Bedrock-soil layer slopes (BSLSs) are widely distributed in nature. The existence of the interface between bedrock and soil layer (IBSL) affects the failure modes of the BSLSs, and the seismic action makes the failure modes more complex. In order to accurately evaluate the safety and its corresponding main failure modes of BSLSs under seismic action, a system reliability method combined with the upper bound limit analysis method and Monte Carlo simulation (MCS) is proposed. Four types of failure modes and their corresponding factors of safety (Fs) were calculated by MATLAB program coding and validated with case in existing literature. The results show that overburden layer soil's strength, the IBSL's strength and geometric characteristic, and seismic action have significant effects on BSLSs' system reliability, failure modes and failure ranges. In addition, as the cohesion of the inclination angle of the IBSL and the horizontal seismic action increase, the failure range of the BSLS gradually approaches the IBSL, which means that the damage range becomes larger. However, with the increase of overburden layer soil's friction angle, IBSL's depth and strength, and vertical seismic actions, the failure range gradually approaches the surface of the BSLS, which means that the failure range becomes smaller.

Cement mixing techniques are widely used to improve the mechanical properties of weak soils in geotechnical engineering. However, due to the influence of various factors such as material properties, mixing conditions, and curing conditions, cement-mixed soil exhibits pronounced spatial variability which is greater than that of natural soil deposits, introducing additional uncertainty into the measurement and evaluation of its unconfined compressive strength. The purpose of this study is to propose a novel framework that integrates image analysis with Bayesian approach to evaluate the unconfined compressive strength of cement-mixed soil. A portable scanner is used to capture high-quality digital images of cement-mixed soil specimens. Mixing Index (MI) is defined to effectively evaluate mixing quality of specimens. An equation describing the relationship between water cement ratio (W/C) and unconfined compressive strength (qu) is introduced to estimate the strength of uniform specimens. To estimate the strength of non-uniform specimens, the equation is developed by integrating MI with the strength of uniform specimens. The coefficients of equations are obtained using Bayesian approach and Markov Chain Monte Carlo (MCMC) method, which effectively estimating the strength of both uniform and non-uniform specimens, with coefficients of determination (R2) of 0.9858 and 0.8745, respectively. For each specimen, a distribution of estimated strength can be obtained rather than a single fixed estimate, providing a more comprehensive understanding of the variability in strength. Bayesian approach robustly quantifies uncertainties, while image analysis serves as a convenient and non-destructive method for strength evaluation, providing accurate method for optimizing the mechanical properties of cement-mixed soil.

The frequent occurrence of extreme rainfall events often triggers levee slope failure (LSF), which, due to the levee effect, significantly damages the roads behind the levee. This paper presents a novel framework for the quantitative risk assessment of roads posed by LSF. Within the framework, the innovative integration of Monte Carlo simulation (MCS) and Material point method (MPM) provides a unique solution for simulating the complicated dynamic relationship between LSF and road destruction. MCS generates precise failure scenarios for MPM simulations, overcoming the limitations of traditional approaches in addressing uncertainty in complex scenario systems. With its technical superiority in capturing post-failure deformations, MPM offers critical insights for assessing road exposure and vulnerability. The framework also accounts for indirect losses from road disruptions, which have long been overlooked. The application of the framework to the risk assessment of the road behind the Shijiao Levee in the Pearl River Basin fully demonstrates its practicality and robustness. Compared to traditional risk assessment methods, the proposed framework provides a more refined dynamic evaluation, facilitating the formulation of more effective disaster mitigation strategies.

Ice and water coexist in frozen soil, and their respective contents (ice content, theta i; unfrozen water content, theta u) are critical factors influencing the mechanical properties of frozen soil. Currently, these two parameters are measured separately. Existing measurement methods require specialized equipment, are time-consuming. To improve measurement efficiency, this paper proposes an inverse analysis surrogate model, which can simultaneously predict both theta i and theta u within one minute. The method process is as follows: 1. A three-dimensional numerical model is established to simulate the transient heat conduction in frozen soil under heat pulse. 2. Six parameters (theta i, theta u, rho s, lambda s, Cs, Gs) need to be determined for each simulation. Through Monte Carlo sampling of six parameters, thousands of numerical simulations are performed. Then, a dataset comprising thermal response curves (TRC) labeled with (theta i, theta u, rho s, lambda s, Cs, Gs) is established. 3. A machine learning algorithm is used, where TRC and soil property parameters serve as inputs, and (theta i, theta u) as outputs. 4. In the laboratory, soil property parameters are measured, and in the field, TRC within one minute of frozen soil is measured in real-time. By inputting soil property parameters and TRC into the machine learning model, (theta i, theta u) can be obtained in real-time.The method was validated through laboratory experiments. The results show that with TRC and rho s, lambda s, Cs as inputs, mean absolute errors (MAE) for theta i and theta u were 2.3 % and 3.1 %, respectively. The proposed method significantly improves measurement efficiency, allowing for the simultaneous measurement of theta i and theta u within one minute.

For offshore platforms installed in seismically active regions, maintaining the safety of operations is an important concern. Therefore, the reliability of these structures, under earthquake ground motions, should be evaluated accurately. In this study, reliability methods are applied to determine the probability of failure of jacket platforms against extreme level earthquake (ELE), considering uncertainties in ground motions and the properties of the structure and soil. They are verified by two variance reduction Monte Carlo sampling methods to find the most efficient method in terms of both accuracy and calculation time. During the ELE event, also called strength level earthquake, structural members and foundation components are permitted to sustain localised and limited nonlinear behaviour, so a force-based criterion is utilized for the limit-state function. The results indicate that all reliability methods, except for FOSM, provide a good approximation of the probability of failure. Also, Point-fitting SORM is the most efficient method.

The structural characteristics of soil-rock mixture (SRM) slopes, including the content, shape, size, and spatial distribution of rock blocks, can significantly influence their failure mechanisms and factor of safety (FOS). Defining the structural characteristics of SRM slopes for stability analysis remains challenging. This study proposes a method for establishing random models and evaluating the statistical properties of the FOS values of SRM slopes. Accordingly, the SRM slope models were constructed by considering the random properties of the shape, size, and spatial distribution of rock blocks in the slope domain. A slope failure criterion based on energy changes and the combined subroutines of USDFLD and URDFIL was implemented in the ABAQUS finite element software to determine the FOS values of the SRM slopes. Monte Carlo simulations were performed to assess the statistical properties of the FOS for random SRM slopes varying rock block properties. The results indicated that when the rock block content was greater than 30%, the stability of SRM slopes considerably increased. For a rock block content of 40%, the effect of rock block size on the SRM slope stability followed two different trends: the mean FOS value tended to decline and subsequently increased as rock block size increased. However, this trend was not observed on SRM slopes with a 30% rock block content. Besides, the dispersion of the FOS values gradually increased with increasing rock content and rock block size. Furthermore, the soil-rock interface strength affected the stability and failure mechanism of SRM slopes. These findings enhance comprehension of the SRM slope stability assessment and demonstrate improved accuracy in predicting and mitigating damage.

In order to investigate the impact of plant root systems on the stability of loess shallow slope, this study conducted plant morphology investigations and direct soil shear tests to analyse the morphological characteristics of alfalfa and the shear characteristics of alfalfa root-loess composites under different soil bulk densities and soil moisture saturation levels. Additionally, the reinforcing effect of the alfalfa root system on the reliability of loess slopes was assessed using the Monte Carlo method. Slope reliability analysis refers to the estimation of the probability of slope failure under specific conditions. The results showed that plant weight and root weight both decreased following an exponential function with increasing soil bulk density. Root weight had a positively linear correlation with plant weight. The cohesion and internal friction angle of both loess samples without roots and with roots increased with increasing soil bulk density. The cohesion and internal friction angle of the two kinds of samples could decreased at less and more than 30% soil moisture saturation. The cohesion and internal friction angle of the root-soil composites were significantly higher than those of the rootless soil. The decrease of soil bulk density and the increase of soil moisture could increase the difference of the two mechanical parameters between the two kinds of samples. Assuming the thickness of the landslide body was 0.3 m, the failure probability of loess slopes covered with alfalfa significantly decreased from 34.97 to 14.51% compared to slopes without vegetation cover. Alfalfa roots significantly increased the reliability of the loess slopes in stability.

The inclusion of calcite precipitates (CaCO3) in soft soil can improve the mechanical properties. Understanding the variability in sand stiffness due to heterogeneous precipitates is crucial for stiffness evaluation and prediction. A novel discrete element-Monte Carlo (DE-MC) method was proposed to quantify the sand stiffness variability induced by stochastic distributions of calcite precipitates, specifically focusing on shear wave velocity (Vs) as an indicator of soil stiffness. A total of 1972 samples were constructed to simulate stochastic spatial distributions of calcite precipitates. Through joint stochastic analysis, the preferential paths formed by calcite clusters were identified as significant contributors to Vs variability. The normalized connectivity per unity distance contact weight (Cd,n) exhibited the most correlated relation with Vs. Two weight selection methods were applicable for using Cd,n to characterize and predict Vs. The results suggest that the DE-MC method has the potential to assess the variability in sand stiffness quantitatively.

On 18 December 2023, a Ms 6.2 earthquake struck the Jishishan area in Northwest China, located at the border of the Qinghai-Tibet and Loess Plateau. The earthquake triggered shallow loess landslides, small rock failures, and soil cracks, mainly along hilly gullies and cut slopes at the edges of terraced fields. A rare large-scale flowslide also occurred in irrigated farmland. These seismic landslides and collapses blocked roads, buried farmland, damaged houses, and resulted in many casualties. Field investigations revealed that these geological hazards were concentrated around cultivated land. Consequently, cultivated land was introduced as an engineering geological zoning factor into the seismic geological hazard risk assessment for Jishishan area. The Newmark cumulative displacement model was refined by incorporating lithological uncertainties via the Monte Carlo method. Comparative analysis of coseismic geohazards with and without considering cultivated land suggests that, in loess-covered areas with cultivation activities, the consideration of the disturbed characteristics of soils provides a more accurate probabilistic risk assessment of seismic geohazards. Human cultivation and irrigation activities affect the physical properties of surface soil, the terraced fields around earthquake prone areas have a risk of earthquake-induced geological hazards. This study may offer valuable insights for hazard prevention and mitigation in high fortification intensity loess covered areas.

Accurate estimation of soil properties is crucial for reliability-based design in engineering practices. Conventional empirical equations and prevalent data-driven models rarely consider uncertainty quantification in both measurement and modelling processes. This study tailors three uncertainty quantification methods including Bayesian learning, Markov chain Monte Carlo and ensemble learning into data-driven modelling, in which support vector regression is selected as the baseline algorithm. The compression index of clay is adopted as an example for model training and testing. In this context, Bayesian learning and Markov chain quantify uncertainty by considering the distribution of function and hyper-parameters, respectively, while different sampled data are employed to explore model uncertainty. These models are evaluated in terms of accuracy, reliability and cost-effectiveness and also compared with Gaussian process regression, etc. The results reveal that based on built-in structural risk minimization, sparse solution and uncertainty quantification, developed models can capture more accurate and reliable correlations from actual measured data over other methods. Their practicability and generalization ability are also verified on a new creep index database. The proposed probabilistic methods are also compiled into a user-friendly platform, showing a significant potential to enrich the data-driven modelling framework and be applied in other geotechnical properties.