Physics-Informed Neural Networks for Heat Transfer Problems

TLDR

Physics‑informed neural networks (PINNs) leverage automatic differentiation and multitask learning to solve realistic engineering problems with noisy or incomplete data, avoiding discretization errors. This study applies PINNs to prototype heat‑transfer problems, focusing on realistic conditions that are difficult for conventional computational methods. The authors demonstrate PINNs on forced and mixed convection with unknown thermal boundaries, a two‑phase Stefan problem to recover moving interfaces and material conductivities, and industrial power‑electronics scenarios, all using sparse temperature measurements to infer full temperature and velocity fields. Results show that PINNs can solve ill‑posed heat‑transfer problems beyond the reach of traditional methods and bridge the gap between computational and experimental heat transfer.

Abstract

Abstract Physics-informed neural networks (PINNs) have gained popularity across different engineering fields due to their effectiveness in solving realistic problems with noisy data and often partially missing physics. In PINNs, automatic differentiation is leveraged to evaluate differential operators without discretization errors, and a multitask learning problem is defined in order to simultaneously fit observed data while respecting the underlying governing laws of physics. Here, we present applications of PINNs to various prototype heat transfer problems, targeting in particular realistic conditions not readily tackled with traditional computational methods. To this end, we first consider forced and mixed convection with unknown thermal boundary conditions on the heated surfaces and aim to obtain the temperature and velocity fields everywhere in the domain, including the boundaries, given some sparse temperature measurements. We also consider the prototype Stefan problem for two-phase flow, aiming to infer the moving interface, the velocity and temperature fields everywhere as well as the different conductivities of a solid and a liquid phase, given a few temperature measurements inside the domain. Finally, we present some realistic industrial applications related to power electronics to highlight the practicality of PINNs as well as the effective use of neural networks in solving general heat transfer problems of industrial complexity. Taken together, the results presented herein demonstrate that PINNs not only can solve ill-posed problems, which are beyond the reach of traditional computational methods, but they can also bridge the gap between computational and experimental heat transfer.