Extended pseudo-spectral physics-informed neural networks for phase-field models
This work provides a data-efficient method for identifying constitutive parameters in phase-field models, which is important for materials science and continuum mechanics applications where such parameters are often unknown.
The authors developed an extended pseudo-spectral physics-informed neural network (ESPINN) to infer unknown bulk chemical potential and gradient coefficients in phase-field models from limited dynamical data. On the 1D Cahn-Hilliard equation, they achieved accurate reconstruction from as few as one snapshot pair in the noiseless regime, with graceful degradation under noise.
Phase-field models play a central role in the continuum description of phase separation, in which the bulk free-energy density and the interfacial thickness parameter determine pattern formation and microstructural evolution. In practice, these constitutive quantities are rarely known a priori and must be inferred from limited dynamical observations. In this work, an extended pseudo-spectral physics-informed neural network (ESPINN) framework is developed for the inverse identification of phase-field models from transient snapshot data. It enables the simultaneous recovery of both the bulk chemical potential and unknown gradient coefficients. Numerical experiments on the one-dimensional Cahn-Hilliard equation demonstrate accurate and statistically stable reconstruction in the noiseless regime, with substantial constitutive information recoverable from even a single snapshot pair. In the presence of noise, reconstruction accuracy degrades gracefully, and increasing the number of snapshots improves robustness by reducing variance across runs. These results establish ESPINN as a data-efficient and physically consistent approach for learning free-energy structure in continuum models of phase separation.