Title: Computationally efficient simulation of polymer electrolyte fuel cells and stacks
Authors : Schumacher, Jürgen
Eller, Jens
Sartoris, Guido
Roos, Markus
Conference details: COMSOL Conference, Grenoble, France, 23 - 24 October 2007
Publisher / Ed. Institution : COMSOL Group
Issue Date: Oct-2007
License (according to publishing contract) : Licence according to publishing contract
Type of review: Not specified
Language : English
Subjects : Fuel cell; Proton exchange membrane; Numerical model
Subject (DDC) : 621.3: Electrical engineering and electronics
Abstract: A computationally efficient model of a proton exchange membrane (PEM) fuel cell is presented. The approach is used for simulation of PEM single cells and stacks. In the framework of this "2+1D"-approach the anodic and cathodic fields are discretised in two dimensions with our finite element software SESES. The coupling between the anodic and cathodic side is established by a one-dimensional model representing the gas diffusion layer (GDL) and the membrane electrode assembly (MEA). The nonlinear 1D model of the MEA and the GDLs is created by starting from the symbolic weak form expressions of the coupled transport phenomena. The tangential stiffness matrix of the coupled FEM problem is computed analytically by the computer algebra software Mathematica. As a next step a numerical FEM solution of the 1D model is calculated with Mathematica. Finally, the 1D model of the MEA and the GDL is implemented with the programming language C and linked to SESES. Coupling between the 1D model and the two 2D models in SESES is achieved by using the values of the degrees-of-freedom (DOF) variables of the 2D model as Dirichlet boundary conditions for the 1D model. The 1D model returns fluxes computed as the non-zero residual at the 1D boundaries and their numerical derivatives with respect to the 1D Dirichlet values to the 2D model. This approach is suitable to take the high aspect ratio between the in-plane and the through-plane dimensions of fuel cells into account. The number of DOF variables is significantly reduced in comparison to a full 3D model approach. Further, this modelling approach can also be applied to other fields, e.g. for the description of coupled electrochemical processes.
Departement: School of Engineering
Organisational Unit: Institute of Computational Physics (ICP)
Publication type: Conference Poster
URI: https://digitalcollection.zhaw.ch/handle/11475/11684
Appears in Collections:Publikationen School of Engineering

Files in This Item:
There are no files associated with this item.


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.