Loess, predominantly distributed in arid and semi-arid regions of central and western China, exhibits low shear strength and structural instability, rendering it prone to geological hazards such as landslides and collapses, which pose significant threats to local infrastructure and safety. This study evaluated the urease activity of soybean and sword bean at different temperatures to screen the optimal enzyme source for enzyme-induced carbonate precipitation (EICP). Methods including single EICP, EICP combined with nano-SiO₂, and EICP combined with both nano-SiO₂ and soil stabilizer (SS) were adopted to enhance the surface strength of loess. The results showed that the EICP technique significantly improved the surface strength of loess, especially with the addition of nano-SiO₂ and soil stabilizer. This study confirmed that using sword bean urease treated at −20 °C for 24 h in combination with 1.5% nano-SiO₂ was both cost-effective and efficient in reinforcement. The incorporation of 5% soil stabilizer further enhanced the surface strength, and the accuracy was further verified by combining the results of SEM and XRD. Future research will focus on optimizing the material ratio to maximize the improvement of surface strength, providing an economical and feasible solution for rapid loess solidification, and evaluating the long-term durability under cyclic wet and dry conditions.

Sampling of training data is the most important step in active learning slope reliability analysis, which controls the analysis accuracy. In this study, a novel surrogate-assisted normal search particle swarm optimization (SANSPSO) was proposed to enhance the accuracy and robustness of existing methodologies. In SANSPSO, the sampling process was considered a minimum problem with an objective function defined as the absolute value of the performance function. Initiated with a normal search paradigm and supplemented by three algorithm strategies, this approach seeks to preserve the continuity of the solution while refining the algorithm’s efficacy and efficiency. To reduce computation cost, surrogate-assistance was used, in which a surrogate model substitutes the objective function in most iterations. This surrogate model evolves during the iteration process and ultimately replaces the actual performance function within Monte Carlo simulation. Finally, this study presents a comparative study with five state-of-the-art methods across four explicit problems and three engineering cases, where test data suggest that the SANSPSO methodology yields a 20% improvement in accuracy and a 30% rise in stability under different dimensional problems relative to the most efficacious of the alternate methods assessed because of the improved and more consistent prediction of limit state function. These findings substantiate the validity and robustness of the SANSPSO approach.

Geostatic stress field procedure is the first and the most important step for the numerical simulation of geotechnical engineering, which greatly influences the simulation results. Traditional methods often fail when the model is complex. In this paper, based on finite element method (FEM) software ABAQUS, failure reasons of four commonly used methods for the geostatic stress field are studied. According to the analysis results, a new set of methods, which can provide reasonable displacement and stress field results under complex conditions, is proposed. The proposed methods follow the principle that the stress of different materials should be obtained separately to avoid stress distortion. Then, the accuracy and applicability of the proposed method are verified through a comparison study and a specific application. This study provides a theoretical basis for the method of geostatic stress field procedure under complex condition and can serve as a reference for relevant studies.