https://ntrs.nasa.gov/search.jsp?

R=19840027014 2018-06-10T18:31:30+00:00Z

NASA Technical Memorandum 83799 NASA-TM-83799 19840027014

Supersonic Jet Shock Noise Reduction

"?"-:4"""". " - ,

j'C£ 7"_ "_:',.:

;'-'--:Z', 22..;7" _:ii3 2_&:(3:.

James R. Stone
Lewis Research Center

Cleveland, Ohio _[_,[__tJ,. O_';V

UU-,"

;,:'. ;..

LANGLEYRESEARCHCENTER
L:ER_RY, IU:SA

Prepared for the

Ninth Aeroacoustics Conference
sponsored by the American Institute of Aeronautics and Astronautics
i Williamsburg, Virginia, October 15-17, 1984
 SUPERSONIC JET SHOCK NOISE REDUCTION
James.R. Stone* -
National Aeronautics and Spac~ Administration
L~wis Research Center
Cleveland, Ohio 44135

Abstract B shock strength, ,1M] -1

Shock-cell noise has been identified as a p density
potentially significant problem for advanced super- A wavelength
sonic aircraft at takeoff. Therefore NASA has e polar angle from inlet axis, deg
conducted fundamental studies of the phenomena
involved and model-scale experiments aimed at e apparent angle of attack 25 , deg
a
developing means of noise reduction. This paper Mach angle, 180 -sin- 1(I/M j ), deg
o
eM
reviews the results of a series of studies con-
ducted to determine means by which supersonic jet Subscripts:
shock noise can be reduced to acceptable levels for a ambient
advanced supersonic cruise aircraft. Theoretical
studies were conducted on the shock associated c convection
noise of supersonic jets from convergent-divergent d design point
(C-O) nozzles. Laboratory studies were conducted
m
on the influence of narrowband shock screech on o downstream
m
N
broadband noise and on means of screech reduction. e equivalent
N
I
The usefulness of C-D nozzle passages was investi-
~ gated at model scale for single-stream and dual- ef effective
stream nozzles. The effect of off-design pressure Ex exit
ratio was determined under static and simulated
flight conditions for jet temperatures up to 960 K. h hydraulic
Annular and coannular flow passages with center ISA international standard atmosphere (288 K
plugs and multielement suppressor nozzles were
evaluated, and the effect of plug tip geometry was and 101.3 kN/m2)
established. In addition to the far-field acoustic
data, mean and turbulent velocity distributions inner
were obtained with a laser velocimeter, and shadow- j fully-expanded jet
graph images of the flow field were obtained. n shock cell number (downstream of nozzle
Nomenclature exit)
(All symbols are in SI units unless noted.) o outer
A area P premerged
c speed of sound p plug
o nozzle diameter o aircraft
F functional relation 1 inner stream
f 1/3-octave-band center frequency 2 outer stream
h annulus height
k ratio of convection velocity to jet Introduction
velocity Shock-cell noise has been identified as a
L shock cell spacing potentially significant problem for advanced super-
sonic aircraft at takeoff. Therefore, the NASA
M Mach number, VIc Lewis and Langley Research Centers have conducted
OASPL overall sound pressure level, dB re 20 ~N/m2 fundamental studies of the phenomena involved and
more applied studies aimed at developing means of
P pressure noise reduction. Early studies indicated that it
R source-to-observer distance might be necessary to employ convergent-divergent
(C-D) nozzle passages in order to reduce or elimi-
S nondimensional frequency parameter nate shock noise. Some benefits have been demon-
SPL 1/3-octave-band sound pressure level, strated statically for single stream C-D circular 1
nozzles at design jet Mach numbers of 1.5 and 2.0;
dB re 20 ~N/m2 however, lower jet Mach numbers will probably be
T total temperature required to achieve acceptable jet noise levels.
In addition, more complicated single- or dual-
V velocity stream nozzle geometries, possibly including multi-
X distance downstream of nozzle exit element suppressor exhaust passages will probably
be required, and the effect of flight should be
a turbulent length scale ratio determined. Existing shock noise theories were
found to apply to a limited degree, but only to
*Member AIAA. convergent circular nozzles, thus indicating a need
for further theoretical advances. Experimental data
were required to determine the degree to which these
theoretical approaches could be applied to the com-
plicated nozzle geometries and to evaluate the merit
of the noise reduction concepts. This paper sum-
marizes the results of the NASA Lewis Research
Center supersonic jet shock noise reduction studies
and relates these results to those of other inves-
tigators, particularly those involved in the NASA
Langley Research Center Program.
A multifaceted approach was employed in the
Lewis program to determine means by which super-
sonic jet shock noise can be reduced to acceptable
levels for advanced supersonic cruise aircraft.
Extensive photographic data on the flo~ fields of
coaxial supersonic jets were obtained. Theore-
tical studies were conducted on the shock assocj-
ated noise of supersonic jets from C-O nozzles. j ,4
The influence of narrowband shock screech on broad-
band shock noise was investiga~ed and means of
screech reduction developed. -. The usefulness
~:r~~~8~BZ!~; ~~!~~~~~e~~r02rye~~~~f~~~ f~~eS~~¥~~t
of off-design pressure ratio was determined under
static and simulated flight conditions for jet tem-
perature up to 960 K. Annular and coannular flow
passages with center plugs and multielement sup-
pressor nozzles were evaluated, and the effect of
plug tip qeometry was established. In addition to
the far-field acoustic data, mean and turbulent
velocity distributions were obtained with a laser
velocimeter, and shadowqraph images of the flow
field were obtained. Other concepts which may
prove useful for shock noise reductioo are porous
plug nozzles l2 - 13 and lined ejectors,14 but these
approaches will not be discussed in this paper.
Circular Nozzles
When a supersonic jet exhausts from a conver-
gent circular nozzle the static pressure at the
nozzle exit is above ambient. Thus, as the jet
expands outside the nozzle expansion waves are
radiated from the nozzle lip and are subsequently
reflected by the jet boundary. These reflected
compression waves converge to form a shock, again
raising the jet static pressure above ambient,
leading to another set of expansion waves, and the
process is repeated, resulting in the familiar
repetitive shock structure shown in Fig. 1. The
viscous mixinq taking place in the jet shear layer
reduces the strength and spacing of these shock
cells in the downstream direction. The eddies
formed by this mixing process are convected down-
stream and interact with the shock cells to produce
shock noise.
A very useful, although rather simplified
approach to predicting shock noise for convergent
Circular jets was proposed by Harper-Bourne and
Fisher. 5 They characterized the far-field noise
as resulting from the summation of the noise gener-
ated by each eddy interacting with each shock cell
as indicated by the "group" spectrum in Fig. 2.
The acoustic waves from the different shocks were
modeled to constructively or destructively inter-
fere depending on the lifetime and convection veloc-
ity of the eddies, resulting in the "interference"
spectrum in Fig. 2. The superposition of these
results produces the characteristic spectrum of
broadband shock noise, as also shown in Fig. 2.
One of the major results of this model is that the
broadband shock noise sound pressur level (SPL)
4
should be proportional to 10 log e , where
2
e ~~. This same result has also been ob-
tained theoreti4alJy, based on different reasoning
by Tam et al. 3, ,1 The predicted trend is shown
by the experimental results for modest values of
e, as shown in Fig. 3, where the peak SPL is plot-
ted against 10 log a for a forward quadrant direc-
tivity angle (referred to the nozzle upstream axis,
9 = 50·). However, at higher values of e a Mach
disc forms in the flow, reducing the strength of the
downstream shocks and causing the noise levels to
become relatively insensitive to further i£9rr§ses
in e, as pointed out by Seiner and Norum. '
Influence of Screech
In addition to the broadband shock noise, dis-
cussed in the preceding section, and jet mixing
noise, narrowband "screech" tones are often ob-
served for imperfectly expanded jets. The feedback
mechanism involved has been the subject of several
investi~ations based primarily on the early work of
Powell.l~ Since screech tones often produce higher
SPL levels, even on a 1/3-octave-band basis, than
broadband shock noise, many investigators LU - 24 have
used nozzle lip modifications to break the feedback
loop and suppress screech in order to investigate
broadband shock noise. However, the devices used
to suppress screech also influenced the flow field
development and consequentjy the broadband noise,
as discussed by von Glahn. This problem motivated
the studies conducted by Nagel et al. to develop
means of screech suppressiog ~hich do not influ-
ence flow field development. -
Nagel et al. 6 devised a new method for screech
tone elimination biged on the feedback mechanism
proposed by Powell and the assumption that the
acoustic waves propagating back to the nozzle exit
are nearly plane waves at the nozzle exit. Under
these conditions it is possible to position a re-
flective surface near the nozzle which establishes
a local standing wave pattern from the incident
and reflected screech tone. As illustrated in
Fig. 4 a minimum pressure occurs at a distance of
1/4 wavelength from the reflector (and also at 3/4,
5/4, etc. wavelengths). If the reflector is posi-
tioned at the proper distance upstream of the
exit, a node will occur at the nozzle exit plane
(Fiq. 4). This node serves to cancel the screech
tone at the nozzle exit and interrupt the feedback
mechanism. 6
Although in theory a hard reflector should be
used, Nagel et al. 6 obtained more consistent re-
sults with a foam disk. A comparison of spectra
obtained in a reverberant room is shown in Fig. 5
for Mj = 1.4 with screech, with screech removed
by an lntrusive tab, and with screech cancelled by
a foam disk located approximately 1/4 screech tone
wavelength upstream of the nozzle exit. Both meth-
ods largely suppress screech. The intrusive tab
however, significantly alters the broadband noise,
while the foam disk does not. When screech is re-
moved with the tab the shock cell spacing is re-
duced by about 10 percent, and severe distortions
caused by the tab are evident in the flow and shear
layer. 6 This is 2~nsistent with the results of
Norum and Seiner. In contrast, when the screech
is removed by cancellation the flow appears un-
altered from the clean nozzle case with screech. 6
Thus, it appears that where feasible, the cancella-
tion approach is the preferred method for removing
screech tones so that broadband shock noise and its
suppression can be investigated.
Effect of C-D Termination layer is thin. Norum and Seiner 23 found that the
domi~ant sources of shock noise lie much further
Yamamoto et a1 8 ,9 investigated the effect of downstream, near the end of the potential core,
a C-D termination designed for shock-free flow at where· the mixing layer is quite thick. Therefore,
Mj = 1.40 (pressure ratio, Pj/P a = 3.12) on circu- the vortex sheet model cannot be very accurate.
lar nozzle shock noise. Experiments were conducted Thus a better shock cell model solution which takes
over a range of pressure ratios from 2.6 to 3.9, into account the spatial evolution of the mean flow
covering both overexpanded and underexpanded condi- is needed and has been developed by Tam et al. 3,4
tions, under static and simulated flight condi- In terms of sound lev 1 sc~l~ng, the peak SPL varies
tions. Results for a typical jet temperature, 2
as 10 log [(Aj/R2)(MJ - Mal ], as also obtained
Tj ~ 960 K, are shown in Fig. 6, where the overall by th3 ~impler formulation. 6 The more complete
sound pressure level, OASPL, is plotted against the model' requires numerical computation of t~5 peak
logarithm of the shock strength parameter, e, for a frequency, instead of the simplified result,
forward-quadrant angle, e = 50·. Data for the C-D
termination are compared with data for the conver-
gent termination. Large noise reductions can be
seen in the vicinity of the C-D design point. It As can be seen in Fig. 9, the numerically calcu-
is an important observation that noise reductions lated peak frequency agrees more closely than the
are obtained over a wide range of pressure ratios,
not just in the immediate vicinity of the design simplified relationship with the experimental data.
point. Directivity comparisons at the C-D design These theoretical findings can be incorporated
point (shown in Fig. 7) indicate that the noise
reductions are obtained throughout the forward into simple empirical predi~Sion methods, such as
quadrant and extend somewhat into the rear quad- that of Stone and Montegani rather easily. The
rant. The peak noise, which should be dominated following expression is obtained for the overall
by jet mixing noise, is essentially unchanged by sound pressure level:
the C-D termination.
The spectral comparisons at the C-D design OASPL
point (shown in Fig. 8) provide further insight.
The C-D nozzle at its design point shows a typical
jet mixing noise spectral shape with very little
evidence of shock noise. Thus, it can be seen that
jet mixing noise sets the "noise floor" when shock
noise is suppressed. This explains the greater
noise reduction, -11 dB, in simulated flight than
log (~
in the static case, -7.5 dB, for the C-D termina-
tion, since the shock noise is amplified in flight
in the forward quadrant while the jet mixing noise
is generally reduced slightly (although in this
particular case there is very little effect). (1a)
Shadowgraph photographs and laser doppler veloci-
meter surveys show no evidence of shock structure where eM is the Mach angle given by
for the C-D termination.
180·-sin-1(1/Mj). The function F is given by
Modeling
F= 0 for e 5.. eM
Reasonably accurate empirical correlations of
broadband shock noise for converqent circular noz- (la)
zles have been obtained by several investigator 8 ,25 F = - 0.75 for e > e m
based on the Har%er-Bourne and Fis~Br approach,15
2
and theoretica1 ,27 and numerical models have
been formulated which predict the essential fea- Th2 mor~ ~omplicated relation than
tures of the noise generation process. However, (M J - Ma) 1[1 + (M2 - M~)2] is introduced
the interest in C.D nozzles led to the need to (rather than e4 ) io account for the Mach disk for-
develop a theoretical model for the shock noise mati?n and 1~n15quent leveling off of shock noise
generation for off-design C-D nozzles. at hlgh e. ,

Tam and Tanna 16 proposed that this noise be The appropriate nondimensional frequency pa-
modeled as weak but coherent interaction between ramete~ again based on the Harper-Bourne and
downstream propagating large turbulence structures Fisher ; model, but also consistent with Tam and
in the mixing layer of the jet and quasi-periodic Tanna, 16 is given by
shock cells. By using simple analytical models to
represent the large turbulence structures and the
shock cells they derived relations for shock noise
S ~~~)~IM~ M~I ~
- - Mo cos (e + e a~
intensity scaling and peak fr~quency. Comparisons
with their experimental data 16 were favorable, pro-
viding support for the general validity of the pro-
posed me %anism. However, the vortex sheet model x (2)
solution1 is adequate only as a first approxima-
tion. It provides a reasonably good description
of the weak shock cells in the region immediately Note that the convection velocity factor k = 0.7,
downstream of the nozzle exit where the mixing and the turbulence length scale factor a = 0.2.

3
The shock noise peaks at S = 1.0 and varies with
log S (as shown in Fig. 10).
Annular Plug Nozzles
The experimental configurations of Yamamoto
et al. 8 ,g included an annular plug nozzle with
both convergent and convergent-divergent termina-
tions (as shown in Fig. 11). The annular radius
ratio at the throat was 0.85. Experiments were
also conducted with screech tabs on the convergent
configuration.
Influence of Screech
A limited study was conducted on screech and
the usefulness of tabs in reducing screech, along
with the resulting effects on broadband noise.
Narrowband data at e = 60· obtained with and
without tabs are compared in Fig. 12, for a pres-
sure ratio, Pj/P a ~ 3.4 and jet temperature,
Tj = 483 K. A strong discrete tone at 937.5 Hz is
apparent for the nozzle without tabs and is barely
discernible with tabs. In addition the broadband
shock noise is somewhat reduced in level and
shifted to a higher frequency with tabs. The cor-
responding 1/3-octave-band spectra are shown in
Fig. 13(a), and it can be seen that the tabs influ-
ence a fairly wide frequency range. The shift of
the broadband peak cannot be seen because the spec-
trum without tabs is so strongly influenced by
screech that the broadband peak cannot be deter-
mined. The corresponding aft quadrant spectra at
e = 140· are shown in Fig. 13(b). The screech tabs
produce a noise reduction over the entire spectrum,
even though jet mixing noise is the dominant source
at this angle. It is clear from these results that
the tabs influence the mixing process as well as
break the feedback loop. At higher temperature,
Ti ~ 960 K, the screech is weaker and the influence
ot the tabs is less than at the lower temperature
(as shown in Fig. 14). From these comparisons it
appears that the high-temperature data without tabs
are the most valid broadband shock noise data,
since the influence of screech is fairly small,
whereas the tabs have a noticeable effect on the
flow field. The foam disc cancellation approach
was not considered practical for these tests be-
cause of the inclusion of simulated flight testing.
Effect of C-D Termination
Yamamoto et al. 8 ,9 investigated the effect of
a C-D termination designed for shock-free flow at
Mj = 1.44 (pressure ratio, Pj/P a = 3.30) on shock
noise for an annular plug nozzle with a throat
radius ratio of 0.85. The exit radius ratio of the
C-D nozzle was 0.79. Experiments were conducted
over a range of pressure ratios from 2.94 to 3.54,
covering both over-expanded and under-expanded con-
ditions, under static and simulated flight condi-
tions. Results for a typical jet temperature,
Tj = 960 K, are shown in Fig. IS, where the OASPL
is plotted against log a for a forward-quadrant
angle, e = 50·. Data for the C-D termination are
compared with data for the convergent termination
and with data for the baseline convergent circular
nozzle. Perhaps the most significant observation
is that the C-D termination does not provide any
significant reduction in noise compared with the
convergent annular plug nozzle, although both
annular plug nozzles provide suppression, -6 dB,
relative to the baseline convergent circular nozzle.
Directivity comparisons shown in Fig. 16 and
4
spectral comparisons at e = 50· (shown in Fig. 17)
support this observation.
The reason for the ineffectiveness of the C-D
termination is that, as shown by laser velocimeter
(LV) measurements, near the C-D design point both
the convergent and C-D nozzles produce shocks down-
stream of the plug, and the shock noise generated
in that region exceeds that generated in the plug
region. As the LV results also showed, even though
shocks on the plug were eliminated by the C-D ter-
mination, the plume velocity at the plug tip was
still supersonic. Consequently the truncated plug
tip produces a series of expansion and shock waves
downstream which interact with tha turbulent mix-
ing layer to produce shock noise. From the spec-
tral comparisons in Fig. 17, it can be seen that
the C-D termination does provide some noise reduc-
tion at high frequency, which is attributable to
the suppression of shock noise in the plug region.
However, in the aft quadrant the directivity com-
parisons of Fig. 16 indicate that the C-D termina-
tion produces increased noise; this result is
consistent with the reduction in nozzle exit radius
ratio.
Effect of Plug Tip Geometry
The tests of Yamamoto et al. 8 ,9 did not in-
clude the effects of plU916iplgeometry. However,
Janardan, Yamamoto et al. ' investigated the
effect of plug tip geometry for a dual-stream co-
annular plug nozzle. The downstream-generated
shock noise was substantially reduced when a sharp
tipped plug was used instead of a truncated plug.
This effect should also be observed for the single-
stream annular plug nozzle.
Modeling
For purposes of predictive modeling, annular
plug nozzle shock noise is broken into two com-
ponents: noise generated in the premerged region
in the vicinity of the plug, and noise generated
downstream of the plug, as shown by Yamamoto et al. 8
This is illustrated in Fig. 18, where the spectra at
e = 50· for both the convergent and C-D termination
are compared with each other and with predictions
based on mo~~fied inputs to the Motsinger-Sieckman
(M-S) Tgdel (which is based on Harper-Bourne and
Fisher ). The aerodynamic conditions correspond
to the C-D design point. At low and middle fre-
quencies the two annular nozzles show little dif-
ference and agree well in spectral shape with
predictions taking the effective diameter to be
the nozzle equivalent diameter, De = ';4Aj/~. (The
level is arbitrary.) This noise 1S therefore
ascribed to the downstream region.
In the high frequency region the convergent
nozzle levels exceed those of C-D configuration,
which indicates that this noise is generated in the
plug region, where the C-D nozzle has no shocks.
The spectral shape agrees with the M-S model cg us-
ing the hydraulic diameter, Dh = 4A/~Do' as the
effective diameter.
Plug reqion. A theoretical study of shock
noise ~eneration in this region was conducted by
Balsa, ~gsed on extension of the Howe and Ffowcs
Williams model. Bals9 found that the SPL should
increase with 10 log a as predicted by other
models. He also found a directional dependence,
SPL ~ 10 log (1 - Mc cos e) as also found by Howe
and Ffowcs Williams 26 but which is not found in where F(e - eM) is given by Eq. (1a), and Md p
predici~ons based on the Harper-Bourne and Fisher is ~qual to the design Mach number if an extended
model. However, because of the difficulty in plug is used and is 1.0 if a truncated plug is
separating out the effect of high frequency jet used. The appropriate nondimensional frequency
mixing noise, it is uncertain whether this rather parameter, to be used in conjunction with Fig. 10,
weak directivity is actually present. is given by
The observations of Yamamoto et al. 8 can be
incorporated into simple empirical prediction meth-
ods, such as the circular nozzle model proposed
So = (!l)
\0.7 l. . /IM -
Vj
2
j
M2 I
d,p
[1 - M0 cos (e + e a )l
J
earlier in this pap
Stone ahd Montegani ZS '
which is based on that of
(w~~ch, in turn, is based on
Harper-Bourne and Fisher ). In order to get
proper limiting behavior with a two-source model,
the constant in Eq. (1) must be reduced by 3 dB.
x E G:)
+ 0.7 cos e] 2 + 0.0196 G:)2 (6)

An additional level correction is also required; Coannular Nozzles

the form 6SPL« log [1 - (Di/Do)Ex] gives the
proper limiting behavior, and the proportionality
constant is about 10. The following expression is Most modern jet engines are of the dual-
then obtained for the overall sound pressure level: stream type and have coannular exhaust nozzles.
Therefore, information on the noise characteristics

OASPL p = 159 , 10 log (~::,)' ~;:A) 1 of such nozzles is quite important. The dual-
streaml0h~1k noise studies of Janardan, Yamamoto
et al. ' included convergent and C-D coannular
plug nozzles with both truncated and extended plugs

r-(~:)£J '
(as shown in Fig. 19). Both the coannular nozzles

(:~
had an inner-stream radius ratio of 0.93 and an
, 10 log 10 log outer-stream radius ratio of 0.85.
Influence of Screech

+ 10 log
(M~J - M~)2 Since screech was shown by Yamamoto et al. 8 to
be less for an annular nozzle than for a circular

11 _ (M~ _ M~)2
nozzle and to decrease with increasing temperature,
it was not expected to be a critical problem in
these experiments, particularly at jet temperatures
+ 10 log [1 - Mo cos (e + e a~ of practical interest. The narrowband spectra
shown in Fig. 20 confirm this expectation. Moder-
ate levels of screech are seen at moderate tempera-
+ F(e - eM) (3) tures under both static (Fig. 20(a)) and simulated
flight (Fig. 20(b)) conditions. The higher temper-
where F(e - eM) is given by Eq. (la). The appro- ature spectra shows some evidence of screech (at
priate nondimensional frequency parameter, to be -1200 Hz) under static conditions (Fig. 20(c)) but
used in conjunction with Fig. 10, is given by none in simulated flight (Fig. 20(d)). Thus, it is
concluded that screech is not a critical problem,
but care must be exercised in analyzing the data,
Sp (dv;) ~II~~ - M~I [1 - Mo cos (e + e a~ especially for temperature effects.
Effect of Temperature

x ~ + 0.7 GD cos ef + 0.0196 G:Y (4)

The effect of outer-stream temperature for
the convergent coannular nozzle is shown in
Fig. 21. OASPL directivities are shown for two
different outer-stream temperatures under both
Downstream region. Again the observations of static (Fig. 21(a)) and simulated flight
Yamamoto et al. 8 are incorporated into the model of (Fig. 21(b)) conditions. The increased aft quad-
Stone and Montegani~5, and limiting behavior is rant noise at high temperature is due to increased
taken into account. The following expression is jet mixing noise. In the forward quadrant the
then obtained for the OASPL: reverse effect is seen, but is of relatively small
magnitude. This increase at lower temperature is

r~ ~ a )41
probably due largely to the increased screech, and
OASPL = 159 + 10 log a )2 the broadband shock noise is relatively insensitive
D l~ISA
P
\cISA
C
J to temperature.
Effect of C-D Termination
Janardan, Yamamoto et al. 10 ,11 investigated the
+ 10 log (:}) + 10 log
1+ (M~ _ M2 \ 2 effect of C-D terminations designed for shock-free
flow at an inner-stream Mach number, Mj 1 = 1.38,
\ J d, p)
and an outer-stream Mach number, Mj 2 ='1.44. At
- 10 log [1 - Mo cos (e + e a)] + F(e - eM) the exit the inner-stream radius raelo was 0.91,
and that of the outer-stream was 0.79. Experiments
(5) were conducted over a range of over-expanded and
under-expanded conditions for both streams, under

5
static and simulated flight conditions. Typical
results for the truncated plug are shown in Fig. 22,
where the OASPL is plotted again~t 10 log Sef for
a forward-quadrant angle, e ~ 60. (The effective
shock strength, Sef, is calculated from an area-
weighted pressure ratio, as defined subsequently in
the "Modeling" discussion.) Data for the C-D con-
figuration are compared with data for the conver-
gent coannular nozzle; also shown for comparison
are baseline convergent circular nozzle data scaled
to the fully-mixed conditions of the dual-stream
nozzles. It can be seen that at the C-D design
point the C-D nozzle provides about 2 dB reduction
under static conditions and about 5 dB reduction in
simulated flight relative to the convergent co-
annular nozzle. The convergent coannular nozzle
itself provides about 5 dB reduction under static
conditions and about 6 dB reduction in simulated
flight relative to the fully-mixed convergent
circular nozzle.
OASPL comparisons for these configurations at
the C-D design point are shown in Fig. 23. The C-D
termination is seen to reduce noise throughout the
forward quadrant. The noise increase in the aft
quadrant with the C-D termination is due to the
effect of the decreased radius ratios at the exit.
The correspondin~ forward quadrant spectral compar-
isons at e ~ 60 are shown in Fig. 24. Broadband
shock noise is clearly present even with the C-D
terminations, but its level is reduced, with little
or no frequency shift. The significant residual
shock noise with the C-D configuration will be
shown to be due to plug tip geometry.
Effect of Plug Tip Geometry
Becaus g of the residual shock noise observed
for annular and coann~6ar10 nozzles with truncated
plugs, Janardan et al. investigated the effect
of plug tip geometry. Typical results are shown
in Fig. 25, where the OASPL is plotted against
10 log Bef for a forward quadrant angle, e ~ 60·.
Data for the nozzle with C-D terminations and an
extended plug are compared with both the c-o coan-
nular nozzle with truncated plug and the convergent
coannular nozzle with truncated plug. It can be
seen that the extended plug provides an additional
1.5 dB (static) to 3 dB (simulated flight) suppres-
sion relative to the C-D nozzle with truncated
plug. Compared to the convergent nozzle with trun-
cated plug, the C-D nozzle with extended plug pro-
vides 2.5 dB suppression under static conditions
and 7 dB under simulated flight conditions.
OASPL directivity comparisons at the C-D
design point are shown in Fig. 26. The extended
plug is seen to provide additional noise reduction
throughout the forward quadrant and has no signifi-
cant effect in the rear quadrant, where jet mixing
noise is dominant. The corresponding forward quad-
rant spectral comparisons at e ~ 60· are shown in
Fig. 27. It appears that there may be some shock
noise still present with the extended plug, but if
so its contribution is relatively small.
Effect of Subsonic Inner Stream
Some candidate engine cycles for advanced
supersonic transports feature coannular exhausts
with a supersonic outer stream and subsonic inner
stream (e.g., Refs. 2, 22, and i8 i£ 33). There-
fore, Janardan, Yamamoto et al. ' conducted ex-
periments at such conditions for the convergent
6
coannular nozzle with the extended plug. The ef-
fect of the subsonic inner stream conditions is
illustrated in Fig. 28, where the OASPL is plotted
against 10 log Sef for a forward quadrant angle,
e ~ 60 , for the convergent coannular nozzle at
subsonic and supersonic inner stream conditions.
The subsonic inner stream provides a significant
noise reduction over a range of effective Mach
number from 1.35 to 1.5 in simulated fliqht and
over a wider range under static conditions. OASPL
directivity comparisons are shown in Fig. 29 for an
effective jet Mach number of -1.4, where the bene-
ficial effect of the subsonic inner stream is most
pronounced. The subsonic inner stream provides a
noise reduction throughout the forward quadrant.
Near the jet axis where jet mixing noise is predom-
inant, there is no significant change. The corre-
sponding spectral comparisons at e ~ 60· are shown
in Fig. 30. A significant reduction in broadband
shock noise is quite evident. These results are
consistent with diagnostic LV measurements, which
indicated a significant weakening of the shock
structure with a subsonic inner stream.
Modeling
For purposes of predictive modeling, coannu-
lar plug nozzle shock noise is broken into two
components, like the annular plug nozzle: noise
generated in the vicinity of the plug, and noise
generated downrtream of the plug, as shown by Ja
nardan et al. 0 Janardan developed modl~ications
to the Harper-Bourne and Fisher approach to ac-
count for his findings, which 2g re now applied to
the Stone and Montegani model using a similar
approach to that used for the annular plug model.
Plug region. Since the shocks on the plug are
due mainly to the outer stream expansion, the pre-
diction is based on outer stream conditions. The
following expression, derived from Eq. (3), is then
obtained for the overall sound pressure level:
~:;sJ ('::AJ']
OASPL, 159 • 10 log
r-
• 10 log (~:::t] ·10 log (:~
rM2 M2]2
+ 10 1og - -lj,2- d,2 \ Z
1 - (Mt2 - M~,2)
(7)
The appropriate nondimensional frequency parameter,
to be used in conjunction with Fig. 10, is given by
S _ (fDh'2 )
P - 0.7 Vj
x ~ + 0.7 (V{;2) cos sf + 0.0196 G:Y (8)
 Downstream region. Janardan 10 showed that the to determine the differences between these two
downstream shock noise depends on the conditions of approaches.
both streams in a rather complicated manner. The
effective pressure ratio is given by the appropri- Multielement Suppressors
ate relation, as follows:
Multielement suppressor nozzles may very well
p. f P. 1 be required to provide sufficiently low jet mixing
J,e J, for P, I >P' 2 ( 9a) noise for. advanced supersonic transports. 14 ,33-37
r;-=~ J, - J,
When a multielement nozzle is used to suppress jet
mixing noise, the shock noise can become a more
important source relative to jet mixing noise.
Therefore, shock noise reduction for suppressors is
of interest. Both Sing1e-lbreamS and outer-stream
(in a dual-stream exhaust) suppressors were in-
vestigated. Because of the complicated nozzle exit
geometries, no strong feedback loop is established,
for P.J, 1 < P.J, 2 and screech is not an important factor. Because of
the rapid mixing with these nozzles, no significant
Using the fully-mixed jet total temperature, the effect of plug tip geometry is expected.
effective Mach number, Mj ef, is calculated along Effect of C-D Termination
with the effective jet ve'ocity, Vj ef' The values
are then substituted, along with total area, The effect of C-D termination of the nozzle
Aj 1 + Aj 2, into Eqs. (5) and (6). The resulting elements was investigated for both single-streamS
expression for OASPL is then given by and dua1-stream lO nozzles.

[~;:A) ~;:S]
2 Single-stream. Yamamoto et a1. 8 ,9 investiga-
OASPlO 159 + 10 109
ted the effect of C-D suppressor elements, designed
for shock-free flow at Mj = 1.42, on a 20-chute
annular plug suppressor nozzle (Fig. 31), having a

+ 10 109 [h.,,: A;.2)1 suppressor area ratio of 1.75 and radius ratio of
0.76 at the throat. Results for a typical jet tem-
perature, Tj ~ 950 K (shown in Fig. 32) where the
OASPL is plotted against 10 log a for a forward
quadrant angle, e = 50·. Data for the C-D termi-
nations are compared with data for the convergent
~
2
M2 _ M2 terminations and with data for the baseline conver-
+ 10 log j,ef d,p] gent circular nozzle. The C-D elements clearly
1 + (M2j,ef - M2)2 provide additional shock noise suppression beyond
d,p that provided by the convergent suppressor in the
simulated flight case. Corresponding spectral com-
parisons at the C-D design conditions (shown in
Fig. 33) show the C-D elements effectively suppress
the shock noise, which peaks at 1250 Hz. However,
( 10)
at middle frequencies the C-D elements produce
some increase in noise. OASPL directivitr compari-
son at these conditions (shown in Fig. 34t and aft-
where Md p. is calculated by substituting the quadrant spectral comparisons at e = 140 (shown
design conditions into Eq. (9) and taking the re- in Fig. 35) clarify the situation. In the aft
sulting Mach number as the effective Md p for the quadrant jet mixing noise is dominant, and the C-D
extended plug case, or by setting Md ~ 1 for a suppressor is significantly noisier than the con-
truncated plug. The appropriate nondfgensional vergent suppressor. Laser ve10cimeter measurements
frequency parameter, to be used in conjunction with confirm the expectation that the weakened shocks
Fig. 10 is given by produce less rapid mixing and, consequently, more
jet noise. This is not surprising since one method
of enhancing jet mixing and noise reduction is to

~
S _ fD e ) promote a strong normal shock near the nozzle. 2 ,34
D- O. 7 Vj ,ef
Dual-stream. Janardan, Yamamoto et al. 10 ,11
investigated the effect of C-D suppressor elements
x and C-D inner-stream nozzle on a 20-chute outer-

r
stream suppressor coannular plug nozzle shown in
Fig. 36. The suppressor area ratio was 1.75, and

,r + 0.7 (' ~;'f) co,
where De is based on the total nozzle area. In-
stead of using formulations of this type for the
C-D termination effect, Janardan calculated the
noise for M~ = 1 and developed a set of empirical
corrections. 0 No comparison has yet been made
+ 0.0196 (V d the suppressor radius ratio was 0.76 at the throat;
the inner-stream nozzle had a radius ratio at the
throat of 0.94 and an area constituting 0.2 of the
total exit area. Results for typical conditions
are shown in Fig. 37, where OASPL is plotted
( 11)
against 10 log 6ef for a forward quadrant angle,
e = 60 • Data for the C-D terminations are com-
pared with data for the convergent terminations and
with data for the baseline convergent circular
nozzle at fully-mixed dual-stream conditions. The
effect of the C-D terminations is that the C-D

7
elements provide no additional shock noise OASPL
suppression beyond that of a convergent suppressor,
although perceived noise levels (not shown) are
slightly reduced. The effect of C-D terminations
on the SPL spectra (shown in Fig. 38) and the OASPL
directivity (shown in Fig. 39) illustrate the same
general effects as for the single-stream suppres-
sor: while there is some reduction in shock noise,
especially downstream, the jet mixing noise is in-
creased. Thus it appears that while C-D suppressor
elements may reduce shock noise, they may not be
practical for aircraft applications because of the
increased jet mixing noise.
Modelinq
For purposes of predictive modeling, the shock
noise generation for the suppressor nozzle is
broken into two components: premerged noise gener-
ated near the nozzle exit where the individual ele-
ments of the flow are discernible, and the noise
generated in the downstream mixing region.
Single-stream suppressors. Because of the
rapid mixing of the flow elements with downstream
distance, the downstream region is subsonic and
there is no dow~stream shock noise generation.
Yamamoto et al. compared the experimental results
with P~9dictions based on modified inputs to the M-S
model. Reasonable agreement was found using Dh
as the characteristic dimension. Thus Eqs. (3) and
(4) should be appropriate, where Dh is the ele-
ment hydraulic diameter, and Di and 0 are the
inside and outside diameters, respective?y of
circles intersecting the innermost and outermost
positions of the suppressor elements.
Dual-stream suppressors. The same approach
should be applicable for premerged shock noise as
used for the single-stream suppressor. Therefore,
Eqs. (7) and (8) should be appropriate. For the
downstream region, only the inner stream should
contribute because of the rapid mixing of the
outer stream. Applying this reasoning, we obtain
[~;;S ~:;S]
OASPLO 159 + 10 " ,
(A~21) f.\M j2,1 2,l)2
- Md
+ 10 log + 10 log
- 10 log ~ - Mo cos (e + ea~ + F(e - e M,l)
(12)
The appropriate nondimensional frequency parameter,
to be used in conjunction with Fig. 10 is given by
So (O.~D~~\)
x IM~,l-M~,ll f-MoCOS(e+ea~
, r 0.7 C;;I))'
+ + 0.0196 (V;;')'
Conclusions
The most important result of these studies is
that supersonic jet shock noise can be reduced by
proper C-D design of nozzle flow passages and cen-
ter plugs, even for the complicated nozzle geome-
tries investigated experimentally. Furthermore,
the noise reduction is obtained over a fairly wide
range of pressure ratios above and below the C-D
shock-free design point. This finding was predic-
ted theoretically for the Single-stream circular
C-D nozzle and also found experimentally to be true
for the more complicated geometries. However, it
was found that the shock noise reductions were
generally accompanied by some increase in jet mix-
ing noise (except for the circular nozzle); this
effect was especially pronounced for the suppressor
nozzles. Thus, it is clear that tradeoffs will be
required in applying the C-D nozzle approach to
noise reduction in most practical situations, where
jet mixing noise must also be considered.
References
1. Seiner, J. M., and Narum, T. D., "Experiments
on Shock Associated Noise of Supersonic Jets,"
AIAA Paper 79-1526, July 1979.
2. Dosanjh, D. S., and McAfee, R. S., Jr.,
"Coaxial Supersonic Jet Flow Shock Structure
and Related Limitations of Comparative Noise
Suppression Assessment Schemes," AIAA Paper
83-0707, April 1983.
3. Tam, C. K. W., and Jackson, J. A., "On Shock
Cell Structure and Noise of Supersonic Jets,"
AIAA Paper 83-0703, April 1983.
4. Tam, C. K. W., "On the Shock Cell Structure
and Noise of Supersonic Jets," Proc. 11th
International Conference on ACOUStlCS, Vol.
l, Paris, France, July 1983, pp. 333-336.
5. Nagel, R. T., and Papathanasiou, A. G., "An
Experimental Study of Supersonic Jet Shock
Associated Noise," AIAA Paper 83-0708, April
1983.
6. Nagel, R. T., Denham, J. W., and
Papathanasiou, A. G., "Supersonic Jet Screech
Tone Cancellation," AIAA J., Vol. 21, No. 11,
\ Nov. 1983, pp. 1541-~
7. von Glahn, U., "New Interpretations of Shock-
Associated Noise With and Without Screech,"
NASA TM-81590, 1980.
8. Yamamoto, K., Brausch, J. F., Balsa, T. F.,
Janardan, B. A., and Knott, P. R.,
"Experimental Investigation of Shock-Cell
Noise Reduction for Single Stream Nozzles in
Simulated Flight," to be published NASA
Contractor Report.
9. Yamamoto, K., Brausch, J. F., Janardan, B.
A., Hoerst, D. J., Price, A. 0., and Knott,
P. R., "Experimental Investigation of Shock-
Cell Noise Reduction for Single-Stream Nozzles
in Simulated Flight, Comprehensive Data
Report, Volumes I-III," General Electric Co.
(13)
Report R82AEB491, May 1984. (NASA CR-168234)
8
10. Janardan, B. A., Yamamoto, K., Majjigi, R.
K., and Brausch, J. F., "Experimental
Investigation of Shock-Cell Noise Reduction
for Dual-Stream Nozzles in Simulated Flight,"
to be published NASA Contractor Report.
11. Yamamoto, K., Janardan, B. A., Brausch, J.
F., Hoerst, D. J., and Price, A. 0.,
"Experimental Investigation of Shock-Cell
Noise Reduction for Dual-Stream Nozzles in
Simulated Flight, Comprehensive Data Report,
Volumes I-II," General Electric Co. Report
R83AEB358, Feb. 1984. (NASA CR-168336)
12. Maestrello, L., "An Experimental Study on
Porous Plug Jet Noise Suppressor," AIAA Paper
79-0673, Mar. 1979.
13. Bauer, A. B., "Jet Noise Suppression by Porous
Plug Nozzles," AIAA Paper 81-1993, Oct. 1981.
14. FitzSimmons, R. D., McKinnon, R. A., and
Johnson, E. S., "Flight and Tunnel Test
Results of the MDC Mechanical Jet Noise
Suppressor Nozzle," su~ersonic Cruise Research
~, Part I, NASA CP-2 08, 1980, pp. 453-478.
15. Harper-Bourne, M., and Fisher, M. J., "The
Noise from Shock Waves in Supersonic Jets,"
Noise Mechanisms, AGARD-CP-131, Mar. 1974,
pp 11-1 to 11-13.
16. Tam, C. K. W., and Tanna, H. K., "Shock
Associated Noise of Supersonic Jets from
Convergent-Divergent Nozzles," J. Sound
Vibration, Vol. 81, 1982, pp. 337-358.
17. Seiner, J. M., and Norum, T. D., "Experiments
on Shock Associated Noise of Supersonic Jets,"
AIAA Paper 79-1526, July 1979.
18. Seiner, J. M., and Norum, T. D., "Aerodynamic
Aspects of Shock Containing Jet Plumes," AIAA
Paper 80-0965, June 1980.
19. Powell, A., "On the Mechanism of Choked Jet
Noise," Proc. Phys. Soc., Sec. B., Vol. 66,
1953, pp. 1039-1056.
20. Tester, B. J., Morris, P. J., Lau, J. C., and
Tanna, H. K., "The Generation, Radiation and
Prediction of Supersonic Jet Noise,"
AFAPL-TR-78-85-Vol. 1, Oct. 1978.
21. Tanna, H. K., "An Experimental Study of Jet
Noise, Part II: Shock Associated Noise," ~
Sound Vibration, Vol. 50, No.3, 1977, pp.
429-444.
22. Kozlowski, H., and Packman, A. B., "Fl ight
Effects on the Aerodynamic and Acoustic
Characteristics of Inverted Profile Coannular
Nozzles," NASA CR-3018, 1978.
23. Norum, T. D., and Seiner, J. M., "Broadband
Shock Noise from Supersonic Jets," AIAA J.,
Vol. 20, 1982, pp. 68-73 (AIAA Paper 80-0983).
24. Norum, T. D., "Screech Suppression in
Supersonic Jets," AIAA Paper 82-0050, Jan.
1982.
9
25. Stone~ J. R., and Montegani, F. J., "An
Improved Prediction Method for the Noise
Generated in Flight by Circular Jets," NASA
TM-81470, 1980.
26. Howe, M. S., and Ffowcs-Williams, J. E., "On
the Noise Generated by an Imperfectly Expanded
Supersonic Jet," Phil. Trans. Roy. Soc.
London, Vol. A289, No. 1358, 1978, PP.
m=TI4.
27. Pao, S. P., and Seiner, J. M., "A Theoretical
and Experimental Investigation of Shock
Associated Noise in Supersonic Jets," AIAA
Paper 81-1973, Oct. 1981
28. Pao, S. P., and Salas, M. D., "A Numerical
Study of Two-Dimensional Shock Vortex
Interaction," AIAA Paper 81-1205, June 1981.
29. Gliebe, P. E., Motsinger, R. E., and
Sieckman, A., "High Velocity Jet Noise Source
Location and Reduction, Task 6 Supplement -
Computer Programs," FAA-RD-78-79, Vol. la,
Mar. 1979.
30. Tanna, H. K., Tam, C. K. W., and Brown, W.
H., "Shock Associated Noise Reduction from
Inverted-Velocity-Profile Coannular Jets,"
Lockheed Report No. LG81EROI62, Aug. 1981.
( NASA-CR-3454)
31. Stone, J. R., "An Empirical Model for
Inverted-Velocity-Profile Jet Noise
Prediction," NASA TM-73838, 1977.
32. Goodykoontz, J. H., and Stone, J. R.,
"Experimental Study of Coaxial Nozzle Exhuast
Noise," AIAA Paper 79-0631, Mar. 1979. (NASA
TM-79090)
33. Stone, J. R., Goodykoontz, J. H., and
Gutierrez, O. A., "Effects of Geometric and
Flow-Field Variables on Inverted-Velocity-
Profile Coaxial Noise and Source
Distributions," AIAA Paper 79-0635, Mar.
1979. (NASA TM-79095)
34. Huff, R. G., and Groesbeck, D. E., "Cold-Flow
Acoustic Evaluation of a Small-Scale,
Divergent, Lobed Nozzle for Supersonic Jet
Noise Suppression," NASA TM-X-3210, 1975.
35. Simcox, C. D., Armstrong, R. S., and Atvars,
J., "Recent Advances in Exhaust Systems for
Jet Noise Suppression of High Speed Aircraft,"
J. Aircraft, Vol. 13, No.6, June 1976, pp.
442-448.
36. Stone, J. R., Miles, J. H., and Sargent, N.
B., "Effects of Forward Velocity on Noise for
a J85 Turbojet Engine with Multi-Tube
Suppressor from Wind Tunnel and Flight Tests,"
NASA TM-X-73542, 1976.
37. Moore, M. T., "Flight Effects on the Jet Noise
Signature of a 32-Chute Suppressor Nozzle as
Measured in the NASA Ames 40x80 Foot Wind
Tunnel," NASA CR-152175, 1979.
OBSERVER

1
Figure 1. - Representation of a family of shock-cells.

+
L INTERFERENCE
SPECTRUM

\
\
\
L GROUP
SPECTRUM

I
\
I
L TOTAL
(COMPOSITE)
SPECTRUM

FREQUENCY, f, kHz
Figure 2. -Illustration of primary components of a
convergent circular nozzle shock-cell noise
spectrum.
 105

100

co
"C
95
....J-
c...
Vl
«
0
....J-
LLJ 90
> (a) Static.
~
w
c:::: 110
=>
Vl
Vl
W
c::::
c...
0
z 105
=>
0
Vl
....J
....J
«
c::::
w 100
(3

95

90~----------~----------~----------~----------~----------~
-1. 5 -1.0 -.5 0 .5 1.0
SHOCK STRENGTH PARAMETER; 10 log ~
(b) Simulated flight, Vo = 122 m/sec
I I I
1.3 1.4 1.5
JET MACH NUMBER, Mj

2. 6 2. 8 3. 0 3. 2 3. 4 3. 8
NOZZLE PRESSURE RATIO, P/P a
Figure 3. - Typical variation of shock noise OASPL in
forward quadrant (8 = 50°) as function of ~ for con-
vergent circular nozzle. Data scaled to 0.903-m 2
nozzle area and extrapolated to 730-m sideli ne; jet
temperature Tj ~ 950 K (ref. 8).
 ....
,..,....,.........
I- 3A 14 - - - - - - . l
STANDING WAVE
PATTERN

,- - - - - - . """----
,, . "."" , " "" "" ",,"
.... ,,
-=""'----

" ...... ---

---t--*'-

/
/
".
---~~><'-~--

. /

...... ,
- - - - - - .:><:.::::-....-, - - - - -
/
/ .",. . . . ,, ..... ...

........
. ...... ':>/!'

I
I
/

I - - -
7(

- --
-
........ I
.... -....L SHOCK CELLS
REFLECTOR SURFACE

Figure 4. - Schemat ic diagram of nozzle exit and reflector showing

the foam cancellation technique (ref. 6).

/
,,-- WITH SCREECH
'"
co
"0

._i
LLI /- CANCELLATION METHOD
g .... """,' ..... L
/

e::::
LLI

e3c..

o 10 20 30 40
FREQUENCY, kHz
Figure 5. - Narrowband noise spectra from the reverberation room
comparing screech cancellation to screech reduction with tab,
Mj = 1. 4 (ref. 6).
 105

----
CONVERGENT
CIRCULAR ""',
100 "

co
-0 95
....J'
C-
Vl
c::(
o
~' 90~----~-----L----~~----~----~
(a) Static.
~
\...LI

§5
Vl
110
Vl
\...LI
~ CONVERGENT
C-
o CIRCULAR --..., . .
5o 105
Vl /
....J
....J I
c::( I
~ I
\...LI I
6 100 I

11. 0 dB I
d
I

95
__ ~-_-A .b/
,l-C-D CIRCULAR
-u- - -.../:l
"'tl-~ 1 t::r
90~----~-----L----~~----~----~
-1. 5 -1.0 -.5 0 .5 1.0
SHOCK STRENGTH PARAMETER; 10 log ~
(b) Simulated flight, Vo = 122 m/ sec
I I I
1.3 1.4 1.5
JET MACH NUMBER, Mj

2. 6 2. 8 3. 0 3. 2 3. 4 3. 8
NOZZLE PRESSURE RATIO, P/P a
Figure 6. - Effectiveness of C-D termination in reducing
circular nozzle shock noise OASPL in forward quad-
ranUS = 500 ); jet temperature, Tj'~ 950 K. Data
scaled to O. 903-m 2 nozzle area and extrapolated to
730-m sideline (ref. 8).
 120

110
r::a
-0
.....J- 100
c..
VI
<C
o

gu:l 90 \
\
\... SUPPRESSION DUE TO C-D FLOW PATH
UJ
a:
:::>
V'l
80 L -__ ~ ____-L____- L_ _ _ _ ~ _ _ _ _L-__~
VI
UJ
a: (a) Static.
c..
~ 110
:::>
o
VI

80 40 80 100 120 140 160

60
ANGLE TO INLET, e, deg
(b) Simulated flight, Vo ~ 122 ml sec
Figure 7. - Comparison of OASPL directivity of convergent
and C-D nozzle at the C-D nozzle design conditions; jet
temperature, Tj = 950 K. Data scaled to 0.903-m 2 nozzle
area and extrapolated to 730-m sideline (ref. 8).
 no

•
co
"'C
100

° ., • ,
/
... -- CONVERGENT

....
...1
Cl... 90 °
VI
°eo
...1
LLJ
>
°
fJJ ~ °o-CONVERGENT
••
LLJ
-I

r:tnCOo • ~.,..
;I
LLJ
e:::: 80
::::> o
VI
VI
LLJ
e:::: CfJo
/-.
-•
Cl...
Cl
Z 70 ,Do C-D ../
/

::::> C-D.../
0
VI
00
0
60
Cb

50
• 05 . 25 1 4 16. 05 • 25 1 4 16
1I3-0CTAVE-BAND CENTER FREQUENCY, f, kHz
(a) Static. (b) Simulated flight (Vo~
122 m/sec.
Figure 8. - Comparison of typical front quadrant (9 = 500 ) spectra
of convergent and C-D nozzles at the C-D nozzle design condi-
tions; jet temperature, T· ~ 950 K. Data scaled to 0.903-m 2
nozzle area and extrapolaled to 730-m sideline (ref. 8).
 TAM AND TANNA 16
Mj· 1. 925~ Md so 1.67
NUMERICAL CALCULATION, 3,4
VORTEX SHEET MODEL16 I

105

co
105
"0

~
!::
VI 105
z
LLJ
I-
~
105

105
.8 3.15 12.5
1/3-0CTAVE CENTER FREQUENCY, kHz
Figure 9. - Variation of shock associated noise peak fre-
quency with angle.

o
o
5 co
0"0
-10
VI •
0-1
Z 0.. -20
«VI
co«
,0
LLJ~
> 0.. -30
~VI
~ -I.
,LLJ -40
""2i
;:;-1
oLLJ

~~ -50
<i~
~o:::
0::: 0.. -60
o
z:
-70UL~~~--~~~---L---L--~---I
-1.4 -1.0 -.6 -.20.2 .6 1.0 .1.4 1.8
LOGARITHMIC FREQUENCY PARANETER, log S

Figure 10. - Recommended l/3-octave band spectrum

for shock noi se.
 (a) Convergent termination.

(b) Convergent-divergent (C-D) termination.

Figure 11. - Annular plug nozzle configurations.

FILTER
BANDWIDTH,
Hz
120
1.25 12. 5
co
"0
100
...1
a.. 80
V')

...1
UJ
> 60
~
UJ (a) Without tabs.
c:::
::::>
V')
V')
UJ 120
c:::
a.. 1.25 12. 5
CI
:z: 100
::::>
0
V')
80
937.5 Hz
60
500 1500 0 5K
FREQUENCY, Hz
(b) With tabs.
Figure 12. - As-measured narrowband spectra at 9 = 600 for con-
vergent annular plug nozzle with and without tabs: moderately
heated jet, Tj ~ 480 K, at pressure ratio, Pj/Pa '" 3.4 (ref. 8).
 Pj/Pa
~. Vj.
m/sec
0 NO TABS 3.44 484 579

100
• WITH TABS 3.41 481 576

0
co 90 0
-c

:~,
J
a..
VI

J
L1J
> 80
~
0 i
•
L1J
c:::
=>
VI
•
•
VI
L1J
c:::
a..
0
:z
70
•
=>
i
0
VI
60 •
50
50 250 lK 4K 16K 50 250 lK 4K 16K
1/3-0CTAVE-BAND CENTER FREQUENCY. f. Hz
(b) e= 140°.
Figure 13. - Effects of screech tabs on SPL spectra: convergent
annular plug nozzle (moderately heated jet). Static data scaled
to O. 903-m 2 nozzle area and extrapolated to 730-m sideline
(ref. 8).
 FILTER
BANDWIDTH,
120 Hz
1.25 12.5
100
~
co
-0

...i
a.. 80
Vl

...i
L.l.J 60
> 975 1275 0 5K
~
L.l.J FREQUENCY, f, Hz
c::::
=>
Vl
Vl
(a) With no tabs.
L.l.J
c::::
a..
120
0
z 1. 25 12.5
=> 100
0
Vl ~
80

60
1025 1275 0 5K
FREQUENCY, f, Hz
(b) With tabs.
Figure 14. - As-measured narrowband SPL spectra at e= 600
for convergent annular plug nozzle with and without tabs;
high temperature jet, Tj =960 K, pressure ratio, Pj/Pa =
3. 34 (ref. 8).
 ANNULAR
PLUG
110 NOZZLES
o CONVERGENT
I:::. C-D
105

100

~ 95 ~ !:::.
J <A
!:::.O~
0-
VI
C§ 90 !:::.
J
o
LLJ C-D DESIGN POINT
>
~
LLJ
c:::
=> (a) Static.
VI
en
LLJ
g: 110
c
Z
=>
o
VI
-I
105
-I
<x:
c:::
LLJ

5 100

95

C-D
90 DESIGN
POINT
~~--~----~~~----~----I
-1. 5 -1. 0 -. 5 0 •5 1. 0
SHOCK STRENGTH PARAMETER, 10 log ~

(b) Simulated flight, Vo = 122 m/sec

Figure 15. - Comparison of C-D annular plug
nozzle OASPL with those of convergent
annular plug nozzle and baseline convergent
circu lar nozzle in forward quadrant, 9 .. 500 ;
jet temperature, T· ~ 960 K. Data scaled to
O. 903-m 2 nozzle atea and extrapolated to
730-m sideline (ref. 9),
 130
o BASELINE CONVERGENT CIRCULAR NOZZLE
o CONVERGENT ANNULAR PLUG NOZZLE
l:::. C-D ANNULAR PLUG NOZZLE
120

o
o ~ 8
~ no o
0
J
c..
0 0 o 0
Vl o (\
C3 100
J
l.J.J
>
::I
l.J.J 90
c:::
::::> (a) Static.
Vl
Vl
l.J.J
g: 120
o
z

••••• • •• a.
::::>
~
---l
no
•
---l
«c:::
: *
l.J.J

6 100 • * ** * •
• BASELINE CONVERGENT CIRCULAR NOZZLE
90 • CONVERGENT ANNULAR PLUG NOZZLE
A C-D ANNULAR PLUG NOZZLE

80 40 60 80 100 120 140 160

ANGLE TO INLET, 8, deg
(b) Simulated flight, Vo = 122 mIsec.
Figure 16. - Comparison of directivity of C-D annular
plug nozzle with those of convergent annular plug
nozzle and baseline convergent circu lar nozzle at
C-D design conditions, T· ~ 960 K. Data scaled to
0.903-m Z nozzle area and extrapolated to 73O-m
sideline (ref. 81.
 o BASELINE CONVERGENT
CIRCULAR NOZZLE
o CONVERGENT ANNULAR
PLUG NOZZLE
6. C-D ANNULAR PLUG
NOZZLE
no

•
......
100

o •
a:l
"'0
...1
a..
VI
...1
I.J..J
90

~e
°Cb
o
o ~ .
"If#: •
• !t\:....
> 0
~
I.J..J
80
~~~ 000

~O
0::
::>
V'l
VI
I.J..J
0:: lS>O
a.. ro

.
0
z lX:lJ A
70
•
::>
0 6.
VI
0
0
lS.
o
60 6.
o
•
o
50
50 250 1K 4K 50 250 1K 4K
1I3-0CTAVE-BAND CENTER FREQUENCY, f, Hz
(a) Static. (b) Simulated flight,
Vo = 122 m/sec.
Figure 17. - Spectral comparison between C-D annular
plug nozzle, convergent annular plug nozzle, and
baseline convergent circular nozzle at C-D design
conditions; forward quadrant angle, e= 500 ; jet
temperature, T· =' 960 K. Data scaled to 0.903-m 2
nozzle area and extrapolated to 730-m sideline
(ref. 8).
 h,
~'
MODEL P/Pa De'
cm cm

CONVERGENT 3. 16 967 14.4 2.06

CONVER GENT-D IVERGENT 3.12 962 14.4';' 2.06';'

';'BASED ON THROAT AREA AND HEIGHT

100

co rSCREECH .
"'C 90
/ r EXPERIMENTAL
J
c.. ! (CONVERGENT)
Vl 80 I

J
I..LJ
> / ", . . . ~ EXPER IMENT
I..LJ
-I 70 J
I..LJ
B/ (C-D)
c:::
=>
Vl
Vl
60 / SHADED AREA
I..LJ PREDICTION PREDICTION
c::: REPRESENTS
c.. BASED ON De BASED ON Dh
0 50 C-D EFFECTIVE-
z
=> NESS
0
Vl 40

30
10 100 1000 10000
1/3-0CTAVE-BAND CENTER FREQUENCY, f, Hz
Figure 18. - Spectral comparison of shock-cell noise between
modified predictions and both convergent and C-D data,
illustrating spectral contribution of shocks formed on the
plug and downstream of the plug~ 9 = 50 0 ; static data scaled
to O. 903-m 2 nozzle diameter and extrapolated to 73O-m
sideline (ref. 8). Lines A and B are not real predictions,
but an illustration of prediction concept
[ ----
__ ---L-J] ~
(a) Inner and outer nozzle terminations convergent. with truncated plug.

(b) Inner and outer nozzle terminations convergent. with extended plug.

(c) Inner and outer nozzle terminations convergent-divergent. with

truncated plug.
C

(d) Inner and outer nozzle terminations covergent-divergent. with

extended plug.
Figure 19. - Coannular plug nozzle configurations (ref. 10),
 SCREECH AT SCREECH AT
1045 Hz 870 Hz

co
"0 ~
...1 10 dB
0...
VI T
...1
I..LJ
> (a) Moderately heated jet, (b) Moderately heated jet,
::I Pj ,2 /Pa= 3.41, \2" Pj,2 /P a = 3.41, \2=
I..LJ
c::
::J
480 K; static. 490 K; simulated flight,
VI
VI
I..LJ
c::
0...
Cl
Z
::J
0
VI

o 2 4 6 8 10 0 2 4 6 8 10
FREQUENCY, f, kHz

(c) High temperature jet, (d) High temperature jet,

Pj ,2 /Pa= 3.32, \2= P·,2 /Pa= 3.33, Tj,2=
940 K; static. 945 K: simulated flight,
Vo= 122 m/sec.
Figure 20. - On-line narrowband data obtained with convergent
coannular nozzle at two outer stream temperatures, with con-
stant inner stream conditions, Pj,l' Pa ~ 3.13, \ 1 ~ 480 K.
 120
o T· 2 =945K
6- T!' 2 = 483 K
J,
co 110
"0 0
......r
C-
Vl
c:x: O 0 6- H
0 100 0 6- ~
8 Q ~ 6-
.....r
I.J.J
6-
G:i
.....J
I.J.J
90
c:::
::l
(a) Static.
Vl
Vl
I.J.J
c::: llO
c-
o 0
Z
::l
0
Vl 100 e ee 8
6-
0 0
0
6- 6-
0
0
~
.....J
.....J
c:x:
c::: 6- 6- 0
I.J.J

0
> 90 6-

80 40 60 80 100 120 140 160

ANGLE TO INLET, e, deg
(b) Simulated flight, V0 = 122 m/sec.
Figure 21. - Effect of outer stream temperature on
OASPL-directivity of convergent coannu lar nozzle
with tru ncated plug.
 llO COANNULAR
TRUNCATED
PLUG
NOZZLES
105
o CONVERGENT
o C-D

100 FULLY-MIXED
CONVERGENT
CIRCULAR
NOZZLE ~-_ o
95 o
co
-c
....J.
o
c...
tr)
<C 90 o
0
....J.
UJ C-D DESIGN POINT
>
~
UJ 85 It I
e::
=> (a) Static.
tr)
tr)
UJ
e:: llO
c...
Cl
z
=>
0
V)
....J
....J
105
<C
e::
UJ
>
0 0
0
100 0 0
0 0
0
~oo
0
o 0
90 0
C-D DESIGN POINT

85~--------~--------~--------~--------~------....J
-1.5 -1.0 -.5 0 .5 1.0
EFFECTIVE SHOCK STRENGTH PARAMETER, 10 log ~ef
(b) Simulated flight, Vo = 122 m/sec.
Figure 22. - Comparison of C-D coannular truncated-
plug nozzle OASPL with those of convergent coannu-
lar truncated-plug nozzle and baseline fully-mixed
convergent circular nozzle in forward quadrant,
8=600 ; inner-stream temperature, T'l:480K;
outer-stream temperature, Tj 2~ 940 k.
Data scaled
to O. 903-m 2 total nozzle area and extrapolated to a
730-m sideline (ref. lll.
 110
o CONVERGENT
o CONVERGENT-
0 8
DIVERGENT
§ 0 8
~. 100 § 8
VI
«
o
d 90
~
LLJ
c::
~ 80L----L----L----L----L---~--~
VI
LLJ (a) Static.
c::
0...

~ 110
::>
o
VI
-l
<i
c::
100
LLJ

6
90L----L----L---~----~--~--~
40 60 80 100 120 140 160
ANGLE TO INLET, 8, deg
(b) Simulated flight, V0;' 122 ml sec.
Figure 23. - Effect of C-D terminations on coannular
nozzle directivity at C-D design point; Tj,l :: 485 K,
Tj 2:: 940 K. Data scaled to O. 903-m 2 total nozzle
area and extrapolated to 730 m sideline (ref. 10).
 100 COANNULAR
--0-- CONVERGENT AND
TRUNCATED PLUG
--0-- C-D AND TRUNCATED
PLUG
90

80

70

co
"C
....,J-
LLI
>
::I
LLI 60
c:::: (al Static.
=>
V')
V')
LLI
c::::
c...
100
C\
Z
=>
0
V')
,Q
\

I b
90 ,, \
\

80

70

60~-- ______ ~ __________ ________ ~ ~

10 100 1000 10000

FREQUENCY, Hz
(bl Simulated flight, V~ 122 m/sec
Figure 24. - Effect of C-D termination on forward quadrant
spectra (9" 60°) for coannular nozzle at C-D desig~ point~
Tj 1 20 480 K. Tj 2 20 940 K. Data scaled to O. 903-m total
nozzle area am]' extrapolated to 730-m sideline (ref. 10),
 110 0 C-D COANNULAR NOZZLE
WITH EXTENDED PLUG
CONVERGENT
105 COANNULAR
NOZZLE WITH
TRUNCATED
100 FULLY-MIXED 0
CONVERGENT ""_"(
CIRCULAR ).~ ....
NOZZLE-- /.-......_~,. .;:,
r::a
95 0 if1ft...-' C-D COANNU-
•/ ~/ 000 LAR NOZZLE
"'C

J
c... / ,.~~ WITH TRUNCA-
Vl
«
0 90 ----- TEDPWG
J
I.J.J C-D DESIGN POINT
>
::I
I.J.J 85
c::::
::J (a) Static.
Vl
Vl
I.J.J
c::::
c...
110
c
z:
::J
0
Vl
...J
...J
105
«
c::::
I.J.J
>
0
100

95

90
C-D DESIGN POINT
85 ~ ____L __ _ _ _L __ _ _ _a __ _ _ _L __ _ ~

-1. 5 -1. 0 -.5 0 .5 1. 0

EFFECTIVE SHOCK STRENGTH PARAMETER, 10 log ~ef
(b) Simulated flight, Vo = 122 m/sec.
Figure 25. - Effectiveness of extended plug termina-
tion in reducing coannular nozzle OASPL in the
forward quadrant, 9" 600 ; inner-stream temper-
ature, Ti 1 ~ 480 K; outer-stream temperature,
Tj, 2 ~ 94D K. Data scaled to O. 903-m 2 total nozzle
area and extrapolated to a 730-m sideline (ref. 11).
 o CONVERGENT, TRUNCATED PLUG
110
o C-D, TRUNCATED PLUG
o C-D, EXTENDED PLUG 0 @
f:l
€I ~ 0
8
90

80 ~ __ ~ __-....J_ _ _ _ _ _
~ ~ _ _ _ _~_ _~
UJ
0:::
=> (a) static.
Vl
Vl

~
a..
110
c
Z
=>
~ 100
....J
....J
«
8J 90
6
80 '--__ __-....J_ _ _ _ _ _
~ ~ ___L_ _ __ ' _ __ __ _ _ '

40 60 80 100 120 140 160

ANGLE TO INLET, 8, deg
(b) Simulated flight, V0 -::; 122 ml sec.
Figure 26. - Effect of plug tip geometry C-D coan-
nular nozzle directivity at C-D design conditions;
\ 1 "" 480 K, Tj, 2 ~ 940 K. Data scaled to
O. 903-m 2 total nozzle area and extrapolated to
730-m sideline (ref. 10).
 100
COANNULAR
--0-- CONVERGENT AND TRUNCATED PLUG
-0- C-D AND TRUNCATED PLUG
---0--- C-D AND EXTENDED PLUG
90

80

70

co
"0

-I
L.U q
~ 60
L.U
0:::
~
(a) Static.
Vl
Vl
L.U
0::: 100
Q..

C
Z
~
0
Vl

90

80

70

60~--------~----------~----~0~~
10 100 1000 10000
FREQUENCY, Hz
(b) Simulated flight, Vo;' 122 m/sec
Figure 27. - Effect of plug tip geometry on forward quadrant
spectra (9 c 600 ) for C-D coannular nozzle at design point;
Tj 1 '" 480 K, T· 2 '" 940 K. Data scaled to 0.903 m2 total
nozzle area and'extrapolated to 73Q-m sideline (ref. 10).
 no CONVERGENT COANNULAR
NOZZLE WITH
EXTENDED PLUG
~ SUPERSONIC INNER STREAM
105 b. SUBSONIC INNER STREAM

FU LLY-MIXED
100 CONVERGENT
CIRCULAR
,-.....
NOZZLE ....... , / ' ~ ~"
[{ ~
cc
95 ~ /'
"C '/ '/ b;. b;. b;.
J ,/'/b;.
a..
V)
<t:
'"~--~ .-( '-CONVERGENT COANNULAR
0 90
b;. NOZZLE WITH TRUNCATED
J
l.J.J PLUG
>
l.J.J
---I
l.J.J
~
85
::::> (a) Static.
V)
V)
l.J.J
~
a.. no
0
z
::::>
0
V)
---I 105
---I
<t:
~
l.J.J
>
0
100 /, ......- .."
I
I " ~ ~

I
95
tlI
_- -- -" b;.
90 ~-
b;.
85~----~----~----~----~----~
-1.5 -1.0 -.5 o .5 1.0
EFFECTIVE SHOCK STRENGTH PARAMETER.
10 log ~ef
(b) Simulated flight. Vo ~ 122 m/sec.

Figure 28. - Effectiveness of subsonic inner stream

conditions in reducing OASPL for convergent
coannular nozzle in the forward quadrant. 9 = 600 :
outer-stream tem~erature. T· 2 ~ 940 K. Data
scaled to 0.903-m total nozz~e area and extra-
polated to a 730-m sideline (ref. 11).
 120 \" WITH SUPERSONIC INNER
\ Pj" 2/P a" 3.14, Tj" 1 = 480 K
BENEFIT DUE TO
\
\
' ,
llO SUBSONIC INNER I \
I \
co I \
"0

~"::-:;;~~~~~ i~~~~j~ ~R710

-l- 100
c....
Vl
<C
0
90 K
-l
l.LJ
>
~
l.LJ 80
c::: (al Static.
~
VI
Vl
l.LJ
c:::
c....
120
0
r WITH SUPERSONIC INNER
z BENEFIT DUE TO / p" 1 = 3.14, T"1 = 483 K
~
0 110 SUBSONIC INNER I I], ],
Vl I I

~
-l
-l
<C
c:::
l.LJ

6
100 !
;p:f\~·~·if;i;;/.(;V;;.~~fiS.:': ".
90 B~~~7;:"dB ',-- WITH SUBSONIC INNER
Pj, I/Pa = 1. 81, \ 1 = 700 K
80
0 ~ M 1~ 1~
ANGLE TO INLET, 8, deg
(bl Simu lated flight, V0 = 122 ml sec.
Figure 29. - Typical static and simulated flight OASPL-
directivities of convergent coannular nozzle with extend-
ed plug for subsonic and supersonic inner streams for a
given underexpandedouter stream; Pj,2/Pa; 3.32,
Tj ,2; 9~ K.
 100

,R\
?: ~\
co
"C 90
_i \ / / "WITH
0-
Vl , b(y/ SUPERSONIC
_i I , INNER
~
I..LJ
2i / Pj, 11 Pa = 3. 14
....J ~ T- 1 =483K
I..LJ
0::: 80
0. J,
\
=>
Vl /
Vl /
I..LJ /
0::: I
0- /
0 L WITH SUBSONIC
z
=>
0
Vl 70
INNER
Pj Pa = 1. 81
T-'1 ~ 700 K
J,
I'
60~--------~-----------L--------~
10 100 1000 10000
FREQUENCY

Figure 30. - Typical simulated flight front quadrant spectra

at 8 = 600 for convergent coannular nozzle with extended
plug for subsonic and supersonic inner streams for a
given underexpandedouter stream; Pj,2/Pa~ 3.32;
\2~ 900 K.

----E--- ----E---
SECTION A-A SECTION B-B
A A

(a) Convergent flow element terminations.

~ -Er
SECTION A-A SECTION B-B

B B

-------
(b) C-D flow element terminations.
Figure 31. - 20 chute annular plug suppressor configurations (ref. 8),
 110
MULTI-ELEMENT
SUPPRESSORS
105 o CONVERGENT
o C-D

100

co 95
"0
J
c..
VI
<t: 90 0 0
OO~~DESIGN POINT
0
J
LLJ
>
~
LLJ
c:::
85
:::J
VI (a) Static.
VI
LLJ
c::: 110
c..
Q
:z:
:::J
0
VI
.....I 105
.....I
<t:
c:::
LLJ
>
0
100

95
00
90
o ~A~O 0
o OL.J'"""
C-D DESIGN POINT

85~----~----~--~~----~--~
-1. 5 -1. 0 -.5 0 .5 1.0
SHOCK STRENGTH PARAMETER, 10 log ~

(b) Simulated flight, Vo= 122 m/sec.

Figure 32. - Comparison of C-D multi-element sup-
pressor nozzle OAS PL with those of convergent
multi-element suppressor and baseline convergent
circular nozzle in forward quadrant, 9 = 500~ jet
temperature, T· "" 950 K. Data scaled to 0.903-m 2
nozzle area and extrapolated to 730-m sideline
(ref. 9).
 0 BASELINE CONVERGENT CIRCULAR
CONIC NOZZLE
110 0 CONVERGENT MULTI-ELEMENT
SUPPRESSOR PLUG NOZZLE

.
0 C-D MULTI-ELEMENT SUPPRESSOR
PLUG NOZZLE
100

co
0

• •••
"0

...J
90

..
c...
VI Cb
...J 0
0 ••
•
L.U

..•. .
0 00
L.U
c::: SO
,.~
V'l
LLJ
c:::
c...
cP
~<IJ;
~O~
00
~~..a ,
c
z
:::>
0
70
OQ OCC~ ~ •
t
VI

~
60
cg
50 L-.....l...---L---L----L.---1...--I
50 250 lK 4K 50 250 lK 4K
lJ3-0CTAVE-BAND CENTER FREQUENCY, f, Hz
(a) Static. (b) Simu lated flight,
Vo:::! 122 m/ sec
Figure 33. - Spectral comparison between baseline
convergent circular nozzle, convergent multi-
element suppressor plug nozzle, and C-D multi-
element suppressor plug nozzle, a = 500 ; Tj'=
960 K. Data scaled to 0.903 m2 nozzle area and
extrapolated to 730-m sideli ne (ref. SI.
 0 BASELINE CONVERGENT
CIRCULAR NOZZLE
0 CONVERGENT MULTI-
ELEMENT SUPPRESSOR
120 PLUG NOZZLE
~ C-D MULTI-ELEMENT SUPPRESSOR
PLUG NOZZLE 0
llO 0
0
, JY~'~
co
-0
....J- o
0..
VI o 0
o 000 ,.c! ''0
<:
0
....J-
LLI
>
::l
LLI
c::::
:::J
VI
VI
LLI (al Static.
c::::
0..
0
z llO
:::J
0
VI
....J
....J
<:
••••
c::::
LLI

6
90

80 ~ __ ~ __-....J_ _ _ _ ~ _ _ ~_ _ _ _~ _ _~

40

(bl Simulated flight, Vo = 122 m/sec.

Figure 34. - OASPL directivity comparisons between
baseline convergent circular nozzle, convergent
multi-element suppressor plug nozzle, and C-D
multi-element suppressor plug nozzle, Ti ~ 960 K.
Data scaled to O. 903-m 2 nozzle area and extra-
polated to 730-m sideli ne (ref. 8),
 ° BASELINE CONVERGENT
CIRCULAR NOZZLE
110
0 CONVERGENT 20-CHUTE
SUPPRESSOR NOZZLE

.
°
0 C C-D SUPPRESSOR
8 Gb
co
"0

-l
a..
VI
100
O~
C ~
°0
CO
CO
.• ..-•
.~

-l
L.LJ
G:i
-l
L.LJ
c:::
90 C

~d)J~Oo
°
.....
~
::J
VI
VI
80
~O
L.LJ
c:::
a..
0
z: ~
::J
0
& i
VI
70
•
g 1
60
50 250 lK 4K 16K 50 250 1K 4K
1I3-0CTAVE-BAND CENTER FREQUENCY, f, Hz
(a) Static. (b) Simulated flight,
Vo = 122 m/sec.
Figure 35. - Spectral comparison between convergent circu-
lar nozzle, convergent multi-element suppressor plug
nozzle, and C-D multi-element suppressor ~Iug nozzle,
8 = 1400 , Ti :: 960 K. Data scaled to 0.903-m nozzle
area and extrapolated to 730-m sideline.
 -E-
SECTION A-A
-E-
SECTION B-B

A A
OUTER FLOW PATH
~ .-------..

(a) Convergent flow element terminations.

-E:f- ~
SECTION A-A SECTION B-B

B B

(b) C-D flow element termination.

Figure 36. - 20-chute outer stream suppressor configurations (ref. 10),
 MULTI-ELEMENT
OUTER-STREAM
SUPPRESSOR
D CONVERGENT
o C-D FULLY-MIXED
CONVERGENT
105
CIRCULAR
NOZZLE "
CONVERGENT
100 COANNULAR
NOZZLE WITH
TRUNCATED ;';'
"-,,
;'
PLUG,
"0"
95
;'
\ ;'
\ ;' 0 D
a)
\ ;' 0
--~-,.Oq:g D
"0

...1 90
C-
Vl
O D
<x:: D C-D DESIGN POINT
0
...1
w
D
> 85
~
w (a) static.
~
::::l
Vl
Vl
no
w
~
c-
o
z
::::l 105
0
Vl
-I
-I
<x::
~
w 100 .;
,.-- ' .....
> ;'
0
, I
/

95 ,
I
I

0 D
,.,."
... ... - _~o
0
90
OD
C-D DESIGN POINT
85 ____ ____ __ ____- L_______
~
~
~
~
-
L
-1. 5 -1. 0 -.5 o .5 1.0
EFFECTIVE SHOCK STRENGTH, 10 log ~ef

(b) Simulated flight. Vo ~ 122 m/sec.

Figure 37. - Effectiveness of C-D nozzle terminations
on OASP:'" in forward quadrant, 9 = 600 , for dual-
stream nozzle with multi-element outer-stream
suppressor; inner-stream temperature, Tj 2 ~ 940 K.
Data scaled to O. 903-m 2 total nozzle area a~d extra-
polated to 730-m sideline (ref. lll.
 COANNULAR
90
0 C-D SUPPRESSOR NOZZLE
l:l. CONVERGENT SUPPRESSOR NOZZLE

A
6~
80
" \~
\
h
\
\
h,
co
"0
....J.
70
,,
\

a.. \
Vl
....J.
\
LLI 1::.
Gi
....J
60
~
::l
VI
VI
LLI
c::
a..
90
0
z
::l
0
VI

80

70

100 1000 10000

FREQUENCY, f, Hz
(bl Simulated flight, V:::: 122 m/sec.
Figure 38. - Typical static and simulated flight front quadrant
spectral comparison at e = 600 between coannular C-D and
convergent suppressor nozzles at C-D design conditions;
Pj,l/Pa~ 2.91; Pj,2/Pa~ 3.14; Tj,1~480K; Tj,2~950K.
Data scaled up to total nozzle area of 0.903 m2 and extra-
polated to 730 m sideline.
 t::. CONVERGENT
llO 0 CONVERGENT -DIVERGENT
010 0
100 666
...J
c...
V'l
~ ~ ~ ~ T6.0 dB
0
6
C§
90
~la ~
...J
l..LJ
~ T1.1 dB
G:i
...J
l..LJ
0:::
80
:::::>
V'l
(al Static.
V'l
l..LJ
0:::
c... llO
o
z:
:::>
oV'l 100 10 0
4~ ~ ¢ 6 0
...J
...J

13.9 6
<t:
0:::
{).
¢ ¢. ¢ 0
l..LJ
> 90 6
o 0.5 dB dB 6

80
40 60 80 100 120 140 160
ANGLE TO INLET, 8, deg
(bl Simulated flight, V0 ~ 122 m/sec.
Figure 39. - Comparison of OASPL-directivities of
coannular suppressor C-D nozzle with those of
coannular suppressor convergent nozzle at C-D
design conditions; inner-stream temperature,
TLl ~ 450 K; outer-stream tefperature, Tj 2 ~
9ju K. Data scaled to 0.903-m total nozzle area
and extrapolated to 730-m sideline (ref. 10).
1: Report No. 2. Govemment Accession No. 3. Recipient's Catalog No.

NASA TM-83799
4. Title and Subtitle 5. Report Date

8. Performing Organization Code

Supersonic Jet Shock Noise Reduction
505-31-38
7. Author(s) 8. Performing Organization Report No.

E-2299
James R. Stone 10. Work Unit No.

9. Performing Organization Name and Address

11. Contract or Grant No.
National Aeronautics and Space Administration
Lewis Research Center
Cleveland, Ohio 44135 13. Type of Report and Period Covered

12. Sponsoring Agency Name and Address Technical Memorandum

National Aeronautics and Space Administration 14. Sponsoring Agency Code
Washington, D.C. 20546

15. Supplementary Notes

Prepared for the Ninth Aeroacoustics Conference sponsored by the American Institute
of Aeronautics and Astronautics, Williamsburg. Virginia. October 15-17, 1984.

16. Abstract
Shock-cell noise has been identified as a potentially significant problem for ad-
vanced supersonic aircraft at takeoff. Therefore NASA has conducted fundamental
studies of the phenomena involved and model-scale experiments aimed at developing
means of noise reduction. This paper reviews the results of a series of studies
conducted to determine means by which supersonic jet shock noise can be reduced to
acceptable levels for advanced supersonic cruise aircraft. Theoretical studies
were conducted on the shock associated noise of supersonic jets from convergent-
divergent (C-D) nozzles. Laboratory studies were conducted on the influence of
narrowband shock screech on broadband noise and on means of screech reduction.
The usefulness of C-D nozzle passages was investigated at model scale for single-
stream and dual-stream nozzles. The effect of off-design pressure ratio was deter-
mined under static and simulated flight conditions for jet temperatures up to
960 K. Annular and coannular flow passages with center plugs and multielement
suppressor nozzles were evaluated, and the effect of plug tip geometry was estab-
lished. In addition to the far-field acoustic data, mean and turbulent velocity
distributions were obtained with a laser velocimeter, and shadowgraph images of
the flow field were obtained.

17. Key Words (Suggested by Author(s)) 18. Distribution Statement

Noise reduction; Shock noise; Supersonic Unclassified - unlimited

jet noise; Convergent-divergent nozzles STAR Category 71

19. Security ClassU. (of this report) 20. Security Classlf. (of this page) 21. No. of pages 22. Price"

Unclassified Unclassified
"For sale by the National Technical Information Service, Springfield, Virginia 221&1
National Aeronautics and SPECIAL FOURTH ClASSMAll
Space Administration BOOK

Washington, D.C.
20546
Official Business
Pl!nallV for Priv.te Use. S300 Poat. and Fees Paid
National Aeronautics and
$pace Administration
NASA~51

NI\S/\ POSTMASTER: If Undrliverahlr (StClion 1511

Postal Manual) no Nut Return