Implications of inherent inhomogeneities in thin carbon fiber-based gas diffusion layers: A comparative modeling study
MetadataShow full item record
Thin porous media are present in multiple electrochemical energy devices, where they provide key transport and structural functions. The prototypical example is gas diffusion layers (GDLs) in polymer-electrolyte fuel cells (PEFCs). While modeling has traditionally been used to explore PEFC operation, this is often accomplished using volume-averaged (VA) formulations, where the intrinsic inhomogeneities of the GDL are smoothed out and the lack of defining a representative elementary volume is an ever-present issue. In this work, the predictions of a single-phase VA PEFC model are compared to those of a pore-scale PEFC model using GDL tomograms as a part of the meshed domain to delineate important aspects that VA models cannot address. The results demonstrate that while VA models equipped with suitable effective properties can provide a good average estimate for overall performance, the lack of accounting for real structures limits their predictive power, especially for durability and degradation behavior where large deviations are found in the spatial distributions. Furthermore, interfacial effects between the GDL and the microporous layer are explored with the pore-scale model to understand the implications of the layered geometry. It is shown that the actual microstructure of the GDL/MPL transition region can significantly affect the fluxes across the sandwich, something that VA models cannot easily consider. Interfacial design is recognized as a key quality control parameter for large-scale MEA manufacturing and assembly.
Cite this version of the work
Pablo García-Salaberri, Iryna Zenyuk, Gisuk Hwang, Marcos Vera, Adam Z. Weber, Jeff Gostick (2019). Implications of inherent inhomogeneities in thin carbon fiber-based gas diffusion layers: A comparative modeling study. UWSpace. http://hdl.handle.net/10012/14258
The following license files are associated with this item: