A coupled SPFEM/DEM approach for multiscale modeling of large-deformation geomechanical problems

Computational modeling in geotechnical engineering frequently needs sophisticated constitutive models to describe prismatic behavior of geomaterials subjected to complex loading conditions, and meanwhile faces challenges to tackle large deformation in many geotechnical problems. The study presents a multiscale approach to address both challenges based on a hierarchical coupling of the smoothed particle finite element method (SPFEM) and the discrete element method (DEM) (coined SPFEM/DEM). In the approach, SPFEM serves as a solver for the global boundary value problem, in which the material constitutive responses are derived from the DEM solution of representative volume elements (RVEs) attached to the SPFEM nodes to avoid phenomenological constitutive assumptions. The approach is capable of modeling large deformation because of use of SPFEM, which discretizes the domain with a set of Lagrangian nodes and employs Delaunay triangulation for efficient remeshing on the nodes. In addition, as the RVEs are associated with the nodes due to the nodal integration technique in SPFEM, the interpolation of RVEs from the old mesh to the new one is bypassed, which is otherwise infeasible. The smoothing operation in nodal integration further offers a remedy for regularizing mesh dependency in simulation of strain localization problems. Two examples, namely, general failure of a footing and flow of an unstable slope, are used to demonstrate the potential of the proposed method in solving large deformation and providing reliable predictions on collapse and failure of geotechnical problems.

Automated summary

The study proposes a multiscale approach based on the smoothing particle finite element method (SPFEM) and the discrete element method (DEM) to solve large deformation problems in geotechnical engineering. This approach avoids phenomenological constitutive assumptions, effectively simulates large deformation, and provides reliable predictions for collapse and failure of geotechnical problems.