Towards Quantification of Local Electrochemical Parameters in Microstructures of Solid Oxide Fuel Cell Electrodes using High Performance Computations

Friday, 28 July 2017: 08:20
Atlantic Ballroom 1/2 (The Diplomat Beach Resort)
T. Hsu, R. Mahbub (Carnegie Mellon University, U.S. DOE National Energy Technology Laboratory), W. K. Epting (Oak Ridge Institute for Science and Education, U.S. DOE National Energy Technology Laboratory), H. Abernathy (AECOM, U.S. DOE National Energy Technology Laboratory), G. A. Hackett (U.S. DOE National Energy Technology Laboratory), A. D. Rollett, S. Litster (Carnegie Mellon University, U.S. DOE National Energy Technology Laboratory), and P. A. Salvador (U.S. DOE National Energy Technology Laboratory, Carnegie Mellon University)
Recent progress in micro-scale three-dimensional (3D) characterization techniques such as focused ion beam – scanning electron microscopy (FIB-SEM) and X-ray nanotomography has brought unprecedented opportunities in linking material microstructure to performance and properties. The link between solid oxide fuel cell (SOFC) electrode microstructure and performance is sometimes established by effective medium theory, where effective performance parameters are calculated from a representative volume element (RVE), an adequate volume that statistically represents global microstructure. However, different from performance, degradation and failure have not been addressed by mean parameters from effective medium theory. With the advent of modern 3D characterization techniques, there are new approaches in linking microstructure and degradation. This work aims to emphasize distributions (and outliers), rather than the mean, in better establishing the link between microstructure and degradation, failure, and performance. Using an open source finite element framework (MOOSE, Idaho National Laboratory) on a high performance computing platform (Joule, National Energy Technology Laboratory), a numerical transport and reaction model is applied to microstructural meshes to compute local distributions of electrochemical parameters. The microstructural meshes are obtained from the plasma sourced FIB-SEM characterization of a commercial SOFC electrode. The RVE size with respect to the computed distributions is determined based on examining the distributions from a series of microstructural meshes with various sizes. To validate the distributions, the averaged numbers from the distributions are compared with the effective medium theories in the literature. Finally, features such as the tails of the distributions and the local gradients of the electrochemical parameters are highlighted and discussed.