close

Вход

Забыли?

вход по аккаунту

?

Jet noise shielding from airframe surfaces

код для вставкиСкачать
UNIVERSITY OF CALIFORNIA,
IRVINE
Jet Noise Shielding from Airframe Surfaces
THESIS
submitted in partial satisfaction of the requirements
for the degree of
MASTER OF SCIENCE
in Mechanical and Aerospace Engineering
by
Salvador Mayoral
Thesis Committee:
Professor Dimitri Papamoschou, Chair
Professor Feng Liu
Professor Manuel Gamero-Castaño
2010
UMI Number: 1476709
All rights reserved
INFORMATION TO ALL USERS
The quality of this reproduction is dependent upon the quality of the copy submitted.
In the unlikely event that the author did not send a complete manuscript
and there are missing pages, these will be noted. Also, if material had to be removed,
a note will indicate the deletion.
UMI 1476709
Copyright 2010 by ProQuest LLC.
All rights reserved. This edition of the work is protected against
unauthorized copying under Title 17, United States Code.
ProQuest LLC
789 East Eisenhower Parkway
P.O. Box 1346
Ann Arbor, MI 48106-1346
© Salvador Mayoral, 2010
Para mi familia
ii
Contents
LIST OF TABLES
vi
LIST OF FIGURES
vii
NOMENCLATURE
xv
ACKNOWLEDGEMENTS
xx
ABSTRACT OF THESIS
xxi
1
INTRODUCTION
1
1.1
Motivation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
2
1.2
Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3
1.3
Previous work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4
1.4
1.3.1
Jet noise . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4
1.3.2
Jet noise Suppression . . . . . . . . . . . . . . . . . . . . . . . . .
9
1.3.3
Jet noise Shielding. . . . . . . . . . . . . . . . . . . . . . . . . . .
11
Objectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
14
iii
1.5
2
3
15
2.1
HWB nozzle-shield configuration. . . . . . . . . . . . . . . . . . . . . .
15
2.2
Jet aeroacoustics facility. . . . . . . . . . . . . . . . . . . . . . . . . . .
17
2.3
Jet mean flow surveys . . . . . . . . . . . . . . . . . . . . . . . . . . . .
20
DATA PROCESSING
22
3.1
22
Acoustic analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3.1.1
Sound pressure level . . . . . . . . . . . . . . . . . . . . . . . . .
22
3.1.2
Effective perceived noise level . . . . . . . . . . . . . . . . . . . .
24
3.1.3
Effect of noise source redistribution on shielding. . . . . . . . . . .
27
Mean Flow Analysis. . . . . . . . . . . . . . . . . . . . . . . . . . . . .
29
RESULTS
31
4.1
Acoustic results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
31
4.1.1
Baseline unshielded and shielded configuration . . . . . . . . . .
32
4.1.2
Variable shield geometry . . . . . . . . . . . . . . . . . . . . . .
33
4.1.3
Variable nozzle configuration . . . . . . . . . . . . . . . . . . . .
36
4.1.4
Noise source maps . . . . . . . . . . . . . . . . . . . . . . . . .
41
Mean flow results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
43
4.2
5
14
EXPERIMENTAL SETUP
3.2
4
Thesis overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
CORRELATIONS
47
5.1
Formulation of average illumination angel . . . . . . . . . . . . . . . . .
47
5.2
Acoustic-mean flow correlations . . . . . . . . . . . . . . . . . . . . . .
49
iv
6
CONCLUSION
50
6.1 Summary. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
50
6.2 Future work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
52
Bibliography
54
Appendix A: Tables
62
Appendix B: Figures
70
Appendix C: Supplementary Data
98
v
List of Tables
2.1
Geometry of wedge fan flow deflectors (FFD) utilized in the study. . . . .
2.2
BPR10 takeoff engine cycle operating conditions for acoustic and mean
flow survey. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
2.3
62
62
Microphones with corresponding array position and sensitivity for
acoustic surveys. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
63
2.4
Microphone array position for noise source measurements. . . . . . . . . .
64
3.1
Conditions for PNL evaluations. . . . . . . . . . . . . . . . . . . . . . .
65
4.1
Acoustic and mean flow results for variable shield geometry. . . . . . . .
65
4.2
Acoustic and mean flow results for variable nozzle modifications. . . . . .
66
5.1
Average illumination angles and corresponding ∆EPNL (dB) from
shielding for all nozzle-shield configurations investigated. . . . . . . . . .
vi
67
List of Figures
1.1
Three dimensional illustration of the Hybrid-Wing-Body aircraft
(HWB). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
68
1.2
FSS and LSS similarity spectra from Tam et al. (1996). . . . . . . . . .
68
1.3
Turbulent mixing noise mechanisms in a jet, fine-scale turbulence and
turbulent large-scale structures. . . . . . . . . . . . . . . . . . . . . . .
69
1.4
Diagram of compressible free shear layer. . . . . . . . . . . . . . . . .
69
1.5
Mach wave radiation by a wavy surface propagating with supersonic
convective velocity, Uc. . . . . . . . . . . . . . . . . . . . . . . . . . .
2.1
Baseline BPR10 nozzle used in subscale experiment. a) Radial nozzle
coordinates; b) picture. . . . . . . . . . . . . . . . . . . . . . . . . . .
2.2
70
Scaling of HWB planform to UCI dimensions and retention of critical
dimensions for shielding (red lines). . . . . . . . . . . . . . . . . . . .
2.3
70
71
(a) Design of UCI HWB shield model with longitudinal traverse; b)
installation with nozzle. . . . . . . . . . . . . . . . . . . . . . . . . . .
vii
71
2.4
Subset of the nozzle configurations tested. (a) Mild chevrons; (b)
aggressive chevrons; (c) W18x3 wedge; (d) W18 wedge; (e)
MC+W18x3 combination; and (f) AC+W18x3 combination. . . . . . . .
72
2.5
UC Irvine Jet Aeroacoustics Facility. . . . . . . . . . . . . . . . . . . .
73
2.6
Diagram of dual-stream flow facility. . . . . . . . . . . . . . . . . . . .
73
2.7
Microphone setup for noise source imaging. . . . . . . . . . . . . . . .
74
2.8
Schematic of Pitot rake and traverse path for mean flow surveys. . . . .
74
3.1
Geometric relations for assessment of aircraft perceived noise level. . . .
75
4.1
Effect of shield on acoustics in the downward and sideline directions.
Shielded plain nozzle (blue) compared to unshielded baseline nozzle
(red). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.2
Insertion loss maps of baseline jet in the (a) downward and (b) sideline
directions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.3
76
77
Effect of shield on acoustics in the downward and sideline directions
with the +2Df nozzle-shield configuration (blue) compared to baseline
nozzle-shield configuration (red). . . . . . . . . . . . . . . . . . . . . .
4.4
78
Effect of shield on acoustics in the downward direction with the -1Df
nozzle-shield configuration with βe = +15° (blue) compared to -1Df
nozzle-shield configuration with βe = 0° (red). . . . . . . . . . . . . . .
viii
79
4.5
Effect of shield on acoustics in the sideline direction with the shield
without the verticals (red) compared to baseline nozzle-shield
configuration (blue) . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.6
79
(a) Noise source distribution of free baseline jet with corresponding (b)
plot of peak intensity location compared to the trailing edge position
(blue). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.7
80
Effect of shield on acoustics in the downward and sideline directions.
Shielded mild chevron nozzle (blue) compared to unshielded mild
chevron nozzle (green) and unshielded baseline nozzle (red). . . . . . .
4.8
81
Effect of shield on acoustics in the downward and sideline directions.
Shielded aggressive chevron nozzle (blue) compared to unshielded
aggressive chevron nozzle (green) and unshielded baseline nozzle (red).
4.9
82
Effect of shield on acoustics in the downward and sideline directions.
Shielded W18x3 nozzle (blue) compared to unshielded W18x3 nozzle
(green) and unshielded baseline nozzle (red) . . . . . . . . . . . . . . .
4.10
83
Effect of shield on acoustics in the downward and sideline directions.
Shielded W30 nozzle (blue) compared to unshielded W30 nozzle
(green) and unshielded baseline nozzle (red) . . . . . . . . . . . . . . .
4.11
84
Effect of shield on acoustics in the downward and sideline directions.
Shielded
AC+W18x3
nozzle
(blue)
compared
to
unshielded
AC+W18x3 nozzle (green) and unshielded baseline nozzle (red) . . . .
ix
85
4.12
(a) Noise source distribution of mild chevron jet with corresponding (b)
plot of peak intensity location. . . . . . . . . . . . . . . . . . . . . . .
4.13
Insertion loss maps of mild chevron jet in the (a) downward and (b)
sideline directions. . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.14
86
(a) Noise source distribution of aggressive chevron jet with
corresponding (b) plot of peak intensity location. . . . . . . . . . . . .
4.15
86
87
Insertion loss maps of aggressive chevron jet in the (a) downward and
(b) sideline directions. . . . . . . . . . . . . . . . . . . . . . . . . . .
. .
4.16
87
(a) Noise source distribution of W18x3 jet with corresponding (b) plot
of peak intensity location. . . . . . . . . . . . . . . . . . . . . . . . .
4.17
Insertion loss maps of W18x3 jet in the (a) downward and (b) sideline
directions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.18
4.20
88
(a) Noise source distribution of AC+W18x3 jet with corresponding (b)
plot of peak intensity location. . . . . . . . . . . . . . . . . . . . . . .
4.19
88
89
Insertion loss maps of AC+W18x3 jet in the (a) downward and (b)
sideline directions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
89
Contours of mean velocity for the baseline jet. . . . . . . . . . . . . . .
90
x
4.21
Contours of mean axial velocity on the symmetry plane. (a) Mild
chevrons; (b) aggressive chevrons; (c) W18x3 wedge; and (d)
AC+W18x3 combination. . . . . . . . . . . . . . . . . . . . . . . . . .
4.22
91
Cross-sectional mean velocity contours at various axial locations. (a)
Mild chevrons; (b) aggressive chevrons; (c) W18x3 wedge; and (d)
AC+W18x3 combination. . . . . . . . . . . . . . . . . . . . . . . . . .
4.23
92
Comparison of axial distribution of maximum mean velocity to baseline
jet. (a) Mild and aggressive chevrons; (b) W18x3 wedge and
AC+W18x3 combination. . . . . . . . . . . . . . . . . . . . . . . . . .
93
5.1
Illustration of the definition of illumination angle, ψ. . . . . . . . . . . .
94
5.2
Reduction of EPNL due to shielding versus average illumination angle
for all the configurations studied. . . . . . . . . . . . . . . . . . . . . .
5.3
Correlation between noise source length and primary potential core
length. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
C.1
95
95
Effect of shield on acoustics in the downward and sideline directions
with the +1Df nozzle-shield configuration (blue) compared to baseline
nozzle-shield configuration (red). . . . . . . . . . . . . . . . . . . . . .
C.2
96
Effect of shield on acoustics in the downward and sideline directions
with the -1Df nozzle-shield configuration (blue) compared to baseline
nozzle-shield configuration (red). . . . . . . . . . . . . . . . . . . . . .
xi
97
C.3
Effect of shield on acoustics in the downward and sideline directions
with βv = 90° (blue) compared to baseline nozzle-shield configuration
(red). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
C.4
98
Effect of shield on acoustics in the downward and sideline directions.
Shielded W15 nozzle (blue) compared to unshielded W15 nozzle
(green) and unshielded baseline nozzle (red). . . . . . . . . . . . . . . .
C.5
99
Effect of shield on acoustics in the downward and sideline directions.
Shielded W18 nozzle (blue) compared to unshielded W18 nozzle
(green) and unshielded baseline nozzle (red). . . . . . . . . . . . . . . .
C.6
100
Effect of shield on acoustics in the downward and sideline directions.
Shielded W21 nozzle (blue) compared to unshielded W21 nozzle
(green) and unshielded baseline nozzle (red). . . . . . . . . . . . . . . .
C.7
101
Effect of shield on acoustics in the downward and sideline directions.
Shielded TW15 nozzle (blue) compared to unshielded TW15 nozzle
(green) and unshielded baseline nozzle (red). . . . . . . . . . . . . . . .
C.8
102
Effect of shield on acoustics in the downward and sideline directions.
Shielded TW18 nozzle (blue) compared to unshielded TW18 nozzle
(green) and unshielded baseline nozzle (red). . . . . . . . . . . . . . . .
C.9
103
Effect of shield on acoustics in the downward and sideline directions.
Shielded TW18x3 nozzle (blue) compared to unshielded TW18x3
nozzle (green) and unshielded baseline nozzle (red). . . . . . . . . . . .
xii
104
C.10
(a) Noise source distribution of W15 jet with corresponding (b) plot of
peak intensity location. . . . . . . . . . . . . . . . . . . . . . . . . . .
C.11
Insertion loss maps of W15 jet in the (a) downward and (b) sideline
directions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
C.12
108
(a) Noise source distribution of TW15 jet with corresponding (b) plot of
peak intensity location. . . . . . . . . . . . . . . . . . . . . . . . . . .
C.19
108
Insertion loss maps of W30 jet in the (a) downward and (b) sideline
directions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
C.18
107
(a) Noise source distribution of W30 jet with corresponding (b) plot of
peak intensity location. . . . . . . . . . . . . . . . . . . . . . . . . . .
C.17
107
Insertion loss maps of W21 jet in the (a) downward and (b) sideline
directions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
C.16
106
(a) Noise source distribution of W21 jet with corresponding (b) plot of
peak intensity location. . . . . . . . . . . . . . . . . . . . . . . . . . .
C.15
106
Insertion loss maps of W18 jet in the (a) downward and (b) sideline
directions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
C.14
105
(a) Noise source distribution of W18 jet with corresponding (b) plot of
peak intensity location. . . . . . . . . . . . . . . . . . . . . . . . . . .
C.13
105
109
Insertion loss maps of TW15 jet in the (a) downward and (b) sideline
directions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
xiii
109
C.20
(a) Noise source distribution of TW18 jet with corresponding (b) plot of
peak intensity location. . . . . . . . . . . . . . . . . . . . . . . . . . .
C.21
Insertion loss maps of TW18 jet in the (a) downward and (b) sideline
directions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
C.22
111
Contours of mean axial velocity on the symmetry plane. (a) W15
wedge; (b) W18 wedge; (c) W21 wedge; and (d) W30 wedge. . . . . . .
C.25
111
Insertion loss maps of TW18x3 jet in the (a) downward and (b) sideline
directions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
C.24
110
(a) Noise source distribution of TW18x3 jet with corresponding (b) plot
of peak intensity location. . . . . . . . . . . . . . . . . . . . . . . . . .
C.23
110
112
Contours of mean axial velocity on the symmetry plane. (a) TW15
wedge; (b) TW18 wedge; (c) TW18x3 wedge; and MC+W18x3 combo.
xiv
113
NOMENCLATURE
Roman symbols
a
speed of sound
A
wave amplitude spatial growth and decay
b
octave band number
BPR
bypass ratio
C
duration correction
d
duration of the time defined by the points within PNLM-10 dB
D
nozzle exit diameter
EPNL
effective perceived noise level
f
frequency
f*
center frequency
FAA
U.S. Federal Aviation Administration
FAR
Federal Aviation Regulations
FFD
fan-flow-deflectors
FSS
fine-scale similarity
GRC
NASA John H. Glenn Research Center at Lewis Field
hw
wedge height
H
normalization time
HWB
hybrid wing body aircraft
k
wavenumber
l
length
xv
L
axial extent of the noise source region
LeRC
NASA Lewis Research Center
LSS
large-scale similarity
M
Mach number
N
number of microphones
NASA
National Aeronautics and Space Administration
NPR
nozzle pressure ratio
OASPL overall sound pressure level
OCT
1/3 octave band spectra
OTW
over-the-wing
p
pressure
p′
pressure fluctuation
PNL
perceived noise level
PNLM maximum perceived noise level
PSF
point-spread-function
QAT
Quiet Aircraft Technology Program
r
radial position
R
gas constant
Rm
microphone radius from the nozzle plug
Sr
Strouhal number
SLA
stereolithography apparatus
SPL
sound pressure level
t
time
xvi
T
temperature
Tmn
array response matrix
umax
local maximum velocity
U
mean velocity
UCI
University of California, Irvine
w
microphone weighting functions
x, y, z
axial, transverse, and spanwise coordinates relative to nozzle exit
Xs, Ys, Zs shield trailing edge location relative to nozzle exit
Greek symbols
αw
wedge half-angle
αa
aircraft angle of attack
βe
elevator flap angle
βv
vertical dihedral angle
∆EPNL change in effective perceived noise level
∆SPL
raw spectra correction
∆t
time increment
∆xw
axial displacement between wedge FFD vertex and fax exit plane
φ
azimuth angle
Φ
coherence-based beamforming output
γ
specific heat ratio
γa
aircraft climb angle
xvii
η
instability wave
λ
complex coherence of the pressure field
θ
polar angle relative to downstream jet centerline
ρ
density
ω
frequency
ξ
spatial coordinate
ψ
illumination angle
Ψ
microphone coherence-based noise source distribution
Subscripts
o
total
1
fast moving stream
2
slow moving stream
aa
atmospheric absorption
b
octave band number
c
convective
e
emission
eng
full-scale engine
exit
exit condition
exp
experimental
f
fan
ff
free field
xviii
fr
actuator frequency-responce
i
individual wave
ns
noise source
p
primary
pc
primary potential core
ref
reference
s
secondary
upper
upper limit
wp
wavepacket
∞
ambient
xix
ACKNOWLEDGEMENTS
I would like to thank Professor Dimitri Papamoschou, my advisor, for his support and
mentorship throughout this project. His guidance and positive reinforcement, especially
during my faults, helped me correct my mistakes and overcome technical challenges.
Second, I would also like to thank the members of my thesis committee, Professor
Feng Liu and Professor Manuel Gamero-Castaño, for their feedback and suggestions.
Very special thanks to Dr. Andrew Johnson for his mentorship throughout my
transition into graduate school and in my time in the Jet Aeroacoustics Lab. I also thank
Dr. Eric Unger (Boeing) for his design of the chevron nozzles.
Finally, my deepest thanks go to my parents, Salvador Mayoral Sr. and Esperanza
Sanchéz de Mayoral, and my sisters, Marisol, Esperanza, Érika, and Yoselín, for their
endless love and support.
This research has been funded by Boeing Subcontract No. 208547 in support of
NASA contract NNL07AA54C “Acoustic Prediction Methodology and Test Validation
for an Efficient Low-Noise Hybrid Wing Body Subsonic Transport.”
xx
ABSTRACT OF THESIS
Jet Noise Shielding from Airframe Surfaces
by
Salvador Mayoral
Master of Science in Mechanical and Aerospace Engineering
University of California, Irvine, 2010
Professor Dimitri Papamoschou, Chair
This work investigates the potential suppression of jet noise by airframe surface
shielding with application to the Hybrid Wing Body aircraft (HWB).
Subscale
experiments utilized a bypass-ratio-10 (BPR10) turbofan nozzle that operated at realistic
cycle conditions using helium-air mixture jets. The shielding surface was a flat plate
shaped in the form of a generic HWB and featured vertical fins and an adjustable
elevator. Chevrons and porous wedge fan flow deflectors were incorporated into the
baseline nozzle to restructure and alter the jet noise source distribution.
Acoustic measurements of the unshielded and shielded jets were conducted inside an
anechoic chamber using a 24-microphone apparatus composed of two polar arrays, each
consisting of twelve condenser microphones. One polar array was mounted at an azimuth
angle φ = 0° (downward) while the second array was mounted at φ = 60° (sideline).
Noise source distribution measurements were performed by densely grouping twentyfour microphones on a linear array in the downward direction. Surveys of the jet mean
flow were conducted with a fully automated three-dimensional Pitot rake traverse. The
xxi
Pitot rake consists of five high-resolution Pitot probes that measure the local jet total
pressure.
The potential for noise suppression was quantified in terms of the estimated Effective
Perceived Noise Level (EPNL), a noise metric used in aircraft certification. Acoustic
results of the baseline BPR10 nozzle with the HWB shield yielded only marginal EPNL
reductions, even when the engine was translated two fan diameters upstream of its
nominal location.
Using the estimated cumulative (downward plus sideline) EPNL
reduction as a figure of merit, shielding of the baseline nozzle with the nominal shield
configuration yielded a 2.5 dB reduction. Acoustic imaging of the baseline jet noise
source revealed that the large fan diameter, inherent to the high bypass ratio, created a
long noise source region. The integration of nozzle devices significantly improved EPNL
reduction with the engine positioned at its nominal location.
Application of the
aggressive chevrons increased the reduction to 6.5 dB, while the best wedge
configuration improved this value to 6.9 dB. Combination of wedge and aggressive
chevrons yielded a benefit of 7.6 dB. Examination of high-definition noise source maps
showed a direct link between insertion loss and axial location of the peak noise.
Ultimately, the enhanced shielding effect from the nozzle devices resulted from
translation of peak intensities upstream closer to the nozzle plug. Thus the compaction
and/or redistribution of the noise source via nozzle devices are essential for effective jet
noise shielding on the HWB.
A proposed correlator for the noise reduction due to shielding is the average
illumination angle. The average illumination angle is defined as the average angular
sector between noise source and observer not intercepted by the shield. For average
xxii
illumination angles less than 45 deg, the EPNL reduction exceeds 3 dB. A correlation
between the acoustic and mean flow results reveal that the mean velocity field by itself
cannot provide useful information for inferring the noise source location.
xxiii
Chapter 1
INTRODUCTION
Aircraft noise is one of the major sources of noise pollution in urban areas. Aircraft have
progressively gotten quieter with the integration of the high-bypass ratio turbofan engine
into the commercial aviation fleet since the 1970s. The low-specific-thrust design of the
turbofan engine yields a lower jet exit velocity that suppresses the jet noise component of
aircraft noise which is proportional to the eighth power of the jet exit velocity1.
However, the reduced aircraft noise levels have not effectively offset the growth of the
commercial aviation fleet to completely eliminate the noise issue. With the anticipated
congestion of cities and expansion of airports to support the future growth of air traffic,
aircraft noise exposure within airport-neighboring communities is expected to rise. To
mitigate this effect along with increasingly stringent aircraft noise regulations, the
development of the next generation of subsonic commercial aircraft aims to eliminate the
emission of aircraft noise outside of airport boundaries.
1
1.1 Motivation
The National Aeronautics and Space Administration (NASA) proposed the Quiet Aircraft
Technology (QAT) Program to reduce aircraft perceived noise levels by 10 dB in 10
years and 20 dB in 25 years, relative to a 1997 baseline2. To meet these goals, noise
reduction has become an integral factor in the airframe design process for the next
generation of aircraft. Thus all components of aircraft noise are under investigation for
potential noise reduction methodologies, including engine noise sources. A feasible
method to suppress engine noise for future aircraft is to mount the engines over-the-wing
(OTW) so that the wing may act as a sound barrier between engine noise sources and
airport communities.
A potential design for the next generation of subsonic commercial aircraft that utilizes
the OTW concept is the Hybrid Wing Body aircraft (HWB), illustrated in Figure 1.1.
The unconventional design of the HWB calls for two top-wing mounted engines, thus in
principle the aircraft airframe will shield ground observers from both the forwardemitting inlet noise sources and the aft-emitting jet noise sources. Experimental work on
quantifying the reduction of inlet noise from shielding by the fuselage has estimated a 14
dB reduction at low frequencies and a 20 dB reduction at high frequencies3. However,
the investigation of the potential jet noise reduction by airframe surface shielding is yet to
be fully explored and is the topic of this thesis.
Proper integration of the engines with the airframe for effective jet noise shielding
necessitates the development of accurate predictive tools. Specifically for jet noise, the
current state-of-the-art in empirical prediction of shielding approximates the jet noise
2
source as a small number of discrete point sources and utilizes barrier insertion loss
formulas which were developed for acoustic point sources4 and are based on Maekawa’s
experiments5. This method is inadequate for jet noise shielding predictions, for jet noise
is a distributed and directive source of finite spatial coherence whose exact nature
remains under investigation6. This study experimentally investigates the impact of jet
noise shielding on the effective perceived noise level (EPNL) through acoustic
measurements of a nozzle-shield combination composed of a dual-stream turbofan nozzle
with a scaled-down planform shield of the HWB; and mean flow surveys of the
corresponding jet plume. Nozzle devices such as chevrons and wedge type fan-flowdeflectors were integrated into the baseline nozzle to alter and restructure the jet noise
distribution. The focus was to gain a preliminary assessment of shielding for application
to the HWB, examine relations between distortions of the jet plume and noise source
distribution, and generate correlations that will assist in the development of more
advanced models of jet noise.
1.2 Background
In 1969 the U.S. Federal Aviation Administration (FAA) issued the Federal Aviation
Regulations (FAR) Part 36, to set limits on the maximum amount of aircraft noise
emitted7. The increase of the bypass ratio (BPR) within turbofan engines has resulted in
a significant reduction of jet noise levels over the last three decades. However, jet noise
continues to be a major contributor of aircraft noise. Subsonic jet noise is generated from
the turbulent mixing in the shear flow between the jet engine exhaust and ambient air.
Turbulent mixing noise consists of two separate noise sources: large-scale eddies in the
3
jet shear layer and fine-scale turbulence. Previous jet noise suppression efforts have
predominantly focused on enhancing the turbulent mixing via modifications to the nozzle
exit geometry as well as reducing the convective Mach number Mc of the emitted Mach
waves by thickening the secondary flow on the underside of the jet. A revived jet noise
suppression method is the OWT concept. Although significant experimental research on
the OTW concept for conventional and short-takeoff airplanes occurred in the 1970s,
commercially the OTW concept did not find support, with application to only the Fokker
VFW614.
A joint effort between the Boeing Company and NASA has led to the development of
the HWB for the next generation of subsonic commercial aircraft. Designed to address
long-term fuel efficiency, emissions, cargo capacity, and noise standards, the
revolutionary fixed-wing design of the HWB consists of a wide triangular-shaped
composite fuselage with high-aspect-ratio wings and features two high-bypass-ratio
engines mounted over the center-body behind the passenger compartment, illustrated in
Figure 1.1. Preliminary estimates indicate a reduction of takeoff weight by 15% and fuel
burn per seat mile by 27%8. Acoustically, it is projected the HWB will provide a
cumulative 52 dB reduction in total aircraft noise. A major contributor to this projection
is the suppression of the engine noise sources by airframe surface shielding.
1.3 Previous Work
1.3.1 Jet Noise
An extensive analysis on all the axisymmetric supersonic and subsonic jet noise spectra
within the database at the Jet Noise Laboratory at NASA Langley Research Center
4
carried out by Tam et al.9 revealed that the measured far-field spectral density is
composed of two similarity spectra, shown in Figure 1.2.
First is the Large-Scale
Similarity (LSS) spectrum that dominates in the peak noise directions. Second is the
Fine-Scale Similarity (FSS) spectrum that fits the data at larger angles relative to the jet
downstream axis. At intermediate angles, a combination of both LSS and FSS matched
the experimental measurements. It is argued that the existence of two similarity spectra
is attributed to two separate noise generation mechanisms. The LSS spectrum sources
from turbulent large-scale structures in the shear layer and the FSS sources from finescale turbulence. Noise generated from large-scale structures is highly directional and
peaks in the aft direction at shallow angles relative to the downstream centerline, shown
in Figure 1.3. Fine-scale turbulence is the primary source for noise radiated in the
direction normal to the jet centerline yet it is less intense than the noise from large-scale
turbulent structures. Further experimental studies by Viswanathan10 and Tam et al.11
supported the concept of two different noise sources for turbulent mixing noise at all jet
operating conditions.
Studies by Brown and Roshko12 were among the first to demonstrate the existence of
large coherent structures in turbulent jets and free shear layers in addition to previously
seen small-scale turbulence.
Experimental work on supersonic jets with low and
moderately-high Reynolds numbers by McLaughlin et al.13 and subsequently by Troutt
and McLaughlin14 demonstrated that the dominant noise generation mechanism in
supersonic jets is large-scale flow instabilities which contribute to the highly directional
behavior of jet noise. These studies measured jet flow fluctuations with a hot-wire and
measured near-field pressure measurements with microphones to detect the presence of a
5
spatially growing instability wave at an externally controlled forced frequency. This
acoustic effect was also detected in the far-field, near the walls of the anechoic test
facility where the experiments were conducted.
Turbulent large-scale structures radiate noise to the far-field when their convective
velocities exceed the ambient speed of sound. Theoretical works by Tam and Burton15,16
and by Crighton and Huerre17 utilize linear stability analysis to develop analytical jet
noise models by treating large-scale turbulent structures as linear instability waves. The
noise emission to the far-field can be analogized to supersonic flow over a wavy wall.
Consider the case of the compressible free shear layer, where the convective Mach
number in both the fast and slow streams is supersonic, illustrated in Figure 1.4. The
convective Mach number of the fast moving stream, Mc1, is given by
M c1 =
U1 − U c
a1
(1.1)
where U1, a1, and Uc are the fast stream velocity, speed of sound in the fast stream, and
convective velocity, respectively. The convective Mach number of the slow moving
stream, Mc2, is given by
M c2 =
Uc − U 2
a2
(1.2)
where U2 and a2, are the slow stream velocity and speed of sound in the slow stream,
respectively. Past the location where the two streams merge and prior to any nonlinear
effects, the perturbed vortex sheet takes on a structure similar to that of a wavy wall. The
flow experiences expansion waves shifted 90° from the crests and compression waves
shifted 90° from the troughs. By shifting the frame of reference to that of the traveling
6
wave, the wavy wall behaves as a propagating instability wave of equal wavelength and
convective velocity.
Therefore if the convective velocity of an instability wave is
supersonic, noise will radiate in the form of Mach waves as illustrated in Figure 1.518. It
is important to reemphasize that the convective Mach number of the fast and slow stream
are different, therefore the emitted Mach waves into the fast stream are not identical to
the Mach waves radiated into slow stream.
Mathematically, large-scale turbulent structures are modeled as instability waves. To
fully capture the true physical behavior of large-scale structures that propagate in shear
layers, the wavy wall analogy is modified to account for the growth and decay of the
instability wave as it propagates downstream. In turbulent free shear flows such as jets,
the shear layer grows with axial position. Near the nozzle exit, the shear layer is thin and
the mean-velocity gradient is large causing the instability wave to grow rapidly. As the
instability wave propagates downstream the shear layer thickens, reducing the meanvelocity gradient. Thus the growth rate of the instability wave also decreases. At some
point further downstream corresponding to the peak amplitude of the instability wave the
growth rate becomes zero. Upon propagating further downstream, the instability wave
experiences a dampening effect that continuously suppresses its amplitude until its effect
becomes negligible.
As previously mentioned, experimental studies by Viswanathan10 and Tam et al.11
demonstrated that noise is radiated from turbulent large-scale structures at all jet
operating conditions. Although it is clear that the supersonic jets emit noise through
Mach waves, this is not so obvious for a subsonic jet. In a parallel flow for a given
frequency, by subsonic wavy wall analogy an instability wave of constant amplitude
7
propagating with a subsonic convective velocity will generate pressure fluctuations that
exponentially decay with distance from the point of disturbance. In a jet however,
instability waves experience a growth then decay process due to the diverging shear flow.
Let η(x,t) describe a vortex sheet in the shear flow between the jet and ambient, and it
assumes the form of a wavepacket given by
η ( x, t ) = A( x)e
(
i k wp x −ω wp t
)
(1.3)
where x is axial position, t is time, kwp is the wavepacket wavenumber, ωwp is the
wavepacket frequency, and A(x) describes the wave amplitude spatial growth and
decay19. The convective Mach number of the instability wave is
Uc
a∞
Mc =
(1.4)
where the convective velocity is given by
Uc =
ω wp
(1.5)
k wp
The wavepacket can then be represented as a superposition of individual waves in Fourier
space19
η ( x, t ) =
1
2π
(
∫ Aˆ (k − k )e
∞
i kx −ω wp t
wp
−∞
)
dk
(1.6)
each with an individual wavenumber k, wave amplitude
Aˆ (k − k wp )dk
2π
and convective velocity
8
(1.7)
U c ,i =
ω wp
k
(1.8)
Therefore, for a given frequency, an instability wave whose amplitude spatially grows
then decays produces a broadband wavenumber spectrum rather than a discrete
wavenumber spectrum as in the case of a constant amplitude wave. The convective
Mach number for each wave is given by
M c ,i =
ω wp
ka∞
(1.9)
Therefore a portion of the broadband wavenumber spectrum may consist of supersonic
convective velocities, |k| < ωwp /a∞, that radiate noise through Mach waves to the far-field.
The waves with a |k| > ωwp /a∞ propagate with a subsonic convective velocity and decay
rapidly with axial distance. Thus the growth and decay envelope of the wave amplitude
is a critical factor in the noise generation process of turbulent large-scale structures.
1.3.2 Jet Noise Suppression
There has been significant progress on the quieting of jet engines. Early efforts include
the addition of corrugated lobes to the nozzle exit in Europe and venting the engine
exhaust through several tubes rather than a single smooth circular exit in the U.S.
However, these methods hindered the engine performance in exchange for the acoustic
benefit1. The major breakthrough in engine fuel efficiency and jet noise reduction was
the development of the turbofan engine. Since then new technologies such as tabs,
chevrons, bluebell nozzles, beveled nozzles, fluidic microjet injection, and fan-flow-
9
deflectors (FFD) have been investigated to suppress jet noise. Both nozzle chevrons and
FFD were investigated in the present study.
Nozzle tabs and chevrons effectively suppress jet noise for both medium-bypass and
high-bypass turbofan engines. Samimy et al.20 and Zaman et al.21 have demonstrated that
tabs which protruded into the jet flow generate a streamwise trail of counter-rotating
vortices that enhance mixing in the nozzle shear layer between the jet mean flow and the
ambient. This results in a reduction of the jet core. Acoustically, the enhanced mixing
reduces low-frequency noise yet amplifies high-frequency noise. Chevrons are triangular
shaped serrations mounted along the perimeter of the nozzle exit that are slightly
immersed into the exhaust stream. As the tabs, chevrons generate pairs of streamwise
counter-rotating vortices that enhance mixing in the corresponding shear layer and
ultimately shorten the jet potential core22. However, chevrons do not penetrate the flow
to the degree of the tabs. The induced vorticity is thus effectively weaker which permits
chevrons to efficiently reduce peak noise emissions with minimal thrust reduction22. In
dual-stream turbofan nozzles for a given thrust loss, the combination of chevrons on both
core and fan nozzles acoustically outperforms the combinations of a chevron core nozzle
only, and a chevron fan nozzle only23.
Thickening the secondary flow on the underside of a dual-stream jet represses jet
noise in the downward direction.
This effect was initially observed in studies
investigating the eccentric dual-stream nozzle configurations by Papamoschou and
Debiasi24 at the University of California, Irvine’s (UCI) Jet Aeroacoustics Laboratory and
later confirmed at NASA’s John H. Glenn Research Center at Lewis Field (GRC)25.
Later studies indicated that the thickening of the secondary fan flow extended the general
10
secondary core and generated higher noise reductions by lowering convective Mach
number26. Offsetting the secondary flow with respect to the primary flow is performed
by FFD. A practical FFD are wedge shaped flaps. Placed at the fan exit duct of a
convergent flow, a wedge FFD reshapes the jet flow by deflecting the fan stream
downward relative to the core stream and thickens the fan flow on the underside of the
jet. This results in a significant reduction in the length of the primary potential core27,28.
Porous wedge FFD mitigate the steep velocity gradients that can cause excess noise at
high polar angles seen in solid wedges29. Preliminary estimates of static-thrust loss for
high-bypass engine caused by a solid wedge projects a 0.5% reduction30.
1.3.3 Jet Noise Shielding
A considerable amount of research was conducted during the 1970’s on jet noise
shielding by airframe surface at what was formerly NASA’s Lewis Research Center
(LeRC); now GRC. Their work focused on understanding the effects of nozzle size and
location relative to the wing using OTW configurations with convergent circular nozzles
that operated at different velocities. Von Glahn et al.31 demonstrated that the shielding
effect reduced the sound pressure level (SPL) for middle and high frequencies, but can
amplify noise at low frequencies. Two sources were attributed to the low-frequency
noise amplification. One was believed to be generated from the interaction between the
jet flow and the trailing edge. The second was attributed to scrubbing noise that results
from the jet flow contacting the shielding surface. The low-frequency noise can be
minimized by positioning the nozzle significantly above the shield, extending the length
of the shielding surface, or deflecting the shielding surface downward. Reshotko et al.32
11
experimentally investigated the potential acoustic benefit of the OWT concept for
conventional and powered lift configurations. Results showed that jet noise reduction
increased with increasing frequency and that the OWT configurations were 10 dB quieter
than nozzle alone configurations at high frequencies.
Other sets of experiments by Von Glahn et al.33 at NASA LeRC examined the effects
of nozzle geometry and forward flight effects on jet noise reduction for the OTW
concept. The study utilized four different types of nozzles of equivalent diameters: single
stream convergent nozzle, dual-stream bypass nozzle, 6 and 8 tube mixer nozzles, and 6tube mixer nozzle with an ejector. For all cases the nozzle position relative to the shield
trailing edge was held constant. Jet and free stream velocities were varied. Results
demonstrated that as the jet velocity decreased the effect of shielding became more
dominant.
The mixer nozzles provided the best acoustic benefit in terms of SPL
reduction. The effect of forward velocity was investigated by adding a free stream, this
lengthened the jet core.
Yet while some tests demonstrated a negative impact on
shielding effectiveness, others showed little to no change in shielding effectiveness. A
more detailed investigation of the effect of tube and multi-lobed mixers on the velocity
decay for the OTW concept was later conducted by Von Glahn and Groesbeck34. Mixers
included 6 and 8 mixer tube nozzles as well as 7 and 8 lobed orifice type nozzles.
Results demonstrated that the alteration of the jet flow field by the nozzle geometry
modification yielded a greater jet noise shielding benefit when compared to the baseline
canonical nozzles. It was concluded that the enhanced mixing between the jet flow and
ambient shortened the jet core and effectively increased the relative shielding surface
area.
12
Further investigations on the effect of wing shielding of high-velocity jets and shock
associated noise for cold and hot flow jets were conducted by Von Glahn et al.35. Results
lead to the conclusion that there were no significant differences in jet noise shielding
effectiveness between hot and cold cases.
Short shielding surfaces that effectively
shielded jet noise for subsonic jets were not sufficiently long enough to fully shield shock
associated noise. This was attributed to shocks present at the tip of the supersonic jet
core what were not shielded by the airframe surface.
Recent studies on a fundamental nozzle–shield configuration composed of simple
single-stream convergent nozzle and rectangular shield were conducted at UCI by
Papamoschou and Mayoral36. The chord length, vertical displacement between the
nozzle centerline and the shield surface, and the axial displacement between the shield
trailing edge and the nozzle exit plane were varied. Additionally, four triangular mixer
tabs were added along the lip of the nozzle exit to modify the jet mean flow and acoustic
behavior. Acoustic results demonstrated that at low frequencies, the shield produced
excess noise which amplified with increasing polar angle. However at higher frequencies,
the shield significantly reduced the sound pressure level. The introduction of mixing tabs
increased the level of noise attenuation; in downward direction the shielded case with the
tabs yielded an EPNL reduction of 9.2 dB when compared to the unshielded case with
tabs. Acoustic imaging of the jets revealed that the addition of the tabs to the singlestream jet compacted the noise source region by translating the peak intensities upstream
closer to the nozzle exit.
13
1.4 Objectives
The research detailed in this thesis is focused on suppressing jet noise by airframe surface
shielding while achieving minimal thrust loss for applications to the HWB.
The
objectives of the study include:
•
Acoustic measurements of the shielded baseline nozzle with variable shield
geometry.
•
Acoustic measurements of shielded nozzle with chevrons, FFD, and combinations
with the nominal shield configuration.
•
Identification of the jet noise distribution and peak noise locations using phased
array beamforming methods for different nozzle configurations.
•
Surveying the jet mean flow of different nozzle configurations to examine
potential relations between the distortion of the jet plume and corresponding noise
distribution.
1.5 Thesis Overview
This thesis is divided into six chapters.
•
Chapter 1: discusses the general background and the objectives of the thesis.
•
Chapter 2: details the nozzles, shield, and facilities used in the experiments.
•
Chapter 3: explains the post-processing of the raw experimental data.
•
Chapter 4: presents and discusses the acoustic and mean flow results.
•
Chapter 5: presents correlations aimed at unifying the data.
•
Chapter 6: summarizes the thesis and its key findings.
14
Chapter 2
EXPERIMENTAL DETAILS
2.1 HWB nozzle–shield configuration
Subscale jet noise shielding experiments were carried out using a nozzle–shield
configuration composed of a dual-stream nozzle with a scaled-down planform shield of
the HWB. The nozzle was a dual-stream turbofan-type nozzle with a fan diameter Df
=3.12 cm and a bypass-ratio of 10 (BPR10). Its design is based on GRC’s 5BB turbofan
nozzle. The radial coordinates of the 5BB nozzle were obtained from GRC and scaled
down to fit within the flow capacity of UCI’s Jet Aeroacoustics Facility. Nominally with
BPR=8, the 5BB nozzle plug was enlarged to reduce the core area and increase the BPR
to approximately 10. The radial coordinates, normalized by Df, of the fan nozzle, core
nozzle, and nozzle plug are plotted in Figure 2.1a. The nozzle was manufactured from
UV-curable photopolymer resin with a stereolithography apparatus (SLA). The SLA
fabricates components in 0.007-inch thick layers by tracing the component cross-section
on the resin surface with a UV laser. Exposure to the UV laser cures the resin, while
simultaneously adhering it to the layer below and solidifying a three-dimensional
15
component. Because the UV-curable resin cannot be fabricated to very fine thicknesses,
the relative thickness of the nozzle exit lip is larger than that of a full-scale nozzle. The
manufactured baseline BPR10 nozzle is shown in Figure 2.1b.
The shield counterpart to the HWB nozzle–shield configuration is a scaled-down
planform of the HWB airframe, depicted in Figure 2.2. Scaling is based on the Df. The
full scale Df corresponds to 279.4-cm while the small scale Df is equal to 3.12-cm. This
yields a scale factor of approximately 90. Using this scale factor, a two-dimensional
HWB planform shield was designed and manufactured from a 3.2-mm thick aluminum
sheet with all of the essential dimensions. The shield features removable/adjustable
verticals and a removable/adjustable elevator flap to acoustically simulate takeoff and
cruise. Additionally, the HWB shield is mounted on a longitudinal traverse that permits
the axial displacement of the nozzle either upstream or downstream of its nominal
position. The detailed design of the shield and its integration with the BPR10 nozzle is
depicted in Figure 2.3.
To modify the jet flow field and alter the acoustic behavior of the BPR10 jet,
chevrons and wedge-type FFD were incorporated into the baseline BPR10 nozzle. Two
different sets of chevrons were integrated into both the fan and core nozzles of the
BPR10 nozzle. Both sets of chevrons were designed by the Boeing Company. The first
“mild” set of chevrons featured 10 serrations with a 10° insertion angle, as shown in
Figure 2.4a. The second “aggressive” set of chevrons featured 10 serrations with a 20°
insertion angle, shown in Figure 2.4b. This study also integrated porous wedge FFD of
variable geometry into the baseline BPR10 nozzle. Wedge half-angles αw ranged from
16
15° to 30°.
Wedge heights hw of 5-mm and 7-mm were used.
Relative axial
displacements between the wedge vertex and fan exit plane ∆xw of 0-mm and 3-mm were
used. The wedge length lw was held constant at 13-mm. The porous wedges were
fabricated from a fine interwoven metal mesh with a mesh size of 0.0223-mm and
porosity of 49.6%. Figure 2.4c-d shows two sets of wedge FFD utilized in the study,
W18x3 and W18. Table 2.1 list the wedge geometry specifications of all the wedges
tested in this study. Combinations of select wedges and chevrons were also investigated.
Figure 2.4e-f shows the combinations of W18x3 wedge with mild and aggressive
chevrons, respectively.
2.2 Jet Aeroacoustics Facility
Subscale jet noise shielding acoustic measurements were conducted at UCI’s Jet
Aeroacoustics Facility, depicted in Figure 2.5. This subscale facility houses a dualstream coaxial jet apparatus, illustrated in Figure 2.6, and two twelve-microphone polar
arrays azimuthally mounted in the downward and sideline directions inside an anechoic
chamber.
To capture the acoustic behavior of a full-scale heated jet under cold operating
conditions, one would attempt to match the nozzle exit conditions: velocity U, Mach
number M, and density ρ. Dimensional analysis of Lighthill’s equation reveals that
acoustic intensity scales to an eight-power law with the jet exit velocity37, 38; therefore it
is highly critical that the subscale exit velocity be matched to the full-scale exit velocity.
Given that the exit temperatures cannot be matched with cold operating conditions, the
exit Mach number may be matched by increasing the product of the specific heat ratio
17
and gas constant γR. This is accomplished by utilizing a gas or mixture of gases with a
molecular weight lower than air. Helium is a prime candidate for it is readily available,
non-toxic, and non-combustible. Previous studies by Papamoschou39 have demonstrated
that matching exit velocity and Mach number conditions with helium-air mixtures is an
effective and inexpensive method to accurately simulate the acoustic behavior of a fullscale heated turbofan jet. Comparisons between full-scale acoustic data at GRC and
subscale acoustic data at UCI for a BPR=8 dual-stream nozzle with integrated FFD,
demonstrates good agreement in spectral shapes and changes in the noise spectral
characteristics40. The coaxial jet apparatus at UCI is designed to deliver helium-air
mixtures at ambient temperatures to the primary (core) and secondary bypass (fan)
nozzles. The helium mass fraction and the total pressure of the helium-air mixture are
determined from the desired exit Mach number and exit velocity. The individual mass
flow rates of air and helium are set by nozzle size. The helium mass fraction is set by
initially running air alone through the nozzle to match the air-alone total pressure,
followed by the addition of helium to match the total pressure. For aeroacoustic surveys,
the BPR10 nozzle operated with helium-air mixtures in both the core and fan nozzles at
the hot take-off engine cycle conditions listed in Table 2.2. The resultant Reynolds
number based on annular height of the core and fan jets were 0.512×105 and 0.778×105,
respectively.
Noise measurements were performed with a 24-microphone apparatus composed of
two polar arrays each consisting of twelve 3.2-mm condenser microphones (Bruel &
Kjaer, Model 4138) with a frequency response of 140 kHz. The polar angle θ relative to
18
the downstream nozzle centerline of the microphones ranged from 20° to 120°. One
polar array is mounted at an azimuth angle φ = 0° (downward) while the other is mounted
at φ = 60° (sideline). Figure 2.5 depicts the downward polar array configuration; the
sideline arm is practically identical.
This arrangement enabled simultaneous
measurement of the downward and sideline noise at all the polar angles of interest. Table
2.3 lists the corresponding θ, radius from the nozzle plug Rm, and φ for each microphone.
The microphones were connected in groups of four to six amplifier/signal conditioners
(Bruel & Kjaer, Model 4138).
The 24 outputs of the amplifiers were sampled
simultaneously, at 250 kHz per channel, by three eight-channel multi-function data
acquisition boards (National Instruments PCI-6143).
National Instruments LabView
software was used to acquire the signals which were filtered through a high-pass filter set
to 300 Hz and a low-pass filter set to 140 kHz. A run time of just over a second
corresponds to 262,144 samples and is performed by each microphone. The temperature
and humidity were recorded to facilitate the computation of the atmospheric absorption
correction. The arrangement of the microphones inside the anechoic chamber and the
principal electronic components are detailed in Figure 2.5.
The Jet Aeroacoustics Facility is housed by a cubic anechoic chamber that is 8-m3 in
internal volume.
The inner walls of the anechoic chamber are lined with 6-in tall
polyurethane foam acoustic wedges (Illbruck SONEXsuper) that consist of an absorption
coefficient greater than 1.0 for frequencies higher than 500 Hz. The exhaust of the
anechoic chamber is lined with polyurethane foam acoustic baffles (Illbruck
SONEXmini).
19
Noise source measurements of each unshielded nozzle configuration were performed
by densely grouping microphones in the downward direction. The microphones were
arranged in a linear array with 1.0-inch axial spacing between each microphone, as
shown in Figure 2.7. The microphone polar angles ranged from φ = 47.5° and φ = 73.0°
which corresponds to a polar aperture of 27.5°. Table 2.4 lists the corresponding θ and
Rm, for each microphone during noise source measurements.
The imaging of the
distributed noise sources are detailed in the following chapter.
2.3 Jet Mean Flow Surveys
A corresponding mean velocity survey of the jet was conducted for each acoustically
surveyed nozzle-configuration.
The mean flow facility is composed of a duplicate
coaxial dual-stream apparatus and a three-dimensional Pitot rake traverse.
The extensive run-times for the mean flow surveys makes the use of helium-air
mixtures financially unfeasible.
Thus only pure air is used for both primary and
secondary streams resulting in the inability to simultaneously match the subscale Mexit
and Uexit to their corresponding full-scale values. Experimental work by Forstall and
Shapiro41 on the mass and momentum transfer of coaxial jets demonstrated that the
spreading rate and turbulent transport are governed by the velocity ratio between the
primary and secondary flows Us/Up. Thus to preserve the fundamental mixing mechanics
in the mean flow surveys, the velocity ratio Us/Up = 0.72 is kept the same as in the
acoustic tests. Table 2.2 lists the cold operating conditions for the BPR10 nozzle during
the mean flow surveys. The resultant Reynolds number based on annular height of the
core and fan jet were 0.292×105 and 0.567×105, respectively.
20
The Pitot rake utilized to survey the total pressure in the jet plume consists of five
high-resolution Pitot probes vertically spaced 10-mm apart, shown in Figure 2.8.
Attached to a streamlined airfoil-shaped support, the probes are 50-mm in length and
consist of a 0.5-mm internal diameter. Each probe is connected to a Setra Model 207
pressure transducer. The pressure transducers are sampled at a rate of 1000 Hz by an
analog to digital data acquisition board (National Instruments PCI-MIO-16E) and
recorded using National Instruments LabView software. The top probe is set as the
reference probe and is initially positioned at the tip of the nozzle plug.
The rake is mounted on a three dimensional traverse system that features motorized
motion in the x, y, and z directions with the negative y direction oriented towards the
ground (φ = 0). The traverse is driven by three IMS MDrive 23 motor drivers, each
individually connected to THK LM Guide Actuators. The traverse system is controlled
remotely with National Instruments LabView over a pre-specified traverse array
consisting of 14 axial planes spanning 6.5 inches. Each axial plane compromises 17
horizontal passes of length 101.6 mm spaced 2.5 mm vertically apart. The horizontal
passes are surveyed at a speed of 10.2 mm/s.
The data of each y-z plane were interpolated on a fixed grid. The measured Pitot
pressure is converted to velocity under the assumptions of isentropic conditions, the static
pressure is equal to the ambient pressure, and the total temperature is equal to ambient
temperature. A Savitzky-Golay filter42 was used to smooth the velocity profiles. The
calculation of the flow Mach numbers and velocities of the jet plume are detailed in the
following chapter.
21
Chapter 3
DATA PROCESSING
3.1 Acoustic analysis
Post-processing of the raw acoustic data for each experimental configuration computes
several acoustic metrics: sound pressure level spectra, overall sound pressure level
(OASPL), effective perceived noise level, and images of the noise source distribution.
3.1.1 Sound pressure level
Using the microphone sensitivities and amplifier gain settings listed in Table 2.3, the
pressure trace is computed from the microphone voltage signal. This is then transformed
into the narrowband sound pressure level spectrum with a 4096-point fast Fourier
transform which provides a spectral resolution of 61 Hz. The raw sound pressure level
(SPLraw) spectrum is corrected for microphone actuator frequency-response ∆SPLfr,
microphone free-field response ∆SPLff, and atmospheric absorption ∆SPLaa. This is
performed in the frequency f domain. The corrected lossless spectrum is given by
SPL( f ) = SPLraw ( f ) − ∆SPL fr ( f ) − ∆SPL ff ( f ) + ∆SPLaa ( f )
22
(3.1)
where the sound pressure level (SPL) and corrections are in decibels.
Both the frequency-response and microphone free-field response corrections are
performed in accordance with correction curves provided by the microphone
manufacturer, Bruel & Kjaer. The actuator frequency-response correction is typically
minor. The free-field correction may be significant for frequencies higher than 50 kHz.
The free-field response correction is used to correct the measured pressure from the
interference and diffraction of sound waves with wavelengths of the same order or less
than the microphone diameter43.
The excess noise from the internal interference and
diffraction produces a higher measured pressure response than the true value.
The
atmospheric absorption correction is utilized to render the spectra into lossless form. The
atmospheric absorption correction depends on the measured values of relative humidity
and temperature of the ambient air.
The overall sound pressure level (OASPL) is calculated by integrating the corrected
lossless SPL spectrum:
OASPL = 10 log10 ∫
f upper
0
100.1SPL ( f ) df
(3.2)
where the upper limit is the highest resolvable frequency which corresponds to 125 kHz.
The OASPL is referenced from a distance of 12-inch from the nozzle plug.
For
assessment of the full-scale spectrum, the corrected SPL spectrum is extrapolated to
frequencies higher than those resolved in the experiment (125 kHz) using a decay slope
of -30 dB/decade for full-scale perceived noise purposes. The full-scale corrected SPL
spectrum is obtained by dividing the measured frequencies by the scale factor of 90 and
is also referenced to a distance of 12-inch from the nozzle plug.
23
3.1.2 Effective perceived noise level
Although the OASPL is a physical value, when working with community noise it is
highly critical to factor in the element of human “annoyance”. The human response to
sound is more sensitive of mid- and high-frequencies thus the resulting OASPL may not
truly reflect the perceived noise heard by an observer. Extensive work on quantifying
human annoyance to sound at different frequencies and levels has lead to the
development of weighted sound pressure level curves A, B, C, D, and N1. Yet when
dealing specifically with aircraft noise the perceived noise level (PNL) is utilized and it is
the basis for determining the effective perceived noise level (EPNL).
The EPNL includes the additional metric of the “duration” of aircraft noise exposure
by recognizing the Doppler effect of an aircraft flyover. EPNL is an internationally
recognized unit that is primarily used for aircraft noise certification purposes. Typically
its calculation is evaluated at three separate flight conditions: takeoff, approach flight
paths, and noise at the side of the runway. However, for the present study only the
takeoff (downward) and noise on the side of the runway (sideline) EPNLs are evaluated.
On approach the engines are generally idled and the resulting aircraft noise is dominated
by airframe noise.
The conversion of SPL spectra to EPNL is detailed in FAR Part 367. A summary
outlining the essential steps for estimating ENPL is provided. The conditions used for
EPNL calculations in the downward and sideline directions are listed in Table 3.1 and
Figure 3.1 presents the geometric relations associated with a typical takeoff profile. The
basic EPNL estimation procedure is as follows:
24
1. The flight path is computed at 0.5-sec intervals. For each observation time t, the
distance r(t), polar emission angle θe(t), and azimuth emission angle φe(t) are
computed for the “retarded” position for an aircraft with Mach number M∞, angle
of attack αa, and climb angle γa using the relations detailed in Figure 3.127,28.
2. For each t, the lossless, scaled-up spectrum corresponding to θe(t) and φe(t) is
obtained from the experimental measurements. This step requires interpolation
between spectra and, for polar angles outside the range covered in the experiment,
moderate extrapolation. To enhance the accuracy of interpolation or extrapolation
the spectra were smoothed using a Savitzky-Golay fillter42 to remove extraneous
fluctuations yet preserve the basic shape of the spectrum.
3. The spectrum is Doppler-shifted to account for the motion of the aircraft. The
relations of McGowan & Larson44 are used:
f flight
f static
=
1 + (M c − M ∞ ) cos θ
1 + M c cos θ
(3.3)
The convective Mach number Mc is obtained from the empirical relations of
Murakami & Papamoschou45.
4. The spectrum is corrected for distance and atmospheric absorption. The distance
correction is
 (r / D f )eng 

− 20 log10 
 (r / D f )exp 
25
(3.4)
The atmospheric absorption is calculated for an ambient temperature of 29°C and
relative humidity of 70% (conditions of least absorption) is used in the present
study. The formulas proposed by Bass et al.46 are used to calculate its value.
5. The spectrum is discretized into 1/3-octave bands. The center frequency of the bth
octave band is determined by
*
1/ 3 *
f oct
f oct ,b−1
,b = 2
(3.5)
*
f oct
,1 = 50 Hz.
(3.6)
where
−1 / 6 *
f oct ,b and 21/ 6 f oct* ,b , respectively.
The lower and upper limits of band b are 2
The sound pressure level in the band b is obtained by integrating the SPL
spectrum as follows:
*
21 / 6 f oct
,b
OCTb = 10 log10 ∫ −1 / 6
2
*
f oct
,b
*
10 SPL ( f ) df *
(3.7)
6. The PNL is computed from the 1/3-octave band spectra in accordance with Part
36 of the Federal Aviation Regulations7.
7. The PNL is corrected for lateral attenuation according to SAE AIR 157147. This
applies only to the sideline estimate.
8. The above steps result in the time history of perceived noise level, PNL(t). From
it, the maximum level of PNL, PNLM, is determined. The duration of PNL
exceeding PNLM-10 dB is calculated and the corresponding “duration correction”
is given by
26
 1 d / ∆t
PNL(t ) 
C = 10 log   ∑ anti log
− PNLM
10 
 H  b =1
(3.8)
where H =10-sec is the normalizing time constant, ∆t =0.5-sec is the time
increments, and d is the duration of the time defined by the points within PNLM10 dB.
9. The effective perceived noise level is the algebraic sum of the PNLM and the
duration correction.
EPNL = PNLM + C
(3.9)
It is important to note that the EPNL estimate does not include the “tone correction”,
a penalty for excessively protrusive tones in the 1/3-octave spectrum which are absent
from our spectra anyway. Additionally, this estimate does not include factors such as
aerodynamic effects of forward flight, installation effects, and noise from sources other
than turbulent mixing noise (e.g., fan tones, airframe sound) that affect the overall EPNL
of a real airplane. The only correction for forward flight is the Doppler shift of Step 4
above. However EPNL does include metrics of community noise, duration of noise, and
human annoyance.
The change in EPNL (∆EPNL) in the downward and sideline
directions were used as a preliminary “figure of merit” in the assessment of shielding for
each nozzle-shield configuration.
3.1.3 Noise Source Maps
Imaging of jet noise sources requires that measurements use a narrow-aperture array in
both the peak emission and lateral direction to distinguish between the two turbulent
mixing noise mechanisms.
The imaging methodology utilized in the present study
27
produces self-consistent maps of the noise source distribution by incorporating a
directionality parameter in the formulation of the noise source model. This means that
the far-field auto-spectrum of the pressure in a particular polar direction is obtained with
the axial integration of the noise source.
Noise source maps are generated from the deconvolution of the delay-and-sum
beamforming output which is based on the complex coherence of the acoustic field. A
detailed discussion regarding the determination of the distribution and noise source maps
can be found in Papamoschou48. In the delay-and-sum beamforming, the coherencebased beamforming output Φ is given by
N
N
*
(x, ω )λmn (ω )
Φ( x, ω ) = ∑∑ wm wnTmn
(3.10)
m =1 n =1
where wm are user specified microphone weighting functions of radian frequency ω and
axial spatial coordinate x, Tmn is the array response matrix, and λ is the complex
coherence of the pressure field. The array response matrix that describes the modeled
coherence of the acoustic field for a point source at x, its defined by
Tmn ( x, ω ) = eiω [τ m ( x )−τ n ( x )]
(3.11)
where τm and τn are the time delays for microphones m and n, respectively.
The
coherence is related to the microphone coherence-based noise source distribution Ψ by
λmn (ω ) = ∫ Tmn (x, ω )Ψ ( x, ω )dx
(3.12)
L
where L is the axial extent of the noise source region. On defining the point-spread
function (PSF) as
28
N
N
*
( x, ω )
PSF ( x, ξ , ω ) = ∑∑ wm wnTmn (ξ , ω )Tmn
(3.13)
m =1 n =1
where ξ is a spatial coordinate. The coherence-based beamforming output is rewritten as
Φ( x, ω ) = ∫ Ψ (ξ , ω )PSF ( x, ξ , ω )dξ
(3.14)
L
The above equation states that the delay-and-sum beamformed array output is the
convolution of the noise distribution with the PSF. Extraction of the coherence-based
noise source distribution is performed iteratively by solving the discretized linear system.
The solution procedure parallels the DAMAS approach taken by Brooks and
Humphreys49, however the Richardson-Lucy (RL) inversion algorithm50,51 is used for the
deconvolution. The noise source maps presented in this thesis are based on the coherence
λmn of the acoustic field, meaning that they are normalized by the autospectrum of the far
field pressure.
3.2 Mean Flow Analysis
The post-processing of the Pitot pressure signals from the mean flow surveys is
performed to resolve the mean axial velocity in the jet plume. To improve the accuracy
calculation of the velocity from the Pitot pressure, the signals were smoothed in order to
eliminate any undesirable noise.
The velocity field of the jet plume is obtained from the measured total pressure. The
local Mach number and velocity are calculated under the following assumptions: the flow
is isentropically brought to rest; the static pressure p inside the jet and downstream of the
nozzle exits is equal to the ambient pressure p∞; and the total temperature To is equal to
29
the ambient temperature T∞ and remains constant throughout the nozzle and jet flow.
Since the BPR10 jet operates at subsonic conditions, the Pitot pressure ppitot is equal to
the total pressure po, and the local Mach number M is obtained from the isentropic
pressure relation
p pitot
p∞
p
 γ −1 2 
= o = 1 +
M 
p∞ 
2

γ / (γ −1)
(3.15)
where γ is the ratio of specific heats of the gas. The velocity U is determined from the
obtained Mach number
U = Ma = M γRT
(3.16)
where R is the gas constant and T is the static temperature which is determined by the
adiabatic temperature relation
T=
To
1+
γ −1
2
M2
(3.17)
For pure air γ = 1.4 and R = 287 kJ/kg K.
The potential core is defined by a velocity threshold of the maximum core exit
velocity. In this study the primary potential core lpc is defined as the distance between the
nozzle plug tip and the point where the local maximum velocity decays to 90% of the
core exit velocity.
30
Chapter 4
RESULTS
4.1 Acoustic results
Acoustic measurements of the unshielded and shielded jets were conducted at polar
angles ranging from 20° to 120° in both the downward (φ = 0°) and sideline (φ = 60°)
directions. Post-processing of the raw acoustic data for each experimental configuration
computed several acoustic metrics: lossless SPL spectra, OASPL, and EPNL. For each
nozzle-shield configuration an “acoustic summary” is presented. An acoustic summary is
composed of the following quantities: full-scale (scale factor of 90) narrowband lossless
SPL spectra at selected polar angles; directivity of OASPL; PNL versus time; PNL versus
observer polar angle; and an estimate of EPNL. First, the acoustic results of the baseline
unshielded and shielded configuration are presented, followed by the acoustic results of
the baseline jet with variable shield geometry and the acoustic results of variable jets with
the shield configured to its nominal position and configuration. A select number of cases
are presented for brevity; however acoustic results for all cases listed yet not mentioned
may be found in Appendix C.
31
4.1.1 Unshielded and shielded configuration with baseline nozzle
The investigation of the potential jet noise reduction from airframe surface shielding
began with taking acoustic measurements of the unshielded baseline jet (BPR10) and the
shielded baseline jet with the HWB shield setup in its nominal configuration. The
nominal shield configuration consist of an elevator deflection angle of βe = 0°, vertical
dihedral angle of βv = 79°, and the fan exit plane is positioned 2.27 Df upstream of the
trailing edge. Figure 4.1 compares the acoustic results from the unshielded baseline jet
(red) to the shielded baseline jet (blue) in the downward and sideline directions. At the
nominal position, the shield modestly attenuates high-frequency noise at large polar
angles in both the downward and sideline directions. The directivity of the OASPL
demonstrates that the shield amplifies the OASPL at low polar angles and reduces the
OASPL for large polar angles with a crossover polar angle between 50° and 60°. The
resulting EPNL reductions are 1.07 dB and 1.29 dB in the downward and sideline
directions, respectively.
To obtain a general idea as to how effective the shield attenuates noise for a given
frequency and polar angle, an insertion loss map is presented. The insertion loss is
defined as the reduction in sound pressure level due to shielding alone for a given nozzle
configuration. The insertion loss is plotted against Strouhal number Sr and polar angle.
A positive value is indicative of noise reduction. The Strouhal number represents a
normalized frequency and is defined by
Sr =
fD f
Uf
32
(4.1)
Figure 4.2 plots the insertion loss maps for the shielded baseline jet in both the downward
and sideline direction. In both azimuthal directions the insertion loss is very small except
at high frequencies and high polar angles. For example, in examining the insertion loss at
θ = 100° the insertion loss is significant for Strouhal numbers greater than 6. Below this
value there are modest insertion loss levels.
4.1.2 Variable shield geometry with baseline nozzle
A parametric investigation of the HWB shield geometry on the shielding of the baseline
jet varied the axial displacement of the BPR10 nozzle from its nominal position, elevator
flap angle, and vertical dihedral angle. The acoustic results are compared to baseline
shielded configuration to isolate the effect of each parameter (red).
Translating the BPR10 nozzle about its nominal location changes the shielding
surface area. The BPR10 nozzle was displaced 1 fan diameter downstream (+1Df), 1 fan
diameter upstream (-1Df), and 2 fan diameters upstream (-2Df) of its nominal position.
For all configurations the βe = 0° and βv = 79°. Figure 4.3 presents the acoustic results of
the shielded baseline jet with the -2Df nozzle-shield configuration (blue) compared to the
baseline shielded configuration (red). In the downward direction, the shield effectively
suppresses mid- and high-frequency noise at all polar angles. However at large polar
angles the shield also amplifies low-frequency noise. In the sideline direction, there is a
significant increase of low-frequency noise at low polar angles with mid- and highfrequency noise reduction at large polar angles. The shield alters the directivity of the jet
by increasing OASPL at low polar angles and attenuating OASPL at large polar angles
with cross over at θ = 80°.
The translation of the nozzle upstream improved the
33
downward and sideline EPNL reductions by 0.65 dB and 0.67 dB when compared to the
shielded baseline jet. In comparison to the unshielded baseline jet the ∆EPNL improved
to 1.72 dB and 1.96 dB in the downward and sideline directions, respectively.
Moving the nozzle upstream about its nominal position intensifies the effect of lowfrequency noise amplification and high-frequency noise suppression, while translating the
nozzle downstream lessens it. Amplification for low-frequency noise is believed to result
from sound waves diffracting off the shield trailing edge. Previous experimental work by
Heavens52 on sound diffraction about trailing edges demonstrated that the intensity of
diffracted sound waves may increase or decrease depending on whether the flow in the
vicinity of the trailing edge is unsteady or steady, respectively. In terms of EPNL,
translation of the nozzle upstream yielded higher EPNL reductions. This is in part due to
the higher shielding surface area that acts as a sound barrier between the jet noise and
ground observer.
The HWB shield features an elevator flap with variable elevator angle that may range
from -15° to +15°. Setting the elevator to βe = +15° arranges the shield in the takeoff
configuration. Figure 4.4 presents the acoustic summary in the downward direction of
the -1Df nozzle-shield configuration with βe = +15° (takeoff, blue) compared to -1Df
nozzle-shield configuration with βe = 0° (cruise, red). At large polar angles the angular
position of the elevator has no effect on the SPL spectra. However at low polar angles
the takeoff configuration reduces SPL levels. Compared to the cruise configuration, the
takeoff configuration boosted the EPNL reduction by 0.65 dB.
34
Designs of the HWB concept have considered the utilization of winglets rather than
verticals, such as the X-48B Blended Wing Body. An investigation of sideline noise
shielding by the verticals was conducted to asses its impact. With the BPR10 nozzle
situated at its nominal position and βe = 0°, the verticals were removed from the shield.
Figure 4.5 presents the acoustic results (red) in the sideline direction; they are compared
to the baseline shielded configuration (blue). Without the verticals, the SPL is higher for
all frequencies at all polar angles. The verticals provide a strong reduction in OASPL.
Thus the verticals provide an EPNL reduction of 1.26 dB when compared to the shield
without the verticals.
Considering that relative to the unshielded baseline case the
sideline EPNL reduction for the shielded nozzle with the verticals is 1.29 dB, it is clear
that the verticals are predominantly responsible for the noise attenuation in the sideline
direction.
Acoustic imaging of the baseline jet, based on the deconvoled beamforming method
outlined in Chapter 3, generates a noise source distribution map.
A noise source
distribution map plots SPL contours versus axial position normalized by fan diameter
(x/Df) and versus Strouhal number. Figure 4.6a presents the noise source distribution
map for the baseline jet where x/Df = 0 and x/Df = -1 correspond to the nozzle plug tip
and the fan exit plane, respectively. The peak sound intensities ranges from Sr ~ 1 to Sr
~ 5 and are located approximately 4Df downstream of the nozzle plug.
A more
quantitative assessment of the noise distribution is gained by tracking the peak noise
source location for each frequency. Figure 4.6b plots the corresponding peak noise
source location for the baseline jet. Highlighting that the trailing edge is located at x/Df =
35
1.28 (blue line), the BPR10 nozzle generates a highly extensive noise source region
whose peak intensities extend further than the trailing edge. This is evident since the
location of the peak intensities for all frequencies are situated at roughly 4Df downstream
of the nozzle plug.
Peak intensities for very low frequencies are located further
downstream. From this result it is no surprise as to why the shield only yielded modest
EPNL reductions, even after the BPR10 nozzle was translated 2Df upstream of its
nominal position.
Table 4.1 lists the acoustic results for all the variable shield
configurations investigated.
4.1.3 Variable nozzle configurations with and without shield
Shielding of the baseline BPR10 nozzle is modest due to the extensive noise source
region generated by the BPR10 nozzle. Significantly translating the nozzle upstream
from its nominal position enhances the acoustic benefit yet this poses aerodynamic and
control challenges to the HWD design.
An alternative approach to increasing the
shielding surface area is to compact the jet noise source with nozzle devices. Chevrons,
wedge FFD, and combinations of chevrons and wedge FFD were integrated into the
BPR10 nozzle to alter and restructure the jet noise distribution. Acoustic measurements
of these shielded jets were performed with the shield setup in its nominal position and
configuration.
Acoustic summaries in the downward and sideline directions are
presented for all the nozzle configurations listed in Table 4.2. The shielded quantities
(green) are compared against their respective unshielded values (blue) and the unshielded
baseline jet (red). Additionally, three ∆EPNL are presented. The first is ∆EPNL due to
the nozzle device only (blue). Second is the ∆EPNL due to the combined effect of the
36
shield and nozzle device (green). The final value is the ∆EPNL due to shielding only
(black). For all quantities, a positive value is indicative of noise reduction while a
negative value denotes noise amplification.
Two sets of chevrons were integrated into the BPR10 nozzle, mild chevrons with a
10° insertion angle and aggressive chevrons with a 20° insertion angle. Figure 4.7
presents the acoustic summaries of the free mild chevron jet (blue), shielded mild
chevron jet (green), and unshielded baseline jet (red) in the downward and sideline
directions. Compared to the baseline case, the mild chevrons slightly attenuate SPL at
low frequencies across all polar angles yet amplifies high-frequency noise at low polar
angles, θ = 20° and θ = 30°. Additionally there is an increase in the reduction of OASPL
with increasing polar angle. These effects are seen in both the downward and sideline
directions. The resulting ∆EPNL due to mild chevrons are 0.24 dB and 0.29 dB in the
downward and sideline directions, respectively. The addition of the shield suppresses
mid- and high-frequency noise for θ ≥ 30°. At polar angles θ = 20° and θ = 30° the
shield has no effect. Consequently, the shield and mild chevrons together generated a
greater reduction in OASPL for θ ≥ 30°. The combined effect of shielding and mild
chevrons improved the ∆EPNL to 1.85 dB and 1.72 dB in the downward and sideline
directions, respectively.
The acoustic summaries of the free aggressive chevron jet (blue), shielded aggressive
chevron jet (green), and unshielded baseline jet (red) in the downward and sideline
directions are presented in Figure 4.8. Compared to the baseline case, at low polar angles
the aggressive chevrons suppress low-frequency noise and amplify high-frequency noise
37
with a full-scale crossover frequency ranging between 100 Hz and 200 Hz. At high polar
angles the aggressive chevrons reduce SPL for low frequencies.
There is a strong
OASPL reduction by approximately 3 dB across all polar angles. This is primarily due to
reductions of low-frequency noise. These trends are evident in both the downward and
sideline directions. The resulting ∆EPNL due to aggressive chevrons are 0.38 dB and
0.04 dB in the downward and sideline directions, respectively.
While aggressive
chevrons suppressed low-frequency noise, the addition of the shield suppressed mid- and
high-frequency noise for θ ≥ 30°. Thus the combined effect of the aggressive chevrons
and shielding surfaces reduced SPL across the entire spectrum for large polar angles. As
with the mild chevrons, at low polar angles the shield had no effect. Compared to the
free aggressive chevron jet, the shielded aggressive chevron jet significantly reduced the
OASPL for θ ≥ 60°.
The combined effect of shielding and aggressive chevrons
drastically improved the ∆EPNL to 3.34 dB and 3.15 dB in the downward and sideline
directions, respectively.
The second nozzle device investigated was porous wedge FFD.
A parametric
investigation of the wedge FFD geometry was conducted. Wedge geometry variables
included wedge half-angle αw, height hw, and wedge axial displacement from the fan exit
plane ∆xw. Table 2.1 list the wedge geometry specifications of the wedge FFD tested in
this study.
When compared to the unshielded baseline jet, the best performing wedge was
W18x3. Wedge W18x3 consists of a αw = 18°, hw = 5-mm, and ∆xw = 3-mm. Figure 4.9
presents the acoustic results of the free W18x3 jet (blue), shielded W18x3 jet (green), and
38
unshielded baseline jet (red) with acoustic summaries in the downward and sideline
directions. In the downward and sideline directions, wedge W18x3 substantially reduces
low- and mid-frequency noise, with no crossover at high frequencies. Similarly, there is
a significant OASPL reduction for low- and- mid polar angles. The resulting ∆EPNL
from the W18x3 wedge alone are 1.70 dB and 0.71 dB in the downward and sideline
directions, respectively. Combined with the shield, there is a considerable reduction of
SPL levels across all polar angles. The shield increased SPL reduction and attenuated
OASPL at large polar angles where the W18x3 wedge was ineffective. Together the
shield and W18x3 wedge produced ∆EPNL of 3.86 dB and 3.05 dB in the downward and
sideline directions, respectively.
A more aggressive wedge half-angle didn’t necessarily correspond with higher noise
reduction, as was the case for wedge W30. Figure 4.10 presents the acoustic summaries
of the free W30 jet (blue), shielded W30 jet (green), and unshielded baseline jet (red) in
the downward and sideline directions. Wedge W30 consists of a αw = 30°, hw = 5-mm,
and xw = 0-mm. Just as the W18x3 wedge, the W30 wedge suppresses low- and midfrequency noise at low polar angles. However at large polar angles the W30 wedge
significantly amplifies high-frequency noise for θ ≥ 90° with a full-scale crossover
frequency ranging between 100 Hz and 200 Hz. In the sideline direction, high-frequency
noise amplification begins at θ = 60°. Similar trends are seen in the OASPL directivity
of the jet. In terms of EPNL, the W30 wedge actually amplified the EPNL in the sideline
direction by 1.17 dB. In the downward direction the resulting ∆EPNL is 0.63 dB. The
addition of the shield suppresses the high-frequency noise amplified by wedge W30, but
39
the combined effect is not sufficient enough to fully offset the noise generated by wedge
W30. The resulting ∆EPNL due to the shield and W30 wedge are 3.72 dB and 2.42 dB in
the downward and sideline directions, respectively.
Aggressive means of suppressing jet noise combines the use of both chevrons and
wedge FFD. A combination of the mild chevrons with the W18x3 wedge (MC+W18x3)
and aggressive chevrons with the W18x3 wedge (AC+W18x3) were surveyed. Figure
4.11 presents the acoustic results of the free AC+W18x3 combination jet (blue), shielded
AC+W18x3 combination jet (green), and unshielded baseline jet (red) with acoustic
summaries in the downward and sideline directions.
The AC+W18x3 combination
significantly attenuates low-frequency noise and notably amplifies high-frequency noise
with a crossover frequency ranging from 100 Hz to 300 Hz. High-frequency noise
amplification is more effective in the sideline direction. However there is a strong
OASPL reduction across all polar angles for the down and sideline directions. The
∆EPNL resulting from the AC+W18x3 combination alone is 1.10 dB and 0.04 dB in the
downward and sideline directions, respectively. The integration of the shield attenuates
the amplified high-frequency noise at the large polar angles but does not reduce the highfrequency noise amplification at low polar angles.
OASPL reductions were further
enhanced for all polar angles in both azimuthal directions. The shield and AC+W18x3
combination produce the greatest EPNL reductions of 4.32 dB and 3.30 dB in the
downward and sideline directions, respectively.
Some basic trends were observed from the nozzle modifications utilized in the study.
Both chevrons and wedge FFD were effective at suppressing low- and mid-frequency
40
noise levels.
However, in some cases high-frequency noise was amplified which
penalized their potential EPNL reduction. Chevrons predominantly amplified noise at
low polar angles while the wedge FFD amplified noise at large polar angles. The shield
alone is effective at attenuating high frequency noise for θ ≥ 50°. So the combination of
the shield and nozzle modification yielded greater EPNL reductions than those measured
with the shielded baseline jet. Table 4.2 lists the ∆EPNL for the all the variable jets
tested.
4.1.4 Effect of noise source redistribution on shielding
Modifications to the baseline BPR10 jet in the form of chevrons or wedge FFD altered
the noise source distribution of the jet in ways that benefited jet noise shielding.
Measurements of the noise source distribution were conducted by densely grouping the
microphones on a linear array at polar angles ranging from θ = 47.5° and θ = 73.0° in the
downward direction.
Noise source distribution maps are generated from the
deconvolution of the delay-and-sum beamforming method described in Chapter 3 to
illustrate the jet noise intensity and axial distribution of the jet for a desired frequency
range. Each noise source distribution is accompanied by its corresponding plot of peak
intensity location and is compared to the baseline jet (red).
Additionally, maps of
insertion loss in the downward and sideline directions are presented.
Results are
presented for a select number of jets: mild chevron, aggressive chevron, W18x3, and
AC+W18x3. Results for all other jets may be found in Appendix C.
The noise source distribution map and peak intensity location plot for the mild
chevron jet are presented in Figure 4.12. Compared to the baseline jet, the mild chevrons
41
slightly decrease peak intensity levels but have a modest effect on the noise source
distribution. In quantifying the noise source distribution, the noise source length lns is
defined as the average axial location of the peak noise within the 100Hz to 400 Hz band
(blue lines). Applying this definition to the mild chevron jet (green) and baseline jet
(red), the mild chevrons slightly reduced the noise source length from lns /Df = 4.37 for
the baseline jet to lns /Df = 4.19 for the mild chevron jet. The insertion loss generated by
mild chevrons is presented in Figure 4.13. Like the baseline case, Figure 4.6, the region
of peak insertion loss is at high polar angles for high frequencies yet the mild chevrons
generate a marginally larger level of insertion loss.
Since the mild chevrons were
ineffective at compressing the noise source region, it yielded modest EPNL reductions
similar to those for the baseline case.
Noise source distribution results for the aggressive chevron jets are presented in
Figure 4.14. Recall from the SPL spectra that the aggressive chevrons suppressed lowfrequency noise while amplified high-frequency noise.
From the noise source
distribution maps, this is arguably accomplished by shifting the peak sound intensities to
higher frequencies and simultaneously translating those high-frequency peak intensities
upstream closer to the fan exit plane. Thus the aggressive chevrons distribute jet noise by
positioning the high-frequency peak intensities within the vicinity of the nozzle plug tip
and with decreasing Sr the corresponding peak intensity location is positioned further
downstream. The aggressive chevrons compressed the noise source length to lns /Df =
2.42. In doing so, the high-frequency peak intensities were effectively attenuated by the
shield. The insertion loss maps of the aggressive chevron jet, shown in Figure 4.15,
indicate a drastic increase in the region and level of insertion loss.
42
Wedge FFD were a different nozzle device studied, Wedge W18x3’s effect on the
noise source distribution is presented in Figure 4.16. The W18x3 wedge induces an
overall suppression of the sound intensity rather than shifting the peak intensities to a
different frequency. Furthermore, in contrast to the abrupt transition in the peak noise
source location previously seen with the aggressive chevrons, wedge W18x3 induces a
more gradual trend of reduction in noise source length. The resulting noise source length
is shortened to lns /Df = 3.40, yet this reduction is not as large as with the aggressive
chevrons. Consequently, the insertion loss levels for the W18x3 wedge, shown in Figure
4.17, are much less intense than those obtained with the aggressive chevrons.
It is evident from these results that the aggressive chevrons offer the largest noise
source compaction and resulting insertion loss. However, the aggressive chevrons, by
themselves, are noisy at high frequencies which offset the insertion loss benefit. On the
other hand, even though the insertion loss of the W18x3 wedge is more modest it is
inherently quieter. Therefore, the W18x3 wedge outperforms the aggressive chevrons in
EPNL reductions.
The AC+W18x3 combination offers moderately larger noise source compaction than
the aggressive chevrons alone. And consequently greater levels of insertion loss as
shown in Figures 4.18 and 4.19.
This combination yielded the best noise source
compaction at lns /Df = 1.93.
4.2 Mean Flow Results
Surveys of the jet mean plume were conducted with a fully-automated three-dimensional
Pitot rake traverse. Results of each jet mean flow are presented as contours of the mean
43
axial velocity on the symmetry plane (z = 0) and cross-sectional mean velocity contours
at various axial locations. The axis is centered at the nozzle plug and the positive xdirection is the downstream direction. Results for a select number of jets are presented:
baseline, mild chevron, aggressive chevron, W18x3, and AC+W18x3. Velocity contours
along the symmetry plane for all other jets may be found in Appendix C.
The mean velocity results for the baseline jet are presented in Figure 4.20. The
velocity contours show that the baseline jet is symmetric about the x-axis which is
expected as there were no modifications to the nozzle. The jet symmetry is also seen in
cross-sectional cuts of the axial mean velocity which indicate good alignment of the
nozzle components. The high-speed region where u>0.9Up is taken as the measure of the
primary potential core. Applying this metric to the baseline jet, the potential core extends
to lpc /Df = 3.53.
Figure 4.21 presents the mean velocity contours along the symmetry plane for the
mild chevron, aggressive chevron, W18x3, and AC+W18x3 jets. The mild chevrons
cause a counter-intuitive effect of elongating the high-speed region to lpc /Df =4.04. The
aggressive chevrons reduce the volume of the high-speed region, but like the mild
chevrons they counter-intuitively elongate the high-speed region to lpc /Df =3.65. It is
hypothesized that the formation of the pair of streamwise counter-rotating vortices by
chevrons stabilize the shear flow and dampen the radial growth rate of the jet which
ultimately extended the primary potential core. The integration of wedge FFD into the
BPR10 nozzle imparts asymmetry to the flow field. The W18x3 wedge reshapes the flow
by deflecting and thickening the secondary flow in the downward (negative y) direction.
Ultimately this effect compressed the high-speed region to lpc /Df =2.92. The AC+W18x3
44
combination consisted of features from both the wedge and aggressive chevrons. The jet
is asymmetric and the volume of the high speed region is compressed. The combined
effect lead to further contraction of the high-speed region to lpc /Df =2.60.
Cross-sectional cuts of the axial mean velocity contours for the mild chevron,
aggressive chevron, W18x3, and AC+W18x3 jets are presented in Figure 4.22. Both the
mild and aggressive chevrons impact the mean velocity pattern only for x/Df <1; the
chevrons induce a “sun-pattern” velocity profile. Downstream, turbulent mixing quickly
dissipates the profile to the circular profile as in the baseline case. The W18x3 wedge
generates an ovular velocity profile that is fatter at the base by concentrating the fan flow
in the downward direction. This results from the downward deflection of the fan flow.
Near the vicinity of the nozzle plug tip plane (x/Df =0) the wedge wake is evident on the
top region of the jet.
Further downstream the wake dissipates yet the overall jet flow
remains deflected and its effect persist longer than those of the chevrons.
The
AC+W18x3 combination features characteristics from both the aggressive chevrons and
the W18x3 wedge. The secondary flow is deflected downward with the sun-pattern
velocity profile and wedge wake are superimposed in the near-field of the jet, x/Df <1.
The axial distribution of the local maximum velocity umax(x) for the mild chevron,
aggressive chevron, W18x3, and AC+W18x3 jets are plotted in Figure 4.23. Both the
mild and aggressive chevrons moderately elongated the high-speed region. In the jet
near-field both chevrons induced a small reduction in umax but reduce the velocity decay
for x/Df >3.5. The mild chevrons suppress the velocity decay even sooner at x/Df =2.5.
In contrast, the W18x3 wedge accelerates the initiation of the velocity decay without
significantly altering the downstream decay rate itself. This contributes to significant
45
contraction of the high-speed region, as previously seen. The AC+W18x3 combination
initiates the velocity decay even sooner and like the wedge has minimal effect on the
decay rate. The AC+W18x3 combination ultimately results in the highest compaction of
the high-speed region.
46
Chapter 5
CORRELATIONS
5.1 Formulation of the average light angle
Jet noise shielding is an extremely complex problem and is thus not amenable to simple
analytical solutions. Accurate numerical predictions of the acoustic field require models
of the jet noise source and of its diffraction about the shielding surface. The resulting
computations are expensive and time-consuming, especially for high frequencies. On the
other hand, it would be useful to have an estimate of the shielding benefit during
preliminary designs without having to resort to lengthy computations. A simple model
based on the illumination angle is presented for preliminary estimates of EPNL
reductions obtained from a shielding surface.
The illumination angle ψ is defined as the angular sector between noise source and
observer not intercepted by the shield, as shown in Figure 5.1a. Given that the jet noise
source is distributed, we define the average illumination angle ψ̄ as follows:
l
1 ns
ψ = ∫ψ dx
lns 0
47
(5.1)
where lns is the previously defined noise source length. For a given azimuth angle φ = φo,
the illumination angle for source position x depends on the location of the shield trailing
edge, relative to the position of the source, as determined by the intersection of the
trailing edge by the plane φ = φo passing through the jet axis. This is illustrated in Figure
5.1b. Denoting the intersection point (Xs, Ys, Zs), the average illumination angle is
1
ψ =
lnc
2
2


−1  Ys (φ ) + Z s (φ ) 
tan
∫0  X s (φ ) − x dx


l nc
(5.2)
This general formulation may include not only the planform trailing edge but also the
trailing edge of the verticals. Clearly the formula assumes a large upstream extent of the
shield, which may break down in reality, particularly for the verticals.
Equation 5.2 was applied to all the HWB nozzle-shield configurations listed in Table
5.1 in the downward (φ = 0°) and sideline (φ = 60°) planes. The verticals were included
in the sideline calculation. The resulting average illumination angle was used as a
correlator for the reduction in EPNL due to shielding alone.
So for each nozzle
configuration (baseline, wedge, chevrons, etc), the calculated EPNL reduction is only due
to insertion of the shield and that value is plotted versus its corresponding ψ̄ . This
∆EPNL value is represented in black in the acoustic summaries of the shielded jets with
nozzle modifications. Figure 5.2 plots the resulting trend. The plot includes all nozzleshield configurations and nozzle modifications plus the nozzle shield configurations from
Papamoschou and Mayoral36. Although there is considerable scatter in the plot, a definite
trend is shown of increasing EPNL reduction with decreasing illumination angle. As a
rule of thumb, an EPNL reduction of 3 dB may be expected if ψ̄ ~ 45°, and an EPNL
48
reduction of as much as 6 dB may be possible by reducing the average illumination angle
to about 20°. It is important to keep in mind that the EPNL was based on a thrust of
70,000 lb. The trends for much smaller or much larger engines may be different.
5.2 Acoustic-mean flow correlation
An objective of the study was to examine potential relations between the distortion of the
jet plume and corresponding noise distribution. From the acquired data, an effort to link
the compaction of the noise source region to the contraction of the high-speed region was
inconclusive. The wedge FFD were successful at compacting both the noise source and
high-speed region, Figure 4.16 and Figure 4.21. However the same is not true for the
chevrons. For the aggressive chevrons there is no straight-forward way to connect the
mean velocity field. While potential core length hardly changed the noise source length
was dramatically reduced, especially at high frequencies, Figure 4.14. Figure 5.3 plots
the noise source length versus the primary potential core length. A linear correlation is
observed between the noise source length and potential core length except for two
distinct points which correspond to the chevron nozzles. Thus it is concluded that for
these complex flows, the mean velocity field alone does not provide enough information
to infer a noise source length.
49
Chapter 6
CONCLUSIONS
6.1 Summary
This study conducted a parametric subscale experimental investigation of shielding of jet
noise. The nozzle-shield configuration was composed of a high-bypass nozzle and a
shield shaped in the form of a generic Hybrid Wing Body aircraft. The nozzle operated
at realistic cycle conditions using helium-air mixture jets. Chevrons, porous wedge fan
flow deflectors, and combinations of both were incorporated into the baseline nozzle to
restructure and alter the jet noise source distribution.
Acoustic surveys of free and shielded jets were conducted inside an anechoic
chamber using a 24-microphone apparatus composed of two polar arrays each consisting
of twelve condenser microphones.
One polar array was mounted in the downward
direction while the second array was mounted in the sideline direction. Noise source
distribution measurements were performed by densely grouping twenty-four microphones
on a linear array in the downward direction.
50
Surveys of the jet mean flow were
conducted with a fully automated three-dimensional Pitot rake traverse. The Pitot rake
consists of five high-resolution Pitot probes that measure the local jet total pressure.
Acoustic results of the baseline BPR10 nozzle with the HWB shield yielded only
marginal EPNL reductions, even with the engine translated two fan diameters upstream
about its nominal location. Using the estimated cumulative (downward plus sideline)
EPNL reduction as a figure of merit, shielding of the baseline nozzle with the nominal
shield configuration yielded a 2.5 dB reduction.
Acoustic surveys in the sideline
direction indicated that the vertical fins are the principal shielding surfaces in that
direction. Acoustic imaging of the baseline jet noise source revealed that the large fan
diameter inherent to the high bypass ratio creates a long noise source region that is
sufficiently longer than the extent of the shielding surface.
The integration of nozzle devices significantly improved EPNL reduction with the
engine positioned at its nominal location.
Application of the aggressive chevrons
increased the reduction to 6.5 dB, while the best wedge configuration improved this
figure to 6.9 dB. Combination of wedge and aggressive chevrons yielded a benefit of 7.6
dB. Examination of high-definition noise source maps shows a direct link between
insertion loss and axial location of the peak noise. There are differences in the ways the
nozzle devices change the location of the peak noise source.
The aggressive chevrons
cause an abrupt contraction at Strouhal number 1.2, while the wedge induces a gradual
contraction with increasing frequency. As a result, the insertion loss with the aggressive
chevrons is stronger than with the wedge. However, because the wedge is inherently
quieter than the chevrons, it gives a slightly better overall benefit. Ultimately, the
enhanced shielding effect from the nozzle devices resulted from contracting the noise
51
source region. Thus the compaction and/or redistribution of the noise source via nozzle
devices are essential for effective jet noise shielding on the HWB.
Surveys of the mean flow field show that the wedge, and its combination with
chevrons, produces a significant reduction in the potential core length.
On the other
hand, the chevrons alone induce modest changes in the length of the high-speed region of
the jet. Therefore, the mean velocity field by itself cannot provide useful information for
inferring the noise source location for these complicated flows.
A proposed correlator for the noise reduction due to shielding is the average
illumination angle. The EPNL reductions due to shielding were plotted against this
correlator for all the cases considered in this study. A clear trend emerges of increasing
EPNL reduction with decreasing average illumination angle. For average illumination
angles less than 45 deg, the EPNL reduction exceeds 3 dB.
6.2 Recommendations for future work
Limitations of the experimental facilities inhibit the study from addressing some critical
factors that engineers face in the actual design of an HWB aircraft.
First, the issue of the associated trust loss for a given nozzle modification is an
important design factor that needs to be taken into account in order to truly determine an
optimal nozzle configuration. Experimental measurements for porous wedge fan flow
deflectors similar to those conducted for porous wedges may be performed to obtain an
estimate of the associated thrust loss. Alternatively, the use of computational fluid
dynamics could highly contribute to the study by computing the thrust loss for all the jets
investigated.
52
Furthermore, CFD methods may be utilized to also obtain turbulent statisitcs that are
beyond the capabilities of the current experimental facilities. One such variable is the
turbulent kinetic energy of the jet which could serve as a possible correlator between the
jet noise distribution and jet mean flow.
Lastly, forward flight effects experienced by aircraft jets were not addressed in this
study. To assess the impact of forward flight, a completely different jet aeroacoustic
facility is required to simulate the effect of the aircraft motion relative to the ambient.
53
Bibliography
[1] Smith, M. J. T., “Aircraft Noise,” 1st Edition, Cambridge University Press,
Cambridge, 1989.
[2] Thompson, E., Barnstorff, K., and Olsen, M. J., “NASA and Industry Team Tests
Aircraft Noise-Reducing Technologies,” NASA News, August 10, 2005.
[3] Gerhold, C. H., and Clark, L. R., “Database of Inlet and Exhaust Noise Shielding for
Wedge-Shaped Airframe,” NASA TM-2001-210840, April 2001.
[4] Russell, J., and Berton, J., “Stone Jet Noise Module (ST2JET)”, ANOPP Theoretical
Manual, ver.25, NASA Langley Research Center, Hampton, VA, 2006.
[5] Maekawa, Z., “Noise Reduction by Screens,” Applied Acoustics, Vol.1, 1968, pp.
157–173.
[6] Huang, C., and Papamoschou, D., “Numerical Study of Noise Shielding by Airframe
Surfaces," AIAA-2008-2999, May 2008.
54
[7] Federal Aviation Regulations, Part 36, Noise Standards: Aircraft Type and
Airworthiness Certification. Federal Aviation Administration, January 2002.
[8] Liebeck, Robert H., “Design of the Blended Wing Body Subsonic Transport,”
Journal of Aircraft, Vol. 41, No. 1, January–February 2004.
[9] Tam, C. K. W., Golebiowski, M., and Seiner, J. M., “On the Two Components of
Turbulent Mixing Noise from Supersonic Jets,” AIAA Paper 96-1716, 1996.
[10] Viswanathan, K., “Analysis of the two similarity components of turbulent mixing
noise,” AIAA Journal, Vol. 40, 2002, pp. 1735–1744.
[11] Tam, C. K. W., Viswanathan, K., Ahuja, K. K., and Panda, J., “The sources of jet
noise: experimental evidence,” AIAA Paper 2007-3641, 2007.
[12] Brown, G.L. and Roshko, A. “On denisty effects and large structure in turbulent
mixing layers.” Journal of Fluid Mechanics, 64(4):775–816, 1974.
[13] McLaughlin, D. K., Morrison, G. L., and Troutt, R. R., “Experiments on the
instability waves in a supersonic jet and their acoustic radiation,” Journal of Fluid
Mechanics, Vol. 69, 1975, pp. 73–95.
[14] Troutt, T.R., and McLaughlin, D.K., “Experiments on the Flow and Acoustic
properties of a Moderate Reynolds Number Supersonic Jet,” Journal of Fluid
Mechanics, Vol. 116, March 1982, pp. 123-156.
55
[15] Tam, C. K. W. and Burton, D. E., “Sound generation by the instability waves of
supersonic flows. Part 1. Axisymmetric jets.” Journal of Fluid Mechanics, Vol. 138,
1984.
[16] Tam, C. K. W. and Burton, D. E., “Sound generation by the instability waves of
supersonic flows. Part 2. Axisymmetric jets.” Journal of Fluid Mechanics, Vol. 138,
1984, pp. 273–295.
[17] Crighton, D.G. and Huerre, P., “Shear-Layer Pressure Fluctuations and
Superdirective Acoustic Sources,” Journal of Fluid Mechanics, Vol. 220, 1990, pp.
355-368.
[18] Tam, C. K. W., “Supersonic Jet Noise,” Annual Review of Fluid Mechanics, 1995.
27: 17-43
[19] Papamoschou, D., “New Method for Jet Noise Suppression in Turbofan Engines,”
AIAA Journal, Vol. 42, No.11, 2004, pp. 2245-2253.
[20] Samimy, M., Zaman, K. B. M. Q., and Reeder, M. F., “Effect of Tabs on the Flow
and Noise Field of an Axisymmetric Jet,” AIAA Journal, Vol. 31, No. 4, April 1993,
pp. 609–619.
[21] Zaman, K. B. M. Q., Reeder, M. F., and Samimy, M., “Control of an Axisymmetric
Jet Using Vortex Generators,” Physics of Fluids, Vol. 6, No. 2, 1994, pp. 778–793.
56
[22] Callender, B., Gutmark, E., and Martens, S., “Far-Field Acoustic Investigation into
Chevron Nozzle Mechanisms and Trends,” AIAA Journal, Vol. 43, No. 1, January
2005, pp. 87–95.
[23] Saiyed, N. H., Mikkelsen, K. L., and Bridges, J. E., “Acoustics and Thrust of Quiet
Separate-Flow High-Bypass-Ratio Nozzles,” AIAA Journal, Vol. 41, No. 3, March
2003, pp. 372–378.
[24] D. Papamoschou and M. Debiasi, “Directional suppression of noise from a highspeed jet”, AIAA Journal, Vol. 39, No.3, 2001, pp. 380-387.
[25] Zaman, K.B.M.Q., 2004, “Noise- and flow-field of jets from an eccentric coannular
nozzle”, AIAA 2004-0005, January 2004.
[26] Papamoschou, D., “Mean Flow and Acoustics of Dual-Stream Jets,” AIAA-20040004, January 2004
[27] Papamoschou, D., “Fan Flow Deflection in Simulated Turbofan Exhaust,” AIAA
Journal, Vol. 44, No. 12, December 2006.
[28] Mayoral, S., and Papamoschou, D., “Effects of Source Redistribution on Jet Noise
Shielding,” AIAA-2010-0652,
48th AIAA Aerospace Sciences Meeting and
Exhibit, Jan.4-7, 2010.
[29] Papamoschou, D., “Pylon Based Jet Noise Suppressors,” AIAA 2008-0040, January
2008.
57
[30] Papamoschou, D., Vu, A., and Johnson, A., “Aerodynamics of Wedge-Shaped
Deflectors for Jet Noise Reduction,” AIAA 2006-3655, June 2006.
[31] Von Glahn, U., Groesbeck, D., and Reshotko, M., “Geometry Considerations for Jet
Noise Shielding with CTOL Engine-Over-The-Wing Concept,” AIAA Paper 74-568,
June 1974.
[32] Reshotko, M., Goodykoontz, J. H., and Dorsch, R. G., “Engine-Over-The-Wing
Noise Research,” NASA TM-X-68246, July 1973.
[33] Von Glahn, U., Goodykoontz, J., and Wagner, J., “Nozzle geometry and forward
velocity effects on noise for CTOL engine-over-the-wing concept,” NASA TM-X71453, October 1973.
[34] Von Glahn, U., and Groesbeck, D., “Influence of Mixer Nozzle Velocity Decay
Characteristics on CTOL-OTW Jet Noise Shielding,” NASA TM-X-71631, January
1975.
[35] Von Glahn, U., Groesbeck, and D. Wagner, J., “Wing shielding of high-velocity jet
and shock-associated noise with cold and hot flow jets,” AIAA Paper 76-547, July
1976.
[36] Papamoschou, D., and Mayoral, S., “Experiments on Shielding of Jet Noise by
Airframe Surfaces,” AIAA-2009-3326, May 2009.
58
[37] Lighthill, M. J., “On sound generated aerodynamically: I. General theory,”
Proceedings of the Royal Society of London, Vol. A 211, 1952, pp. 564–587.
[38] Lighthill, M. J., “On Sound Generated Aerodynamically. II. Turbulence as a Source
of Sound,” Proceedings of the Royal Society of London, Vol. A 222, 1954, pp. 1–32.
[39] Papamoschou, D., “Acoustic Simulation of Hot Coaxial Jets Using Cold Helium-Air
Mixture Jets,” AIAA-2005-0208, AIAA 43d Aerospace Sciences Meeting, January
2005, Reno, NV.
[40] Zaman, K., Bridges, J., and Papamoschou, D., “Offset Stream Technology Comparison of Results from UCI and GRC,” AIAA-2007-0438, 45th AIAA
Aerospace Sciences Meeting and Exhibit, Jan.8-11, 2007, Reno, NV.
[41] Forstall, W., Jr., and Shapiro, A. H., “Momentum and Mass Transfer in Coaxial Gas
Jets,” Journal of Applied Mechanics, Vol. 10, 1950, pp. 399–408.
[42] Savitzky, A. and Golay, M.J.E. “Smoothing and Differentiation of Data by
Simplified Least Squares Procedures,” Analytical Chemistry, Vol. 36, No. 8, 1964,
pp. 1627-.
[43] Mueller, T. J., “Aeroacoustic Measurements,” 1st Edition, Springer-Verlag, Berlin,
2002.
59
[44] McGowan, R.S. and Larson, R.S., “Relationship Between Static, Flight and
Simulated Flight Jet Noise Measurements,” AIAA Journal, Vol. 22, No. 4, 1984, pp.
460-464.
[45] Murakami, E. and Papamoschou, D., “Eddy Convection in Supersonic Coaxial Jets,”
AIAA Journal, Vol. 38, No.4, 2000, pp. 628-635.
[46] Bass, H.E., Sutherland, L.C., Zuckerwar, A.J., Blackstock, D.T., and Hester, D.M.,
“Atmospheric Absorption of Sound: Further Developments,” Journal of the
Acoustical Society of America, Vol. 97, 1995, pp. 680-683
[47] Society of Automotive Engineers Aerospace A21 committee. Prediction of Lateral
Attenuation of airplane noise during takeoff and landing. Information Report - SAE
AIR-1751. March 1981.
[48] Papamoschou, D., “Imaging of Distributed Directional Noise Sources,” AIAA-20082885, May 2008.
[49] Brooks, T.F., and Humphreys, W.M., “A Deconvolution Approach for the Mapping
of Acoustic Sources (DAMAS) Determined from Phased Microphone Arrays,”
AIAA-2004-2954, May 2004.
[50] Richardson,W.H., “Bayesian-Based Iterative Method of Image Restoration,” Journal
of the Optical Society of America, Vol. 62, No. 1, 1972, pp. 55-.
[51] Lucy, L.B., “Iterative Technique for Rectification of Observed Distribution,”
Astronomical Journal, Vol. 79, No. 6, 1964, pp. 745-754.
60
[52] Heavens, S. N., “An experimental study of sound diffraction at an airfoil trailing
edge,” J . Fluid Mech., vol. 84, part 2, 1978, pp. 331-335.
61
Appendix A
Tables
Table 2.1: Geometry of wedge fan flow deflectors (FFD) utilized in the study.
Wedge FFD
αw (°)
W15
W18
W21
W30
TW15
TW18
W18x3
TW18x3
15
18
21
30
15
18
18
18
hw (mm) ∆xw (mm)
5
5
5
5
7
7
5
7
0
0
0
0
0
0
3
3
All wedges consisted of a solidity of 49.6% and side length of 13-mm. Fan exit height was 4-mm.
Table 2.2: BPR10 takeoff engine cycle conditions for acoustic and mean flow surveys.
Quantity
Core
Fan
Nozzle diameter (mm)
Plug diameter (mm)
Height of exit annulus (mm)
Lip thickness (mm)
Hot
NPR
NTR
To (K)*
T (K)*
Velocity
Mach number
Bypass ratio
Cold
Velocity
Mach number
Bypass ratio
15.32
12.24
1.54
0.47
31.2
4.15
0.71
1.376
2.95
864
781
387
0.691
-
1.55
1.139
334
291
279
0.817
9.3
286
0.9
-
206
0.625
3.54
Hot: Acoustic surveys
Cold: Mean flow surveys
*Equivalent conditions using helium-air jets
62
Table 2.3: Microphones with corresponding array position and sensitivity for acoustic
surveys.
Sensitivity
Microphone
R m / Df
φ (°)
θ (°)
(V/Pa)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
0
0
0
0
0
0
0
0
0
0
0
0
60
60
60
60
60
60
60
60
60
60
60
60
29.58
30.84
32.17
36.4
39.75
35.12
32.29
30.95
30.61
31.26
32.9
34.99
29.7
30.89
32.22
36.45
39.96
35.35
32.53
31.19
30.85
31.48
33.11
34.98
19.82
25.52
30.10
40.13
50.35
60.63
71.39
81.44
91.28
101.78
111.54
119.00
20.21
25.54
30.11
40.12
50.51
60.75
71.45
81.42
91.19
101.61
111.32
118.17
63
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.3163
0.3163
0.3163
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.3163
0.3163
0.3163
Table 2.4: Microphone array position for noise source measurements.
Microphone
R m / Df
θ (°)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
41.77
41.23
40.69
40.16
39.65
39.14
38.64
38.16
37.68
37.22
36.77
36.33
35.91
35.49
35.10
34.71
34.35
34.00
33.66
33.34
33.04
32.75
32.49
32.24
47.53
48.37
49.23
50.11
51.01
51.94
52.89
53.86
54.86
55.89
56.94
58.01
59.12
60.24
61.40
62.58
63.78
65.02
66.27
67.55
68.86
70.19
71.54
72.91
All microphones were mounted in the downward (φ = 0°) direction with a sensitivity of 0.1-V/Pa.
64
Table 3.1: Conditions for PNL evaluations.
Parameter
Downward
Sideline
0
1500
10
4
0.3
1476
1000
12
10
0.2
Lateral distance, zo
Altitude at x = 0, (ft)
Angle of attack (°), αa
Climb angle (°), γa
Flight Mach number, M
Table 4.1: Acoustic results for variable shield geometry.
Shield
∆x/Df
βe (°)
βv (°)
∆EPNL (dB)
∆EPNL (dB)
downward
sideline
S00
S01
S02
S03
S04
S05
S06
0
+1
-1
-1
-2
0
0
0
0
0
0
0
0
15
79
79
90
79
79
off
79
1.07
1.30
1.33
1.35
1.74
1.07
2.00
1.29
0.85
1.13
0.65
1.20
0.03
1.00
Sxx: Shield configuration
Downward (φ = 0°)
Sideline (φ = 60°)
65
Table 4.2: Acoustic and mean flow results for variable nozzle modifications.
nozzle only
Nozzle
N00
N01
N02
N03
N04
N05
N06
N07
N08
N09
N10
N11
N12
Device
BB
MC
AC
W15
W18
W21
W30
TW15
TW18
W18x3
TW18x3
MC+W18x3
AC+W18x3
shield + nozzle
∆EPNL
∆EPNL
∆EPNL
∆EPNL
downward
sideline
downward
sideline
0.24
0.38
1.41
1.59
1.70
0.63
1.75
1.93
1.70
1.96
1.59
1.10
0.29
0.04
0.67
0.13
0.30
-1.17
0.37
0.09
0.71
0.98
0.28
0.04
1.07
1.85
3.34
3.13
3.80
3.74
3.72
3.47
3.84
3.86
3.93
3.86
4.32
1.29
1.72
3.15
2.14
2.30
1.98
2.42
1.80
1.65
3.05
2.31
2.50
3.30
Nxx: Nozzle configuration
BB: Baseline
MC: Mild chevrons
AC: Aggressive chevrons
Wxx and TWxx: perforated wedge FFD with half-angle xx-degrees (Table 2.1)
All cases utilized the nominal (S00) shield configuration.
Downward (φ = 0°)
Sideline (φ = 60°)
66
lpc / Df
lns / Df
4.37
4.19
2.42
3.49
2.91
3.10
2.30
3.40
3.51
3.40
3.77
2.80
1.93
3.53
4.04
3.65
3.10
2.87
2.97
2.70
2.84
3.07
2.92
3.22
2.92
2.60
Table 5.1: Average illumination angles and corresponding ∆EPNL (dB) from shielding
for all nozzle-shield configurations investigated.
Case
ψd (°)
ψs (°)
∆EPNLd
∆EPNLs
N00S00
N00S01
N00S02
N00S03
N00S04
N00S05
N00S06
N01S00
N02S00
N03S00
N04S00
N05S00
N06S00
N07S00
N08S00
N09S00
N10S00
N11S00
N12S00
113.29
137.84
84.86
84.86
57.21
113.29
116.18
109.88
113.52
104.71
87.15
94.93
67.37
97.46
106.61
99.19
111.23
88.24
80.33
119.84
139.11
95.39
95.64
70.56
129.88
119.84
117.17
120.01
113.20
100.10
105.85
85.11
107.73
114.66
109.04
118.23
100.90
95.05
1.07
1.30
1.33
1.35
1.74
1.07
2.00
1.61
2.96
1.72
2.21
2.04
3.09
1.72
1.91
2.16
1.97
2.27
3.22
1.29
0.85
1.13
0.65
1.20
0.03
1.00
1.43
3.11
1.47
2.17
1.68
3.59
1.43
1.56
2.34
1.33
2.22
3.26
NxxSxx: Nozzle-shield configuration (Tables 4.1 and 4.2)
67
Appendix B
Figures
Figure 1.1: Three dimensional illustration of the Hybrid-Wing-Body aircraft (HWB).
10
FSS
LSS
SPL (dB)
0
-10
-20
-30
-1
10
10
0
10
1
f / fpeak
Figure 1.2:
Fine-scale similarity (FSS) and large-scale similarity (LSS) similarity
spectra from Tam et al. (1996).
68
Fine-scale turbulence noise
Large-scale structures noise
Figure 1.3: Turbulent mixing noise mechanisms in a jet, fine-scale turbulence and
turbulent large-scale structures, Tam et al. (1996).
U1, ρ1, a1, γ1
Uc
U2, ρ2, a2, γ2
Figure 1.4: Diagram of compressible free shear layer.
69
Mach waves
Uc
Mc > 1
Figure 1.5:
Mach wave radiation by a wavy surface propagating with supersonic
convective velocity, Uc.
r (mm)
30
20
10
0
0
10
20
30
40
x (mm)
50
60
70
(a)
80
(b)
Figure 2.1: Baseline BPR10 nozzle used in subscale experiment. (a) Radial nozzle
coordinates; (b) picture.
70
Side View
Top View
4346.5 cm
48.29 cm
6492.2 cm
71.12 cm
4.95 cm
Nozzle fan diameter, Df
Df = 279.4 cm
Df = 3.12 cm
Scale factor ~ 90
Full-scale
UCI-scale
Figure 2.2: Scaling of HWB planform to UCI dimensions and retention of critical
dimensions for shielding (red lines).
(a)
Figure 2.3:
(b)
(a) Design of UCI HWB shield model with longitudinal traverse; b)
installation with nozzle.
71
(a)
(b)
(c)
(d)
(e)
(f)
Figure 2.4: Subset of the nozzle configurations tested. (a) Mild chevrons; (b) aggressive
chevrons; (c) W18x3 wedge; (d) W18 wedge; (e) MC+W18x3 combination; and (f)
AC+W18x3 combination.
72
Helium-Air
Mixtures
PC with three
PCI-6143
DAQ boards
Anechoic Chamber 1.9 x 2.2 x 2.2 m
BPR10 Jet Nozzle
HWB Shield Surface
24 B&K-4138 Microphones
(Simultaneous Acquisition)
Six Nexus 2690-A-OS4
Conditioning Amplifiers
Figure 2.5: UC Irvine Jet Aeroacoustics Facility.
Compressed
Air 1 MPa
Secondary flow
Metering Valves
Primary flow
Helium from
cylinders
3 – 17 MPa
BPR10 Nozzle
Pressure Solenoid valves
regulator
Figure 2.6: Diagram of dual-stream flow facility.
73
θ
R=1 m
θ =73.0°
θ=47.5°
Figure 2.7: Microphone setup for noise source imaging, utilized 24 densely spaced
microphones.
y
x
z
Figure 2.8: Schematic of Pitot rake and traverse path for mean flow surveys.
74
Aircraft flying with Mach number M, angle of attack α, and climb angle γ
RETARDED
POSITION
y
Observation distance and polar emission angle
ACTUAL
POSITION
r ' = ( x '− xo ) 2 + y '2 + zo
(x’,y’,0)
r = ( x − xo ) 2 + y 2 + zo
(x,y,0)
Mr
b
γ +α
r
tan(θ / 2) =
x
a
z
2
( p − b)( p − r )
p( p − a)
where
x1 = x − y / tan(γ + α )
a = ( xo − x1 ) 2 + z o
OBSERVATION
POINT
x1
Figure 3.1:
θ
r’
2
2
b = y / sin(γ + α )
(0,0, zo)
p=
1
(a + b + r )
2
Geometric relations for assessment of aircraft perceived noise level,
Reference [28].
75
HWM257D
90
θ = 19.9
o
90
90
θ = 60.3
o
90
80
80
80
70
70
70
70
70
60
60
60
60
60
50
50
50
0.1 0.2 0.5 1 0.02
50
0.1 0.2 0.5 1 0.02
0.1 0.2 0.5 1 0.02
ffull (kHz)
PNL(dB)
125
120
115
110
105
ffull (kHz)
90
80
70
60
-20
20 40 60 80 100 120
-10
θ (deg)
0
90
θ = 20.5
90
θ = 30.2
90
∆EPNL=1.07 dB
70
50
20
θ, deg
θ = 60.3
o
90
θ = 90.3
o
90
80
80
80
80
80
70
70
70
70
70
60
60
60
60
60
50
0.02
50
0.1 0.2 0.5 1 0.02
ffull (kHz)
50
0.1 0.2 0.5 1 0.02
0.1 0.2 0.5 1 0.02
ffull (kHz)
PNL(dB)
125
120
115
110
105
ffull (kHz)
90
PNL(dB)
50
80
70
60
20 40 60 80 100 120
θ (deg)
-10
0.10.2 0.5 1
80
Time (sec)
o
o
EPNL=91.77 dB
EPNL=90.7 dB
HWM257S
o
θ = 118.5
ffull (kHz)
90
60
140 110 80
10
HWM237D
50
0.1 0.2 0.5 1 0.02
ffull (kHz)
Sideline
SPL(dB/Hz)
90
80
ffull (kHz)
OASPL(dB)
θ = 90.7
o
80
0.02
OASPL(dB)
θ = 30.1
o
PNL(dB)
SPL(dB/Hz)
Downward
0
10
Time (sec)
HWM237S
θ = 117.2
o
50
0.1 0.2 0.5 1 0.02
ffull (kHz)
0.10.2 0.5 1
ffull (kHz)
EPNL=91.46 dB
EPNL=90.17 dB
90
80
∆EPNL=1.29 dB
70
60
140 110 80
50
20
θ, deg
Figure 4.1: Effect of shield on acoustics in the downward and sideline directions.
Shielded plain nozzle (blue) compared to unshielded baseline nozzle (red).
76
10
Differential SPL map HWM257D-HWM237D
10
4
5
3
2
Sr
Sr
5
3
2
6
1
2
0.5
0.3
0.2
0
6
4
1
2
0.5
0.3
0.2
0.1
0.05
20
Differential SPL map HWM257S-HWM237S
0
0.1
40
60
80
100
-2
θ (deg)
0.05
20
40
60
80
100
-2
θ (deg)
(a)
(b)
Figure 4.2: Insertion loss maps of baseline jet in the (a) downward and (b) sideline
directions.
77
HWM237D
90
θ = 19.9
o
90
90
θ = 60.2
o
90
80
80
80
70
70
70
70
70
60
60
60
60
60
50
50
50
0.1 0.2 0.5 1 0.02
50
0.1 0.2 0.5 1 0.02
0.1 0.2 0.5 1 0.02
ffull (kHz)
PNL(dB)
125
120
115
110
105
ffull (kHz)
90
80
70
60
-20
20 40 60 80 100 120
-10
θ (deg)
0
90
θ = 20.4
90
θ = 30.0
90
∆EPNL=0.67 dB
70
50
20
θ, deg
θ = 60.0
o
90
θ = 89.7
o
90
80
80
80
80
80
70
70
70
70
70
60
60
60
60
60
50
0.02
50
0.1 0.2 0.5 1 0.02
ffull (kHz)
50
0.1 0.2 0.5 1 0.02
0.1 0.2 0.5 1 0.02
ffull (kHz)
PNL(dB)
125
120
115
110
105
ffull (kHz)
90
PNL(dB)
50
80
70
60
20 40 60 80 100 120
θ (deg)
-10
0.10.2 0.5 1
80
Time (sec)
o
o
EPNL=90.7 dB
EPNL=90.03 dB
HWM237S
o
θ = 118.3
ffull (kHz)
90
60
140 110 80
10
HWM040D
50
0.1 0.2 0.5 1 0.02
ffull (kHz)
Sideline
SPL(dB/Hz)
90
80
ffull (kHz)
OASPL(dB)
θ = 90.6
o
80
0.02
OASPL(dB)
θ = 30.1
o
PNL(dB)
SPL(dB/Hz)
Downward
0
10
Time (sec)
HWM042S
θ = 116.7
o
50
0.1 0.2 0.5 1 0.02
ffull (kHz)
0.10.2 0.5 1
ffull (kHz)
EPNL=90.17 dB
EPNL=90.26 dB
90
80
∆EPNL=-0.09 dB
70
60
140 110 80
50
20
θ, deg
Figure 4.3: Effect of shield on acoustics in the downward and sideline directions with
the -2Df nozzle-shield configuration (blue) compared to baseline nozzle-shield
configuration (red).
78
HWM030D
90
θ = 20.0
o
90
θ = 30.0
90
θ = 60.0
o
90
θ = 91.0
o
80
80
80
80
70
70
70
70
70
60
60
60
60
60
50
50
50
0.1 0.2 0.5 1 0.02
50
0.1 0.2 0.5 1 0.02
ffull (kHz)
0.1 0.2 0.5 1 0.02
ffull (kHz)
PNL(dB)
125
120
115
110
105
ffull (kHz)
90
80
70
60
-20
20 40 60 80 100 120
-10
θ (deg)
0
HWM060D
90
80
0.02
OASPL(dB)
o
θ = 118.0
o
50
0.1 0.2 0.5 1 0.02
ffull (kHz)
0.10.2 0.5 1
ffull (kHz)
EPNL=90.42 dB
EPNL=89.77 dB
90
PNL(dB)
SPL(dB/Hz)
Downward
80
∆EPNL=0.65 dB
70
60
140 110 80
10
Time (sec)
50
20
θ, deg
Figure 4.4: Effect of shield on acoustics in the downward direction with the -1Df
nozzle-shield configuration with βe = +15° (blue) compared to -1Df nozzle-shield
configuration with βe = 0° (red).
HWM112S
90
θ = 20.0
o
θ = 30.0
90
θ = 60.0
o
90
θ = 91.0
o
90
80
80
80
80
80
70
70
70
70
70
60
60
60
60
60
50
50
50
50
0.02
0.1 0.2 0.5 1 0.02
ffull (kHz)
0.1 0.2 0.5 1 0.02
125
120
115
110
105
20 40 60 80 100 120
θ (deg)
0.1 0.2 0.5 1 0.02
ffull (kHz)
PNL(dB)
OASPL(dB)
90
o
ffull (kHz)
90
PNL(dB)
SPL(dB/Hz)
Sideline
80
70
60
-20
-10
0
10
Time (sec)
HWM237S
θ = 118.0
o
50
0.1 0.2 0.5 1 0.02
ffull (kHz)
0.10.2 0.5 1
ffull (kHz)
EPNL=91.43 dB
EPNL=90.17 dB
90
80
∆EPNL=1.26 dB
70
60
140 110 80
50
20
θ, deg
Figure 4.5: Effect of shield on acoustics in the sideline direction with the shield without
the verticals (blue) compared to baseline nozzle-shield configuration (red).
79
Noise source distribution map HWM275
Sr
5
3
2
10
7
Peak noise location HWM275
5
3
2
6
5
Sr
10
1
1
4
0.5
0.3
0.2
3
2
0.5
0.3
0.2
0.1
1
0.1
0.05
-2
0
0.05
-2
0
2
4
6
8
10
x/D
Baseline
0
2
4
6
8
10
x/D
f
f
(a)
(b)
Figure 4.6: (a) Noise source distribution of free baseline jet with corresponding (b) plot
of peak intensity location compared to the trailing edge position (blue).
80
HWM257D
90
θ = 19.9
o
90
90
θ = 60.3
θ = 90.7
o
90
80
80
80
70
70
70
70
70
60
60
60
60
60
50
50
50
0.1 0.2 0.5 1 0.02
50
0.1 0.2 0.5 1 0.02
0.1 0.2 0.5 1 0.02
ffull (kHz)
PNL(dB)
125
120
115
110
105
ffull (kHz)
90
80
70
60
-20
20 40 60 80 100 120
-10
θ (deg)
0
90
θ = 20.5
90
θ = 30.2
90
80
60
140 110 80
50
20
∆EPNL=1.61 dB
θ, deg
θ = 60.3
o
90
HWM249S
θ = 90.3
o
90
80
80
80
80
70
70
70
70
70
60
60
60
60
60
50
50
0.1 0.2 0.5 1 0.02
ffull (kHz)
50
0.1 0.2 0.5 1 0.02
0.1 0.2 0.5 1 0.02
ffull (kHz)
PNL(dB)
125
120
115
110
105
ffull (kHz)
90
PNL(dB)
0.02
80
70
60
20 40 60 80 100 120
θ (deg)
-10
0.10.2 0.5 1
∆EPNL=0.24 dB
∆EPNL=1.85 dB
70
80
50
o
EPNL=91.77 dB
EPNL=91.53 dB
EPNL=89.92 dB
Time (sec)
o
θ = 118.5
ffull (kHz)
90
HWM257S
o
HWM250D
50
0.1 0.2 0.5 1 0.02
ffull (kHz)
10
Sideline
SPL(dB/Hz)
90
HWM249D
80
ffull (kHz)
OASPL(dB)
o
80
0.02
OASPL(dB)
θ = 30.1
o
PNL(dB)
SPL(dB/Hz)
Downward
0
10
Time (sec)
HWM250S
θ = 117.2
o
50
0.1 0.2 0.5 1 0.02
ffull (kHz)
0.10.2 0.5 1
ffull (kHz)
EPNL=91.46 dB
EPNL=91.17 dB
EPNL=89.74 dB
90
80
∆EPNL=0.29 dB
∆EPNL=1.72 dB
70
60
140 110 80
50
20
∆EPNL=1.43 dB
θ, deg
Figure 4.7: Effect of shield on acoustics in the downward and sideline directions.
Shielded mild chevron nozzle (blue) compared to unshielded mild chevron nozzle
(green) and unshielded baseline nozzle (red).
81
HWM257D
90
θ = 19.9
o
90
90
θ = 60.3
θ = 90.7
o
90
80
80
80
70
70
70
70
70
60
60
60
60
60
50
50
50
0.1 0.2 0.5 1 0.02
50
0.1 0.2 0.5 1 0.02
0.1 0.2 0.5 1 0.02
ffull (kHz)
PNL(dB)
125
120
115
110
105
ffull (kHz)
90
80
70
60
-20
20 40 60 80 100 120
-10
θ (deg)
0
90
θ = 20.5
90
θ = 30.2
90
80
60
140 110 80
50
20
∆EPNL=2.96 dB
θ, deg
θ = 60.3
o
90
HWM253S
θ = 90.3
o
90
80
80
80
80
70
70
70
70
70
60
60
60
60
60
50
50
0.1 0.2 0.5 1 0.02
ffull (kHz)
50
0.1 0.2 0.5 1 0.02
0.1 0.2 0.5 1 0.02
ffull (kHz)
PNL(dB)
125
120
115
110
105
ffull (kHz)
90
PNL(dB)
0.02
80
70
60
20 40 60 80 100 120
θ (deg)
-10
0.10.2 0.5 1
∆EPNL=0.38 dB
∆EPNL=3.34 dB
70
80
50
o
EPNL=91.77 dB
EPNL=91.39 dB
EPNL=88.43 dB
Time (sec)
o
θ = 118.5
ffull (kHz)
90
HWM257S
o
HWM254D
50
0.1 0.2 0.5 1 0.02
ffull (kHz)
10
Sideline
SPL(dB/Hz)
90
HWM253D
80
ffull (kHz)
OASPL(dB)
o
80
0.02
OASPL(dB)
θ = 30.1
o
PNL(dB)
SPL(dB/Hz)
Downward
0
10
Time (sec)
HWM254S
θ = 117.2
o
50
0.1 0.2 0.5 1 0.02
ffull (kHz)
0.10.2 0.5 1
ffull (kHz)
EPNL=91.46 dB
EPNL=91.42 dB
EPNL=88.31 dB
90
80
∆EPNL=0.04 dB
∆EPNL=3.15 dB
70
60
140 110 80
50
20
∆EPNL=3.11 dB
θ, deg
Figure 4.8: Effect of shield on acoustics in the downward and sideline directions.
Shielded aggressive chevron nozzle (blue) compared to unshielded aggressive chevron
nozzle (green) and unshielded baseline nozzle (red).
82
HWM257D
90
θ = 19.9
o
90
90
θ = 60.3
θ = 90.7
o
90
80
80
80
70
70
70
70
70
60
60
60
60
60
50
50
50
0.1 0.2 0.5 1 0.02
50
0.1 0.2 0.5 1 0.02
0.1 0.2 0.5 1 0.02
ffull (kHz)
PNL(dB)
125
120
115
110
105
ffull (kHz)
90
80
70
60
-20
20 40 60 80 100 120
-10
θ (deg)
0
90
θ = 20.5
90
θ = 30.2
90
80
60
140 110 80
50
20
∆EPNL=2.16 dB
θ, deg
θ = 60.3
o
90
HWM260S
θ = 90.3
o
90
80
80
80
80
70
70
70
70
70
60
60
60
60
60
50
50
0.1 0.2 0.5 1 0.02
ffull (kHz)
50
0.1 0.2 0.5 1 0.02
0.1 0.2 0.5 1 0.02
ffull (kHz)
PNL(dB)
125
120
115
110
105
ffull (kHz)
90
PNL(dB)
0.02
80
70
60
20 40 60 80 100 120
θ (deg)
-10
0.10.2 0.5 1
∆EPNL=1.70 dB
∆EPNL=3.86 dB
70
80
50
o
EPNL=91.77 dB
EPNL=90.07 dB
EPNL=87.91 dB
Time (sec)
o
θ = 118.5
ffull (kHz)
90
HWM257S
o
HWM261D
50
0.1 0.2 0.5 1 0.02
ffull (kHz)
10
Sideline
SPL(dB/Hz)
90
HWM260D
80
ffull (kHz)
OASPL(dB)
o
80
0.02
OASPL(dB)
θ = 30.1
o
PNL(dB)
SPL(dB/Hz)
Downward
0
10
Time (sec)
HWM261S
θ = 117.2
o
50
0.1 0.2 0.5 1 0.02
ffull (kHz)
0.10.2 0.5 1
ffull (kHz)
EPNL=91.46 dB
EPNL=90.75 dB
EPNL=88.41 dB
90
80
∆EPNL=0.71 dB
∆EPNL=3.05 dB
70
60
140 110 80
50
20
∆EPNL=2.34 dB
θ, deg
Figure 4.9: Effect of shield on acoustics in the downward and sideline directions.
Shielded W18x3 nozzle (blue) compared to unshielded W18x3 nozzle (green) and
unshielded baseline nozzle (red).
83
HWM257D
90
θ = 19.9
o
90
90
θ = 60.3
θ = 90.7
o
90
80
80
80
70
70
70
70
70
60
60
60
60
60
50
50
50
0.1 0.2 0.5 1 0.02
50
0.1 0.2 0.5 1 0.02
0.1 0.2 0.5 1 0.02
ffull (kHz)
PNL(dB)
125
120
115
110
105
ffull (kHz)
90
80
70
60
-20
20 40 60 80 100 120
-10
θ (deg)
0
90
θ = 20.5
90
θ = 30.2
90
80
60
140 110 80
50
20
∆EPNL=3.09 dB
θ, deg
θ = 60.3
o
90
HWM242S
θ = 90.3
o
90
80
80
80
80
70
70
70
70
70
60
60
60
60
60
50
50
0.1 0.2 0.5 1 0.02
ffull (kHz)
50
0.1 0.2 0.5 1 0.02
0.1 0.2 0.5 1 0.02
ffull (kHz)
PNL(dB)
125
120
115
110
105
ffull (kHz)
90
PNL(dB)
0.02
80
70
60
20 40 60 80 100 120
θ (deg)
-10
0.10.2 0.5 1
∆EPNL=0.63 dB
∆EPNL=3.72 dB
70
80
50
o
EPNL=91.77 dB
EPNL=91.14 dB
EPNL=88.05 dB
Time (sec)
o
θ = 118.5
ffull (kHz)
90
HWM257S
o
HWM243D
50
0.1 0.2 0.5 1 0.02
ffull (kHz)
10
Sideline
SPL(dB/Hz)
90
HWM242D
80
ffull (kHz)
OASPL(dB)
o
80
0.02
OASPL(dB)
θ = 30.1
o
PNL(dB)
SPL(dB/Hz)
Downward
0
10
Time (sec)
HWM243S
θ = 117.2
o
50
0.1 0.2 0.5 1 0.02
ffull (kHz)
0.10.2 0.5 1
ffull (kHz)
EPNL=91.46 dB
EPNL=92.63 dB
EPNL=89.04 dB
90
80
∆EPNL=-1.17 dB
∆EPNL=2.42 dB
70
60
140 110 80
50
20
∆EPNL=3.59 dB
θ, deg
Figure 4.10: Effect of shield on acoustics in the downward and sideline directions.
Shielded W30 nozzle (blue) compared to unshielded W30 nozzle (green) and unshielded
baseline nozzle (red).
84
HWM257D
90
θ = 19.9
o
90
90
θ = 60.3
θ = 90.7
o
90
80
80
80
70
70
70
70
70
60
60
60
60
60
50
50
50
0.1 0.2 0.5 1 0.02
50
0.1 0.2 0.5 1 0.02
0.1 0.2 0.5 1 0.02
ffull (kHz)
PNL(dB)
125
120
115
110
105
ffull (kHz)
90
80
70
60
-20
20 40 60 80 100 120
-10
θ (deg)
0
90
θ = 20.5
90
θ = 30.2
90
80
60
140 110 80
50
20
∆EPNL=3.22 dB
θ, deg
θ = 60.3
o
90
HWM255S
θ = 90.3
o
90
80
80
80
80
70
70
70
70
70
60
60
60
60
60
50
50
0.1 0.2 0.5 1 0.02
ffull (kHz)
50
0.1 0.2 0.5 1 0.02
0.1 0.2 0.5 1 0.02
ffull (kHz)
PNL(dB)
125
120
115
110
105
ffull (kHz)
90
PNL(dB)
0.02
80
70
60
20 40 60 80 100 120
θ (deg)
-10
0.10.2 0.5 1
∆EPNL=1.10 dB
∆EPNL=4.32 dB
70
80
50
o
EPNL=91.77 dB
EPNL=90.67 dB
EPNL=87.45 dB
Time (sec)
o
θ = 118.5
ffull (kHz)
90
HWM257S
o
HWM256D
50
0.1 0.2 0.5 1 0.02
ffull (kHz)
10
Sideline
SPL(dB/Hz)
90
HWM255D
80
ffull (kHz)
OASPL(dB)
o
80
0.02
OASPL(dB)
θ = 30.1
o
PNL(dB)
SPL(dB/Hz)
Downward
0
10
Time (sec)
HWM256S
θ = 117.2
o
50
0.1 0.2 0.5 1 0.02
ffull (kHz)
0.10.2 0.5 1
ffull (kHz)
EPNL=91.46 dB
EPNL=91.42 dB
EPNL=88.16 dB
90
80
∆EPNL=0.04 dB
∆EPNL=3.30 dB
70
60
140 110 80
50
20
∆EPNL=3.26 dB
θ, deg
Figure 4.11: Effect of shield on acoustics in the downward and sideline directions.
Shielded AC+W18x3 nozzle (blue) compared to unshielded AC+W18x3 nozzle (green)
and unshielded baseline nozzle (red).
85
Noise source distribution map HWM284
Sr
5
3
2
10
7
Peak noise location HWM275 - HWM284
5
3
2
6
5
Sr
10
1
1
4
0.5
0.3
0.2
3
2
0.5
0.3
0.2
0.1
1
0.1
0
0.05
-2
0.05
-2
0
2
4
6
8
10
Baseline
Mild chevrons
0
2
x/D
4
6
8
10
x/D
f
f
(a)
(b)
Figure 4.12: Noise source distribution of mild chevron jet with corresponding plot of
peak intensity location.
10
Differential SPL map HWM249D-HWM250D
10
4
5
3
2
Sr
Sr
5
3
2
6
1
2
0.5
0.3
0.2
0
6
4
1
2
0.5
0.3
0.2
0.1
0.05
20
Differential SPL map HWM249S-HWM250S
0
0.1
40
60
80
100
-2
θ (deg)
0.05
20
40
60
80
100
-2
θ (deg)
(a)
(b)
Figure 4.13: Insertion loss maps of mild chevron jet in the (a) downward and (b)
sideline directions.
86
Noise source distribution map HWM287
Sr
5
3
2
10
7
Peak noise location HWM275 - HWM286
5
3
2
6
5
Sr
10
1
4
0.5
0.3
0.2
3
2
0.5
0.3
0.2
0.1
1
0.1
0
0.05
-2
0.05
-2
0
2
4
6
8
10
1
Baseline
Aggressive chevrons
0
2
x/D
4
6
8
10
x/D
f
f
(a)
(b)
Figure 4.14: (a) Noise source distribution of aggressive chevron jet with corresponding
(b) plot of peak intensity location.
10
Differential SPL map HWM253D-HWM254D
10
4
5
3
2
Sr
Sr
5
3
2
6
1
2
0.5
0.3
0.2
0
6
4
1
2
0.5
0.3
0.2
0.1
0.05
20
Differential SPL map HWM253S-HWM254S
0
0.1
40
60
80
100
-2
θ (deg)
0.05
20
40
60
80
100
-2
θ (deg)
(a)
(b)
Figure 4.15: Insertion loss maps of aggressive chevron jet in the (a) downward and (b)
sideline directions.
87
Noise source distribution map HWM278
Sr
5
3
2
10
7
Peak noise location HWM275 - HWM278
5
3
2
6
5
Sr
10
1
4
0.5
0.3
0.2
3
2
0.5
0.3
0.2
0.1
1
0.1
0
0.05
-2
0.05
-2
0
2
4
6
8
10
1
Baseline
W18x3 wedge
0
2
x/D
4
6
8
10
x/D
f
f
(a)
(b)
Figure 4.16: (a) Noise source distribution of W18x3 jet with corresponding (b) plot of
peak intensity location.
10
Differential SPL map HWM260D-HWM261D
10
4
5
3
2
Sr
Sr
5
3
2
6
1
2
0.5
0.3
0.2
0
6
4
1
2
0.5
0.3
0.2
0.1
0.05
20
Differential SPL map HWM260S-HWM261S
0
0.1
40
60
80
100
-2
θ (deg)
0.05
20
40
60
80
100
-2
θ (deg)
(a)
(b)
Figure 4.17: Insertion loss maps of W18x3 jet in the (a) downward and (b) sideline
directions.
88
Noise source distribution map HWM287
Sr
5
3
2
10
7
Peak noise location HWM275 - HWM287
5
3
2
6
5
Sr
10
1
4
0.5
0.3
0.2
3
2
0.5
0.3
0.2
0.1
1
0.1
0
0.05
-2
0.05
-2
0
2
4
6
8
10
1
Baseline
AC+W18x3 combo
0
2
x/D
4
6
8
10
x/D
f
f
(a)
(b)
Figure 4.18: (a) Noise source distribution of AC+W18x3 jet with corresponding (b) plot
of peak intensity location.
10
Differential SPL map HWM255D-HWM256D
10
4
5
3
2
Sr
Sr
5
3
2
6
1
2
0.5
0.3
0.2
0
6
4
1
2
0.5
0.3
0.2
0.1
0.05
20
Differential SPL map HWM255S-HWM256S
0
0.1
40
60
80
100
-2
θ (deg)
40
60
80
100
-2
θ (deg)
(a)
Figure 4.19:
0.05
20
(b)
Insertion loss maps of AC+W18x3 jet in the downward and sideline
directions.
89
u(x,y) /Uf contours for HWP006
1
1
0.8
0.5
y/D
f
0.6
0
0.4
-0.5
0.2
-1
0
1
2
3
0
4
x/D
f
1
u(y,z) /U f contours for HWP002
x/Df =0
0.5
x/Df=1.6
x/Df =3.3
x/Df =4.9
y/D
f
0.6
0
0.4
-0.5
-1
-1 -0.5
0.8
0.2
0
z/D
f
0.5
1 -1 -0.5
0
z/D
0.5
1 -1 -0.5
f
0
z/D
0.5
f
Figure 4.20: Contours of mean velocity for the baseline jet.
90
1 -1 -0.5
0
z/D
f
0.5
1
1
u(x,y) /Uf contours for HWP007
0.8
0.5
y/D
f
0.6
0
0.4
-0.5
0.2
-1
0
1
2
x/D
3
4
f
(a)
1
u(x,y) /Uf contours for HWP006
0.8
0.5
y/D
f
0.6
0
0.4
-0.5
0.2
-1
0
1
2
x/D
3
4
f
(b)
1
u(x,y) /Uf contours for HWP004
0.8
0.5
y/D
f
0.6
0
0.4
-0.5
0.2
-1
0
1
2
x/D
3
4
f
(c)
1
u(x,y) /Uf contours for HWP009
0.8
0.5
y/D
f
0.6
0
0.4
-0.5
0.2
-1
0
1
2
x/D
3
4
f
(d)
Figure 4.21: Contours of mean axial velocity on the symmetry plane. (a) Mild chevrons;
(b) aggressive chevrons; (c) W18x3 wedge; and (d) AC+W18x3 combination.
91
1
u(y,z) /U f contours for HWP007
x/Df =0
0.5
x/Df=1.6
x/Df =3.3
x/Df =4.9
y/D
f
0.6
0
0.4
-0.5
0.2
-1
-1 -0.5
0
0.5
z/D
1 -1 -0.5
f
1
0
z/D
0.5
1 -1 -0.5
f
0
z/D
0.5
1 -1 -0.5
f
0
z/D
0.5
1
f
(a)
u(y,z) /U f contours for HWP006
x/Df =0
0.5
x/Df=1.6
x/Df =3.3
x/Df =4.9
f
y/D
0.4
-0.5
0.2
-1
-1 -0.5
0
0.5
z/D
1 -1 -0.5
f
0
z/D
0.5
1 -1 -0.5
f
0
z/D
0.5
1 -1 -0.5
f
0
z/D
0.5
1
f
(b)
u(y,z) /U f contours for HWP004
x/Df =0
0.5
x/Df=1.6
x/Df =3.3
x/Df =4.9
f
y/D
0.4
-0.5
0.2
-1
-1 -0.5
0
0.5
z/D
1 -1 -0.5
f
0
z/D
0.5
1 -1 -0.5
f
0
z/D
0.5
1 -1 -0.5
f
0
z/D
0.5
1
f
(c)
u(y,z) /U f contours for HWP009
x/Df =0
0.5
x/Df=1.6
x/Df =3.3
x/Df =4.9
0
0.4
-0.5
-1
-1 -0.5
0.8
0.6
f
y/D
0.8
0.6
0
1
0.8
0.6
0
1
0.8
0.2
0
z/D
f
0.5
1 -1 -0.5
0
z/D
0.5
1 -1 -0.5
f
0
z/D
f
0.5
1 -1 -0.5
0
z/D
0.5
1
f
(d)
Figure 4.22: Cross-sectional mean velocity contours at various axial locations. (a) Mild
chevrons; (b) aggressive chevrons; (c) W18x3 wedge; and (d) AC+W18x3 combination.
92
Distribution of u ( x)max for HWP002, HWP007, and HWP006
u/Up
1
0.8
Baseline
Mild chevrons
Aggressive chevrons
0.6
0
1
2
3
4
5
x/Df
(a)
Distribution of u ( x)max for HWP002, HWP004, and HWP009
u/Up
1
0.8
Baseline
W18x3 wedge
AC+W18x3 combination
0.6
0
1
2
3
4
5
x/Df
(b)
Figure 4.23: Comparison of axial distribution of maximum mean velocity to baseline
(red) jet. (a) Mild (green) and aggressive chevrons (blue); (b) W18x3 wedge (green) and
AC+W18x3 combination (blue).
93
y
ln
Noise source region
ψ = illumination angle
x
Ys
Shield
Xs
2-dimensional
y
φ0
Plane φ =φ0
Nozzle
Noise source region
x
ψ
ln
z
(Xs,Ys,Zs)
Trailing edge
3-dimensional
Figure 5.1: Illustration of the definition of illumination angle, ψ.
94
x
10
HWB - Downward
HWB - Sideline
Reference [36]
∆EPNL (dB)
8
6
4
2
0
0
20
40
60
80
100 120 140
ψ
Figure 5.2: Reduction of EPNL due to shielding versus average illumination angle for
all the configurations studied.
5
pc
l /D
f
4
3
2
1
0
0
1
2
3
4
5
l /D
ns
f
Figure 5.3: Correlation between noise source length and primary potential core length.
95
Appendix C
Supplementary Data
HWM257D
90
θ = 19.9
o
90
90
θ = 60.3
o
90
80
80
80
80
70
70
70
70
60
60
60
60
60
50
50
50
50
0.1 0.2 0.5 1 0.02
0.1 0.2 0.5 1 0.02
0.1 0.2 0.5 1 0.02
ffull (kHz)
PNL(dB)
125
120
115
110
105
ffull (kHz)
90
80
70
60
-20
20 40 60 80 100 120
-10
θ (deg)
0
90
θ = 20.5
90
θ = 30.2
90
∆EPNL=1.30 dB
70
50
20
θ, deg
θ = 60.3
o
90
θ = 90.3
o
90
80
80
80
80
80
70
70
70
70
70
60
60
60
60
60
50
0.02
50
0.1 0.2 0.5 1 0.02
ffull (kHz)
50
0.1 0.2 0.5 1 0.02
0.1 0.2 0.5 1 0.02
ffull (kHz)
PNL(dB)
125
120
115
110
105
ffull (kHz)
90
PNL(dB)
50
80
70
60
20 40 60 80 100 120
θ (deg)
-10
0.10.2 0.5 1
80
Time (sec)
o
o
EPNL=91.77 dB
EPNL=90.47 dB
HWM257S
o
θ = 118.5
ffull (kHz)
90
60
140 110 80
10
HWM120D
50
0.1 0.2 0.5 1 0.02
ffull (kHz)
Sideline
SPL(dB/Hz)
90
70
ffull (kHz)
OASPL(dB)
θ = 90.7
o
80
0.02
OASPL(dB)
θ = 30.1
o
PNL(dB)
SPL(dB/Hz)
Downward
0
10
Time (sec)
HWM122S
θ = 117.2
o
50
0.1 0.2 0.5 1 0.02
ffull (kHz)
0.10.2 0.5 1
ffull (kHz)
EPNL=91.46 dB
EPNL=90.61 dB
90
80
∆EPNL=0.85 dB
70
60
140 110 80
50
20
θ, deg
Figure C.1: Effect of shield on acoustics in the downward and sideline directions with
the +1Df nozzle-shield configuration (blue) compared to baseline nozzle-shield
configuration (red).
96
HWM257D
90
θ = 19.9
o
90
90
θ = 60.3
o
90
80
80
80
70
70
70
70
70
60
60
60
60
60
50
50
50
0.1 0.2 0.5 1 0.02
50
0.1 0.2 0.5 1 0.02
0.1 0.2 0.5 1 0.02
ffull (kHz)
PNL(dB)
125
120
115
110
105
ffull (kHz)
90
80
70
60
-20
20 40 60 80 100 120
-10
θ (deg)
0
90
θ = 20.5
90
θ = 30.2
90
∆EPNL=1.35 dB
70
50
20
θ, deg
θ = 60.3
o
90
θ = 90.3
o
90
80
80
80
80
80
70
70
70
70
70
60
60
60
60
60
50
50
50
50
50
0.1 0.2 0.5 1 0.02
ffull (kHz)
0.1 0.2 0.5 1 0.02
0.1 0.2 0.5 1 0.02
ffull (kHz)
PNL(dB)
125
120
115
110
105
ffull (kHz)
90
PNL(dB)
0.02
80
70
60
20 40 60 80 100 120
θ (deg)
-10
0.10.2 0.5 1
80
Time (sec)
o
o
EPNL=91.77 dB
EPNL=90.42 dB
HWM257S
o
θ = 118.5
ffull (kHz)
90
60
140 110 80
10
HWM030D
50
0.1 0.2 0.5 1 0.02
ffull (kHz)
Sideline
SPL(dB/Hz)
90
80
ffull (kHz)
OASPL(dB)
θ = 90.7
o
80
0.02
OASPL(dB)
θ = 30.1
o
PNL(dB)
SPL(dB/Hz)
Downward
0
10
Time (sec)
0.1 0.2 0.5 1 0.02
ffull (kHz)
HWM032S
θ = 117.2
o
0.10.2 0.5 1
ffull (kHz)
EPNL=91.46 dB
EPNL=90.81 dB
90
80
∆EPNL=0.65 dB
70
60
140 110 80
50
20
θ, deg
Figure C.2: Effect of shield on acoustics in the downward and sideline directions with
the -1Df nozzle-shield configuration (blue) compared to baseline nozzle-shield
configuration (red).
97
HWM257D
90
θ = 19.9
o
90
90
θ = 60.3
o
90
80
80
80
70
70
70
70
70
60
60
60
60
60
50
50
50
0.1 0.2 0.5 1 0.02
50
0.1 0.2 0.5 1 0.02
0.1 0.2 0.5 1 0.02
ffull (kHz)
PNL(dB)
125
120
115
110
105
ffull (kHz)
90
80
70
60
-20
20 40 60 80 100 120
-10
θ (deg)
0
90
θ = 20.5
90
θ = 30.2
90
∆EPNL=1.33 dB
70
50
20
θ, deg
θ = 60.3
o
90
θ = 90.3
o
90
80
80
80
80
80
70
70
70
70
70
60
60
60
60
60
50
50
50
50
50
0.1 0.2 0.5 1 0.02
ffull (kHz)
0.1 0.2 0.5 1 0.02
0.1 0.2 0.5 1 0.02
ffull (kHz)
PNL(dB)
125
120
115
110
105
ffull (kHz)
90
PNL(dB)
0.02
80
70
60
20 40 60 80 100 120
θ (deg)
-10
0.10.2 0.5 1
80
Time (sec)
o
o
EPNL=91.77 dB
EPNL=90.44 dB
HWM257S
o
θ = 118.5
ffull (kHz)
90
60
140 110 80
10
HWM020D
50
0.1 0.2 0.5 1 0.02
ffull (kHz)
Sideline
SPL(dB/Hz)
90
80
ffull (kHz)
OASPL(dB)
θ = 90.7
o
80
0.02
OASPL(dB)
θ = 30.1
o
PNL(dB)
SPL(dB/Hz)
Downward
0
10
Time (sec)
0.1 0.2 0.5 1 0.02
ffull (kHz)
HWM022S
θ = 117.2
o
0.10.2 0.5 1
ffull (kHz)
EPNL=91.46 dB
EPNL=90.33 dB
90
80
∆EPNL=1.13 dB
70
60
140 110 80
50
20
θ, deg
Figure C.3: Effect of shield on acoustics in the downward and sideline directions with βv
= 90° (blue) compared to baseline nozzle-shield configuration (red).
98
HWM257D
90
θ = 19.9
o
90
90
θ = 60.3
θ = 90.7
o
90
80
80
80
80
70
70
70
70
60
60
60
60
60
50
50
50
50
50
0.1 0.2 0.5 1 0.02
0.1 0.2 0.5 1 0.02
0.1 0.2 0.5 1 0.02
ffull (kHz)
PNL(dB)
125
120
115
110
105
ffull (kHz)
90
80
70
60
-20
20 40 60 80 100 120
-10
θ (deg)
0
90
θ = 20.5
90
θ = 30.2
90
80
60
140 110 80
50
20
θ, deg
θ = 60.3
o
90
HWM224S
θ = 90.3
o
90
80
80
80
80
70
70
70
70
60
60
60
60
60
50
50
50
50
0.1 0.2 0.5 1 0.02
ffull (kHz)
PNL(dB)
125
120
115
110
105
ffull (kHz)
90
PNL(dB)
ffull (kHz)
0.1 0.2 0.5 1 0.02
80
70
60
20 40 60 80 100 120
θ (deg)
-10
0.10.2 0.5 1
∆EPNL=1.72 dB
70
0.1 0.2 0.5 1 0.02
o
∆EPNL=1.41 dB
∆EPNL=3.13 dB
70
80
0.02
θ = 118.5
EPNL=91.77 dB
EPNL=90.36 dB
EPNL=88.64 dB
Time (sec)
o
HWM225D
ffull (kHz)
90
HWM257S
o
0.1 0.2 0.5 1 0.02
ffull (kHz)
10
Sideline
SPL(dB/Hz)
90
HWM224D
70
ffull (kHz)
OASPL(dB)
o
80
0.02
OASPL(dB)
θ = 30.1
o
PNL(dB)
SPL(dB/Hz)
Downward
0
10
Time (sec)
HWM225S
θ = 117.2
o
50
0.1 0.2 0.5 1 0.02
ffull (kHz)
0.10.2 0.5 1
ffull (kHz)
EPNL=91.46 dB
EPNL=90.79 dB
EPNL=89.32 dB
90
80
∆EPNL=0.67 dB
∆EPNL=2.14 dB
70
60
140 110 80
50
20
∆EPNL=1.47 dB
θ, deg
Figure C.4: Effect of shield on acoustics in the downward and sideline directions.
Shielded W15 nozzle (blue) compared to unshielded W15 nozzle (green) and unshielded
baseline nozzle (red).
99
HWM257D
90
θ = 19.9
o
90
90
θ = 60.3
θ = 90.7
o
90
80
80
80
70
70
70
70
70
60
60
60
60
60
50
50
50
0.1 0.2 0.5 1 0.02
50
0.1 0.2 0.5 1 0.02
0.1 0.2 0.5 1 0.02
ffull (kHz)
PNL(dB)
125
120
115
110
105
ffull (kHz)
90
80
70
60
-20
20 40 60 80 100 120
-10
θ (deg)
0
90
θ = 20.5
90
θ = 30.2
90
80
60
140 110 80
50
20
∆EPNL=2.21 dB
θ, deg
θ = 60.3
o
90
HWM258S
θ = 90.3
o
90
80
80
80
80
70
70
70
70
70
60
60
60
60
60
50
50
0.1 0.2 0.5 1 0.02
ffull (kHz)
50
0.1 0.2 0.5 1 0.02
0.1 0.2 0.5 1 0.02
ffull (kHz)
PNL(dB)
125
120
115
110
105
ffull (kHz)
90
PNL(dB)
0.02
80
70
60
20 40 60 80 100 120
θ (deg)
-10
0
0.10.2 0.5 1
∆EPNL=1.59 dB
∆EPNL=3.80 dB
70
80
50
o
EPNL=91.77 dB
EPNL=90.18 dB
EPNL=87.97 dB
Time (sec)
o
θ = 118.5
ffull (kHz)
90
HWM257S
o
HWM259D
50
0.1 0.2 0.5 1 0.02
ffull (kHz)
10
Sideline
SPL(dB/Hz)
90
HWM258D
80
ffull (kHz)
OASPL(dB)
o
80
0.02
OASPL(dB)
θ = 30.1
o
PNL(dB)
SPL(dB/Hz)
Downward
10
Time (sec)
HWM259S
θ = 117.2
o
50
0.1 0.2 0.5 1 0.02
ffull (kHz)
0.10.2 0.5 1
ffull (kHz)
EPNL=91.46 dB
EPNL=91.33 dB
EPNL=89.16 dB
90
80
∆EPNL=0.13 dB
∆EPNL=2.30 dB
70
60
140 110 80
50
20
∆EPNL=2.17 dB
θ, deg
Figure C.5: Effect of shield on acoustics in the downward and sideline directions.
Shielded W18 nozzle (blue) compared to unshielded W18 nozzle (green) and unshielded
baseline nozzle (red).
100
HWM257D
90
θ = 19.9
o
90
90
θ = 60.3
θ = 90.7
o
90
80
80
80
70
70
70
70
70
60
60
60
60
60
50
50
50
0.1 0.2 0.5 1 0.02
50
0.1 0.2 0.5 1 0.02
0.1 0.2 0.5 1 0.02
ffull (kHz)
PNL(dB)
125
120
115
110
105
ffull (kHz)
90
80
70
60
-20
20 40 60 80 100 120
-10
θ (deg)
0
90
θ = 20.5
90
θ = 30.2
90
80
60
140 110 80
50
20
∆EPNL=2.04 dB
θ, deg
θ = 60.3
o
90
HWM226S
θ = 90.3
o
90
80
80
80
80
70
70
70
70
70
60
60
60
60
60
50
50
0.1 0.2 0.5 1 0.02
ffull (kHz)
50
0.1 0.2 0.5 1 0.02
0.1 0.2 0.5 1 0.02
ffull (kHz)
PNL(dB)
125
120
115
110
105
ffull (kHz)
90
PNL(dB)
0.02
80
70
60
20 40 60 80 100 120
θ (deg)
-10
0
0.10.2 0.5 1
∆EPNL=1.70 dB
∆EPNL=3.74 dB
70
80
50
o
EPNL=91.77 dB
EPNL=90.07 dB
EPNL=88.03 dB
Time (sec)
o
θ = 118.5
ffull (kHz)
90
HWM257S
o
HWM227D
50
0.1 0.2 0.5 1 0.02
ffull (kHz)
10
Sideline
SPL(dB/Hz)
90
HWM226D
80
ffull (kHz)
OASPL(dB)
o
80
0.02
OASPL(dB)
θ = 30.1
o
PNL(dB)
SPL(dB/Hz)
Downward
10
Time (sec)
HWM227S
θ = 117.2
o
50
0.1 0.2 0.5 1 0.02
ffull (kHz)
0.10.2 0.5 1
ffull (kHz)
EPNL=91.46 dB
EPNL=91.16 dB
EPNL=89.48 dB
90
80
∆EPNL=0.30 dB
∆EPNL=1.98 dB
70
60
140 110 80
50
20
∆EPNL=1.68 dB
θ, deg
Figure C.6: Effect of shield on acoustics in the downward and sideline directions.
Shielded W21 nozzle (blue) compared to unshielded W21 nozzle (green) and unshielded
baseline nozzle (red).
101
HWM257D
90
θ = 19.9
o
90
90
θ = 60.3
θ = 90.7
o
90
80
80
80
70
70
70
70
70
60
60
60
60
60
50
50
50
0.1 0.2 0.5 1 0.02
50
0.1 0.2 0.5 1 0.02
0.1 0.2 0.5 1 0.02
ffull (kHz)
PNL(dB)
125
120
115
110
105
ffull (kHz)
90
80
70
60
-20
20 40 60 80 100 120
-10
θ (deg)
0
90
θ = 20.5
90
θ = 30.2
90
80
60
140 110 80
50
20
∆EPNL=1.72 dB
θ, deg
θ = 60.3
o
90
HWM228S
θ = 90.3
o
90
80
80
80
80
70
70
70
70
70
60
60
60
60
60
50
50
0.1 0.2 0.5 1 0.02
ffull (kHz)
50
0.1 0.2 0.5 1 0.02
0.1 0.2 0.5 1 0.02
ffull (kHz)
PNL(dB)
125
120
115
110
105
ffull (kHz)
90
PNL(dB)
0.02
80
70
60
20 40 60 80 100 120
θ (deg)
-10
0
0.10.2 0.5 1
∆EPNL=1.75 dB
∆EPNL=3.47 dB
70
80
50
o
EPNL=91.77 dB
EPNL=90.02 dB
EPNL=88.3 dB
Time (sec)
o
θ = 118.5
ffull (kHz)
90
HWM257S
o
HWM229D
50
0.1 0.2 0.5 1 0.02
ffull (kHz)
10
Sideline
SPL(dB/Hz)
90
HWM228D
80
ffull (kHz)
OASPL(dB)
o
80
0.02
OASPL(dB)
θ = 30.1
o
PNL(dB)
SPL(dB/Hz)
Downward
10
Time (sec)
HWM229S
θ = 117.2
o
50
0.1 0.2 0.5 1 0.02
ffull (kHz)
0.10.2 0.5 1
ffull (kHz)
EPNL=91.46 dB
EPNL=91.09 dB
EPNL=89.66 dB
90
80
∆EPNL=0.37 dB
∆EPNL=1.80 dB
70
60
140 110 80
50
20
∆EPNL=1.43 dB
θ, deg
Figure C.7: Effect of shield on acoustics in the downward and sideline directions.
Shielded TW15 nozzle (blue) compared to unshielded TW15 nozzle (green) and
unshielded baseline nozzle (red).
102
HWM257D
90
θ = 19.9
o
90
90
θ = 60.3
θ = 90.7
o
90
80
80
80
70
70
70
70
70
60
60
60
60
60
50
50
50
0.1 0.2 0.5 1 0.02
50
0.1 0.2 0.5 1 0.02
0.1 0.2 0.5 1 0.02
ffull (kHz)
PNL(dB)
125
120
115
110
105
ffull (kHz)
90
80
70
60
-20
20 40 60 80 100 120
-10
θ (deg)
0
90
θ = 20.5
90
θ = 30.2
90
80
60
140 110 80
50
20
∆EPNL=1.91 dB
θ, deg
θ = 60.3
o
90
HWM232S
θ = 90.3
o
90
80
80
80
80
70
70
70
70
70
60
60
60
60
60
50
50
0.1 0.2 0.5 1 0.02
ffull (kHz)
50
0.1 0.2 0.5 1 0.02
0.1 0.2 0.5 1 0.02
ffull (kHz)
PNL(dB)
125
120
115
110
105
ffull (kHz)
90
PNL(dB)
0.02
80
70
60
20 40 60 80 100 120
θ (deg)
-10
0
0.10.2 0.5 1
∆EPNL=1.93 dB
∆EPNL=3.84 dB
70
80
50
o
EPNL=91.77 dB
EPNL=89.84 dB
EPNL=87.93 dB
Time (sec)
o
θ = 118.5
ffull (kHz)
90
HWM257S
o
HWM233D
50
0.1 0.2 0.5 1 0.02
ffull (kHz)
10
Sideline
SPL(dB/Hz)
90
HWM232D
80
ffull (kHz)
OASPL(dB)
o
80
0.02
OASPL(dB)
θ = 30.1
o
PNL(dB)
SPL(dB/Hz)
Downward
10
Time (sec)
HWM233S
θ = 117.2
o
50
0.1 0.2 0.5 1 0.02
ffull (kHz)
0.10.2 0.5 1
ffull (kHz)
EPNL=91.46 dB
EPNL=91.37 dB
EPNL=89.81 dB
90
80
∆EPNL=0.09 dB
∆EPNL=1.65 dB
70
60
140 110 80
50
20
∆EPNL=1.56 dB
θ, deg
Figure C.8: Effect of shield on acoustics in the downward and sideline directions.
Shielded TW18 nozzle (blue) compared to unshielded TW18 nozzle (green) and
unshielded baseline nozzle (red).
103
HWM257D
90
θ = 19.9
o
90
90
θ = 60.3
θ = 90.7
o
90
80
80
80
80
70
70
70
70
60
60
60
60
60
50
50
50
50
50
0.1 0.2 0.5 1 0.02
0.1 0.2 0.5 1 0.02
0.1 0.2 0.5 1 0.02
ffull (kHz)
PNL(dB)
125
120
115
110
105
ffull (kHz)
90
80
70
60
-20
20 40 60 80 100 120
-10
θ (deg)
0
90
θ = 20.5
90
θ = 30.2
90
80
60
140 110 80
50
20
θ, deg
θ = 60.3
o
90
HWM234S
θ = 90.3
o
90
80
80
80
80
70
70
70
70
60
60
60
60
60
50
50
50
50
0.1 0.2 0.5 1 0.02
ffull (kHz)
PNL(dB)
125
120
115
110
105
ffull (kHz)
90
PNL(dB)
ffull (kHz)
0.1 0.2 0.5 1 0.02
80
70
60
20 40 60 80 100 120
θ (deg)
-10
0
0.10.2 0.5 1
∆EPNL=1.97 dB
70
0.1 0.2 0.5 1 0.02
o
∆EPNL=1.96 dB
∆EPNL=3.93 dB
70
80
0.02
θ = 118.5
EPNL=91.77 dB
EPNL=89.81 dB
EPNL=87.84 dB
Time (sec)
o
HWM235D
ffull (kHz)
90
HWM257S
o
0.1 0.2 0.5 1 0.02
ffull (kHz)
10
Sideline
SPL(dB/Hz)
90
HWM234D
70
ffull (kHz)
OASPL(dB)
o
80
0.02
OASPL(dB)
θ = 30.1
o
PNL(dB)
SPL(dB/Hz)
Downward
10
Time (sec)
HWM235S
θ = 117.2
o
50
0.1 0.2 0.5 1 0.02
ffull (kHz)
0.10.2 0.5 1
ffull (kHz)
EPNL=91.46 dB
EPNL=90.48 dB
EPNL=89.15 dB
90
80
∆EPNL=0.98 dB
∆EPNL=2.31 dB
70
60
140 110 80
50
20
∆EPNL=1.33 dB
θ, deg
Figure C.9: Effect of shield on acoustics in the downward and sideline directions.
Shielded TW18x3 nozzle (blue) compared to unshielded TW18x3 nozzle (green) and
unshielded baseline nozzle (red).
104
Noise source distribution map HWM276
5
3
2
Sr
10
7
Peak noise location HWM275 - HWM276
5
3
2
6
5
Sr
10
1
4
0.5
0.3
0.2
3
2
0.5
0.3
0.2
0.1
1
0.1
0
0.05
-2
0.05
-2
0
2
4
6
8
10
x/D
1
Baseline
W15 wedge
0
2
4
6
8
10
x/D
f
f
(a)
(b)
Figure C.10: (a) Noise source distribution of W15 jet with corresponding (b) plot of
peak intensity location.
(a)
(b)
Figure C.11: Insertion loss maps of W15 jet in the (a) downward and (b) sideline
directions.
105
Noise source distribution map HWM277
5
3
2
Sr
10
7
Peak noise location HWM275 - HWM277
5
3
2
6
5
Sr
10
1
4
0.5
0.3
0.2
3
2
0.5
0.3
0.2
0.1
1
0.1
0
0.05
-2
0.05
-2
0
2
4
6
8
10
1
Baseline
W18 wedge
0
2
x/D
4
6
8
10
x/D
f
f
(a)
(b)
Figure C.12: (a) Noise source distribution of W18 jet with corresponding (b) plot of
peak intensity location.
10
Differential SPL map HWM258D-HWM259D
10
4
5
3
2
Sr
Sr
5
3
2
6
1
2
0.5
0.3
0.2
0
6
4
1
2
0.5
0.3
0.2
0.1
0.05
20
Differential SPL map HWM258S-HWM259S
0
0.1
40
60
80
100
-2
0.05
20
θ (deg)
40
60
80
100
-2
θ (deg)
(a)
(b)
Figure C.13: Insertion loss maps of W18 jet in the (a) downward and (b) sideline
directions.
106
Noise source distribution map HWM279
5
3
2
Sr
10
7
Peak noise location HWM275 - HWM279
5
3
2
6
5
Sr
10
1
4
0.5
0.3
0.2
3
2
0.5
0.3
0.2
0.1
1
0.1
0
0.05
-2
0.05
-2
0
2
4
6
8
10
1
Baseline
W21 wedge
0
2
x/D
4
6
8
10
x/D
f
f
(a)
(b)
Figure C.14: (a) Noise source distribution of W21 jet with corresponding (b) plot of
peak intensity location.
10
Differential SPL map HWM226D-HWM227D
10
4
5
3
2
Sr
Sr
5
3
2
6
1
2
0.5
0.3
0.2
0
6
4
1
2
0.5
0.3
0.2
0.1
0.05
20
Differential SPL map HWM226S-HWM227S
0
0.1
40
60
80
100
-2
0.05
20
θ (deg)
40
60
80
100
-2
θ (deg)
(a)
(b)
Figure C.15: Insertion loss maps of W21 jet in the (a) downward and (b) sideline
directions.
107
Noise source distribution map HWM280
5
3
2
Sr
10
7
Peak noise location HWM275 - HWM280
5
3
2
6
5
Sr
10
1
4
0.5
0.3
0.2
3
2
0.5
0.3
0.2
0.1
1
0.1
0
0.05
-2
0.05
-2
0
2
4
6
8
10
1
Baseline
W30 wedge
0
2
x/D
4
6
8
10
x/D
f
f
(a)
(b)
Figure C.16: (a) Noise source distribution of W30 jet with corresponding (b) plot of
peak intensity location.
10
Differential SPL map HWM242D-HWM243D
10
4
5
3
2
Sr
Sr
5
3
2
6
1
2
0.5
0.3
0.2
0
6
4
1
2
0.5
0.3
0.2
0.1
0.05
20
Differential SPL map HWM242S-HWM243S
0
0.1
40
60
80
100
-2
0.05
20
θ (deg)
40
60
80
100
-2
θ (deg)
(a)
(b)
Figure C.17: Insertion loss maps of W30 jet in the (a) downward and (b) sideline
directions.
108
Noise source distribution map HWM281
5
3
2
Sr
10
7
Peak noise location HWM275 - HWM281
5
3
2
6
5
Sr
10
1
4
0.5
0.3
0.2
3
2
0.5
0.3
0.2
0.1
1
0.1
0
0.05
-2
0.05
-2
0
2
4
6
8
10
1
Baseline
TW15 wedge
0
2
x/D
4
6
8
10
x/D
f
f
(a)
(b)
Figure C.18: (a) Noise source distribution of TW15 jet with corresponding (b) plot of
peak intensity location.
10
Differential SPL map HWM228D-HWM229D
10
4
5
3
2
Sr
Sr
5
3
2
6
1
2
0.5
0.3
0.2
0
6
4
1
2
0.5
0.3
0.2
0.1
0.05
20
Differential SPL map HWM228S-HWM229S
0
0.1
40
60
80
100
-2
0.05
20
θ (deg)
40
60
80
100
-2
θ (deg)
(a)
(b)
Figure C.19: Insertion loss maps of TW15 jet in the (a) downward and (b) sideline
directions.
109
Noise source distribution map HWM282
5
3
2
Sr
10
7
Peak noise location HWM275 - HWM282
5
3
2
6
5
Sr
10
1
4
0.5
0.3
0.2
3
2
0.5
0.3
0.2
0.1
1
0.1
0
0.05
-2
0.05
-2
0
2
4
6
8
10
1
Baseline
TW18 wedge
0
2
x/D
4
6
8
10
x/D
f
f
(a)
(b)
Figure C.20: (a) Noise source distribution of TW18 jet with corresponding (b) plot of
peak intensity location.
10
Differential SPL map HWM232D-HWM233D
10
4
5
3
2
Sr
Sr
5
3
2
6
1
2
0.5
0.3
0.2
0
6
4
1
2
0.5
0.3
0.2
0.1
0.05
20
Differential SPL map HWM232S-HWM233S
0
0.1
40
60
80
100
-2
0.05
20
θ (deg)
40
60
80
100
-2
θ (deg)
(a)
(b)
Figure C.21: Insertion loss maps of TW18 jet in the (a) downward and (b) sideline
directions.
110
Noise source distribution map HWM283
5
3
2
Sr
10
7
Peak noise location HWM275 - HWM283
5
3
2
6
5
Sr
10
1
4
0.5
0.3
0.2
3
2
0.5
0.3
0.2
0.1
1
0.1
0
0.05
-2
0.05
-2
0
2
4
6
8
10
1
Baseline
TW18x3 wedge
0
2
x/D
4
6
8
10
x/D
f
f
(a)
(b)
Figure C.22: (a) Noise source distribution of TW18x3 jet with corresponding (b) plot of
peak intensity location.
10
Differential SPL map HWM234D-HWM235D
10
4
5
3
2
Sr
Sr
5
3
2
6
1
2
0.5
0.3
0.2
0
6
4
1
2
0.5
0.3
0.2
0.1
0.05
20
Differential SPL map HWM234S-HWM235S
0
0.1
40
60
80
100
-2
0.05
20
θ (deg)
40
60
80
100
-2
θ (deg)
(a)
(b)
Figure C.23: Insertion loss maps of TW18x3 jet in the (a) downward and (b) sideline
directions.
111
1
u(x,y) /Uf contours for HWP019
0.8
0.5
y/D
f
0.6
0
0.4
-0.5
0.2
-1
0
1
2
x/D
3
4
f
(a)
1
u(x,y) /Uf contours for HWP020
0.8
0.5
y/D
f
0.6
0
0.4
-0.5
0.2
-1
0
1
2
x/D
3
4
f
(b)
1
u(x,y) /Uf contours for HWP021
0.8
0.5
y/D
f
0.6
0
0.4
-0.5
0.2
-1
0
1
2
x/D
3
4
f
(c)
1
u(x,y) /Uf contours for HWP012
0.8
0.5
y/D
f
0.6
0
0.4
-0.5
0.2
-1
0
1
2
x/D
3
4
f
(d)
Figure C.24: Contours of mean axial velocity on the symmetry plane. (a) W15 wedge;
(b) W18 wedge; (c) W21 wedge; and (d) W30 wedge.
112
1
u(x,y) /Uf contours for HWP023
0.8
0.5
y/D
f
0.6
0
0.4
-0.5
0.2
-1
0
1
2
x/D
3
4
f
(a)
1
u(x,y) /Uf contours for HWP024
0.8
0.5
y/D
f
0.6
0
0.4
-0.5
0.2
-1
0
1
2
x/D
3
4
f
(b)
1
u(x,y) /Uf contours for HWP025
0.8
0.5
y/D
f
0.6
0
0.4
-0.5
0.2
-1
0
1
2
x/D
3
4
f
(c)
1
u(x,y) /Uf contours for HWP011
0.8
0.5
y/D
f
0.6
0
0.4
-0.5
0.2
-1
0
1
2
x/D
3
4
f
(d)
Figure C.25: Contours of mean axial velocity on the symmetry plane. (a) TW15 wedge;
(b) TW18 wedge; (c) TW18x3 wedge; and (d) MC+W18x3 combination.
113
Документ
Категория
Без категории
Просмотров
0
Размер файла
8 017 Кб
Теги
sdewsdweddes
1/--страниц
Пожаловаться на содержимое документа