Microstructure Generation via Generative Adversarial Network for Heterogeneous, Topologically Complex 3D Materials
Microstructure Generation via Generative Adversarial Network for Heterogeneous, Topologically Complex 3D Materials
复制标题
DOI：
10.1007/s11837-020-04484-y
复制
发表时间：
2020-12-02
期刊：
影响因子：
2.6
通讯作者：
Holm, Elizabeth A.
中科院分区：
文献类型：
--
作者：
Hsu, Tim;Epting, William K.;Holm, Elizabeth A.
利用大规模的实验采集的三维微结构数据集，我们实现了生成对抗网络(GAN)框架来学习和生成固体氧化物燃料电池电极的三维微结构。生成的显微组织在视觉上、统计上和拓扑上都是逼真的，其体积分数、颗粒尺寸、比表面积、弯曲度和三相界面密度等微结构参数的分布与原始组织的分布高度相似。将这些结果与已建立的基于颗粒的生成算法(DREAM.3D)的结果进行了比较。重要的是，使用局部解析有限元模型对电化学性能进行了模拟，结果表明，GaN生成的微结构与原始的性能分布非常接近，而DREAM.3D则导致了显著的差异。产生式机器学习模型重建高保真微结构的能力表明，复杂微结构的本质可以被捕捉到，并以紧凑和可操作的形式表示。
Using a large-scale, experimentally captured 3D microstructure data set, 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.