Advanced 3D Fuel Cell Analysis and Condition Diagnostics

In the project, a multichannel, scalable measurement system was jointly developed at TU Graz and AVL, using S++ shunt sensor plates for spatially selective acquisition of voltage, current, spectral impedance, nonlinearities and temperature within the active cell area. In order to differentiate between various failure modes, the THDA technique (Total Harmonic Distortion Analysis) was applied. This method makes use of the fact that the voltage response to small superimposed sinusoidal current signals get distorted and form harmonics under critical conditions in the fuel cell. The analysis of these harmonics and spectral impedance characteristics enables the differentiation between different failure modes that cause performance loss or irreversible cell damage. For the first time this was done on a cell surface.

The multi-physics software package AVL FIRE® was applied to the 3D CFD simulation of an air cooled PEMFC stack. The fuel cell simulation module of AVL FIRE® was extended by a semi-empirical degradation model describing changes of geometry and material parameter as a function of operating time and local operating conditions. The model was developed at TU Graz. The degradation model predicts the fuel cell performance by taking into account the influence of operating conditions on degradation processes in the gas diffusion layer, catalyst layer and polymer electrolyte membrane. In order to predict the mechanical durability of the membrane under humidity cycling, TU Graz developed the necessary theoretical framework that takes into account the maximum deformation energy which the membrane can withstand and considers the influence of humidity cycling conditions on the membrane mechanical properties.

In order to investigate the highly dynamic and nonlinear behaviour during these failure modes, a dynamic large signal equivalent circuit was modelled, parameterized and validated with measurements by TU Graz. The model enables the rapid prediction of specific failure signal patterns based on electrochemical and diffusion kinetics by correlating them with the THDA system response.

For validation and data collection, a test environment for fuel cell stack testing was set-up by means of an open architecture software based on AVL PUMA Open®. This Platform is capable of testing PEMFC-Stacks under real world conditions. Extensive measurements were done at Intelligent Energy and University College London, leading to a significant amount of novel, in-situ data during cell operations. Finally, based on analyses of the data set, the simulation and degradation models were validated and favourable operating strategies with special focus on fuel cell performance and lifetime were derived.