-occur during the water-loess interaction, particularly in the early seepage stages, fundamentally altering the material composition and weakening the soil structure. (3) These reactions result in significant microstructural changes, including increases in pore number, porosity, and pore area, which correspond to macroscopic strength deterioration.

The permeability and compressibility of a saturated tailing materials are important parameters in the field of mining safety and geotechnical engineering. The geometric characteristics of a porous medium are key factors in the prediction of its permeability and compressibility. In this paper, the compression and hydraulic characterizes of different gradation tailings through high-stress permeability compression tests were herein investigated. Then, the relationship between the geometric parameters and the high-stress permeability and compressibility of tailings is established. Based on PFC numerical simulation, the non-spherical cluster particles with different fractal dimension and other geometric parameters were constructed, and the compression simulations considering particle breakage were carried out. Using tests and simulations analysis, the influence degree on the compressibility of tailings is as follows: Ultrafine content (F_c) > Roundness (R_c) > Sphericity (S_k) > Fractal dimension (F_d). Numerical simulation proves that particle breakage and water film are closely related to particle shape, and the prediction formula proposed has a wider scope of application on high-stress permeability and compressibility of saturated tailings materials.

To improve the efficiency and safety of shield machines when cutting pile foundation rebars, abrasive waterjet (AWJ) assisted cutting technology has been identified as a feasible solution. However, the cutting performance of AWJ is highly influenced by abrasive concentration. This paper presents a systematic theoretical and experimental study to quantify the effect of abrasive concentration on the cutting performance of rebars and to determine the optimal concentration range. Using an established effective kinetic energy model of abrasive particles, and supported by experimental results, it was found that abrasive concentration significantly influences the cutting depth of rebars. As the concentration increases, the cutting performance of AWJ initially rises rapidly and then gradually declines, with optimal performance occurring at concentrations between 8%-10%. The effect on cutting width is relatively minor, remaining stable in the range of 2.5-3 mm. Furthermore, abrasive particle size plays a critical role: smaller particles (80 mesh) produce deeper grooves, whereas larger particles (24 mesh) lead to a 67%-69% reduction in groove depth. These research findings offer valuable insights and guidance for the design and construction of AWJ assisted shield machines in combined cutting of pile foundation rebars.

Soft rock tunnels under high in-situ stress and complex hydrogeological conditions are highly susceptible to large deformations, posing serious risks to construction safety and efficiency. This study provides a systematic review of current research and development trends in soft rock large deformation, using bibliometric and visualization analysis. The review focuses on deformation mechanisms and corresponding control strategies. Studies show that large deformations are driven by stress redistribution, structural degradation, and environmental influences. These deformations typically evolve in a nonlinear, irreversible, and staged manner. Mechanistic investigations have advanced in key areas such as nonlinear deformation paths, fracture propagation, time-dependent behavior, multifield coupling, and rock-support interaction. In terms of control, yielding supports, energy-absorbing components, high prestress anchors, and intelligent monitoring systems have shown significant effectiveness. Machine learningbased prediction models have also demonstrated potential for deformation identification and early risk warning. Nevertheless, significant limitations remain. Mechanistic analyses are largely macroscopic and phenomenological, and dynamic multi-physical coupling is insufficient. Control strategies lack standardization and long-term validation, while predictive models are constrained by data quality and interpretability. Future work should develop multi-scale models, establish open case repositories, and implement intelligent closed-loop control to enable accurate prediction and active management, enhancing tunnel resilience and sustainability.

This study employed distributed fiber-optic sensors (DFOSs) to examine twin-tunnelling-induced pile bending responses and to identify the primary influencing factors across different tunnelling stages. The high spatial resolution of DFOS measurements enabled detailed analysis of the bending strain energy (U) along the pile. Throughout each tunnel advancement, both U and negative peak bending moment (NPBM) underwent one loading and one unloading process, exhibiting a consistent positive correlation. This investigation introduced two crucial parameters: the negative peak energy density (U_n), derived from NPBM, and the mean strain energy density (U_m), derived from U. The long-term measurements revealed that the U_m/U_n ratio remained within a narrow interval (width < 0.10) throughout most L_pt intervals, allowing a zero-intercept linear function to serve as the development line for normal strain energy concentration. The level of strain energy concentration, which is correlated with risk, are quantifiable via the deviation value of U_n from the development line. In particular, positive and negative deviation values of U_n occurred at the peak and steady states, respectively, corresponding to high and low strain energy concentrations. Additionally, the development of pile-soil interface contact stress was investigated using three-dimensional numerical modelling, providing deeper insight into the loading and unloading mechanisms.

Slope stability analysis, traditionally formulated in two-dimensional (2D) under plane strain conditions, requires three-dimensional (3D) analysis where plain strain condition is violated such as in case of corner failures or variations in the longitudinal direction. This study presents 3D slope stability analysis using a Finite Element Method (FEM) program based on the Strength Reduction Technique (SRT). Stress redistribution is achieved through a visco-plastic algorithm, and the Mohr-Coulomb strength criterion is applied to predict stress states. Slope failure is simulated when iterative calculations show non-convergence, indicating that the equilibrium of the forces could no longer be achieved. The program has been validated by analysing problems, including slopes subjected to earthquake forces and water loading, and the obtained result is compared with the results of existing literature. For ponding water, equivalent nodal loads are derived for slopes discretized using 20-noded brick elements. A novel extrapolation method calculates nodal stresses from sampling points, minimizing fitting errors while preserving the trends. Results, such as deformed meshes, contour plots of visco-plastic strain, yield function, and pore-water pressure, illustrate failure states. Comparison with other studies demonstrates strong agreement, confirming the program's accuracy and robustness in capturing complex slope failure mechanisms.