SKM 2023 – wissenschaftliches Programm

Bereiche | Tage | Auswahl | Suche | Aktualisierungen | Downloads | Hilfe

MM: Fachverband Metall- und Materialphysik

MM 12: Poster I

MM 12.24: Poster

Montag, 27. März 2023, 18:15–20:00, P2/OG1+2

An efficiently automated method to sample the energies of grain boundaries — •Timo Schmalofski¹, Martin Kroll^2,3, Rebecca Janisch¹, and Holger Dette³ — ¹ICAMS, Ruhr-University Bochum — ²Department of Mathematics, Physics, and Computer Science, University of Bayreuth — ³Department of Mathematics, Ruhr-University Bochum

Grain growth and microstructure evolution depend on the anisotropy of the energy of grain boundaries, which is a function of the five geometric degrees of freedom (DOF) of the grain boundaries. To access this parameter space in an efficient way and discover energy cusps in unexplored regions, a method was established, which combines atomistic simulations with statistical methods [1]. It has been successfully applied to sample the 2D subspace of GB plain inclinations for fixed misorientations. The poster explains the main features of the algorithm: Initial design, sequential design, the stopping criterion and the final interpolation of the energy. The algorithm draws its strengths from two aspects, the choice of the next point, which balances a homogeneous distribution of points with a precise sampling of the cusps, and the stopping criterion, which monitors the error of the prediction as well as the number of cusps which have been found. With these features, the method is able to outperform a regular high-throughput sampling.