Microstructure Generation via Generative Adversarial Network for Heterogeneous, Topologically Complex 3D Materials
This work addresses the challenge of accurately modeling heterogeneous materials for materials science applications, representing an incremental improvement over existing methods.
The authors tackled the problem of generating realistic 3D microstructures for solid oxide fuel cell electrodes by implementing a generative adversarial network (GAN) on an experimental dataset, resulting in microstructures that closely matched the original in visual, statistical, and topological properties, with simulations showing performance distributions similar to the original while an established algorithm led to significant differences.
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.