A tungsten parameter set for Kinetic Monte Carlo simulations

An initially perfectly cubic tungsten cluster, will evolve by atom diffusion into its most energetically favourable shape, as shown here. Simulations with both the new Kinetic Monte Carlo (KMC) model and Molecular Dynamics (MD) give the same result.

In a recently published paper, we describe a new parameter set for tungsten surfaces that should be able to simulate long-term evolutions of arbitrarily rough tungsten surfaces and nanoscale structures such as e.g. clusters, nanotips or maybe nanowires. The parameter set, which consists of three subsets of migration energy barriers for first-, second- and third-nearest neighbor atom transitions, can be used with the opensource Kinetic Monte Carlo (KMC) code Kimocs.

In the paper we show that the tungsten model relaxes nanoclusters to their most energetically favourable shapes (see the figure above), in very good agreement with Molecular Dynamics simulations (MD). The simulations shown in the figure started as identical cubes. Even though they show a very similar evolution and end shape, the KMC simulation is about two orders of magnitude faster than the MD simulation. This is of course the main advantage of KMC simulations.

The new parameter set for tungsten, jansson2020tungsten, is included in the Supplementary Materials of the main paper and as ancillary files to the eprint, but also distributed with the source code of Kimocs: https://gitlab.com/vjansson/Kimocs/-/tree/master/parameters/W_jansson2020tungsten

Paper reference: Ville Jansson, Andreas Kyritsakis, Simon Vigonski, Ekaterina Baibuz, Vahur Zadin, Alvo Aablo, and Flyura Djurabekova 2020 Modelling Simul. Mater. Sci. Eng. 28 035011 https://doi.org/10.1088/1361-651X/ab7151. arXiv:1909.03519 [cond-mat.mtrl-sci]

The Kimocs source code: https://gitlab.com/vjansson/Kimocs

Some other animations of Kimocs simulations and a list of available parameter sets for other materials.

Leave a comment