The study deals with reliability analysis of strip foundation on spatially variable c - phi soil. The spatial variability of soil strength parameters, namely cohesion c and friction angle phi is modelled using anisotropic uncorrelated random fields, generated with the Fourier series method. Random finite element limit analysis (RFELA) providing a rigorous lower and upper bound for bearing capacity for individual Monte-Carlo simulations is employed. Additional use of adaptive meshing refinement algorithm leads to a significant reduction of the relative difference between statistical moments of obtained lower and upper bound results. The influence of the horizontal and vertical scales of fluctuation and foundation depths on the mean and standard deviation of the obtained bound moments is investigated. Additionally, the rigorousness of the mean and standard deviation of both considered bounds estimation is checked. As a result of the analysis, a novel approach based on a mixed distribution that combines lower and upper bound moments is introduced. As shown, this approach offers significant benefits by providing conservative and relatively precise measures of reliability which can be obtained in reasonable computation time. The proposed method seems to be adequate for practical engineering reliability analysis of foundation bearing capacity and other limits states problems.

The determination of the factor of safety (FoS) of slopes during seismic excitation can be complex if the relevant effects of pore water pressure accumulation, nonlinear material response and variable shear strength are duly accounted for. A rational two-step approach to tackle this task based on a hydro-mechanically coupled dynamic simulation and finite element limit analyses is henceforth introduced. To ensure accurate transfer of the hydro-mechanical soil state, a mapping concept is presented, accounting for spatial distributions of stresses, excess pore water pressures, inertial forces and shear strength. The proposed approach is compared to limit equilibrium method (LEM) for the case of a large-scale water-saturated open cast mine slope subjected to seismic loading. In comparison with LEM, the new approach to assess seismic slope stability proves to be simpler in its implementation and straightforward, which could be an important asset for practitioners.

The aim of this study is to investigate the influence of rock variability on the failure mechanism and bearing capacity of strip footings. A probabilistic analysis of the bearing capacity of footings on rock masses is conducted in this paper, where random adaptive finite-element limit analysis (RAFELA) with the Hoek-Brown yield criterion and the Monte Carlo simulation technique are combined. The stochastic bearing capacity is computed by considering various parameters, such as the mean values of the uniaxial compressive strength of intact rock, Hoek-Brown strength properties, coefficient of variance, and correlation lengths. In addition to the RAFELA, this study introduces a novel soft-computing approach for potential future applications of bearing capacity prediction by employing a machine learning model called the eXtreme Gradient Boosting (XGBoost) approach. The proposed XGBoost model underwent thorough verification and validation, demonstrating excellent agreement with the numerical results, as evidenced by an impressive R2 value of 99.99%. Furthermore, Shapley's analysis revealed that the specified factor of safety (FoS) has the most significant influence on the probability of failure (PoF), whereas the geological strength index (GSI) has the most significant effect on the random bearing capacity (mu Nran). These findings could be used to enhance engineering computations for strip footings resting on Hoek-Brown rock masses.

In underground space technology, the issue of tunnel stability is a fundamental concern that significantly causes catastrophe. Owing to sedimentation and deposition processes, the strengths of clays are anisotropic, where the magnitudes of undrained shear strengths in the vertical and horizontal directions are different. The anisotropic undrained shear (AUS) model is effective at considering the anisotropy of clayey soils when analyzing geotechnical stability issues. This study aims to assess the stability of rectangular tunnels by adjusting the dimensionless overburden factor, cover-depth ratio, and width-depth ratio in clay with various anisotropic strength ratios. The stability analysis of these tunnels involves employing finite element limit analysis and the AUS model to identify the planes of soil collapse in response to the aforementioned variations. In addition, this study presents the development of soft-computing models utilizing artificial neural networks (ANNs) to forecast the stability of rectangular tunnels across various combinations of input parameters. The findings of this study are presented in the form of design charts, tables, and soft-computing models to facilitate practical applications.