Filtered Rayleigh scattering thermometry in premixed flames, Part II: Implementation and accuracy in turbulent flames

Abstract

An important issue for conversion of measured filtered Rayleigh scattering (FRS) signals into temperature in combustion systems is the degree to which the local chemical state needs to be known or approximated. To avoid the need for simultaneous multi-species measurements, Part I of this paper series introduced a framework based on state relationships from one-dimensional laminar flame simulations to account for composition effects. Following the demonstration of accurate and reliable FRS-based thermometry under laminar flame conditions (Part I), this paper uses both computational and experimental data to evaluate the accuracy of this methodology in turbulent premixed flames. First, synthetic results from direct numerical simulations (DNS) of CH₄/air flames under varying turbulence levels are used to directly evaluate the proposed FRS approach. Results demonstrate accurate temperature determination using the proposed FRS methodology with uncertainties of less than 3%. Analysis shows that, even when the local species composition fluctuates significantly in space and time, the conditional relationship between a given species and temperature shows relatively little scatter compared to the most probable value. Furthermore, there exists an apparent stable behavior where the under-prediction of some species by the presumed state relationship is counter-balanced by an over-prediction of other species in such a manner that the composition-dependent portion of the FRS signal-temperature relationship remains relatively unchanged and well represented by the laminar flame state relationship, regardless of temperature or turbulence level. Experimental results are also presented in a series of turbulent premixed flames with variations in reactant composition, equivalence ratio, and Reynolds numbers. Simultaneous laser Rayleigh scattering (LRS) and FRS measurements converge to common temperatures, indicating accuracy of the proposed FRS methodology within turbulent premixed flames. Overall, results demonstrate that laminar flame state relationships sufficiently capture the net effect of species formation, mixing, and transport with respect to their influence on the local FRS signal. This allows accurate FRS-based temperature measurements in turbulent premixed flames without the need for additional species measurements.