Open access
Aug 2026
Thermomechanical Field Prediction in Subbase Materials Using Physics-Informed Neural Networks
A two-dimensional axisymmetric physics-informed neural network model integrating thermomechanical behavior, phase transition, and deformation was developed, offering a novel framework for multi-field reconstruction and frost heave prediction in subgrade materials under conditions of limited observational data.
An-Shuang Su, Heng Zhou, Mingwei Hai et al.
· Applied Sciences · 2 citations