Analysis of the dynamic response of premixed flames through chemiluminescence cross-correlation maps

This work proposes a novel methodology to extract useful information on flame dynamics from instantaneous chemiluminescence images in terms of the 'effective local response', by means of the so-called cross-correlation maps (CCM). Theoretical considerations suggest that CCM presents some advantages with respect to other options: it allows filtering out some spurious heat release rate fluctuations, not related to the acoustic excitation (either natural or forced) and generates a map of the 'effective gain', which can be interpreted as the actual contribution of each flame parcel to the global dynamic response. The method has been applied to an experimental dataset collected for premixed V flames of methane and CO2/methane blends, covering a wide range of operating conditions. The results are fully consistent with the physical interpretation proposed for the cross-correlation maps, and confirm their potential for diagnosing the effective contribution of the different flame regions to the global dynamic response, as quantified, for example, in terms of the flame transfer function. With further hypotheses (e.g., n-τ formulation), CCM can provide further information about some features of the dynamic flame response, such as the characteristic flame length related to convective time lag and, hence, to the phase of the FTF. Cross-correlation maps are also compared with the spatial distribution of the local Rayleigh index, revealing a qualitatively similar pattern but also some essential differences, related to the different nature of both magnitudes. CC maps reflect the intrinsic dynamic response of the flame, whereas the Rayleigh index is related to the spontaneous instability and, hence, depends on the coupling between flame and the rest of the system.