Microstructure Generation via Generative Adversarial Network for Heterogeneous, Topologically Complex 3D Materials

TL;DR

This paper presents a GAN-based method for generating realistic 3D microstructures of solid oxide fuel cell electrodes, outperforming traditional algorithms in fidelity and electrochemical performance simulation accuracy.

Contribution

The study introduces a GAN framework for 3D microstructure generation that closely replicates real microstructures and their electrochemical performance, surpassing existing methods.

Findings

Generated microstructures are visually, statistically, and topologically realistic.

GAN-generated microstructures match the original's electrochemical performance distribution.

Compared to DREAM.3D, GAN produces more accurate and representative microstructures.

Abstract

Using a large-scale, experimentally captured 3D microstructure dataset, we implement the generative adversarial network (GAN) framework to learn and generate 3D microstructures of solid oxide fuel cell electrodes. The generated microstructures are visually, statistically, and topologically realistic, with distributions of microstructural parameters, including volume fraction, particle size, surface area, tortuosity, and triple phase boundary density, being highly similar to those of the original microstructure. These results are compared and contrasted with those from an established, grain-based generation algorithm (DREAM.3D). Importantly, simulations of electrochemical performance, using a locally resolved finite element model, demonstrate that the GAN generated microstructures closely match the performance distribution of the original, while DREAM.3D leads to significant differences. The ability of the generative machine learning model to recreate microstructures with high fidelity suggests that the essence of complex microstructures may be captured and represented in a compact and manipulatable form.