Influence of coherent structures on Heat Release and Noise Production of Premixed Flames

I. Background and Motivation

Turbulent flows play a significant role in the field of power generation. The noise of combustion systems is quite an important problem in the industry, which manufactures stokers and gas turbines. The stricter noise regulations of airports demand the development of more silent engines. Next to higher efficiency in terms of combustion processes, turbulent flows applied in modern burners help to increase technical reliability and enable a reduction of pollutant gases at the same time, i.e. NOx. For these reasons, almost all technical applications today use turbulent flows with a strong swirl. Furthermore, combustion noise can cause engine failures. The interaction of turbulence, mixture and reaction is an important subject in combustion research, particularly since environmental aspects became more and more important. New methods of high speed investigations allow us to acquire both the detection of the flame front and the resulting velocity field. Our research of noise will be concentrated on open flames first and later on flames in closed conditions. In most areas, numerical simulation techniques based on Reynolds averaged Navier-Stokes equations (RANS) are primarily used to design and investigate new applications. Since only averaged quantities are calculated here and all fluctuations are represented by an appropriate model based on mean values, the computational cost is relatively low compared to techniques which include an explicit computation of low frequent phenomena, i.e. Large Eddy Simulation (LES). As a consequence, numerical simulations are frequently many orders of magnitude less expensive than experimental research. Furthermore, a variation of geometrical quantities and different configurations are much easier to realise than in experiments.

II. Project Goals

Noise is produced in turbulent flames by local temporal fluctuations of volume change. Hence, alterations are caused by chemical reactions and convective transport of turbulence into the flame front. The broadband continuous frequency spectrum of the turbulence results in a broadband production of noise. Near the flame, a complex mechanism of propagation and interference of acoustic waves occurs. A more global and easier approach permits treating the flame effect as a source of acoustic waves. It has been shown that a flame, as a source of combustion noise, has the nature of a monopole, respectively, a collective of monopoles with different intensities and frequencies. Thus, the noise emission does hardly depends on the direction and can be described by a pulsating sphere. The main goal of this project is to develop a model, which can compute the spectral connections between the turbulence spectrum in the reaction zone and the produced noise spectrum in premixed jet and swirling flames. In addition the model should be able to compute in which case coherent structures arise in burners and how the changing of the volume fluctuations is occurred and thereby how the turbulent noise spectrum is modified. The research has to cover flames with different intensity of the coherent structures. For this the swirl number – except the jet flame – is varied within a wide range. The results of the calculations serve to formulate rules for the noise reduction which are useful for simple application in practice, without extensive numeric investigations.

This project is funded by Bundesministerium für Bildung und Forschung - BMBF -