![Page 1: [American Institute of Aeronautics and Astronautics 17th AIAA/CEAS Aeroacoustics Conference (32nd AIAA Aeroacoustics Conference) - Portland, Oregon ()] 17th AIAA/CEAS Aeroacoustics](https://reader035.vdocuments.pub/reader035/viewer/2022080405/575093351a28abbf6bae1b00/html5/thumbnails/1.jpg)
American Institute of Aeronautics and Astronautics
1
Large-Eddy Simulation of High-Subsonic Jet flow
with Microjet Injection
Shunji ENOMOTO1 and Kazuomi YAMAMOTO
2
Japan Aerospace Exploration Agency, Chofu, Tokyo, JAPAN
Kenshi YAMASHITA3
Advanced Science & Intelligence Research Institute Corp., Chiyoda-ku, Tokyo, JAPAN
and
Nozomi TANAKA, Yoshinori OBA and Tsutomu OISHI4
IHI Corporation, Nishitama-gun, Tokyo, JAPAN
Large-Eddy Simulation of high subsonic jet with microjet injection was performed
using UPACS-LES code which is developed in JAXA. Large scale (476M grid point)
simulation was executed on JAXA Supercomputer System using 980 processors. The
result shows good agreement with the experimental data in terms of velocity fluctuation
and far-field noise level. Far-field noise prediction using FW-H method from the LES
results show that the LES successfully predict noise reduction by microjets for lower
frequency component emitted in 30 deg. observation angle, while it still has difficulty in
predicting reduction of higher frequency noise emitted in 90 deg. observation angle.
Nomenclature
a = speed of sound in the ambient
D = diameter of main nozzle
Ma = acoustic Mach number = Uj / a
r = radial coordinate
R = distance from the center of nozzle exit to observation point of far-field noise
St = Strouhal number = (frequency D) / Uj
U = axial velocity (time averaged)
u’ = axial turbulent velocity
Uj = main jet velocity
X = axial coordinate
Y or Z = transverse coordinate
θ = observation angle of far-field noise from the downstream jet axis
I. Introduction
One of the main noise sources of airplane is jet noise. Jet noise reduction method, such as microjet injection,
is widely investigated. Experimental investigation1 of microjet effects on subsonic M=0.9 unheated jet showed
that microjet injection reduced turbulent intensities up to 20% and far field noise reduction was about 0.5 – 2 dB.
Alkislar examined the effect of streamwise vortex generated by microjets2, and he measured the flow field of the
Chevron-Microjet combination concept in detail3. Gutmark et al.4 applied microjets to double stream
configuration. Zaman5 showed that the smaller diameter microjet ports with higher driving pressure can produce
better noise reduction. Castelain et al.6 made detail measurements of flow field and far field noise on microjets,
and studied the noise reduction as a function of the mass flux of microjet injection.
Numerical simulation of jet flow and jet noise is expected to be utilized as a design tool for those jet nozzles.
Several studies of jet noise simulation have been carried out using large-eddy simulation (LES), and some of
1 Associate Senior Researcher, Aerospace Research and Development Directorate, 7-44-1 Jindaiji-Higashi,
Chofu, Tokyo 182-8522, JAPAN, AIAA member 2 Senior Researcher, AIAA senior member
3 Engineer, Science Engineering Division, 1-18-14 Uchi-Kanda, Chiyoda-ku, Tokyo 101-0047, JAPAN
4 Aero-Engines & Space Operations, 229 Tonogaya, Nishitama-gun, Tokyo 190-1297, JAPAN
17th AIAA/CEAS Aeroacoustics Conference (32nd AIAA Aeroacoustics Conference)05 - 08 June 2011, Portland, Oregon
AIAA 2011-2883
Copyright © 2011 by the author(s). Published by the American Institute of Aeronautics and Astronautics, Inc., with permission.
![Page 2: [American Institute of Aeronautics and Astronautics 17th AIAA/CEAS Aeroacoustics Conference (32nd AIAA Aeroacoustics Conference) - Portland, Oregon ()] 17th AIAA/CEAS Aeroacoustics](https://reader035.vdocuments.pub/reader035/viewer/2022080405/575093351a28abbf6bae1b00/html5/thumbnails/2.jpg)
American Institute of Aeronautics and Astronautics
2
them showed good accuracy. Bodony and Lele7 performed LES of subsonic and supersonic circular jet. Mach
numbers were from 0.5 to 1.5, Reynolds numbers were 105-10
6. Sixth order compact scheme were used for
radial and axial direction, and spectral method for circumferential direction. They used a dynamic Smagorinsky
model for the calculation. Bogey and Bailly8 used high-cut filter instead of SGS model for their LES of subsonic
jet Ma = 0.9. They investigated the influence of inflow condition and inlet disturbance. Uzun and Lyrintzis9
performed LES of subsonic jet (Ma = 0.9, ReD = 105) using dynamic Smagorinsky model and 6
th order compact
scheme. Shur et al.10
uses 5th order upwind scheme for LES simulation without subgrid scale model. PC clusters
were used for their calculation and LES of practical nozzles, such as chevron nozzles and slanted nozzles, were
carried out.
Numerical simulations are also applied to microjet flow. Huet et al.11
numerically investigated the efficiency
of microjets. Their LES showed that the microjets led to a reduction of the turbulence in the shear layer and a
decrease of the far field noise. LES of Shur et al.12
successfully predicted the difference between baseline jet
noise and noise reduction by microjets, and showed that the microjets concept might not effective in flight
conditions.
UPACS is a CFD code developed in JAXA13
. It is a multi-block structured grid solver for general curvilinear
grids. Third order MUSCL and Roe method are used for spatial discretization. UPACS-LES is a modified
version of UPACS. Sixth-order compact scheme are used for spatial discretization. Imamura et al.14,15
uses this
code for vortical flows around high lift device, such as a flap and a slat.
We have been trying to simulate jet flow with microjet injection using UPACS-LES. Previous computation16
was not fully satisfactory to simulate details of flow phenomena, and it's reason seemed to be insufficient grid
resolution. In the present study, we had an opportunity to use large number of CPUs, so we used denser grid to
simulate jet flow.
II. Numerical Method
UPCAS-LES is an unsteady three dimensional filtered compressible Navier-Stokes solver based on finite
volume method using multi-block structured grids. The convection term is discretized by high-order compact
scheme for finite volume method, and a low storage four-stage Runge-Kutta method is implemented for the time
integration. Smagorinsky model is used for the subgrid scale stress model of LES.
Symbols in this section are not included in nomenclature.
Governing equations used in UPACS-LES are spatial filtered Navier-Stokes equations17
,
(( ) )
( )
where the overbar ( ) denotes a filter operation, tilde denotes Favre filter operation ( ⁄ ). Here, f is an
arbitrary variable.
(
)
(
)
For the subgrid scale stress model of LES, Smagorinsky model are used.
( )
![Page 3: [American Institute of Aeronautics and Astronautics 17th AIAA/CEAS Aeroacoustics Conference (32nd AIAA Aeroacoustics Conference) - Portland, Oregon ()] 17th AIAA/CEAS Aeroacoustics](https://reader035.vdocuments.pub/reader035/viewer/2022080405/575093351a28abbf6bae1b00/html5/thumbnails/3.jpg)
American Institute of Aeronautics and Astronautics
3
| | | |
The convection term was discretized by Kobayashi’s compact scheme for finite volume method18
. In the
compact scheme for finite volume method, cell-face quantities are reconstructed from cell-averaged quantities,
and the fluxes on the cell-face are calculated from the reconstructed quantities:
where on the left hand side denotes interpolated value at cell faces, and on the right hand side denotes
cell-averaged value. Sixth order is achieved when the coefficients are:
Since this equation uses interpolated values at the adjacent cell faces implicitly, a tri-diagonal matrix has to
be solved. The scheme maintains 6th order only when the computational grid is equally spacing orthogonal grid.
On the general curvilinear grid, it maintains 3rd order.
At the interface between each block, 4th order explicit scheme is used in order to avoid accessing the
variables of adjacent blocks:
To prevent numerical odd-even oscillation, a compact filter, whose order of accuracy is up to 14, was used:
∑
on the right hand side denotes original value, and on the left hand side denotes filtered value. The
coefficients α and aj are shown in Reference [19]. The essentials of the filter are the same as the method of
Gaitonde et al.20
, but it does not use one-sided filter near the boundaries, instead, it use lower order central filter
near the boundaries in order to avoid phase error.
III. Numerical Simulation
As the verification data for our computational method for jet flow with microjet injection, we referred to the
experiment of Castelain et al. 6,21,22
. In their experiment, the diameter of the main jet nozzle was D = 50mm, the
air was heated to maintain the temperature of the expanded jet close to the ambient temperature, Mach number
based on the jet velocity Uj and the ambient speed of sound a is Ma= Uj / a = 0.9, and the Reynolds number ReD
= 106. Eighteen microjet nozzles were placed around the main jet nozzle. The diameter of microjet nozzle was
1mm each, and injection angle was 45 degree. The mass flow rates of each microjet to main jet are shown in
Table 1. There are seven test points in the experiment of Castelain et al21
and we selected two test points, rm(3)
and rm(7)
from them, because noise reduction levels were relatively large comparing to the other test points. The
baseline jet (without microjet) is also computed and referred as rm(0)
.
Table 1 Microjet condition
test point rm(0)
rm(3)
rm(7)
microjet mass flow ratio 0 3.36 8.86 (×10-4
)
Figure 1 shows computational grids. Figure 1a is a side view of the grid. The physical part of computational
region extends −0.1 ≤ X/D ≤ 20. The sponge zone, where grid spacing is extended gradually, is extended outside.
Freund method23
is applied in the sponge zone to attenuate outgoing waves. Table 2 shows the specification of
the computational grids. “fine” is the present case. “coarse” is our previous computation16
which is shown here
as a reference. Nozzle exit region is shown in Fig. 1b. The main nozzle is straight pipe and it does not represent
the nozzle profile of the experiment. The main nozzle exit position is X/D = 0, and it extends upstream to X/D =
−1. Grid is almost equally spaced for all direction near nozzle exit region. Detail of microjet injection region is
shown in Fig. 1c, where grid lines are tilted 45°. Microjet flow is specified as an inflow boundary condition, and
![Page 4: [American Institute of Aeronautics and Astronautics 17th AIAA/CEAS Aeroacoustics Conference (32nd AIAA Aeroacoustics Conference) - Portland, Oregon ()] 17th AIAA/CEAS Aeroacoustics](https://reader035.vdocuments.pub/reader035/viewer/2022080405/575093351a28abbf6bae1b00/html5/thumbnails/4.jpg)
American Institute of Aeronautics and Astronautics
4
is represented by 4x4cells, which are shown as red line in Fig. 1c. Figure 1d is bird's-eye view of the grid near
main nozzle exit.
Table 2 Specification of computational grids
Case Grid points Grid spacing Grid spacing for
microjet
fine (present) 476M 0.005D 4x4
coarse (reference) 28M 0.013D 2x2
Inflow velocity profile is constant velocity Uj. For laboratory jets, inflow boundary layer thickness is very
thin. Our grid is not clustered to nozzle wall and is spaced about 0.005D, so physical boundary layer thickness is
less than one grid spacing. Hence, we did not specify any boundary layer profile at inflow boundary of jet
nozzle.
In LES of jet flow, it is necessary to impose inflow disturbance for laminar-turbulent transition in very early
stage of the shear layer. Present simulation used an inflow forcing based on Bogey’s method24
. Inflow forcing
was added near the nozzle wall (X/D = −0.2 and r/D=0.48) in order to simulate the turbulence in the boundary
layer of nozzle inside wall. The amount of disturbance should be greater enough to make the shear layer
turbulent, and should be as small as possible so that noise emitted from the disturbance does not affect far-field
SPL estimation.
For each simulation, we ran 200,000 time steps for initial transients, which is 200 non-dimensional time
based on speed of sound in the ambient a and diameter of the main nozzle D. Then we collected the flow
statistics over 150,000 time steps (150 non-dimentional time).
The computation was executed on JAXA Supercomputer System (JSS). The system has total 3008 nodes,
and 4 cores are in one node. MPI and openMP are used for the parallel computation. Three cases are computed
(rm(0)
, rm(3)
, and rm(7)
), and 980 nodes (3920 cores) × 12 days were used for each case.
IV. Results and discussion
A. Baseline jet without microjet injection
First, we compared the baseline jet result (rm(0)
) with existing data in order to verify the accuracy of the code.
In Fig. 2, axial mean and turbulent velocities on the centerline and on the lipline of the LES are compared
with experimental data of Bridges & Wernet25
and Lau et al26
. The data of Bridges do not include isothermal
condition. Hence, cold jet of Ma = 0.9 (set point 7) is used instead. The data of Lau are isothermal jet of Ma = 0.9.
LES result is indicated as “fine” and “coarse”, “fine” is present result and “coarse” is our previous16
result with
coarser grid. Figure 2a shows centerline axial mean velocity. “coarse” exhibit an earlier potential core collapse
than the experiment. On the other hand, the potential core of “fine” is longer than the experiment. The mean
velocity decay rate of “fine” is close to Bridges data. Lipline axial mean velocity is shown in Fig .2b. “fine”
exhibits good agreement with the experiment. “coarse” shows lower lipline velocity. In “coarse” simulation,
thick boundary layer profile was set at the nozzle inflow boundary, so lipline does not correspond to the center
of the shear layer. Figure 2c is centerline axial turbulent velocity. “coarse” exhibits low turbulence level. “fine”
has very similar profile to Bridges data, but it shifts downstream about X/D=1 compared to the data. Bridges
data exhibits u’/Uj=0.015 at the beginning of jet, but LES result exhibits lower value of 0.05. Inflow disturbance
of LESs were added near the nozzle wall, but no disturbance were added to whole jet. To simulate centerline
turbulence correctly, adding small inflow disturbance to whole jet might be needed. Figure 2d shows lipline
axial turbulent velocity. “fine” exhibits a little higher level than the Bridges data. Figure 2e shows centerline
transverse turbulent velocity. “fine” exhibits similar peak value to Lau’s data except for the location of the peaks.
Figure 3 is axial mean and turbulent velocity profiles at X/D = 4, 8, 12, and 16. Axial mean velocities are
shown on the left column, Fig. 3a, 3c, 3e, and 3g. “fine” exhibits good agreement with Bridges, but it’s shear
layer expands slower than the experiment. The profiles of “coarse” are wider and the peak value is lower than
the experiment. For axial turbulent velocity profiles, shown on the right column of Fig. 3b, 3d, 3f, and 3h, “fine”
shows very good agreement with Bridges.
The above results shows that the LES result of “fine” grid exhibits good agreement with the experiment,
especially with Bridges25
, and “coarse” does not. So further discussion in this paper will be made with “fine”
data.
Next, far-field sound pressure level (SPL) is compared with an experiment. The porous Ffowcs Williams-
Hawkings (FW-H) method28,29
is used to study the far-field noise of our LES result. The FW-H integral surface
does not include downstream jet exit region. Far-field noise level of LES are computed at R/D = 40. Figure 4
shows sound pressure levels from LES and the experiment of Tanna27
. The experimental data have been scaled
to the same observation distance of R/D = 40. Overall SPL of the LES exhibits 2dB lower than the experiment
(see Fig. 4a). Figure 4b and 4c show the 1/3 octave band sound pressure levels at observation angles θ =30° and
![Page 5: [American Institute of Aeronautics and Astronautics 17th AIAA/CEAS Aeroacoustics Conference (32nd AIAA Aeroacoustics Conference) - Portland, Oregon ()] 17th AIAA/CEAS Aeroacoustics](https://reader035.vdocuments.pub/reader035/viewer/2022080405/575093351a28abbf6bae1b00/html5/thumbnails/5.jpg)
American Institute of Aeronautics and Astronautics
5
90°. At θ = 30°, the profile of LES agrees well with the experiment, but peak level at St = 0.2 is 3dB lower than
the experiment. At θ = 90°, the profile of LES is about 5dB lower than the experiment, and peak frequency is
higher. These results are not very accurate as Shur et al.12
, especially for θ = 90°, but the differences between
LES and the experiment are not very large. Hence, the simulation code seems to be eligible for an evaluation
tool for microjet effects.
For both angles, higher frequency of St > 4 show lower SPL, that means the cut-off frequency of the
prediction is St = 4.
B. Flow field with microjets
From Fig. 5 to Fig. 7 compare the computational results with microjet injection with the experimental data
by Castelain22
. Figure 5 shows u-velocity fluctuation at X/D=1 cross section. Left column, Fig. 5a, 5c and 5e are
LES results, and right column, Figs.5b, 5d, and 5f, are PIV data of Castelain22
. These figures for the experiment
are the same data shown in Castelain22
. The color map is adjusted to be the same between the experiment and
the LES. We can see deformation of shear layer and increase of fluctuation caused by microjet injection from
these figures. The shear layer of rm(0)
is thicker than the experiment. But shear layer thickness and the
deformation of the shear layer of rm(3)
and rm(7)
are closer to the experiment. The peak locations of the turbulence
are between microjet injection points in the experiment, and the fact is the same in the LES results (Fig. 5c and
5e).
Figure 6 shows axial velocity fluctuation at X/D=3. In this section, the deformation caused by microjet
injection is not clearly seen. But it is apparent that microjet decreases turbulence level in the experiment, and the
LES results also shows the same trend. LES shows lower turbulence level for rm(0)
(Fig. 6a) than the experiment
in this section, and microjet have a little effect to decrease turbulence (Fig. 6c and 6e). The shear layer
thicknesses for all cases have good agreement with the experiment.
Figure 7 shows the Reynolds stress like component at X/D=1 defined in Ref.[22]. It represents turbulent
mixing in the jet shear-layer in radial direction. Again, shear layer width of the LES results is wider than that of
the experiment, but the shape of the shear layer and the level of Reynolds stress agree with the experiment.
Figure 8 is the Reynolds stress at X/D=3. Overall value of the LES is less than the experiment, but there can be
seen that microjets make the Reynolds stress lower in LES as well as the experiment. From these comparisons,
it is shown that the LES successfully captures the flow physics of shear layer deformation due to the microjet
injections in detail, which enhances the mixing in the jet shear-layer then reduces turbulence level in
downstream region.
Figure 9 shows turbulent kinetic energy along X-Y plane which is not obtained in the experiment. Despite
the difference in Fig. 5 to Fig. 7, the LES result for rm(3)
does not show much difference with this global view. In
contrast, TKE for rm(7)
clearly shows the effect of microjet injection; TKE increases near the nozzle exit in
X/D<1, then decreases in 2≤X/D≤4. Further downstream beyond X/D=4, the LES results show little difference
among them.
In Fig. 10, three cross sections X/D=0.12, 0.2, and 0.6 are shown to observe process that the microjet
injection generates shear-layer deformation and turbulence which is difficult to measure in experiments. The left
column is rm(3)
and the right column is rm(7)
. In each figure, axial velocity (U) is shown on the left, and turbulent
kinetic energy (TKE) is shown on the right. At X/D=0.12 (Fig. 10a), microjets impingement make “concave
regions” on the shear layer between main jet and the ambient air, and higher TKE region appears at the bottom
of the concave regions. Note that microjets impinge at around X/D=0.06 to the shear layer, and they merge into
the main jet after the impingement. Figure 11 shows closer view of the Z=0 section near microjet injections.
This result clearly shows that the turbulence is not generated by the microjet impingement itself but generated in
the strong shear in outer side of the shear-layer downstream of the impingement. In Fig. 10b, the regions of
higher TKE develops along concave shaped shear-layer. At a downstream location of X/D=0.6 (Fig. 10c), high
TKE regions extend circumferentially and connect with each other. For rm(7)
, on the other hand, microjets
deform the shear layer deeper, but not wider like rm(3)
at X/D=0.12 (Fig. 10d). In Fig. 10e, the shear layer
deforms and shows omega (Ω) shape, and TKE exhibits high value along the shear layer. In Fig. 10f the higher
TKE region spreads along the shear-layer between the microjet locations. It is interesting that in the rm(3)
case
TKE spreads almost uniformly along the deformed shear-layer. On the other hand, in the rm(7)
case, it spreads
outer side of the shear-layer after the omega shape deformation due to strong microjet injection. To enhance the
turbulent mixing in the shear-layer, it is preferred that the TKE spreads along the middle in the shear-layer
thickness like rm(3)
results. Therefore, injection for rm(7)
may be too much to obtain optimal efficient condition
that realizes noise reduction with less mass flow injection. Actually, the experimental results in Castelain22
shows rm(3)
obtained a comparable level of noise reduction to rm(7)
which mass flow rate is more than twice as
much.
![Page 6: [American Institute of Aeronautics and Astronautics 17th AIAA/CEAS Aeroacoustics Conference (32nd AIAA Aeroacoustics Conference) - Portland, Oregon ()] 17th AIAA/CEAS Aeroacoustics](https://reader035.vdocuments.pub/reader035/viewer/2022080405/575093351a28abbf6bae1b00/html5/thumbnails/6.jpg)
American Institute of Aeronautics and Astronautics
6
C. Sound Pressure Level
Figure 12 compares frequency weighted power spectral density (Pa2) of far-field sound pressure level
obtained at R/D=40 between the experiment6 and the LES. Figure 12a is observation angle of θ = 30° from the
downstream jet axis. Low frequency component is dominant for this angle. The experiment shows 30%
reduction with microjet (both for rm(3)
and rm
(7)) at the peak frequency of 1.5kHz. The LES result shows half
value of the experiment at the peak, while peak frequency is close to the experiment. Although in the spectra for
rm(3)
, LES fails to obtain noise reduction, for rm(7)
it successfully shows noise reduction. Absolute level of spectra
is different but ratio of noise level between microjet on and off is close to the experiment. For θ = 90° (Fig. 12b),
the SPL of the experiment exhibits peak frequency of 5kHz, and about 40% of noise reduction are achieved with
both rm(3)
and rm(7)
. Slight increase of SPL of the experiment are seen at 20k-40kHz of rm(7)
, and this is caused by
strong interaction between the microjets and the jet-mixing layer6. On the other hand, the LES results show peak
frequency at 10k-20kHz, and they do not show noise reduction distinctly with the microjet injection. Excessive
peak are seen at 30kHz on rm(7)
, and this may be also caused by the interaction between microjets and the mixing
layer. This over-prediction might also be related to the imposed inflow disturbance. It is still under investigation.
The 1/3 octave band SPL of the LES result is shown in Fig. 13. For the observation angle θ = 30° (Fig. 13a),
noise reduction is 1.5dB at the peak frequency (St = 0.2) for rm(7)
, but noise reduction is not clearly seen for rm(3)
.
High frequency noise, St > 2, is increased for both cases. For the observation angle θ = 90° (Fig. 13b), 0.5-1dB
noise reduction is seen for rm(7)
, and no noise reduction is seen for rm(3)
. Again, high frequency noises are
increased. Directivity of overall sound pressure level (OASPL) is shown in Fig. 14. For rm(7)
, sound pressure
level decreases at low observation angle θ < 55°, and increases at high observation angle θ > 55°. No noise
reduction is seen for rm(3)
.
V. Conclusion
Large-Eddy Simulations of jet flow with microjet injection were executed using 476 million grid points on
980 nodes of JAXA Supercomputer System (JSS). Flow field and far-field noise prediction of the baseline jet
(without microjet) was within 2-5dB accuracy compared with the experimental data. Two test points of microjet
injection, rm(3)
and rm
(7) were numerically simulated. Shear layer deformation and turbulent velocity change
caused by microjet injection were simulated satisfactory. The detail figures of microjets injection suggest how
microjets increase turbulence kinetic energy.
Far-field noise prediction using FW-H method from the LES results show that the LES successfully predict
noise reduction by microjets for lower frequency component emitted in 30 deg. observation angle, while it still
has difficulty in predicting reduction of higher frequency noise emitted in 90 deg. observation angle.
Acknowledgments
We are indebted to many people for this research. First of all, we would like to thank Prof. Castelain of Univ.
of Lyon and Prof. Bogey of Ecole Centrale de Lyon for detail discussion about microjet experiment and
simulation, allowing us to use experimental data in this paper to compare with the LES result. We also would
like to thank Mr. M. Huet and Dr. F. Vuillot of ONERA for the discussion on the computational method for
microjet. Our colleagues in JAXA, Dr. Imamura and Dr. Amemiya helped us with the FW-H code for the far-
field noise calculation.
The computational resource was provided by "Strategic Large-Scale Simulation scheme" of JAXA Super-
computer System (JSS).
References 1 V. H. Arakeri, et al., “On the use of microjets to suppress turbulence in a Mach 0.9 axisymmetric jet”, J. Fuic Mech.
vol.490, pp.75-98, 2003 2 M. B. Alkislar, et al., “The effect of streamwise vortices on the aeroacoustics of a Mach 0.9 jet”, J. Fluid Mech.
vol.578, pp.139-169, 2007 3 M. B. Alkislar, “Flow Characteristics of a Jet Controlled with Chevron-Microjet Combination for Noise Reduction”,
AIAA 2009-851 4 S. M. Harrison et al., “Jet Noise Reduction by Fluidic Injection On a Separate Flow Exhaust System”, AIAA 2007-
439 5 K. B. M. Q. Zaman, “Subsonic jet noise reduction by microjets – parametric study”, Int. J. of Aeroacoustics, vol.9,
No.6, 2010 6 T. Castelain, M. Sunyach, D. Juve, J.C. Bera, “Jet-Noise Reduction by Impinging Microjets: An Acoustic
Investigation Testing Microjet Parameters”, AIAA Journal Vol.46, No.5, 2008 7 Bodony and Lele, “On using large-eddy simulation for the prediction of noise from cold and heated turbulent jets”,
Physics of Fluids, 17, 085103, 2005 8 Bogey and Bailly, “Effects of Inflow Conditions and Forcing on Subsonic Jet Flows and Noise”, AIAA Journal,
Vol.43, No.5, 2005 9 Uzun, Blaisdell, Lyrintzis, “3-D Large Eddy Simulation for Jet Aeroacoustics”, AIAA 2003-3322
![Page 7: [American Institute of Aeronautics and Astronautics 17th AIAA/CEAS Aeroacoustics Conference (32nd AIAA Aeroacoustics Conference) - Portland, Oregon ()] 17th AIAA/CEAS Aeroacoustics](https://reader035.vdocuments.pub/reader035/viewer/2022080405/575093351a28abbf6bae1b00/html5/thumbnails/7.jpg)
American Institute of Aeronautics and Astronautics
7
10 Shur, Spalart, Strelets and Garbaruk, “Further Steps in LES-Based Noise Prediction for Complex Jets”, AIAA 2006-
485 11 M. Huet et al., “Numerical investigation of the micro-jets efficiency for jet noise reduction”, AIAA 2009-3127 12 M. L. Shur et al., “LES-Based Evaluation of a Microjet Noise Reduction Concept in Static and Flight Conditions”,
Procedia Engineering, 6 (2010) 44-53 13 Takaki et al., “The Development of the UPACS CFD Environment”, High Performance Computing, Proceedings of
ISHPC 2003, Springer, pp.307-319, 2003 14 Imamura et al., “Numerical Simulation of NACA0012 Wingtip Flow Leading to Noise Generation”, AIAA 2005-
2864 15 Imamura et al., “Designing of Slat Cove Filler as a Noise Reduction Device for Leading-edge Slat”, AIAA 2007-
3473 16 Enomoto et al., “Large-Eddy Simulation of High-Subsonic Jet Noise Reduction with Microjet Injection”, JSASS-
2009-0191-A (in Japanese) 17 Vreman, Geurts and Kuerten, “Subgrid-modeling in LES of Compressible Flow”, Direct and Large-Eddy Simulation
I, Kluwer Academic Publishers, 133-144, 1994 18 Kobayashi, “On a Class of Pade Finite Volume Methods”, J. Comp. Phys. 156, 127-180, 1999 19 S.Enomoto et.al., “Large Eddy Simulation of Jet Noise using Multi-Block Structured Grid”, IGTC-ABS-148,
International Gas Turbine Congress, 2007 20 Gaitonde and Visbal, “Pade-Type Higher-Order Boundary Filters for the Navier-Stokes Equations”, AIAA Journal
Vol.38, No.11, 2000 21 T. Castelain, M. Sunyach, D. Juve, “Jet Noise reduction by impinging microjets: an aerodynamic investigation
testing microjet parameters”, 13th AIAA/CEAS Aeroacoustics Conference, AIAA 2007-3419 22 Castelain, T., 2006, “Control of jet by impinging microjets. Measurement of radiated noise and aerodynamic
analysis”, Ph.D. thesis, ECL-No.2006-33 23 J. B. Freund, “Proposed In-flow/Outflow Boundary Condition for Direct Computation of Aerodynamic Sound”,
AIAA Journal Vol.35, No.4, pp.740, April 1997 24 C. Bogey, C. Bailly, and D. Juve, “Noise Investigation of a High Subsonic Moderate Reynolds Number Jet Using a
Compressible Large Eddy Simulation”, Theoret. Comput. Fluid Dynamics 16: 273–297, 2003 25 J. Bridges and M. P. Wernet, “Establishing Consensus Turbulence Statistics for Hot Subsonic Jets”, AIAA 2010-
3751 26 J. C. Lau, P. J. Morris and M. J. Fisher, “Measurements in subsonic and supersonic free jets using a laser
velocimeter”, J. Fluid Mech., vol. 93, part 1, pp.1-27, 1979 27 H. K. Tanna, “An Experimental study of Jet Noise, PART I: Turbulent Mixing Noise”, J. of Sound and Vibration,
50(3), pp.405-428, 1977 28 Ffowcs Williams, J. E., and Hawkings, D. L., "Sound Generation by Turbulence and Surfaces in Arbitrary Motion,"
Phil. Trans. Roy. Soc. (London), Ser. A, 264, 321-342. (Philosophical Transactions of the Royal Society of London,
Vol. 264A, May 1969, pp. 321-342.) 29 A.S.Lyrintzis, "Surface integral methods in computational aeroacoustics - From the (CFD) near-field to the
(Acoustic) far-field", Inter. J. Aeroacoustics, Vol. 2, No. 2, 2003, pp.95-128.
![Page 8: [American Institute of Aeronautics and Astronautics 17th AIAA/CEAS Aeroacoustics Conference (32nd AIAA Aeroacoustics Conference) - Portland, Oregon ()] 17th AIAA/CEAS Aeroacoustics](https://reader035.vdocuments.pub/reader035/viewer/2022080405/575093351a28abbf6bae1b00/html5/thumbnails/8.jpg)
American Institute of Aeronautics and Astronautics
8
(a) conputational grid (every 16
th grid line is shown)
(b) nozzle exit region (every 4
th grid line is shown)
(c) microjet injection region
(d) Main nozzle exit and microjet injection (bird’s view)
Figure 1 Computational Grid
X/D
Y/D
0 10 20
-4
-2
0
2
4
X/D
Y/D
-1 -0.5 0 0.5 1 1.5 2-1.5
-1
-0.5
0
0.5
1
1.5
X/D
Y/D
-0.1 0 0.1 0.2
0.4
0.5
0.6
![Page 9: [American Institute of Aeronautics and Astronautics 17th AIAA/CEAS Aeroacoustics Conference (32nd AIAA Aeroacoustics Conference) - Portland, Oregon ()] 17th AIAA/CEAS Aeroacoustics](https://reader035.vdocuments.pub/reader035/viewer/2022080405/575093351a28abbf6bae1b00/html5/thumbnails/9.jpg)
American Institute of Aeronautics and Astronautics
9
(a) Centerline axial mean velocity (r/D=0.0)
(b) Lipline axial mean velocity (r/D=0.5)
(c) Centerline axial turbulent velocity (r/D=0.0)
(d) Lipline axial turbulent velocity (r/D=0.5)
(e) Centerline transverse turbulent velocity (r/D=0.0)
Figure 2 Comparison with the experimental data: axial profiles. Axial mean and turbulent velocities on the centerline and on the lipline are compared with the experimental data
of Bridges25
and Lau26
X/D
U/U
j
5 10 15 200
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
coarse
fine
Bridges
Lau
X/D
U/U
j
0 5 10 15 20 250
0.2
0.4
0.6
0.8
coarse
fine
Bridges
X/D
u'/U
j
0 5 10 15 20 250
0.02
0.04
0.06
0.08
0.1
0.12
0.14
0.16coarsefineBridgesLau
X/D
u'/U
j
0 5 10 15 20 250
0.05
0.1
0.15
0.2
0.25
coarse
fine
Bridges
X/D
v'/U
j
0 5 10 15 20 250
0.02
0.04
0.06
0.08
0.1
0.12
0.14
0.16
coarse
fine
Lau
![Page 10: [American Institute of Aeronautics and Astronautics 17th AIAA/CEAS Aeroacoustics Conference (32nd AIAA Aeroacoustics Conference) - Portland, Oregon ()] 17th AIAA/CEAS Aeroacoustics](https://reader035.vdocuments.pub/reader035/viewer/2022080405/575093351a28abbf6bae1b00/html5/thumbnails/10.jpg)
American Institute of Aeronautics and Astronautics
10
(a) Axial mean velocity, X/D = 4
(b) Axial turbulent velocity, X/D = 4
(c) Axial mean velocity, X/D = 8
(d) Axial turbulent velocity, X/D = 8
Figure 3 Comparison with the experimental data: radial profiles.
Axial mean and turbulent velocities at X/D = 4, 8, 12 and 16 are compared
with the experimental data of Bridges25
and Lau26
Y/D
U/U
j
-1 -0.5 0 0.5 1
0.2
0.4
0.6
0.8
1coarsefineBridgesLau
Y/D
u'/U
j
-1.5 -1 -0.5 0 0.5 1 1.5-0.02
0
0.02
0.04
0.06
0.08
0.1
0.12
0.14
0.16
0.18
0.2coarsefineBridgesLau
Y/D
U/U
j
-1 -0.5 0 0.5 1
0.2
0.4
0.6
0.8
1coarsefineBridgesLau
Y/D
u'/U
j
-1.5 -1 -0.5 0 0.5 1 1.5-0.02
0
0.02
0.04
0.06
0.08
0.1
0.12
0.14
0.16
0.18
0.2
coarsefineBridgesLau
![Page 11: [American Institute of Aeronautics and Astronautics 17th AIAA/CEAS Aeroacoustics Conference (32nd AIAA Aeroacoustics Conference) - Portland, Oregon ()] 17th AIAA/CEAS Aeroacoustics](https://reader035.vdocuments.pub/reader035/viewer/2022080405/575093351a28abbf6bae1b00/html5/thumbnails/11.jpg)
American Institute of Aeronautics and Astronautics
11
(e) Axial mean velocity, X/D = 12
(f) Axial turbulent velocity, X/D = 12
(g) Axial mean velocity, X/D = 16
(h) Axial turbulent velocity, X/D = 16
Figure 3 Comparison with the experimental data: radial profiles. (cont’d) Axial mean and turbulent velocities at X/D = 4, 8, 12 and 16 are compared
with the experimental data of Bridges25
and Lau26
Y/D
U/U
j
-1 -0.5 0 0.5 1
0.2
0.4
0.6
0.8
1coarsefineBridgesLau
Y/D
u'/U
j
-1.5 -1 -0.5 0 0.5 1 1.5-0.02
0
0.02
0.04
0.06
0.08
0.1
0.12
0.14
0.16
0.18
0.2coarsefineBridgesLau
Y/D
U/U
j
-1 -0.5 0 0.5 1
0.2
0.4
0.6
0.8
1coarsefineBridgesLau
Y/D
u'/U
j
-1.5 -1 -0.5 0 0.5 1 1.5-0.02
0
0.02
0.04
0.06
0.08
0.1
0.12
0.14
0.16
0.18
0.2
coarsefineBridgesLau
![Page 12: [American Institute of Aeronautics and Astronautics 17th AIAA/CEAS Aeroacoustics Conference (32nd AIAA Aeroacoustics Conference) - Portland, Oregon ()] 17th AIAA/CEAS Aeroacoustics](https://reader035.vdocuments.pub/reader035/viewer/2022080405/575093351a28abbf6bae1b00/html5/thumbnails/12.jpg)
American Institute of Aeronautics and Astronautics
12
(a) Overall SPL(dB)
(b) 1/3 octave SPL, θ = 90°
(c) 1/3 octave SPL, θ =30°
Figure 4 Comparison with the experimental data : Far-fleld noise level
angle
OA
SP
L
40608010012098
100
102
104
106
108
110
112
114
LES
Tanna
St
SP
L(d
B)
10-1
100
101
70
75
80
85
90
95
100
105
110
LES
Tanna
St
SP
L(d
B)
10-1
100
10170
75
80
85
90
95
100
105
110
LES
Tanna
observation angle θ
![Page 13: [American Institute of Aeronautics and Astronautics 17th AIAA/CEAS Aeroacoustics Conference (32nd AIAA Aeroacoustics Conference) - Portland, Oregon ()] 17th AIAA/CEAS Aeroacoustics](https://reader035.vdocuments.pub/reader035/viewer/2022080405/575093351a28abbf6bae1b00/html5/thumbnails/13.jpg)
American Institute of Aeronautics and Astronautics
13
(a) LES, baseline (rm
(0))
(b) Experiment, baseline (rm
(0))
(c) LES, rm
(3)
(d) Experiment, rm
(3)
(e) LES, rm
(7)
(f) Experiment, rm
(7)
Figure 5 Axial turbulent velocity with microjets, X/D=1 left: present LES, right: experiment of Castelain
22
![Page 14: [American Institute of Aeronautics and Astronautics 17th AIAA/CEAS Aeroacoustics Conference (32nd AIAA Aeroacoustics Conference) - Portland, Oregon ()] 17th AIAA/CEAS Aeroacoustics](https://reader035.vdocuments.pub/reader035/viewer/2022080405/575093351a28abbf6bae1b00/html5/thumbnails/14.jpg)
American Institute of Aeronautics and Astronautics
14
(a) LES, baseline (rm
(0))
(b) Experiment, baseline (rm
(0))
(c) LES, rm
(3)
(d) Experiment, rm
(3)
(e) LES, rm
(7)
(f) Experiment, rm
(7)
Figure 6 Axial turbulent velocity with microjets, X/D=3 left: present LES, right: experiment of Castelain
22
![Page 15: [American Institute of Aeronautics and Astronautics 17th AIAA/CEAS Aeroacoustics Conference (32nd AIAA Aeroacoustics Conference) - Portland, Oregon ()] 17th AIAA/CEAS Aeroacoustics](https://reader035.vdocuments.pub/reader035/viewer/2022080405/575093351a28abbf6bae1b00/html5/thumbnails/15.jpg)
American Institute of Aeronautics and Astronautics
15
(a) LES, baseline (rm
(0))
(b) Experiment, baseline (rm
(0))
(c) LES, rm
(3)
(d) Experiment, rm
(3)
(e) LES, rm
(7)
(f) Experiment, rm
(7)
Figure 7 Reynolds stress (√ ), X/D=1
left: present LES, right: experiment of Castelain22
![Page 16: [American Institute of Aeronautics and Astronautics 17th AIAA/CEAS Aeroacoustics Conference (32nd AIAA Aeroacoustics Conference) - Portland, Oregon ()] 17th AIAA/CEAS Aeroacoustics](https://reader035.vdocuments.pub/reader035/viewer/2022080405/575093351a28abbf6bae1b00/html5/thumbnails/16.jpg)
American Institute of Aeronautics and Astronautics
16
(a) LES, baseline (rm
(0))
(b) Experiment, baseline (rm
(0))
(c) LES, rm
(3)
(d) Experiment, rm
(3)
(e) LES, rm
(7)
(f) Experiment, rm
(7)
Figure 8 Reynolds stress (√ ), X/D=3
left: present LES, right: experiment of Castelain22
3
3
3
![Page 17: [American Institute of Aeronautics and Astronautics 17th AIAA/CEAS Aeroacoustics Conference (32nd AIAA Aeroacoustics Conference) - Portland, Oregon ()] 17th AIAA/CEAS Aeroacoustics](https://reader035.vdocuments.pub/reader035/viewer/2022080405/575093351a28abbf6bae1b00/html5/thumbnails/17.jpg)
American Institute of Aeronautics and Astronautics
17
(a) baseline (rm
(0))
(b) rm
(3)
(c) rm
(7)
Figure 9 Turbulent Kinetic Energy, Z=0
![Page 18: [American Institute of Aeronautics and Astronautics 17th AIAA/CEAS Aeroacoustics Conference (32nd AIAA Aeroacoustics Conference) - Portland, Oregon ()] 17th AIAA/CEAS Aeroacoustics](https://reader035.vdocuments.pub/reader035/viewer/2022080405/575093351a28abbf6bae1b00/html5/thumbnails/18.jpg)
American Institute of Aeronautics and Astronautics
18
(a) rm
(3), X/D=0.12
(d) rm
(7), X/D=0.12
(b) rm
(3), X/D=0.2
(e) rm
(7), X/D=0.2
(c) rm
(3), X/D=0.6
(f) rm
(7), X/D=0.6
Figure 10 Axial velocity and turbulent kinetic energy distribution
![Page 19: [American Institute of Aeronautics and Astronautics 17th AIAA/CEAS Aeroacoustics Conference (32nd AIAA Aeroacoustics Conference) - Portland, Oregon ()] 17th AIAA/CEAS Aeroacoustics](https://reader035.vdocuments.pub/reader035/viewer/2022080405/575093351a28abbf6bae1b00/html5/thumbnails/19.jpg)
American Institute of Aeronautics and Astronautics
19
(a) rm
(3)
(b) rm
(7)
Figure 11 Velocity magnitude (color) and TKE (line) Closer view of the Z=0 section near microjet injections.
0.01
0.01
0.02
0.02
0.030.04
0.04
X/D
Y/D
0 0.1 0.2 0.3
0.4
0.45
0.5
0.55
0.6
Vmag
0.85
0.75
0.65
0.55
0.45
0.35
0.25
0.15
0.05
rm3
0.01
0.01
0.01
0.01
0.02
0.02
0.02
0.02
0.03
0.03
0.03
0.04
0.04
X/D
Y/D
0 0.1 0.2 0.3
0.4
0.45
0.5
0.55
0.6
Vmag
0.85
0.75
0.65
0.55
0.45
0.35
0.25
0.15
0.05
rm7
![Page 20: [American Institute of Aeronautics and Astronautics 17th AIAA/CEAS Aeroacoustics Conference (32nd AIAA Aeroacoustics Conference) - Portland, Oregon ()] 17th AIAA/CEAS Aeroacoustics](https://reader035.vdocuments.pub/reader035/viewer/2022080405/575093351a28abbf6bae1b00/html5/thumbnails/20.jpg)
American Institute of Aeronautics and Astronautics
20
(a) θ = 30°
(b) θ = 90°
Figure 12 Far-field Sound Pressure Level (Frequency weighted PSD)
LES and the experiment7
Freq. (Hz)
St
Fre
q.w
eig
hte
dP
SD
(Pa
2)
103
104
10-1
100
0
5
10
15
20
25
30
35
40
45
50
Exp-rm0
Exp-rm3
Exp-rm7
LES-rm0
LES-rm3
LES-rm7
Exp. : Castelain et al.,AIAA J. Vol46, No.5, 2008
Freq. (Hz)
St
Fre
q.w
eig
hte
dP
SD
(Pa
2)
103
104
10-1
100
0
0.5
1
1.5
2
2.5
3
Exp-rm0
Exp-rm3
Exp-rm7
LES-rm0
LES-rm3
LES-rm7
Exp. : Castelain et al.,AIAA J. Vol46, No.5, 2008
![Page 21: [American Institute of Aeronautics and Astronautics 17th AIAA/CEAS Aeroacoustics Conference (32nd AIAA Aeroacoustics Conference) - Portland, Oregon ()] 17th AIAA/CEAS Aeroacoustics](https://reader035.vdocuments.pub/reader035/viewer/2022080405/575093351a28abbf6bae1b00/html5/thumbnails/21.jpg)
American Institute of Aeronautics and Astronautics
21
(a) θ = 30°
(b) θ = 90°
Figure 13 Far-field Sount Pressure Level, 1/3 octave band SPL
St
SP
L(d
B)
10-1
100
10170
75
80
85
90
95
100
105
110
LES(rm0)
LES(rm3)
LES(rm7)
St
SP
L(d
B)
10-1
100
101
70
75
80
85
90
95
100
105
110
LES(rm0)
LES(rm3)
LES(rm7)
![Page 22: [American Institute of Aeronautics and Astronautics 17th AIAA/CEAS Aeroacoustics Conference (32nd AIAA Aeroacoustics Conference) - Portland, Oregon ()] 17th AIAA/CEAS Aeroacoustics](https://reader035.vdocuments.pub/reader035/viewer/2022080405/575093351a28abbf6bae1b00/html5/thumbnails/22.jpg)
American Institute of Aeronautics and Astronautics
22
Figure 14 Far-field Sound Pressure Level, overall SPL
angle
OA
SP
L
40608010012098
100
102
104
106
108
110
112
114
LES(rm0)
LES(rm3)
LES(rm7)
observation angle θ