Mathematical and Computational Modeling of Polymer Exchange Membrane Fuel Cells

In this thesis a comprehensive review of fuel cell modeling has been given and based on the review, a general mathematical fuel cell model has been developed in order to understand the physical phenomena governing the fuel cell behavior and in order to contribute to the efforts investigating the optimum performance at different operating conditions as well as with different physical parameters. The steady state, isothermal model presented here accounts for the combined effects of mass and species transfer, momentum conservation, electrical current distribution through the gas channels, the electrodes and the membrane, and the electrochemical kinetics of the reactions in the anode and cathode catalyst layers. One of the important features of the model is that it proposes a simpler modified pseudo-homogeneous/agglomerate catalyst layer model which takes the advantage of the simplicity of pseudo-homogenous modeling while taking into account the effects of the agglomerates in the catalyst layer by using experimental geometric parameters published. The computation of the general mathematical model can be accomplished in 3D, 2D and 1D with the proper assumptions. Mainly, there are two computational domains considered in this thesis. The first modeling domain is a 2D Membrane Electrode Assembly (MEA) model including the modified agglomerate/pseudo-homogeneous catalyst layer modeling with consistent treatment of water transport in the MEA while the second domain presents a 3D model with different flow filed designs: straight, stepped and tapered. COMSOL Multiphysics along with Batteries and Fuel Cell Module have been used for 2D & 3D model computations while ANSYS FLUENT PEMFC Module has been used for only 3D two-phase computation. Both models have been validated with experimental data. With 2D MEA model, the effects of temperature and water content of the membrane as well as the equivalent weight of the membrane on the performance have been addressed. 3D COMSOL simulation results showed that the fuel performance can be improved by using flow field designs alleviating the reactant depletion along the channels and supplying more uniform reactant distribution. Stepped flow field was found to show better performance when compared to straight and tapered ones. ANSYS FLUENT model is evaluated in terms of predicting the two phase flow in the fuel cell components. It is proposed that it is not capable of predicting the entire fuel cell polarization due to the lack of agglomerate catalyst layer modeling and well-established two-phase flow modeling. Along with the comprehensive modeling efforts, also an analytical model has been computed by using MathCAD and it is found that this simpler model is able to predict the performance in a general trend according to the experimental data obtained for a new novel membrane. Therefore, it can be used for robust prediction of the cell performance at different operating conditions such as temperature and pressure, and the electrochemical properties such as the catalyst loading, the exchange current density and the diffusion coefficients of the reactants. In addition to the modeling efforts, this thesis also presents a very comprehensive literature review on the models developed in the literature so far, the modeling efforts in fuel cell sandwich including membrane, catalyst layer and gas diffusion layer and fuel cell model properties. Moreover, a summary of possible directions of research in fuel cell analysis and computational modeling has been presented.