| The high efficiency, high power density, and low operating temperatures of polymer electrolyte membrane fuel cell (PEMFC) differentiates itself from other fuel cell technologies. The major challenges facing commercialization of PEMFC are high cost, durability and the requirement of high purity hydrogen as fuel. To improve the power densities, and thus reduce cost per kilowatt and improve the durability mathematical models are used extensively in PEMFC research.; In this work a detailed steady-state isothermal two-dimensional model of a proton exchange membrane fuel cell has been developed. A finite element method is used to solve this multi-component transport model coupled with flow in porous medium, charge balance, electrochemical kinetics and a rigorous water balance in the membrane. The model-predicted fuel cell performance curves are compared with published experimental results and a good agreement is found. The complex water balance in the membrane is investigated, and the effects of channel width, bipolar plate shoulder size, porosity, and the relative humidity of the inlet streams on the fuel cell performance are also evaluated. It is found that smaller width channels and smaller bipolar plate shoulders are required for high current density operations.; This model is used in a full factorial experimental design to investigate the main and interaction effects of PEM fuel cell design parameters, and a new algorithm is then developed to integrate component balances along the PEMFC channels to obtain three dimensional results from the two dimensional model. This showed that hydrogen and air flow rates and their relative humidity are critical for current density, membrane dry-out, and electrode flooding. Furthermore relative humidity and flow rates are also critical for obtaining uniform current densities along the channels which impacts thermal management, reliability, and life of the fuel cell. This new integration approach, together with a detailed two dimensional across-the-channel model, is found to be a promising analysis method due to its low computational cost, its applicability to a wide range fuel cell designs and its scalability to stack models. |
