Sie sind auf Seite 1von 13

JOURNAL OF AIRCRAFT

Design and Analysis of a Supersonic Jet Noise Reduction Concept


Anthony R. Pilon∗
Lockheed Martin Aeronautics Company, Palmdale, California 93599
and
Russell W. Powers,† Dennis K. McLaughlin,‡ and Philip J. Morris§
Pennsylvania State University, University Park, Pennsylvania 16802
DOI: 10.2514/1.C033977
The design and analysis of a simple jet noise reduction concept is discussed. The concept is intended for use on the
type of supersonic exhaust nozzles typically employed on tactical aircraft. The concept addresses both broadband
shock-associated noise and turbulent mixing noise. Broadband shock-associated noise is addressed through a
reduction in the effective exhaust area, which reduces the strength of the shock cell structure in the jet. Turbulent
mixing noise is reduced by enhancement of mixing in the jet shear layer, which creates a shorter region of strongly
turbulent flow and breaks up the large-scale turbulent structures. This paper describes an active control device that
Downloaded by UNIV. OF ARIZONA on March 14, 2017 | http://arc.aiaa.org | DOI: 10.2514/1.C033977

deflects a fraction of the adjustable seals in the divergent section of the engine exhaust nozzle. After takeoff, the
deflected seals are returned to their undeflected position. Very favorable jet noise reductions are demonstrated at
multiple observer angles for nominal takeoff conditions.

I. Introduction and breaks up the large-scale structures that are responsible for noise
radiation in the peak noise direction. This then leads to an overall
T HE high-speed jet flows produced by modern high-performance
military aircraft engines generate high-amplitude noise. These
high noise levels are of increasing concern to the U.S. military
reduction in the intensity of the noise sources.
The issues discussed previously have encouraged the development
because of the hearing loss and other physiological impacts to of many noise reduction technologies [2,3]. To date, these technologies
personnel who work in close proximity to the aircraft. Additionally, have seen limited real-world application on military aircraft. One noise
communities near military air bases are increasingly concerned about reduction technology that is planned for transition to full-scale military
the annoyance and impact to quality of life due to the noise from these engines is exhaust nozzle chevrons [4], which introduce longitudinal
aircraft. One measure of the financial impact of high noise levels is vortices into the jet shear layer to increase mixing of the core and
the cost of hearing loss benefits paid to U.S. military veterans. freestream flows, and thus reduce peak noise as discussed previously.
Doychak [1] highlighted the recent rising costs associated with these The high-frequency noise is generally increased slightly, though this
claims, which are now more than 1 billion U.S. dollars per year. These increase is countered by generally higher atmospheric attenuation at
costs must be countered with effective jet noise reduction those frequencies. The numerous design variables associated with
technologies as well as improved hearing protection. The weight chevron development require extensive analysis and experimentation
and thrust requirements of the high-performance aircraft involved, to produce designs that have acceptable noise reduction performance
and the low bypass ratio engines they employ, make development of and minimal thrust performance penalties [5]. The addition of chevrons
these noise reduction technologies very difficult. could also adversely impact the radar signature of a tactical aircraft,
The noise emitted by the jet flows from military-style converging– adding to the penalties associated with their use.
diverging nozzles is due to the turbulent mixing of air in the Additional noise reduction technologies are also in development.
supersonic exhaust jet, as well as unsteady motion of shock waves One technique that has shown promise is the corrugated seal inserts
and shock wave–turbulence interaction. The reduction of supersonic developed at NASA Langley Research Center and the University of
jet noise typically involves different methodologies to reduce the Mississippi by Seiner et al. [6,7] and Murray and Jansen [8]. When
turbulent mixing noise and the broadband shock-associated noise used in a military-style converging–diverging nozzle, these inserts
(BBSAN). In general, a direct approach to reduce BBSAN is the have been shown to reduce both large-scale mixing noise and
alteration of the nozzle geometry to produce a more fully expanded broadband shock-associated noise. The inserts change the area
flow so that the exhaust and freestream pressures are nearly matched progression in the divergent section of the nozzle to more ideally
and there is minimal strength in the shock cell structure of the jet. expand the flow and weaken the resulting shock cell structure, and
Turbulent mixing noise reduction is often achieved by enhancement thus reduce the BBSAN. The inserts also generate streamwise
of mixing in the jet shear layer. Enhancement of mixing between the vortices in a similar fashion to chevrons, which leads to similar
jet core flow and the low-speed ambient air adds low-energy ambient mixing noise reductions. However, the corrugated seal inserts have a
air to the jet. This results in a shorter region of strongly turbulent flow major drawback in that their presence will unacceptably degrade thrust
and mission performance away from the takeoff operating point.
Researchers at the Pennsylvania State University have recently
Presented as Paper 2014-0525 at the AIAA Science and Technology Forum developed a noise reduction technique whereby the noise reduction of
and Exhibition, National Harbor, MD, 13–17 January 2014; received 14 April
2016; revision received 20 December 2016; accepted for publication 23
the hard-wall corrugated seal inserts developed by Seiner et al. [6,7] is
December 2016; published online 8 March 2017. Copyright © 2016 by achieved with “fluidic inserts” [9,10]. They demonstrated 3–5 dB
Lockheed Martin Corporation. Published by the American Institute of noise reductions through the use of blowing in the nozzle divergent
Aeronautics and Astronautics, Inc., with permission. All requests for copying section such that the nozzle flow replicated the flow in a nozzle with the
and permission to reprint should be submitted to CCC at www.copyright.com; hard-wall corrugations. Because the blowing was only employed at
employ the ISSN 0021-8669 (print) or 1533-3868 (online) to initiate your takeoff, the thrust and mission performance penalties associated with
request. See also AIAA Rights and Permissions www.aiaa.org/randp. permanent hard-wall inserts were nearly eliminated.
*Lockheed Martin Fellow, Advanced Development Programs, 1011 This paper describes a simple jet noise reduction concept that was
Lockheed Way. Associate Fellow AIAA.

Graduate Research Assistant, DoD SMART Scholar, Department of
developed in a similar fashion to the University of Mississippi and
Aerospace Engineering; currently Aerospace Engineer, Naval Air Systems Pennsylvania State University techniques. The concept discussed
Command, Patuxent River, MD 20670. Member AIAA. here seeks to reduce the noise generated by jets issuing from military-
‡ style exhaust nozzles by reducing both the turbulent mixing noise and
Emeritus Professor, Department of Aerospace Engineering. Fellow AIAA.
§
Boeing/A.D. Welliver Professor of Aerospace Engineering. Fellow AIAA. the broadband shock-associated noise of the jets. Both of these
Article in Advance / 1
2 Article in Advance / PILON ET AL.

objectives are met through a modification to the position of a number (CAD) models of a baseline (unmodified) nozzle, a “deflected seals”
of the divergent seals in a variable-geometry nozzle. After takeoff, nozzle, and an approximate configuration where the deflected seals
when the aircraft is at altitude and noise reduction is no longer are replicated with solid wedges. The use of solid wedges allows for
required, the divergent seals are returned to their normal operating more structural stability in the wind-tunnel model nozzles discussed
locations. As a result, jet noise is reduced with very minimal mission in the following. Approximate configurations, where solid wedges
performance penalties. replicate deflected seals, are used for all computational fluid
The following section describes the concept in detail. This is dynamics (CFD) analyses and experiments in this study.
followed by a description of a computational fluid dynamics analysis of
the concept viability. Then, small-scale experimental data are presented.
This is followed by a discussion of future work and conclusions. III. Computational Fluid Dynamics Simulations
The effectiveness of the proposed concept and the associated thrust
losses have been estimated through the use of steady-state Reynolds-
II. Jet Noise Reduction Concept Description averaged Navier–Stokes CFD simulations. The CFD solver employed
High-performance tactical aircraft are powered by low-bypass- was CFD++ from Metacomp Technologies, Inc. [12]. Multiblock,
ratio turbofan engines and variable-geometry convergent–divergent structured grids were generated for a baseline nozzle, as well as several
nozzles. A photograph of a common military supersonic exhaust deflected seals nozzles. The grids were generated using the exact CAD
nozzle is shown in Fig. 1. The jet flows that emanate from these nozzles geometries used in the construction of nozzles for the experimental
are normally supersonic and overexpanded (static pressure at the study discussed in the following. The same resolution was used in each
nozzle exit is lower than ambient static pressure) at high power of the grids to allow for qualitative comparisons between respective
takeoff conditions. The variable-geometry convergent–divergent flow solutions. A standard k–ϵ turbulence model was employed in the
Downloaded by UNIV. OF ARIZONA on March 14, 2017 | http://arc.aiaa.org | DOI: 10.2514/1.C033977

nozzles are typically constructed of hinged metal “flaps” and “seals” solver.
arranged around the exhaust axis. The flaps are actuated through
linkages so that the cross-sectional flow area is reduced along the axis A. Qualitative Flow Analysis
in a “convergent” section, and then the area is increased in a Figure 3 shows Mach number contours for both the baseline
“divergent” section. The circular arrangement of hinges between nozzle and a deflected seals nozzle where every other seal (six in
these sections is referred to as the nozzle “throat.” total) is deflected 4.99 deg toward the centerline, referred to in the
In the present concept [11], some of the seals in the divergent following as the “6DS5” nozzle. The top half of the figure shows the
section are deflected into the exhaust flow. The primary baseline configuration, whereas the bottom shows the deflected
implementation investigated to date has been to deflect every other seals configuration. The cutting plane passes directly through the
seal or every fourth seal into the exhaust flow by a small amount: 5– center of a deflected seal. The design Mach number of the baseline
10 deg. The additional vorticity created by this deflection increases nozzle is 1.65, and the exit diameter is 0.708 in. For all simulations
mixing with the low-energy external air, thereby reducing turbulent shown in this section, the nozzle pressure ratio (NPR) is 3.25, the
mixing noise in the near and far fields. In addition, the effective exit nozzle temperature ratio (hereafter referred to as TTR) is 3.00, and
area of the nozzle is changed, which reduces the strength of the shock the jet is static; i.e., there is no forward flight stream. Figure 4 is
cell structure (for a typical takeoff overexpanded condition) and the from the same simulation, with the cutting plane passing through
associated broadband shock-associated noise. Nozzle seals are the nondeflected seal midway between two deflected seals. Evident
generally supported by the adjacent flaps. In the current nozzle in these figures is the significant weakening of the shock cell
concept, the support from adjacent flaps is unconstrained, and the structure in the jet issuing from the 6DS5 nozzle when compared
divergent seals are deflected into the exhaust flow. This system will with the baseline nozzle.
be employed only during takeoff, where the throat and exit areas are Figure 5 shows calculated Mach number contours at a series of
at their minimum settings. The system is designed to deflect the seals downstream axial stations for the baseline jet flow. As expected, the
past the normal minimum for takeoff, and then to operate the nozzle flowfield is essentially circular beyond 0.5Dj . Figure 6 shows Mach
normally throughout the rest of the flight envelope. Thus, the weight number contours under the same conditions and at the same
and performance penalties inherent with many jet noise reduction downstream stations for the 6DS5 deflected seals configuration.
technologies are minimized. Figure 2 depicts computer-aided design Strong mixing of the jet flow with freestream air is evident to 2Dj
downstream; after which, the flowfield becomes more circular.
Figure 7 shows Mach number contours for both the baseline nozzle
and a deflected seals nozzle where every fourth seal (three in total) is
deflected 9.95 deg towards the centerline, referred to in the following
as the “3DS10” nozzle. The top half of the figure shows the baseline
configuration, whereas the bottom shows the deflected seals
configuration. The cutting plane passes directly through the center of
a deflected seal. Figure 8 is from the same simulation showing the
other half of the cutting plane (i.e., the portion directly opposite from
the deflected seal), whereas in Fig. 9, the cutting plane passes through
a nondeflected seal adjacent to a deflected seal. Evident in these
figures is the significant weakening of the shock cell structure in the
jet issuing from the 3DS10 nozzle, as well as extensive modification
of the jet flowfield when compared with the baseline flow.
Figure 10 shows Mach number contours at a series of downstream
axial stations for a jet issuing from the 3DS10 nozzle. Strong mixing of
the jet flow with freestream air is evident, and the flowfield remains
noncircular beyond 8Dj . The enhanced mixing and shock strength
reductions evident in these CFD simulations provide confidence in the
noise reduction potential of the proposed methodology.

B. Thrust Estimation
The decrease in effective exit area with the deflected seals concept
Fig. 1 Photograph from the first author's personal files of a GE F414 should increase thrust at an overexpanded takeoff condition because
nozzle installed on an F/A-18E aircraft. the jet flow becomes closer to ideal expansion. However, the increased
Article in Advance / PILON ET AL. 3

Fig. 2 Baseline, deflected seals, and approximate nozzle configurations.


Downloaded by UNIV. OF ARIZONA on March 14, 2017 | http://arc.aiaa.org | DOI: 10.2514/1.C033977

Fig. 3 Comparison of calculated jet flows, with Mach number contours: baseline nozzle (top), and 6DS5 nozzle (bottom). NPR  3.25, TTR  3.00, and
Mj  1.415. Cutting plane passes through a deflected seal.

Fig. 4 Comparison of calculated jet flows, with Mach number contours: baseline nozzle (top), and 6DS5 nozzle (bottom). NPR  3.25, TTR  3.00, and
Mj  1.415. Cutting plane passes through a nondeflected seal.
4 Article in Advance / PILON ET AL.
Downloaded by UNIV. OF ARIZONA on March 14, 2017 | http://arc.aiaa.org | DOI: 10.2514/1.C033977

Fig. 5 Cross-stream Mach number contour plots at axial locations x∕Dj  0.05, 0.5, 1.0, 2.0, 4.0, 6.0, and 8.0. Baseline nozzle with NPR  3.25,
TTR  3.00, and Mj  1.415.

Fig. 6 Cross-stream Mach number contour plots at axial locations x∕Dj  0.05, 0.5, 1.0, 2.0, 4.0, 6.0, and 8.0. 6DS5 nozzle with NPR  3.25,
TTR  3.00, and Mj  1.415.

mixing and increased surface area (in the CFD model and experimental Nozzle thrust was estimated from the results of CFD calculations
nozzles discussed in the following) will lead to viscous losses. Note for the model-scale experimental nozzles discussed in the following.
that the nozzles analyzed here have replaced deflected seal surfaces Thrust was calculated by integrating the axial momentum and static
with solid wedges in order to maintain structural stability in the pressure at the nozzle exit plane for four nozzles: the baseline, 6DS5
experimental nozzles discussed in the following. Therefore, thrust and 3DS10 discussed previously; and a deflected seals nozzle with
estimates on these nozzles should be conservative for implementation three seals deflected 7.95 deg toward the centerline, referred to as the
of the full-scale deflected seals concept. 3DS8 nozzle. Three nozzle pressure ratios were calculated for each
Article in Advance / PILON ET AL. 5

Fig. 7 Comparison of calculated jet flows, with Mach number contours: baseline nozzle (top), and 3DS10 nozzle (bottom). NPR  3.25, TTR  3.00,
Downloaded by UNIV. OF ARIZONA on March 14, 2017 | http://arc.aiaa.org | DOI: 10.2514/1.C033977

and Mj  1.415. Cutting plane passes through a deflected seal.

Fig. 8 Comparison of calculated jet flows, with Mach number contours: baseline nozzle (top), and 3DS10 nozzle (bottom). NPR  3.25, TTR  3.00,
and Mj  1.415. Cutting plane passes directly opposite of a deflected seal.

Fig. 9 Comparison of calculated jet flows, with Mach number contours: baseline nozzle (top), and 3DS10 nozzle (bottom). NPR  3.25, TTR  3.00,
and Mj  1.415. Cutting plane passes through a seal adjacent to a deflected seal.
6 Article in Advance / PILON ET AL.
Downloaded by UNIV. OF ARIZONA on March 14, 2017 | http://arc.aiaa.org | DOI: 10.2514/1.C033977

Fig. 10 Cross-stream Mach number contour plots at axial locations x∕Dj  0.05, 0.5, 1.0, 2.0, 4.0, 6.0, and 8.0. 3DS10 nozzle with NPR  3.25,
TTR  3.00, and Mj  1.415.

nozzle: 3.00, 3.25, and 3.50. All thrust calculations were performed noise characteristics of actual heated jets. This has also been
for a nozzle total temperature ratio of 3.0, at static conditions. Table 1 demonstrated by Doty and McLaughlin [14] and Papamoschou [15].
highlights the change in net static thrust associated with the various Additionally, a comparison between experimental data from the
deflected seals nozzles. Note that the thrust coefficient is laboratory helium–air jet noise facility at Penn State and the moderate-
CT  T∕p0 At , where T is the net thrust, p0 is the nozzle total scale heated jet noise facility at NASA John H. Glenn Research Center
pressure, and At is the cross-sectional area at the nozzle throat. It is has been made. McLaughlin et al. [16] documented how the acoustic
important to note that the seal deflections in the three nozzles studied data measured in the two facilities compare, with very good
were chosen based purely on the associated change in nozzle exit engineering accuracy. All data presented here use the acoustic velocity
area for a particular nozzle pressure ratio. No optimization of the matching method of Kinzie and McLaughlin [13].
deflection angle or number of deflected seals has been performed in The Penn State jet noise anechoic chamber is a 16.5 × 19.8 ×
this preliminary investigation. A full optimization would allow for 9.2 ft room covered with Fiberglass wedges to absorb sound and
the concept to reduce both mixing noise and BBSAN while prevent reflections from the walls. The theoretical cutoff frequency of
minimizing any thrust loss. Additionally, a full-scale implementation the chamber is 250 Hz. An exhaust fan, installed in the downstream
of the concept would not employ solid wedges in place of deflected section of a collector, captures the jet exhaust and minimizes air
seals. The full-scale design could employ novel structural supports, recirculation and possible local helium accumulation in the anechoic
flow control techniques, and/or thermal treatments, so that viscous chamber. Acoustic measurements are typically performed using 11
losses are minimized and thrust is increased. 1∕8-in: pressure field microphones, type 40DP from G.R.A.S. Each
of the microphones is supported by a boom such that the diaphragm is
IV. Experimental Facility, Setup, and Procedure at a grazing incidence to the centerline of the jet exhaust. The
microphones are positioned on an arc at 10 deg intervals from 60 to
A. Jet Noise Facility and Instrumentation
160 deg when measured from the upstream axis. The physical
The effectiveness of the current noise reduction technique has been distance from each microphone to the nozzle exit is approximately
tested in the Pennsylvania State University (Penn State) high-speed 70 in. This distance is sufficient to ensure that the microphones are in
aeroacoustics facility. To produce acoustic measurements that can be the far field when testing nozzles up to 0.7 in. in diameter. A
compared directly to aircraft engine measurements, the temperature of photograph of the facility is shown in Fig. 11.
the jet needs to be replicated. A hotter jet results in different acoustic The Penn State jet noise facility also has a forward-flight
characteristics, due to the increase in jet exit velocity and decrease in simulation capability. When this capability is used, a low-speed
jet density. At Penn State, in order to simulate the flow and acoustic (M  0.17) cold-air jet surrounds the supersonic jet. This capability
properties of a hot jet, Kinzie and McLaughlin [13] demonstrated that is important for the current study, as noise reduction technologies that
the use of a mixture of helium and air was able to capture the dominant appear promising at static conditions are often severely degraded in
the presence of a forward-flight stream. This capability will be
Table 1 Calculated nozzle thrust and percentage change discussed in more detail in the following.

NPR  3.0 NPR  3.25 NPR  3.50 B. Data Acquisition


Baseline thrust lbf 17.70 19.89 22.10
Analog time-domain signals from the calibrated microphones are
Baseline CT 0.845 0.876 0.904
6DS5 change, % 1.1 0.5 0.1 amplified and filtered, both for antialiasing and to remove
3DS8 change, % 0.7 1.1 0.0 undesirable low-frequency noise below 500 Hz. Digital time-domain
3DS10 change, % 0.3 −0.1 −0.5 data are acquired at a sampling rate of 300 kHz. Datasets of 102,400
or 409,600 data points are collected, with the smaller dataset being
Article in Advance / PILON ET AL. 7

Figure 12 shows acoustic power spectral densities for five


microphones at different polar angles measured with the Gen1A
3DS8 nozzle. In this case, the nozzle pressure ratio is 3.0, and the
simulated total temperature ratio is 3.0, corresponding to a nominal
takeoff condition. Two azimuthal angles are shown. Zero degrees
corresponds to the microphone in plane with a deflected seal, and
30 deg corresponds to the microphone being on a plane midway
between two deflected seals. Note that the polar angle is measured
from the upstream jet axis. For some of the jet conditions presented
within this study, the baseline nozzle has large screech tones
evident in the narrowband acoustic spectra. The feedback
mechanism responsible for screech is amplified by the large lip
thickness of the baseline nozzle, but it is not present in the deflected
seals nozzle data. Furthermore, the screech tone feedback
mechanism breaks down with full-scale heated jets. Thus, for more
realistic overall SPL (OASPL) reduction and spectral representa-
Fig. 11 Photograph of the Penn State high-speed jet noise facility. tion, the screech tone is removed manually from the data in the
baseline nozzle case. (The presence of the screech tones could also
increase the broadband jet noise by a small amount. So, the baseline
used for helium–air mixture jets in order to minimize the amount of
levels could be slightly higher than in a nonscreeching jet.)
Downloaded by UNIV. OF ARIZONA on March 14, 2017 | http://arc.aiaa.org | DOI: 10.2514/1.C033977

helium used during an experiment. The raw data are processed using
Figure 13 shows the measured overall sound pressure level
MATLAB to produce power spectral densities with a frequency
reductions achieved with the 3.95 deg deflected seals in this
resolution of either Δf  74 Hz or Δf  296 Hz, as well as a
preliminary test. Note that a 3–4 dB noise reduction is observed in
decibel amplitude referenced to 20 μPa. Three corrections are then
the maximum polar radiation direction. Small amounts of reduction
applied to the raw sound pressure level (SPL) to compute the lossless
of the broadband shock-associated noise are also evident in the
SPL as explained by Kuo et al. [17]. The frequency range of the Penn
sideline and forward arc.
State (PSU)-measured spectra shown in this paper is between 500 Hz
Figures 14 and 15 show acoustic results for the Gen1A 3DS8
and 100 kHz. All spectra shown here are nondimensionalized to the
nozzle. The pressure ratio and temperature ratio are the same as
SPL per unit Strouhal number, where the Strouhal number is defined
those in Figs. 12 and 13. More significant reductions in the mixing
as St  f∕fc , with fc being the characteristic frequency of the jet
noise are observed: as much as ∼5.5 dB. As this nozzle has fewer
defined by fc  Uj ∕Dj. Uj is the fully expanded jet velocity, and Dj
seals deflected, the larger deflection angle likely induces stronger
is the fully expanded diameter of the jet plume. The processing of mixing vortices that reduce the mixing noise more than the nozzle
experimental data into lossless spectra per unit Strouhal number with twice as many deflected seals. Another apparent effect is that
makes comparison easier across different measurement scales. the nozzle with six deflected seals shows very little azimuthal
variation. When the number of deflected seals is reduced to three,
C. Nozzle Description levels in the plane midway between deflected seals show ∼1–2 dB
Small-scale nozzles were built using rapid prototyping and tested more noise reduction than the levels on a plane aligned with a
in the Penn State Anechoic Jet Noise facility to assess the noise deflected seal.
reduction potential of the deflected seals concept. Several nozzles,
with solid wedges replicating deflected seals, were tested and E. Follow-On Acoustics Experiments
compared with a baseline nozzle: the “GE 404” configuration tested The Gen1A nozzles demonstrated significant reduction in mixing
extensively at Penn State. For this research, an exit-to-throat area ratio noise while showing only marginal reduction in broadband shock-
was selected at a typical configuration for a takeoff scenario. The associated noise. This is believed to be due to the nozzles not
design Mach number of the baseline nozzle, based on the area ratio of operating at the optimal nozzle pressure ratio for the effective area
1.29, was 1.65. More details about these military-style supersonic ratio. The success of these initial measurements led to a follow-on set
nozzles can be found in the work of Kuo et al. [18]. Several jet of more detailed measurements with a 25% increase in nozzle exit
pressure ratios and (simulated) total temperature ratios were tested for diameter: the “Gen1B” nozzles. Gen1B nozzle characteristics are
the overexpanded conditions experienced at takeoff. The seal documented in the bottom portion of Table 2.
deflection caused the effective area ratio of the nozzle to decrease. In an attempt to find the optimum nozzle pressure ratio at which
Two different effective area ratios, along with two different numbers each nozzle is most perfectly balanced, measurements at 10 different
of deflected seals, were tested. The specific seals deflected were nozzle pressure ratios and two different azimuthal angles for each
spaced equally around the nozzle azimuth. Gen1B nozzle were performed. These results are summarized in
Figs. 16 and 17.
D. Initial Acoustics Experiments Figure 16 shows the noise reduction average of the downstream
An initial set of acoustic measurements were performed in the polar microphones (130–160 deg), giving a good representation of
Penn State facility on a set of 0.708-in.-diam “Gen1A” nozzles at a peak mixing noise reduction. Measurements for each nozzle and
small number of pressure ratios and simulated total temperature azimuthal angle are plotted as separate lines. The comparison used
ratios. Gen1A nozzle characteristics are documented in the top for each nozzle pressure ratio was the baseline nozzle data, such that a
portion of Table 2. negative number indicates noise reduction. It should be noted that

Table 2 Nozzle list and descriptions


Nozzle identification Exit diameter, in. Number of deflected seals Deflection angle, deg Effective area ratio Effective design Mach number
Baseline 0.708 0 0.00 1.29 1.65
Gen1A 3DS8 0.708 3 7.95 1.15 1.47
Gen1A 6DS4 0.708 6 3.95 1.15 1.47
Baseline 0.885 0 0.00 1.29 1.65
Gen1B 3DS8 0.885 3 7.95 1.15 1.47
Gen1B 3DS10 0.885 3 9.95 1.12 1.41
Gen1B 6DS5 0.885 6 4.99 1.12 1.41
8 Article in Advance / PILON ET AL.
Downloaded by UNIV. OF ARIZONA on March 14, 2017 | http://arc.aiaa.org | DOI: 10.2514/1.C033977

Fig. 12 Far-field acoustic power spectral densities measured for the


baseline and Gen1A 6DS4 nozzle from heat-simulated jets with Fig. 14 Far-field acoustic power spectral densities measured for the
NPR  3.0, Mj  1.36, TTR  3.0, and scaled R∕D  100 (re., baseline and Gen1A 3DS8 nozzle from heat-simulated jets with
reference pressure). NPR  3.0, Mj  1.36, TTR  3.0, and scaled R∕D  100 (re.,
reference pressure).

Fig. 13 OASPL reduction comparison of heat-simulated jets issuing


from the Gen1A 6DS4 nozzle with NPR  3.0, Mj  1.36, and
TTR  3.0. Fig. 15 OASPL reduction comparison of heat-simulated jets issuing from
the Gen1A 3DS8 nozzle with NPR  3.0, Mj  1.36, and TTR  3.0.

these results are for cold jets, and the baseline nozzle produced
significant screech tones, whereas the deflected seals nozzles did not. nozzles, the mixing noise reduction increases with the nozzle
Thus, for this figure, the specific reduction levels attributed to the pressure ratio until an apparent optimal condition is reached at
deflected seals are realistically 1–2 dB lower in magnitude. For all NPR  3.5. The Gen1B 3DS10 nozzle showed the largest amount of
Article in Advance / PILON ET AL. 9

Fig. 16 Mixing noise reduction of cold jets as a function of pressure ratio


for all of the Gen1B deflected seals nozzles. Reduction is measured from
the baseline nozzle at the same pressure ratio and averaged over the polar
microphones from 130 to 160 deg.
Downloaded by UNIV. OF ARIZONA on March 14, 2017 | http://arc.aiaa.org | DOI: 10.2514/1.C033977

Fig. 18 Far-field acoustic power spectral densities measured for the


baseline and Gen1B 3DS10 nozzle from cold jets with NPR  3.3,
Mj  1.43, TTR  1, and scaled R∕D  100 (re., reference pressure).

shock-associated noise peak is reduced by ∼8 dB in conjunction


with an increased peak frequency.
For further analysis of these cold jet measurement results, flow
visualizations with a standard z-type shadowgraph system were
performed. Each nozzle, including the baseline, was visualized at two
azimuthal angles and five different nozzle pressure ratios. Only two
pressure ratios are shown for the baseline and Gen1B 3DS10 nozzle to
Fig. 17 Broadband shock-associated noise reduction of cold jets as a complement and support the conclusions derived from the acoustic
function of pressure ratio for all of the Gen1B deflected-seal nozzles.
Reduction is measured from the baseline nozzle at the same pressure ratio
results.
and averaged over the polar microphones from 60 to 100 deg. Figure 19 shows shadowgraph visualization from the baseline
nozzle at two nozzle pressure ratios of 3.3 and 4.58, on the top and
bottom, respectively. These shadowgraphs are time averaged both
mixing noise reduction when measured in the plane midway between optically and digitally to minimize the viewing of turbulence and to
deflected seals. highlight the shock structure of the jet. The nozzle is overexpanded at
Figure 17 shows the noise reduction average of the sideline and the lower pressure ratio, with a large Mach disk and subsequent
forward polar microphones (60–100 deg), giving a representation of strong shock cells downstream. The higher-pressure ratio is chosen
the broadband shock noise reduction. At these polar angles, there is
less dependence on the azimuthal angle than for the downstream
polar angles. This trend was noticed with the Gen1A nozzles as
well. The Gen1B 3DS10 nozzle again performs the best for most
nozzle pressure ratios, with the optimum at NPR  3.3. This is
expected because the nozzle pressure ratio for the effective design
Mach number of this nozzle would be 3.22. The minimum for the
3DS8 nozzle also correlates well with the effective design Mach
number. However, the magnitude of reduction is significantly less
than that for the larger deflection angle nozzle. This could be due to
a number of factors: most likely the oblique shocks and shock cell
structure created by the specific deflection angle.
For the cold jets, the best noise reduction across all angles and
nozzles occurs approximately at a NPR of 3.3 for the Gen1B
3DS10 nozzle. Figure 18 shows acoustic power spectral densities
for five different polar angle microphones measured with this
nozzle at this jet condition for cold jets. Examination of the
narrowband acoustic spectra shows the reasons for the large Fig. 19 Ensemble-averaged shadowgraph flow visualization of the
reductions observed in the OASPL. The peak of the mixing noise is baseline nozzle. The nozzle was operating at NPR  3.3 (top) and
reduced by ∼8 dB. Additionally, the magnitude of the broadband NPR  4.58 (bottom): both with cold jets.
10 Article in Advance / PILON ET AL.

Fig. 20 Ensemble-averaged shadowgraph flow visualization of the


Gen1B 3DS10 nozzle. The nozzle was operating at NPR  3.3 (top) and
NPR  3.7 (bottom), both with cold jets. The nozzle was oriented such
that a deflected seal was directly down in the image.
Downloaded by UNIV. OF ARIZONA on March 14, 2017 | http://arc.aiaa.org | DOI: 10.2514/1.C033977

such that the baseline nozzle is perfectly expanded. Due to the


straight sections of the nozzle, oblique shocks are still present in the
jet plume; however, there are no apparent normal shocks or a strong
repeating shock structure.
Figure 20 shows shadowgraph visualization for cold jets issuing
from the Gen1B 3DS10 nozzle operating at NPR  3.3 and
NPR  3.7. These nozzle pressure ratios were chosen because the
first is the optimum for the observed broadband shock-associated
Fig. 21 Far-field acoustic power spectral densities measured for the
noise reduction, and the higher NPR depicts a case with increased baseline and Gen1B 3DS10 nozzle, without screech removed, from heat-
BBSAN. The flow visualization correlates well with the acoustic simulated jets with NPR  3.3, Mj  1.43, TTR  3.0, and caled
results because the strength of the downstream shock cells is much R∕D  100 (re., reference pressure).
less apparent in the upper image when compared to the lower image.
Additionally, both images show the wake flow generated by the
deflected seals.
Helium–air jet mixtures were tested following the completion of
the cold jet study. Two different nozzle pressure ratios (3.3 and 3.5)
were chosen based upon the results of the cold jet measurements. A
simulated total temperature ratio of 3.0 was chosen to correspond to
nominal takeoff conditions. These jet conditions were again used for
the three Gen1B deflected seals nozzles at two azimuthal angles.
Figure 21 shows acoustic power spectral densities for five different
polar angle microphones measured with the Gen1B 3DS10 nozzle at
a NPR  3.3 and simulated TTR  3.0, whereas Fig. 22 shows the
measured overall sound pressure level reductions achieved. Note that
the large reductions in OASPL are due in part to the deflected seals
nozzle eliminating the screech tones evident in the baseline nozzle.
As noted previously, hot full-scale jets generally do not exhibit the
screech phenomenon. Therefore, it is not realistic to claim noise
reductions due to elimination of screech in model-scale tests. To gain
a more realistic understanding of the noise reduction potential of the
deflected seals concept, screech tones were removed manually from
the baseline nozzle data shown in Figs. 21 and 22 to produce Figs. 23
and 24.
Figures 23 and 24 show acoustic results for heat-simulated jets
for the Gen1B 3DS10 nozzle. Figure 23 shows acoustic power
spectral densities for five different polar angles from both the
baseline nozzle and the deflected seal nozzle at two different
azimuthal angles. Figure 24 shows the noise reduction in OASPL
for both azimuthal angles for a range of polar angles. Despite
removal of the screech tones in the baseline nozzle data,
significant noise reduction is evident for this nozzle with
heat-simulated jets. Reductions of ∼4 dB OASPL are noticeable in
the downstream direction. Also evident is a larger reduction in Fig. 22 OASPL reduction comparison of heat-simulated jets, without
BBSAN. An approximately 2 dB noise reduction is evident in the screech removed, issuing from the Gen1B 3DS10 with NPR  3.3,
forward arc, with a significant reduction in the BBSAN peak level. Mj  1.43, and TTR  3.0.
Figures 25 and 26 show acoustic power spectral densities and
OASPL reductions, respectively, for jets issuing from the from the six reductions are observed when compared to the Gen1B 3DS10 nozzle.
deflected seals Gen1B 6DS5 nozzle. The jet conditions presented The variation with azimuthal angle is again decreased, as was
here are again NPR  3.3 and simulated TTR  3.0. Lower overall observed with the Gen1A 6DS4 nozzle.
Article in Advance / PILON ET AL. 11
Downloaded by UNIV. OF ARIZONA on March 14, 2017 | http://arc.aiaa.org | DOI: 10.2514/1.C033977

Fig. 23 Far-field acoustic power spectral densities measured for the Fig. 25 Far-field acoustic power spectral densities measured for the
baseline and Gen1B 3DS10 nozzle, from heat-simulated jets with baseline and Gen1B 6DS5 nozzle, from heat-simulated jets with
NPR  3.3, Mj  1.43, TTR  3.0, and scaled R∕D  100 (re., reference NPR  3.3, Mj  1.43, TTR  3.0, and scaled R∕D  100 (re.,
pressure). reference pressure).

Fig. 24 OASPL reduction comparison of heat-simulated jets, issuing Fig. 26 OASPL reduction comparison of heat-simulated jets, issuing
from the Gen1B 3DS10 with NPR  3.3, Mj  1.43, and TTR  3.0. from the Gen1B 6DS5 with NPR  3.3, Mj  1.43, and TTR  3.0.

F. Gen1B in the Presence of Forward Flight of a forward-flight stream. In this case, the noise produced by the
The Penn State high-speed jet noise facility has the ability to supersonic jet is significantly affected by the existence of the
perform acoustic measurements of supersonic jets in the presence forward-flight stream. The effect on the jet noise and how to correct
12 Article in Advance / PILON ET AL.

the data for the existence of the outer shear layer were described in
detail by Viswanathan and Czech [19]. The first measurements in the
Penn State facility with forward flight, along with the specific
equations and processing methodology used for the shear layer
correction, were described by Veltin et al. [20]. The facility has since
been used to quantify the noise reduction potential of several nozzle
modifications in the presence of forward flight: specifically with
beveled nozzles in the work of Powers et al. [21] and with hard-walled
corrugation nozzles in the work of Powers and McLaughlin [22]. The
results and trends from these studies compare well to studies
performed on similar noise reduction techniques in the presence
of a forward-flight stream by Viswanathan and Czech [23] and
Viswanathan et al. [24].
When data recorded in the Penn State facility were processed and
corrected for the outer shear layer, the jet noise was assumed to be
produced at the nozzle exit plane. This estimation is not ideal but
provides a practical method to obtain an acceptable accuracy for the
correction. The data shown for forward-flight measurements have
been corrected for the shear layer refraction caused by the forward-
flight stream to produce data that would be measured in a wind
Downloaded by UNIV. OF ARIZONA on March 14, 2017 | http://arc.aiaa.org | DOI: 10.2514/1.C033977

tunnel of larger size than the distance from the jet to the observer
microphones, such as if an observer was moving with the nozzle.
For the forward-flight measurements, one additional microphone
boom angle was used to allow for raw data ranging from θ  55 to
160 deg in increments of 5 deg. This allows for a more accurate
interpolation of data during the correction for shear layer refraction.
Figure 27 shows power spectral densities measured with
microphones at five polar angles and two azimuthal angles for the
baseline and 3DS10 deflected seals nozzles operating at NPR  3.3
Fig. 27 Forward flight far-field acoustic power spectral densities
and TTR  3.0. Both the baseline and deflected seals jets were measured from heat-simulated jets from the baseline and Gen1B 3DS10
operating in the presence of a forward-flight stream of Mf  0.17, nozzles with Mf  0.17, NPR  3.3, Mj  1.43, TTR  3.0, and
and the data have been corrected for the outer shear layer refraction as scaled R∕D  100 (re., reference pressure).
previously described. Figure 28 shows the OASPL reductions
achieved with the 3DS10 deflected seal nozzle and the baseline
nozzle for several polar angles at the same jet and forward-flight
conditions. The presence of the forward-flight stream decreased the
noise reduction benefit of the deflected seals concept. However,
significant noise reductions are still apparent for both the mixing
noise (∼3 dB) and the BBSAN (∼2 dB). This demonstrates the
validity of the concept for both static and forward-flight conditions.
Note that the minor amount of screech in the baseline case was
removed manually so that the observed noise reductions are as
realistic as possible, i.e., no noise reduction is claimed for elimination
of screech. Note also that, at this point, no optimization of the
deflected seals concept has been performed for the static or forward
flight cases.

V. Future Work
This paper has outlined the results of a preliminary investigation of
the deflected seals concept. Logical follow-on efforts include
optimization of the number of deflected seals and seal deflection angle,
as well as computational aeroacoustics efforts to gain greater
understanding of the noise reduction mechanisms possible with this
concept. Fairly simple extensions of the present experimental study
could include fabricating new deflected seals nozzles that would allow
flow under the deflected seal surfaces. These more realistic nozzles
would likely change the optimization for noise reduction to occur at
different deflection angles. Additionally, asymmetric positioning of
the deflected seals could “steer” the high-amplitude noise to directions
away from ground crew or airport communities. In a similar manner,
limited thrust vectoring could be possible through moderated
application of deflected seals. In numerical studies, the potential
damage to exposed deflected seals due to the hot exhaust flow would
be analyzed. Efforts to design and integrate this concept into a full-
scale nozzle for a high-performance aircraft are warranted. Such an
effort will include planning and executing experiments with medium- Fig. 28 Forward-flight OASPL reduction comparison for heat-
scale (∼5–6 in: diameter) nozzles and a similar matrix of parameters to simulated jets from the baseline and Gen1B 3DS10 nozzles with
provide a less expensive path to full-scale deployment. Mf  0.17, NPR  3.3, Mj  1.43, and TTR  3.0.
Article in Advance / PILON ET AL. 13

VI. Conclusions Aeroacoustics, Vol. 11, Nos. 7–8, Dec. 2012, pp. 937–956.
doi:10.1260/1475-472X.11.7-8.937
A simple supersonic jet noise reduction concept has been [9] Morris, P. J., McLaughlin, D. K., and Kuo, C.-W., “Noise Reduction in
described. The concept could be employed on the nozzles of Supersonic Jets by Nozzle Fluidic Inserts,” Journal of Sound and
contemporary fighter aircraft with minimal additional hardware and Vibration, Vol. 332, No. 17, Aug. 2013, pp. 3992–4003.
associated weight penalties, as the concept largely employs hardware doi:10.1016/j.jsv.2012.11.023
already existing on the nozzle. Very little aircraft performance [10] Powers, R. W., Kuo, C.-W., and McLaughlin, D. K., “Experimental
penalties are envisioned because the concept is only employed at Comparison of Supersonic Jets Exhausting from Military Style Nozzles
takeoff and allows the nozzle to operate normally throughout the rest with Interior Corrugations and Fluidic Inserts,” AIAA Paper 2013-
of the flight envelope. 2186, May 2013.
[11] Pilon, A. R., “Noise Reduction Technique for Converging-Diverging
Although simple in nature, laboratory experiments have Exhaust Nozzles,” U.S. Patent 9261046, Feb. 2016.
demonstrated the concept’s ability to reduce significantly both [12] “CFD++,” Metacomp Technologies, Inc., 2013.
broadband shock-associated noise and turbulent mixing noise. [13] Kinzie, K. W., and McLaughlin, D. K., “Measurements of Supersonic
Employment of the concept on current and future tactical aircraft Helium/Air Mixture Jets,” AIAA Journal, Vol. 37, No. 11, Nov. 1999,
could lead to major reductions in noise levels at all polar angles. pp. 1363–1369.
Reduction of these noise levels would lead to significant cost savings doi:10.2514/2.634
for operators of high-performance aircraft. [14] Doty, M. J., and McLaughlin, D. K., “Acoustic and Mean Flow
Measurements of High-Speed, Helium-Air Mixture Jets,” International
Journal of Aeroacoustics, Vol. 2, No. 3, July 2003, pp. 293–333.
Acknowledgments doi:10.1260/147547203322986151
[15] Papamoschou, D., “Acoustic Simulation of Coaxial Hot Air Jets Using
The experiments presented in this paper were funded by Lockheed
Downloaded by UNIV. OF ARIZONA on March 14, 2017 | http://arc.aiaa.org | DOI: 10.2514/1.C033977

Cold Helium-Air Mixture Jets,” Journal of Propulsion and Power,


Martin Independent Research and Development efforts. The second Vol. 23, No. 2, March 2007, pp. 375–381.
author (Russell W. Powers) would like to thank the U.S. Department doi:10.2514/1.21776
of Defense Science, Mathematics, and Research for Transformation [16] McLaughlin, D. K., Bridges, J., and Kuo, C.-W., “On the Scaling of
program, with mentor Allan Aubert, for its sponsorship. The authors Small, Heat Simulated Jet Noise Measurements to Moderate Size
would like to thank Alex Karns for his assistance in the laboratory Exhaust Jets,” International Journal of Aeroacoustics, Vol. 9, No. 4,
June 2010, pp. 627–654.
experiments. [17] Kuo, C.-W., Veltin, J., and McLaughlin, D. K., “Methods to Imporve the
Accuracy of Acoustic Measurements in Small Scale High Speed Jets,”
AIAA Paper 2009-3251, 2009.
References [18] Kuo, C.-W., Veltin, J., and McLaughlin, D. K., “Acoustic Measurements
[1] Doychak, J., “Department of Navy Jet Noise Reduction Project of Models of Military Style Supersonic Nozzle Jets,” AIAA Paper 2009-
Overview,” Technical Session 3B, ‘Aviation and the Environment: 0018, 2009.
Deicing and Noise’, Strategic Environmental Research and Develop- [19] Viswanathan, K., and Czech, M. J., “Measurement and Modeling of
ment Program (SERDP) and Environmental Security Technology Effect of Forward Flight on Jet Noise,” AIAA Journal, Vol. 49, No. 1,
Certification Program (ESTCP), Dec. 2010. Jan. 2011, pp. 216–234.
[2] Greska, B., Krothapalli, A., Seiner, J. M., Jansen, B. J., and Ukeiley, L. S., doi:10.2514/1.J050719
“The Effects of Microjet Injection on an F404 Jet Engine,” AIAA Paper [20] Veltin, J., Day, B. J., and McLaughlin, D. K., “Forward Flight Effect on
2005-3047, 2005. Small Scale Supersonic Jet Acoustics,” AIAA Paper 2010-3924,
[3] Anderson, B. A., Wygnanski, I. J., and Gutmark, E. J., “Noise Reduction June 2010.
by Interaction of Flexible Filaments with an Underexpanded Supersonic [21] Powers, R. W., Senft, M., and McLaughlin, D. K., “Acoustics
Jet,” AIAA Paper 1999-0080, 1999. Measurements of Scale Models of Military Style Supersonic Beveled
[4] Martens, S., and Spyropoulos, J. T., “Practical Jet Noise Reduction for Nozzle Jets,” AIAA Paper 2011-2702, June 2011.
Tactical Aircraft,” Proceedings of the ASME Turbo Expo 2010: Power [22] Powers, R. W., and McLaughlin, D. K., “Acoustics Measurements of
for Land, Sea and Air, American Society of Mechanical Engineers Scale Models of Military Style Supersonic Beveled Nozzle Jets with
Paper, GT2010-23699, Fairfield, NJ, June 2010. Interior Corrugations,” AIAA Paper 2012-2116, June 2012.
[5] Henderson, B., and Bridges, J., “An MDOE Investigation of Chevrons [23] Viswanathan, K., and Czech, M. J., “Adaptation of the Beveled Nozzle
for Supersonic Jet Noise Reduction,” AIAA Paper 2010-3926, 2010. for High Speed Jet Noise Reduction,” AIAA Journal, Vol. 49, No. 5,
[6] Seiner, J. M., Ukeiley, L. S., and Jansen, B. J., “Aero-Performance May 2011, pp. 932–944.
Efficient Noise Reduction for the F404-400 Engine,” AIAA Paper doi:10.2514/1.J050409
2005-3048, 2005. [24] Viswanathan, K., Krothapalli, A., Seiner, J. M., Czech, M. J., Greska, B.,
[7] Seiner, J. M., “Method and Device for Reducing Engine Noise,” U.S. and Jansen, B. J., “Assessment of Low-Noise Nozzle Designs for Fighter
Patent 7475550 B2, Jan. 2009. Aircraft Applications,” Journal of Aircraft, Vol. 48, No. 2, March–
[8] Murray, N. E., and Jansen, B. J., “Performance Efficient Jet Noise April 2011, pp. 412–423.
Reduction for Supersonic Nozzles,” International Journal of doi:10.2514/1.C000285

Das könnte Ihnen auch gefallen