Modelling of solidification problems with FEniCS
Florian Arbes (IFE, 🇳🇴)
Kent-Andre Mardal (UiO, 🇳🇴)
Øyvind Jensen (IFE, 🇳🇴)
Jørgen S. Dokken
(University of Cambridge, 🇬🇧)
You can cite this talk by using the following BibTeΧ:
@incollection{fenics2021-arbes,
title = {Modelling of solidification problems with FEniCS},
author = {Florian Arbes and Kent-Andre Mardal and {\O}yvind Jensen and J{\o}rgen S. Dokken},
year = {2021},
url = {http://mscroggs.github.io/fenics2021/talks/arbes.html},
booktitle = {Proceedings of FEniCS 2021, online, 22--26 March},
editor = {Igor Baratta and J{\o}rgen S. Dokken and Chris Richarson and Matthew W. Scroggs},
doi = {10.6084/m9.figshare.14495358},
pages = {371}
}
Hide citation infoModelling the solidification of alloys remains challenging, as there is a multi-physics problem to be solved involving fluid flow, heat transfer as well as phase changes. Shape casting processes are constantly sought to be improved, which can potentially be achieved through accurate and fast modelling of the alloy solidification process. We present in a short overview our approach of an application relying on FEniCS that will be used to model such solidification problems. The model, which is currently under development, follows a volume-averaged two-phase approach, as proposed by Beckermann et al [1].
Simplified models have been tested and used to generate snapshots for a non-intrusive reduced order model. As this application is being built in Python, we hope its flexible core will open up new opportunities for fast and accurate reduced order models.
References
- [1] J. Ni and C. Beckermann. A volume-averaged two-phase model for transport phenomena during solidification, Metallurgical Transactions B 22, 1991. [DOI: 10.1007/BF02651234]