Search

Water management is of significant importance to achieving high performance of proton exchange membrane fuel cells. In recent years, droplets emerged from the rib surface and accumulated at the channel corner have been found to be a crucial part of water flooding. In this study, an analytical model is first proposed to quantitatively estimate the variation in the morphology and dynamic behavior of growing droplets with consideration of the channel sidewall interaction. In order to predict the water geometry, the flow channel with compressed gas diffusion layer (GDL) is described mathematically, and water behavior at steady-state and dynamic state are both evaluated through the geometric and force analysis. The model results indicate that the droplet profile transforms from concave to convex when its size grows, in which process contact angles and channel shape play an important role. Compared with the graphite channel, the droplet in the metallic channel is more inclined to be adsorbed on the sidewall and GDL, resulting in a higher adhesion force and a lower gas shear force. The critical gas velocities for the detachment of droplets are quantitatively predicted to avoid water flooding. The model is helpful to understand the droplet behavior in the presence of channel sidewall interaction.

Proton exchange membrane fuel cell (PEMFC) stacks are assembled by numerous bipolar plates (BPPs) and membrane electrode assemblies (MEAs). Component manufacturing errors are inevitably caused during the production process and brought into the stack, affecting the mechanical consistency of every cell. Aiming to improve the uniformity of the assembled stack effectively, this paper establishes a novel assembly deviation model based on the asymptotic homogenization (AH) method and proposes an assembly sequence optimization approach for the stack with dimensional errors using the genetic algorithm. The results reveal that dimensional errors significantly affect the equivalent mechanical properties of the stack. Reasonable matching of components with different types of dimensional errors is an effective method to avoid the cumulative negative impacts on the mechanical behavior of the stack. Statistically, the optimization of the assembly sequence can improve contact pressure uniformity in a 10-cell stack by more than 9.00%, while reducing the maximum contact pressure by 15.55%. This methodology is also applicable to improving the assembly uniformity of other stacked structures.