In-Situ Fatigue Lifetime Modeling of a Reinforced Membrane by Projecting Critical Accumulated Plastic Dissipation Energy from Pressure Differential-Accelerated Mechanical Stress Tests

Abstract

Polymer electrolyte fuel cells (PEFCs) have become increasingly appealing over internal combustion engines because of their high efficiency, low operating temperature, and zero CO2 emissions. Nevertheless, the transportation sector necessitates high durability and reliability, which may be difficult to predict for emerging technologies. A crucial aspect is the ability of the thin membranes that conduct ions in PEFCs to endure the chemical and mechanical stresses that arise during dynamic operations. For mechanical fatigue, temperature and relative humidity (RH) fluctuations induce dynamic stresses that lead to the formation and propagation of microcracks in the membrane. As the membrane is confined by other components in the membrane-electrode assembly (MEA), changes in temperature and humidity can generate thermal and swelling strains in the membrane, leading to dynamic and residual stresses [1].The US Department of Energy sets a passing criterion of 20,000 RH cycles for mechanical fatigue assessment. However, many modern reinforced membranes have already passed this threshold without failing [2]. Therefore, Ref [3] introduced a pressure differential between the cathode and anode sides of the membrane at 80°C to speed up the testing. In this research, the pressure differential-accelerated mechanical stress test (ΔP-AMST) method was applied to a reinforced membrane at two different temperatures and with four times faster humidity cycles in a wider range of ΔPs. This objective is to project the mechanical fatigue lifetime from the ΔP-AMST to the membrane under its in-situ conditions by using the critical accumulated plastic dissipation energy (CAPDE) in ΔP-AMSTs and the plastic dissipation energy (PDE) during a single cycle of humidity that the membrane experiences under complete fuel cell settings. The first step involves performing a series of ∆P-AMST, and in the second step, a finite element model (FEM) for ∆P-AMST based on the developed constitutive model for the tensile tests that covers temperature, humidity, and swelling strain impacts is built, and therefore its S-N curve is extracted. Next, a FEM model for a complete fuel cell is created, and the mechanical fatigue life is estimated by dividing the CAPDE in ∆P-AMST by the PDE in one cycle of the in-situ modeled membrane by considering amplitude stress as a link between the full fuel cell model and ∆P-AMST, as illustrated in Figure 1 and verified by previous studies [4,5]. In the final step, we will also discuss opportunities to integrate the present mechanical fatigue model with a chemical degradation module to simulate its impacts on fatigue lifetime. This integration includes the two main effects of chemical degradation, which are reflected in the thickness of the reinforced membrane [6] and its updated CAPDE [7]. Acknowledgments The authors gratefully acknowledge AVL Fuel Cell Canada and Mitacs for supporting this project. The authors also thank Roger Penn and Amy Nelson for technical advice.