Testing several combinations of structured grid and model shows that the models do not work with every grid, e.g. H-grid, O-grid or butterfly grid, i.e. the computation does not converge in many cases. The choice of the right type of grid and turbulence model needs a lot of experience, time and effort. Hence, it is strongly recommended to test meshing procedure, grid type and model with a simplified manifold or fuel cell fluidic system before a CFD-simulation of a (real) fuel cell stack is tried. Furthermore, experimental validation of the CFD results is necessary, as will be shown in the following sections.