Electronic International Standard Serial Number (EISSN)
Multiscale modeling of porous media in polymer electrolyte fuel cells is of paramount importance to improve predictions and assist the design of new materials. In this work, a composite-continuum-network formulation is presented to model species diffusion and convection in gas diffusion layers (GDLs). The model can be incorporated into CFD codes with moderate computational cost. The macroscopic model is based on a structured mesh composed of parallelepiped control volumes (CVs) and differential connectors (with negligible volume). The CV mesh embeds an internal structured pore network, which is used to determine analytically local anisotropic effective transport properties (effective diffusivity and permeability). The global structural parameters and effective transport properties predicted by the model are in good agreement with previous experimental data. Moreover, the results show that heterogeneities in the GDL can have significant influence on the fluxes from/to the catalyst layer, thus affecting local degradation rates.