Summary
Predicting how tiny 3D structures inside materials change over time usually takes a lot of computer time because it requires complex simulations. The authors developed a new method that uses a special type of deep learning to predict these changes much faster while still keeping the important physical rules in mind. Their approach looks at the 3D shapes over multiple future time steps all at once and can produce accurate predictions even with less past information. This method speeds up the prediction process by more than 30 times compared to traditional simulations, making it useful for studying materials efficiently.
What this means in practice
- •For materials engineers: Rapidly forecast 3D microstructure evolution during material processing to optimize treatments without costly simulations.
- •For computational physicists: Speed up repeated 3D phase-field simulations by replacing them with efficient physics-guided neural network predictions.
Abstract
Accurate prediction of three-dimensional (3D) microstructure evolution remains computationally demanding because high-fidelity phase-field simulations require repeated numerical integration over large volumetric domains and long temporal horizons. This study develops an efficient and scalable physics-guided fully convolutional spatiotemporal framework for direct multi-frame prediction of complete 3D microstructure sequences. The model combines shared 3D spatial encoding and decoding with a factorized latent translator that integrates temporal, local 3D spatial, and channel interactions. A discrete Cahn--Hilliard (CH) residual is incorporated during training to regularize the learned evolution toward the governing dynamics without altering the inference pathway. The framework is evaluated on high-resolution 3D spinodal-decomposition trajectories under nominal, long-horizon, and reduced-temporal-context forecasting. Under full temporal context, the model accurately reproduces volumetric evolution, with average 3D structural similarity remaining above 0.97 over the nominal prediction horizon. Physics guidance becomes increasingly beneficial as temporal information is reduced, improving predictive robustness and preservation of interface-level morphology. The framework also achieves more than a 30-fold wall-clock speedup relative to the reference spectral phase-field solver, while physics guidance introduces no additional inference cost. These results establish direct multi-frame, physics-guided fully convolutional learning as a high-throughput surrogate strategy for dense 3D phase-field dynamics and repeated microstructure forecasting.