A three-dimensional particle finite element model for simulating soil flow with elastoplasticity

Area of Science:

• Geotechnical Engineering
• Computational Mechanics
• Earth Surface Processes

Background:

• Soil flow phenomena, including debris flows and landslides, are crucial earth surface processes.
• Modeling these events is challenging due to extreme material property changes and large deformations.
• Existing models often rely on 2D simplifications, which do not fully represent 3D reality.

Purpose of the Study:

• To develop a novel 3D particle finite element method (PFEM) for direct simulation of complex soil flows.
• To overcome the limitations of existing 2D models in capturing the full 3D nature of soil flow.
• To provide a robust computational tool for analyzing large-deformational soil flow problems.

Main Methods:

• Implementation of a fully implicit solution framework based on a generalized Hellinger-Reissner variational principle.
• Utilization of a mixed quadratic-linear element to prevent volumetric locking and ensure accuracy.
• Validation through benchmarks against analytical and independent numerical solutions.

Main Results:

• The developed 3D PFEM allows for large time steps without sacrificing numerical stability.
• The model accurately handles extreme changes in material configuration and properties during soil flow.
• Demonstrated correctness and robustness in simulating large-deformational soil flow problems.

Conclusions:

• The 3D PFEM is a powerful and accurate tool for simulating complex soil flows in three dimensions.
• This method effectively addresses the limitations of 2D simplifications in geotechnical and earth surface process modeling.
• The model can significantly support the assessment of catastrophic soil slope failures and their subsequent runout behaviors.