Investigation of the onset of acoustic instability in turbulent combustion systems

点击翻译按钮获取中文摘要

Combustion instability is a major problem in the design and development of the combustor. The key elements in the study of this phenomenon are a model for how the flame structure responds to acoustic oscillations, solutions for the acoustic quantities in the combustor for the given flame response and appropriate acoustic boundary conditions at the combustor inlets and exit. Available literature indicates either a system approach, wherein flame transfer functions are obtained experimentally, or with rudimentary flame dynamics models. A few other works have performed computational simulation of compressible flow, in order to predict dominant acoustic modes excited in the combustor. While the former approach is either empirical or rudimentary, the CFD approach is either lacking in details of the turbulent combustion process or is too expensive. Moreover, there have been no investigations on combustors under conditions that approach instability-like conditions, in order to investigate the onset of this phenomenon. The proposed work is based on a 'tunable' combustor geometry developed at IIT Madras, India, that can be made to excite oscillations over a range of frequencies from broadband, relatively low amplitude roar (resembling noise) to discrete, high-amplitude tones (resembling instability conditions). The goal of the present proposal is to identify the operating conditions that trigger the discrete high-amplitude tones in a more systematic manner than has been done so far, and to computationally simulate turbulent premixed and diffusion flames under these conditions using large eddy simulation at the Fachgebiet Energie- und Kraftwerkstechnik (EKT) of TU-Darmstadt, Germany. Detailed laser diagnostic measurements on instantaneous velocity field by PIV and temperature measurements by Rayleigh scattering would be performed at IIT Madras and TU Darmstadt respectively, in order to validate the LES simulations. One dimensional acoustic calculations will be combined with the LES results in order to predict the acoustic pressure field, and the acoustic field would be coupled to the LES code in a feedback loop. The proposed work is expected to enable a better understanding of acoustic instabilities induced by turbulent combustion in engines.