Modeling of PEM Fuel Cells

We model PEM fuel cells on all scales from the electrode structure to the fuel cell system level. Our various modeling methods shed light onto the performance and degradation in the electrode, the flow of reaction gases in the flow field and the optimization of the system behavior. We place particular emphasis on the experimental validation of our models, which provide you with detailed insights into the physical effects during fuel cell operation, with regard to cell performance and aging behavior.

 

We offer:

  • modeling of the catalyst layer, e.g. to investigate local performance and aging as a function of operating conditions, potential or load cycles and the materials used
  • flow simulation and electrochemical modeling (3D-CFD) for the design of flow fields, single cells and stacks
  • system modelling to optimize system efficiency. We can investigate the effects of different operating strategies.
  • use of machine learning algorithms to predict cell performance as a function of various process parameters in CCM production

 

 

Your benefits:

  • Our experimentally validated models provide you with detailed insights into the physical effects during fuel cell operation with regard to performance and aging behavior.
  • Our modeling helps you to determine and select optimal fuel cell materials, geometries and operating conditions.
  • We can assess the effects of changes in materials, operation and design based on our many years of experience. 
Oxygen concentration fuel cell
© Fraunhofer ISE
Oxygen concentration in the through-plane direction of the cathode catalyst layer (CCL) at different applied cell voltages. Oxygen enters the CCL at the gas diffusion layer/CCL interface (at CCL depth 0 µm) and diffuses toward the CCL/membrane interface (at CCL depth 12 µm). At working points with low cell voltage and correspondingly high cell current densities, oxygen starvation occurs around a CCL depth of 10 µm.
Ageing simulations for three different fuel cells
© Fraunhofer ISE
Degradation simulations for three fuel cells with different platinum to carbon ratio. The cells are subjected to an accelerated stress test protocol consisting of 30,000 potential cycles between 0.6 and 0.95 V (DoE catalyst ageing). Each color represents one cell. Solid lines and bars denote the status at begin of test, dashed lines and hatched bars the status after ageing.
Neuronal Network trained to forecast fuel cell performance
© Fraunhofer ISE
A neuronal network trained to forecast fuel cell performance over degradation, ink and MEA production properties and prior characterizations.

Research Projects on the Topic Modeling of PEM Fuel Cells

 

FC-CAT

Fuel Cell CFD and Through-Plane Modeling

 

FC-RAT

Fuel Cell Realistic Aging Trend Modelling

More Information on this Topic

R&D Infrastructure

Fuel Cell Lab

Research Topic

Fuel Cell