Determination of the gas temperature profile in a large-scale furnace using a fast/efficient inversion scheme for the SRS technique

Abstract

Measurements of gas temperature profiles in a large-scale test furnace are made by inverting the measured spectral data at 4.3mum band with inversion algorithms. A 3.4m-long stainless steel tube is employed as the test furnace through which flue gas flows. Three inversion algorithms are tested: modified constrained inversion method (MCIM), base function-based inversion method (BFIM) and BFIM-based MCIM (BCIM). The correlated-k (CK)-based WNB model is employed for calculation of spectral intensities during the inversion process. The overall performances of algorithms are scrutinized with the inversion results. To verify the accuracy, the inverted temperatures are compared with measured ones using thermocouples. BFIM is good for initial guess and MCIM is good for fine touch, while BCIM gives fairly good results for all cases considered. Thus, BCIM is considered to be applicable, with good accuracy and capability, to inverse problems in large-scale furnaces. (C) 2004 Published by Elsevier Ltd.