Water Distribution in Fuel Cell Gas Channels Using a Mechanistic Discrete Particle Model

Abstract

Water removal and accumulation in fuel cell gas channels is a complex phenomenon which is a consequence of the strength of the capillary forces and inertial forces (1 < Re < 1000). The presence of water in the channel restricts the access of oxygen to diffuse through the gas diffusion layers to the catalyst layer. The air supplied exchanges momentum to the droplets and water clusters, causing detachment when the adhesion force is overcome.Water emerging from discrete locations as droplets interact with both the channel walls and each discrete water cluster, depending on the configuration and operating conditions. Complexity increases as multiple detachment and coalescence events occur over large channel lengths and time scales. Modelling these systems with traditional method computational fluid dynamic (CFD) methods is computationally expensive so we have developed a new mechanistic discrete particle model to handle the liquid water two-phase flow.Taking inspiration from interface resolving volume of fluid (VoF) simulations of water flow in gas channels, we use analytically equations determined by the equilibrium contact angle in different configurations on the channel wall to predict detachment times and coalescence events. We compare the prediction of water area coverage ratio and channel saturation to the VoF simulations and can also estimate dynamic pressure drop.Using this model, we are able to investigate the dynamic effects of liquid water accumulation and distribution in the channel, with often periodic detachment events arising from the formation of liquid plugs. Consequently, the model is able to investigate the effect of the channel wall wettability, from hydrophilic to hydrophobic and the impact that has on the distribution and accumulation of water in the system.We compare the speed of this model against the VoF simulations to show that this type of modelling of two-phase flow in the channel can enable higher resolution accuracy when coupled with continuum scale approaches. Furthermore, with this approach it is possible to investigate the effect of: channel dimensions, liquid water flow rate and spatial location of liquid water emergence. This should allow for designed fuel cell structures to be evaluated. Figure 1