This paper addresses optical-based techniques for measuring soot particulate loading in the exhaust stream of gas turbine engines. The multi-angle scattering and multi-wavelength extinction of light beams by ensembles of submicrometer soot particles was investigated as a diagnostic means of inferring particle field characteristics. This is, the particle size distribution function and particle number density were deduced using an innovative downhill simplex inversion algorithm for fitting the deconvolved Mie-based scattering/extinction integral to the measured scattering/extinction signals. In this work, the particle size distribution was characterized by the widely accepted two-parameter log-normal distribution function, which is fully defined with the specification of the mean particle diameter and the standard deviation of the distribution. The accuracy and precision of the algorithm were evaluated for soot particle applications by applying the technique to noise-perturbed synthetic data in which the signal noise component is obtained by Monte Carlo sampling of Gaussian distributed experimental errors of 4, 6, and 10 percent. The algorithm was shown to yield results having an inaccuracy of less than 10 percent for the highest noise levels and an imprecision equal to or less than the experimental error. Multi-wavelength extinction experiments with a laboratory bench-top burner yielded a mean particle diameter of 0.039 μm and indicated that molecular absorption by organic vapor-phase molecules in the ultraviolet region should not significantly influence the measurements. A field demonstration test was conducted on one of the JT-12D engines of a Sabre Liner jet aircraft. This experiment yielded mean diameters of 0.040 μm and 0.036 μm and standard deviations of 0.032 μm and 0.001 μm for scattering and extinction methods, respectively. The total particulate mass flow rate at idle was estimated to be 0.54 kg/h.
