Physics-Informed Neural Networks (PINNs) have shown considerable potential in solving both forward and inverse problems governed by partial differential equations (PDEs) for a wide range of practical applications. This study leverages PINNs for modeling nonlinear large-strain consolidation of soft soil, including creep behavior. The inherent material and geometric nonlinearities associated with soft soil consolidation pose challenges for PINNs, including precision and computational efficiency. To address these issues, we introduce self-adaptive physics-informed neural networks (SA-PINNs), featuring an adaptive loss function weighting and a slope scaling method for the activation functions. Additionally, a sensitivity analysis exploring the influence of monitoring data on the parameter inversion accuracy is presented. Two engineering case studies are used to benchmark the settlement prediction capabilities of the present SA-PINN method with traditional techniques, demonstrating the superior prediction accuracy and consistency of the SA-PINN approach. The findings highlight the significant potential of SA-PINN in practical geotechnical engineering problems.

The large-strain geometric assumptions and nonlinear compressibility and permeability have significant effects on the consolidation of soft soils with high compressibility. However, analytical solutions for large-strain nonlinear consolidation of soft soils with partially penetrating PVDs have been rarely reported in the literature. A double logarithmic model is adopted to describe the nonlinear compressibility and permeability of soft soils with high compressibility, and a large-strain consolidation model for soft soils with partially penetrating PVDs under the condition that the excess pore water pressure at the interface between the improved and unimproved layers is equal is established based on Gibson's large-strain consolidation theory. The analytical solution for the large-strain nonlinear consolidation model for soft soils with partially penetrating PVDs is obtained. The reliability of the analytical solution obtained in this study is verified by comparing it with the existing solutions under different conditions, and the maximum deviation between the two methods does not exceed 5 %. On this basis, consolidation behaviors of soft soils with partially penetrating PVDs under different conditions were analyzed by extensive calculations. Finally, the proposed analytical solution for the large strain consolidation model is applied to the settlement calculation of the Bachiem Highway Project, which further demonstrates the applicability of the consolidation model.

A system of vacuum preloading combined with partially penetrating prefabricated vertical drains (PP-PVDs) is an effective solution for promoting the consolidation of the dredged marine clay. However, a significant and traditionally challenging-to-predict amount of deformation or settlement occurs. Therefore, it is necessary to introduce a three-dimensional large-strain consolidation model to consider the length of the vertical drain to determine the consolidation time and degree of consolidation (DoC), and associated settlement. The predictions using the proposed analytical model provide fair agreements with the field data and those in the literature. Parametric studies reveal that to achieve a 90% DoC within 100 days in soft soil, the optimal penetration depth for PP-PVDs would be 0.7 times the depth of the soil layer. With the increase of DoC, the ratio of excess pore water pressure to applied vacuum pressure (u/P) in the whole soil layer moves toward 1.0. With the increase of PVD penetration ratio (H1/H), more DoC is required to dissipate the excess pore water pressure in the top improved soil layer.