Advances in Modeling and Simulation of Transport

B. D'Aguanno, G. Murgia, L. Pisani and M. Valentini
Center for Advanced Studies, Research and Development in Sardinia, IT

Keywords: PEM Fuel Cell

Abstract:
ORAL Mathematical models and numerical simulations treatment of PEM fuel cells have been developed to predict both the overall cell behavior, as typified by the cell potential versus the cell current density curve, and the local spatial behavior of fuel cell variables, such as the chemical species concentrations, the temperature, the pressure, the electrical potential, and their associated flow fields. Although in many specific cases the prediction aim has been fulfilled, water flooding and concentration phenomena, which are relevant at high current densities, are not yet included in mechanistic models. Aim of this presentation is to show under which circumstances the inclusion of the porous structure of all Membrane-Electrode-Assembly regions is essential to qualitatively and quantitatively describe the cell potential at any cell current density. This will be achieved by introducing capillary theory considerations in the diffusion regions, and by considering the interplay between the "diffusion" lengths of the chemical species with the characteristic length scales of a porous medium (pore-to-pore distance, pore size, tortuosity, etc.), especially in the reactive regions. The derived Bernardi-Verbrugge-like model is presented together with extensive applications and comparisons with experimental data. The advantages of the model in terms of solution algorithm performances are also presented and discussed.

NSTI Nanotech 2003 Conference Technical Program Abstract