Abstract
Proton-exchange membrane fuel cell degradation is one of the factors that limit the deployment of this technology in the automotive market. Accelerated Stress Tests (ASTs) are the standard tool for assessing durability during design and development phases, thanks to their shorter duration, but there are concerns about their representativeness of degradation phenomena appearing in real driving conditions. The present study couples a semi-empirical multi-layer degradation model with a validated reduced-order physical model of the cell to evaluate degradation occurring in both kinds of conditions. First, the phenomenological model is used to predict the decay in the polarization curve when running continuous real driving cycle and accelerated stress profiles up to a total of 1000 h. Then, the physical reduced-order model is calibrated to the produced polarization curves at each cumulative time by introducing three degradation factors: two associated with the electrochemical surface area of the anode and cathode catalyst layers, and another one related to the membrane conductivity. Additionally, the results of the physical model are evaluated to identify the occurrence of local conditions that can induce degradation. The results show that a properly designed accelerated stress test provides approximately 10 times faster degradation while maintaining similar degradation factors and physical representativeness.