This study presents a novel framework for evaluating slope stability in spatially variable soils by integrating a newly developed sequential limit analysis based on the Hellinger-Reissner functional, utilizing the node-based smoothed finite element method (NS-FEM), with a newly proposed deep learning (DL) approach termed multi-downsampling hybrid Linformer-convolutional neural networks (CNNs). The NS-FEM-based mixed formulation of limit analysis (MFLA) enhances computational accuracy and convergence by smoothing strain fields and mitigating numerical discontinuities commonly encountered in standard finite element methods (FEMs). This method generates reliable datasets for stochastic simulations of slope stability under both static and seismic loading conditions. To address the computational expense of specific simulations, we propose the multi-downsampling hybrid Linformer-CNN model, a sophisticated DL architecture that employs dual parallel pathways with distinct downsampling strategies - AveragePpooling1D for medium-scale feature extraction and MaxPooling1D for coarse-scale feature extraction. Each pathway integrates one-dimensional (1D) CNNs for local feature extraction and Linformer-based self-attention mechanisms to efficiently capture global dependencies. The parallel downsampling strategies balance computational efficiency with feature granularity, enabling the model to leverage both local and global data characteristics effectively. The extracted multi-scale features are concatenated and further processed through fully connected networks (FCNs) to accurately predict the factor of safety (FoS) of slopes. Comparative analyses demonstrate that the hybrid Linformer-CNN model outperforms traditional FCN and CNN architectures, achieving robust and precise predictions with a mean absolute percentage error (MAPE) below 10 %. Additionally, the proposed framework significantly reduces computational time, highlighting the potential of integrating NS-FEM-based MFLA with advanced DL architectures for rapid and reliable slope stability assessment in geotechnical engineering.

Sub-level caving (SLC) is a mass mining method suitable for large, steeply dipping orebodies. The particle size distribution (PSD) of blasted material affects material flow through the stope. Improving blast-induced fragmentation can enhance draw point extraction, increasing ore recovery, reducing dilution, and lowering costs in loading and crushing. Numerical simulations using the Mechanistic Blasting Model (MBM) explored these improvements. MBM simulates the explosive loading, rock fracturing, and dynamic explosive gas effects. It addresses uneven explosive distribution from fan-shaped blast holes and complex broken ground conditions. The simulations used Ernest Henry Mine (EHM) data to define the baseline blast design and rock mass and compared field and modelled fragmentation sizes for varying explosive densities and burden sizes. Then, MBM simulations incorporated different rock mass fracture densities, tensile strengths and in-situ stresses, and further blast design changes in the blasthole diameter and charge spacings. A total of 34 scenarios were modelled. Multivariate regression analysis identified key parameters, and new regression models for P20, P50, and P80 passing sizes were developed and validated against the EHM and MBM simulation data. Additional simulations confirmed that while regression predictive models were slightly less accurate, they provided efficient predictions with acceptable accuracy.

Current in-situ stress determination methods are typically conducted inside a drillhole after its creation. However, the drilling process itself is not utilized for measuring in-situ stress or rock strength, despite being a form of direct mechanical testing on the rock mass. Crucially, drilling contains valuable information about in-situ stress and rock strength, as rocks under high compressive stresses exhibit greater strength. This paper presents a novel in-situ stress determination method, supported by the experimental result of rock drilling monitoring tests using a mine hydraulic-rotary drilling machine. Key drilling parameters-including thrust force, rotation speed, torque and drilling speed-are monitored in real time to determine the drilling specific energy per unit volume of rock. A concave-upward relationship between drilling specific energy and rotation speed is identified, which can characterize rock compressive strength and tensile strength with consistent regularity. Further drilling tests are conducted on the same rock samples under varying confining pressures. Results indicate that as confining pressure increases, the concave-upward curve of drilling specific energy shift upward, reflecting enhanced rock strengths due to confinement. The paper outlines the complete methodology for in-situ stress determination using drilling monitoring techniques, bridging the research gaps among drilling monitoring, rock mechanics, and in-situ stress analysis.

The ISRM-suggested Brazilian disc (BD) test using split Hopkinson pressure bar (SHPB) for dynamic rock tensile strength requires central crack initiation and stress equilibrium. This study aims to re-evaluate the critical strain rate, ensuring a valid dynamic Brazilian disc test, and to analyse the reliable dynamic tensile behaviour of granite using high-speed digital image correlation (DIC). The comparison between the measured strain obtained through high-speed DIC analysis and the strain gauge allowed for determining the optimal subset parameters used to obtain the real-time deformation field and the stress-strain curve from DIC data. Crack initiation, crack velocity, and failure process are studied to reveal the rate dependence of granites. A unified dynamic increase factor (DIF) model is proposed for the tensile strength of rocks, and the reason for the sudden drop in DIF for high strain rates is discussed. The results reveal that the upper limit of the valid strain rate, which ensures the validity of the ISRM-suggested dynamic BD test, is co-determined by the conditions of stress equilibrium and crack initiation from the centre of the disc. At higher strain rates (75 s^-1), BD test results fail to capture the actual tensile behaviour of rocks, and the potential factors influencing the critical valid strain rate (CVSr), such as sample radius and boundary crack length, should also be considered.

The mixed rock-ice deposits in high-latitude areas usually come from the accumulation of rock-ice avalanches. Melting tests and temperature-controlled triaxial tests were conducted on rock-ice mixture samples with varying volumetric ice contents (0.25, 0.5, and 0.75), environment temperatures (0.5 ℃, 1 ℃, and 1.5 ℃), and temperature control durations (20 min, 40 min, and 60 min), aiming to investigate the thermodynamic behavior of rock-ice mixed deposits under an ablation environment. Our findings reveal that the melting behavior of rock-ice mixtures mainly occurs in the surface layer; then, ice meltwater transports along the surface seepage path to the bottom of the sample. Notably, the basal meltwater gathering zone leads to accelerated melting of the ice debris, and the cohesion and occlusion between the rock debris in this zone almost disappear, ultimately resulting in severe basal damage. Triaxial test results indicate that the basal damaged zone leads to an easy initial compression process of the rock-ice mixture samples and forms a basal strain effect. Statistics show that ice content, environment temperature, and temperature control duration all show a negative impact on the mixture's peak strength in thaw environments, which also exacerbates the basal strain effect, with a maximum basal strain of 8.61% and a corresponding mass loss ratio of 28.69%. Finally, the mechanisms of the secondary sliding of the mixed deposits and the failure mode of the rock-ice deposit dams induced by ice debris melting were discussed.

Rock mass discontinuities arise from tectonic movements and other geological processes, reflecting the evolution of the Earth's crust. These discontinuities significantly influence the physical properties, deformation characteristics, and energy release mechanisms of the crust. Therefore, recognizing discontinuities is crucial for understanding the evolution of geological structures, analyzing the physical and mechanical properties of geological bodies, and investigating geological hazards. Traditionally, discontinuity recognition has relied on manual interpretation or automated algorithms based on pixel brightness. However, these methods often struggle to strike a balance between efficiency and robustness. To overcome these limitations, we leveraged deep learning techniques that integrate the strengths of both approaches, enabling the recognition of automated discontinuity with expert-level accuracy. To accomplish this objective, we developed and open-sourced the first large-scale deep learning database for rock mass discontinuities, featuring over 300,000 annotated discontinuities. The YOLOv8x-seg model was extensively trained on this database and evaluated across diverse and complex scenarios. The results demonstrated the model's capability to accurately recognize discontinuities even under challenging conditions. Furthermore, we expanded the test set to include rock masses from various global locations, as well as underground rock masses, soils, and artificial structures, where the model consistently achieved effective recognition. The model consistently delivered accurate results, highlighting its strong generalization capability. A comparative analysis revealed that its performance closely aligns with expert manual interpretations. Our open-source database enables researchers to train various deep learning models and achieve equally high-performance results.

Tunnel portal sections have historically been more susceptible to earthquakes than other components, exhibiting significant seismic damage. However, critical seismic behaviors of portal sections remain unrevealed owing to insufficient consideration of actual topography. Moreover, the extent of asymmetric seismic responses induced by topography remains unclear, which is essential for seismic design. To overcome these limitations, this study replicated the actual geological conditions of a tunnel portal section, including the portal slope, topography, slope and tunnel supports, and the often-overlooked portal wall using large-scale shaking table tests. The asymmetric seismic responses and their impact ranges identified in the experiments were validated through numerical simulations. The results revealed that the seismic damage to the slope is attributable to the presence of the tunnel, with slope acceleration near the tunnel portal increasing by 20 %-40 % compared to slopes without a tunnel. Additionally, the tunnel facilitates seismic wave propagation in specific directions, leading to further seismic damage across the portal section. Portal walls, being exposed structures, are susceptible to higher seismic strain and acceleration than tunnel linings and thus warrant increased attention. Importantly, the asymmetric seismic response was found to vary based on different sides and influence ranges. Within 15 m of the portal, the tunnel was dominated by the open-side asymmetric response of acceleration, strain, and displacement. From 15 to 35 m range, the seismic response of the mountainside was more pronounced, exhibiting increased seismic earth pressure and stress on the right sidewall. Seismic earth pressure and stress diminished within the 35-75 m range and steadily decreased beyond 75 m. This enhanced understanding of seismic behaviors facilitates the targeted establishment of future seismic fortifications based on these classified ranges.

Water-rich sand layers are frequently encountered as adverse geological conditions during underground construction. Polymer slurry grouting has been widely recognized as an effective technique for reducing permeability and enhancing the stability of such strata. In this study, a mathematical model is established to describe the diffusion behavior of polymer slurry in porous media under dynamic water conditions and is further validated through laboratory experiments. The theoretical formulation of the slurry permeation process is developed based on Darcy's law, the Hagen-Poiseuille flow principle, and the physicochemical characteristics of the slurry. The derivation primarily focuses on analyzing the dynamic response of the slurry under the influence of water flow, considering the effects of flow velocity, grouting pressure, and sand-layer porosity on diffusion behavior. To verify the proposed model, a visualized grouting simulation system was designed to observe the diffusion process of polymer slurry in water-rich sand layers. The results demonstrate that slurry diffusion is significantly affected by grouting pressure, porosity, and water flow velocity. The observed staged diffusion characteristics, dynamic evolution patterns, and directional effects are in good agreement with theoretical predictions. Furthermore, the average relative deviations between the theoretical and experimental results for diffusion pressure and diffusion distance are both less than 25 %, confirming the reliability of the proposed model. Additionally, this study identifies distinct differences in slurry diffusion between porous and void media. In porous media, slurry propagation encounters greater hydraulic resistance, leading to rapid pressure attenuation and a limited diffusion range. Conversely, diffusion in void media occurs more smoothly due to the continuous cavity structure, resulting in slower pressure decay and a substantially larger diffusion radius. These findings elucidate the mechanisms governing slurry diffusion under dynamic water conditions and provide a theoretical basis for optimizing grouting parameters and improving construction efficiency in water-bearing strata.

Traditional deterministic numerical simulation often has a poor prediction performance for landslide-induced wave run-up (LIWR) hazards, as it neglects the effects of uncertainty. The limitation for efficiently quantifying the uncertainties in primary parameters remains largely unsolved. In this study, we propose a probabilistic evaluation method, integrating the adaptive Kriging (AK) metamodel method and probability density evolution method (PDEM) based on generalized F-discrepancy. A Taylor expansion-based adaptive design strategy is applied to construct the global AK model over representative points generated by generalized F-discrepancy, thereby approximating the numerical physical response (i.e., maximum LIWR). Using these approximate responses, the PDEM is used to compute the exceedance probabilities that LIWR heights exceed elements at risk based on a construction of virtual time, and then a probabilistic criterion is introduced to classify hazard zones. The proposed method is demonstrated via two examples: Example Ⅰ, which possesses risk element (building), and Example Ⅱwith water-level variations. The results indicate that the proposed method has an acceptable performance (showing a 1.7 % difference in exceedance probability compared to Monte Carlo simulation with 50,000 samples) with low computation cost (requiring 284 deterministic analyses). For two specific scenarios in this study, the wave induced by the landslide exhibits a solitary-like leading wave. The proposed probabilistic method provides promising prospects for quantifying LIWR uncertainties, and is helpful for direct, efficient, and low-cost quantification assessment of cascading hazards.

This study investigates the effectiveness of combined tip-and-side post-grouting on large-diameter bored piles in deep fine sand layers. Field tests were conducted on nine piles for the Shishou Yangtze River Highway Bridge project. A detailed comparison of pile performance pre- and post-grouting assessed the technique's influence on ultimate bearing capacity and side resistance. The distribution and effectiveness of the cement grout were analysed using core drilling and the standard penetration test (SPT). An equation correlating post-grouting side resistance with the pre-grouting SPT index (N_SPT) was established. Results demonstrate a substantial improvement in pile bearing capacity after grouting. Ultimate bearing capacity increased by 76 %-152 % after grouting. Longer piles on the main bridge exhibited more pronounced enhancement, achieving ultimate capacities 145 %-206 % higher than those of the shorter approach bridge piles. This is attributed to the greater total cement volume applied along their sides. Critically, combined grouting outperformed side-only grouting, enhancing both side and tip resistance. Core drilling confirmed the spread of cement grout around the piles, confirming the method's effectiveness. SPT results indicated significant increases in the soil N_SPT adjacent to the piles following grouting. These findings provide directly applicable data for designing the bridge pile foundations and offer essential guidance for comparable projects in deep fine sand layers.