A versatile microstructure tool for large-scale atomistic simulations
Atomistic simulations of polycrystalline systems require realistic microstructure models with resolved atomic coordinates. This work presents a tool that efficiently creates such models. Besides the construction of conventional Voronoi-based polycrystalline models, the tool interfaces with existing software for microstructure generation and analysis, thereby allowing realistic microstructure models based on experimentally determined or synthetic grain geometries obtained from phase field simulations or equivalent ellipsoidal grain (EEG) generation with constrained grain packing, implemented in the DREAM.3D code. The MedeA Microstructure Builder is verified to produce atomistic models of both equiaxed and textured microstructures as well as key statistical information on grain size distributions, geometry, and orientation. Precipitates at grain boundaries, voids, and interstitial atoms may also be inserted to create more complex microstructures. Periodic polycrystal systems with millions of atoms can be generated for use in MD or MC simulations. A complementary database and building tools facilitate the generation of high-symmetry grain boundaries amenable for density-functional theory calculations as well as large-scale simulations using interatomic potentials.
Authors
- Jonathan L. Wormald (ORCID: https://orcid.org/0000-0001-7727-5967)
- Leonid Kahle (ORCID: https://orcid.org/0000-0002-4440-5996)
- Benjamin S. Anglin (ORCID: https://orcid.org/0000-0001-5501-116X)
- E. Wimmer (ORCID: https://orcid.org/0000-0002-7079-9589)
- Clint B. Geller (ORCID: https://orcid.org/0000-0003-2414-5928)
- Mikael Christensen (ORCID: https://orcid.org/0000-0002-7078-124X)
- Kyle P. Starkey (ORCID: https://orcid.org/0000-0001-9798-4541)
Institutions
- Naval Nuclear Laboratory (US)
Publication Details
- Journal
- Computational Materials Science
- Published
- 2026-10-05
- DOI
- https://doi.org/10.1016/j.commatsci.2026.115119
- Primary Topic
- Machine Learning in Materials Science
- Type
- article
- Field-Weighted Citation Impact
- 0.00
Funders
- U.S. Department of Energy