A massive parallel simulation approach to 2D and 3D grain growth

Mießen, Christian; Gottstein, Günter; Bleck, Wolfgang; Melcher, Christof Erich

doi:10.18154/RWTH-2017-10148

A massive parallel simulation approach to 2D and 3D grain growth

Mießen, Christian^RWTH*

2017 & 2018

Verantwortlichkeitsangabevorgelegt von Dipl.-Math. Christian Mießen

ImpressumAachen 2017

Umfang1 Online-Ressource (v, 142 Seiten) : Illustrationen

Dissertation, Rheinisch-Westfälische Technische Hochschule Aachen, 2017

Veröffentlicht auf dem Publikationsserver der RWTH Aachen University 2018

Genehmigende Fakultät
Fak05

Hauptberichter/Gutachter
Gottstein, Günter (Thesis advisor) ; Melcher, Christof Erich (Thesis advisor) ; Bleck, Wolfgang (Thesis advisor)

Tag der mündlichen Prüfung/Habilitation
2017-11-06

Online
DOI: 10.18154/RWTH-2017-10148
URL: https://publications.rwth-aachen.de/record/709678/files/709678.pdf

Einrichtungen

Inhaltliche Beschreibung (Schlagwörter)
grain growth (frei) ; large scale simulation (frei) ; level-set (frei) ; magnetically driven grain boundary motion (frei) ; parallelization (frei) ; recrystallization (frei)

Thematische Einordnung (Klassifikation)
DDC: 620

Kurzfassung
online nicht verfügbar

A highly efficient simulation model for 2D and 3D grain growth and recrystallization was developed based on the level-set method. The new model introduces modern computational concepts to achieve excellent performance on parallel computer architectures. Strong scalability was found on ccNUMA architectures underlining maximum parallel efficiency of the implementation. For this purpose, the model considers the application of local level-set functions at the grain level. The model was utilized to simulate ideal and non-ideal grain growth in 2D and 3D with the objective to study the evolution of statistical representative volume elements in polycrystals. The novelty of the proposed level-set approach to grain growth resides in the explicit consideration of structural interfacial elements of the microstructure. The extensions allow to consider anisotropic grain boundary energies and triple junction drag in polycrystalline materials. In addition, microstructure evolution under the influence of secondary driving forces, i.e such as resulting from stored elastic energies or such as occur in anisotropic magnetic materials affected by an external magnetic field, was modeled and simulated considering very large volume elements composed of half a million of grains in 3D. The gain in computational performance is essential to conduct simulation to investigate rare events in microstructure evolution, such as nucleation sites during recrystallization.

OpenAccess:
PDF
(additional files)