No abstract

The present paper describes a Large Eddy Simulation modelling framework for the simulation of oscillating flames in practical flow configurations. The unresolved subfilter scale motion is modelled using the dynamic Smagorinsky model in combination with the Probability Density Function method. It is shown that the Large Eddy Simulation method is capable of reproducing the characteristic shape of the reaction zone as well as the nonlinear evolution of the total heat release rate in a bluff-body stabilised combustor. Commonly used measures for quantifying the variation of the total heat release rate are evaluated and examined in the present flow configuration of a lean-premixed ethylene-air flame. It was found that formaldehyde-based measures do not appropriately reproduce the amplitude and phase of the total heat release rate. A significantly improved correlation was achieved by employing the product of the mass fractions of molecular oxygen (O 2 ) and the ketenyl radical (HCCO) as a means of characterising the variation of the total heat release rate.

Modern low-emission gas turbine combustion systems often experience thermo-acoustic instabilities at certain operating conditions, which adversely affect the performance of the engine. One way to mitigate the detrimental effect of such instabilities is to place passive damping devices along the wall of the combustion chamber. To achieve greatest overall damping requires good understanding of the acoustic properties of the damping devices at engine conditions and determination of the undesired acoustic mode shapes for optimal placement at the wall. This paper presents an experimental study which characterises the acoustic properties of bias flow liners operating at frequencies in the low kilohertz regime (> 1 kHz). The engine conditions are simulated in the experiment at ambient conditions by maintaining dynamic similarity, i.e. by matching a number of non-dimensional parameters in the experiment which characterise the engine conditions. The present experimental study contributes to the existing measurement database by taking into account the strong gradient in characteristic impedance between grazing and bias flow medium. The acoustic properties of the investigated damper configurations are assessed in terms of the surface impedance at the interface between grazing and bias flow. An impedance model is suggested which accounts for the strong gradient in characteristic impedance between grazing and bias flow medium. The impedance model may serve conveniently as input to an acoustic mode shape prediction in the combustion chamber to identify the optimal placement of the damping devices.

No abstract