Oxide (Cr2O3) scale growth on metallic interconnects and its impact on ohmic resistance: Combined study of image analysis and modeling

2014 
Oxide formation on metallic interconnects (MIC) represents a major source of SOFC stack degradation. Studies of scale growth mechanisms and their relationship with power degradation thus attract major attention in the context of lifetime improvements. MIC degradation is evaluated here by comparing the oxide scale thickness and microstructure evolution with the corresponding ohmic losses, and finite-element- (FE)modeling is used to explain differences between the rates of scale growth and the resistivity increase. The growth of the oxide scales and the simultaneous increase of area specific resistance (ASR) are generally described by a parabolic rate law (i.e. exponent n = 0.5). However, the Cr2O3 scale growth measured by image analysis and as well as the experimentally determined ASR evolution exhibit evident deviations, which are consistently sub-parabolic (i. e. n 0.5), for the ASR evolution. Such deviations become particularly pronounced for investigations > 5000 hours. In order to explain the difference for these two time-dependent behaviors, FE-modeling is performed on electron microscopy images. The distribution of current densities within oxide layers of real samples is simulated and generates ASR values for specific time steps. Deviations between scale growth and ASR evolution are explained by the influence of scale morphology, which is a time-dependent factor. This morphology factor (M) has a high impact during short-terms ( 5000 h) the scale morphology loses its impact on ASR increase because higher scale thicknesses preclude the ‘bridging-effect’. Errors are thus generated when ASR evolutions are predicted only based on growth rates of (measured) average oxide scale thickness. Consequently, for reliable predictions of the ASR evolution based on scale thickness growth, a time dependent morphology factor has to be considered.
    • Correction
    • Cite
    • Save
    • Machine Reading By IdeaReader
    1
    References
    0
    Citations
    NaN
    KQI
    []