Results 1  10
of
62
Direct Computation of the noise induced by a turbulent flow through a diaphragm in a duct at low Mach number
 Computers & Fluids
"... diaphragm in a duct ..."
(Show Context)
DEVELOPMENT OF NONCENTERED WAVENUMBERBASED OPTIMIZED INTERPOLATION SCHEMES WITH AMPLIFICATION CONTROL FOR OVERLAPPING GRIDS ∗
"... Abstract. A noncentered interpolation technique has been constructed to perform simulations using overlapping grids for complex geometries. Highorder centered Lagrange polynomial interpolations and interpolations optimized in the Fourier space are first generalized to the noncentered case. These no ..."
Abstract

Cited by 3 (0 self)
 Add to MetaCart
(Show Context)
Abstract. A noncentered interpolation technique has been constructed to perform simulations using overlapping grids for complex geometries. Highorder centered Lagrange polynomial interpolations and interpolations optimized in the Fourier space are first generalized to the noncentered case. These noncentered interpolations either generate significant dispersion errors or strongly amplify highwavenumber components. Accordingly, a noncentered highorder wavenumberbased optimized interpolation method is developed with the addition of a nonlinear constraint for the control of the amplitude amplification induced by decentering. Highorder piecewise polynomial regressions of the obtained interpolation coefficients are performed. The time stability of the method is investigated in the 1D case when the interpolation method is used in conjunction with explicit highorder differencing, filtering schemes, as well as a 6step Runge–Kutta time integration algorithm. A criterion is formulated to predict its stability as a function of the filtering strength and the Courant–Friedrichs– Lewy constant. Finally, 1D convection simulations are presented to illustrate the stability and the accuracy of the developed noncentered interpolations. Key words. space noncentered interpolations, decentering, high order, overlapping grids, Fourier
Optimized explicit RungeKutta schemes for the spectral difference method applied to wave propagation problems
 SIAM J. Sci. Comput
, 2013
"... Explicit Runge–Kutta schemes with large stable step sizes are developed for integration of high order spectral difference spatial discretizations on quadrilateral grids. The new schemes permit an effective time step that is substantially larger than the maximum admissible time step of standard expli ..."
Abstract

Cited by 3 (3 self)
 Add to MetaCart
(Show Context)
Explicit Runge–Kutta schemes with large stable step sizes are developed for integration of high order spectral difference spatial discretizations on quadrilateral grids. The new schemes permit an effective time step that is substantially larger than the maximum admissible time step of standard explicit Runge–Kutta schemes available in literature. Furthermore, they have a small principal error norm and admit a lowstorage implementation. The advantages of the new schemes are demonstrated through application to the Euler equations and the linearized Euler equations. 1
Influence of the nozzleexit boundarylayer thickness on the flow and acoustic fields of initially laminar jets
 15th AIAA/CEAS Aeroacoustics Conference, AIAA Paper
"... Round jets originating from a pipe nozzle are computed by LargeEddy Simulations (LES) to investigate the effects of the nozzleexit conditions on the flow and sound fields of initially laminar jets. The jets are at Mach number 0.9 and Reynolds number 105, and exhibit exit boundary layers characteri ..."
Abstract

Cited by 2 (1 self)
 Add to MetaCart
Round jets originating from a pipe nozzle are computed by LargeEddy Simulations (LES) to investigate the effects of the nozzleexit conditions on the flow and sound fields of initially laminar jets. The jets are at Mach number 0.9 and Reynolds number 105, and exhibit exit boundary layers characterized by Blasius velocity profiles, maximum rootmeansquare axial velocity fluctuations between 0.2 % and 1.9 % of the jet velocity, and momentum thicknesses varying from 0.003 to 0.023 times the jet radius. The farfield noise is determined from the LES data on a cylindrical surface by solving the acoustic equations. Jets with thinner boundary layer develop earlier but at a slower rate, yielding longer potential cores and lower centerline turbulent intensities comparing well with measurements at high Reynolds numbers. In all jets the shearlayer transition is dominated by vortex rollingups and pairings, which generate strong components in the noise spectra. Just adding random disturbances of low magnitude in the nozzle however leads to weaker rollingup and pairing processes, thus significantly reducing their contributions to the sound field. This high sensitivity to the initial conditions is in good agreement with experimental observations. I.
Modelling of sound generation by turbulent reacting flows
, 2010
"... The two components of combustion noise, namely the direct noise induced by heat release fluctuations in the flow and the indirect noise caused by acceleration of entropy perturbations through mean flow gradients resulting in acoustic radiation, are examined by considering different types of inhomoge ..."
Abstract

Cited by 2 (0 self)
 Add to MetaCart
(Show Context)
The two components of combustion noise, namely the direct noise induced by heat release fluctuations in the flow and the indirect noise caused by acceleration of entropy perturbations through mean flow gradients resulting in acoustic radiation, are examined by considering different types of inhomogeneous wave equations deduced from the balance equations of fluid dynamics. This is accomplished by adapting acoustic analogies initially derived by Lighthill (1952), Phillips (1960) and Lilley (1972) for nonreactive flows to the reacting case. This reformulation is based on a standard set of assumptions borrowed from combustion. It is shown that a consistent formulation of a wave operator and its associated source terms is required.
Analysis of stretched grids as buffer zones in aeroacoustic simulations
 Proc. 15th AIAA/CEAS Aeroacoustics Conference, 43 of Conference Proceeding Series, AIAA
, 2009
"... A zone of increasingly stretched grid is a robust and easytouse way to avoid unwanted reflections at artificial boundaries in aeroacoustic simulations. There are two main damping mechanisms, dissipation and underresolution a traveling wave, respectively. In this paper the behavior of acoustic w ..."
Abstract

Cited by 2 (0 self)
 Add to MetaCart
(Show Context)
A zone of increasingly stretched grid is a robust and easytouse way to avoid unwanted reflections at artificial boundaries in aeroacoustic simulations. There are two main damping mechanisms, dissipation and underresolution a traveling wave, respectively. In this paper the behavior of acoustic waves and wave packages, respectively, are studied via analysis of solutions to a semidiscrete linear boundary value problem. The solutions of the linear problem is compared to solutions of the full nonlinear Euler equations. I.
Flow and acoustic fields of Reynolds number 105, subsonic jets with tripped exit boundary layers, AIAA Paper
, 2010
"... Compressible LargeEddy Simulations of five isothermal round jets at a Mach number of 0.9 and a diameterbased Reynolds number of 105 originating from a pipe are reported, to examine the feasibility of computing the flow and acoustic fields of initially nominally turbulent jets. In the pipe the boun ..."
Abstract

Cited by 2 (2 self)
 Add to MetaCart
Compressible LargeEddy Simulations of five isothermal round jets at a Mach number of 0.9 and a diameterbased Reynolds number of 105 originating from a pipe are reported, to examine the feasibility of computing the flow and acoustic fields of initially nominally turbulent jets. In the pipe the boundary layers are tripped in order to obtain at the exit section laminar mean velocity profiles of momentum thickness δθ = 0.018 times the jet radius, yielding a Reynolds number Reθ = 900, as well as peak turbulent intensities around 9 % of the jet velocity. Two methods of boundarylayer tripping, and four grids containing from 50 to 252 million points, are considered. The results are found to vary negligibly with the tripping procedure, but appreciably with the grid resolution. A high azimuthal resolution appears in particular necessary to simulate the initial shearlayer development properly. For the present tripped jet, nθ = 256 points in the azimuth are for instance insufficient, which has led us to specify nθ = 1024 points to reach a high level of confidence in the numerical solutions. A fine discretization of the mixing layers also seems to be required from the nozzle exit to the end of the potential core to get jet features similar to those measured at high Reynolds numbers. I.
Aerodynamic excitation of the diametral modes of an internal axisymmetric cavity
, 2008
"... TITLE: Aerodynamic excitation of the diametral modes of an internal axisymmetric cavity ..."
Abstract

Cited by 2 (0 self)
 Add to MetaCart
TITLE: Aerodynamic excitation of the diametral modes of an internal axisymmetric cavity
Noise sources from a cylindrical cavity
"... The expansion of air tra±c operations is nowadays limited by environmental constraints on noise. Advances in jet noise reduction have increased the importance of landing noise from the airframe as a signi¯cant contribution to the e®ective perceived noise level around airports. The most acoustically ..."
Abstract

Cited by 1 (1 self)
 Add to MetaCart
(Show Context)
The expansion of air tra±c operations is nowadays limited by environmental constraints on noise. Advances in jet noise reduction have increased the importance of landing noise from the airframe as a signi¯cant contribution to the e®ective perceived noise level around airports. The most acoustically active airframe components in a civil aircraft are the high lift systems and the landing gear. Nonetheless, other components, such as fuel vents or ailerons, also contribute to the overall noise emissions. This study considers a cylindrical cavity as a low ¯delity fuel vent model. Di®erent diameter to depth ratios and in°ow velocities are simulated by means of an inhouse compressible Euler solver. The predictions for a diameter to depth ratio of 0.714 show an unsteady asymmetric vortex structure at the in°ow Mach numbers of 0.235 and 0.3. The alternate impingement of this vortex on the right and on the left of the cavity trailing edge produces pressure waves and the °ow instability is selfsustained. The simulations of a cavity with a length to depth ratio of 2.5 at the same Mach numbers show a similar selfsustained instability where the °ow recirculation is symmetric about the cavity midplane. To identify and localize the most acoustically active regions in the inviscid °ow model, the double divergence of the Lighthill stress tensor was computed from the aerodynamic predictions. This work sets the basis to perform a Ffowcs Williams and Hawkings acoustic analogy to predict the fuel vent contribution to landing noise. Nomenclature c Speed of sound ± (f) Dirac delta function e Internal energy F Inviscid °uxes, F = [u; u u+ p
A 2D Discontinuous Galerkin Method for Aeroacoustics with Curved Boundary Treatment
"... A Discontinuous Galerkin Method is applied to unstructured grids to simulate aeroacoustic propagation, modelled by the Linearized Euler Equations. The quadraturefree form of the Discontinuous Galerkin Method is used on triangular elements with straight edges. Computations of a freely propagating pr ..."
Abstract

Cited by 1 (1 self)
 Add to MetaCart
(Show Context)
A Discontinuous Galerkin Method is applied to unstructured grids to simulate aeroacoustic propagation, modelled by the Linearized Euler Equations. The quadraturefree form of the Discontinuous Galerkin Method is used on triangular elements with straight edges. Computations of a freely propagating pressure pulse demonstrate that the implementation achieves the theoretical order of accuracy. With increasing order, the method becomes more efficient. However, simulation of acoustic scattering by a cylinder shows that the linear treatment of the geometry can limit the accuracy at high order. Two higherorder treatments of curved boundaries are presented to overcome this restriction. It is found that only the use of curved elements, on which the quadraturefree technique cannot be applied, provides satisfying accuracy; further implications on stability and computational cost are discussed. 1