Cold start processes are of high importance for the investigation and development of PEM fuel cell systems in dynamic operation. In particular, the formation of liquid water and condensate at low operating temperatures influences the performance and ageing process of fuel cells. Therefore, an existing fuel cell model with a 2D-1D resolution from the previous project (Fuel Cell System Simulation - Membrane Humidification Management, FVV project number 6012982) is extended at the ZBT by the consideration of liquid water and the differentiation of diffusion effects in the channel and web area. In order to ensure real-time capability, a predictive method for determining the starting conditions of the deterministic, numerical solver will be developed using the machine learning approach. At the Institute for Combustion Engines (vka) of the RWTH Aachen University, the resulting fuel cell model is integrated into a complete system simulation and supplemented by detailed multi-dimensional CFD simulations.These CFD simulations focus on the investigation of the water distribution within the cell layers, which is of particular interest for cold starts due to phase change phenomena as well as condensation and evaporation. The simulation results are afterwards validated at ZBT by means of experimental measurements on segmented cells in the cold start test rig. Furthermore, the fuel cell system is being developed at vka with special attention to the membrane humidifier, which is the critical peripheral component with regard to cold start.
In this project a model is developed which is able to investigate the cold start behaviour of PEM fuel cells in a system environment. This model thus enables a faster and more cost-effective development process of fuel cell vehicles.
