Hostname: page-component-7b7567c8c-7pwfn Total loading time: 0 Render date: 2026-11-16T14:14:04.705Z Has data issue: false hasContentIssue false

Characteristics of acoustic and hydrodynamic waves in under-expanded supersonic impinging jets

Published online by Cambridge University Press:  04 November 2020

Shahram Karami*
Affiliation:
Laboratory for Turbulence Research in Aerospace and Combustion (LTRAC), Department of Mechanical and Aerospace Engineering, Monash University, Melbourne 3800, Australia
Daniel Edgington-Mitchell
Affiliation:
Laboratory for Turbulence Research in Aerospace and Combustion (LTRAC), Department of Mechanical and Aerospace Engineering, Monash University, Melbourne 3800, Australia
Vassilis Theofilis
Affiliation:
School of Engineering, University of Liverpool, Liverpool L69 7ZX, UK
Julio Soria
Affiliation:
Laboratory for Turbulence Research in Aerospace and Combustion (LTRAC), Department of Mechanical and Aerospace Engineering, Monash University, Melbourne 3800, Australia
*
†Email address for correspondence: shahram.karami@monash.edu

Abstract

In this study large-eddy simulations of under-expanded supersonic impinging jets are performed to develop a better understanding of the characteristics of the acoustic and hydrodynamic waves. Time history, dispersion relation and autocorrelation of the velocity and pressure fluctuations are used to investigate the propagation velocity, time and length scales of the dominant flow structures in the shear layer and near field. The mechanism by which the initial high-frequency instabilities change to low-frequency coherent structures within a short distance is investigated utilising Mach energy norm and linear spatial instability analysis with streamwise varying mean flow profiles. It is shown that the hydrodynamic and acoustic wavepackets have different propagation velocities and length scales while having a similar dominant frequency. It is also observed that the hydrodynamic wavepackets form approximately one jet diameter downstream of the nozzle lip. No evidence has been found to support the ‘collective interactive’ mechanism proposed by Ho & Nosseir (J. Fluid Mech., vol. 105, 1981, pp. 119–142). The ‘vortex pairing’ proposed by Winant & Browand (J. Fluid Mech., vol. 63, 1974, pp. 237–255) is observed near the nozzle; however, it has an insignificant role in the sharp reduction of the most unstable frequency of disturbances. Nonetheless, both Mach energy norm and linear spatial instability analyses show that the most unstable frequency of disturbances decreases rapidly in a very short distance from the nozzle lip in the near-nozzle region through the spatial growth of instabilities where the linear instability analysis overpredicts the frequency of the most unstable instabilities downstream of the nozzle.

Information

Type
JFM Papers
Copyright
© The Author(s), 2020. Published by Cambridge University Press
Figure 0

Figure 1.(a) Collective interaction, figure from Ho & Nosseir (1981) with permission from the authors. (b) ‘Schematic diagram of entrainment process as a function of downstream distance, or alternatively, as a function of time while riding with the mean speed’, figure and caption from Winant & Browand (1974) with permission from the authors.

Figure 1

Figure 2.Schematic of the domain and the configuration of this study.

Figure 2

Table 1.Computational grid of LES.

Figure 3

Figure 3.(a) Radial and (b) axial mesh spacing.

Figure 4

Figure 4.Instantaneous contour plots of the density gradient with a time interval of 0.26 acoustic time units (solid circles show the tracked vortical structures, dashed red curves show the tracked acoustic wave and $y$ is related to the radius by $-r < y < r$).

Figure 5

Figure 5.Instantaneous contour plots of the density gradient with a time interval of 1.3 acoustic time units (dashed curves show the acoustic waves and $y$ is related to the radius by $-r < y < r$).

Figure 6

Figure 6.(a–d) Snapshots in the ($x, r$) plane of vorticity downstream of the nozzle exit with a time interval of 0.26 acoustic time units (blue solid arrow shows a rolling/pairing process and the black dashed arrow shows the rolling/without pairing event).

Figure 7

Figure 7.Time history of (a) streamwise velocity fluctuations along the shear layer of the jet and (b) pressure fluctuations on a straight streamwise line at $r = 2.0$ for the nozzle-to-wall distance of $2d$.

Figure 8

Figure 8.Frequency-wavenumber spectra of (a) streamwise velocity fluctuations extracted at the shear layer, (b) pressure fluctuations extracted at the shear layer, and (c) pressure fluctuations extracted at $r = 2.0$ and $0.0 < x < 1.5$ of the impinging under-expanded supersonic jet with the nozzle-to-wall distance of $2d$.

Figure 9

Figure 9.Time history of (a) streamwise velocity fluctuations along the shear layer of the jet and (b) pressure fluctuations on a straight streamwise line at $r = 2.0$ for the nozzle-to-wall distance of $5d$.

Figure 10

Figure 10.Frequency-wavenumber spectra of (a) streamwise velocity fluctuations extracted at the shear layer, (b) pressure fluctuations extracted at the shear layer, and (c) pressure fluctuations extracted at $r = 2.0$ and $0.0 < x < 4.1$ of the impinging under-expanded supersonic jet with the nozzle-to-wall distance of $5d$.

Figure 11

Figure 11.Autocorrelation of the (a) streamwise, and (b) radial velocity fluctuations for the sample points at the shear layer for the nozzle-to-wall distance of $2d$.

Figure 12

Figure 12.Autocorrelation of the (a) streamwise and (b) radial velocity fluctuations for the sample points at the shear layer for the nozzle-to-wall distance of $5d$.

Figure 13

Figure 13.Autocorrelation of the radial velocity random fluctuations for sample points in the shear layer for the nozzle-to-wall distance of (a) $2d$, (b) $5d$ and (c) integral temporal scales along the shear layer of both cases of the nozzle-to-wall distance of $2d$ and $5d$.

Figure 14

Figure 14.The contour maps of non-normalised cross-correlation for the observation reference of the pressure fluctuations at $x = 1.0$ and $r = 2.0$ and (a) pressure ($C_{p'p'}$), (b) radial velocity ($C_{p'u'_{r}}$) for the nozzle-to-wall distance of $2d$.

Figure 15

Figure 15.Colour contours of non-normalised cross-correlation for the observation reference of the pressure fluctuations at $x = 2.5$ and $r =4.25$ and (a) pressure ($C_{p'p'}$), (b) radial velocity ($C_{p'u'_{r}}$) for the nozzle-to-wall distance of $5d$.

Figure 16

Figure 16.Profiles of ensemble-averaged streamwise velocity at different axial locations for the cases with a nozzle-to-wall distance of (a) $2d$ and (b) $5d$.

Figure 17

Figure 17.The spatial growth rates obtained from solution of compressible Rayleigh equation at different axial locations for the cases with a nozzle-to-wall distance of (a) $2d$ and (b) $5d$.

Figure 18

Figure 18.Strouhal number of the maximum growth rate as a function of axial location for both cases with the nozzle-to-wall distance of $2d$ and $5d$.

Figure 19

Figure 19.The energy of disturbances for the axisymmetric mode and $St$ of 1.2 (the colour map is normalised between 0 and 1 for the purpose of demonstration). The lower and upper limits of the integration in (3.2) are presented by red and black dots, respectively.

Figure 20

Figure 20.The norm of energy, $A$, of the axisymmetric mode ($m=0$) for the cases with the nozzle-to-wall distance of (a) $2d$ and (b) $5d$. (The red circles are the Strouhal number of the maximum growth rate obtained from linear spatial instability analysis.)

Figure 21

Figure 21.Profiles of autocorrelation of the radial velocity fluctuations for the sample points at the shear layer (a) $h = 2d$ (profiles are for axial locations $0.8 < x < 1.3$ with an increment of 0.1), (b) $h = 5d$ (profiles are for axial locations $1.5 < x < 2.5$ with an increment of 0.1).

Figure 22

Figure 22.The contour maps of normalised cross-correlation for the observation reference of the pressure fluctuations at $x = 1.0$ and $r = 2.0$ and (a) pressure ($\langle C_{p'p'}\rangle$), (b) radial velocity ($\langle C_{p'u'_{r}}\rangle$) for the nozzle-to-wall distance of $2d$.

Figure 23

Figure 23.Colour contours of normalised cross-correlation for the observation reference of the pressure fluctuations at $x = 2.5$ and $r =4.25$ and (a) pressure ($\langle C_{p'p'}\rangle$), (b) radial velocity ($\langle C_{p'u'_{r}}\rangle$) for the nozzle-to-wall distance of $5d$.

Figure 24

Figure 24.The normalised energy of disturbances as a function of azimuthal mode number for both cases, $h=2d$ and $5d$.