Runout of Submarine Landslide Simulated with Material Point Method

Youkou Dong, Dong Wang, Mark F. Randolph

    Research output: Chapter in Book/Conference paperConference paperpeer-review

    17 Citations (Scopus)


    Most of the present knowledge on submarine landslides relies upon back-analysis of post-failure deposits identified using geophysical techniques. In this paper, the runout of slides on rigid bases is explored using the material point method with focus on the geotechnical aspects of the morphologies. In MPM, the sliding material and bases are discretised into a number of Lagrangian particles, and a background Eulerian mesh is employed to update the state of the particles. The morphologies of the slide can be reproduced by tracking the Lagrangian particles in the dynamic processes. A real case history of a submarine slide is back-analyzed with the MPM and also a depth-averaged method. Runout of the slides from steep slopes to moderate bases are reproduced. Then different combinations of soil and basal parameters are assumed to trigger runout mechanisms of elongation, block sliding and spreading. The runout distances predicted by the MPM match well with those from large deformation finite element analysis for the elongation and block sliding patterns. Horst and grabens are shaped in a spreading pattern. However, the current MPM simulations for materials with high sensitivities are relatively mesh sensitive.

    Original languageEnglish
    Title of host publicationProcedia Engineering
    EditorsA Rohe, K Soga, H Teunissan, B Zuada Coelho
    Place of PublicationNetherlands
    Number of pages8
    Publication statusPublished - 2017
    Event1st International Conference on the Material Point Method: MPM 2017 - Delft, Netherlands, Delft, Netherlands
    Duration: 10 Jan 201713 Jan 2017


    Conference1st International Conference on the Material Point Method
    Internet address


    Dive into the research topics of 'Runout of Submarine Landslide Simulated with Material Point Method'. Together they form a unique fingerprint.

    Cite this