Simulation von Brennstoffzellen am Fraunhofer Institut für Solare Energiesysteme

Mathematical modeling with the aid of partial differential equations is increasingly being used in many areas of chemical technology and above all to estimate and assess a large number of physical-chemical phenomena. The Fraunhofer Institute for Solar Energy Systems of Freiburg models fuel cells with the aid of a commercial finite element program. The success of modern modeling software resides in its flexibility. Traditional finite element or finite volume packages are based on programmed multi-physical couplings that often are tied together and in so-called coupling elements or by predetermined user programs. Because however not all possibilities of coupled phenomena can be predicted, a suitable software enables equation based input and also can solve equation systems with almost all multi-physical couplings that are possible. The modeling of fuel cells is a multi-physical task because it encompasses electrostatics, flow dynamics, mass and heat transport, and electroc hemical reactions. The coupling in such a model is often a high degree non-linear equation. The Fraunhofer Institute is interested in the distribution of the current density in proton exchange membrane fuel cells. One of the greatest challengers is the development of the fuel cell stacks because they control energy losses and efficiency of the fuel cell. A discussion covers several models that have been developed using the finite element package FEMLAB. The studies presented indicate the ability to perform of equation based programs such as FEMLAB in the treatment of non standardized multiphysical couplings. Equation based modeling offers a quick and efficient method to make trustworthy catalyst structures and electrode kinetics because the kinetic expressions can be placed directly into the balance equations. The fact that the equations can be completely coupled and not be calculated individually as is often the case in defined functions of the traditional programs, provides numerical stability and