Characterization of the acoustic interactions in a two-staged multi-injection combustor fed with liquid fuel.
Résumé
Burners operating in lean premixed prevaporized (LPP) regimes are considered as good candidates to reduce pollutant emissions from gas turbines. Lean combustion regimes result in lower burnt gas temperatures and therefore a reduction on the NOx emissions, one of the main pollutant species. However, these burners usually show strong flame dynamics, making them prone to various stabilization problems (combustion instabilities, flashback, flame extinction). To face this issue, multi-injection staged combustion can be envisaged. Staging procedures enable fuel distribution control, while multipoint injections can lead to a fast and efficient mixing. A laboratory-scale staged multipoint combustor is developed in the present study, in the framework of LPP combustion, with an injection device close to the industrial one. Using a staging procedure between the primary pilot stage and the secondary multipoint one, droplet and velocity field distributions can be varied in the spray that is formed at the entrance of the combustion chamber. The resulting spray and the flame are characterized using OH-Planar Laser Induced Fluorescence (OH-PLIF), High Speed Particle Image Velocimetry (HSPIV) and Phase Doppler Anemometry (PDA) measurements. Three staging values, corresponding to three different flame stabilization processes, are analyzed, while power is kept constant. It is shown that mean values are strongly influenced by the fuel distribution and the flame position. Using adequate post-processing, the interaction between the acoustic field and the droplet behavior is characterized. Spectral analysis reveals a strong acoustic-flame coupling leading to a low frequency oscillation of both the velocity field and the spray droplet distribution. In addition, acoustic measurements in the feeding line have shown that a strong oscillation of the acoustic field leads to a change in fuel injection, and hence droplet behavior.