Water–sand inrush disasters are frequently encountered during underground engineering construction in karst terrain. The objective of this paper is to study the failure mechanism and evolution characteristics of water–sand inrush caused by the instability of filling medium in karst cavity, as well as the impacts of soil compactness, hydraulic pressure and confining pressure on the instability process. In response to this purpose, a stress-controlled seepage test apparatus in consideration of particle loss was designed, and a series of seepage tests were performed correspondingly. The test results indicate that: (1) Based on the nonlinear feature analysis of water-outflow pattern, the water–sand inrush process can be divided into the “slow flow” stage, “transition flow” stage and “pipe flow” stage by Transition Point I, II. (2) The decreasing soil compactness and increasing hydraulic pressure both exponentially facilitate the seepage-erosion process by increasing the particle-erosion ability; the increasing confining pressure extends the “slow flow” stage and shortens the duration of the “transition flow” stage, ultimately advancing the occurrence of the “pipe flow” stage; the existence of critical hydraulic pressure for the seepage-erosion progress is confirmed, the occurrence of the “pipe flow” stage is significantly advanced once the hydraulic pressure over the critical value. (3) The particle loss caused by the seepage-erosion process is the internal mechanism of water–sand inrush, the variation characteristics of water-outflow pattern are crucial external manifestations correspondingly. Therefore, with the monitoring of water-outflow pattern variation tendency as indicators, the critical status of water–sand inrush can be near-real-time identified, which offers experimental foundation for the early warning and forecast of the occurrence of water–sand inrush.

The topography and the incident angle of seismic waves both have considerable effects on the seismic ground motions of canyons in a half-space. In this paper, the theory of wavefield decomposition and the artificial boundary is used to develop a method for inputting obliquely incident SV waves. Formulas for the equivalent nodal forces applied to the truncated boundary are derived and implemented in the finite element method. The validity of the proposed method is verified by a test case. A parametric study is then performed to investigate the influence of canyon geometry and incident angle of SV waves on the seismic response of trapezoidal canyons. The numerical results indicate that the canyon inclination has a more significant effect on the ground motion amplification than its height and width. The amplification effects are strongly related to the canyon inclination and the incident angle of SV waves. Additionally, the dominant frequency corresponding to the acceleration of the canyon crests is not sensitive to the incident angle of SV waves.

In this paper, a series of quasi-static cyclic shear tests were performed on artificial rock joint specimens with natural morphologies at different normal stress levels. After shearing, combined with 3D scanning technology and image processing, the 3D morphology parameter, $A_0{\theta }_{max}^{\ast }/\left(C+1\right)$, of 36 analyzed directions was measured to investigate the damage and degradation laws of the joint surfaces. The polar curves of the roughness parameter shrank from a circular to an elliptical shape, with the shear direction as the minor axis. The roughness degradation rates and the shear cycles were mutually interrelated with the normal stress levels, which had a tendency to be fast initially and then to slow down under higher normal stress. Influenced by the cyclic shear direction, the 3D roughness of the joint surfaces degenerated anisotropically and sustained the most serious damage along the shear direction. Two damage and degradation fitting formulas were proposed, which could aid the assessment of residual roughness under different stress histories and the cumulative wear degree.

Western Guangdong Province in China is frequently affected by intense rainfall and related geological hazards, yet studies on debris flow disasters in this area remain limited. In June 2025, four consecutive debris flow events occurred along the Lianzhou-Sanshui section of the G55 Erenhot-Guangzhou Expressway, attracting significant attention. To better understand the spatial distribution and controlling factors of these hazards, this study conducted a debris flow susceptibility assessment for this section. Based on regional topographic, meteorological, and geological conditions, combined with historical disaster data, an evaluation index system and database were established. Four machine learning algorithms, Back Propagation Neural Network (BPNN), Random Forest (RF), Support Vector Machine (SVM), and Extreme Gradient Boosting (XGBoost), were applied and systematically compared. Algorithm performance was assessed using 3-fold spatially independent validation. Among the tested algorithms, XGBoost achieved the best predictive performance, with an accuracy of 0.923, recall of 0.879, F1-score of 0.794, and ROC-AUC of 0.927. The debris flow susceptibility was classified into low, medium, and high levels. Medium and high susceptibility areas exhibited clear spatial clustering, mainly in the northern section of the expressway, while the southern section showed generally low susceptibility. Debris flows along the expressway were primarily triggered by shallow surface collapses and landslides that formed cascading hazard chains, while the slipperiness of surface weathering products showed an observed association with debris-flow occurrence and may represent a potential conditioning factor. These findings provide a scientific basis for debris flow risk identification and mitigation and offer insights for improving susceptibility assessment methodologies.