Integrating Neural Network-Based Drag Force Models in MFIX

Hello everyone,

I am interested in knowing whether it is feasible to conveniently call a drag force model based on neural network predictions within the MFIX process. This would be useful for enhancing the simulation accuracy by incorporating machine learning techniques into the traditional CFD framework.

So far, I have not come across any literature that deploys AI model in MFIX. If anyone has insights or suggestions on how to implement such a feature, I would greatly appreciate your advice. Additionally, if there are any relevant papers, case studies, or examples that you could share, they would be invaluable resources for my work.

Thank you in advance for your time and assistance. Your expertise and guidance are much appreciated!

Best regards,