Exploration of optimal microstructure and mechanical properties in continuous microstructure space using a variational autoencoder

Data-driven approaches enable a deep understanding of microstructure and mechanical properties of materials and greatly promote one's capability in designing new advanced materials. Deep learning-based image process-ing outperforms conventional image processing techniques with unsupervised learning. This study employs a variational autoencoder (VAE) to generate a continuous microstructure space based on synthetic microstructural images. The structure-property relationships are explored using a computational approach with microstructure quantification, dimensionality reduction, and finite element method (FEM) simulations. The FEM of representa-tive volume element (RVE) with a microstructure-based constitutive model model is proposed for predicting the overall stress-strain behavior of the investigated dual-phase steels. Then, Gaussian process regression (GPR) is used to make connections between the latent space point and the ferrite grain size as inputs and mechanical properties as outputs. The GPR with VAE successfully predicts the newly generated microstructures with target mechanical properties with high accuracy. This work demonstrates that a variety of microstructures can be can-didates for designing the optimal material with target properties in a continuous manner. (c) 2021 The Author(s). Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

Automated summary

Data-driven approaches with deep learning-based image processing, utilizing variational autoencoders to generate continuous microstructure spaces, help explore structure-property relationships for designing new advanced materials. By using finite element method simulations and Gaussian process regression with VAE, accurate predictions of microstructures with target mechanical properties can be achieved in a continuous manner.