Erhard, Linus C. ; Utt, Daniel Thomas ; Klomp, Arne Jan ; Albe, Karsten (2024)
Crystal structure identification with 3D convolutional neural networks with application to high-pressure phase transitions in SiO2.
In: Modelling and Simulation in Materials Science and Engineering, 32 (6)
doi: 10.1088/1361-651X/ad64f3
Artikel, Bibliographie
Kurzbeschreibung (Abstract)
Efficient, reliable and easy-to-use structure recognition of atomic environments is essential for the analysis of atomic scale computer simulations. In this work, we train two neuronal network (NN) architectures, namely PointNet and dynamic graph convolutional NN (DG-CNN) using different hyperparameters and training regimes to assess their performance in structure identification tasks of atomistic structure data. We show benchmarks on simple crystal structures, where we can compare against established methods. The approach is subsequently extended to structurally more complex SiO2 phases. By making use of this structure recognition tool, we are able to achieve a deeper understanding of the crystallization process in amorphous SiO2 under shock compression. Lastly, we show how the NN based structure identification workflows can be integrated into OVITO using its python interface.
| Typ des Eintrags: | Artikel |
|---|---|
| Erschienen: | 2024 |
| Autor(en): | Erhard, Linus C. ; Utt, Daniel Thomas ; Klomp, Arne Jan ; Albe, Karsten |
| Art des Eintrags: | Bibliographie |
| Titel: | Crystal structure identification with 3D convolutional neural networks with application to high-pressure phase transitions in SiO2 |
| Sprache: | Englisch |
| Publikationsjahr: | 5 August 2024 |
| Verlag: | IOP Publ. |
| Titel der Zeitschrift, Zeitung oder Schriftenreihe: | Modelling and Simulation in Materials Science and Engineering |
| Jahrgang/Volume einer Zeitschrift: | 32 |
| (Heft-)Nummer: | 6 |
| DOI: | 10.1088/1361-651X/ad64f3 |
| Kurzbeschreibung (Abstract): | Efficient, reliable and easy-to-use structure recognition of atomic environments is essential for the analysis of atomic scale computer simulations. In this work, we train two neuronal network (NN) architectures, namely PointNet and dynamic graph convolutional NN (DG-CNN) using different hyperparameters and training regimes to assess their performance in structure identification tasks of atomistic structure data. We show benchmarks on simple crystal structures, where we can compare against established methods. The approach is subsequently extended to structurally more complex SiO2 phases. By making use of this structure recognition tool, we are able to achieve a deeper understanding of the crystallization process in amorphous SiO2 under shock compression. Lastly, we show how the NN based structure identification workflows can be integrated into OVITO using its python interface. |
| Freie Schlagworte: | crystal structure identification, machine learning, silica |
| ID-Nummer: | Artikel-ID: 065029 |
| Zusätzliche Informationen: | D U gratefully acknowledges funding by the NHR4CES research project as part of SDL ‘Materials’. The research was supported by the Bundesministerium für Bildung und Forschung (BMBF) within the project FESTBATT under Grant No. 03XP0174A. |
| Fachbereich(e)/-gebiet(e): | 11 Fachbereich Material- und Geowissenschaften 11 Fachbereich Material- und Geowissenschaften > Materialwissenschaft 11 Fachbereich Material- und Geowissenschaften > Materialwissenschaft > Fachgebiet Materialmodellierung Zentrale Einrichtungen Zentrale Einrichtungen > Hochschulrechenzentrum (HRZ) Zentrale Einrichtungen > Hochschulrechenzentrum (HRZ) > Hochleistungsrechner |
| TU-Projekte: | PTJ|03XP0174A|FestBatt-Daten |
| Hinterlegungsdatum: | 30 Aug 2024 07:40 |
| Letzte Änderung: | 12 Dez 2024 08:35 |
| PPN: | 524504865 |
| Export: | |
| Suche nach Titel in: | TUfind oder in Google |
 |
Frage zum Eintrag |
Optionen (nur für Redakteure)
 |
Redaktionelle Details anzeigen |
Drucken
Impressum