Berlin 2024 – scientific programme
Parts | Days | Selection | Search | Updates | Downloads | Help
MM: Fachverband Metall- und Materialphysik
MM 11: Data Driven Material Science: Big Data and Workflows II
MM 11.8: Talk
Monday, March 18, 2024, 17:45–18:00, C 243
Adaptive-precision potentials for large-scale atomistic simulations — •David Immel1,2, Ralf Drautz1, and Godehard Sutmann1,2 — 1ICAMS, Ruhr-Universität Bochum, Bochum, Germany — 2JSC, Forschungszentrum Jülich, Jülich, Germany
Large-scale atomistic simulations rely on interatomic potentials providing an efficient representation of atomic energies and forces. Modern machine learning (ML) potentials provide the most precise representation compared to electronic structure calculations while traditional potentials provide a less precise, but computationally much faster representation and thus allow simulations of larger systems.
We combine a traditional and a ML potential to a multi-resolution description, leading to an adaptive-precision potential with an optimum of performance and precision in large complex atomistic systems. The required precision is determined per atom by a local structure analysis and updated automatically during a simulation. We use Copper as demonstrator material with an embedded atom model (EAM) as traditional and an atomic cluster expansion (ACE) as ML potential, but any material and potential combination can be used for an adaptive-precision potential. The approach is developed for the molecular dynamics simulator LAMMPS and includes a load-balancer to prevent problems due to the atom dependent force-calculation times, which makes it suitable for large-scale atomistic simulations.
In this contribution strategies for the creation of an adaptive-precision potential are discussed. First results from Copper nanoindentations are reported and further improvements are outlined.
Keywords: adaptive precision; large-scale atomistic simulations; Copper; ACE; EAM