Influence of Temperature Gradients on the Sound radiated from Flames.

Technische Fachhochschule Berlin, Univ. of Applied Sciences, Luxemburger Str. 10, 13353 Berlin, Germany, .
The Journal of the Acoustical Society of America (Impact Factor: 1.5). 06/2008; 123(5):3405. DOI: 10.1121/1.2934111
Source: PubMed


The far field pressure of a turbulent flame can be determined using the standard boundary element method (BEM) if the sound pressure or its derivative is known at a closed surface (control surface) surrounding the flame, as long as the medium outside the control surface is homogeneous. If temperature gradients are present, the homogeneous Helmholtz equation is no more valid. In that case, the wave equation can be rewritten in form of an inhomogeneous Helmholtz equation with a source term that depends also on the unknown pressure. Using the "Dual Reciprocity BEM" the integral form of this wave equation can be solved involving only surface integrals, so that the sound field can still be computed from field values at the control surface. The cases under study consider a volume of hot gas with a temperature distribution that is prescribed or obtained from a CFD simulation. The influence of the temperature gradients on the sound field can be evaluated by comparison of characteristic quantities like sound power and radiation patterns, with and without temperature gradient.

1 Follower
11 Reads
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Hybrid methods, coupling CFD codes and acoustic methods like the acoustic analogy, the linearised Euler equations and the Kirchhoff-Method are being used to predict the sound produced by turbulent combustion, since the present computational power only allows an accurate estimation of the physical quantities, in a reasonable computational time, near the source. This means that the acoustic far field cannot be determined by a direct simulation. This research attempts to show that the Equivalent Source Method (ESM) and the Boundary Element Method (BEM), which are considered as Kirchhoff-Methods, can also be used to deter-mine the sound generated by combustion. These two methods have the advantage that only one acoustic variable must be known at a surface surrounding the source zone and that the far field can be directly com-puted. The sound power generated from two open diffusion flames have been calculated with both the ESM and the BEM, using the velocity distribution over cylindrical control surfaces computed with a Large Eddy Simulation. Results of the calculations are presented and compared with the measured sound power of the same flames. For one configuration good agreement between measurement and simulation at low and middle frequencies is obtained. Possible reasons for the differences for the other configuration will be discussed.
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Equivalent sources have been successfully used to calculate the sound radiation and the sound scattering from solid bodies lying in a homogeneous medium without flow. For the field determination an acoustic boundary condition at the body surface must be known. In an earlier work, the application of this method to compute the sound radiation of open turbulent flames was investigated in order to extend the range of use of this basic method to aero-thermoacoustic problems. It was assumed that outside a region surrounding the flame, the flow and temperature gradient had strongly decayed and approximate homogeneous conditions existed. The necessary data at a control surface (Kirchhoff surface) surrounding the combustion zone was delivered by an incompressible Large Eddy Simulation (LES). Measurements carried out of two simulated flame configurations showed that while the spectrum of one flame was well reproduced, the spectrum of the second flame was satisfactorily matched only in some positions. In the present work, additional calculations are made trying to explain and reduce these differences, for example, calculations using different control surfaces, using open surfaces and including a constant background flow. The results obtained are presented and discussed.
  • Applied Mechanics Reviews 01/2003; 56(2). DOI:10.1115/1.1553431 · 2.65 Impact Factor


11 Reads
Available from