This paper presents a finite element methodology to predict the thermoacoustic eigenmodes of combustion chambers using the linearized Navier-Stokes equations (LNSE) in frequency space. The effect of the mean flow on the acoustics is accounted for. Besides scattering and refraction of acoustic waves in shear layers, this set of equation describes two main damping mechanisms. One is related to the generation of entropy waves, so called hot-spots, in flame regions. The other is related to the transformation of acoustic energy into vorticity waves at sharp leading or trailing edges. Both fluctuation types, i.e. entropy and vorticity, are convected by the mean flow, leading to significant damping when the fluid discharges into an open outlet. In combustion chamber environments these waves are accelerated in the downstream high pressure distributor and are partially transformed back into acoustic waves constituting to the feedback loop of thermo-acoustic instabilities. Accurate prediction of the eigenmodes and eigenfrequencies of instability require therefore to take these interaction effects into account.
First, the accuracy of the LNSE approach, to capture the damping generated by the first mechanism of entropy generation and convection, is investigated for a generic premixed flame configuration. Solutions of the LNSE are compared to the analytic solutions as well as eigenvalues determined by an Helmholtz ansatz. Later methodology assumes a quiescent medium and neglects all interactions of acoustics with the mean flow. It is shown that large errors are introduced with increasing Mach-number.
To illustrate errors assuming a quiescent medium for realistic combustion chambers, the LNSE are used to assess the eigenmodes of a two-dimensional aero-engine combustor including strong shear regions, in the next step. The non-isothermal mean flow field is obtained performing an incompressible RANS simulation. It features an expanding jet with inner and outer recirculation zones. The acoustic computations using LNSE reveal a set of unstable and neutral hydrodynamic modes in addition to acoustic modes. Both damping mechanisms are present and contribute to the overall system stability. Again the obtained solution is compared to the solution of an Helmholtz code and differences are discussed.
When predicting combustion instabilities in gas turbine combustion chambers, the complex geometry and three dimensional flow configurations are often neglected. However, these may have significant influence on the overall acoustic damping behavior of the system. An important element governing the flow inside a combustion chamber is the swirl atomizer nozzle. Therein, the flow is accelerated and a swirling fluid motion is imposed. At its exit considerable high flow velocities are reached and multiple shear layers are formed which discharge into the combustion chamber. To predict damping effects in these environments, acoustic-flow interaction processes need to be taken into account. These involve scattering and refraction of incident acoustic waves in shear layers, acoustic interaction with the unstable hydrodynamic shear layers as well as acoustic wall interaction processes. Their combined effect can be studied using acoustic scattering matrices. In this paper the acoustic scattering behavior of a lean injection system developed by Avio is predicted under non-reactive conditions and compared to experiments. The numerical method is very general and works as follows: First, the fluid dynamic field is computed using a Reynolds averaged Navier-Stokes turbulence model. Then, the linearized Navier-Stokes equations are solved in frequency space around the previously computed mean flow state. The complex three dimensionality of the nozzle configuration is taken into account as well as its corresponding flow field. Results are compared against experimental measurements of a swirl atomizer nozzle at atmospheric and elevated inlet temperatures. It is shown that the scattering behavior and therefore the acoustic-flow interactions are captured accurately.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.