With the development of data collection and storage capabilities in recent decades, abundant data have been accumulated in geotechnical engineering fields, providing opportunities for the usage of machine learning approaches. Thus, a rising number of scholars are adopting machine learning techniques to settle geotechnical issues. In this paper, the application of three popular machine learning algorithms, support vector machine (SVM), artificial neural network (ANN), and decision tree (DT), as well as other representative algorithms in geotechnical engineering, is reviewed. Meanwhile, the applicability of diverse machine learning algorithms in settling specific geotechnical engineering issues is compared. The main findings are as follows: ANN, SVM, and DT have been widely adopted to solve a variety of geotechnical engineering issues, such as the classification of soil and rock types, predicting the properties of geotechnical materials, etc. Based on the collected relevant research, the performance of random forest (RF) in sorting soil types and assessing landslide susceptibility is satisfying; SVM has high precision in classifying rock types and forecasting rock deformation; and backpropagation ANNs and Hopfield ANNs are recommended to forecast rock compressive strength and soil settlement, respectively.

The dynamic anisotropy and failure mechanism of shales are greatly affected by bedding surfaces. To reveal the influence of beddings on anisotropic characteristics of shales under dynamic impact, the Brazilian splitting tests were conducted by the split Hopkinson pressure bar system. The fracturing process were monitored by the high-speed camera. Meanwhile, to understand crack initiation and propagation mechanism, the stress buildup, stress shadow and stress transfer were modelled based on the digital image processing and the rock failure process analysis method. The effect of dip angle and bedding spacing on crack initiation, propagation and coalescence was analyzed. Simultaneously, the spatial distribution and energy magnitude of crack-induced acoustic emissions were captured numerically. The results show that the shale discs continue to produce parallel cracks and cambered cracks induced by the high stresses at the tips of initial cracks; the tensile strength under dynamic splitting changes in the U-shaped trend with the bedding dip angle increasing; the cracking percentage of bedding surfaces decreases, and the cracking percentage of rock matrix increases with the bedding dip angle increasing. In addition, the acceleration of crack growth and the rapid growth of AE energy can be regarded as the effective precursors of shale failure.

Discrete fractures are pre-existed in rock masses, which have been proved influencing the mechanical behaviours of rock masses. Modeling the network of discrete fractures is a crucial aspect in studying fluid flow within fractured rock masses. A three-dimensional rough discrete fracture network (RDFN3D) model, developed using the three-dimensional fractal Weierstrass-Mandelbrot relationship and compared with the discrete fracture network (DFN3D) model, was discussed in terms of its influence on both 2D and 3D structural models. The methodology for modeling RDFN3D based on MATLAB code was described. The impact of fracture roughness on seepage characteristics was analyzed and compared for both RDFN3D and DFN3D models. Results indicate that rough structures have a significant effect on the seepage behavior in rock masses. Differences between flow velocities of DFN3D and RDFN3D models increase with increasing pressure gradient and fracture width. Fluid flow velocity in rough structures is lower than that in DFN3D models, which becomes increasingly evident with growing flow pressure and joint aperture.

Locked-in stress refers to internal stress present within rock formations that can influence the failure process of rocks under specific conditions. A simplified mechanical model is applied, drawing on elasticity and the hypothesis of locked-in stress, to explore the influence of locked-in stress on the mechanical properties of loaded rocks. An analytical solution is obtained for the stress distribution in a failure model of rocks that include locked-in stress. The findings demonstrate that the geometry and orientation of stress inclusions within the rock influence the initiation and propagation of cracks under the combined influence of locked-in stress and high-stress conditions. Moreover, the presence of locked-in stress substantially reduces the rock’s capacity to withstand maximum stress, thereby increasing its susceptibility to reaching a state of failure. The increase in closure stress leads to a significant increase in the magnitude of the maximum stress drop and radial strain variation within the rock, resulting in reduced strength and a shortened life of the ageing failure of the rock. In addition, the influence of stress inclusions on energy dissipation is investigated, and a novel relationship is established between the roughness coefficient of the rock structure surface and the angle of the rock failure surface. This relationship serves to characterize the linear dynamic strength properties of rock materials containing locked-in stress. This investigation not only advances the comprehension of stress distribution patterns and the effects of locked-in stress on rock failure patterns but also facilitates a more precise portrayal of the nonlinear features of alterations in the rock stress-strain curve under the influence of confined stress. These findings provide a solid theoretical foundation for ensuring the safety of excavations in various deep engineering projects.

Rocks with multi-shaped fractures in engineering activities like mining, underground energy storage, and hydropower construction are often exposed to environments where stress and seepage fields interact, which heightens the uncertainty of instability and failure mechanisms. This has long been a long-standing challenge in the field of rock mechanics. Current research mainly focuses on the mechanical behavior, seepage, and energy evolution characteristics of single-fractured rocks under hydro-mechanical coupling. However, studies on the effects of multi-shaped fractures (such as T-shaped fractures, Y-shaped fractures, etc.) on these characteristics under hydro-mechanical coupling are relatively scarce. This study aims to provide new insights into this field by conducting hydro-mechanical coupling tests on multi-shaped fractured sandstones (single fractures, T-shaped fractures, Y-shaped fractures) with different inclination angles. The results show that hydro-mechanical coupling significantly reduces the peak strength, damage stress, crack initiation stress, and closure stress of fractured sandstone. The permeability jump factor (ξ) demonstrates the permeability enhancement effects of different fracture shapes. The ξ values for single fractures, T-shaped fractures, and Y-shaped fractures are all less than 2, indicating that fracture shape has a relatively minor impact on permeability enhancement. Fracture inclination and shape significantly affect the energy storage capacity of the rock mass, and the release of energy exhibits a nonlinear relationship with fracture propagation. An in-depth analysis of energy evolution characteristics under the influence of fracture shape and inclination reveals the transition pattern of the dominant role of energy competition in the progressive failure process. Microstructural analysis of fractured sandstone shows that elastic energy primarily drives fracture propagation and the elastic deformation of grains, while dissipative energy promotes particle fragmentation, grain boundary sliding, and plastic deformation, leading to severe grain breakage. The study provides important theoretical support for understanding the failure mechanisms of multi-shaped fractured sandstone under hydro-mechanical coupling.