AI for Chemical Engineering
At KIT we unite a vast diversity of research disciplines. We teamed up with Bradley Ladewig (Institute for Mirco Process Engineering) and Alexander Stroh (Institute for Fluid Dynamics) to improve fluid processes in Chemical Engineering.
Our goal is to build integrated simulation and AI workflows in which AI models:
- ...can predict the properties of microfluidic devices
- ...obtain new training data on the fly by running simulations for predictions with high uncertainty
- ...thereby provide knowledge to generative models to design even better devices!
In late 2020 we already started a proof of concept study which is available as preprint on arXiv and which we summarized for you in the following video:
Machine learning for rapid discovery of laminar flow channel wall modifications that enhance heat transfer
The calculation of heat transfer in fluid flow in simple flat channels is arelatively easy task for various simulations methods. However, once the channelgeometry becomes more complex, numerical simulations become a bottleneck inoptimizing wall geometries. We present a combination of accurate num…
In 2021 we obtained a DFG funding through the SPP "Machine Learning in Chemical Engineering" (2331) for lab equipment at Brad's department, one PhD student at Alex's and one PhD student in our group. ?