Electrochemical impedance spectroscopy (EIS) is an extremely important tool for in-situ measurements and modelling of fuel cells (FC). Despite long research efforts in the field of EIS modelling, the interpretation of experimental data remains a challenge as physicochemically consistent models, which use parameters with a clearly defined physical meaning, have no yet to be fully developed. With the purpose of accellerating the developement in this field, in this thesis an implementation of such a physically consistent model is described. The model is valid for all current densities and takes into account the cathode and the anode side of the FC. The developed model was interconnected with a time domain model with state-of-the-art extrapolation capabillities. The model was verified and validated, reduced versions of the model were tested for accuracy and optimal design of experiments (DoE) was conducted. In addition, an analysis of sensitivity of calibration parameters with respect to frequency was conducted and the influence of each seperate operational condition was asessed. The results show a high accuracy of fit as the R-squared parameter never falls below 0.965. The model also shows good extrapolation capabilities. DoE has shown that experiments with fewer measurements can yield much higher information of calibration parameters if they are meticulously planned. This proposed methodology is used with the aim of reducing the number of required measurements of impedance spectra for acquiring a sufficient ammount of information about individual calibration parameters.
|
