The discrete element method (DEM) is used to model particulate solids (granular materials) with particle sizes typically larger than 1 micron. These materials are highly diverse, including soils, pharmaceutical powders and many food ingredients. DEM models individual particles and their interactions over time, and is algorithmically similar to molecular dynamics.
The presenter has been running DEM simulations on national HPC facilities since 2012 (HECToR). He has principally used the Granular package of LAMMPS, adapting the source code as required. In this webinar, he will present examples of his and collaborators’ research in two areas which have been enabled by DEM on national HPC facilities: the simulation of triaxial tests to gain insights into sand behaviour, and current research on abrasion in particle systems.
The second part of the webinar will discuss broader trends in DEM and their effect on HPC usage. The desire to simulate larger systems and increase particle shape fidelity are both motivating the adoption of HPC.
DEM code developments are being undertaken to ensure that codes can efficiently run on GPUs, which will be essential to exploit the next generation of HPC systems.
The webinar will close with a summary of CCC-ParaSolS: an STFC-funded project to create a multi-disciplinary Collaborative Computational Community in particulate solids simulations for the UK. A code development project is being undertaken within CCC-ParaSolS to enable the MercuryDPM code to be run efficiently on GPUs. Since early 2025, CCC-ParaSolS has delivered bespoke training at network events to reduce the barriers to adopting open-source DEM codes and accessing HPC systems such as ARCHER2.