Development of compressible large-eddy simulations combining high-order schemes and wall modeling

Bras, Sophie Le; Deniau, Hugues; Bogey, Christophe; Daviller, Guillaume

doi:10.2514/6.2015-3135

Cited by 8 publications

(10 citation statements)

References 35 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…LESs of jets at Re D 50;000 with thick transitional and turbulent boundary-layer profiles have been carried out in Bogey and Marsden [34]. Finally, Brès et al [35] and Le Bras et al [36] very recently simulated initially turbulent jets at Re D > 500;000 using wall modeling inside the nozzle.…”

Section: Introductionmentioning

confidence: 99%

Simulations of Initially Highly Disturbed Jets with Experiment-Like Exit Boundary Layers

Bogey

Marsden

2016

AIAA Journal

Self Cite

View full text Add to dashboard Cite

Two isothermal round jets at a Mach number of 0.9 and a diameter-based Reynolds number of 2 × 10 5 have been computed by compressible large-eddy simulation using high-order finite differences on a grid of 3.1 billion points. At the exit of a straight pipe nozzle in which a trip forcing is applied, the jet flow velocity parameters, including the momentum thickness and the shape factor of the boundary layer, the momentum-thickness-based Reynolds number, and the peak turbulence intensity, roughly match those found in experiments using two nozzles referred to as the ASME and the conical nozzles. The boundary layer is in a highly disturbed laminar state in the first case and in a turbulent state in the second. The exit flow conditions, the shear-layer and jet flowfields, and the far-field noise provided by the large-eddy simulation are described. The jet with the ASME-like initial conditions develops a little more rapidly, with slightly higher turbulence levels than the other. Overall, however, the results obtained for the two jets are very similar, and they are in good agreement with measurements available for Mach 0.9 jets. In particular, this similarity holds for the far-field spectra. Because the ASME nozzle has been reported to yield higher noise levels than the conical nozzle, this suggests that the nozzle-exit conditions in the large-eddy simulation do not adequately reflect those in the experiments and/or that the link between the noise differences and the jet initial conditions using the two nozzles is not as simple as was first thought, and that other parameters, associated for instance with the nozzle geometry such as the presence of pressure gradients, may also play an important role.

show abstract

Section: Introductionmentioning

confidence: 99%

Simulations of Initially Highly Disturbed Jets with Experiment-Like Exit Boundary Layers

Bogey

Marsden

2016

AIAA Journal

Self Cite

View full text Add to dashboard Cite

show abstract

“…In relation to this work, the modeling of the nozzle should be further improved, going deeper inside the nozzle in order to obtain a better turbulent boundary layer at its exit. In order to accomplish this goal, new wall-turbulence models adapted to high-order compact schemes 47 should be applied in order to avoid a costly wall-resolved mesh but yet, achieving a good representation of the boundary layer. Moreover, a less dissipative filter of order eight should be used in the future to better capture the pressure perturbations.…”

Section: Discussionmentioning

confidence: 99%

Large Eddy Simulation of Shock-Cell Noise From a Dual Stream Jet

Arroyo

Puigt

Airiau

et al. 2016

22nd AIAA/CEAS Aeroacoustics Conference

View full text Add to dashboard Cite

This study presents the shock-cell noise results obtained with a large eddy aeroacoustic simulation from a dual stream jet. The primary stream is cold and subsonic with an exit Mach number of Mp = 0.89 and a Reynolds number of Rep = 0.57 × 10 6. The secondary stream is cold supersonic and under-expanded with a perfectly expanded Mach number of Ms = 1.20 and Res = 1.66 × 10 6. The computations are performed with the structured multiblock solver elsA. The aerodynamics and aeroacoustics of the flow are studied and analyzed in detail showing good agreement with experimental fits for the lengthscale and shear layer development. An acoustic-hydrodynamic filtering is used in order to compute the characteristic wavelengths of each component and estimate the shock-cell noise frequency. Moreover, the broadband shock-cell noise is captured in the near and far fields.

show abstract

“…Since the first developments in the field of computational aeroacoustics in the early nineties, 1 considerable progress has been made in the simulation of the flow and acoustic fields of high-speed turbulent jets, [2][3][4][5] which should help us to better describe the underlying noise generation mechanisms. For subsonic jets, in particular, a number of research teams [6][7][8][9][10] have been able to obtain far-field pressure spectra in good agreement with experimental measurements using different numerical methodologies. Nevertheless, for these flows, unlike other flows such as turbulent boundary layers, there is still no clear rule concerning the resolution required to obtain trustworthy solutions.…”

Section: Introductionmentioning

confidence: 87%

A study of the grid dependence of the flow field and noise of subsonic jets

Bogey

Marsden

2016

54th AIAA Aerospace Sciences Meeting

Self Cite

View full text Add to dashboard Cite

Three isothermal round jets at a Mach number of 0.9 and a diameter-based Reynolds number of 10 5 are computed by large-eddy simulation using three grids with increasing resolution in order to investigate the grid dependence of the jet flow field and noise. The jets correspond to two initially fully laminar jets and one initially strongly disturbed jet considered in previous numerical studies. At the exit of a pipe nozzle of radius r0, at z = 0, they exhibit laminar boundary-layer mean-velocity profiles of thickness 0.2r0, 0.025r0 and 0.15r0, and peak turbulence intensities close to 0.2%, 0.3% and 9%, respectively. The grids contain up to one billion points, and, compared to the grids used in previous studies, they are much finer in the axial direction for z ≥ 5r0, and in the radial direction in the outer region of the jet mixing layers. The main jet flow field and noise characteristics given by the simulations, including the mixing-layer thickness, the centerline mean velocity, the turbulence intensities on the nozzle lip line and the jet axis, velocity spectra in the jets, and near-field and far-field pressure spectra, are presented. For the initially laminar jet with thin boundary layers and the initially disturbed jet, significant differences are found with respect to the results from previous studies. The jet development is more rapid, the turbulence intensities just upstream and downstream of the end of the potential core are higher due to the presence of stronger large-scale structures, and more low-frequency noise is generated. For the three jets, however, the results obtained using the present grids are very similar.

show abstract

Development of compressible large-eddy simulations combining high-order schemes and wall modeling

Cited by 8 publications

References 35 publications

Simulations of Initially Highly Disturbed Jets with Experiment-Like Exit Boundary Layers

Simulations of Initially Highly Disturbed Jets with Experiment-Like Exit Boundary Layers

Large Eddy Simulation of Shock-Cell Noise From a Dual Stream Jet

A study of the grid dependence of the flow field and noise of subsonic jets

Contact Info

Product

Resources

About