What type of study is this?

September 10, 2025Physics of Fluids

Flow field reconstruction of cylinder water entry using graph neural networks

View Full Paper

Ask AI

Bookmark

Authors

Graph neural networks accurately predict flow fields during cylinder water entry, suggesting enhanced modeling.

The GNN model successfully predicts water entry flow, capturing cavity formation and dynamics effectively.
At a Froude number of 15.65, the model shows maximum prediction error near the flow boundary and surface.
Adding similar cases to the training set helps reduce prediction errors in the water volume fraction.
Future improvements in accuracy may come from integrating physical constraints into the model design.

Shi et al. (2025) studied this question.

View Full Paper

Ask AI

Bookmark