Effects of moderate Reynolds numbers on subsonic round jets with highly disturbed nozzle-exit boundary layers

Abstract

Particle transport in a turbulent boundary layer: Non-local closures for particle dispersion tensors accounting for particle-wall interactions Phys. Fluids 24, 103304 (2012) Convection and reaction in a diffusive boundary layer in a porous medium: Nonlinear dynamics Chaos 22, 037113 (2012) Symmetry analysis and self-similar forms of fluid flow and heat-mass transfer in turbulent boundary layer flow of a nanofluid Phys. Fluids 24, 092003 (2012) Detuned resonances of Tollmien-Schlichting waves in an airfoil boundary layer: Experiment, theory, and direct numerical simulation Phys. Fluids 24, 094103 (2012) Additional information on Phys. FluidsThe effects of moderate Reynolds numbers on the flow and acoustic fields of initially highly disturbed isothermal round jets at Mach number M = 0.9 and diameter-based Reynolds numbers Re D between 2.5 × 10 4 and 2 × 10 5 are investigated using largeeddy simulation under carefully controlled conditions. To the best of our knowledge, this is the first comprehensive study of its kind. The jets originate at z = 0 from a pipe nozzle of radius r 0 , in which a tripping procedure is applied to the boundary layers. At the nozzle exit, laminar-like mean velocity profiles of thickness δ 0.15r 0 and momentum thickness δ θ 0.018r 0 , yielding Reynolds numbers Re θ varying from 256 to 1856 depending on Re D , and peak turbulence intensities around 9% of the jet velocity, are thus obtained. As the Reynolds number increases, the mixing layers develop more slowly, with smaller integral length scales and lower levels of velocity fluctuations. The axial profiles of turbulence intensities become smoother, showing a clear overshoot around z = 2r 0 at Re D = 2.5 × 10 4 , but a monotonical growth at Re D = 2 × 10 5 . Velocity spectra downstream of the nozzle exit also broaden with Re D , as expected. Large-scale components usually observed in turbulent boundary layers and shear layers, characterized by Strouhal numbers St θ 0.013 around z = r 0 and by azimuthal spacings λ θ δ, remain dominant, although the contribution of fine-scale structures with λ θ ≤ δ/2 strengthens. Moreover, with rising Re D , the jet potential core lengthens slightly, but the flow properties do not change significantly farther downstream. Finally, lower sound pressure levels are generated, with a decrease of about 2 dB over the range of Re D considered. C 2012 American Institute of Physics.

Influence of nozzle-exit boundary-layer profile on high-subsonic jets

“…For z ≤ 25r 0 and r ≤ 6.5r 0 , it is identical to that used in recent simulations [27][28][29][30][31] for subsonic jets with laminar boundary-layer profiles. There are 169 points along the pipe nozzle between z = −2r 0 and z = 0, 77 points within the jet radius, and 41 points between r = r0 − δ BL = 0.75r 0 and r = r 0 .…”

“…The physics of the larger turbulent structures is therefore unlikely to be governed by numerical or subgrid-modeling dissipation. This allows the effective flow Reynolds number not to be artificially decreased, and viscosity effects to be captured, as was the case in Bogey et al 30 for jets at Re D between 2.5 × 10 4 and 2 × 10 5 . These remarks certainly equally hold true for the present LES dealing with jets at Re D = 5 × 10 4 .…”

“…In most cases, they are identical to those used in recent jet simulations, which have been thoroughly described in previous references. [27][28][29][30][31] A great amount of information about the boundary-layer tripping procedure, the discretization quality and the LES reliability is moreover available. 27 …”

Wall pressure fluctuations in rectangular partial enclosures

“…These Re numbers ensure low-frequency acoustic wave emissions within the region 1.2 and 2.5 jet diameters [17,18]. Such perturbations excite the RPE first modes, ensuring the achievement of the self-excited oscillations.…”

Petascale large eddy simulation of jet engine noise based on the truncated SPIKE algorithm

“…References [39][40][41] use a multiblock solver with overset subgrids to perform high-fidelity simulations of subsonic jets with nozzles (with and without chevrons), on grids with up to 500 million points. References [3,4] use high-order methods on meshes with 252 million points to study the effect of important parameters on the noise in subsonic conditions. There are comparatively fewer high-fidelity supersonic jet simulations.…”