Deconvolution of Temperature and Soot Volume Fraction in a Turbulent Ethylene Flame by Inverse Spectral Radiation Analysis

We report a new non-intrusive diagnostics technique for the simultaneous reconstruction of temperature (T) and soot volume fraction (f_v) profiles in axi-symmetric turbulent luminous flames. Line-of-sight spectral radiation intensities (I_λ) for one diametric and nine chord-like radiation paths from a representative horizontal plane of a turbulent ethylene jet flame were measured by a fast infrared array spectrometer. By inverse analysis of the measured mean I_λ at four wavelengths where continuum radiation from soot particles dominates, four local scalar statistics, including mean and root-mean-square (rms) of T and f_v, were de-convoluted. Powell’s conjugate-gradient algorithm and Brent’s line minimization algorithm were adopted in solving the present four-variable inverse problem. The calculated mean I_λ matched the experimental data very well within a 3% difference in general. The reconstructed local mean/rms T and f_v distributions were in reasonably good agreement with sampling data from similar turbulent flames.