FDF-based combustion instability analysis for stabilization effects of a slotted plate in a multiple flame combustor

In the present study, new promising methods are suggested to analyze thermoacoustic instability and the stabilization effects in the multiple flame combustor with a slotted plate. Using the generalized regression neural network (GRNN), the flame describing function (FDF) is effectively modeled from a limited number of experimental data. This neural-network based FDF method is able to generate more refined FDF data in an extended range. These refined FDF data are utilized in a Helmholtz solver for thermoacoustic instability analysis. According to the velocity perturbation ratio, eigenfrequencies are investigated to know the unstable regimes of the combustor. To take account of the effects of plate thickness, the present approach has slightly modified the Dowling method for modeling the impedance of a slotted plate. To find the effective damping conditions of a slotted plate, parametric studies have been carried out with the help of simulated annealing (SA) algorithm in wide-range operating conditions. It is identified that the absorption bandwidth becomes wider by decreasing slit width, and narrower width yields the higher average absorption coefficient. All the numerical results confirm that these new methodologies are quite reliable and widely applicable for the analysis of combustion instability encountered in many practical combustion systems.