A trained Physics-Informed Neural Networks (PINNs) method for phase-field model in Allen-Cahn framework

Le Song, Xuyang Chen, Zhiyong Huang

Keywords: Asaro-Tiller-Grinfeld (ATG) instability, Physics-Informed Neural Networks (PINNs), Phase-field model, Allen-Cahn Framework, Semiconductor thin films, Computational efficiency

TL;DR: This study integrates PINNs with a phase-field model to predict ATG instability in thin films. PINNs reduce computation time 10× and lower error rates from ~8.9% to ~2.4%, outperforming traditional methods in accuracy and efficiency.

Confirmation Of Submission Requirements: I submit an abstract. It uses the template provided on the submission page and is no longer than 2 pages.

Submission Number: 264