It is known from theory that growing oxide layers on metal surfaces effect interferences of the spectral emissivities.
These interferences can strongly change with increasing thickness of the oxide layer, because their maximum and
minimum values shift to longer wavelengths and their intensities become larger or smaller. The changes of spectral
emissivities also lead to changes of the band-emissivities used as values for emissivity correction of pyrometers or IRcameras.
Primarily, these effects depend only on emissivity and can occur without changes of metal's temperature. For
the working ranges of the optical temperature instruments follows, that the values of band-emissivities may change
constantly with growing oxide layers. This will result in problems for an accurate temperature measurement.
At the University of Duisburg-Essen it is now possible to carry out in-situ measurements of the spectral emissivities of
growing oxide layers on metals. These investigations can be done under technically relevant conditions of heating time,
and temperature. Low alloyed steel with a growing oxide layer was investigated. The steel sample was heated up to
1150°C within a period of 2 minutes and, further, tempered for 1 minute more at this temperature. During this period,
the oxide layer was steadily growing and the changes of spectral emissivities were measured at several increments of
time. It was found, that the theoretically predicted interference effects, their changes and shifts in the spectral properties
are readily seen and could be measured.
Results will be presented of the spectral emissivities from 0.7 to 25 μm and of based on it calculated band-emissivities
between 0.7 to 4.5 μm, which are in correlation with the working ranges of three chosen pyrometers. Strong and
relatively fast changes of the spectral emissivities can be achieved with values between 0.6 and 0.9. Therefore, during
the growth of the oxide layer accurate temperature measurements with optical instruments make no sense because of
the permanent change in the values of band-emissivities. An exact determination of the temperature under these
circumstances is not possible. Only weak influence on the values of band-emissivities were found when correlated with
the spectral response functions of the chosen pyrometers.