Peridynamics-based large-deformation simulations for near-fault landslides considering soil uncertainty
Ruohan Wang,Shaofan Li,Yong Liu,Xuan Hu,Xin Lai,Michael Beer
DOI: https://doi.org/10.1016/j.compgeo.2024.106128
IF: 5.218
2024-02-05
Computers and Geotechnics
Abstract:Landslides are widely acknowledged as among the most prevalent natural disasters. Peridynamics (PD), a mesh-free computational method, offers distinctive advantages in circumventing mesh distortion issues. However, limited attempts to employ PD in landslide simulation. Utilizing the features of non-ordinary state-based peridynamics (NOSBPD), we propose a computational method to analyze the entire process of slope run-out. Moreover, the occurrence and progression of landslides are notably affected by soil strength uncertainties. Hence, a coupling procedure is proposed to integrate random fields with NOSBPD, investigating the impact of spatial variability in soil strength on landslides. Results indicate that considering soil heterogeneity leads to a 12% increase in run-out distance compared to homogeneous soil analyses. This highlights the significance of accounting for soil spatial variability to avoid underestimating landslide run-out distances. Additionally, this study compares the influence of ground motion types containing non-pulse ground motions and pulse-like ground motions (PLGMs) on entire landslide process. The findings suggest that landslides under PLGMs exhibit larger run-out distances and demonstrate a more concentrated spatial distribution, indicating higher susceptibilities to landslides under PLGMs. Lastly, we explored the interaction of two uncertainty sources on landslides. The findings can guide engineers in implementing assessments of potential uncertainties associated with landslides.
computer science, interdisciplinary applications,geosciences, multidisciplinary,engineering, geological