Understanding the training of PINNs for unsteady flow past a plunging foil through the lens of input subdomain level loss function gradients
This work provides incremental insights into PINN training dynamics for computational fluid dynamics problems with moving boundaries, aiding researchers in optimizing such models.
The study investigated how different spatial subdomains (moving body, wake, outer zones) influence the training of moving boundary-enabled physics-informed neural networks (MB-PINNs) for unsteady flow past a plunging foil, using novel metrics based on zonal loss gradients and point proportions to confirm that training is driven by these combined effects.
Recently immersed boundary method-inspired physics-informed neural networks (PINNs) including the moving boundary-enabled PINNs (MB-PINNs) have shown the ability to accurately reconstruct velocity and recover pressure as a hidden variable for unsteady flow past moving bodies. Considering flow past a plunging foil, MB-PINNs were trained with global physics loss relaxation and also in conjunction with a physics-based undersampling method, obtaining good accuracy. The purpose of this study was to investigate which input spatial subdomain contributes to the training under the effect of physics loss relaxation and physics-based undersampling. In the context of MB-PINNs training, three spatial zones: the moving body, wake, and outer zones were defined. To quantify which spatial zone drives the training, two novel metrics are computed from the zonal loss component gradient statistics and the proportion of sample points in each zone. Results confirm that the learning indeed depends on the combined effect of the zonal loss component gradients and the proportion of points in each zone. Moreover, the dominant input zones are also the ones that have the strongest solution gradients in some sense.