PROGRAM | Mechanical Engineering

By: Mohammad Morshed Moshiur Hasan Chair: Michael H. Santare

ABSTRACT

Limited durability remains one of the key hindrances to the widespread commercialization of polymer electrolyte membrane fuel cells (PEMFCs). Perfluorosulfonic-acid (PFSA) membranes, commonly used as solid electrolytes in PEMFCs, degrade chemically, mechanically, and/or due to the synergistic effects of the two. This dissertation investigates the mechanical durability of the PFSA membranes through characterizing and predicting fatigue crack initiation and propagation using several numerical schemes.

In operating PEMFCs, the alternating nature of the electrochemical loads causes variations in the water, temperature, and pressure distribution in the membrane. These variations cause the hydrophilic membrane material to swell and de-swell, inducing alternating stresses which can induce fatigue damage. This damage can gradually degrade the material leading to fatigue crack propagation and eventual failure of the membrane. This dissertation develops and details four numerical schemes to (i) approximate constitutive material behaviors of reinforced PFSA membranes, (ii) capture membrane fatigue stresses in the PEMFC in response to various cyclical electrochemical loads and, (iii) investigate fatigue crack initiation and propagation in PFSA membranes under a range of fuel cell operating conditions.

Since the environmental conditions in operating PEMFCs continually change with the electrochemical loads, modeling the membrane’s in-situ mechanical response requires an environment-sensitive constitutive material model. Typically, extensive experimentation is required to fully characterize membrane materials under a range of environmental conditions. However, this dissertation outlines a methodology to develop an approximate material model with a limited amount of testing. This is done by simulating limited experiments of a state-of-the-art reinforced membrane and approximating its behaviors for ranges of temperature, humidity, and loading conditions using existing test results for similar materials.

In another investigation, mechanical responses of a commercial PFSA membrane are explored in the PEMFC for various electrochemical loads. This research couples two separate models- a transport model created by the researchers at the University of Michigan and a mechanical model created by the researchers at the University of Delaware- to translate the complex electrochemical phenomena, occurring during fuel cell operation, into membrane mechanical stresses. Simulation results reflect that voltage and current fluctuations in the fuel cell play an important role in inducing mechanical stresses.

To understand environmental sensitivity of fatigue crack resistance in a commercial PFSA membrane, a numerical scheme is developed to simulate ex-situ fatigue crack propagation under various fuel cell operating conditions. This model calculates fatigue crack resistances and, uses them to predict crack growth rates under the given conditions. The predictions are then verified with experimental data reported in the literature. Simulation findings reflect that fatigue crack resistance is sensitive to temperature and humidity, showing a decrease with temperature and a nonmonotonic interplay with humidity.

Lastly, a finite-element-based computational framework is developed to predict fatigue crack initiation on the membrane surface from a pre-existing electrode crack in the PEMFC. This model calculates in-situ fatigue plastic energy production and relates this to the measured fatigue lifetimes of a state-of-the-art reinforced membrane under three different mechanical accelerated stress tests (ASTs). The predicted fatigue lifetimes are found to be surprisingly close to experimental failure times. Simulation results show that holding the membrane at a constant humidity for a longer time increases mechanical degradation per cycle. Using this model, temperature, humidity, and rate of change of humidity have been varied in hypothetical operating conditions to understand their individual as well as combined impacts on the membrane mechanical degradation. Within the range of studied conditions, temperature is found to have a greater impact on damage than humidity changes or humidity rate changes.

These models have the potential to screen membrane mechanical durability under real-life fuel cell operating conditions and thereby, can reduce experimental resources and time. Parallel to that, these models can contribute to understanding membrane material behaviors and developing new materials.

Back >