No abstract

The purpose of the first part of this study was to compare four different temperature measuring methods. The application of these tools for possible temperature monitoring or calibration of monitors of microtubular solid oxide fuel cells (MT-SOFCs) is explored. It was found that a thermographic camera is very useful to visualize the temperature gradient on the outside of a cell, while an electrochemical impedance spectroscopy method was useful for estimating the core temperature of a test cell. A standard thermocouple was also used in combination with the previous two methods. Furthermore, an inexpensive laser guided thermometer was also tested for MT-SOFC temperature measurement. This initial study has opened up a range of questions not only about the effect of the experimental apparatus on the measurement results but also about the radial temperature distribution through a MT-SOFC in a working mode. Both these topics will be further investigated in part II of this study through a computational fluid dynamics study. This should provide additional interesting information about any differences between testing single cells and those within a bundle of cells. The discussed results are expected to be mainly temperature related, which should have direct consequences on power output and optimized gas inlet temperatures.

Micro-tubular solid oxide fuel cells (MT-SOFCs) are a much smaller version of larger tubular SOFCs. They are operational within seconds and allow a higher power density per volume than the larger version. Hence they are a potential technology for automotive, auxiliary and small scale power supply devices. In this study a commercially available computational fluid dynamic (CFD) software program was used to predict a MT-SOFCs performance when located inside a high temperature wind tunnel experimental apparatus. In Part I, experimentally measured temperature profiles were recorded via thermo-graphic analyses and I/V curves. These measurements were used in this study to establish the predictability and validity of the CFD code and furthermore understand the MT-SOFC attributes measured in Part I. A maximum 4% I/V curve deviation and 6 K temperature deviation between the experimentally measured and model predicted results was observed. Thus, the model predicted the MT-SOFCs performance in the experimental environment very accurately. A very critical observation was the current density and temperature profile across the MT-SOFC that was strongly dependent on the distance from the hydrogen/fuel inlet. Not only was the model validated but also a grid and quantitative solution analysis is explicitly shown and discussed. This resulted in the optimum grid density and the indication that a normally undesirable high grid aspect ratio is acceptable for similar MT-SOFC modeling. These initial simulations and grid/solution analysis are the prerequisite before performing a further study including multiple MT-SOFCs within a stack using different fuels is also envisaged.

Three dimensional (3D) computer aided X-ray tomography (CT) has proven to be an extremely useful tool in developing our own as well as in examining commercially available solid oxide fuel cells. The results of 3D-CT measurements became very important for understanding the functionality of our first generation and improving the development of our second fuel cell generation. Also geometrical measurements, especially the roundness and the straightness of the tube, can be evaluated, both critical parameters when the stack is heated and mechanical stress has to be avoided. By using this technique the structure of the first generation cells proved to be of insufficient quality. Problems like the variation in thickness of the electrolyte tube as well as the homogeneity in thickness of the electrodes deposited can easily be detected by this nondestructive technique. Microscopic investigations of this problem of course provide equal results, but only after cutting the samples in many slices and many single measurements of different areas of the fuel cell. Using cells with inhomogeneous thickness of course results in drastic variations of the current densities along a single cell. Electrolyte layers that are too thick will result in power loss due to the increased resistance in the ionic conductivity of the electrolyte. If the electrolyte of an electrolyte supported cell is too thin, this can cause mechanical instability. Problems can also occur with the leak tightness of the fuel cell tube. Gas diffusion through the electrode layer can become a problem when the thickness of the electrode layer is too high. On the other hand, if the layers are too thin, the result can be a discontinuous layer, leading to a high electrical series resistance of the electrode. Besides determining the thickness variations also the porosity of the electrolyte needs careful attention. Larger cavities or shrink holes form insulating islands for the ion-stream and are therefore limiting the ionic conductivity. They are also diminishing the mechanical stability and provide problems for depositing a closed electrode film in electrode supported cells.

No abstract