Master Thesis: Physics-Informed Machine Learning in Computational Fluid Dynamics

I am starting a master thesis project on Machine Learning in Computational Fluid dynamics in cooperation with Intel and BOSCH Corporate Research. We will work on coupling OpenFOAM with Physics-Informed Neural Networks. For more information and a small hands-on example, check out my Youtube talk at the OpenFOAM workshop. The minimal working example is on GitHub, this will be our starting point. We will collaborate with Intel's Advanced AI laboratory to see if we can train Physics-Informed Neural Networks in OpenFOAM efficiently, then use the models in microfluidics simulations in cooperation with BOSCH Corporate Research.

I'm looking for a Computational Engineer/Scientist with a strong affinity for research, interested in learning the details of the unstructured Finite Volume Method, OpenFOAM, and Physics-Informed Machine Learning. You will be working closely with me on this topic, and collaborating with industrial partners and PhD candidates in my research group. Email your grade transcripts and CV to maric@mma.tu-darmstadt.de

Bumping this thread, we're still searching for a candidate. In case the Mathematical Modeling and Analysis Institute is scaring you off, this is a Computational Engineering topic to be organized at CE!