aeroacoustics volume 3 á number 1 á 2004 Ð pages 43 Ð 66 43

On the near field pressure of a transonic

axisymmetric jet
Lawrence S. Ukeiley
Jamie L. Whitten National Center for Physical Acoustics,
University of Mississippi, University, MS 38677
E-mail: lukeiley@olemiss.edu

Michael K. Ponton
School of Education, Regent University, Virginia Beach, VA 23464-9800

ABSTRACT
The near field fluctuating pressure of a cold subsonic jet (Mach 0.85) issuing from a contoured
convergent nozzle was studied. Conventional time series analysis (i.e., correlation and spectral
analysis) as well as a wavelet analysis were used to characterize time-dependent events thereby
revealing features consistent with current understandings of jet noise while also uncovering modal
features consistent with previous measurements of the jet flow field itself. Specifically, the
measurements reveal extensive sources of multiple frequencies in the streamwise direction. The near
field information of the present investigation should help to bridge our understanding of the
relationship between the velocity field and the far field acoustic emission associated with subsonic
jet noise as well as supply a data set that can be used for verification of computational efforts.

1. INTRODUCTION
The noise due to engine exhaust of current subsonic transports continues to be a major
environmental concern to airport communities and hence manufacturers. The source of
this noise is the turbulent mixing of the exhausted jet plume with the ambient
environment and is particularly dominant in the acoustic far field during takeoff. This
mixing creates pressure fluctuations in the hydrodynamic near field that propagate to
the acoustic far field regions of jet flow. One of the keys to understanding and hence
being able to reduce noise sources is to link flow field events with noise production.
While the majority of literature on jet noise involves studying either the far field
acoustics or the flow field properties of the jet plume, studying the acoustic near field
may lead to a better understanding of hydrodynamically induced pressure fluctuations
that ultimately propagate to the far field through a complicated process.
Since the early 1950s, there have been analytical, numerical, and experimental studies
that attempt to link the far field acoustics with the hydrodynamic field. This began with
the seminal work of Lighthill (1952) where the now famous acoustic analogy relates
properties of the turbulent field with the far field acoustics. Details of much of the work
44 On the near field pressure of a transonic axisymmetric jet

on jet noise theories, based on Lighthill’s work as well as other theories, can be found
in the many comprehensive and review articles such as Goldstein (1984), Lilley (1991)
and Tam (1998), naming just a few. The far field acoustics of transonic jets have been
studied extensively with some of the earlier studies (Ahuja and Bushell, 1974; Lush, 1971)
concentrating on verifying the aforementioned theoretical efforts and demonstrating
some success. Although there has not been a tremendous number of studies conducted
to measure the velocity field in transonic jets, those such as Morris (1976), Stromberg
et al. (1980), Ukeiley et al. (1998, 1999), Seiner et al. (1999, 2001), and Narayanan et al.
(2002) are among some of the ones that have acquired detailed measurements of the
velocity field in transonic jets. These studies along with others have led to a general
understanding of the main features of the statistical behavior of this class of jets. Works
such as those by Seiner and Reethof (1974), Schaffar (1979) and Juve et al. (1980) were
among the first to use causality principles by directly measuring the correlation between
the velocity measurements and the far field acoustics. They showed the dominance of
“shear noise” over “self noise” as the primary source of radiation to the peak jet noise
radiation angle. Additionally, these studies showed that the dominant noise sources
occur near the end of the potential core. Recent studies that use the advanced
experimental techniques discussed in Thurow et al. (2002) have continued to link
events in the flow field, through correlation and direct observation techniques, in an
attempt to isolate the sound producing events.
Even though the bulk of studies have either involved the measurement of the far field
acoustics or the velocity field, a handful of studies have concentrated on measuring the
fluctuating pressure inside the jet or in the near field. Measurements of the fluctuating
pressure inside the jet have been reported in studies such as Fuchs and co-workers
(1972, 1975), Jones et al. (1979), and George et al. (1984). From the studies by Jones
et al. and George et al., the spectral distribution of the fluctuating pressure was examined
experimentally and shown to agree with theoretical predictions while differing in the
decay rate of the fluctuating velocity. The studies of Fuchs have concentrated on relating
the in-flow pressure and velocity measurements. One of the more interesting results in
the difference between the velocity and pressure fields was reported in Michalke and
Fuchs (1975) where the pressure field was shown to be dominated by the lower azimuthal
mode numbers while the velocity field is dominated by higher azimuthal modes at three
diameters downstream. It should be noted that in more recent experiments, Ukeiley
et al. (1999) in a transonic jet and Gamard et al. (2002) in an incompressible jet have
shown how the velocity field evolves in the downstream direction to become dominated
by the lower azimuthal modes.
Although the near field acoustics cannot be directly linked to the far field recent
work of Picard and Delville (1999), Arndt, Long, and Glauser (1997) clearly show the
link between near field pressure measurements and the hydrodynamic velocity field. In
Picard and Delville, the near field pressure was used as the condition in the linear
stochastic estimation procedure to estimate the time dependent velocity field of an
incompressible jet. They clearly showed that the near field pressure provides an
estimate of the temporal evolution of the spatial velocity field for a streamwise plane.
Furthermore, they showed that the velocity estimated from the near field pressure is
aeroacoustics volume 3 á number 1 á 2004 45

associated with the shear noise component of the far field noise. However in that study,
they did not examine the azimuthal discrepancies between the velocity and pressure
fields. In Arndt, Long, and Glauser, the near field of an incompressible jet was examined
using Proper Orthogonal Decomposition. Their results allowed for the analysis of the
streamwise and azimuthal directions where it was shown that the azimuthal direction is
dominated by low mode numbers while the streamwise direction is dominated by a
structure which grows until saturation then decays.
In the present work, detailed microphone array measurements in the near field of a
Mach 0.85 jet will be presented. These measurements were done with both streamwise
and spanwise aligned arrays. What follows will be the experimental details and an
introduction to the continuous wavelet transform methodology used to examine the data.
Results will then be presented on the azimuthal and streamwise structure of the near jet
pressure region. This will include comparisons to some of the known features of the jet
flow field and far field acoustics. These measurements and their analysis not only
provide insight into the relationship between the hydrodynamic and acoustic fields of a
high Reynolds number transonic jet but they also provide key information that can be
used to help validate and develop numerical simulations and low-dimensional models.

2. EXPERIMENTAL DETAILS AND TECHNIQUES

The experiments were conducted in the Small Anechoic Jet Facility (SAJF) located at
the NASA Langley Research Center. Model hardware and instrumentation are available
within the SAJF for research directed at measuring the fluid dynamic and acoustic
characteristics associated with internal and external airflows. The interior walls of the
SAJF are anechoically treated with acoustic wedges that absorb in excess of 99% of the
incident sound for frequencies above 100 Hz. The internal dimensions of the SAJF
(within the wedge tips) are 3.2 meter by 3 meter by 3.7 meter along the streamwise (or x)
direction. Research models are connected to an air delivery system capable of supplying
1.13 kg/s of continuous dry air. A 275 kW resistance heater is used to create a constant
cold stagnation temperature of 40°C thereby establishing an operating temperature
independent of variations in supply temperatures. An electronically controlled valve
maintains the nozzle pressure ratio to within 0.5% of the desired set point and pressure
transducers used by the flow control system receive frequent in-situ calibration.
The subsonic nozzle consists of a contoured transition section connecting a
152.4-mm settling chamber to a 50.8-mm inside exit diameter convergent nozzle.
Metallic honeycomb is installed in the settling chamber to diminish large scale
propagating disturbances. The nozzle is designed for parallel exit flow. A 0.61-meter
diameter circular duct treated with fine mesh screen is positioned concentric to the
nozzle assembly thus facilitating co-flow aspiration.
The linear and semi-circular arrays (Figure 1) incorporated seven 6.35-mm phase-
matched microphone systems. The separation distance between each microphone on the
linear array was 1 D (i.e., 50.8 mm) and the array was positioned approximately parallel
to the edge of the jet shear layer, i.e., 6° from the jet centerline. On the semicircular
array, a constant azimuthal separation of 30° was used with each sensor positioned at
r/D = 2. The plane of this array was perpendicular to the jet centerline. For data
46 On the near field pressure of a transonic axisymmetric jet

B
Figure 1. Photographs of linear and azimuthal microphone array.
aeroacoustics volume 3 á number 1 á 2004 47

acquisition, both arrays were traversed in the x-direction while only the linear array was
traversed in the r-direction. Data were acquired using the semicircular array at x/D = 0,
1, 2, 3, 4, 5, 6, 7 and 8. For the linear array, data were acquired to map out the region
between 3 and 18 diameters in the radial direction and from the nozzle exit to 19 jet exit
diameters in the streamwise direction. For the linear array data, each microphone signal
was digitized at 50 kHz and bandpass filtered between 160 Hz and 20 kHz. For the
semicircular array data, the signals were digitized at 30 kHz and bandpass filtered
between 160 Hz and 12.5 kHz. For all microphone data, 128 independent blocks of
1024 usable data points were recorded.

Continuous wavelet transforms

The use of wavelet transforms for the analysis of digital time series data relevant to
fluid mechanic applications has been discussed in many places such as Jordan et al.
(1997) and Lewalle (1994). The use of wavelet transforms techniques offers an
advantage over traditional time series analysis techniques since they retain temporal
information in concert with the scale of the function that can be related to different
properties depending upon the mother function. In the work reported here, wavelet
analysis will be used to examine the existence of variations in frequency along the
streamwise extent of the jet. Wavelet analysis has been used in a similar fashion for
pressure data on the floor of a resonating cavity (Cattafesta et al., 1998) and for
microphone measurements around a screeching jet (Walker et al., 1997) thereby
showing the usefulness of the time localization properties of the wavelet transform. In
the first study, the authors were able to study the switching between Rossiter modes in
the cavity; in the second study, the authors were able to verify the simultaneous
existence of multiple screech modes.
Since discussions of how to apply continuous wavelet transforms can be found in
many sources, only a brief discussion is included here. The coefficients of the continuous
wavelet transforms are defined as the inner product of the time signal with the basis
function and can be written as
¥
W (t , s) = ò p (t ) × Y *
t ,s ( t )dt (1)
-¥

where p(t) is the time history from the microphones, Yt,s(t) is the wavelet function, s
represents the scale of the wavelet function, t is a time lag, and the asterisk denotes the
complex conjugate.
The choice of the wavelet mother function determines the features in the time series
that are highlighted by the analysis. As discussed in Cattafesta et al. (1998), the use of
a Morlet wavelet can be viewed as a bandpass filter in the frequency domain. This
feature is important for the analysis that is presented below since the attempt is to
localize certain frequency events in the near field of the jet.
The Morlet wavelet used is defined by the following mother function,
2
Y( t ) = e iw Y t e it /2
(2)
48 On the near field pressure of a transonic axisymmetric jet

where w Y is a constant that forces admissibility (see Jordan et al., 1997). In the application
of the wavelet transform, the above mother function is applied by shifting and dilatations
that involve the aforementioned scale and time lag as follows,
t -tö
Yt , s (t ) = s -0.5 Y æ . (3)
è s ø
The normalization factor s-0.5 ensures that all of the dilated versions have the same energy.
Further details of the constraints on a function to be a wavelet and the algorithm used to
solve for the wavelet coefficients can be found in Jordan et al.. The specific algorithm
used for this work was verified against several artificially generated time signals.

3. RESULTS
In the following sections, aspects of the near field pressure will be presented that can
be used to study the structure of the jet in both the azimuthal and streamwise evolution
directions. This will involve examining several aspects of the pressure signals such as
time histories, spectral distributions, and wavelet transforms and comparing this near
field pressure to the velocity field and far field noise.

Jet characteristics
Figure 2 shows the evolution of the mean streamwise velocity profile of the jet. The
velocity field was computed using total pressure probe measurements and isentropic
relationships by assuming the jet to be pressure balanced with the ambient. From this
figure one can clearly see the growth of the jet where the end of the potential core was
found to be approximately 6.2 diameters downstream of the nozzle exit, in good
agreement with historical data (v. Morris, 1976). The momentum thickness of the jet
was calculated and shown to grow in a linear fashion over the region examined (see
Ponton, Ukeiley, and Lee, 1999). Figure 3(a) shows the Overall Sound Pressure Levels
(OASPL) from the microphone measurements in the near field. From this figure, the
overall radiation pattern is evident and suggests the presence of distributed acoustic
sources through the streamwise extent of the jet. Figure 3(b)-(e) show the same data
bandpass filtered around Strouhal numbers based on the jet exit parameters (StD = fD/U)
of 0.2, 0.4, 0.8, and 1.2 with an integration range of 488.3 Hz. The primary frequency
at StD = 0.20 was chosen as the peak frequency of the emitted jet noise at x/D = 19,
r/D = 8, which is close to a polar angle of 30° with respect to the jet centerline. From
these plots, one can see the general trend of higher frequencies being emitted closer to
the jet exit. The dominant jet noise frequency appears to be emitted in the region from
7 through 12 jet diameters downstream of the jet exit plane. These results are consistent
with the commonly held understanding of where different frequencies observed in the
far field noise spectrum are predominantly generated based upon dominant sound
radiation directions for a jet operated at these exit conditions.

Azimuthal dependence
The azimuthal dependence of the near field pressure was examined at r/D = 2 and at
increments of one nozzle exit diameter from the nozzle exit through 8 diameters
downstream of the nozzle exit plane.
aeroacoustics volume 3 á number 1 á 2004 49

0.9
x/D = 10
x/D = 9
0.8
x/D = 8
x/D = 7
0.7
x/D = 6
x/D = 4
0.6 x/D = 1
U/Ue

0.5

0.4

0.3

0.2

0.1

0
-2 -1 0 1 2
r/D
Figure 2. Streamwise evolution of mean velocity profiles with mach 0.85 exit
condition.

Figure 4(a) presents the azimuthal correlations of the microphones for all of the
streamwise positions. There is a clear evolution as a function of the downstream position.
In the exit plane of the nozzle and at 1 diameter downstream the correlation levels decay
quite rapidly. As the measurement region progresses downstream, the azimuthal extent of
the correlations increase, i.e., a broadening of the curves until approximately 6 diameters
downstream. At that point, approximately the end of the potential core, the azimuthal
extent of the correlations tends to reach a maximum. Further downstream the azimuthal
extent of these correlations begins to decay again although the curves are broader at
small azimuthal separations as compared to those before the end of the potential core.
This shows that the dominant high frequency near the jet exit has little azimuthal
correlation. The velocity field has also been shown to have a change in behavior in the
azimuthal extent around the end of the potential core by Ukeiley et al. (1999) in this jet
and Gamard et al. (2002) in an incompressible jet. In both of these studies, the flow field
seemed to asymptote from being dominated by higher azimuthal mode numbers in the
initial region of the jet to being dominated by a lower azimuthal mode number as one
traverses through the end of the potential core. The behavior observed in this pressure
field is similar to what has been observed in the velocity field detailed in the previous
50 On the near field pressure of a transonic axisymmetric jet

(A) Overall (B) Filtered at StD = 0.20

(C) Filtered at StD = 0.40 (D) Filtered at StD = 0.80

(E) Filtered at StD = 1.20

Figure 3. Radial and streamwise distribution overall sound pressure levels.
aeroacoustics volume 3 á number 1 á 2004 51

1 x/D = 0.0
x/D = 1.0
0.9 x/D = 2.0
x/D = 3.0
x/D = 4.0
0.8
x/D = 5.0
x/D = 6.0
0.7
Correlation Coefficient

x/D = 7.0
x/D = 8.0
0.6

0.5

0.4

0.3

0.2

0.1

0
0 50 100 150
Azimuthal Separation (deg.)
(A) Raw Data

1 x/D = 0.0
x/D = 1.0
x/D = 2.0
0.9
x/D = 3.0
x/D = 4.0
0.8
x/D = 5.0
x/D = 6.0
0.7
Correlation Coefficient

x/D = 7.0
x/D = 8.0
0.6

0.5

0.4

0.3

0.2

0.1

0
0 50 100 150
Azimuthal Separation (deg.)
(B) Filtered at StD = 0.20
Figure 4. Azimuthal pressure correlations at r/D = 2.
52 On the near field pressure of a transonic axisymmetric jet

studies and is consistent with the low azimuthal mode number in flow pressure
correlations reported in Michalke and Fuchs (1975). This implies that the azimuthal
structure of the dominant sound producing events are not associated with the small scale
azimuthal features of the velocity field and perhaps more closely linked with lower
azimuthal mode number events. Lower azimuthal mode number events which may be
the noise producing events have been previously reported by Taylor et al. (2001) in a
Mach 0.6 jet and Citriniti and George (2000) in an incompressible jet. This is also
consistent of with the results of Freund and Colonius (2002) where they linked the far
field noise to the nearly axisymmetric low azimuthal mode number events. Figure 4(b)
shows the azimuthal correlation plots for the data in the band centered at the primary
jet noise frequency. For this case, there is no marked difference in the behavior as a
function of streamwise position. It also appears that the shape of the correlation
function is quite similar to that observed in the region between 5 and 7 diameters
downstream of the exit plane. This implies that the azimuthal behavior observed in the
dominant jet noise frequency range exists throughout the jet while not always being the
dominant behavior. This last point is important because it suggests that the flow may
not be reorganizing axially; rather the high frequency, high azimuthal mode number
behavior becomes less dominant in this direction.
Figure 5 presents the fluctuating pressure measured by each of the microphones on
the azimuthal array (ordinate) as a function of time (abscissa) with the amplitude being
pressure in Pa. There appears to be distinctly different behaviors near the jet exit as
compared to that farther downstream. As one might expect, the behavior near the exit
is dominated by fluctuations of short temporal duration as well as stretching only over
a couple of microphones, which would be characteristic of the small scale events that
are usually attributed with being dominant at the jet exit plane. As the microphone array
progresses farther downstream, the temporal duration of the fluctuations begins to increase.
The increase in the temporal extent is also followed by an increase in azimuthal extent
of the coherent fluctuations that are consistent with the correlation plots shown earlier.
Additionally, the frequency of the fluctuations decrease, which is also evident in the
spectra presented in Figure 6. Here one can clearly see that in the exit plane, the energy
content in the higher frequencies is on the same order as that for the lower frequencies.
However, by 6 diameters downstream there is a dominance of the lower frequencies.
These plots show evidence that the high frequency (small scale) noise is emitted
throughout the jet but is overshadowed by the lower frequency (large scale) noise
farther downstream. It is also apparent from these plots that although the high frequency
noise is usually attributed to events near the nozzle lip, the sound pressure levels in the
high frequency range are greater farther downstream, even beyond the end of the
potential core. This implies that to correctly model the high frequency noise, the region
beyond the end of the potential core must also be considered.
Figure 7-10 show the magnitude of the wavelet transforms for azimuthal angles
around half the jet from the exit plane to 8 diameters downstream. In all plots, the
abscissa is time and the ordinate is a 1/time scale, which can be related to frequency as
previously discussed. In all contour plots, the gray scale goes from white being small
amplitudes to black being large amplitudes. It also should be noted that at each of the
streamwise locations, the amplitude of the contour levels for all of the angles are
 aeroacoustics volume 3 á number 1 á 2004 53

180 180 180

160 41.0 160 44.0 160 67.0
Azimuthal angle

Azimuthal angle

Azimuthal angle
34.3 35.8 55.2
140 27.6 140 27.6 140 43.4
20.9 19.4 31.6
120 14.2
120 120
11.2 19.8
100 7.5 100 3.0 100 8.0
0.8 -5.2 -3.8
80 -5.9 80 -13.4 80 -15.6
-12.6 -21.6 -27.4
60 -19.3 60 -29.8 60 -39.2
-26.0 -38.0 -51.0
40 40 40
20 20 20
0 0 0
0 0.001 0.002 0.003 0 0.001 0.002 0.003 0 0.001 0.002 0.003

Time (sec) Time (sec) Time (sec)

(A) 0 D (B) 1 D (C) 2 D
180 180 180
160 100.0 160 147.0 160 178.0
Azimuthal angle

Azimuthal angle

Azimuthal angle
80.4 118.0 145.3
140 60.8 140 89.0
140 112.6
41.2 60.0 79.9
120 120 120
21.6 31.0 47.2
100 2.0 100 2.0 100 14.5
-17.6 -27.0 -18.2
80 -37.2 80 -56.0 80 -50.9
-56.8 -85.0 -83.6
60 -76.4 60 -114.0 60 -116.3
-96.0 -143.0 -149.0
40 40 40
20 20 20
0 0 0
0 0.001 0.002 0.003 0 0.001 0.002 0.003 0 0.001 0.002 0.003

Time (sec) Time (sec) Time (sec)

(D) 3 D (E) 4 D (F) 5 D
180 180 180
160 248.0 160 304.0 160 422.0
Azimuthal angle

Azimuthal angle

Azimuthal angle

201.7 247.0 352.4

140 155.4 140 191.0
140 282.8
120 109.1 120 134.5 120 213.2
62.8 78.0 143.6
100 16.5 100 21.5 100 74.0
-29.8 -35.0 4.4
80 -76.1 80 -91.5 80 -65.2
-122.4 -148.0 -134.8
60 -168.7 60 -204.5 60 -204.4
-215.0 -261.0 -274.0
40 40 40
20 20 20
0 0 0
0 0.001 0.002 0.003 0 0.001 0.002 0.003 0 0.001 0.002 0.003

Time (sec) Time (sec) Time (sec)

(G) 6 D (H) 7 D (I) 8 D
Figure 5. Time dependence of azimuthal pressure distribution at r/D = 2.

fixed; however, the amplitude ranges are different for each streamwise location. This
was necessary because the acoustic levels increase with increasing downstream position
over the range being studied. Behaviors observed via the wavelet transforms are
consistent with other analyses while also providing additional insights.
In the exit plane, one observes the preponderance of short duration high frequency
events beginning at approximately 5 kHz. Also evident is that these high frequency
events typically only span over 30-60 degrees. Additionally, in these plots there is
evidence of lower amplitude events at frequencies associated with the large scale
events, at approximately 1 kHz. These events also tend to extend only over 30 or so
degrees. At 2 diameters downstream of the jet exit plane, the wavelet plots are
dominated by events around 5 kHz although other frequency events also exist. The
events at 5 kHz tend to have the same azimuthal extent of 60 degrees. The relative
amplitudes of the lower frequency events have increased to almost the amplitude of the
higher frequency events. At 5 diameters downstream of the jet exit plane, the wavelet
plots are dominated by events from 1-3 kHz. The higher frequency events do exist
54 On the near field pressure of a transonic axisymmetric jet

120

110
SPL, dB re 20 micro Pa

100

90

80

x/D = 0.0
70 x/D = 2.0
x/D = 6.0

60
102 103 104
Frequency, Hz
Figure 6. Downstream evolution of pressure spectra at r/D = 3.0.

although their relative amplitudes are significantly smaller. Some of the events in the
1-2 kHz range have azimuthal extents on the order of 90 degrees, in some cases.
Somewhat apparent are possible helical structures associated with large amplitude, low
frequency events. This is evident in high amplitude events that continuously change
their time of presence as a function of angle for a given frequency. At 8 diameters
downstream of the jet exit plane, the wavelet plots show all of the dominant events
being less than 2 kHz. The behavior is quite similar to that at 5 diameters downstream
with all of the information above 3 kHz not being visible. At both 5 and 8 diameters
downstream, there is some evidence of a helical mode existing by noting the behavior
of a large amplitude event (between 1.5 and 2 kHz) shifting in time. This frequency
range corresponds to the peaks in the mass-flux spectra reported in Ukeiley and Seiner
(1998). The existence of a helical mode in the near field pressure is consistent with
results from Freund and Colonius (2002) where the dominant POD mode, when applied
to the pressure field, was found to be the first helical mode.
aeroacoustics volume 3 á number 1 á 2004 55

10000 10000
9000 9000
8000 8000
1/Scale, Hz

1/Scale, Hz
7000 7000
6000 6000
5000 5000
4000 4000
3000 3000
2000 2000
1000 1000

0.0025 0.005 0.0075 0.01 0.0025 0.005 0.0075 0.01

Time, sec Time, sec
(A) 0 deg. (B) 30 deg.

10000 10000
9000 9000
8000 8000
1/Scale, Hz

1/Scale, Hz

7000 7000
6000 6000
5000 5000
4000 4000
3000 3000
2000 2000
1000 1000

0.0025 0.005 0.0075 0.01 0.0025 0.005 0.0075 0.01

Time, sec Time, sec
(C) 60 deg. (D) 90 deg.

10000 10000
9000 9000
8000 8000
1/Scale, Hz

1/Scale, Hz

7000 7000
6000 6000
5000 5000
4000 4000
3000 3000
2000 2000
1000 1000

0.0025 0.005 0.0075 0.01 0.0025 0.005 0.0075 0.01

Time, sec Time, sec
(E) 120 deg. (F) 150 deg.
Figure 7. Azimuthal dependence of wavelet transforms at x/D = 0.0 and r/D = 2.
56 On the near field pressure of a transonic axisymmetric jet

10000 10000
9000 9000
8000 8000
1/Scale, Hz

1/Scale, Hz
7000 7000
6000 6000
5000 5000
4000 4000
3000 3000
2000 2000
1000 1000

0.0025 0.005 0.0075 0.01 0.0025 0.005 0.0075 0.01

Time, sec Time, sec
(A) 0 deg. (B) 30 deg.

10000 10000
9000 9000
8000 8000
1/Scale, Hz

1/Scale, Hz

7000 7000
6000 6000
5000 5000
4000 4000
3000 3000
2000 2000
1000 1000

0.0025 0.005 0.0075 0.01 0.0025 0.005 0.0075 0.01

Time, sec Time, sec
(C) 60 deg. (D) 90 deg.

10000 10000
9000 9000
8000 8000
1/Scale, Hz

1/Scale, Hz

7000 7000
6000 6000
5000 5000
4000 4000
3000 3000
2000 2000
1000 1000

0.0025 0.005 0.0075 0.01 0.0025 0.005 0.0075 0.01

Time, sec Time, sec
(E) 120 deg. (F) 150 deg.
Figure 8. Azimuthal dependence of wavelet transforms at x/D = 2 and r/D = 2.
aeroacoustics volume 3 á number 1 á 2004 57

10000 10000
9000 9000
8000 8000
1/Scale, Hz

1/Scale, Hz
7000 7000
6000 6000
5000 5000
4000 4000
3000 3000
2000 2000
1000 1000

0.0025 0.005 0.0075 0.01 0.0025 0.005 0.0075 0.01

Time, sec Time, sec
(A) 0 deg. (B) 30 deg.

10000 10000
9000 9000
8000 8000
1/Scale, Hz

1/Scale, Hz

7000 7000
6000 6000
5000 5000
4000 4000
3000 3000
2000 2000
1000 1000

0.0025 0.005 0.0075 0.01 0.0025 0.005 0.0075 0.01

Time, sec Time, sec
(C) 60 deg. (D) 90 deg.

10000 10000
9000 9000
8000 8000
1/Scale, Hz

1/Scale, Hz

7000 7000
6000 6000
5000 5000
4000 4000
3000 3000
2000 2000
1000 1000

0.0025 0.005 0.0075 0.01 0.0025 0.005 0.0075 0.01

Time, sec Time, sec
(E) 120 deg. (F) 150 deg.
Figure 9. Azimuthal dependence of wavelet transform at x/D = 5 and r/D = 2.
58 On the near field pressure of a transonic axisymmetric jet

10000 10000
9000 9000
8000 8000
1/Scale, Hz

1/Scale, Hz
7000 7000
6000 6000
5000 5000
4000 4000
3000 3000
2000 2000
1000 1000

0.0025 0.005 0.0075 0.01 0.0025 0.005 0.0075 0.01

Time, sec Time, sec
(A) 0 deg. (B) 30 deg.

10000 10000
9000 9000
8000 8000
1/Scale, Hz

1/Scale, Hz

7000 7000
6000 6000
5000 5000
4000 4000
3000 3000
2000 2000
1000 1000

0.0025 0.005 0.0075 0.01 0.0025 0.005 0.0075 0.01

Time, sec Time, sec
(C) 60 deg. (D) 90 deg.

10000 10000
9000 9000
8000 8000
1/Scale, Hz

1/Scale, Hz

7000 7000
6000 6000
5000 5000
4000 4000
3000 3000
2000 2000
1000 1000

0.0025 0.005 0.0075 0.01 0.0025 0.005 0.0075 0.01

Time, sec Time, sec
(E) 120 deg. (F) 150 deg.
Figure 10. Azimuthal dependence of wavelet transform at x/D = 8 and r/D = 2.
aeroacoustics volume 3 á number 1 á 2004 59

Streamwise dependence
Figure 11 displays the instantaneous pressure distributions as a function of time and
streamwise position. As with the data displayed for the azimuthal information, the
abscissa for this figure is time in seconds whereby the dark regions represent large
positive pressures and the white regions represent large negative pressures relative to

5
Streamwise Position (x/D)

0
0.001 0.002 0.003 0.004 0.005
Time, sec
(A) Exit Plane to 6 Diameters Downstream (r/D for the upstream microphone = 2.0)
9

8
Streamwise Position (x/D)

3
0.001 0.002 0.003 0.004 0.005
Time, sec
(B) 3 to 9 Diameter Downstream (r/D for the microphone = 1.8)
Figure 11. Time dependence of streamwise pressure distributions.
60 On the near field pressure of a transonic axisymmetric jet

ambient conditions. It should be noted that microphones used to collect these data along
with other data in this section had the microphones approximately parallel to the edge
of the shear layer, i.e., a 6° half angle. In Figure 11(a), the microphone array spanned
from the exit plane to 6 jet exit diameters downstream while in Figure 11(b) it spanned
from 3 to 9 diameters downstream. For Figure 11(a), r/D for the upstream microphone
was 2.0 while for Figure 11(b) it was 1.8. From these figures, the change in the radiated
sound field at approximately 3 diameters downstream is quite apparent. The change in
behavior of the streamwise structure of the near field pressure in this region is consistent
with the previously mentioned velocity field measurements where it was shown that the
azimuthal structure changes throughout this region. In the region between the exit plane
and 3 diameters downstream, the amplitude of the near field pressure is significantly
less than that observed farther downstream. Additionally, it appears to be radiating
uniformly to the microphone array aligned with the edge of the shear layer, i.e.,
perpendicular to the shear layer edge. In the region starting near 3 diameters downstream
and continuing through 9 diameters downstream, the acoustic field is characteristic of
a convecting disturbance. It has been pointed out in Picard and Delville (1999) that the
angle of the amplitude distribution corresponds to the convection velocity in the mixing
layer where present measurements support those assertions (the convection velocity is
roughly 183 m/s). In the downstream region, there appears to be some acoustic waves
in the frequency range of the dominant jet mixing noise although it is difficult to
pinpoint the frequencies from this type of plot especially since conventional spectra
indicate an overwhelming presence of broadband information. However, SPL autospectra
presented in Ponton, Ukeiley and Lee (1999) show the emergence of broadband peaks
at the typical jet noise frequencies between 2 and 3 diameters downstream of the jet exit
plane implying that these radiating source are associated with the dominant far field
jet noise.
Figure 12 displays the magnitude of the wavelet transforms for the streamwise
aligned microphone array spanning 3 to 9 jet exit diameters downstream of the exit

5000 5000

4000 4000
1/Scale, Hz

1/Scale, Hz

3000 3000

2000 2000

1000 1000

0.0025 0.005 0.0075 0.01 0.0025 0.005 0.0075 0.01

(A) 3 D Time, sec (B) 4 D Time, sec
Figure 12. Streamwise evolution of wavelet transform (r/D for the upstream
microphone = 1.8).
aeroacoustics volume 3 á number 1 á 2004 61

5000 5000

4000 4000
1/Scale, Hz

1/Scale, Hz
3000 3000

2000 2000

1000 1000

0.0025 0.005 0.0075 0.01 0.0025 0.005 0.0075 0.01

(C) 5 D Time, sec (D) 6 D Time, sec

5000 5000

4000 4000
1/Scale, Hz

1/Scale, Hz

3000 3000

2000 2000

1000 1000

0.0025 0.005 0.0075 0.01 0.0025 0.005 0.0075 0.01

(E) 7 D Time, sec (F) 8 D Time, sec

5000

4000
1/Scale, Hz

3000

2000

1000

0.0025 0.005 0.0075 0.01

(G) 9 D Time, sec
Figure 12. Continued

plane (r/D for the upstream microphone is 1.8). Similar to what was presented for the
azimuthal array, these plots have time as the abscissa and one over the scale factor as
the ordinate. However, these plots have been truncated at 5 kHz since it was previously
determined that there was very little information that could be observed above this
frequency in this downstream region. One of the features that is first observed from
62 On the near field pressure of a transonic axisymmetric jet

these plots is the expected increase in amplitude of the events as one progresses into the
dominant jet noise mixing region, i.e., the increase in dark regions in the contour plots.
It is also apparent from these plots that there is an increase in the events that occur in
the 1 kHz and lower region in the range of 7 to 9 jet diameters downstream of the jet
exit plane. It should be pointed out that these events do occur in the upstream region
also but are not nearly as prominent. The streamwise extent of the events observed in
these plots appears to be on the order of 1 to 2 jet exit diameters. This is consistent with
the streamwise correlation plots presented in Ponton, Ukeiley, and Lee (1999), the
contours of the instantaneous pressures previously discussed, and the streamwise extent
of streamwise correlations of the velocity field calculated from particle image velocity
measurements and reported in Seiner et al. (1999). These plots also shed information
about the frequency content of the waves observed to be propagating at the convection
velocity that can be approximated from the results presented in Figure 11. The
propagating waves typically correspond to frequencies in the range of 1 to 1.5 kHz. This
corresponds to Strouhal numbers between 0.18 and 0.27, which are typically regarded
as the peak frequencies at the radiation angles dominated by jet mixing noise.

4. CONCLUSIONS
The near field acoustics of a transonic jet have been studied through extensive
microphone measurements. These measurements were used to study the azimuthal and
streamwise structure of this field using several different time series analysis techniques.
This data set provides salient features of the near field pressure of a high Reynolds
number transonic jet that can be used for the validation of numerical simulations in
addition to the insight gained from the analysis reported here. Examination of the mean
velocity and the OASPL show the mean jet properties behave in an expected manner.
Examination of the azimuthal correlations of the near field show that the correlation
with the largest azimuthal extent is near the end of the potential core. This azimuthal
extent of large correlation coefficients appear to decay more rapidly in the upstream
direction versus the downstream direction, from the end of the potential core, implying
that a lower azimuthal mode number structure persists as the jet evolves in the
downstream direction. The altering of the large scale behavior of the azimuthal structure
around the end of the potential core is consistent with the Proper Orthogonal
Decomposition results of the flow field reported in Ukeiley et al. (1999) and Gamard et
al. (2002). Furthermore, since the near field pressure in the jet noise frequency range is
dominated by lower azimuthal mode number events, it is postulated that similar events
in the flow field should be responsible for the radiated noise. Low azimuthal mode
number events that originate near the center of the jet and evolve to the outer part of the
jet-mixing layer, which could be these events, have been observed by Citriniti and
George (2000) and Taylor et al. (2001). Events associated with the ejection of fluid with
large values of streamwise momentum to the outer part of the mixing layer have also
been linked to the noise source through a simplified Lighthill approach by Ukeiley
(2000). Additionally, the measurements show that most frequency components in the
acoustic near field of the jet radiate throughout the jet thereby highlighting the
importance of modeling the whole jet when attempting to characterize specific
aeroacoustics volume 3 á number 1 á 2004 63

frequencies in the far field. However, the existence of convecting sources at the typical
jet noise frequencies do not appear until between two and three jet exit diameters
downstream of the jet exit plane.

ACKNOWLEDGEMENTS
This work was performed while LSU held a NASA Langley Research Associateship
administered by the National Research Council. Current support for LSU has been
supplied by Air Force Office of Scientific Research, Dr. John Schmisseur Technical
Monitor, and is gratefully acknowledged. The authors would also like to acknowledge
the staff of the NASA Langley Jet Noise Laboratory and especially Misty Rich and
Sang Lee who were students in the laboratory while this work was initiated.

REFERENCES
Ahuja, K. K. and Bushell, K.W., “An Experimental Study of Subsonic Jet Noise and
Comparison With Theory,” Journal of Sound and Vibration, 30(3), pp. 317-341, 1973.
Arndt, R., Long, D. and Glauser, M., “The Proper Orthogonal Decomposition of
Pressure Fluctuations Surrounding a Turbulent Jet,” Journal of Fluid Mechanics,
340, pp. 1-33, 1997.
Cattafesta, L. N., Garg, S., Kegerise, M. A. and Jones, G. S., “Experiments on
Compressible Flow-Induced Cavity Oscillations,” AIAA Paper 98-2912, 1998.
Citriniti, J and George, W. K., “Reconstruction of the Global Velocity Field in the
Axisymmetric Mixing Layer Utilizing the Proper Orthogonal Decomposition,”
Journal of Fluid Mechanics, 480, pp. 137-166, 2000.
Freund, J. B. and Colonius, T. “POD Analysis of Sound Generation by a Turbulent Jet,”
AIAA Paper 2002-0072, 2002.
Fuchs, H. V., “Space Correlations of the Fluctuating Pressure in Subsonic Turbulent
Jets,” Journal of Sound and Vibration, 23(1), pp. 77-99, 1972.
Gamard, S., George, W. K., Jung, S. and Woodward, S., “Application of Slice Proper
Orthogonal Decomposition to the Far Field of an Axisymmetric Jet,” Phys. Fluids,
14(6), 2002.
George, W. K., Beuther, P. D. and Arndt, R. E. A., “Pressure Spectra in Turbulent Free
Shear Flows,” Journal of Fluid Mechanics, 148, pp. 155-191, 1984.
Goldstein, M. E., “Aeroacoustics of Turbulent Shear Flows,” Annual Review of Fluid
Mech., 16, pp. 263-285, 1984.
Jones, B. G., Adrian, R. J., Nithiandan, C. K. and Planchon, H. P., “Spectra of
Turbulent Static Pressure Fluctuation in Jet Mixing Layers,” AIAA Journal, 17(5),
pp. 449-457, 1979.
Jordan, D., Miksad, R. W. and Powers, E. J., “Implementation of the Continuous
Wavelet Transform for Digital Time Series Analysis,” Rev. Sci. Instrum. 68(3),
March 1997.
64 On the near field pressure of a transonic axisymmetric jet

Juve, D., Sunyach, M. and Comte-Bellot, G., “Intermittency of the Noise Emission
in Subsonic Cold Jets,” Journal of Sound and Vibration, 71(3), pp. 319-332,
1980.
Lau, J. C., Fisher, M. J. and Fuchs, H. V., “The Intrinsic Structure of Turbulent Jets,”
Journal of Sound and Vibration, 22(4), pp. 379-406, 1972.
Lewalle, J., “Wavelet Analysis of Experimental Data: Some Methods and the
Underlying Physics,” AIAA Paper 94-2281, June 1994.
Lighthill, M. J., “On Sound Generated Aerodynamically. I. General Theory.” Proc. Of
Royal Society A. 211, pp. 564-587, 1952.
Lilley, G. M., “Jet Noise Classical Theory and Experiments,” In Aeroacoustics of
Flight Vehicles: Theory and Practice, Vol. 1, (ed. H.H. Hubbard), pp. 211-289,
NASA RP-1258, 1991.
Lush, P. A., “Measurement of Subsonic Jet Noise and Comparison with Theory,”
Journal of Fluid Mechanics, 40(3), pp. 477-500, 1971.
Michalke, A. and Fuchs, H. V., “On Turbulence and Noise of an Axisymmetric Shear
Flow,” Journal of Fluid Mechanics, 70(1), pp. 179-205, 1975.
Morris, P. J., “Turbulence Measurements in Subsonic and Supersonic Axisymmetric
Jets in a Parallel Stream,” AIAA Journal, 14(10), pp. 1468-1475, 1976.
Narayanan, S., Barber, T. J. and Polak, D. R., “High Subsonic Jet Experiments: Turbulence
and Noise Generation Studies,” AIAA Journal, 40(3), pp. 430-437, 2002.
Picard, C. and Delville, J., “Pressure Velocity Coupling in a Subsonic Round Jet,” In
Engineering Turbulence Modeling and Experiments – 4, (eds. Rodi, W. and
Laurence, D.) Elsevier Sciences LtD., 1999.
Ponton, M. K., Ukeiley, L. S. and Lee, S. W., “Aeroacoustic Data for a High Reynolds
Number Axisymmetric Subsonic Jet,” NASA/TM-1999-209336, 1999.
Schaffar, M., “Direct Measurements of Correlation Between Axial In-Jet Velocity
Fluctuations and Far Field Noise Near the Axis of Cold Jet,” Journal of Sound and
Vibration, 64(1), pp. 73-83, 1979.
Seiner, J. M. and Reethof, G., “On the Distribution of Source Coherency in Subsonic
Jets,” AIAA Paper No. 74-4, 1974.
Seiner, J. M., Ukeiley, L. S., and Ponton, M. K., “Jet Noise Source Measurements Using
PIV,” AIAA Paper 99-1869, 1999.
Seiner, J., Ukeiley, L., Ponton, M. and Papp, J., “Jet Turbulence Measurements of
Aeroacoustic Source Using Stereo PIV,” AIAA Paper 2001-2184, 2001.
Stromberg, J. L., McLaughlin, D. K. and Trout, T. R., “Flow Field and Acoustic
Properties of a Mach 0.9 Jet at Low Reynolds Number,” Journal of Sound and
Vibration, 72(2), pp. 159-176, 1980.
Tam, C. K. W., “Jet Noise: Since 1952,” Theoretical and Computational Fluid Dynamics,
10, pp. 393-405. 1998.
Taylor, J., Ukeiley, L. and Glauser, M., “A Low-Dimensional Description of
Compressible Axisymmetric Shear Layer,” AIAA Paper 2001-0292, 2001.
aeroacoustics volume 3 á number 1 á 2004 65

Thurow, B., Hileman, J., Lempert, W. and Samimy, M., “A Technique for Real-Time
Visualization of Flow Structure in High-Speed Flows,” Phys. Fluids, 14,
pp. 3449, 2002.
Ukeiley, L. S. and Seiner, J. M., “Examination of Large Scale Structures in a Transonic
Jet Mixing Layer,” ASME Paper FEDSM98-5234, 1998.
Ukeiley, L. S., Seiner, J. M. and Ponton, M. K., “Azimuthal Structure of an
Axisymmetric Jet Mixing Layer,” ASME Paper FEDSM99-7252, 1999.
Ukeiley, L., “Examination of Quadrupole Sources in a Transonic Jet,” AIAA Paper
2000-0084.
Walker, S. H, Gordeyev, S. V. and Thomas, F. O. “A Wavelet Analysis Applied to
Unsteady Aspects of Supersonic Jet Screech Resonance,” Experiments in Fluids,
22(3), pp. 229-238, 1997.