Model-Based Analysis of PFSA Membrane Mechanical Response to Relative Humidity and Load Cycling in PEM Fuel Cells

Abstract

The hydration/dehydration cycling of the membrane during the fuel cell operation and the resulting mechanical stress are in part responsible for the mechanical failures of a polymer electrolyte membrane. To thoroughly investigate the mechanical behaviors of the membrane under in-situ cyclic conditions, in this paper, we have interfaced a comprehensive two-dimensional transient fuel cell transport model with a viscoelastic-plastic membrane mechanical model. The transport model is used to produce the spatiotemporal profiles of membrane water content and temperature in an operating cell, which are then sent to the membrane mechanical model to calculate the mechanical parameters of interest. This provides the extended capability of studying the membrane mechanical response under in-situ conditions, which was not possible with the original mechanical model. The effects of cycling relative humidity and voltage and current at different temperatures on the membrane stresses are studied using the coupled model. It is found that the location of maximum in-plane tensile stress can vary significantly with the operating temperature during voltage cycling, whereas the highest in-plane compressive stress occurs under the land for all cases, particularly near the cathode. The simulation results confirm the need for coupling the two models to capture comprehensive transport phenomena in studying the membrane mechanical behaviors, and represent an important step toward improved understanding of various synergistic mechanical failure mechanisms that affect the membrane in an operating fuel cell. Mechanical failure of the polymer electrolyte membrane in the form of cracks, pin-holes and delamination has been identified as a limiting factor in the durability of the fuel cell. The hydration/dehydration cycling of the membrane during the fuel cell operation and the resulting mechanical stress are in part responsible for the mechanical failures.1,2 For example, the cyclic mechanical stresses have been shown to play a significant role in propagating a crack through the thickness of a membrane.3 Recent in-situ experiments have shown that membrane damage can be dominated by these cracks. 4 Since operation of a fuel cell causes hydration/dehydration cycling along with spatiotemporal variations of temperature that can influence the membrane response as well, 5 it is critical to thoroughly understand the mechanical behavior of the membrane under in-situ conditions, so that alleviating strategies focusing on the fuel cell operating conditions can be developed.Previously, a 2D viscoelastic-plastic membrane model of a representative fuel cell unit volume has been developed in ABAQUS and preliminarily validated. [6][7][8][9][10] In this prior work, the hydration states (water volume fraction) and temperatures at the interface of gas diffusion media and the gas channels are the model inputs. The spatiotemporal water transport in the electrodes and membrane is governed by diffusion only, without the in-situ features including electro-osmotic drag ...

Through-the-Membrane Transient Phenomena in PEM Fuel Cells: A Modeling Study

“…Therefore, this model is 98 expected to offer higher fidelity than the state of the art models for 99 the purpose of studying transient phenomena that impact performance and durability. 21 ε…”

“…Therefore, this model is expected to offer higher fidelity than the state of the art models for the purpose of studying transient phenomena that impact performance and durability. 21 The rest of the paper is organized as follows: first, the model formulation is presented along with a detailed review of the literature relevant to each sub-model to justify the choices made during model development. Simulation results and discussions are provided next, followed by a brief summary and concluding remarks.…”

A Mathematical Model toward Real-Time Monitoring of Automotive PEM Fuel Cells

“…21 Moreover, it has enabled a closer look at some of the degradation issues that limit the lifetime of the stacks. 22,23 Nevertheless, this all comes at a significant computational cost that inhibits the use of these models for on-line estimation and diagnostics. On the other hand, the available computationally efficient models typically disregard or oversimplify the underlying physics to the point that they miss many critically important phenomena, such as the twophase flow 24 and non-isothermal effects.…”

A Mathematical Model toward Real-Time Monitoring of Automotive PEM Fuel Cells

“…This ever-increasing complexity of the models has shed some light into the transport phenomena in the porous layers of the cell that were previously unidentified [21]. Moreover, it has enabled a closer look at some of the degradation issues that limit the lifetime of the stacks [22,23]. Nevertheless, this all comes at a significant computational cost that inhibits the use of these models for on-line estimation and diagnostics.…”