Search papers, labs, and topics across Lattice.
This paper introduces a fully convolutional spatiotemporal neural network to predict microstructure evolution, offering a computationally efficient alternative to traditional phase-field simulations. The model is trained in a self-supervised manner on sequential images of grain growth and spinodal decomposition, learning underlying physical dynamics. Results demonstrate accurate prediction of both short-term local behaviors and long-term statistical properties, with generalization to unseen domains and improved computational efficiency compared to recurrent architectures.
Skip the computationally expensive simulations: a fully convolutional network can accurately predict microstructure evolution, offering a faster and more scalable approach.
Understanding and predicting microstructure evolution is fundamental to materials science, as it governs the resulting properties and performance of materials. Traditional simulation methods, such as phase-field models, offer high-fidelity results but are computationally expensive due to the need to solve complex partial differential equations at fine spatiotemporal resolutions. To address this challenge, we propose a deep learning-based framework that accelerates microstructure evolution predictions while maintaining high accuracy. Our approach utilizes a fully convolutional spatiotemporal model trained in a self-supervised manner using sequential images generated from simulations of microstructural processes, including grain growth and spinodal decomposition. The trained neural network effectively learns the underlying physical dynamics and can accurately capture both short-term local behaviors and long-term statistical properties of evolving microstructures, while also demonstrating generalization to unseen spatiotemporal domains and variations in configuration and material parameters. Compared to recurrent neural architectures, our model achieves state-of-the-art predictive performance with significantly reduced computational cost in both training and inference. This work establishes a robust baseline for spatiotemporal learning in materials science and offers a scalable, data-driven alternative for fast and reliable microstructure simulations.
