Acta Geotechnica最新文献

Soil drying–wetting (D–W) cycles, a widespread natural phenomenon, pose significant challenges to numerical modeling due to coupled multi-physical interactions, dynamic crack evolution, and degradation of hydro-mechanical properties. To simulate continuous D–W cycles, this study develops a fully coupled thermo-hygro-mechanical ordinary state-based peridynamic (THM OSB-PD) model. The proposed framework innovatively incorporates a fracture-healing criterion and a reduction function to model drying-induced cracking, wetting-phase healing, and progressive deterioration of mechanical strength during D–W cycles. Using a sequential staggered implicit solution algorithm to ensure convergence and efficiency, the model reproduces experimental observations, such as crack memory, sequential healing, and crack evolution characteristics. Meanwhile, hydraulic hysteresis is captured by embedding dynamic soil–water characteristic curve (SWCC) parameters into moisture diffusion equations. Moreover, comparative analysis of isothermal and non-isothermal D–W cycles reveals differences in crack network morphology, moisture content and matric suction distribution, vertical strain, and SWCCs, underscoring the critical influence of temperature. The results demonstrate that the proposed THM OSB-PD model effectively captures crack propagation and healing during D–W cycles, providing a foundation for investigating soil behavior under hydraulic cycling.

Driving the shield machine at a longitudinal inclination angle is necessary in complex geological conditions, to avoid existing underground structures as well as at entry and exit from the ground surface. This paper investigates the face stability of inclined shield tunnels using a refined 3D Finite Difference Method (FDM) model. The strain-hardening behavior of the muck is captured by the UBCSAND model. The actual mechanical response of the muck is simulated through a two-phase approach, which is subsequently employed to simulate the progressive collapse of the excavation face. The parametric analysis based on the refined FDM model is carried out over a range of effective friction angles, cohesion values and longitudinal inclination angles to evaluate their effects on the excavation face stability. A dimensionless equation for predicting the critical earth chamber pressure is proposed, which shows good agreement with the model test. Finally, a case study on the Ultra Rapid Under Pass (URUP) tunnelling project in Nanjing, China is chosen to verify the proposed 3D FDM model.

The transport characteristics of heavy metals within CCL-based liner systems are influenced by variations in temperature and stress, yet the models capable of accounting for these dual effects are comparatively absent and not well developed. Additionally, all the existing theoretical investigations are proposed based on the Darcy’s law and fail to capture the non-Darcian seepage characteristics under coupled thermo-mechanical loading. To this end, the Hansbo’s seepage law and thermo-osmosis are jointly introduced to establish a new coupled one-dimensional model of heavy-metal transport in CCL for the first time. Current model can comprehensively incorporate the interactions between nonlinear consolidation, heat transfer, and contaminant transport. Subsequently, employing the heating and external loading schemes close to engineering reality, the coupled governing equations and numerical solutions are obtained. Then the precision and dependability of the proposed model are corroborated through degeneration analysis and case studies. Furthermore, an in-depth examination for the influence of several crucial factors is carried out. The results reveal that the presence of temperature difference can significantly accelerate the consolidation rate and promote the transport process. As the Soret and thermo-osmosis coefficients rise, the promotional effect of thermal diffusion and advection is enhanced, which becomes more evident with the increasing final temperature. Moreover, the consolidation and transport behavior are greatly influenced by the non-Darcian seepage, and the neglect of which can result in an underestimation for the service life of CCL. Finally, the growth in the final value of external loading can increase the final settlement and retard the transport process. This study advances the understanding and design of CCL-based liner systems, crucial for environmental safety and protection.

In this paper, a radial finite strain thermal consolidation model is extended to further consider the viscous behavior of soft soils. A newly developed thermal elastic visco-plastic (TEVP) constitutive model for the temperature and time-dependent stress–strain behavior of soft soils is implemented in a radial finite strain thermal consolidation model, which is solved through a numerical solver. This solving method is verified by comparing the calculated results by the numerical solver with two existing studies. After this, a particle swarm optimization (PSO) algorithm is employed to enhance the prediction performance of the proposed consolidation model by automatically calibrating the model parameters during the consolidation process. Two physical model tests were conducted to examine the validity of the proposed model and PSO-assisted method. The results indicate that the proposed radial consolidation model captures the viscous characteristics (including creep) of the time-dependent settlements of soft soils, which could not be simulated by the former reported models. The PSO-assisted method has demonstrated practical applicability when compared with physical model tests. The longer the observation time, the better the predictive performance of our radial finite strain thermal consolidation model. It is recommended that the observation time should not be shorter than the time required for primary consolidation.

Reliable debris-flow susceptibility mapping (DFSM) hinges on both optimal topographic characterization and unbiased data sampling. This study systematically quantifies the effect of four DEM resolutions (6.5–90 m) and sampling strategy uncertainty on advanced deep learning (DL) models, including convolutional neural network (CNN1D, CNN2D), recurrent neural network (RNN), and long short-term memory (LSTM). A field-based comprehensive debris flow inventory (108 debris flow watersheds) and 13 contributing factors derived from multi-resolution DEMs and remote sensing data were utilized to construct the datasets. To assess the impact of sampling on model performance, an extensive uncertainty analysis was conducted through 100 symmetrical sampling iterations (debris flow and non-debris flow), resulting in a 6.7–10.5% improvement in mean accuracy by varying sampling points. Feature importance was further examined using a Random Forest classifier and frequency ratio (FR), identifying rainfall as the most significant factor. Model performance was evaluated through multiple metrics such as accuracy, recall, Jaccard, precision, kappa, AUC, and F1-score. Among the models, LSTM consistently outperformed other compared models across all resolutions, achieving the maximum debris flow susceptibility accuracy of 0.929 and AUC of 0.973 at 12.5 m resolution. The proposed multi-resolution DL framework provides a reproducible pathway for uncertainty-aware debris-flow susceptibility assessment in complex mountainous terrains.

In this study, drainage and imbibition morphologies were quantitatively evaluated during a water retention test using X-ray computed tomography (CT) scanning and image analyses. Drainage and imbibition clusters were extracted by subtracting two CT images with different saturation states after image registration of the CT images, which accounted for the differences in position and size, using affine transformation. By examining the shape complexity and configuration of the drainage and imbibition clusters within the pores, it was revealed that, during the drying process, relatively simple drainage occurs from the pore centers. In contrast, during the wetting process, complex water imbibition occurs from the surface of the soil particles at high suction levels, followed by simpler shape imbibition in the pore centers, filling up the pores at low suction levels. Furthermore, a theoretical water retention curve model, with an assembly of unit cells that can express scanning curves, was constructed by appropriately modeling pore draining and water filling based on the findings obtained from the image analyses.

A good knowledge of subsurface conditions is crucial for underground construction and risk management. When boreholes are sparse, geophysical data can be a valuable supplement for stratigraphic delineation. However, geophysical results can be affected by external factors, and a reliance solely on geophysical and borehole data may introduce errors and uncertainties in geological interpretation. This study proposes an approach that utilizes multi-source geodata and the interpretation enhanced with machine learning (ML) and spatial correlation for geological profiling and uncertainty assessment. In this approach, an iterative interpolation method is first employed to augment the training dataset quantity. The augmented dataset with geophysical and spatial correlation features is then used for geological profiling through stacking ensemble learning. The prediction uncertainty can be assessed using the probability outputs from the ML models involved in stacking ensemble learning. The proposed approach was first validated using a synthetic case and applied subsequently to a site in Singapore. For the synthetic case, the determined geological cross-section achieves a 93.10% accuracy compared to the predefined cross-section. The spatial correspondence between low integrated confidence (IC) values and misclassification areas confirms the IC plot as an effective tool for uncertainty assessment and identification of likely prediction errors. For the site in Singapore, the proposed approach can delineate the boundaries of four geological formations and provide valuable insights into rockhead variations. The discrete cells with low IC values reasonably indicate potential variations of stratigraphic interfaces. Validation using three boreholes shows a prediction accuracy of 87.85%, demonstrating the benefit of multi-source geodata integration in improving geological profiling reliability.

In deep geological engineering, anisotropic shale is widely distributed and frequently subjected to both three-dimensional in situ stress and dynamic loading. Understanding its failure mechanisms under these conditions requires both experimental investigation and theoretical modeling. In this study, combined triaxial quasi-static compression and dynamic impact tests were conducted to examine the mechanical response of shale under varying confining pressures, strain rates, and loading directions. The experimental results provide key insights into the effects of these factors on shale strength and failure modes, which serve as the basis for model development. Building upon these findings, an anisotropic strength criterion and a damage constitutive model are proposed to account for the coupled effects of confining pressure, strain rate, and material anisotropy. The model parameters have clear physical interpretations, reflecting the contributions of different stress components in the material coordinate system to yielding. Validation against experimental data demonstrates that the proposed model can reliably capture the stiffness, strength, and failure behavior of shale under a wide range of loading conditions. This study advances the understanding of anisotropic shale failure mechanisms and offers a theoretical framework applicable to deep geological engineering.