Characterizing and overcoming spectral artifacts in imaging Fourier-transform spectroscopy of turbulent exhaust plumes

The midwave and shortwave infrared regions of the electromagnetic spectrum contain rich information enabling the characterization of hot, rapid events such as explosions, engine plumes, flares and other combustion events. High-speed sensors are required to analyze the content of such rapidly evolving targets. Cameras with high frame rates and non-imaging spectrometers with high data rates are typically used; however the information from these two types of instruments must be later fused to enable characterization of the transient targets. Imaging spectrometers have recently become commercially available for general scientific use, thus enabling simultaneous capture of both spatial and spectral information without co-registration issues. However, their use against rapidly-varying sources has traditionally been considered problematic, for even at moderate spatial and spectral resolutions the time to acquire a single spectrum can be long compared to the timescales associated with combustion events. This paper demonstrates that imaging Fourier-transform spectroscopy (IFTS) can successfully characterize the turbulent combustion exhaust from a turbojet engine. A Telops Hyper-Cam IFTS collected hyperspectral video from a Turbine Technologies SR-30 turbojet engine with a spectral resolution of δν = 1/cm-1 on a 200×64 pixel sub-window at a rate of 0.3 Hz. Scene-change artifacts (SCAs) are present in the spectra; however, the stochastic fluctuations in source intensity translate into high-frequency "noise." Temporal averaging affords a significant reduction of the noise associated with SCAs. Emission from CO and CO2 are clearly recognized in the averaged spectra, and information about their temperature and relative concentrations is evident.