Some considerations for a method that simultaneously measures the temperature and emissivity of a metal in a high temperature furnace

This article describes some considerations for designing a practical radiation thermometry system for a glossy metal moving through a high temperature furnace, such as a continuous annealing furnace. In order to accomplish this task, two problems must be solved. The emissivity compensation of an object must be calculated and the furnace's background radiation noise must be eliminated. The authors have proposed a method that uses the radiance's polarized directional properties to simultaneously measure the emissivity and temperature to solve the first problem and a technique using a pseudo-blackbody installed in the furnace to solve the second problem. During heating, there is a one-to-one correspondence between the emissivity and the ratio of p- and s-polarized radiances for metals. This characteristic has successfully led to the development of a method for simultaneously measuring the emissivity and temperature of metals regardless of a potential large change in emissivity. Introducing a pseudo-blackbody radiator into a furnace removes the background radiation noise. Moreover, the blackbody radiator supplies a constant reference radiance. This reference plays an important role in maintaining the principle of emissivity-compensated radiation thermometry inside the furnace. Experimental results have simultaneously measured the emissivity and temperature of stainless steel at 1300 K with errors of 12% and 0.96%, respectively. These values were attained even though the s-polarized emissivities change from 0.25 to 0.75 at a wavelength of 0.9 mum. These errors can be achieved by designing the apparatus to have a solid angle, the aperture of the pseudo-blackbody subtended by a measuring point of the specimen, of more than 0.02pi steradians. The accuracy of this method is heavily dependent upon the specimen's surface roughness. The maximum surface roughness that allows for the successful utilization of this method is Ra=0.12 mum. (C) 2004 American Institute of Physics.