Temperature-Dependent Multiphysics Modeling of Additive Friction Stir Deposition Using Multi-Task Coupled Physics-Informed Neural Networks

License: See LICENSE file · Updated 2026-10-06
My rating

The paper targets additive friction stir deposition, where frictional heating, severe plastic deformation, and tool-imposed boundary conditions create strongly coupled thermal and material-flow fields. It uses multi-task coupled physics-informed neural networks to address the strong temperature dependence of these fields. High-fidelity finite-volume simulations are presented as an accurate but computationally costly approach for repeated evaluation across process conditions.

Tags: