Spectral remote sensing for furnaces and flames

Abstract

The spectral remote sensing (SRS) technique to measure the gas temperature profile in a combustion system is introduced from fundamental concept to lab-size bench test. As the bases of the technique, computationally working models of narrow, band emission in the CO2 4.3 mu-wide band are tested, and inversion algorithms to convert the spectral intensity data into a temperature profile are proposed. Actual applications to three lab-size test benches are made to confirm the practical applicability of the technique. With the latest inversion methods and instruments, the SRS technique can retrieve the temperature profile within a 20-30 K error for peak temperature of 1500 K within seconds. A scheme called BCIM incorporating base function approach and nodal temperature approach is recommended. It is also shown that a pure concentration problem cannot be handled by SRS. However, SRS can be applied to combustion flames thanks to a strong temperature/concentration correlation.