Noise Immune Absorption Profile Extraction for the TDLAS Thermometry Sensor by Using an FMCW Interferometer

Tunable diode laser absorption spectroscopy (TDLAS) techniques usually utilize direct absorption spectroscopy (DAS) to measure absorbance profiles for temperature measurements. The DAS method is sensitive to thermal noises and fails to work in harsh application cases. A noise-immune method is proposed for robust temperature measurement by using coherent detection and retroflected optical alignment. The absorption spectrum is shifted to a high-frequency band of MHz level away from ambient noises. The optical alignment suppresses thermal radiations along the laser path. Gas temperatures were measured in cases of both laboratory and porotype setup of aeroengine, i.e., the steady and the harsh states. Transition lines of H2O at 7182.94, 7185.59, and 7444.35 cm−1 were selected for performance evaluations of target molecules, i.e., of water vapor. Performance comparisons were made with the DAS method for temperature detection. In a steady-state case, standard deviations of the proposed method were below 0.9 K at SNR of 25 dB, while those of the DAS method were above 3.9 K. In an onsite harsh application case, the fluctuations of path-averaged temperatures obtained by the proposed method were also smaller. The coherent detection and retroflected alignment enable better performance in hostile cases.