Evaluation of Thermoacoustic Instability for Chemically Reacting Flows Using Large-Eddy Simulations

Abstract

Thermoacoustic instability for chemically reacting flows was investigated using large-eddy simulations coupled with a lookup table for turbulence-chemistry closure. The onset of instability was evaluated from pressure fluctuations, as well as standard and extended Rayleigh criterion, as suggested in literature. Two configurations were considered, namely a canonical Rijke tube and a simplified can combustor with a swirling flow injector representing a complex generalized geometry. For the Rijke tube, premixed and non-premixed combustion models were applied for identical fuel flow rate, resulting in different thermoacoustic outcomes due to differences in reaction rates of the two flame regimes. Results from the Rijke-tube case agree with analytic thermoacoustic theory. For the can combustor, only premixed chemistry was considered as it better represents the experimental conditions, and the first resonant pressure mode aligns reasonably with published experimental data. Findings suggest that, if thermoacoustic instability is detected, the resonant frequency can be deduced from the fluctuations of the pressure, heat release, or acoustic source term. However, even though the resonant frequency is correctly identified, fluctuation data alone is insufficient to identify the onset of thermoacoustic instability, requiring the additional application of Rayleigh criterion. Finally, this study concludes that, for the range of configurations studied here, the standard Rayleigh criterion is sufficient to determine the onset of thermoacoustic instability, so the extended Rayleigh criterion is not always necessary, in contrast to suggestions from previous work. This conclusion is significant because the standard Rayleigh criterion is the only practical evaluation for physical experiments.