Project


FNOs for Predicting Fuel Sloshing

Developed a Fourier Neural Operator to predict the flow field of fuels sloshing inside an axisymmetric tank under varous loading conditions in real time. The model was analyzed for any susceptibilities to radiation upsets that are quite common in space applications.

The above video compares the true variation of the axial velocity with the predicted flow field under a rectangular pulse loading condition. The GitHub Repository containing the project code and results is linked.