TY - JOUR

T1 - Peridynamics-based large-deformation simulations for near-fault landslides considering soil uncertainty

AU - Wang, Ruohan

AU - Li, Shaofan

AU - Liu, Yong

AU - Hu, Xuan

AU - Lai, Xin

AU - Beer, Michael

N1 - Funding Information: This research is supported by the National Natural Science Foundation of China (Grant No. U22A20596 ) and the Natural Science Foundation Innovation Group Project of Hubei Province, China (Grant No. 2023AFA017 ). Ruohan Wang has received financial support from China Scholarship Council, China (CSC No. 202206270125 ).

PY - 2024/4

Y1 - 2024/4

N2 - 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.

AB - 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.

KW - Landslide

KW - Large-deformation simulation

KW - Peridynamics

KW - Pulse-like ground motion

KW - Run-out assessment

KW - Spatial variability

UR - http://www.scopus.com/inward/record.url?scp=85183942015&partnerID=8YFLogxK

U2 - 10.1016/j.compgeo.2024.106128

DO - 10.1016/j.compgeo.2024.106128

M3 - Article

AN - SCOPUS:85183942015

VL - 168

JO - Computers and geotechnics

JF - Computers and geotechnics

SN - 0266-352X

M1 - 106128

ER -