Operation of proton-exchange
membrane fuel cells is highly deteriorated
by mass transfer loss, which is a result of spatial and temporal interaction
between airflow, water flow, channel geometry, and its wettability.
Prediction of two-phase flow dynamics in gas channels is essential
for the optimization of the design and operating of fuel cells. We
propose a mechanistic discrete particle model (DPM) to delineate dynamic
water distribution in fuel cell gas channels and optimize the operating
conditions. Similar to the experimental observations, the model predicts
seven types of flow regimes from isolated, side wall, corner, slug,
film, and plug flow droplets for industrial temporal and spatial scales.
Consequently, two-phase flow regime maps are proposed. The results
suggest that an increase in water accumulation in the channel is related
to the increase in the water cluster density emerging from the gas
diffusion layer rather than the increased water flow rate through
constant water pathways. From a modeling perspective, the DPM replicated
well volume-of-fluid channel simulation results in terms of saturation,
water coverage ratio, and interface locations with an estimated 5
orders of magnitude increase in calculation speed.