NASA TECHNICAL NOTE NASA TN D-7075

CASE FILECOPY

NOISE GENERATED BY IMPINGEMENT

OF A JET UPON A LARGE FLAT BOARD

by William A. Olsen, Jeffrey H. Miles,

and Robert G. Dorsch

Lewis Research Center

Cleveland, Ohio 44135

NATIONAL AERONAUTICS AND SPACE ADMINISTRATION • WASHINGTON, D. C. • DECEMBER 1972

https://ntrs.nasa.gov/search.jsp?R=19730004299 2020-03-19T11:34:31+00:00Z

1. Report No.

NASA TN D-7075

2. Government Accession No.

4. Title and Subtitle

NOISE GENERATED BY IMPINGEMENT OF A JET UPON A

LARGE FLAT BOARD

7. Author(s)

William A. Olsen, Jeffrey H. Miles, and Robert G. Dorsch

9. Performing Organization Name and Address

Lewis Research Center

National Aeronautics and Space

Cleveland, Ohio 44135

Administration

12. Sponsoring Agency Name and Address

National Aeronautics and Space Administration

Washington, D.C. 20546

3. Recipient's Catalog No.

5. Report DateDecember 1972

6. Performing Organization Code

8. Performing Organization Report No.

E-698310. Work Unit No.

741-72

11. Contract or Grant No.

13. Type of Report and Period Covered

Technical Note

14. Sponsoring Agency Code

15. Supplementary Notes

16. Abstract

Data were obtained on the noise generated by an air jet impinging on a large flat board. Theboard was large enough so that the flow leaving the edges of the board generated no significantnoise. The impingement angle, nozzle shape and size, jet velocity, and the distance from thenozzle to the board were varied in the experiment. Far-field noise data are presented. Thenozzle-alone noise contribution to the total noise was generally small and was subtracted fromthe total, leaving the impingement-only noise. The impingement-only noise was adequatelycorrelated by eighth power of the peak impingement velocity and first power of the impingementarea. The spectral data were correlated by a Strouhal number based on the peak impingementvelocity and a characteristic impingement diameter.

17. Key Words (Suggested by Author(s))

Aerodynamic noise

Jet impingement

Large surfaces

19. Security Classif. (of this report)

Unclassified

18. Distribution Statement

Unclassified - unlimited

20. Security Classif. (of this page) 21 . No. o

Unclassified 3f Pages 22. Price*

5 $3.00

' For sale by the National Technical Information Service, Springfield, Virginia 22151

CONTENTS

Page

SUMMARY 1

INTRODUCTION 1

APPARATUS AND PROCEDURE 2Flow System and Board 2Acoustic Instrumentation 3Test Procedure 4

RESULTS AND DISCUSSION 7Noise Radiation Patterns 8Effect of Nozzle Exhaust Velocity on Power Spectra 11Effect of Nozzle Diameter 16Effect of Impingement Angle 19Effect of Nozzle Shape 20Effect of Impingement Distance 22Overall Correlation of the Impingement-Only Noise 24

Total impingement-only sound power level 24Impingement-only noise spectra 26

SUMMARY OF RESULTS 30

APPENDIX - SYMBOLS 31

REFERENCES 33

iii

NOISE GENERATED BY IMPINGEMENT OF A JET UPON A LARGE FLAT BOARD

by William A. Olsen, Jeffrey H. Miles, and Robert G. Dorsch

Lewis Research Center

SUMMARY

Data were obtained on the noise generated by an air jet impinging on a large flatboard. The board was large enough so that the flow leaving the edges of the board gen-erated no significant noise. The impingement angle, nozzle shape and size, jet velocity,and the distance from the nozzle to the board were varied in the experiment. Far fieldnoise data are presented. The nozzle-alone noise contribution to the total noise wasgenerally small and was subtracted from the total, leaving the impingement-only noise.

The noise generated by an air jet impinging on a very large flat plate has the follow-ing major characteristics.

(1) The shape of the total noise (impingement and nozzle) radiation pattern is essen-tially independent of nozzle velocity and reaches its maximum near 160° from the up-stream end of the board.

(2) The impingement-only noise (total sound power level) was correlated by theeighth power of the peak impingement velocity and first power of the impingement area.The spectral data were correlated by a Strouhal number based on the peak impingementvelocity and a characteristic impingement diameter. When the jet impingement is notnormal, the impingement angle is also brought into the correlation. The total soundpower level is proportional to the square of the sine of the impingement angle. TheStrouhal number is multiplied by the reciprocal of the square root of the sine of the im-pingement angle.

INTRODUCTION

In one of the short takeoff and landing aircraft (STOL) concepts, the engine exhaustis deflected downward by a wing with trailing-edge flaps. References 1 to 3 showed thatthe impingement of the exhaust jet on the flap surfaces results in considerable additionalnoise. Jet impingement noise also results when an exhaust jet is turned by a thrustreverser or spoiler, m vertical takeoff and landing (VTOL) applications additional noiseis generated when the engine exhaust jet is directed at the ground.

As part of the aeronautical research program at the Lewis Research Center, jetimpingement noise is being studied. This report contains the results of an investigationwherein a jet was directed at a very simple surface, a large flat board.

Far-field noise data were taken for a jet impinging on the board and for the nozzlealone. The variables of the study were the nozzle exhaust velocity, nozzle diameter andshape, distance from the nozzle to the board and the angle of the board from the nozzlecenterline. For most of the data the board was normal to the ground, and the micro-phones were in a plane parallel to the ground that passed through the nozzle centerline.Limited data were taken to examine the impingement noise for axial symmetry.

APPARATUS AND PROCEDURE

Flow System and Board

Proceeding downstream the flow system (fig. 1) consisted of an orifice for flowmeasurement, a 10-centimeter globe type of flow control valve, a valve noise quieting

Lperfo ratedv \plates t-7-m long inlet

of 10-cm pipe

Flow orifice

CD-11338-02

Figure 1. - Flow system and board.

section, a long straight run of 10-centimeter pipe and finally the nozzle. The nominalnozzle stagnation temperature was near ambient. The nozzle jet was directed at nearthe center of a large flat smooth board (2. 4 by 2. 4 m). The board rested on the groundand was moved to change the jet impingement distance and angle. It was made of 2. 5-centimeter-thick smooth plywood with 10-centimeter-thick foam rubber glued to the backside of the board.

Valve noise was not a problem in this experiment for the following reasons. Theflow system and quieting section are the same as those used in reference 2. In thatstudy it was shown that for nozzle exhaust velocities v\T greater than 190 meters per

8second, the noise generated by the nozzle alone followed a V,r relation. (Symbols aredefined in the appendix.) The spectral distribution followed that reported in reference 4.In addition, the noise levels of the board noise are generally much greater than thenozzle-alone noise levels.

Acoustic Instrumentation

The noise data were measured by 1. 3-centimeter (half inch) condensor microphoneswith windscreens. The microphones were located in a horizontal plane that passedthrough the nozzle centerline. The microphones were placed on a 3-meter-radiuscircle (see fig. 2) that enclosed the board and nozzle. Eight microphones (on one sideof the centerline) were used when the board was normal to the nozzle centerline (a; = 90°)or when the nozzle-alone noise data were taken. Fourteen microphones were placedaround the entire microphone circle when the board was not normal to the centerline(a < 90°). The angle a is the minimum angle from the board to the nozzle centerline.In both cases the microphones were more concentrated on the microphone circlewherever the noise level was highest. The microphones were kept out of any high ve-locity exhaust jet.

Most of the noise data were taken with the board normal to the ground and micro-phone circle (<?•& = 0°) because this is the orientation of maximum noise. The noise isaxisymmetric for the board at a = 90° and for the nozzle alone. The noise in the mini-mum noise plane (<pB = 90°) for the nonaxisymmetric case of a = 60° was also meas-ured. This was accomplished by rotating the board counterclockwise 90 about the noz-zle centerline, while holding a = 60° (fig. 2). This means that the board is at an angleof 60° to the ground. This <PJ, = 90° case is geometrically equivalent to leaving theboard normal to the ground at a = 60°, as in the <£>B = 0° case, and then rotating themicrophone plane 90° (clockwise) about the nozzle centerline.

Area element3m lifor power

integration

9 = 180°-

CD-11339-02

Angle from nozzleinlet, 9 Microphone of

mike circle inhorizontalplane

Figure 2. - Microphone setup and area element used in the noise power integrations.Board is shown perpendicular to the microphone plane,<PQ • 0°.

Test Procedure

Far field noise and flow data were taken for as many as five nozzle exhaust veloc-ities ranging from about VN = 140 to 350 meters per second for the board and nozzlealone. In addition, the distance from the nozzle to the board Lrp and impingementangle a. were varied. These geometric parameters were set up with respect to thenozzle lip by using a set of templates. The circular nozzle diameter djr and nozzleshape were also varied. The variables for each test run are listed in table I.

Three noise data samples were taken at each microphone location for each run con-dition. These data were analyzed directly by a one-third octave band spectrum analyzer.The analyzer calculated sound-pressure-level spectra in decibels referenced to 2x10newtons per square meter (0. 0002 jubar). The three samples were then arithmeticallyaveraged, and a correction was applied for any microphone, cable, and atmosphericlosses. No spectral corrections were deemed necessary for ground reflections because,

TABLE I. - EXPERIMENTAL CONVERGENT NOZZLE AND ORIFICE PLATE PARAMETERS

Run day-

run num-ber

T-80

7-836-50

13-04

13-05

13-0613-07

6-52

14-4514-4614-5414-49

9-069-075-61

12-8812-8912-9012-9112-92

12-93

16-9316-94

d!6-9516-9616-97

Nozzle description

Type

Conve rgentb

Orifice platesof samehole area

Conve rgent

Nozzlediam-eter,

cm

5

4

4

22

7

2

11

77

8

—

5 2

Orifice plate description

Orificeshape

Four holesFour holes16 holesVerticalslot1.04X12.7

_cm

Holediam-eter,

cm

4. 1

1.03

Holespac-

ing,cm

6.41.91.0

Nozzleexhaustvelocity,

m/sec

292195130

344236

289

193

191238

196

290

296295287

138286338

340

289193236289141

240192295350

138

Impingementlength to noz-zle diameter

rat o,

Nozzleno boar

Nozzleno boa

7.7.7.

9.

7.

alone,d

alone,-d

050505

7

05

Impingementangle,

o,deg

Nozzno bo

Nozzlno bo

666

9

e alone,ird

e alone,ird

000

0

Azimuthal

angle,

Nozalonboar

Nozalon

boar

lee, nod

e, nod

0

Conditions,

holding a and bconstant

Typnoz

stagti

ternatui

T

?(

2

2

3

222

2

3

ical

zlena-

anper-e,a

N'<

4

)0

i4

)0

88

!4

30

30

Peak Im-

pingementvelocity,

m/sec

—

—

168256322

290

206137181

206103

221

171264334125

Character-istic im-

pingementdiameter,

•Vm

:::::;

0.0374.0374.0426

.0426

.044

.0374

.0374

.0374

.0426

.0412

TABLE I. - Concluded. EXPERIMENTAL CONVERGENT NOZZLE AND ORIFICE PLATE PARAMETERS

Run day-

run num-ber

9-79-89-99-10

12-94e 12-95

12-9612-9712-89, -92

d!2-8714-5114-5216-95

12-8912-9813-02

9-611-369-12

16-90

9-19-139-11

14-4214-32

14-4114-53

14-50

(e). (i)

16-8816-9116-89IK -09

Nozzle description

Type

Conve rgent

Orifice plates

of samelole area

Convergent

1

Nozzlediam-eter,dN.cm

5.

2.4.

5.2.4.7.

5.4.2.

2

7

1

27

18

217

—

5.2

1

Orifice plate description

Orificeshape

Four holes

Four holesFour holes16 holesVerticalslot

1.04x12.7cmOne hole

Holediam-eter,cm

4.14.1

1.03

4.13

Holespac-

ing,cm

6.46.4

1.91.0

Noexhvelc

Vm/

2

2

22

222

222

111

1

222

szleaust

"city,

N'sec

36

33

J939

373735

393939

38

363830

3633

31

290290

296295

287

295

293293192199

Impingement

length to noz-zle diameter

ratio,LT/dN

7.

1.4.

14.259.7.

2.1.7.

9.9.9.

7.

05

887

5

7

05

84305

77

7

05

7.3524.5

7.3514.7

14.5

3.7

7 05

Impingement

angle,

o>deg

631

9

6

<

f

005

0

0

0

0

Azimuthalangle,

<"B-deg

0

0

090

0

90

Condi

<A>a,b =holding

cons

(»'VN,U

<LT>o,V

^Uv,

Nozzle sconstant

LT, andare?

ions,Vary A

i and btant

PdN

M' d N

,, VN

tape atVN-

nozzle

<"B>VN,LT/dN

1

Typicalnozzlestagna-

t ontemper-

ature, a

VK

21

3t3C3C

2'2"2£

2i2"2"

3(

2f2'

2

.3(

8

000

880

9580

988

0

300

1

Peak im-

pingementvelocity,

m/sec

256

29029015588

206267

294294264

206206206

167166167

170

256253251

'230g127

f246BtfiO

'173

"^SO

262262171171

Character-istic im-

pingementdiameter,

<Vm

0.0374

.052

.042

.044

.07

.044

.0374

.046

.052

.0374

.044

.022

.034

.0374

.019

.0296

.0556

.0374

.0296

.019

0.016.056.016.052

.042

.0296

0. 0374

Ambient air temperature is typically 2 K lower than stagnation temperature.bThe nozzle has a 0. 32-cm lip.cValve noise present,thistle.eScreech.

Four circular jets.**One well mixed circular jet.One circular jet.

1One rectangular jet.i No noise data taken.

generally, the major cancellations and reinforcements occur at much lower frequenciesthan the region of interest. The overall sound pressure level OASPL at each micro-phone position 9 was computed from the averaged and corrected sound-pressure-levelspectrum SPL. The sound-power-level spectrum PWL and total sound power levelPWLt were computed by spatial integration of the SPL data from each microphone.The values of SPL and OASPL at each microphone position already include the groundreflection contribution. Therefore, the spatial integrations for sound power is madeover a hemisphere. Because most of the noise data are axisymmetric (e. g., nozzle-alone and board at a = 90°) the integration was performed using the "half-bread-slice"area elements shown in figure 2. When the noise was axisymmetric the correct, PWLwould be computed. But when a < 90° the noise is not axisymmetric and only an ap-proximate measure of the sound power level is computed PWL'. This is evaluatedfrom data taken in the <p-Q plane by assuming that there is no azimuthal variation (with<pB) of the noise.

The condenser microphones were calibrated before each day of running with a stand-ard piston calibrator (a 124-dB, 250-Hz tone). The one-third octave band analyzer wascalibrated before each test with a constant voltage source and checked during the exper-iment with an electronic pink noise generator. The data acquisition and data analysissystems were further checked by the experimenter by repeated runs and by a uniformdirectivity orifice noise source of a 1. 5-decibel repeatability. Considering these cali-brations and checks, repeated data, and the averaging of three widely spaced (in time)samples of data, it is estimated that the reported data are repeatable to within 1. 5 deci-bels from day to day. Most of the directly compared data were obtained on the same dayso that those data are repeatable to about 1/2 decibel.

RESULTS AND DISCUSSION

Placing a large flat board in the high velocity jet of a nozzle affects the noise meas-ured in essentially four ways:

(1) The impingement of the jet on the board surface generates additional noise to thenozzle-alone noise. But no additional noise is generated by the air leaving the edge ofthe board because its velocity is too low (measurements showed that it was an order ofmagnitude lower than the impingement velocity).

(2) The board redirects (by reflection and shielding) the noise generated by the noz-zle alone and any internal noise (e. g., valve noise).

(3) Less low-frequency nozzle-alone jet noise is generated as the board is broughtcloser to the nozzle.

(4) A feedback noise (narrowband whistle or screech) can occur for certain flowand configuration conditions (ref. 5).

The primary purpose of this report is to determine the effect of the experimentalparameters (nozzle exhaust velocity, impingement angle and distance, and nozzle shape)on the additional aerodynamic noise generated by impingement of the jet on a large flatboard. The best way to accomplish this is to work with the sound power level spectrumbecause, ideally, it is not affected by the redirection of noise. These spectra are thencorrected to exclude the noise from the nozzle alone (i. e., PWL generated by the noz-zle without the board), and also to exclude any additional narrowband noise caused byfeedback whistle or screech. Unfortunately, a power spectrum requires sound meas-urements in all azimuthal planes if a < 90°. Most of the board noise data in this re-port were taken for cases where there was normal jet impingement, a = 90°. For thesecases and for the nozzle-alone cases the noise is axisymmetric and measurements needto be taken in only one azimuthal plane (<p-r> = 0°).

Complete tables of the one-third octave SPL and PWL spectra are available fromthe authors on request.

Noise Radiation Patterns

In this section some representative noise radiation patterns are shown to acquaintthe reader with the data before the power spectra are discussed. These plots show theeffect of nozzle exhaust velocity and impingement angle on the noise patterns that resultwhen the jet from a nozzle is directed at a very large flat board. Most of the noisepatterns are in the plane of maximum noise (<p-n = 0°). Some data were also taken in theminimum noise plane (cp-r, = 90°).

Figure 3 shows the variation of OASPL with 9 for the board normal to the ground(<£>„ = 0°) for various velocities at LT = 37 centimeters and d,^ = 5. 2 centimeters.Figure 3(a) shows the data for a = 90° (normal impingement) and figure 3(b) shows thedata for a = 60°. In both cases the noise radiation patterns are similar in shape overthe range of velocity shown. The noise level for normal impingement (fig. 3(a)) is es-sentially independent of 9 on the nozzle side of the board and falls off rapidly behind it.

Figure 4 contains a plot of the noise radiation patterns that result, in the <p-Q = 0plane at LT = 37 centimeters and VN = 286 meters per second, when the jet impinge-ment angle a is varied. These patterns are plotted with respect to the board (B + a)rather than with respect to the nozzle inlet 9. This figure shows that as 01 increasesthe noise level increases, especially for (9 + oi) less than 140°. But the maximumOASPL, which occurs at about (9 + a) = 160°, does not change much. The OASPLplotted here is the sum of the impingement and nozzle-alone noise. The nozzle-alone

130

120

no

100

Iir\

a

90

80

Nozzle exhaust velocity,

VN,m/sec

O 350a 338D 295O 286A 240D 192a 188O 138

Board and nozzle noiseNozzle-alone noize

Open symbols aenote nozzle-alone noisesSolid symbols denote board and nozzle

(total) noise |

(a) impingement angle, a, 90°.

270°

60 90 120 150 180 210 240Angle from nozzle inlet, 8, deg

(b) Impingement angle, a, 60°.

270 300 330 360

Figure 3. - Noise radiation patterns for jet impinging on board at several nozzle exhaust velocities. Azi-muthal angle, 9$, 0°; nozzle diameter, dR, 5.2 centimeters; impingement distance, LT, 37 centimeters.

30 60 90 120 150 180 210 240Angle from board, (9 + a), deg

270 300 330 360

Figure 4. - Effect of impingement angle, a, on noise radiation pattern of board. Azimuthal angle, (Og,0°; nozzle exhuast velocity, VN, 286 meters per second; impingement distance, LT, 37 centimeters;nozzle diameter, d^, 5.2centimeters.

noise pattern is also plotted in this figure, with a taken as 0 . Figure 4 shows thatthe nozzle-alone noise pattern is approached as the impingement angle a approaches0°. The nozzle-alone noise has its maximum near 0 = 160°, and this noise is of highfrequency, which reflects off the large board somewhat like light rays. Therefore, itis understandable that the total noise of the board and nozzle would tend to peak at(9 + a) = 160° also.

Limited data were taken to show the azimuthal variation (with cp-g) of the noiseradiation pattern. Noise data were taken in the maximum noise orientation (<pB = 0°)where the board was at a = 60°. Then the board at a = 60° was rotated 90° counter-clockwise about the nozzle center line to </?„ = 90°, and the data were taken for thisminimum noise orientation. The difference in OASPL between the maximum noiseorientation and minimum noise orientation is plotted on figure 5 as a function of B fortwo nozzle exhaust velocities. The curves drawn through both sets of data show that,except near 9 = 90°, the azimuthal variation is insensitive to the nozzle exhaust veloc-ity. An average curve drawn through the data has skew symmetry about the indicatedpoint of symmetry. The location of this point imply, that on the average the maximumnoise orientation (<p = 0°) is noisier than the minimum noise orientation (<p = 90°).

10

<:o

co o

* a

24

16

-16

--D--r—|

Nozzle exhaustvelocity,

VN.m/sec293192

-Symmetrical curvethrough data

Point of symmetry

30 60 90 120 150 180 210 230Angle from nozzle inlet , 6, deg

270 300 330 360

Figure 5. - Change in overall sound pressure level from maximum noise plane l(0g = 0°) to minimumnoise plane (<flg - 90°l. Impingement angle, a, 60°; impingement distance, LT, 37 centimeters; noz-zle diameter, dn, 5.2 centimeters.

Effect of Nozzle Exhaust Velocity on Power Spectra

Data that show the effect of nozzle exhaust velocity on noise will be analyzed in de-tail in this section. The power spectra will be corrected for nozzle-alone noise, andone data point will be corrected for feedback whistle. From reference 6 it may be ex-pected that the noise generated by the impingement of a jet on a surface would be relatedto the impingement velocity. The relationship with noise can also be with the nozzleexhaust velocity VN because VN is proportional to the impingement velocity for thesame nozzle diameter, impingement angle, and distance (ref. 7). The nozzle-alonenoise (i. e., PWL generated by the nozzle without the board) for a 5.2-centimeter nozzleis plotted in figure 6 for several values of VN- Figure 7 contains the PWL for thetotal noise generated by the nozzle and by the impingement of the jet on the board ata = 90° and LT = 37 centimeters.

Before the nozzle noise is subtracted from the total noise to obtain the impingement-only noise, there are some items to discuss about the results plotted in figures 6 and 7.

The significant ground-ref lection cancellations and reinforcements occur within theone-third octave filters denoted, respectively, for Cj and Rj in figures 6 and 7. Thelargest effect is at the lowest velocity. No spectral correction was made to smooth outthese spectra because the effect on the results was found to be generally insufficient towarrant the effort of correcting the spectra. The spatial integration to obtain the PWLspectra was made over a hemisphere. This accounts for the fact that at high frequency,

11

(M£I.Ol "J3J) 3P 'IMd '|3A3| jafflod punos

*nS IS ^£^•° -o ._c . - c o;i » ro

7-Z "-22 E .2 =

'5 g<J

3S

CD

-IoTD

1" S

roo

•1.C

r "7*• 0)

Co

oCDO.

1CD

CD5aC3oCO

1

>oCD

^.S1

Lu.

iXo>

ooCD>

3TO.CXCD

CD

M

8"ro

sCDt^

OCD

CD

•oC3

eo

sd_coCD

_E

cCDO

CNJ

iri

ocCD

e»TO

QJ

OCCD3

g"

CD

Esc

>^

.1TJCDO.ooCD

(M£I.OI -jaJ) 9P 'IMd |J8/«od punos

12

where most of the significant data occurs, the microphone output includes the ground-reflection contribution. Because the impingement of thejet is normal (a = 90°) for fig-ures 6 and 7, the sound field is axisymmetric. The nozzle-alone noise is also axisym-metric; therefore, the spatial integration performed, which assumed axial symmetry,results in a true PWL spectrum for these cases.

The high-frequency bump in the nozzle-alone spectrum in figure 6 at V,^ = 344 me-ters per second is due to broadband shock noise (ref. 8). This nozzle-alone shock noisedoes not show up strongly in the total noise in figure 7 because the impingement-onlynoise greatly exceeds the nozzle-only noise. The high-frequency bump at the lowestvelocity in figures 6 and 7 is due to unattenuated high-frequency valve noise (shadedarea). These are the only data showing any significant valve noise, and its affect on thelower frequency impingement only noise is negligible.

The board at Lrp = 37 centimeters is about 7 nozzle diameters from the nozzle exit.Therefore, only the very low-frequency part of the nozzle-alone noise spectrum couldbe cut off by the presence of the board. This effect is therefore not important. It isdiscussed further in the section Effect of Impingement Distance.

A slight feedback noise (whistle) occurred at 2 kilohertz for the case of the board atVN = 295 meters per second. Since feedback noise is essentially an add-on narrowbandnoise, it is easily corrected when it makes only a small increase in the one-third octavefrequency band. The data point with this whistle is plotted in figure 7 as a tailed symbol,which is corrected (-1 dB), based on narrowband data, to the double-tailed symbol.

The sound power spectra (in watts) for the nozzle-alone is now subtracted from thescreech corrected sound power spectra for the total noise (board + nozzle). This differ-ence defines the impingement-only noise PWLp. The nozzle-alone noise data were cor-rected for small discrepancies with the nozzle exhaust velocities of the board data, by

Q

scaling the nozzle-alone noise by VN. Only the lowest velocity case in figures 6 and 7required a significant correction (2 dB). Figure 8 contains the resulting corrected soundpower level spectra PWL,-, for the impingement-only noise. Whenever the differencein PWL (e. g., difference in PWL in figs. 6 and 7) is greater than 4 decibels, an opensymbol is used. If the difference is between 2 and 4 decibels, a solid symbol is used.A 4-decibel difference would mean that the nozzle-alone noise adds about 2 decibels tothe total noise. For most of the peak noise data this difference is greater than 7 deci-bels (i. e., the nozzle adds less than 1 dB to the total). If the difference is less than2 decibels, the curve is estimated (dashed line) such that its shape is similar to theother curves in the figure. (This symbolism will be used for all the data that follow.)Similar data were taken over a range of V^ for a = 90° and Lrp = 51 centimeters.These were similarly corrected and the resulting values of PWLr are plotted in fig-

V>

ure 9.

13

140i—

O

ODAao

Nozzle exhaustvelocity,

VN,m/sec

350295240192138

Open symbols denote nozzle noiseat least 4 dB lower than totalnoise

Solid symbols denote nozzle noiseonly 2 t o 4 d B lower than totalnoise

Nozzle noise within2 dB of total noise

Double-tailed symbol denotes pointcorrected for feedback whistle

. 4 1 2 4 1 0One-third octave band center freouency, f, kHz

20

Figure 8. - Impingement-only sound power level spectra for the board. Impingement angle, a,90°; impingement distance, LT, 37 centimeters; nozzle diameter, dN, 5.2 centimeters.

90? C 2 R 3 , 3 R 4i i I i i i I i i

ODAD

Nozzle exhaustvelocity,

VN,m sec

340289236193

Open symbols denote nozzlenoise at least 4 dB lowerthan total noise

Solid symbols denote nozzlenoise only 2 t o 4 d B lowerthan total noise

Nozzle noise with-in 2 dB of totalnoise

. 2 . 4 1 2 4 1 0 2 0One-third octave band center frequency, f, kHz

Figure 9. - Impingement-only sound power level spectra at impingement distance, Lj,of 51 centimeters. Impingement angle, a, 90°; nozzle diameter, dN, 5.2 centimeters.

14

These corrected spectra for the impingement only noise are then summed up to ob-tain the impingement only total sound power level PWLpT. Figure 10 contains a plotof PWLCT as a function of VN for the data in figures 8 and 9, where a - 90° andL,,, = 37 and 51 centimeters. Figure 10 also contains a plot of PWLl^ (evaluated at

o o 8<PB = 0 ) results for a = 60 and LT = 37 centimeters. An eighth power curve VN

fits through each individual set of data in figure 10 fairly well at low velocity. A grad-ually higher power is required at higher velocity. The best straight-line fit over the

Q ccomplete range of VN would be about VN' .

One might expect jet impingement on the board to be a dipole noise source such thata sixth power curve would fit the data. Figure 10 clearly shows that V^ will not fit thedata. The analyses by A. Powell (ref. 9) and O. A. Phillips (ref. 10) showed that jetimpingement on a surface need not be a dipole noise generator. Both authors essentiallyconcluded that a large, flat, rigid surface could be a quadrupole noise generator (V?j)provided no significant noise is generated by the flow leaving the edges of the surface.These requirements appear to describe the experiment performed in this report.

Consider those points and the fact that the impingement only noise follows V»T and6 ft finot Vjg-. Let us turn the argument around. The noise data followed V£T not "VTr;

therefore, it is possible to have a flat, rigid surface that is large enough to result in aquadrupole noise source.

150o

\- oA

140

is•o *-'C p>

^ ^8 CO

T35 i6 u

130

120

110

.100

Impingement Impingementangle, distance to noz-

zle diameter,a,deg

909060

100

ratio,

} Pwk:T7.059.77.05 PWl^B-tfl/CT^Vi/VN

i i i i n ii ii200 300

Nozzle exhaust velocity, VN, m/sec400

Figure 10. - Variation of impingement-only total sound power level withnozzle exhaust velocity. Nozzle diameter, 5.2 centimeters.

15

In light of this discussion it would be profitable to consider some results fromreference 2. In that paper the impingement noise from the board of this study, and twoexternally blown flap wings with the nozzle under the wing (a slotted wing and a slotlesswing) were compared at the same a and LT (a = 60°, LT = 37 cm). In both of thewing configurations the total sound power level changed with the sixth power of the jetvelocity. But unlike the board, neither of the wing cases have large surfaces, comparedwith the characteristic wavelength of the noise generated. Also, unlike the board, theexhaust jet from these surfaces was of high velocity. Based on reference 10 it would ap-pear that these are the conditions necessary for the wings to be predominantly dipole

C

noise sources and follow VN. In reference 2 it was shown that the OASPL at different9 follows VN for the slotted blown flap wing at all values of B. The slotless wing fol-lowed VN everywhere but below the aircraft (70° < B < 120°), where VN would be acloser fit. The OASPL for the board follows V^ at all values of 9.

In summary, the impingement-only noise for the large board is essentially propor-tional to the eighth power of the nozzle exhaust velocity for a given nozzle diameter andimpingement distance (or the eighth power of the peak impingement velocity). The largeDoard is therefore predominantly a quadrupole type of noise source.

Effect of Nozzle Diameter

From reference 7 it can be expected that the peak impingement velocity V- isproportional to the nozzle exhaust velocity V^ and is a function of Lrp/d,^. In the pre-vious section it was established that the impingement-only noise for the board was essen-

8 8tially proportional to V^. Therefore, the noise would be proportional to V? for the5. 2-centimeter-diameter nozzle at a given LT- In this section the nozzle diameter d^will be varied for a fixed V. and a. To keep V- constant, it is necessary that V^be held constant, and the impingement distance LT must also be changed with eachchange in dN in order to maintain a constant LT/dN. Under these conditions the char-acteristic impingement diameter d. would be proportional to d,-. This diameter d.is discussed in a later section.

Figure 11 is a plot of the impingement-only sound power level spectra PWLp forthe board, for d^. = 2. 7, 4. 1, and 5. 2 centimeters, where a = 90°, VN = 289 metersper second, and LT/dN = 9.7. The spectra are essentially similar in shape, but thenoise level increases as d^ and the peak noise frequency varies with 1/djj. Figure 12is a similar plot where 01 = 60°, LT/dN = 7. 05, and "V\j = 188 meters per second. Thetotal sound power levels for these two sets of impingement only noise are plotted in fig-

oure 13 as a function of nozzle diameter to show the effect of dN more clearly. A d-^(or dj) curve fits each set of data.

16

130 p—

Jr>

QOO

Nozzlediameter,

IN.cm

5.24.12.7

Open symbols denote nozzlenoise at least 4 dB lowerthan total noise

Solid symbols denote nozzlenoise only 2 to 4 dB lowerthan total noise

Nozzle noise with-in 2dB of totalnoise

90• 2 .4 I 2 4 10 20

One-third octave band center frequency, f, kHz

Figure 11. - Variation of impingement-only sound power level spectra with nozzle diam-eter. Nozzle exhaust velocity, VN, 289 meters per second; impingement angle, o, 9fjP;

impingement distance to diameter ratio, LT/dN, 9.7.

Nozzlediameter,

IN,cm

7.85.24.12.7

Open symbols denote noz-zle noise at least 4 dBlower than total noise

Solid symbols denote noz-zle noise only 2 to 4 dBlower than total noise

Nozzle noise with-in 2 dB of totalnoise

80.2 . 4 1 2 4 1 0 2 0

One-third octave band center frequency, f, kHz

Figure 12. - Variation of impingement-only sound power level (evaluated at (0g = 0°)with nozzle diameter. Nominal nozzle exhaust velocity, VN, 188 meters per sec-ond; impingement angle, a, 6CP; impingement distance to nozzle diameter ratio,LT/dN, 7.05.

17

140

130

1— W Q)

£3^C Q. CD 120

110

Nominal nozzleexhaust velocity,

VN.m sec289283

Impingementangle,

a,deg

906060

Impingement distance to Type of PWLnozzle diameter ratio,

9.77.057.05

4 5 6Nozzle diameter, dN, cm

PWL•CT

8 9 10

Figure 13. - Variation of impingement-only total sound power level with nozzlediameter.

+10 —

'§T S%'S££ >-•OJ <U

s -3 ,b-"•

CQe "•la-5

DO

Nozzle exhaust velocityVN.

m/sec293192

Open symbols denote nozzle noise atleast 4 dB lower than total noise

Solid symbols denote nozzle noise only2 to 4 dB lower than total noise

J I. 2 . 4 1 2 4 1 0 2 0

One-third octave band center frequency, f, kHz

Figure 14. - Change in impingement-only sound power levelspectrum when measured and evaluated at maximum noiseplane, (c% = 0°) compared to when at minimum noise plane(<zjj • 90°). Impingement angle, a, 60°; impingement dis-tance to nozzle diameter ratio, LT/dN, 7.05.

18

In figure 12 the impingement is not at (?B = 90°; therefore, the noise is not axisym-metric. The impingement -only sound power level spectra plotted in figure 12 are thusonly a measure of the power spectra PWLU because the noise measurements were takenin one plane (<?B = 0°) and PWL£, was calculated by assuming that the noise was axisym-metric. To get an idea of the azimuthal variation of PWL£, with cp~, the board at a =60 was rotated 90 about the nozzle centerline, and the noise was measured again. The

A, was computed the same way again for this minimum noise orientation ((pB = 90°).Figure 14 contains a plot of the difference between the PWL^ spectra for <?„ = 0° and

n o<?B = 90 (where a. = 60 , LT/dN = 7. 05), and for two velocities. This curve indicatesthat this difference is independent of VN (or V. ) and that the <pB = 0° (maximum noise)orientation is no more than 4 decibels noisier than the <p~ = 90° orientation at a = 60°.Since the peak noise occurs at a frequency where this difference is about 4 decibels, itis estimated that at a = 60° the azimuthally averaged total power is about 2 decibelsless than the maximum noise orientation (<?B = 0°) value PWL'p. The azimuthal dif-ference can be expected to increase with decreasing impingement angle a.

Effect of Impingement Angle

The effect of the jet impingement angle a on the impingement-only noise is dis-cussed in this section. The loss of axial symmetry in the noise distribution when the jetimpingement is not normal (a < 90°) means that extensive noise measurements would berequired in several azimuthal planes in order to determine the azimuthally averagedimpingement-only sound power level. Instead, the measurements at various a wereessentially limited to the maximum noise plane ((p~ = 0°). Figure 15 shows the changein PWLL,, which is evaluated at cpR = 0°, as the impingement angle a decreases atconstant V. and d. (i.e., constant LT/dN, dN and VN). The noise level decreasedwith decreasing a. The center frequency also apparently decreased, but this is dis-cussed further in a later section. Figure 16 shows the variation with a of the measureof the total sound power level PWL{~,T which was derived from the spectra plotted infigure 15. Figure 16 shows PWL£,T evaluated at the maximum noise plane (cpB = 0°)and at the minimum noise plane (cp-o = 90°). A sin a curve correlates the maximumnoise plane data points well for a > 15°. The sin a function fails as a approacheszero because the noise could not go to zero. The square symbol at a - 60° and thedata point at a = 90° describe part of the variation with a in the minimum noise plane(cp-rj = 90°). The impingement noise could not be measured for a < 60° because thenozzle-alone noise floor was reached. The impingement-only noise appears to decreasewith decreasing a. much more rapidly in the <^ = 90° plane than in the <? = 0° plane.

19

ar

1

130

120

no

100

DoAO

Impingementangle,

a,cleg90603015

Open symbols denote noz-zle noise at least 4 dBlower than total noise

Solid symbols denote noz-zle noise only 2 to 4 dBlower than total noise

Nozzle noisewithin dB oftotal noise

Double tailed symboldenotes point correctedfor whistle

. 2 . 4 1 2 4 1 0 2 0One-third octave band center frequency, f, kHz

Figure 15. - Variation of impingement-only sound power level spectra evaluated at<PQ = 0°, with impingement angle. Nominal nozzle velocity, Vp^ 286 meters persecond; impingement distance to nozzle diameter ratio, Lj/dN, 7.05.

140

S.S 130c p.£ 5*c SQJ Q_

QJ —TO1 OJ

.£ SO.

- 120

O —

Azimuthal angle, (pgdeg

0 (max. noiseplane)

90(min. noiseplane)

30 60Impingement angle, a, deg

90

Figure 16. - Variation of impingement-only total sound power level, evaluated at </>Q, with im-pingement angle. Nominal nozzle velocity, VM, 286 meters per second; impingement distance tonozzle diameter ratio, Lj/dpj, 7.05.

Effect of Nozzle Shape

The effects of nozzle exhaust velocity and nozzle diameter on the impingement-onlynoise have been established previously. But the shape of the nozzle should also affectthe noise because it will affect the impingement velocity. To determine this effect airjets from orifice type nozzles of different shapes, but the same total hole area, weredirected at the board. The nozzle exhaust velocity was nearly the same (VN = 290m/sec), and the board was at LT = 15.2 centimeters and a = 90°. The shapes were asingle hole of diameter 4.15 centimeters, four holes (diam 2.07 cm), four holes of the

20

same size spaced further apart, a slot (1.04x12. 7 cm), and 16 holes (1.03 diam). Ve-locity profiles were measured without the board in place at the impingement distance(LT = 15. 2 cm) for these nozzles. These profiles are shown in figure 17. The impinge-ment velocity profiles are quite different. The jets for the four-hole orifices hit theboard as four separate jets; the jets of the 16-hole orifice have formed a well mixedlow-velocity stream. The jet from the slot orifice is similar to that of a single hole butof lower velocity at the point of impingement.

300

. 200>

100

Four Fourholes far holes

apart close 16 holes

oob~y Velocity-5S85-— probe

00p° path

295 Nozzle exhaustvelocity, V^,

m/sec

/-Four holes far apart

0 20 40 60 80 100Distance from nozzle plane of symmetry, y, mm

Figure 17. - Measured impingement velocity profiles for five nozzleorifice shapes. Impingement distance, LT, 15.2 centimeters; noz-zle total hole area, 13.2 centimeters.

The impingement-only sound power level spectra for four of these cases is plottedin figure 18. The single hole spectrum was not included because a strong feedbackscreech noise occurred. The low-frequency low-level impingement noise of the 16-holeand slot orifices go with their low-impingement velocity; the high frequency higher levelnoise goes with the high impingement velocity of the four hole orifices.

These noise and impingement velocity data will be correlated in a later section.

21

Nozzle orifice shapeFour holes closeFour holes far apartSlot nozzle16 holes

Open symbols denote at least4 dB lower than total noise

Solid symbols denote nozzlenoise 2 t o 4 d B lower thantotal noise

Nozzle noise within2 dB of total noise

".2 .4 1 2 4 10 20One-third octave band center frequency, f, kHz

Figure 18. - Variation of impingement-only sound power level spectra with nozzle orificeshape. Nominal nozzle exhaust velocity, VN, 292 meters per second; nozzle total holearea, 13. 2 square centimeters; impingement angle, o, 90; impingement distance, LT,15.2 centimeters.

Effect of Impingement Distance

The distance between the 5. 2-centimeter single hole nozzle and the board LT wasvaried from LT/dN = 1. 9 to 14. 5 diameters at fixed a and VN (a = 90° and VN = 295m/sec). Not only will the impingement velocity change, but feedback whistle andscreech, largely avoided in the previous data, may occur. In addition, the noise fromthe nozzle alone, which is generated within about 10 diameters, should be affected bybringing the board in close to the nozzle. This last expectation is nicely illustrated infigure 19, where the sound power level spectra PWL for the nozzle-alone noise andtotal noise (impingement + nozzle) at various LT/dN are plotted. Compare the nozzle-alone noise to the total noise at low frequency. When Lrp/cL, is less than about 7 di-ameters, a significant part of the low-velocity nozzle-alone noise (low-frequency noise)is apparently being cut off by the board. Feedback noise occurred (tailed symbols) atLrp/dN = 4. 7 and 7. 05 diameters and was approximately corrected to the double-tailedsymbols. The impingement only sound power level spectra PWL-, were determined.Wherever the total noise is less than the nozzle-alone noise, the impingement-only spec-tra were extrapolated. Figure 20 contains the impingement-only sound power levelspectra. It shows that the peak impingement-only noise is about the same for LT/dN

22

c

O < ODD > <§•

tl

S . i

."F — O)

? =1' fD E

c -C .2« £^=l a . ™

l > 00°S ~ z .

• c Z SO ? c en

i; o> t=

8

(M t l.OI ) 8PI8A3I J8"'"' punos ^|uo-)uaujs6u!dui|

(M£I_OI 'J3J) 9P 'IMd '|3A3|J8Mod punos

23

less than about 7. The noise level then drops for increasing Lrp/cLj. When the boardis close (LT/dN < 5), the low-frequency part of the spectrum is being cut off. This ispartly due to the fact that at small L~,/dN very little of the impingement velocity profileis at low velocity. Another part of the cause is the cut off of the production of low-frequency nozzle-alone noise by the board. Whatever the cause, the low-frequency partof the impingement-only noise spectrum shown in figure 20 is at best approximate.

Overall Correlation of the Impingement-Only Noise

In this section the total sound power level and spectral distribution of theimpingement -only noise are correlated with the velocity profile data at the impingementdistance from the board.

Total impingement-only sound power level. - In a previous section the jet from agiven nozzle was directed at the board at various exhaust velocities while holding the im-pingement distance and angle constant. It was observed that the impingement-only totalsound power level PWLpT correlated well with the eighth power of either the nozzleexhaust velocity or the peak impingement velocity. In the last two sections the nozzleshape and impingement distance were varied at constant exhaust velocity. These resultsshowed that nozzle velocity was not the correct velocity to correlate impingement noise.Impingement velocity was the better choice. The impingement velocity profiles weremeasured with no board present. Reference 11 indicated that the best correlations offlow and shear stress along the board were achieved by using the impingement velocitymeasured in this way. It was also observed that the impingement-only noise level wascorrelated by the square of the nozzle diameter, when the impingement angle and veloc-ity were held constant. This implies that the impingement-only noise would be corre-lated by the impingement area or the square of some convenient impingement diameter.In addition, the impingement angle a was varied at constant impingement velocity andarea. The total sound power level evaluated at the maximum noise plane ((p-r, = 0°),

, was correlated by sin a.Putting these separate correlations together it is reasonable to assume that the total

Q

sound power of the impingement -only noise PWLrT, would correlate with V; A- when\s J. lp 1

a = 90 . When a < 90 ° it should be expected PWLU™ would correlate with an im-o o '-' •*•

pingement parameter given by sin a V° A.. The impingement area is defined byip 1 oA^ = I d. for the slot nozzle and A. = (n/4)d- for the other cases. The characteristicimpingement diameter or width d. is arbitrarily taken as the dimension, measured fromthe impingement velocity profile where the velocity is 80 percent of the peak impinge-ment velocity V, . The noise at 80 percent of the peak velocity would be down about

24

8 /* 88 decibels, based on V.. The integral / V. dA could have been used instead ofA8 J&V- A.. But for a given impingement distance and nozzle shape VrLA. only differs from

/ V- dA by a multiplying constant for all the cases of this report. Therefore, V?LA.A.

is preferred because it requires only two numbers (V. and d.). Impingement velocityprofiles were measured for most of the cases in this report. The resulting peak im-pingement velocities and diameters (Vip and d^ are listed in table I. The impingementdata are in excellent agreement with the correlations described in reference 12. Thesecorrelations permit the reader to estimate V. from VN, LT/dN, etc., for manyshapes of single nozzles and multiple hole nozzles.

The impingement-only noise total sound power for all the cases in this report are2 8plotted in figure 21 as a function of sin a V° Aj. The figure shows data for normal im-

>AOD

O£|

D

Varied Vip atconstant Aj

Varied LjVaried nozzle

orifice shapeVaried dpjVaried dfjVaried a

VariedVaried

293292

289188286

3751

Varied15.2

VariedVaried37

5.25.25.2

VariedVaried

5.2

RoundRoundRound

See

RoundRoundRound

79

05 9(7

Varied

fig.977

17

70505 Va

150

140

S 130

C <DZ3 ^

. 2 - 1 2 0

no

100

Description Nozzle Impinge- Nozzle Nozzle Impinge- Impinge- Type of total soundexhaust ment diam- shape ment ment powervelocity distance, eter, distance, angle,

VN, LT, dfj, to nozzle a,m/sec m cm diameter deg

ratio,

PWLCT

(fOB • 0°)

/-'-Four holes, close; LT/dN • 7.35

'V--^Four holes, far ; LT/dN = 7.35

'.-Slot LT/dN - 14.5

^16 holes; LT/dN • 14.7

_L1013 1014 1015 U16 1017 1018

7 o oImpingement parameter, sirraVjpAj, (m/sec) m

Figure 21. - Correlation of all board impingement only total sound power level data with the peak impinge-ment velocity, Vjp, and the impingement area, A-r

25

opingement, where sin a = 1 and PWLpT is used because the noise field is axisym-metric, and for the a < 90° case where the total power PWL£T is evaluated at themaximum noise plane (cp^ - 0°). The correlation is good except for the cases wherethe nozzle was very close to the board and for 16 -hole nozzle. In these cases thespectral shape is significantly different than for the other cases considered. Compari-

son of figures 18 and 20 with the other cases shows that the low frequency part of thespectra for the close in board is lower; while the high frequency part of the 16-holespectra is low. This decrease accounts for their lower total power. A slightly better

95 8correlation over the whole range of velocity is achieved if V. ' is used instead of V. .Impingement -only noise spectra. - The next task is to correlate the spectral, or

frequency, distribution of the impingement only noise. This is accomplished by a nor-malized correlation where the dimensionless power spectral density of the impingementonly noise DPSD,-, is plotted as a function of a Strouhal number. From the previousdiscussion a Strouhal number based on the peak impingement velocity V- and charac-teristic impingement diameter d, would seem to be a good choice. Based on refer-ence 2, DPSDp is defined (for or = 90°) as

\s

/V. \DPSDp = 10 log in(— ±2- + PWLP - PWLPTC

and the impingement Strouhal number is given by

diS . = f .v.ip

The first DPSDC plot to consider is figure 22, where the spectral data associated withvariations of VN and dN at a = 90° are plotted. No extrapolated PWLp spectraldata are plotted on these Strouhal correlation plots. The continuous-line curves definethe bands these data lie within. These bands will be compared with subsequent Strouhalcorrelation plots. The dashed curves enclose the nozzle-alone noise spectral data re-ported by reference 4 and this report. The board spectra are sharper than the spectrafor the nozzle.

In a previous section it was noted that the impingement-only noise center frequencydecreased with decreasing impingement angle a. To better estimate the effect of a,a Strouhal correlation plot is shown in figure 23(a). The correlation is not too good.But if V. /dj is multiplied by sin ' a then the correlation becomes much better andfalls within the band (see fig. 23(b)). To complete the effect of impingement angle, some

26

10

•8

-10

^ -20

-30

Nozzle Impingementdiameter distance to

dN nozzle diam-eter ratio,

0 3. 01 7.(

D

A

D

O

0oV

b.

9.7

o 2.7o 4.1

alone data in ref. 4Open symbols denote nozzle noise at

least 4 dB lower than total noiseSolid symbols denote nozzle noise only

2 to 4 dB lower than total noise

Double-tailed symbols denote pointcorrected for whistle

1 1 1 , 1 1 1 1 1 1 1 1 1 1 1 , 1.01 .1 1 10

Impingement Strouhal number, fdj/Vj.

Figure 22. - Strouhal correlation of impingement-only noise data for round nozzles and normal impingement (a - 90°).

27

10

-10

-20

* *

From fig. 22

Impingement Strounal number, fdj/Vjp

(a) Correlation with d/V..

10r—

a

.9

-10

-20

-30.01

r From fig. 22

DOAO

Impingement angle,o,

deg90603015

Open symbols denote nozzlenoise at least 4 dB lowerthan total noise

Solid symbols denote nozzlenoise only 2 to 4 dB lowerthan total noise

Doubled tailed symbols denotespoint corrected for whistle

"nj1 , l i i i i , 1 , 1 I 1 L , ,

£ .01 .1 1 1

, 1 , 1 , 1 1 , 1 . 1 , 1 , 1 , 1 , 1 , 1 ,

10- 1/7Modified impingement Strouhal number, fdj/Vjpsin1"Tj

(b) Correlation with dj/Vjpsin1'2a.

Figure 23. - Effect of impingement angle on Strouhal correlation. Nominal nozzle velocity, V^, 286 meters per second; im-pingement distance diameter, L-[-/dN, 7.05; nozzle diameter, 5. 2 centimeters.

28

data taken at a = 60 to show the effect of nozzle diameter are correlated withd./(V. sin1/2 a) in figure 24.

The Strouhal correlation for the cases where the nozzle shape was changed areshown in figure 25. The correlation is good for S. > 0.1; the low-frequency data di-verge from the band. Figure 26 contains the Strouhal correlation for the cases whereLT/dN was varied. Here again, the correlation is good for S. > 0. 1 and diverges forthe low-frequency data where L,-,/dN < 5.

10

-8

i -20O

cQJ

O

-30.01

Nozzlediameter.

Oaoo

From fig. 22

i 1 i _L _L i I I i I i I i

Modified impingement Strouhal number,10

Figure 24. - Effect of nozzle diameter at 60° impingement angle where correlation uses dj/VjpSin^o. Nominalnozzle velocity, VN, 188 meters per second; impingement distance to nozzle diameter ratio, Lj/dN, 7.05.

10

-10

I

i -20.a

-30

Shape of orifice

D Slot nozzleA Four holes close togetherA Four holes far apartO 16 holes

Open symbols denote nozzle noiseat least 4 dB lower than total noise

Solid symbols denote nozzle noiseonly 2 to 4 dB lower than total noise

•From fig. 22

, 1 , 1 1 1 1 , 1 , 1 , 1 1 1 I I I , !.01 .1 1 10

Impingement Strouhal number, fdj/Vjp

Figure 25. - Effect of nozzle shape on Strouhal correlation. Impingement angle, a, 90°; nozzle orifice area, 13.2square centimeters, impingement distance, Lf, 15.2 centimeters.

29

10

Impingement distance tonozzle diameter ratio,

. LT/"N1.884.77.059.7

14.5

Open symbols denote nozzle noiseat least 4 dB lower than total noise

Solid symbols denote nozzle noise2 to 4 dB lower than total noise

/-From fig. 22

1 . 1 , 1 , 1 , 1 , 1 , 1 , 1 1 , 1 , 1 , 1 ,

-10

£ -20

.01 .1 1 10Impingement Strouhal number, fdj/Vjp

Figure 26. - Effect of impingement distance on Strouhal correlation. Impingement angle, a, 90°; nozzle diameter,dN, 5.2 centimeters; nominal nozzle exhaust velocity, VN, 293 meters per second.

SUMMARY OF RESULTS

The noise generated by an air jet impinging on a very large flat plate has the follow-ing major characteristics:

1. The shape of the total noise (impingement and nozzle) radiation pattern is essen-tially independent of nozzle velocity and reaches its maximum near 20° from the board(downstream end).

2. The impingement-only noise (total sound power level) was correlated by theeighth power of the peak impingement velocity and first power of the impingement area.The spectral data were correlated by a Strouhal number based on the peak impingementvelocity and a characteristic impingement diameter. When the jet impingement is notnormal, the total sound power level is reduced by the square of the sine of the impinge-ment angle, and the Strouhal number is multiplied by the reciprocal of the square rootof the sine of the impingement angle.

Lewis Research Center,National Aeronautics and Space Administration,

Cleveland, Ohio, July 6, 1972,741-72.

30

APPENDIX-SYMBOLS

A

Ai

UNdi

Al.

sOASPL

PWL

PWLT

PWLCT

PWL'.PWLlp

area normal to nozzle centerline at LT, m

impingement area at L™. based on d., that is normal to the nozzle2centerline, m

first, second, etc. , cancellation frequencies for ground reflections,Hz

nozzle diameter, m

diameter of impingement velocity V- profile where velocity hasdropped to 80 percent of the peak impingement velocity, m

frequency, Hz

width of one -third octave band frequency, Hz

impingement distance; distance along nozzle centerline from nozzleto board, m

length of slot

overall sound pressure level, dB

sound power level at a given f , dB

impingement-only noise; result of subtracting PWL for nozzle-alonenoise from PWL generated by jet striking board, dB

total sound power level, dB

total PWLC, dB

same as PWL and PWL™ but evaluated at (pR for nonaxisymmetricnoise, dB

for nonaxisymme-tric noise, dB

PWLJ-,, PWL!-,T same as PWLp and PWLpT but evaluated at

» R> ' ' '

SPL

Si

V:

ip

first, second, etc., reinforcement frequencies for ground reflections,Hz

sound pressure level (averaged and corrected for losses), dB

impingement Strouhal number, f d./V-

impingement velocity; velocity profile measured at impingement dis-tance LT without presence of board, m/sec

peak impingement velocity, m/sec

nozzle exhaust velocity, m/sec

31

y position of velocity probe relative to nozzle cross section axis of symmetry, mm

a impingement angle; smallest angle of nozzle inlet from board, deg

0 angle of microphone from the nozzle inlet (in microphone plane), deg

azimuthal angle; angle of microphone plane with respect to plane defined by thenozzle centerline and A-A in fig. 2, deg

32

/ REFERENCES

1. Dorsch, R. G.; Krejsa, E. A.; and Olsen, W. A.: Blown Flap Noise Research.Paper 71-745, AIAA, June 1971.

2. Olsen, William A.; Dorsch, Robert G.; and Miles, Jeffrey H.: Noise Produced bya Small-Scale, Externally Blown Flap. NASA TN D-6636, 1972.

3. Dorsch, R. G.; Kreim, W. J.; and Olsen, W. A.: Externally-Blown-Flap Noise.Paper 72-129, AIAA, Jan. 1972.

4. Howes, Walton L.: Similarity of Far Noise Fields of Jets. NASA TR R-52, 1960.

5. Wagner, F. R.: The Sound and Flow Field of an Axially Symmetric Free Jet UponImpact on a Wall. NASA TT F-13942, 1971.

6. Curie, N.: The Influence of Solid Boundaries on Aerodynamic Sound. Proc. Roy.Soc. (London), Ser. A, vol. 231, no. 1187, Sept. 20, 1955, pp. 505-514.-

7. Dealy, J. M.: The Confined Circular Jet With Turbulent Source. Symposium onFully Separated Flows. Arthur G. Hausen, ed., ASME, 1964, pp. 84-91.

i6. Simcox, C. D.: Studies of Shock Related Noise Fields Generated by Hot and ColdChoked Jets. Rep. D6-24486, Boeing Co., 1969.

;€9. Powell, Alan: Aerodynamic Noise and the Plane Boundary. J. Acoust. Soc. Am.,vol. 32, no. 8, Aug. 1960, pp. 982-990.

10. Phillips, O. M.: On the Aerodynamic Surface Sound from a Plane Turbulent Bound-ary Layer. Proc. Roy. Soc. (London), Ser. A, vol. 234, no. 1198, Feb. 21,1956, pp. 327-335.

11. Donaldson, Coleman DuP.; and Snedeker, Richard S.: A Study of Free Jet Impinge-ment. I. J. Fluid Mech., vol. 45, Jan. 30, 1971, pp. 281-319.

12. von Glahn, U. H.; Groesbeck, D. E.; and Huff, R. G.: Peak Axial-Velocity Decaywith Single- and Multi-Element Nozzles. Paper 72-48, AIAA, Jan. 1972.

NASA-Langley, 1972 2 E-6983 33

NATIONAL AERONAUTICS AND SPACE ADMISTRAT1ON

WASHINGTON. D.C. 20546

OFFICIAL BUSINESS

PENALTY FOR PRIVATE USE $300FIRST CLASS MAIL

POSTAGE AND FEES PAID

NATIONAL AERONAUTICS AND

SPACE ADMINISTRATION

NASA 451

POSTMASTER: If Undeliverable (Section 158Postal Manual) Do Not Return

"The aeronautical and space activities of the United States shall beconducted so as to contribute . . . to the expansion of human knowl-edge of phenomena in the atmosphere and space. The Administrationshall provide for the widest practicable and appropriate disseminationof information concerning its activities and the results thereof."

— NATIONAL AERONAUTICS AND SPACE ACT OF 1958

NASA SCIENTIFIC AND TECHNICAL PUBLICATIONS

TECHNICAL REPORTS: Scientific andtechnical information considered important,complete, and a lasting contribution to existingknowledge.

TECHNICAL NOTES: Information less broadin scope but nevertheless of importance as acontribution to existing knowledge.

TECHNICAL MEMORANDUMS:Information receiving limited distributionbecause of preliminary data, security classifica-tion, or other reasons.

CONTRACTOR REPORTS: Scientific andtechnical information generated under a NASAcontract or grant and considered an importantcontribution to existing knowledge.

TECHNICAL TRANSLATIONS: Informationpublished in a foreign language consideredto merit NASA distribution in English.

SPECIAL PUBLICATIONS: Informationderived from or of value to NASA activities.Publications include conference proceedings,monographs, data compilations, handbooks,sourcebooks, and special bibliographies.

TECHNOLOGY UTILIZATIONPUBLICATIONS: Information on technologyused by NASA that may be of particularinterest in commercial and other non-aerospaceapplications. Publications include Tech Briefs^Technology Utilization Reports and

Technology Surveys.

Details on the availability of these publications may be obtained from:

SCIENTIFIC AND TECHNICAL INFORMATION OFFICE

NATIONAL AERONAUTICS AND SPACE ADMINISTRATIONWashington, D.C. 20546