Through the membrane & along the channel flooding in PEMFCs

Abstract

Listen

Translate article

Neutron imaging of a polymer electrolyte membrane fuel cell (PEMFC) revealed distinct patterns of water fronts moving through the gas diffusion layers (GDL) and channels. The PEMFC was operating with dead-ended, straight and almost vertically-oriented anode channels; hence the gravity driven accumulation of liquid water at the end of the channel caused flooding in an upward direction. In order to predict the spatiotemporal evolution of water patterns inside severely-flooded fuel cells, various distributed parameter models of the water transport through the membrane and GDLs to the cathode and anode channels have been developed by the authors and others. In this paper, a zero-dimensional moving front model is presented which captures the location of the water phase transition inside the GDL, instead of using the standard partial differential equation (PDE) approach for modeling liquid water in porous media which is numerically difficult to solve. This model uses three nonlinear states (the anode and cathode GDL front location and the membrane water content) and three inputs (the anode and cathode vapor concentration and the current density) to predict the slowly evolving front locations in both anode and cathode side GDLs during flooding and drying as well as the dynamic changes in membrane water content. The unit cell model is finally formulated with three hybrid modes and their transition laws. The hybrid-state model will be parameterized in the future using experimentally observed front evolutions. This parameterized unit cell model will be used to model the water accumulation along the channel in order to predict and avoid severe flooding conditions.