This is for a 2-year Post-Doc position.
 
Missing or incorrectly modeled physics contributes to the lack of a widely accepted drag model formulation 
for universal use across all fluidization regimes and particle types. Overall fluidization simulation accuracy 
cannot improve without a mechanistic understanding of what physical interactions are not adequately 
captured in the current models. To address this, machine learning (ML) / artificial intelligence (AI) can spot 
profound and elucidate ambiguous relations that can be generalizable. Previous ML/AI studies for drag offer 
new formulations that outperform typical ones for targeted problems. What remains amiss is a widely 
accepted drag model that is generally applicable, or a set of drag models that work well for different 
parameter spaces. This motivates the current project. Other novel ideas along this line are welcomed.