Condensation in PEM Fuel Cell Gas Diffusion Layers: A Pore Network Modeling Approach

Abstract

A two-dimensional dynamic pore network model is developed and employed to simulate the through-plane transport of liquid water originating from condensation in hydrophobic gas diffusion layers (GDLs) for polymer electrolyte membrane (PEM) fuel cells. The model tracks viscous and capillary forces over a range of specified condensation rates and nucleation positions. A simplified mass transport assumption of a uniform water vapor flux between the cathode catalyst layer and the liquid water cluster allows for a computationally inexpensive model. Stochastically generated steady-state saturation profiles are compared to investigate the effects of nucleation position, channel rib presence, coalescence assumptions, and condensation rates on liquid water distribution. Results indicate that GDL saturation conditions become increasingly more desirable as nucleation sites are placed further away from the catalyst layer, and saturation profiles are significantly higher when nucleation is adjacent to a hydrophobic rib compared to the gas channel or a hydrophilic rib. Meanwhile, the trapping assumption that affects liquid water coalescence in small throats has little impact on saturation patterns but has a large impact on the system’s viscous forces. Finally, the model can predict the limiting water cluster growth rates of capillary dominated growth for a given pore network.