On data benchmarking and verification of discrete granular simulations
Jose Salomon1, Catherine O'Sullivan1, Fernando Patino-Ramirez1
1Department of Civil and Environmental Engineering, Imperial College London, London SW7 2BX, United Kingdom.
Data in Brief
|March 27, 2024
Summary
This study validates the LAMMPS Discrete Element Method (DEM) code using theoretical benchmarks for granular materials. The findings ensure the reliability of DEM simulations for researchers studying material behavior.
Area of Science:
- Computational physics and geomechanics.
- Numerical modeling of granular materials.
Background:
- The Discrete Element Method (DEM) is crucial for granular material research.
- DEM code credibility relies on replicating physical observations and theoretical expressions.
- Theoretical validation of DEM codes is often overlooked, impacting simulation reliability.
Purpose of the Study:
- To present a dataset for validating the open-source DEM code LAMMPS.
- To verify LAMMPS against established theoretical approaches at macro and particle levels.
- To provide benchmark data for new or existing DEM codes.
Main Methods:
- Simulating a face-center-cubic (FCC) array of spheres under shear.
- Analyzing particle motion and forces for a clump rolling down an inclined plane.
- Characterizing stress-strain behavior of Toyoura sand under drained and undrained conditions.
Main Results:
- LAMMPS successfully captured macroscopic granular behavior in simulations.
- Particle-level calculations of motion, forces, and torques were verified.
- Stress-strain responses for Toyoura sand under different shearing conditions were accurately reproduced.
Conclusions:
- The presented dataset and validation approach confirm LAMMPS's capability for granular material simulations.
- This work enhances confidence in using LAMMPS for complex DEM research.
- The public dataset facilitates benchmarking and promotes reproducible research in computational granular mechanics.
