FAA FRD--77-29 (-44-

tV

AIWFRAME NOISE
PREDICTION METHOD

MARTIN R. FINK

MARCH 1977
FINAL REPORT

Docuknert is available to the U.S. public through

the National Technical Information Service,
Springfield, Virginia 22161.

Prepared for

U.S. DEPARTMENT OF TRANSPORTATION

FED)ERAL, AVIATION ADMINISk.AATION
CSD stems Research & Development Service )C
Washington, D.C. 209 AOri9
LJJr
MWAY ~111L

Ii ~LL LL
 N~frICE

This docum*,nt i's disseminated

Department of Transportatior, inunderteSoiSrhpode
thcnees of
exchange. Lhorato
The United States, Golvertrrnent
for its CJontents or use thereof. assumnes no lIjabti~t,
 Technical Ik.port D)ocumuentation Page
1. epot N 2.GovrorrotAcession No. 3. kecip cot . CotologV No.

f irram Noise PrditonNeth~od1Z v rq~n,,e! or ;,d

/ ( .7 77-9126W7-11
--f -rng'Cr~o
-~ ~ f 10. Work Unit No T AS

United Technologies Research Center

Silver Lane v-.,Gs~~~any?4b

E. Hartford, CT. 06108 fl- )DO"T-FA76WA-1Wý21

3.
TT ye o R CoEred
Perio~d
12Agency Nome an'd Addres-s F inal Repcrrf.r
Federal AviatLion Administration (Juiin 1976 -April 1,977) e
800 independence Avenue, S.W.L
Wahigon ).C. 20591 14. Spons$oring Agency Code
WashigtonAR D-550
is. S mentotry Notos
lup

11, C. Th'ue, DOT/FA.A Technical. Monitor

16.
A nois;e compýonent method is presented for calculating airframe n o i,-e Noise
fromn clean wing- and tail surface is rep~re.s-:,nted as, trailing edge itoi.;e cauised by
the turbalent bou-ndairy l~ayer, Land~i-ng, gear poise is given by an empi rica.]
I epre,;entation of' model data. Trailing edge flap noi se J.,is odjel e(i ustij I1iltt
dipole no~rrsl. to 'i-le dleflected flap, with ampilitude aind .3pectrwn given by a
c~orr-elation of flyover data.
Meas axed fliyov ~r data for the Prue -? sail plane, Aere) Comsiwdeýr Sh5.rke
greneral aviaLion airplane, Lockheedi Jet,,tfr busi ne.s jet, xioelng yb' conmitrc~i n
Jet., -Ind Convair F-106B deltan w.i-ritairplane are used. as test ca:;osý iTiheý;(,stwt

are, comparedi with predic;tion,; by this; method and the NALIA Af'NfO M Il1 ru It

Niot I Lir irf

rimm nol :'e edp- noi II~ t, wi ?w1 t,.1 avvi I i
F 1K
vi i)
1" Iý
ý I501 14', feIý'e fjIap '1 0 :,i' i illi 1l 111 1jiiourk ii huh-iI
rt
1, 1'~ON v i
I~.'e ý~ jr
i~nt'I , I V/~

19 (of li 1 ptN

Forms DOT F )7110.1 ''tedrir of rom-pleti A pngrt tehrid

S-- - = -.

* -
° - I. _ '0 - -- •

*C - • • 4

U 3

3-

S. . .. , . \.
 TABLE OF CONTENTS

Page

1.0 INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

2.0 LIST OF SYMBOLS . .. .. . . . . . . . . . .. . . . . . .. . ..

3.0 DEVELOPMENT OF AIRFRAME NOISE PREDICTION METHOD ....... .......... 5

3.1 Clean Wing and Tail Surfaces ............. ................. 5

3.1.1 Overall Sound Pressure Level ..... ................ 5
3.1.2 Spectrum Shape for Trailing Edge Noise ............ . i..11
3.1.3 Diiectivity ............. ..................... ... 16
3.2 Noise Component Method ............. ................... .. 20
3.3 Landing Gear Noise ............... ...................... .. 23
3.4 Trailing Edge Flap Noise ................... 31
3.5 Leading Edge Slat and Flap Noise ....... ............... 35
..
3.6 Graphical Method ............... ...................... 39
..
3.6.1 Clean Airframe ............ ..................... ... 39
3.6.2 Leading Edge Devices, .............. ............. . 40
'.6.3 Trailing Edge Flaps ............... .................. 40
3.6.4 Landing Gear. ................. ...................... 40

3.6.5 Annoyance-Weighted Noise Levels ....... ........... 41

4.0 NASA ANOPP METHODS ................ ......................... .. 52

5.0 COMPARISON OF MEASURED AND PREDICTED FLYOVER NOISE .............. 54

5.1 Limitations of Available Data .............. ................. 54

5.2 Aircraft for Noise Comparisons ............................... 55
).3 Prue--2 Sailplane..................... ....................... 57
I .). Aeocommand etar Srk....................................C
5. Aero Conmd der Shrike ............. ...................... 6

. oIn 747. .................................... '

5.' Convair i-106B ................... ........................ 77
5.4i Addi tional Aircraft .. . . . . . . . . . . . . . . . . . . . .. . I
TABLE; 1 - FLYOVER NOISE TEST CASES ..... .............. ....
TABLI'; 2 - COMPARISON OF MEAS'URED AND PREDICTEI) MAXIMUM NOSE-

WI L
1. AIL- ao z- 1L

S!-
 6.0 POTENTIAL CONCEPTS FOR AIRFRAME NOISE REDUCTION .......... 90

6.1 Basic Geometry .................... ........................ 90

6.2 Trailing Edge Flaps ........... . .................... 9.
6.3 Landing Gear ...... ...... ......................... .. 93
6.4t Clean Aerodynamic Surfaces ........................... 9.

7.0 CONCLUSIONS ........................................ .. 0

8.0 REFERENCES (17..............................'

-9.O APPENDIX I: MATHEMATICAL CONVENTIONS FOR COMPUTER PROGRAM . . . 10

10.0 APPENDIX II: COMPUTER PROGRAM FOR CALCULATING AIRFRAME NOISE . . ou

10.1 General Description .......... .................. . (.

10.2 Program Input Format ......... .................... .....
1(0.3 Computer Program Listing ............................... ...1(o
10.4 Program Output ................. ....................... 122

ii.0 APPENDIX III: TABULATED AIRFRAME FLYOVER NOISE SPECTRA. ...... 12(-

TABLE 3 - AERO COM14ANDER S}HRIKE, 500 FT ALTITUDE o..........120

TABLE 4 - LOCKHEED JETSTAR, 500 FT ALTITUDE ........ ............. 129
TABLE 5 - CONVAIR 990, 500 FT ALTITUDE ..... .............. ...
TA13LE 6 - BOEING 747, 500 FT ALTITUDE ........ ............... 5:
TABLE 7 - VICKERS VC 10 FLYOVER SPECTRA ........ .............. .
TABLE 8 - AIRF.RAME GEOMETRIC PROPERTIES ........ .............. .

vi
 1.0 INTRODUCTION

Noise emission from current airplanes is dominated by propulsion-system

no~ise. Certification levels for aircraft noisey as specified under Federal
Air Re•'ulation (FAR) 36, have been achieved by use of either existing high
bypass ratio turbofan engines or existing low bypass ratio turbofan engines
plus noise-attenuating nacelles. Systems studies of noise levels for future
advanced-technology transports (e.g., Ref. 1) have concluded that engine
noise levels 10 effective perceived noise decibels (EPNdB) lower than those
of FAR 36 could be achieved with early 1970's technology high bypass ratio
fturbofan engines combined with extensive inlet and discharge duct acoustic
suppression. An additional 5 EPNdB noise reduction was predicted for advanced
(1985) acoustic liners. Still another 5 EPNdB (that is, 20 EPNdB below FAR
36) was predicted by use of special noise abatement procedures combined with
advanced liner technology. More recently, acoustic tests have been conducted
with full-scale high bypass ratio turbofan engine hardware (Ref. 2). These
tests demonstrated that noise levels corresponding to 12 EI4dB below FAR 36
at approach and sideline, and 9 EPNdB below at takeoff, can be achieved with
1975-tech-nology acoustic liners and sonic inlets. However, these reductions
of propulsion-system noise would not cause equal reductions of total aircraft
lnoise at the certification points. Airframe noise, generated by motion of
aircraft external surfaces through the air, is believed (Ref. 3) to be only 10
to 15 dB below FAR 36. Airframe noise may impose a fundamental noise floor
roughly equal to the demonstrated noise from high bypass ratio turbofans
with ertensive inlet and exhaust acoustic suppression. Future certification
levels must be based on what can be achieved with economically viable air-
frames and propulsion systems. Thus it is necessary to understand the
fundamental processes of airframe noise radiation.

Several methods for predicting airframe noise had been examined by NASA
Sunder their Aircraft Noise Prediction Program (ANOPP). Two airframe ncise
prediction methods, the clean aircraft methc)d (also called total aircraft
method) and the drag elemert method, were recommended (Ref. 4) and were

1Brines, G. L.:
Studies fo ý 'eiOermininin the Optimum Propulsion System
Charactýeristics for Use in a Long Riai•i Tran,;port Aircraft. NALxA CR-120950,
July 172.

SOfr'n, T. G. a'ud Riloff, N. Jr.: Two-Ztage, Low-Noise Advanced

'"echrioiy Fan . V. AcFUosi 1 inril Report. NA[IA CR-1348'31, Sept. 1975

J Mo)rgu, [I. S]
.and Hardin J. C.: Airframe Noi:,e - The Next Aircraft
Nol 2• ] rr:ir. Journal of A irora ft,, Vol. ]2", No. 7, July yf97T , pp 522-621i
S9tlar!,
!' ri, 1 C., Fratel 1, JT. Hayden, R. E., Kdadman, Y., and Africk,
8
S. Pr'( 1. (t,, ( A'f Ai f'rvfrie 14 iso . NAW
A TN D--823, , lob. 1975
subsequently programmed by NASA Langley Research Center. Verification by
comparisons with measured airframe noise spectra have not been published for
the clearn aircraft method and werc available for only a few cases for the
drag element method. A lengthier component analysis method was described in
Ref. 4 but its predictions were nol compared with data. For several air-
frame components, differences exist between the dominant noise mechanism
assumed by the drag element method, the component analysis method, and other
investigations.

This program was conducted to (Task I) identify airfraLe noise

generation mechanisms and sources, and develop both a graphical method and a
digital computer program for predicting noise from each source, (Task Ii)
verify the resulting airframe noise model by comparing its predictions and
those of the NASA ANOPP airframe noise models with data for a wide range of
aircraft type and size, flight speed, flap and landing gear position, and
wing planform, (Task III) identify concepts for airframe noise reduction, arid
(Task IV) develop a program plan to verify empirically the assumed airframe
noise generating mechanisms and potential suppression concepts. Results of
the firsIt three Tasks are reported herein.
 2.0 SYMBOLS

Symbol Description

A Reference area

A(M) Convective amplification factor defined by Eq. (A-14)

b Span of wing, horizontal tail, or vertical tail

ca Atmospheric speed of sound

CF Trailing edge flap gross chord

D Landing gear wheel (tire) dimneGer

D(6,0) Normalized directivity factor defined by Eq. (A-8)

f One-third octave band center frequency, Hz

fmax One-third octave band center frequency at which sound pressure

level is maximum, T{z

GI -G5 Empirical functions of S'rouhal number in Eqs. (11) through (20)

for landing gear noise

,G
G-7 Empirical functions of Strouhal number in Eqs. (21) through (28)
for trailing edge flap noise

h Al ti tude

t length of lai~ding gear .-trut exposed beneath wing or fuselage

J Ac tiJ.c i nten.,;ity

Nn;qu
Mi are acous•t.ic pre.;.;ure

P' ACoA(:;Aic power

t
,1)1 .i mice tfrom airplane to observer, evaluated at retarded time

ii /!Iii li
i ii'( a t

IS t'l ) l,,'m I i te:d 'tue-- I i rd C l,:Iay• ;<ec i' Wvl fane to

 Symbol Description

t Time

V Airspeed

a Root mean square normalized turbulence intensity

/*,yE Sweepback angle at trailing edge

Trailing edge flap deflection, deg

8W Turbulent boundary layer thi.ckness at wing trailing edge as

calculated for a flat plate

Azimuth angle in plane defined by flight path and observer

position, evaluated at retarded time und measured from approach

direction

Kinematic viscosity of air

Pa Atmospheric density

9) Direction angle from vertical plane through flight path,

evaluated at retarded time

1 Angular frequency

Subscripts

E Edge noise

f Evaluated within one frequency band

F Trailing edge flap

( kanding gear noice

Low(ur

I 1iift fiictuatiori noise

u IJpp,-:.r

W WIjnjz (extqerdetl to center i ne)

4
 3.0 DEVELOPýENT OF AIRFRAME NOISE PRET)TCTION METHOD

3.1 Clean Wing and Tail Surfaces

5.I.i Overall Sound Pressure level

The basic concept of the method developed herein for clean airframes was
first presented in Fef. 5. An updated description is given here. Noise
radiation from clean airframes is assumed to be caused by cbnvection of the
wing and tail turbulent boundary layers past the trailing edges of those sur-
faces. The resulting trailing edge noise radiation has been studied analyt-
ically (e~g., Refs. 6 and 7). Functional dependence given by those analyses
has been verified experimentally (Ref. 8). It is assumed that noise caused
by an upper-surface and a lcwer-surface boundary layer is randomly phased so
that separately calculated acoustic intensities can be directly added. Far-
field acoustic pressure then has the dependence

5
p2 - aU2(6b/r2)V cos2 ATE cos2 pcos2(8/2) (1)

where a ind 5 are the boumnary layer turbulence intens;ity and scale length,
respectively. Maximum overall s;ound pressure level would occur in the fly-
over pl tine where oýileli ne tintle iis zoro and far-field di,,;tance r i:" related
to ill !J (e
hI 11 )y

h r sn & (2)

"t"iIlk, 14. k . Apj)ruxxirnAte Prfd1('t.ioh ot' AjirtrnwTA N1`46 ;,. J, urliil of

A tirt' lt , V,,I. I- , N(,. 11 , "ov. 10Th, pp - b d . Al]: , AlAA 1',,iper -',t

ii* ..wt .1 W 11i ten.;

m;, 4. tol li ii ti .[. i, . Ii .: e'd riii O l~ ;le-lt n b
'I'l bl il ,.t. F1ow ill lt, Vi,.i •it.y ()I ti :.ct ,i ilft] I 'lal, ,. LA)II'lr tftil " Hlli i

M. .:
I(;kh.::', ',,It'j tý,iI.illtc(t l y lorbtijeijt1 -l ow (O ' a RIi.i±1 llu.,j
}t.i '
asý; Ott a • t,11 Fl iti W alv"(A'ptt.o, -jt•m' ( i'J[vhrolxrii. Jtlu a (t)f"
/1. ' Ii:: I , s• e'i
'y ,I' tine ' ",,
ui V I '. , Nt,
ll( - I-, ti 7,' At-pI) I('¢'
i( , 1(' 1

Pti'.k, !. 1H.: N.!,'rlIlttt !ti .:I,.*•t S i. It ,'I lit!', K lo, I'•• lraii I~ir• fftifft tin I
A lt •t
*" t itq H AiAA I.Vr, , I] .',
No i I 1i(,v. Iq l,
Also, most airplane wings have small tr&iling edge sweepback angles A so
the dependence of far-field acoustic pressure on flight condition6 an(Adirec-
tion angle during a constant-altitude flyover is

S2 (Bb/h )v Sin2 e COS 2 (8/2) (3)

The antle from the approach horizontal direction, 0, is measured at the

retarded time (the time when the measured sound had left the airframe),
motion effects are temporarily neglected because of the low subsonic SSource
Mach numbers of interest. MaximUm far-field acoustic pressure due to trail-
ing edge noise then occurs when this angle is approximately 710, that is, 190
before tihe trailing edge is overhead. Depending on flight Mach number, the
wing pos,,ition when this maximum noise reaches the observer could be ahead or
behind the overhead direction. This result agrees with avail&ble data. In
contrast, if a lift. dipole directivity is assumed, maximum noise measured by
a fixed ob,orver would leave the airframe at the overhead position and would
-always• roach the observer when the airframe was considerably past that
po:; t, i on. Suchi directivit y has not been reported.

Nexi , it Wý1:. noted that tilrbuience intensity within a turbulent boundar:y

•yer i:, :•pIpprtxixaSily indeperident. of Reync. ds nrimber for conditions typical
:•runtt'1.
sl w . Boundliry
-' layer thieKne;s varies with chord t.o the O.1
piwer ikad vel,,city t tthe -'0 power. For simplicity, those exponent!- were
:ipproximn1,ae inahle'. ,by tlhe cloise5t integers. Far-field maximum acoustic
p V-O;:;Sre, k• t1 (11 be expetd ti
(1 ,,o be prooportiional to the product of' wing
i- ,e rd a xi ::pan (thhe wi n• area, div(led by altitude squared, and to vary
&i)
wi lh I'ly(,vr vtel[oc ty t(, the tfifth pIower.

II'hi a"ptprx i ma V im wai

w: c hedoe I by adjut i.laj, a l t.ge number of mea:;ured
lv aly*•i
I );oiti pre2:.x lvel.; as UOAPtI,-lt xi, (i/hi) llnd p1 ott.
, iip,
! iLL +1t;A1 It ' :istj illlt veli.i'ity. Anx kxpdlt edi ver.;ion of this ca mjx xrxi:;t+n i:;
II iý'l1r I CIii siX Itcti ýw( tii yxxin.v cxxi ly cleain iti rfrlvom ::. 'Vit* lixta
C '' Lam
'['tL ' kt• 1'c. aire Jevel:; 1'for thth ideal i ('.l (:;sm ot,hdd) ,;pectra ratlht r
ttiAit j i t,,:; ot. it ) :it•.'tra . TWi:; :Trumoitihirig proct.ess; was:; tie:crih e~i il,
St,'l'. ai. allI•lv il tko ".. .. i '1 ut. oett'-cts (tile :;,( e ly to the ýrc l'eratl
thoste
ittt li
iixiiiI*tI n so 'ol't thai beiry- pecaliar to any i(dAividlua al.crui "
I -,*xf,li tlnd ['
leathe o- t p'r 's vs , l w-I-re quexxcy gr uond ret
l 1,Ct . ,
*-,, t h,- na ,l," r r• re-m' v',l
' t r'r0v O:, :spectra. Iln the 'i A'
I he

,,,~ tl!u
A+. t' , i.,,. NA.1il ,'/, [h , .
Doc,' P )Y/.
 120

-t'

0 110

13.

< CONVENTIONAL LOW SPEED

100 AIRCRAFT 0Oi

< A -JET AIRCRAFT

0
90
o DSYMBOL AIRFRAME
uj
H
o C CV 240

-0- 0 DC 3
J A SHRIKE

>0 80 2- 33
u JJ I &C 5A

an
D
0 F 106B

IJJ
0 0 JETSTAR
Xn A CV 990
o 70 B147
B-
z:D 0 HIGH PERFOORMANC
• HS 125
0
0 SAII-PLAN[ .AC 111

w 1• -VC(10
L. HP 115
o 60 1l, IBL t I
O ,'0 i-f u E 2
41'×z 2 3•2

40 60 80 100 2?00 300 400

_ItLY)VI: u Vi C'I Y KN IT;

t4W'CM. 1 31 I I KN,
\:

S4t JIt /14

 Prue-2 sailplane, not only tones caused by laminar boundary layer inustability
(Ref. 10) but the entire low-frequency part of the spectrum below 200 1iz had
been removed by the smoothing process. The resulting .smoothed ad.justed
levels for this airframe probably are too low. In contrast, composite
(actual) levels were used in the NASA ANOPP ccrrelation given in Ref. 4.
Data for the Cessna 150 from the Ref. 9 study are not shown in Figure I
because that airframe had a fixed l.andirn, gear. Additional data shown by open
symbols were taken from Tables I and III of Ref. 4., Only about half the
tabulated data points for the Schweizer 2-32 sailplane are plotted; the others
were dominated by laminar instability tones.

All data points plotted as open symbols were measured with microphones
mounted on posts or tripods 4 ft above the ground, as specified for noise
certification measurements conducted under FAR 36. Resulting one-third octave
spectra tend to be oscillatory at low frequencies because of sound wave can-
cellation and reinforcement. At higher frequencies the directly-radiated
flyover noise, and flyover noise reflected from the ground to the microphone,
combine to give a measured level approximately 3 dB above directly-radiated
noise. Spectrum irregularities caused by phase differences between directly-
radiated and ground-reflectea waves can be eliminated by use of microphones
mounted flush with the ground. For frequencies of practical interest, the
direct and. reflected waves should be in phase, causing measured levels to be
6 d.B above directly-radiated noise. Airframe noise data published by NASA
Dryden Flight Research Center (e.g., Ref. 11) and Royal Aircraft Establishment
(Ref. 12) have been measured with flush-mounted microphones. These data,
decreased 3 dB to allow direct comparison with data measured for certification
purposes, are plotted in Figure 1 as solid symbols.

The data ;ymbols generally fall within three groups. Adjusted OASPL for
two of the three high-performance sailplanes, the F-106B jet interceptor, and
the BAC 1-11 jet transport, are closely matched by a solid line given by

10
Fink, N. R.: Predietion of Airfoil Tone Frequencies. Journal of'
Aircraf't,, Vol. 12, No. 2, keb. 1975, pp .118--10.

.'Ittnam T.W.
W." Lasavna, P. L and White, K. C tasurements
W and
Ananlyse,, of' Aircraft Airframe Noise. Aeroacoustic,s DTOI, Noise; Airframe and
Hise , V01 . 14'), prcgres s in Aeronautics and As;tronautics,,, M.1 T.
Airfo I 1'oI
Pres",, C(imbridge, Mas•s , , (, th3-ý",78 . Also, AIAA Piper 7%5-'10, March ] O7K
:.. . , 1. An ExT.,er.merCen',a 1 ýtudy of A irframne ].f'Naise. Aer -
½0t t( t 2,',TC4, Noi.-e; Air Cramn and Air:'ail Noise, Voi.. WiC Drs res'• in

Al :,; A. AA lltij " (" ' ,r , r 1 '{ .

) . .. . . . . . . . ..... .. .. •... . . . , , . . • , , . . .h . . .a , : - : • . .. .: :t • t 4 : -•i• : • , • : l , ' . .. ' , ,, , ' : : l ', ' ' i ' : • •. . 1' ' • • • . • ~ . .I
 OAS PL- 50 log (V/100kt)+1 l-~og (S/h 2 )+ 84dB

Note that aspect, ratio does riot, occur in this equati.on, andt that aspect. ratio
for, the Libeile and F-i.ObB dIiffer by a factor of more than ten. In contrast,
prediction equations given in Refs. 3 and 9 include a variation with aspect
ratio to the -2 and -4 powers, respectively.

Data for conventional low-speed aircraft such as three retractable-

landinig-gear prope.Ler--driven aircraft of' Ref. 0arid the Schweizer 2-_33
.. trult-braced trainingf sailplanie f'all on a liine d3b above that, given by Eq.
(4i). Als;o mratched by this line are data for the Handiley Page HP 115 ilender
delta-wing reseatrch airplane, wh~ich ha~s s;mall streamlined fixed landingv gear.
Tibis airplane has an aspect. rati~o of 0.9, and equaations; which predic a,
dependence of. nois;e onl induced dragý or aspect -ratio gr-_eaitly overpredict (Ref.
12) itUs mea.sured nos.Data for ,Jet aircraft generally lie between there
two curves-. A reas;onable prediction for business jets (Jet~tar and H3I~125)
and large commaercial 'jets (B-74i7 and CV-590)) is, 4 dB above the level t,4ven by
V ~~Eq.(t)
TPhese ai~r(iýr.t and flight condlitionsý corre~spond to a factor of about. 40
in Reymolds, number anid therefore ti factor of twc in the rat'io of flat plate
turbulenl, boundary layer- thickness, *to wing, chord. The calculation method for
noise of' clean airframes as developed in this report assume~s that the noise
intensity is, proport~onal to( the product of' flat-plate turbulent boundary
layer thickness,- anid wing spari, as express-ed in Eq. (3). This thickness was
calculated for a chord l(n,-th equai to the wing gross geometric chord (ratio
of' gross,- wingr area to grossý- wing span) . This, parameter was, chosen becaus,-e
thes-e two, ruanti t~ies, are more readily available than the wing mean aerodynamic
chordt, 0 xpo5&ie area:, or lxp()lied span . Tur bulent boundary lay-er thickness;- was
t~aken as;

(5W n0 3 7(SW /b,)(VS w/bwi) 0

()

M('14ý_Wt'u IWX1115141U fjAyover Val e: oftA:.Pl,

(,)I for the same data that were usdill
Hu
.I 1) tadIj14' t.i fo11ipe~ee
a on the producl of01 boundary layer Wi. k -

ij 0). f 1)" 1 rl ii i I r li,'

4,t t ;(t 1 1 1att-.
co l
y i t, ( i
rcI:tt ve ti h 11e :;ýiii 1c r
1,o()

It~iji01 II t~, t))lii

ii ty ktyti .r l~O~ ti.0 Wiire -)Iori. I atitA for a-n
i1w !v~~ti
, t44t il i14410 :4i i411 40 O ~
4po" u~~ 1y
 140

-j

< 130 Ile

0C
co

C14

Cd)

a- l OC :A
U)9
>0 00

10
U) *H 2

W~i 01 ANES

(1) V

90 00 v0 R 04

0 YFV-I i ). N1

i~~~1•~~~iI~~~~H1'ýKVj!K Ic T u 1 2tK

HIGHHFO NIANC
 OASPLz 50 1og (V/i00kt)+ io log (Swbw/h 2 )+ 101.5dB3 (6)

Data l'ur convent~ional low -sp;eed aircraft were 8 dB higher as before, but dat
for jet aircraf.t, hadi an aver-age value roughly 6 dB above the line for aero-
aynamica. ly Cl ean saidlplanle:lO. ore teutmt test of this pred~ci a in
roe thud iý rinot, the ag~reement with mreasured OASPL but the agreemenit, be--tween
predi~eted anid nirsre pectra. The actlual. equatilon used for calc ul ating,
trail int so :i:, given i n the d~i seus iJon of di roctiv ity ,
rwegeni fCt~er the
I~~~io
1 I5 III01 ~ fxli~.rum s-,hape,

.1 . ' I r' I vilhpe to r Tra ilini l,,F Polie

If the MiUj or noin- 5- gerie L,,)I inn 150chan 1 511 f or ai1r frimle n0-11)1- cef
uerodynani cally clean airt'ra.mes7 is traili ng edffL noise, the ispectrum s;ha.pe
For aiirframe~ no)Ci se sýhouldl be given by the exi sting) sollutions,- for tra.13ingjr
edgWe nulo;e .A emerieinpi ri cal e(Iluati on ilof normalized -spectral dens;ity of
trail~ing, edge nioi;,e from externally blown flaps, was given us,- Eq. (12)) Of IRef.
3 eqaL onhad been found to predlict thýe n~olse s-pectra of' upper
.Thi olr -
lace bilown flap.. f)r, measulrement di rections;, exhaus-t, vel ocities ;, and frequenicy
ranglfes where suIch nois.,e was; not dominated by directly-radiated quadrupole
noisie from the deflec.t~ec Jet. Converting, from normalized spectral density and
2,'tZOuihua number ba.;ed on Jet. diameter to third-octave Sound pressure level.
relative tou overall. sýound pressýure level QASPE and center frequency f
relative to the center fre-qu-ency fma whicc ilsmxiu PI

SL1/3 -- OASPL : 10 og{ýO 613 (f / f ,0,)' 1 [f/frnax) 3/2 + 0 51i} (7~)

Tile .3c t1111Silpeg.Veilý by this: eqilai~on is- compured in Fig~ure j wi th the

ralolih 1en,1; ic(01.1 d Ift'rame pect rUM rec oinrnended
-a.ie for us;e wi (11 the NAS;A
ANIP1 li'f rvai. Aalyis(Fig~ure )i of' Ref. 'i.). The two dlotted curves;
ax",e buiin Iar-ifi )t slilothed
ýI iiidI
)Ii lels onul' ýipet Pa from twenty -ei ght fli A htis
- t'iiV il'et ri tetd hsjibonly
vl: (Ref. (). Thll: ý;oliti curve Ir-awn
to I lii'.'' ljI)nar I .I; was; taken trosi Reif'). The ()opn smo a iia

L >~
i (( ak P,(in I l Ký i ,f' Exc 'Iil-11 ly PI1lhmn W1111iail) mnr Treu lei
 BOUN DA IF S5OFWSMOOT(HEDI
cf) SPECIRA, HEAL Y'S DATA

01
Lu

- N
LuJ
20 /
CL TRAIL.ING EDGE NOISE-ý

Z NASA FOTAL AIRCRAF T MAE

TH00
</
i -30 1#4

u CALCUL-ATLD TRIAlt ING EDGýE\NOISE

o PLUS HIGH f RFOOENCY DECAY

40
0.125 0.25 0.5 1 2 4 8 16 31.5
RATIO OF 1/3 OCI AVE CENTER FREQUENCY TO
MAXIMUM AMPLITUD)E FRFOIUENCY (IIl,)

FIGURJ~E ~ A
AlI l'RM\N : NO[l:'F M)NMM/\~'.E! :j :(l A l ~H~'jI~
:I;VI-,RP~l, 101):~
Mv-H'I

104
 ratios,, up to 8. Al larger frequency ratios they decay less rapd].y than the
(ldata. The measured rapid spectrum decay at large frequency ratios might be
"a real property of the noise mechanism but more likely was caused by atmo-
!;pheric attenuation of the radiated flyover noise.

Thiý! difficulty in predicting the measured spectra at large frequency

ratios had been encountered by Revell (Ref. 16). His semniempir:ical drag
analysis' within the NASA ANOPP method (Ref. 4, pp 33-36) approximates the
rtornm].ized airframe noise spectrum shown in Figure 3 by four straight lifies
multiplied by an exponential decay of mean square acoustic pressu.tre. Without
that, added decay, his equation for frequency ratios greater than 4 gives a
spectrum shape that varies inversely with frequency ratio squared. The
Sasymptote of Eq. () at Jarge frequency ratios has the same variation.
Revel. ',, correction factor can be expressed as an added term

ASPL.1/ 3 :--0.02 (f/f,.,ox-i)1 3/2 log e (8)

Agreement with the solid curve in Figure 3 would be improved if approximately

J.5 times this correction were added to Eq. (7). The resulting normalized
spectrum given, by

SPLI/ 0ASPL~ 1oog{0.6I3(f/fmo,) ~fta)/+ 0.5] --0 0 3'(f /f .. (9)

is shown in Figire 3 as solid symbols. They are within 0.1 dB of Eq. (7) for
frequency ratios less than 3 and closely approximate the measured spectrum
decay at large frequency ratios. For far-fiel.d distances much different
from 500 ft, use of tabulated atmospheric attenuation properties rather than
"this simple approximation is recommended.

The nondimensional :spectrum given by Eq. (9) should be valid only for an
unrtapered (cons.tant-chord) wing. To determine the effect of wing taper ratio
on ;pectrum :;hape, this; equwtion waý; applied to calculating the acou.stic

, . D. , liealy, G. J ., and Gibson, J. : Methods for the

I]'redli cji n of' Airframe Aeruldynamic Noise. AeroacoustJCls: Acous tic Wake,
ct,,,tn; Aircraft NGoise 'Prudict,(ln; Airoacoustir Instrumentation, Vol. 4.),
i'ru)r,~
;L11 A(e t'(ranttics ard A.,troriauti c;, M.l .1. Pre:;;.
s Cambridge, Mass., lgY(
lip A c-A , . A.].), AIAA Paper '/' ',-,,)1,, Mar.
 0
(.J

.10 WING TAPEH RA1IO

uj.,> ' /4

L-J

-J
CL)

zz<\
uo

40 - I _ --- -l 1
0.125 0.25 0.5 1 2 4 8 16 31.5
RATIO OF 1/3 OCTAVE CENTER FREQUENCY TO MAXIMUM
AMPLITUDE FREQUENCY, f/fyax

,'lIWRE -. CALCtIjAtEI) F'IFCT 01' WIN( TAItEP RATIO ON NORMA li , /) 9oCTAl,.

AIRIRAME NOI.:;E ,PLCI'Rt)M

/
I, '
spectrum from each of a large number of spanwi se segments comprising tapered
vil.nr;;. Spectra were calcuilated for wings having the same mean geometric
chord (ratio of wi.ng area to wing .span) and fli],ght conditions. Resulting
in Figure 4 for taper ratios of 0, i/I,
nondimensional spectra are plotted
and ,L. The spectrum for a taper ratio of 1/6i- was within 1.2 dB of that for a
taper ratio of' I. The spectrum for a taper ratio of 1/2 (not shown) was
within 0.4 dB of that, for a taper ratio of 1. In contrast, the spectrum
calculated for zero taper ratio was, about 3 dB higher at very low frequencies,
L.2 dB lower at, maximum-ampliitude frequency, and more than 8 dB higher at
very high frequenci.cs Maximum amplitude occurred within the same one-third-
octave band for all taper ratios. However, the calculQated maximum amplitude
decreased as taper ratio was decreased. This occurred because the frquency
was far below the tip region's maxim-am-amplituae frequercy. At large
frequencies rel]ative to the maximum-amplitude frequency, some portion of the
tip region was at its local frequency for maximum amplitude and therefore
radiated. more noise than the large-chord inboard retions. Because few civil
aircraft have taper ratios, less than 1/4, the nondimensional, spectrum calcu--
lated from Eq. (9) will be assumed to apply for all taper ratios.

Use of a calculated flat-plate boundary layer thickness corresponds to

the assumption that, the wing upper surface and lower surface boundary layers
are unaffected by wing lift coefficient. To examine the effect of li-ft
coefficient, boundary layer properties were calculated for a typical business-
jet airfoil section, chord length, and approach speed. The airfoil pressure
distribution had a leading edge suction peak on the upper surface at lift
coefficients larger than 0.6 and on the lower surface at lift coefficients
less than 0.2. Increasing the lift coefficient in the range between 0.2 and
0.6 caused a -,mall increase of upper surface boundary layer thickness and a
smiall decrease on the lower surface. Resulting calculated amplitudes and
essentially independent of lift coefficienL. fý further
spectrum shapes were
increase of lift coefficient to 0.8 caused a relatively larger inci ease of
upper surface boundary layer thickness. Calculated OASPL and low-frequency
noise were increased about 2 dB, but the spectrum at higher frequencies varied
less than 1 LB for the range of lift coefficients from 0.2 to 0.8. This
range covers the low-speed flight conditions of practical interest for air-
craft with leading-edge and trailing-edge high-lift devi.ces retracted.
Therefore the effect of liftt coefficient was neglected in further calcula-
,ions.

By t.rlal anid ,error, it, was I'(ui1ld L, at the' ftUliC'tLIed 20petra t-ab
t:Lt i a ted in
. 9t) wtq)e tat Ilti I ay Eq. () if" he peak f'r•qi'eriv w:',; a it
,W (). I timets
the raItat oV' v,'lOL
(Ayr ; Irialary flu .kaess . That is,, ieak Si(ýr
pL
stL er' wa;; kit ak, I
Ut.; t t' eaeei• 1{ Its•:. ti,'t '.r;; •al, t ....
t Io (!Ill
3.1-3 Directivity

Airframe noise directivity in the flyover plan has been measured at

Douglas Aircraft Co. for the DC-iO aircraft (Ref. I5). Special acoustIc
treatment was used with the high bypass ratio turbofan engines to reduce
engine noise below the airframe noise. Acoustic data were corrected for the
variation of far-field radial distance with direction angle; distance and
angle were evaluated at the retarded time. The resulting measured variation
of distance-adjusted OASPL with angle for a clean DC-1O aircraft was compared
in Figures 17 and 18 of Ref. 15 with two analytical models. These figures
are reproduced in Figure 5 herein, A lift dipole without con'rection ampiifi-
cation, as used in the drag element method of Ref. 1 4 , matches the data
between about 600 and 110i angles from the approach horizontal direction.
Adding the source motion effect, as i.s done in the NASA ANOPP version of that
method, improves agreement at smaller arngles but worsens the agreement at
larger angles. The directivity function for trailing edge noise, without
source motion, matches the data within +5 dB for all angles less than 140(.
Convective amplification worsens the agreement. it was concluded in Ref. 15
that a combination of a lift dipole, a drag dipole which i.s correlated with
the lift dipole, and an uncorrelated drag dipole (all with convective amplifi-
cation) gave the best fit to measured directivity. No method was given for
generalizing the relative strengths and phasings of these dipoles.

Adjusted OASPL directivity data for several flights of the clean DC-lu
airframe had also been given in Figure 14 of Ref, 16. These data are compared
in Figure 6 with the directivity predicted for trailing edge noise without
convective amplification. This prediction closely matches the data for these
three flights. The convective amplification effect on trailing edge noise
directivity therefore was omitted from the method presented herein. However,
,thr convective effect, on frequency (Doppler shift.) was included.

Airframe flyover noise for airoraft. in the cloan configuration should be

a sum of traoi ing edge noise from the wingq and horizontal tail° For typical
rat. ios0 of tail area to winI,'; area, the result.tinF tot a 1irfframe noi :1
ssh ould
be 0.'5 to I dB greater tihan (AOVP, from the wing . However, calculated peak
frequency of horizontal ti.ail noise !Ci larger than that for wing nol se. Thus
the portion of' the calculated spectrum which is heavily weighted in predictinrig
armoyance-weiplhted noi:;e was almost. equally !31'Vected by noise from the wing,
and hori zonltal tail. It was: Coun I that, mea:;ured spectra were be!, predicted

J1 ur,, in, A. '}. A Modei iFa Appro, ch to No1)propu1 ive Nol •,. AIAA

Coltr' .1i- . iucr ul), ,i Airnrc f't, V,)i . I V .'. , ,i\' );1,
) , p[ * . -,[

IWd.,
 aJ0: :)

a. <AMPLIDIIO-
0 c

OIiuVvE IVE
>0 0 80 L F2
I1CT60O

ri

POLAR ANGLE AT RETARDED TIME. BEG

LU
a:

Cc,,
00

zCv I-<
cA H)c

(t) <

I-I /\D I0Ni A I TIFI fM l

 LUI

0CONVEC:TI yE AMPL, E CAT ION

a- LUC
T
10dB

U)><

<>0

1 40301 6

I P0,A WA AIRfIA M1

E C- - - 1- ýI -
i j-- - - ' -! ~ l ý,,E ' i ' 11-1
 by assuming that wing OASPL was 2 dB lower than the values that would match
the solid lines on Figure 2 for total airframe OASPL. Including the direc-
tivity effect, OASPL for the wing was therefore taken as

2
OASPLw : 5Olog(V/lOOkt) + 1O log (8wbi/h) + 8(ND) + clog (Cosq 5in GcoS e/2) 2 + 04.3dB (10)

Here, ND is set equal to one for conventional-construction low subsonic speed

aircraft and for aircraft which have extensive trailint edge flap track
shields. It is set equal to zero for aerodynamically clean aircraft such as
high-performance sailplanes and for ,jet aircraft, that. have relatively simple
trailing edfye flaps. Horizontal tail OASPL is obtained by replacing the wing
span and boundary layer thickness by values for the horizontal tail.

Thi a difference between noise from an aerodynamically clean sailplane

and a conventional airplane with retracted slats, flaps, and landing gear also
occurs in Revell's Drag Element Method (Ref. i14). In that, method, OASPL
usually is dom~iiatbed by the calcul ated wing profile drag noise. Mean ý;qunre
acous1, ic pretissure for ;,uch noise i:, predicted by that method tIo vary wit~h wing
profile drag, e),t-'Vicient cLubed. Thi coefficient can be about half a:; largto
for a saiplpiane wingt as for a convtntional low-:;peed airplane winu,. Thus the
a it'frname nroise ca.lcutla ted by the metthtod of Ref. .1 cam1 vary within about ý) dli,
uiependint,, on a a arbitrary e t.imnat e of aerodyanmnic clTain s The mllethod iJv tol
h1ere.i n UW.e anantrbitrary cllice of' _it hlier thc lower or uppetr textrrme of t,h~i;
ran•,,e.

'Ir'a .1 t tigt
. Il soi tm
frma thte vt,rtI it(l tail i: ra.idl t d t.iwil l' the s;iAhq-
I Illlt.; allmt ;, Le'-'ieus
to) e' , .illtlell:.iy .in th.e 'lyov "r plane . Ampr i taldte ýil
IIi , alk)i ::, 1:; (1; llt.11 Itd by re p A,'ici IiI t ell wi fit.' PI',t' rt, it.,:; ill El1 . (I(,)) by thI I l:

(It the vrti :c'I tail and lt allotwi•q, t.lit IreI ttivity t.r'mati l'(I:; (f) by 111W-
Iinb.
(,'V ,'I' Ihp, : '''tl 'l y [Wc
:;, i• il l 1,1W .I: ; It l' • Ti Eq . (,)') t'(,,r • t.i~ , ý11:t~
,l
t
:lI~lr~xllx :,~t. 11 , ,lltaln ;pheri 't0 e llnI tt Ofl i.,1 tb. 1 boak 'rtelntjcu.y (,t' wixihg trail -

i llt't
'l" :, p '" (,rl " till hle T- ';tI(,(,((• I {' ) '( ,i I[•I w
1)(,Ik i<ll} l ,t{ w('1: rtt llti ý b (t,.

I' '!::~ tl ' : (l'il, • :t t', ''!l iI[ tw t hi[ i~i ,w n k w h~t <ich tt , ! iwlt ; c. h lt" .
'ti it

:•' ,. :: • I ! z1 I<. r yv v. ! I i t i 1''i l : r q,. ' ', w h• wi

W, r e..
t., T,I! t, i 11,, 1 :,,- :1 1, It i

I*," W• •It: " i!1••

' .> ý'':
l• 1Z
1 it'' t il ''1 '.•,
i 'r, !11 ! h - i v•ltýY 1V
: 1A
•,,w it. '

l 11 11
 Figure 9 of Ref. 17 foýr the Lockheed JetStar and by RAE in Figure C..4 of
Ref. 12 for the Vickeýrs VC 10. From both Sets of 0ASPL data, it was concluded
that sideline noise of -lclean airframe varies only inversely, with radius
squared, without an additional dependence on sideline angle. Unpublished
NASA sideline noiso measurements for the Convair 990 in the clean configura-
tion (run 3, 314+ knots) were obtained. for detailed. examrLnation. These data
include spectra measured at intervals ol' 0.1 to 0.2 sec on a line perpendicu-
lar tuo the flight path, at s;ideline distances to 1~476 ft, for a relatively
low-altitude (230 ft) flight. For the following analysis, the time at which
the airplane parssed the instrumentation line was back-figured from direction
V angles, meas;ured at the nominal flyover position. Acoustic travel times to
each microphone were added to this time, and measuredý spectra were examined
"for the ret~ardedl time at. wh~ich the Doppler effect on, frequency was closes~t
to zero.

Res-ultingr vafiLrations- of overall sound pres-sure level, and one-third

octave "e,(un~i pres.-ure level.; at '00) and, 1600 Hiz center frequencies, wi~th
s,idelinrio II ance are plotCAte A In c Als-o sonare variations that
would bie expected. f1or acoustiJc radiation from a h'orizontal surfaceý and for a
Yarn ~i rio inversýely w i thr lI taiebt inidependent, of sidel ine angle
Not. et t hat ftlr the lower Cre-q ieri.y , where airtframie noise shoul~d greatly
o xeee i ia lAo i so , Ill
he ap~~1 A ,Irta'e prediiclod, for, :,ur face -radialted aioiles
ti~t occur. );ATl,, antl thie hierflcp oise aittribut~ed here to eng 'i Ile
iio ,em Lied thew urni tI er 1preiti c t~ed AIcaiy "'at~e. Airframe aol s,-e gliven by t~he
de1, ojpt.' liiiiet- tin 5! (uolii et, therefIore Io;the t~heoretically-
Iprli.c't,t.'i Vuvirit. loll w-it :LIisdellie ,in11.Ie rit her thiin the slower decreasýe
.A lid i cat eA bk[y theI it'(jiti a A1 1he i:ý; .1: ro l A t

I eit l t,
C , Iljl 'q M, itii 11

V~.'ii ll i1tt ii p, i i soec i.le iriethw et lve kjutte htreiri i:: liii

it.l t u Inplit ll I 1151 . IhO ~it'' t'un iiii,.i~

p tilt'
lt-tik~rýý ()I' t'. a lot i r rii
 30- z*"AIRPL ,.
/ 7
--
ED FOR SURFACE AIRPLANE
SEXPECT
30
RADIA1ION, 20 LOG
z > (/h COS,,))

20-
'

20 LOG h /
0
0
D IPJDEPENDENT OF

0F R T 0 SIDELINE ANGLE
0 1

uiu C0 200VEHAL
0 f

0 2 4 6
RATýO OF SlIM- LIM: [DISTANCEL TO ALTITUDE, Y/h

( ,' Nr ,J'.1 ;.. +•:

!i, 'Il' ,T1,:

01, ~ ~~~~ P, AN2Iý E IIX] W 1i

 IVERTICAL TAIL

0 SLATS - CLEAN

NOSEIN

LANDING GEAH FLAPS HORIZONTAL

MAIN IAIL
LA [IN( GEArt

FlI;i}UR; -, 1N)IVIiDIAl, No:1:,-RAl /, "l ALF

SN(; PlRAMiF (•hNENT:;

,v

I1
 The inaividual noise-radiating airframe components represented by the
method developed herein are sketched in Figure 8. Noise from the clean wing,
horizontal tail, and vertical tail were previously discussed. Noise contri-
butions from the nose landing gear and main landing gear are calculated
separately because generally each has a different size and therefore a dif-
ferent peak frequency. Noise from the trai.ling edge flaps is calculated
Independent of whether the landing gear are extended or retracted. As pointed
out in Ref. 4, this would seem to be a questionable assumption. Extending
the landing gear generates turbulence from the wheels, struts, and open
cavities. This turbulence is convected past, the trailing edge flaps and would
be expected to generate incidence fluctuation noise. This type of inter-
action noise could be calculated if the intensity, scale length, and lateral
extent of the landing gear turbulent wake was known. Noise from leading edge
slats was found tc be most easily represented as a sum of noise from the slat
itself plus increased trailing edge noise from the wing<. This was the one
noise component for which an interaction had to be included in the calcula-
tions described herein,.

The assumed absence of interactions between various noise ccmoonents can

be checked by comparing measured flyover noise spectra, at constant airspeed,
for an airframe with components extended i'ndividually and ir. combination.
Flights of this type had been conducted for the Vickers VC ].0 jet transport
at 60o fl altitude and approximately 160 knots airspeed (Ref. 12). Tabulated
one-third octave spectra for these flyovers were obtained from the RAE. The
spectrum measured for the clean airframe with idling engines was logarithmic-
ally subtracted from those for the airframe with only the leading edge slats
extended, only the landing gear extended and wheelwell doors closed, and
only the trailing edge flaps at 450 deflection,. Resulting noise increments
from each of these components are plotted in Figuz'e 9. Also plotted as a
solid line is the logarithmic sum of measured noise for the clean airframe
and thoý'e three noise increments. This experimental prediction of total air-
frame noise in the approach configuration is seen to be in excellent agree--
merit with tne measAred spectrum for the approach configuration, up to 4000 Hz
center frequency. The comparison shown in Fi.gure 9 therefore validates the
usie of o, n•i••ie component method in which no interaction occurs between noise
from lamiuiig gear, trailing edge high-lift devices, and leadinr, edge high -
lift devices.

3.3 Landing,, Gear Noise

MI from extended landing f;ear has been irive st.igted experiimenta ily
a t DltiVIi (t ,I f. 1H). From a compauriý,on ofi re lative diimrns;io :)f for ro!' e p#ear
arid mainx ei r of :several jet. trarv;port:;, [,hey IY rd that. tI-e ra iti(c of s;1rut

P ,ie I-1.,', H. IfH and D.obr11..i , W 2.oind Radial,i aon Pi rcrft

[o
Wheel-Well 'ianlinbrig (;ear (2orfigurat, is. AIAA Paper ' vI)
July /'.
 90 -_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _

SUM Of NOISE COMPONENTS

m1

> 80 A
-jD

0 0 0
70 SYMBOL CON!ýI(URAfI-IN 0 o
0 CL EAN AIRFRAME WITH ENGIN~ES 0 0
0 INCREMEN r DOUE TO) SI ATS

w~c Q NCREMEN1J DUE TO LANDaINC, GEAR~

< 60 Ii IN(;HHzEN1 DUE 10 45" F LAPS
I,- C) (IP TY C01F I`1600ATION

40 100 20)0 l(K) 1000 2000 4000 10000o

1/3 0C FAVI- BAND CEN TER FREOUENCY, H,,

FIGURE 9 -COM11ARM2ON OF :IMM OF N*:Aý;BEI) N01r F COMPONENT"; WITH ?VEASTh'D TOTA!,

N015E FO)R VICKER?'-3 VC 10 AIRHCRAFTl 11 APPROACH C'1)NP1COTRATION
diameter, exposed strut length, wheel lateral spacing, and wheel width to
wheel diameter were fairly constant. Test models therefore consisted of a
two-wheel. configuration typical of nose gear for large airplanes and main
gear for moderate-size airplanes, and a four-wheel configuration typical of
the main gear for large airplanes. By testing several components separately,
it was found that the strong tones and higher-frequency broadband noise
radiated by an isolated landing-gear cavity were greatly suppressed by the
presence of bluff-body landing gear. Also, most of the noise was found to be
associated with the struts, drag braces, and actuators rather than the wheels
themselves. Use of these measured spectra scaled as expected for dipole
JetStar, Hawker Siddley HS 125, and Vickers VC 10 airplanes and to over-

predict that for the British Aircraft Corp. BAC 1-I.

Free-field spectra were given only for different azimuthal angles in a

plane perpendicular to the flight direction. Spectra for the two-wheel land-
ing gear models in the flyover and sideline positions, and 45 between them,
were given in Figure 7 of Ref. 18. These spectra are pl.oted in the upper
left portion of Figure 10. They appear to be the sum of one noise process
that is independent of azimuth angle and another that varies with sine
..quared of the angle from the flyover plane. Fluctuations of side force on
the cylindrical. stiut are likely to be the cause of the second noise process.
These curves were approximated by the arbitrary empirical equation

SPL :60 bg(V/t94 kt) + 20 Iog(D/r)-4-I0log(I OG+ ,OG2) (11)

where Gi: 13 + tog 4.5 (fD/v)2 [12.5 +(fD/v)2- 2

25 (12)

G2:13+ 0q 2(fD/V)2 [30 + (fD/V)8] T (0.34 H/D) s~n2 (13)

and the normalizing velocity 194 kt is equal. to 100 m/sec. Spectra calculated
by this equation are shown in the remainder of Figure 10 to match the data
within about 1 riB. The ratio of exposed strut length H to wheel diameter D
is included in the -ideline noi:se term to allow prediction of noise from long-
strut nose landing gear and short-strut tail wheel:s. By nuyerica] integra-
tion of Eq. (1)

OASPI_ : 60toq(V/•194kt) + 20 log(D/r) + ,0loq(g0'? 2 + 0 GI) (ia)

5 •
 OVERHEAD

450 SIDELINE

90" SSIDELINE
0aL 133 ---- CALCULATED
LU "" \.5900
..SIDELINE
.... -CALCULATED
. 1020 12

o
.CL•'v OVERHEAD

25 1 20 50
""o ,., - - -. I I loo
5 10 20 50 05 1 2 5 10
( 0.5
-"D 1 2
STROUHAL NUMBER, fD/U STROUHAL NUMBER, tD'LI

0<cc

130 90 0S ID E- LIN[
130 4 SID E L IN E
5 0° M EASUR ED
- - M EASU R4ED
0" ' "• . - A C
-J C L. A T ED
UL,
oJ' CD . . . C A C L-E

o~~f u11 0,11

2ii .1 I, A, A hL (C OMIjA R :. W
N ( F 1,j I 1,
I ,I A ( UI ' I' I
~-U=i ~ r 'B( I t R E 0 P- IRE. E t A ' I IO
-AT
" ~•eI ~ -

I MbIIC)N , i()t
E MPTR1CA h A PP H U)X N uoI~ i Fy1 0L
-widtl itl iA l 1N ; HA N
 OVERHEAD

450 SIDELINE

Z . 900 SIDELINE
CC 140 140
m OVERHEAD
0I0 S - MEASURED
(n 1_ CALCULATED

F-
>
O-
l' "u
120 -) 120 -
CL
N--
- -
-1 0 N 110450 SIDELHINE <
13: 110- •

-- 2 8 ______________oo________________________ ,oop I I I
Z 0.2 0.5 1 2 5 10 20 0.2 0.5 1 2 5 10 20
V)
STROUHAL NUMBER, fD/U STROUHAL NUMBER, fD/LI

140 140
m 450 SIDELINE 900 SIDF LINE
__"_-- MEASURED " MEASURED
---- CALCULATED - - CALCULATED
- 130 D 130
IF- I--
wJ uJ

cin 120 V)n12

0
NN

4 110 <~ 110

0
zo 100 1A 1 _. _-l_ z lool I I ! t __ _
0,2 0.5 1 2 5 10 20 02 05 1 2 5 10 20

STROUHAL NUMBER, ID/U STROUHAL NUMBER, fD U

FiGURE 11 - PREJENTATION OF ,'EE I,'II,] ACOU:''iI' ',V :(HTtA, ANI CONT•AI•I.;0G WI I'ii
EMPIRICAL. APPROXIMATION, F.I(F NOIF:;F OF MODEL, L'(IfH-WIETL IAMrI)IN(; GEAR

.k : )h 0') 1 s.l1t
 where G3 :12.79 +1og(O.34 H/D) sin 2 o (15)

These equations can be generalized for landing gear having either one or two
wheels by arbitrarily taking the nonsideline intensity for a one-wheel land-
ing gear as half of that for a two-wheel landing gear. That is, 0.3 is
subtracted from GI, and from the exponent 12.52 in Eq. (12), if a landing
gear has one rather than two wheels.

Spectra for the four-wheel main landing gear models, taken from Figure
10 of Ref. 18, are plotted in the upper left portion of Figure ii. These
spectra were arbitrarily fitted by the sum of one term which was independent
of azimuth angle and another which corresponds to a fluctuating side force.
The spectrum for this sideline-directed noise was flatter and weaker than
that for sideline noise from two-wheel landing gear. Spectra calculated from
the empirical equations

4
SPL:60log (V/194 k)+ 2o log (D/r) +0loq(IOG +- oG
0 5 ) (16)

G4 :12 +l0g(fD/V)2J[0 4 + (f0/v)j]-6 (17)

'1+log
7(f D/V)3[I or(f/V2 -3}1

are compared wi th the mean ured :;pec t, ra i n the remnaindet-r ()' I-'i are
t_ II 0vt-ra-.i.
:;ound preisktre level i_- given by

9
OA•ýOL - 60 log (V/194 t) 20 log(D/r) 0loq(1012,' + i(0G) 0.[, )

Calculate•I 8;pectru forI li.::0 radi t ed frolr i lýOlr-wht'o.] !:1I I< t'e ha l
paret ili Figý,ure2: 1:'
1) rvi .t1 with th id ;ý irirrellrit i-:I ai:;I'd Iv f'xte,,yi- i .:I'iciilm'
ge,ar tdl the Vici-er.; VC JO) %rid with ado!:'e ()I' Ihfe B} eiinc AY withl l j ei
r ×iit.,
"SPectra calculat~e~d from Er;. (.t,) .zi (I'(') are pit't. i'is ,ii we:, e

Mtnpliti•e ie ; overpredi ctted at, low 1'reqienciC : WlI( 8i rk Iy l0I

tll ,rIe(''i eL Ht

, __. _- _. __ _. _- ==_ J
 PH[WCt ), lS ANALYSIS
HI
)

w
D 2~~
0A

>0 50 20a0 00 200 ~ 00

Z/OTV RADCNuOFFUIJYH

cIa JH TM'I I i Ih N II I)jI}

iuX'lNh A I)IN KN N V wih[ I

>' 4
 "13_[ 900 PHDIC TED, THiS ANALYSIS
PF
•00

, 0" Q 0 0
00
o 0
7 "00 r S 1 "1(O
2 030

PREDIC'TED,
Qr ME rHOD OF FIE 1 18
z 70'-
0 _ _

0 RUN , LANDING
3Z GEAR ',)ONVN,
A I LAPS RETRRACIEWL), COEFFEC] EDA

S60 -- TO FREE FIELD

50 100 200 500 1000 2000 WN00 10 00

1/3()(:IAV E BAN ()C IN I R R (Rt)(tNcY. 1

l,'(;I•
1 : .?(MIAR[, OF *1
HIlA,2lI!Rl,':I) ANlI) • 'tI '' NolI:;E V"{)Iý IBu()iV:N: "
A IRCRAPT WI'[TH EXTE•NIDE!D IANN•NG(; •IAH
high frequencies. It was observed that the low-~frequency portion of the data
was closely predicted by use of Eqs. (11) and (12) for two-wheel landing gear,
byt teq (17ethrere by
e wash-replancy eda uhsalr i ucinG ie
but tEq measuhreord high- reqpencyd decywsmuhsalr TefncinGgvn

G2+ lg06(f D/V)2[I 0. ( V)2]-1, (20)

which has essentially no effect on Eq. (19) for OASPL. The resulting calcu-
lated spectra, shown in Figures 12 and 13 as dash lines, generally match the
oscillatory spectra. The high--frequency decay term calculated from Eq. (8),
which approximates atmospheric attenuation, should be added to -these calcula-
ted spectra to obtain a closer match to data such as that of Figure 13 at
high frequency.

3.4 Trailing Edge Flap Noise

Noise radiation from trailing edge flaps had been represented by the NAS1A
ANOFP noise component analysis (Ref. 4, p. 147) as the separately calculated
trailini- edg-,e noise from each flap segment. The drag, element method (Ref. 11)i
represex.it; this noise as a sum of' two components. 3Dne is proportional to flap
proflle dratg LIcIefficient squared and is associated with wake turbulence. The
other component, associated with the wing and. flap trailing vortex system, is
proportional to induced drag coefficient cubed. The prediction method
developed herein regards the fl-ap panels as being immersed in the tu~rbullent
wake of the wing and upstream flap segmente. Flap noise thus is assumed to
De lift. fluctuation noise caused by incident. turbulence. Such iioise could
b~e predicted explicitly if the turbulence spectrum, scale length, and
i ntlns~i L weýre known. Because this information is nct known, an emp~iri cal
approach wit!; ti~ed.

Nol:;e ;;pec tx-u (enerated by incident turbulence at low arid moderat c

ýub ;;onio feti h t, : peels, shouldI 'oal e ace it' the amrpl itude and frequcrney sa
'iare propertl y Yidj i:; I ed . For fconvenience, flap noi se was ftrbi ti-ar i .yas;ud
t~o 'oe i nii-pentident. of' Uth number of chordwi so flap segments. Amplitude should
vary 1i rect. Iy with flap area , inversely wi th far-field di s tuane ;: ;quared , and
ti Iruectly w itth "11juhi. :;peedI to t~he a,Ixt~h power. Frequency ",1011,' 1)- saledA
"e as
It troulal niunler re 1~ itiye to fin p total ,,hordi.
- Hfe.e , fi up chord wats taken
FA: t.he rniiu )I t ta I Itra iiing edgýe flfip area (usual ly known for eaeh a ir -
p)1 i e ) 1(.) ! d ;alili~~Ia: l :t'rom FA t~hree -view sAketch of the aiirpl ano . No i cc
.:ýpoet rn ~ 1'' Vutp It-Ibet. i or; wor~e ob~tained from the a1tia Of' He'. I?1 for the
1. -erV 1at6.U'),an )4 f~lap deflecti on, and( the dath of Bef'. 11
1(t r thf iiacýinji (%i(vit. t'jtIap deflection, the Conva;.- 1)1)() at, -2,( Plap
I
te Ileu! ioen, nr i t. lit!. Lolkht-eIl Jet:ýtnar ait 00O flap dIeft tM-t ion. Thne VC 10 "And
 7147 have triple-slotted flaps; the other two airplanes have double-slotted
flaps. For the VC 10 and CV 990, the measured spectra for the clean air-
frame with engines at flight idle were subtracted from spectra measured with
flaps deflected. This procedure neglects the fact that cle~n-airframe noise
from the undeflected portion of the wing will be somewhat smaller than noise
radiated by the entire clean wing. The portion of the VC 10 spectra near
1600 Hz center frequency was not included because of 3arge contributions from
engine noise at fan blade passing frequency. The CV 990 spectra were not
analyzed beyond 14000 Hz because the di~fference between spectra with flaps
extended and retracted then decreased below 1 dB. Spectra were not available
for the JetStar in these two conf-igurations and nearly the same flight s~peed.
A spectrum for the clean airframe was scaled to the flaps-extended flight.
speed, assuming an amplitude dependence or' velocity to the fifth power an~d
either S;trouhal scaling of frequency directly with velocity or (Ref. 11) no
variation of frequency with velocity. The resulting levels never were clýoser
than 6 dB below the flaps-deflected spectrum and generally were at least 10 dB
below it. Therefore the JetStar spectrum with flaps deflected was completely
attributed to trailing edge flap noise.

Spectra also were available from Ref. 17 for the Aero Commander Shirike
with landing gear extendled anid the trailing edge flaps either deflected or
retracted, at constant airspeed an~d altitude. Below 2'-0 Hz centor freqluency,
deflecting- the fliaps decreasedI the apparent landing-gear cavity tone andi
therefore reduc(,ed the noi se. At larger frequencie!- the measuired1 nofbise
increase casdby flap del'icction wa8' only riboot 2 MB The resualting, noi :.(e
due to flap defte,ýtion 1-s inrherenil1y much le.,ss iccuirate than the other flap
noi se spec tra.

Arnpll -i f.ie: of t~he rc ii tirig uiIjov,,ted -poctro wcýre found t. o Inrcreaes

1,ithIncrea:sing flap deflecet-cm . 'Phis increa.fe seemed t~o be, much le;rapitd
than tha t. fjvon by the dlragr element methodl ( ~ef . 14i). -In thatO. methodt t.rail -
ing- edgetr Can 101 e( at. eonl;; aill, "airs.peed anrd lif t, c oeftFi cient. 1 ;pr ls 1ý t()
vary with I'Dp profile~t ;;mired rg
P1u8ott1i (ut. sl Cla
IJo Iett W~I I-
approx ilwt t (' y w it'l it,IAp de t'lct. o i k ii IIi( tll( :;t('~ ()I-o hi, r~i po(wer1. There V )55,
Jhr~it- (,I emlit, Itt, met t !I( )( prf,'ii ct t;I I: 1. 1,f otp Ilk) i r~It,(nI I: y Vii rc e : i Itt, I F I Zi
iet~lcti to
10hhrlte it iili jIOtWer 11*L11i 25 tii Liid nci
;;eemedtovary rogj y w~itli :Jn :a(tmaured ol' the, fl1ap dieflectli n angple.

The
c;;11 i. t odAe -Ridj U;; ",t (nd
IT~I iij I'cqec ucju ;jec trlt for
maX mnun
Ilove'I()i-(2 by trai iint, edre 11 ap:; aire
CMu:;e.d lotdiiiFi-u'
Th
ItIw pct~ra 1"taecurlly avreed at I low ',t~rcuhii.I niurmbcrs . Ii)wevtr , tl~ i 1 h
troquc o rtc.;tk [01o li t 0 t ot. ktirrir dtIr uhal liuIrilwer;; br'(p1. Itit
flap; thr t 5 ilflplvr CI ip:ý Thl:; ;etrI;haips, ltivAi' rig a ,1dIotii lt'(iry It,

] I
ýjeiith s5)1y i ;
ini111 ok
iii.t tl 1 n aib r11011Il
, !, Nhr)eii~.th ta
i I ii~t

'5() u iI*,i ý 'l ~ r ~ l nt tt , . A ,) : i iv

 110 IMPIRICAL. APPROXIMATricN

wU -IL L

W 0

0 - zt

.- DI4 oa

>H0 -j0 0

0 0j 0 2 0C05 1003!

a.~HA Si IiWSi SIN(,[ I H R I) I INI[l

 noise radiation at high frequencies. The observed spectra would be expected
if double-slotted flaps have two flap segments of roughly equal chord, and
triple-slotted flaps have a small-chord vane followed by two larger flap
segments.

Directivity was arbitrarily assumed to be that 'or a lift dipole normal

to the deflected flap. For steady level flight, the combination of inverse-
square dependence on far--field distance and the deflected lift-dipole
directivity causes the flyover noise to vary with sin2 0 sin2 (0+'), where
8 It can be shown that this quantity is
F is the flap deflection angle.
largest for a retarded-time angle 0 equal to 900-8F/2. That is, the noise
which will be measured as maximum flyover noise leaves the trailing-edge flaps
before they pass over the observer. The resulting one-third octave spectrum
due to deflection of trailing edge flaps, decreased 3 dB for comparison with
data from post-mounted rather than ground-board microphones, were arbitrarily
approximated by straight lines in Figure 1-4. The result becomes

OASPL:1,20 + 10 log (SF Si'

2 8
F/h 2 og'n(&
) + 60 log(v/ IOOkt) 4- 2Cl1g~s, 9 col 2 + F)] (2

fIor n'A on
'•1i• Il .0e-:l]olted flops, arn 1[.0 All more l'()r ,r ip e-s,!olte(l 'i'ap:;.

o: t• :G( •'+ IO(:) SF SUl 2 ýF/h 2t+ (S-log (V/iOOkt) + 20 log sinOcos2 sin(U+ •) (;o-?

F- F

99
-!trG. • Olog (t;F/V), f( F!/V 2 "(

G '103 82 - .0o 9 (f(F / V), 2< ft F /v ' 20

G;:13504 - 300oC(f(F/V), 20 <f /V, ( *,

' :, , I i ut, , I I 1 i ,, :

! .
 G7 - 99,0 + IO Iog(fcF / V), fcF/V < 2 (26)

G7: 102 61 --- 21 og(fcF /V), 2< fc F/V<75 (27)

G7= 155.11- 30 Iog(fcF/V), 755 fCF/V (28)

for triple-slotted flaps.

The assumed simple variation of trailing edge flap noise awplitude on

sine squared of the deflection angle csn be illustrated by use of the measured
spectra for the Vickers VC 10 aircraft (Ref. 12). Spectra are plotted in
Figure 15a for the clean airframe with flaps letracted and for 200, 350, and
45° flap deflection angles. These spectra inulude 6 dB increase above free
field. Noise increments caused by flap deflection were determined by
logarithmic subtraction of the flaps-retracted spectrum. These increments
are plotted in Figure 15b for the three flap angles, Next, thc, quantity
20 log sin&F was subtracted from each spectrusn to account for thc effect of
flap deflection on noise amplitud.-. The small quantity 20 lok, cos (SF/2)
aLho wa:s subtracted to account for L,,e forward rotation of tlap noise direc-
tivity. The resulting three deflection-adjusted spectra are plotted in
Figure 15c. Data symbols for all trnee flap angles are intermixed and
generally are coalesced by this aijLstment. Also shown is the mlh,Lricai
curve from Figure 14 for triple-slotted flaps. The spectra are generally in
good agreement with this curve. Worst agreement oecurred at frequencies near
2500 Hz, the fan blade pas:;ing frequency. This comparison validates the
assUMption that trailing edge flap noise varies with sine squsred of the flap
deflection angle.

3.5, leil.tn E(Ige 21]at and 'l.up 0NN se

Lead intg edge slats would be expected to cause: noi:,e by three process-es.
Trailinh euge noise would be caused by the sFat; boundary layer a-, *t flows-
past the s;lat '.railing edt,;,'. Bluff-body noise would be generated by ;iigh-
speed airtflow pasi exposed actuators and tracks protruding, through the slat,
gap. in addition, the winr, upper-st rface boundary jayer thickness- and turbo -
l en-e ints-,ity would be changed by the slat wake.. Only two s'ets of datu were
vaflabe eX
y •,, rI 01 :IIII1
4t,, n)',e Vlyover nim se hod been measIured with the
ickers VC 1P) aircraft. (R(,H'. 1:1) with rid w I.•iout, leridiJnt edgie slat; dep •oyed.
21 a s; i /crease OA2P!, by atlbout, 11 Fit. constan. a:[t,
F. dE rspced, aIRi engine
tli-. NOi :,- ,-tili i'f r a wilni t, nne] mode' of th o, o' '(i were reported
 90 -- __

"A A A AA
AA
80 0o 0

o C) 0 0 0 •cz A
'00 0 0
70 0 SYMBOL CONFIGURATION Z
<) FLAPS RE-TRACTED
0 "VO F LAPS
0 35" F LAPS

- 45' FLAPS
. .
60
aa) EFFECT OF TRAILING EDGE FLA? DEFLECTION ON FLYOVER SPLECTRt.M
j 90
,L.U
>
UJ
-J
) A A
V)
A
80
8 Ll
u3 0A !:
0 6] A
•A , AA A A
"U) 0 0 0
1 A [] A A CN
0_
or C C]
z 00 5 C) A

0 0 El
V) 70 -- C)
L SYMBOL F LAP ANGLE, DEG o 0]
V . 0 20 A
Do ...z 7 3520
> ., [ A 45 00 0l
< 60 L 1 ! I I 0
0 (h) INCREMENT OF NOISE CAUSED , LAP DEFLECTION

- 30

In 0 [MPIH '(A1 ý:L.R\E0

80 CI

70 L\
,V'N.ROt FIAPANGLL-DFG A(
I -! 20 A.

I
60F-.]

40 100 200 400 10)0 ?000 4000 10000

1!/3 0CAVE: AND CENTER F R[ OUENC'Y. fH,,

(,AMPLI I UDES ADJUfYfD f OR ASSUMELD DEPE-NDEINCEL ON F LAP Wt E-i - Cl;T tN

l JpI J•f, -:VAIJATION 0F tM I)IMTC'E I)E1KN E)ENCE I, 'ý /Ai l iNf; : p IlF ' 1.1; q

NE; 2QIARED
'I (4F DEFLECTION AN('I,E FOR VC.KER:" \VC Ij A
)I
12 1]

4,
 in Ref. 19. Airframe noise for th_ clean configuration could not be
detected above the tunnel background noise. Extending the leading edge flaps
raised the airframe noise to detectable levels. Wh,:n a gap existed between]
the leading edge flaps and the wing, the configuration resembled a leading
edge slat. This slat configuration w.as about 3 dB lotider at all frequencies
than the model with gap sealed. A general comparison with noise data for the
full-scale aircraft. was shown, but absolute levels were not given. Comparing
1,be spectra shown in Figures 12 and 13 of Ref. 19, it is apparent that noise
from leading edge high-,lift devices does not extend to fyequencies as high as
those for trailing edge flap noise.

Because of the absence of det'ailed data and low levels of leading edge
slat and flap noise relative to trailing edge flap noise, a simple approxi-
mation was used. Leading edge flaps deflected to an angle appropriate for
the aircraft lift coefficient have been assumed to raise the wing noise to
that from Eq. (10) with ND equal to one, and to generate no additional noise.
Of course, use of highly deflected leading edge flaps at low lift coefficients
in an early part of desoent can increase the noise. At that flight condition
the airspeed is high, lift coefficient is low, and separated flow can occur
on the lower surface of a leading edge flap.

The one-third octave spectrum of the measured noise increase causer' by

VC 10 slat deployment (Ref. 12) is plotted in Figure 16. Also shown are the
measured engine noise spectrum, the increment between that and the clean-air-
craft flyover noise, and a prediction of that noise by the method developed
herein. Measured levels shown are 3 dB lower than the original data taken
with flush microphones. For most of the frequency range, engine noise and
clean-aircraft noise increment were less than 3 dB apart. Clean-airframe
Snoise was predicted to have peak amplitude within the one-third octave band
centered at 125 Hz. It is possible that the higher-frequency noise increment
attributed to the clean aircraft was actually produced by the engine. If so,
then the peak frequency for slat noise was higher than that for noise of' the
clean airframe. The calculated amplitudes shown in this figure are for con-
ventional low-s-peed Pircraft with an OASPL 8 dB above that given by Eq. (6).

SSlat noise was,, app-oximated as the sum of two spectra, each having the
normalized ,;pectrum shape appropriate for trailing edge noise. Leading edge
slats typically have a chord about 15% of the wing chord. Peak frequency of
slat trailing edge noise should -,hen be 0.15-0.8 or about 4.56 times that for
the wing nois-e. Measured sllat noise for the VC 10 was approximately matched
between 315 axis. i600 Hz if' amplitude of this peak of slat noise was taken to

<•ýeai:L , ,]. i . , Fralel]o, I). J., Bohn, A, ,J., and Burggraf, W. D.-
Model .uid l'Ui--:';cne lIarge Transport Airframe Noi. ;e. ? iAA Paper 7(-.5.,
,July 1'

I!<•
• o i•7
V

CALCULATED

0 0
z co
0v- -j 70
00
LLI
> Al

<(
c-) &0- SYMBOL CONFIGURATION

O r A ENGINES ^LONE
c. IROM CLEAN AIRFRAME '

INCREMENT FROM SLATS

50 00 00 100
200 400 1000 2000 4000 10000
40 100
1/3 OCTAVE BAND CENTER FREQUENCY, H/

V i GU
U.., 1 - MEASI AND CAL, CI ILAi EJ) NO I 6X
•REL) I NCRiMEI jrE' TO V, KER:; -

CLEAN AIRFRAME AND IZLEADING EI)(GE .ILAT

be 3 dB above that for the clean wing. The low-frequency portion of the slat
noise spectrum was then matched by assuming that the slat also produced an
increment of' wing trailing edge noise 3 dB above the clean-wing noise. The
resulting presentation of the slat noise spectrum, plotted in Figure 16,
gives a general approximation to those data on which it was based. Rather
than represent the high-frequency portion of this spectrum as having an
intensity proportional to slat area, all slats are assumed to have roughly the
same ratios of slat area to wing area and slat chord to wing chord. Thus the
only geometric parameters used in calculating slat noise were those for the
wing.

3.6 Graphical Method

A rraphical method was developed for predicting airframe noise. Predicted

levels are 3 dB above free field, as with measurements conducted for noise
certification purposes as described in FAR 36. This method neglects all
source motion effects (convective amplification and Doppler shifts) on noise
measured by a fixed observer. Figures are presented for determining OASPII and
normalized spectrum for various noise components of airframes flying at 500 ft
altitude in a sea-level standard atmosphere. The change of noise amplitude
caused by flight at other altitades Is given by an inverse dependence on
altitude squared. The resulting correction in decibels is plotted in Figure
17 for the range from 100 ft to 2000 ft altitude.

3.6.1 Clean Airframe

Maximum overall sound pressure level for clean configurations was

approximated by neglecting the effect of wing aspect ratio and assuming an
arbitrary ratio of horizontal tail area to wing area. The maximum flyover
OASPL for a very clean airframe (sailplane, supersonic transport, subsonic
jet transport without external flap track fairings) can be determined from
Figure 18 ab a function of wing area and flight velocity. For conventional
subsonie propeller-driven airframes and for jet tran.-ports with numerous
large trailing edge flap truck fairings, add 8 dB to these levels.

Spectrum of clean-airframe noise is a function of tht- flat-plate turbu-

.ent boundary layer thickness at the wing trailing edge. This qIuantity can be
obtained from Figure 19 as; a function of flight velocity and wing mean
geometric chord. The wing mnean geometric chord is the ratio of gross wing
area to total win., span (equivalent to the square roo6 .f the ratio of grosr
witg area to wing as:pect ratio). For each one-third octave center frequency,
the nondimens onal 2trouhal number must be calculated. Thi! is the product
of' center f'r*Qtqu,n, y (Hz ) and bouidoary ]ay r, thickne,;s (ft) obtained from
Fiu•xre l9, div iei by fliight velolity in I't/:zec (1.()( times flight velocity,
 knots). The amount by which the one-third octave sound pressure level of that
center frequency is less than OASPL is given by the normalized spectrum
labeled Clean Wing in Figure 20.

3.6.2 Leading Edge Devices

Leading edge flaps are assumed to raise 0ASPL to that for a conventional
low-subsonic-speed clean wing, 8 dB above that given by Figure 18. Normalized
spectrum is given by the Clean Wing spectrum of Figure 20.

Leading edge slats are assumed to raise the wing OASFL to that for a
conventional low-subsonic-speed clean wing. Also, they are assumed to gener-
ate a noise increment 6 dB larger than that for the low-subsonic clean wing.
The normalized spectrum of this noise increment relative to OASPI, of the low-
subsonic clean wing is given by the curve in Figure 20 labeled Leading Edge
Slat.

3.6.3 Trailing Edge Flaps

Maximum flyover OASPL for the noise increment caused by trailing edge
flaps is obtained from two figures. OASPL amplitude due to flight velocity
and trailing edge flap area is obtained from Figure 21. This quantity must be
decreased by the adjustment for flap deflection angle given in Figure 22.
Deflection angle affects both the flap noise amplitude and directivity, so
maximum flyover noise is predicted to occur near 70° rather than at 900 flap
deflection. These values of flap OASPL are independent of the type of flap.
However, the normalized spectrum shape for trailing edge flap noise (Figure
23) is different for triple-slotted flaps. Note that Strouhal number for
trailingedge flaps is referenced to the total flap chord (ratio of flap plan-
form aroa to flap span as scaled from a top or bottom view of the wing plan-
form).

.').6.-1 Landint Gear

NMaximum OASP], foi flyover of a pair of two-wheel or four-wheel main

landing gear ausemblies can be obtained from Figure 214 in terms of the tire
diameter aný! filight velocity. These values should be decrea:;ed 3 dI3 for
aircraft which have a one-wheel main landing gear. They should then be
increa.s;ed in proportion to the total number of main landing gear on an air-
craft (3 dB for four main gear as on same wide-body jet transports). Noi-,;e
from the nose landing gear can be neglected relative to that. from the main
gear. Normalized one-third octave spectrum, which i.; different for four-wheel
gear than for one- or two-wheel gear, -an, be obtained from Figotre "

••ý
 3.6.5 Annoyance-Weighted Noise Levels

Variations of calculated A-weighted noise level (dB(A)) and perceived

noise level (PNL) with overall sound pressure level, as determined by the
noise component method for a range of test cases, are plotted in Figure 26.
As a rule of thumb, it seems reasonable to approximate airframe noise dB(A)
as 4.6 dB less than OASPL with a standard deviation of 1.3 dB and PNL as 7.1
dB more than OASPL with a standard deviation of 2.1 dB. A more accurate pre-
diction would be obtained by (i) use of the preceding method to calculate the
spectrum for each noise component, (2) adding these spectra logarithmically to
obtain the aircraft noise spectrum for the particular configuration, and
(3) calculating dB(A) and PNL by standard methods.

P..

L4
 20

100

E U-

. o1 TUD Fl I,
w0 I9 16 2
cr <-10 ...
ozD

1"I C Wý
('T .1Y
01., 1ý~ ~I ~,]IP, N!O
i, I I ) A11 ! C1 ' ' ;
-2 0
< .. ___. . ... .. . . .

100 200 300 500 800 1000 2000

AL.TIIrUDe, FT

-
H, I l,
F ~~~~100--__ __

LLI 9

-J

_Z 10000

C<o:

wZIL

:D

-< ao
SLL
LL

jC )

Hi U' .

uj LL
 0,6

0.4
0.5 I j i I.

Si VELOCITY,

0.3-
1.- ' 0 . ....
U-0
LL_
0.2 ' ; "
z
20
, C,,

,< 0.08

-LJ
D
on 0.06
m

-0.05
cc 0.04-

(j.03

0.02

0.01- I , .. I
1 3 4 5 3 8 10 20 30 40
WING MEAN GL OMW THIC CHOR), S/t' (S/ARi 1
<F
2

FIGURE [9 - GRAPH FOR J)ETERMININ(G W[N(; TRAIIING EIX;E BoUNDARY LAYER TUICKI,E'S,•),
FOR UJ,2E IN OBTA•IV NG S;PEC,'RA FOR CJEPN AIRFRAMES;

/ 1, ¶2 ' '

.yn m [imr I•'I-I

 0
N z

e uj

F-i

LO CO

CU )

ci'

00 <2
z~ to_

0
Lii a CIS

oc
2

U] 0<

cc
C)V
I z; r
0 1-
<2

C)ri

C>
0 0 0 -
m2
UP' Id)' &,'VI]A II ')NM
IM SS Li'
ONM)'
TA) ON9
/I OZI
AA IAHo
 110 -

in

C i1ROARLNG EDGE
it CC F LAP AREAS

0 9
CD < 2000 4

1000-
Cc(300

OW * 400

n-J 8
Lt
>]
L:

co H
LI 70
CU)

C- CO

0<w
u)
W

LU 5

O;it
it

40
5 O101!- 0 0

OH1 iNlT E iý)FýI A R All; F ,l' l" 11

oV
RAA ANE, HA ,( 1C A.'-'C FI IY FlAl 'E A ,;+

'1'(
I20

LU
C() 16

LU LD

LUw

00c
i i 12 i -----

u0

<- ----
---

4
4
 0 <

Lu-i

CD

F-

LL- >

LU (U

00
-J Ho

(I) -

LDU

t LU
.i :o

D U.
.r rr

L IU

CD( U
 1100

wi 90 TIRE

00 1 DIAMETEA.
2; 1INCHES

0 60,

< 40 x
80
C) D

Z-¼-

070
LC
/ / 20
I-

1.460 ~//rnSBRC

VLLCIYccOT

<1 D

f14 -
 'C)
Ivr

1.i

- -i tj-

wJ z

LU~

oI -j
/1 S LU
_____
-- 0< > c

~~14-

1w x

< LL
D LU
CI H -

LU )

1---- 4

HPI
S OId'] l ]H SSI
-- -'- I , )ElI I I V;i Nc
 120

ANN-,Ž ANCE:

z "YMBOL. MEASUOL

Lu

0u 8
Lii

C)

< PEHCF.IVED NOISE EVEL,

PNdJ*1 OASPL 7 1 dB
__A WEIGH1TFD NOSl ~EVý-
m 60 A IR iA) OASP t 4.6 d R

Ln

C)

I' 4

40/

20 '10 60 80 100 120

*OVERALL SOUND PRFSS!JRE EFyI- dB11

FIGURE 26 - VARIATTOPJtN OP~A-41[r:)NMSE) lFVPl ANI ECEVi NL,ý'F 1,FV'Vl

WITH 0VFRAlL: ",(lrIr) ESl EVI

j-1
 4~.0 NAý.A ANOFP METHODS

Two NASA Aflrcraft Noise Prediction P1rogram 'ANO.PP) methods for

calculating airfra~se' noise wecZ Cobt&ifl.:d from DNA&A within one oomputer, pro-
grin-m Option 1 of that computer program is the tfotal. aircraft noise method
develor~ed by Hardin (Ref. 4+) for clean airframes. It was derived fromn a
re2gression analysis of measured. peek OASFL for selected aircraft flyoveýrs.
Y4ean sqaaxre acourtic pressure was assumeri to) vary inversely with diistance
squar-ed. E-ponents fro' velocity, wing Prev, 'Wing. aspect ratio, and gro~ss
weight wele determined whicia min-imize the rms error. The data base d~d not,
include any jet, aircraft ot~her than ý,he C-5A , This method was shown iii Ref. 5
to oinder~o~timate OASPL of business jets by about 8 dB and low aspect ratio
[ ~delta-~wing aircraft by about 20 dB. Comparisons between measured and calca-
lated spectra for this method had not. been published.

Opticis 2 through 8 of the NASA-supplied computer program calculate- air-

frame noiLse from seven components as given by Revel-i's drag element Method
(Ref. 2-4). ThIese components are the noise caused by wing profile drag
(incLuding trailing edge flaps), wing induced drag, and pxofile drag of the
fuselage-, nacelles, horizontal tail, landing gear, and leadinj edge slat,
input d&ta for those calculat ions include flignt speed, altitixde, airfrar_,ý
Ceocet~ry, lift coefficient, and the drag co_)fgicient, for each component. Flow
velocity at the wing upper surface trailing edg,,e, as cal-culated from the tnag-
n'.t~de and chordwise location of wing rmaximumn velocity by use of Eq. (4) of
Ref. i14, ilso is needod. For prediP.Jon of airframe flyover noise, spectra
calcul1ated for the selected options, must be logarithmicatly summed by the user
and adJL1.st.,d for the van at~ion of far-field distance with airframe direction
angrlE?

The comjuter prog-ram devel-oped by INASA for the drag element metiiod
contains- M5ch nornbei convection terms which were not included in Ref. 14. Use
of the( low h num'.xQr approximation was, discussed on pp) 1-6 of Ref. 1.4.
Howver, the resul'k! -, eqaatý..ons were applied J.. Ref. l6 to flyover noise
predict i o for the Lockheed Jet.Star aircraft at Mach numbers,, frow 6.38 to 0),Y
where teoeffoct's ar.e not. small . Convectiocn. effects on di rc ýt.i vi ty poti~lern
otre inc-l titled t.n the AHOPP method, causing, cal co lated wax ~munm OAL3L to o:cc or at
aI ret~carled -ti me pos),,Jilon upst~ream of the overhead positi on-. 1!oweVer , thle
Doppler sWh ft. of cal cutl ated s-pectra was mitedfocr all Aiirecti on anglýes,
,Phu,; ',he :ipectrrta o 1ae by the corre cted NASA AN(OPP drip,, e .1 ment. molR ohl
ind pr,;ensettld inri Ii:; report wculrI mat ch thc,,e jse by d1i rect ap ain
ýCIrt ion: in Be"'. lbý for t he limitrd of' very ilow sub;',mni c Mtich numbers- (lie. -s
fl~frl ().i) tiat woulA ri'ti recfs:nýri ly matcoh anfy ri her v .rsi on Ci thaitt methodl.
 Con, i derable aerodynam~i c :inforima t~oi Is re quired asirpal; for calecal at'ic a
')f airframne noLise by the drag element method~. Profile Ir( coeffi cients, for'
aifrmeWagtd~l, o el~, and nacelle surfacesl Were esiatdby use of
the table on p) 95 of' Ref. 20. The increment, of profil~e drz:g, caused by trail-
ring edge flap deflection was; obtained from Figure 2-468 of that re~ferer.ce.
Lift. coefficient for each fl~ight.-test c:ase was calculated from the known
gross weight, wineý, area, flight spead, and st-andL~rd-atmosphere air 'tnsity.
Induced drag coefficient was calcul~ated from the lift coefficient and wing
j~eometry by use of Eq. (2-86) of Ref. 20. Resulting calculated variations of
* ~ tOTai drag coefficient with lift coefficient, at lift coefficients correspon-
din
figit t oniton, enerally were wsithin 10%A Of availal un1.1ipubli ,hed
flight test data for o:.-e of th~esfe aircra~ft with landing gear retractedi and.
* flaps rooth retracted and extended. As had been recommended in Ref. 14,.drag'
2oefficient of the landing gear (referenced to wing plarforin area) was taken
equal- to the rat~io of tot~al Dih;ýding -gear fro~ntal area to Wing planforln ar'ea.
IA~irfoil veloci1-ty distributions, and therefore the t'railing edge -vlocity,
wer2 calculated for each virfoil shape and lift coefficient by use of tables
in Ref. 21.

U~~~.p~r;C. '..o;~-. Airsrafti. Plrfornioivee , 8tihj , nrid

Conr, ol oriWi.1ey &
J.ý1 n2 ro . New York, 14,C).
1
Lot
-AI . fl. an.d vor Doenhk <,1, A.. E . 11hyo(ry ot, Wi rlt' 8.
~ird114rof at ;Jýluiso 'i (it, Ai r 1' i. Data. TX*V(r P h at as,1Ii' , New Yo-k , I¼
 *)COMARIS110ON OF MEAStUIED AND PREDICTED FLYOVER NOTISE

ý).l Limitations of Available Data

Airframe noise dat'a utilized for these comparisons were measured by three
different. organizations. The Measurements and data reduction processes used
-n each t~est program IT'e discussed here. Most of these data were obtained by
PIASA Dryden Flight. Research Center; overall- sound pressure level- (OASPL) and
so,(me spectra were reported in Refs. 1-1, 17, and 22. Tabulated flyover spectra
were supplled- by NAS)A for the 0.1 second time increment which yielded maximum
OtA," _1,. These sipectra were adjusted by NASA for the difference between atmo-
spheric attenuiation at, the measured temperature and humidity and that for a
.staýndardl atmosýphere, over a path length equal to flight altitude h.. They were
aiihr
jdu:;tced by NAS'A to a 500 ft. altitude by adding the increment
o0 iioc l/52 at all frequencies, plus the change in standard -atmosphere
uttciatonover the distance (h-500) at each center frequency. Values of
uASPJ, aiirl perc.eived noise level (PNL) were calculated by NASA for the result-
In; pectra. Such spectra have been described by NAS'A as maximum OASPL con-
dition, corrected *to standard atmosphere and 500 ft altitude. The aircraft
jo,,ometrtc position at this measurement time was not specified. "These data
had been measured with1 microphones flush-mounted in large flat plate- laid
over the g_,round. To obtain sound levels that could be compared with predic-
t~ions fur tripod or post-mounted microphones as for FAA noise certification
me,-surements-, t~he tabulated values of OA3PL, PNL, and one-third octave 6P.1
were decreas-ed 3 dB. The assumption that a 3 dB decrease in all. one-third
oct~ave SPL's produces a 3 dB decrease of' PINh is not rigorous but introduces
less, than I PNdB error at the maximum amplitude for these data. Values of
A-weighted sound pressure l1evel, dB(A), were calculated for thes3e decreased-
ampliltude spectra as part of this contract.

Skpect0ra measured by Lockheed-California Co. and tabulated in Ref. o)were

ohlai netl as, Ihe maxi mum measuredi values in each one -t hird octave band during,
fi rht,
1`cl The, res ltii iro compos itIe s-pec tra therefore consi--sts, of indi vi d-
ual one - t hi ri (COt ave ma xi ma which Ud c not, al 1 occur at. the same ins tarit. of
ie .
m 114lowvr. (tf,"5Pj (, the:e composite s7pec t m only:A y exceedeod the
htrji
a t roc t si Ival.ue of' hA SPI. . Mea sured spectr a andl (ASPI were pro' ýfltod,
1n lmw ca: 11hoctu nl composit e spect~ra andA their s (A'SPV' were tabul cited.
Al: Sm.
0, I fl el pec ira were pl~o tied. 'Thes-e smooth curve.- eliminated s7tmont,
~iii
citi ' oi~on:,: at. low Creq icenc le s, namwad-admpeaks caus-eui ny
IsathereWTC prop I ler ý.C,, and crl peak: cuse by airtframe pmtr : Ltl:
Cch

T~cni
I .lc.i~aP. Y an. iii urchemn, F. ., 'Ir.
eu -cco
p ii ; srrame( jtli ::e Yrelii (,Ai oct Mcdiel 1kIAA Puper V %(,July 1ioYi
as radio antennas. OAPI,'' s, for the smoothed, ideal. ized spectra also were
tabulated. The smoothed data were regarded in Ref. 9 as representing funda-
mental airframe noise excluding the peculiarities of each specific airframe
and measurement installation. These data were taken with tripod-mounted
microphones and correspond to 500-ft flight altitude, unspecified atmospheric
properties, and no corrections fo2 atmospheric attenuation. Values of PNI,
and dB(A) for the selected spectra were caluulated from both tabulated and
smoothed spectra as part of this contract.

Spectra measured by NASA Lewis Research Center (Refs. 23 arid P.4) were
presented as functions of the retarded-time direction angle. These one-third
octave spectra exceed the engine noise for only a limited frequency range.
NASA had fitted the normalized spectrum cucrve of Ref. 9 to these data. The
resulting smoothed extrapolated spectra were utilized by NASA to calculate
airframe-noise maximum OAS.L and the variation of PNL with direction angle.
The spectra were measured with tripod-mounted microphones and were riot
corrected for ground reflection.

Calculated spectra and integrated amplitudes utilized for comparison with

these data were evaluated for the retarded-time angle, in 10 deg increments,
that gave maximum OASPL. Calculated maximum PNL always occurred at the same
angle as maximum OASPI. Individual calculated one-third octave sound pressure
levels could exceed these levels because of Doppler shifts between the air-
frame-fixed and ground-fixed coordinate system and different directivity
shapes for various noise mechanisms. Composite spectra comprising the largest
calculated one-third octave SPL's during a flyover would be within I dB of the
spectra presented.

.2 Aircraft for Noise Comparisons

The aircraft for which measured and calculated noise spectra were
compared had been chosen to provide a large range of type and shape. Within
each type, the specific aircraft chosen was one for which data existed over a
range of flig~ht, configurations. Another con•-traint was the need to choose
aircraft. for which propulsion-system noise did not, overwhelm the airfrfame
noise.

'ý41urley, R. H.: Preliminary Measuremes t.of the Ailrframe Noise From an

F-lO(,B Delta W.inw, Aircraft at Low Flyover Speed;;. NAS3A TM X-'l]Q7, March

tiurle , R. R. :2uppress :<r Nozzt and Aiframe Nod ,Se Meas;urement:; tiurin,
Flyover ofU a M ii }F-1l)4B AircraF't With tiUnderwint, Nact Ic s .
F A,1IE Pipe r
74-WA/Ae r -I
-- Nov. . )'(I.
 Data are available for the Boeing 747 (Ref. 11) and Lockheed C-5A (Ref.
<) large wide-body jet transports and the Convair 990 (Ref. II) and Vickers
VC 10 (Bef. 12) four-engine and. BAC 1-11 (Ref. 12) two-engine narrow-body
jet transports. Spectra for the C--5A have the disadvantage that engine noise
at and above fan blade passing frequency was dominant at frequencies that
strongly affect, annoyance-weighted noise levels. Also, flyover altitude was
roughly
, equal to the wing span so the data may not have been far-field. The
iAC 1-11 and Convair 90 wero, only tested at three of the four combinations
nfup
C and if,ear poý-,'iton (clean airframe, landirng gear extended, trailing
flaps extended, and both gear and flaps extended). Spectra were avail-
able for the Vickers VC 10 over the largest range of configurations (three
*f'lp deflections). However, calculated flyover spectra are available for the
-1eing 747 ax,, given by the highly detailed airframe noise component prediction
Sof Ref. 4. That method had not been evaluated by any publish-ed comparisons
with Iata. Therefore, use of the Boeing 747 as a test airframe would allow
Sevallit, airframe noise prediction method. Data from the
ion of !,,n additional
Vickers VC 10 and Convair 990 were used, however, in developing the airframe
noise component method given herein.

Dita are available for two business jets, the Lockheed JetStar (Refs. 11
and 17) and Hawker Siddley 125 (Ref. 12). The same range of configurations
(clean, gear extended, flaps extended, flaps and gear extended) were available
for both aircraft. However, spectra for the Hawker Siddley 125 were clearly
ion::.nated by engine noise at and above 1.600 Hz center frequency even with the
flaps and landing gear extended. Also, spectra calculated by the drag element
method had been published (Ref. 14) for the JetStar. Proper use of the NASA
A.40111 computer program for this method could be checked by seeing whether an
Sillepen, tent estimate of aerodynamic inputs for this aircraft would yield the
-sane predicted spectrum.

Another important clacs of general-aviation airframe is the light, twin-

ený,ine propel]er-driven aircraft. The only aircraft for which spectra were
avai able both in the clean configuration (Ref. and
I with landing gear and
t'iaps; extended (Ref. F7) wa:; the Aero Commander Shrike. Data wr'a'e supplied
Lv NASA iCr votar extended and both tgear and flaps extended at on airspeed,
S11n1 tuear cxlc lol• ý,t a hig-,hor airspeed. Clean-airframe data existed at both

ai p1 ac o110oie d ata were ()f interest because there is no question of

,af an-n ; t Aon f'rom pripu]i:l1ve-sy.stem noise. A. noted in Ref. , much of the
,allplace nolo;e data are re]aLively old and had been meias ured at low flight

"J. Noneý .utineAerodyiiamic Noi,;e Icnv c tii,:,tio,.

1 "a 1,irI'e
,S•raf , PAiA :.-..23 72, (,t l] "11.
all itudes. The dat~a of Ref. 9 for the Pruc-2 sailplane avoided t~hese
di fficul, Lies . Comparli sons- were made at, three airs-peeds,- to evalutat~e the
unexpected poor a(greement, found between data and pre-d~ictions which usedI one
of the !pc-Ci fied methods.

Finally, the Convair F-10613 delta-wing supersonic aircraft, (Refs. )'3 a--nd
24) was picked as (A.smyall-scale representative supersonic transport. configur-
aatAon. Aircraft 'with low aspect ratio- wing.-, such asi this arid the Handlecy
Page 113 Rf 12). tend to have their airframe nois~e .'evels greatly over-
e.est ima ted by siimple_ prediction methods. Directivity hand been measured, for the
F-106B.

The renal t~ing~ ai rcraft configurationtelcssadiunmbrar

lisýted In Table 1. They comprise a total- of thirteen configurations- and Live
airframes3. Becau.se comparl.sons, were made for somne configurat ions3 au m ~ore
than one airspeed, a tot~al of seventeen cases were examined. The airspeedJs
for the:,e cases,, alongý, with measured and preýdicted OASPL, d_13(A), and PNL, ar,'
given in Table 21. Meosured and predicted spectra are plotted in Figlures' )`(
tburough 61.~ Ot~her compa ri sons are griveni in Figures 11P throatgh Is5. Tab'f"llat Hel
spec tra aNri h i r,f'rame geme ir'is prope rt, ies a re ri yenI n Tat) le:; - -8o
Appendix),fT-1

> ~Prue -ý ",allpl ane

Opectra meas;ured with this, aircraft in two runs at, each of three nominal
airspeeds are plotted i-n Figre 2 through 29. The lower ran 1riombter for each
set of data denotes a gr'ross,, weight 81% of that, focr the_ hig~her run number.
PredictedI spectra are shown for the NAO;A ANOPP total aircraft method and dlrag,
element method, and for the FAA noise component method developed under this
-ontract,. Thie difference in g-ross weights, war, calculate(, to caus-e a 0.0 dB
change in spectrum levels, fcorz the total aircraft, methesi and less thian 0.1 d13
for the other methods Only one cal cul ated curve !:is shown for each' meirhodi
.

At. the 1ower two ye] )c1[ties,, the s;tronges.,t inridiv i daul. one -third octave
bandsi wer tcw ow '00 Ifzs center frequency and had amp] it. de.- aipproximately
independent, ol' yelIocit~y. Mlea, ared OA,(Pl, s, were (lomi na ted by the I oo-t reuluercy
background nois,,e, whi cli had been fa ired oit of the smoot~hed s;pectra. dIpectra
calull e b 4he NA?.A ANON, total aircraft, noise method gýenerally mavtchedl
hle se msa s tred hig,,h valuie:n of' OA,2P],. However , they were aboti4. '8 lb1 above thie
inearscred :noes tra f or center frecluencie,-s greater than 100(4 Hz/. PNI. and dii( A)
were overpredic t~ed by aboiut the :same increment . This, method h1ad previ ouis y
be ne a u td it o); 1Ad2bPo p r nt ny teL-eaiit( ii~ ,ir lt~

0A, P1.<1~ ýIe mnlh( e n o~an dni ýiw nT~

 5 A
Is

A---NASA ANOPP TOI AL AI IP1,1HAF M IH(k.;

.j
> 40-- 0
w,., 30 ..
/
.A NOISE COMPONENT METHOD

m 30 0

D o -c
,x0 • -- -
20
"> 0 0-AlfOO
<i 20

0 NASA ANOPP ORAG,

10 SYMHOL HUN VKT

0, 2.' 0.71lN

0 1 1 1 1
50 100 200 500 1000 2000 5000 10,000

13' OCTAVF BAND CENTFR F RFOU-NCY,H/

FIGURE P7 -- COMPAEA20N OF WEAS.ItiEV) AN[D CAIC'IATED FIYUVER NOI .E 8PFC",., FOP

PIRUTE-? IIAI LPlANE AT 61 KNOTS, A RI8PEEIl

,, 1
 50

{
-~

NASA ANOPP,
0 i OTAL AIRCRAF 4NOISE
U- 40Q a A

od 0
of

c- 0K
0 000 20,0 100 00100

-?ý z'~J 0. NS
I AWlP
VA I f AA ME~ I?
 60-

0
00 50- NASA ANOPP TOTAL
V AIlRCRAFTI METHOD-'

LU

LUJ

~r40
(/))

.1-- 00 FAA NJOISE COMPONENT METHOD

LU__

..- NASA ANOPP DRAG ELEMENTIMETHOD

S20-
SYMBOL RON V, KT
0 21 99 8
A24 999

10L ___II_ _ -__

50 100 200 1500 1000 2000) 5000 10,000

1/3 021 AVE BAND CE NI I H Ri DUE NCY, ki,

1111 '-COMPARJ '04 OF 1 MEA :;iIRFD AND CA ],(.1 ý1ATh I 11 YOV-ER No I K:, PECR
IM1(pO
-2 VAJ TIPANE AT P f) KNOTi:, AIR:;,PEPJI

I1 1 ,
 The NASA ANOPP drag element method underpredicts the data at low
frequencies and overpredicts at high frequencies. It gives good predictions
of annoyance-weighted noise levels even though the spectrum shape is not
closely matched. The FAA noise component method comes closer -to the general
trenC. of the data. 'The calculated curve tends to be parallel to the data at
highly weighted frequencies (i000 to 4oo00 Hz). Thlis method also gives good
predictions of annoyance-weighted noise levels.

Data for the highest airspeed of 100 knots (Fig-ure 29) are dominated by
a laminar instability tone that protrudes 20 dB above the lower-frequency
base. This combination of roughly 1)400 Hz, frequency and 100 knots airspeed
was calculated by the method of Ref. 10 to be associated with Toilmein-
Schlichti~ng laminar ir~stabil-ity at the trailing, edge of a 17( in. chord sur-
face. That dimension corresponds to the Th'ue-2 horizontal tailplane mean
geometric chord, rail er than the Ilarger (28.5 in.) wing tip chord. Such t, ines,
are usually associat Žd with low Reynolds number and therefore low flig~ht speed,
if radiated by the wing. Evidently this tone can occur only for the proper
range of horizontal tail deflection, which is a function of airspeed. OASPL
g~iven by the NASA i/NOPP total airframe method approximately- matches, that of
tL-his: narrowband -random peak. Both t-'he total airframe method and dragý element
method generri ly matc& the measured one-third octave levels, at. frequencies,
above this peak, betwe-en 2000 and 5ý000 Itz center frequency. Thele measured
level-,, may be domrinated by na~rmonics of the laminar ins-tability peak. T[he FAA
Vnoise compcmaerif mrethod0 Underpredici sý those levels but is- cl~os-e to t~he diata
below 200 1Hz ind above 6300 Hiz center frequiencies-.

ins at)i _J tYv tunes canl be ci iminated by tri ppina th1,e 1uundary
Laminar
layer ups,tr1( M 0' tbe trilinig edge. inability of' the rioi:- curnpiienlt, iriethud.
to predict ; Leh ul se i.,- unimp~ortant . 'lone nioise can be idn fedby the
frequeiicy jaud et fihef. 1.0 and el~iminated at, nietr i giL:J (,,co.t inl wel ght.
an.d frict~iouin a

Mea:,ured spcIa(Pet. I,) fel i, th Aeco (Xnuomnanlecr ""hrdloi at.t R,

p.1st~nt.j'ri~;welre (111 au-I the pc~l cswelre teatllered. P'wa I i i, I-

wel givt. w~'-; 78%t' (X the heavier we gllut. N( vseat) itet of i

o(-curredi in tlie dataz. Peaik ampl i tle2ill t'nesietmicci :. psilu

p~c(Lrid r~f'.Iftct.uii) near 100 ~ and' nuoi a tw n the er o ps r Ill tilt'
ýP arid '4l If., un1e -tb il-d us [ave L.anis
I~. '[he :jwt i-tiim tv IL-l
1,yiiate :;
ANOPP t aIrcla Iin hi- lli:- an1(OAS,1. v'~hi (T 1ah1hwVtIua st-a

- W, -, I ):,%
iii It i'l, Ih tis I rod i i-t a c' e I t '' 4 oI1. t.1. t
 several dB and peak frequency had been reduced by about two one-third octave
bands. Spectra calculated by both t-he di-ag element method and the noi1se com-
ponent method are in good agreement with each other and with -the dgata between
500O and 400O0 Hz frequencies. Both methods therefore closely predict annoy-
ance-weigrhted noise levels.

An Aero Commander Shrike had been tested b.y NASA Dryden Flight 2esearch
Center (Ref. 17) at. this airspeed and the smaller g~ross weight with landing
fgear extended. Both the nose and main gear have single 'wheels. Acoustic data
V.were measured with flush-mouinted microphones. Reported levels were decreased
3 dP for comparison with these predictions, Resulting adjusted levels are
plott~ed in FiwLxce 31. This spectrum is considerably highEr than that for the
(-lean airframe (Figure 30) below 125 Ifz because of landing gear cavity noise.
Yt1 matches- the propeller-noise peak of the clean airframe at 315 and 1400 Hz,
nhqu 0 dP higher near 1000 Hz, and decays to the clean--air craft levels
abov~e 00(, Hz,. Landing gear noise is calculated by the drag element method to
peak at. !a relajtively low frequency. Thus the cavity noise is qualitatively
[ predicted near 50 Hz, but the calcuilated spectrum matches Lhat Calculated for
t~he clJean airframe above 400 H.The drag elertent method therefore under-
pred~i cl annoyance -weigrhtedl noise levels: for this configuration. In contrast,
"LIv- lnoise, compotient method underpredicts the low-frequency cavity noise but
.ui(),o21iy prediict~s the landing gear -ioise spectrum at high-annoyance frequencies.
BothA metlhods, overpredicted the measured noise above 5)000 Hz frequiency.

Specl ra, for t~he clean configZuration (Ref. 9) at, 113 knots are plotted in
1i IrLllFO ;.) . AmispLa tuceS of' he low-frequency peaks at Pt00 and 3P) H1z center
J.requienciS 11s match those shown in Fit tire 30 for this, aircraft at 1 53 knots air -
s;peed. Thc spectrum predi cted by the total aircraft method rang.,es from about
to( t0 LIN above t~he data above 5,00 Hzý cýenter frequency. As; with the ccompari -
Is() al. t~ho 100 jher a ir.-peed , th~i preli1 etc] s,pec truin ippcýar s to be dJi,-placedl
:thofl two otie-tIhi ri oct ave bairds tooI highl- in frequency In nulditi on t~o being-:
aboult.f " lb 1)igh it)i ampi i title. The drg lementi antI noise comonepný_lt mth
ci
tI ma B III( IW! i 1t t'requieric itcc above t he Io!lllp(~' t i.-f

N0 .1 I tIn :- ,I i rs o. irt~I tw, "I i'ti!., or;* w. !1itI le 1l'i it 'oarI

II II yw WI I i ii ti l , I''t a r t~ll 5 lit' I f- ýI it t .- I I ra" '
ecu ( Iilg 71)!
'p
~!tii , I t, waul it l i m '11. m':x t emid i-
r t le
h flt:ijs tt e e t he IAI -

:i I01011' !i' ti i > 1 l" cemit en !trfoil ieicv . Thu:ý a e m t iiit

IAI1 PNI.
W II~~i,!O :e'i h.j Iili p ~x!ti is n11 w I Il t tno I zllýI i igý gent di wli.. A: lit the Fit-ig le
V'
it, I ra
r.) I t I i v t l ie aim1
1 ip aiie w it hi i e i' t,)
iwl

Ii ''s i ,, tf- . 1,y !I'll( i sto t ITIIlkI ' :11 wi h cl e ,: ici p! ' -1 c I by

1 ! 1i1;' P it.-' ' ! 1 - ! . ' T0 IfKiI,L1 1 , . 0 ' u I e '' 5 i I ' 1 I . I

 10~~1 A A NOIS~I:,LF(N NT IM i'0FI)

AN A NiV, I
LU Vý 0 H-.A
FCHAF I
%l I4 1)

LLi

cl50. - -- NS N I \,A

I M NT %iA
I Mk MI l(I

LU-00 ,;YloIA ( HON C¼)NI 1H LAA lt N

40 -- a ~ (FA'H

o Al T I FAN I H1t? )t

300
50 10 200 S0() 000 20000 WO0

x;~j~X:.(W~ N IA~
AIVi 1 I1)I~X
If N I NP (A 1; *:
 /0 - - -AA
W)ISF COMPONt NT

60

CNASA Ar\,(Hr)P HA(, t I Mt N I Mi I11H1D

HA i \li)

C ý 0 0 0 00 0n50 00
I ):A [H NPC W IALFN :Y, ,

fw"kýjE)IY ~,N ". T :' lA Fl

1" P10 . !i ýAN1 f11 A1, F T. 'E: T i KN '' AI f:
 70 - _ _ __ _ _ _ _ _ _

C)N'A SA A N o)Pf 1 0'( -

r..0 A IIH H A I ' M E T HO D) 7

0 0
cr c

7i 0

0 40 NASA ANOi'PF 1HA(

1±1 F FMEN] MLIHOO11

C0 SYMFOI HUN COWI IGUFIA I ION

'7 30 0 1/ Cl 1-ANH AVY

A 20 CLFAN~ldHI

60 100 200 500 1000 2000 !3000 100

1/3OC(T A Vf BAND CF NFFR RFOM
10 NCYH/

i 7 IGIJR -2 - 01TAR- SON OF WEA.'1JREI ANil) CAL IfLATEI) F1YOVEP Nolý) K 8PECTIA
! 1,05
AER() COMM'ANDERS HRI KE IN CLEAN CO(N F (; ihAF ()N A"' 1Iý KN( ']T8" A lh 2PEEl)
 I-AA NOISE CVMN ONENT ME Il 00F

CID ILAPS
L AND GEAR DOWVN

>0 3

o (j AR DO*N
I Ails AND) (,FAR DOVVN NAA ANNor'P
al~~

>40 DH AG, F I E MEN \IMET 1401)

;Or 0 HUN ((,iA I)OH

I O

I(AP-S AND (,f Ali DOWVN

200 500 1000 '?OCJ 1,( 1

50 100

13 OICTAVE 13ANI): riJT1 k M~ 001NCY If/

IJ~A11
;IH~1~~s ;u 11't4~~~iriiuANI) CALC(2iATI":P 1H' yovy No!, Jr ý;Im1,VRA FHI
lN
A PID NCTCEA R WX, T N DIL), A NI1 WI" Ii1A 1) PC, IANI
FBA
ALP MN ,1i
1M 11 1pIK 1 W lITf
i KEI l1,A ýI ,,
FXiT; NDEl) AC i1 K 14()T AlI F
I!-, 1)
:'AI,,A11I
 lifference between piredictio~ns above 5,000 Hz. Note that ft),: the f:L'equency
range o-f good agreement, noise froan the tra-1ii,4 edge flapr, as predicued by
the d,,.-ag element rnethoc: is 'n coincidental good agyreerient with the sum of
measured flap and lanrli~ng gear noise.

5 5Lockhieed Jet.Star

Noise uiea,ý;urernents for the Lock~ieed Jet.St(a-r buslness JeT. werc- ubt. i ned
by NASA Dryden I'lighat Research Center and were reported ini Pels. 1,1 inui
A fvvrnoi-se spectrum for the clean coiafiguýration at 24~7 knoi~s saed1 is
plotted in Fig.ure 34t. This aircraft and i&irspeed wiere of inrteres-t b~t;
ýfý(agreement had beer, shown in Figure 1.0 of Rcef. 114 between these dato n
predict~ions bl, ,tue drag_, element miethod. The spectruzi cu cul chteIl by th- Ilit<11A
AlNJ0F'P c,,.4..uter progýram for that method, -,hown in FC ur 34c, generally wltchesý
ch~at. calculated si: ceir un. The total ai~rframe Ynethord ,eneraly ý)z ci cU ('ne
4tqird &1-taL. , -ý PL', I')3 dOB above tlhe
f tac andl the noi c,( coinp n:n met~ho;
gei2 I Y i S ') GO .1C dh 1oW.

Ooe pecalipirlt~y nf s;pectra pr-otiin Rafs. Ii and i'7 for clean Jt.,
a rir---C1t . a.s a peaýk 1oca. near 1600 hIf for a -range, of lircraftsr
(:1,cckhc ad J3et-tar , ConlvaPir ()90, an)i iBo-j-nr 74Y() at. anil fligh~t. speeds . Ieca
tc l here fore were, repeat. (l '0o-' Lh Jaiýil;a~r in ci Cr' ca~nt ijgoratioli ými
½8knt ,te highest speet- 'o~r whi~ch data Were available . Th'e conmpa
o~f cac lan1at ion,- a~id canta foir nti.s condi - ioOn s given iii Fig,.ure '115 If(,
spcta deterilpý,nod from the Ora(, element; rcethod and the noise c(mponeit,
ro, tiacd are: a Ooat. 1.0 dF, apairt. The mneass red span trum i-s conb.u st, itl v -Icw
11san tlfta C fromn the drvug ele--ment mei-1od. :-tis joe the noi. se compont nl
me.tho~d e-xcept for i-oodl 4agrenerner near- 500) Hz and ahorte ' l()
iz frequer c I,

These data wqere obtaind by NAJ\wu hnegnssc twohyrth

.~~eve1I
prvie :adur level fl.4ght . T1hrust AcoA -rhere'ore erif >1 ( h
been increa-es adU) achieve tuie highIer urj d.The -tlrn; V
* e~~kitgi~ne;aWhich tgene rate ,con:;iderablte loe rqnn x ha us;t. rio c a;e o(d. il hj
f requeancy compre a a o-r ti eIt, j., .-i ke.Ly thatI thýe broadi peatý rlfer (,I)( II
soni~n Jguc 14ad whicoh d orina tea, both ( I, And plO-t t I:;
engiJne compre.s so(r noi so rather thran ai rframnenoat

A fute ol aio ciasc the impurtLance of enJrigi tc I i I

tmants fuIr the ,ieacn conffigurtationz. Thie varilati on oif uASM,1 with 0I
,arice for this, aicaf -adi beeni plo i.ted in Figture 0) of' icI 1'( . t wit.; n tA
that, GAISPL varu.i ad i.-)ve-rsely With meRasureinent dis-tance Iquarid , W1 Qii th an
anguk,,lar diirecti vity dependence a,. wouldl be en;pected for ai rl's one ii- , . 1
Ihcs
spectra had been requ'e :;t~ed from NALIA IDrydon F1 I ght Res:earch tter f-r I-iniI
Sy, Cs in t~hA s; - ind~y . lhe-y were not -suppiieci becausef thiey we re !;I ;ii r to Ie
contami rio ted by eýng~ine nois-e- at, rsmeasuLremernt 0 srect1.1on:ý
 10 0
NASA ANOPP TOTAL AIRF~Fk-ME METHOD

K~ ~ ~ l 90 ---
> 0 NASA ANOPP [DRA(,

ELEMENT METHOO

(n 0
800
LL
/0

Q 0

fAA NOI SE COMPONE N T ME HoD0

50 10 2050 00 20050 00
11>CAEHN [NFHFEt OD

- rMP R iE
2N F'IEA2
AP HEI ICU IYCERNO 2 VECTR ~ 4

1 ADC-TI
ROINYH I3
 100
NAS.A ANOPP IOTA[ AIRCHAFI- MEIHODJ

wj 90-
>N
uJI NASA ANOPP
/ UDHAG ELF MEN f ME THOD-

0L 0
D 0/
0 800 0
ELn

< 0

U 0
cfl 70

S50- 0---

50 100 200 300 I 000 2000 Co!)0 1M(000

i,ý OCI AVI BAN[) SO)JNI PHE!FSSUHI iAfV1 (IB

~'IG.RE~<1MAL
N ~ IA0II]A ~1W 'Y!K kF
1J)C1{EI)N IVIA
i1A N!I A AW 'N T # K]R 1. Al 110.
ER
 i 80 - -_ _ _ _

70
C)0
00
0

coF ,A NOISE 'COMPO NEN f ME T IOF

0l 6

50 100 200 500 1000 2000 5000 10000

1/3 OCTAVE BAND CFNTFR FRFOUENCY, H,,

- ~ n;;~~2 if N Pu Y( VER NIýIE~ 11,11FCRA

F!~h FolB
~:}KuL1:~,:
I ' IIANION I (;PiKAH EXTENDI)hJ Al' I"ý EP 1 A HPF

7')
'90

NASA ANOPP IOAU ILE, MEN T ML- ~HAD

LU 80-

u-i

:D

0 F~~AAN,-)iS[ CMPONEN1 METHODNI

z
14 lAlS AND) (LAH DOWN
0 0-

< I I Ar'S DOWN

<YMIL FU V, I) CONl FGLJFA I ION

u 0 ?A 1/0 W' I LAFPS D()OVVN

o 5C
10~ FI APS AND) OEAliF DOWN

50 100 200 500 1 000 2000o 5ooo 10 000

1 '3 OCT AV[ HANDIU

CNTN F HF MJ1 NY, H:

b;A~A;~-.C(oMPApI 2'(IN (IF M'AIH AN!I'A IA! A >,Y

ýl OVE N( I.I]2ý hIlE
IOCKhBETI JET:I'AH WI I'H TRAI IJNI; HFx;P PI-1A 1 EIXI:xiNV!+ AT Y() KNoTI! AN!) WIT
FIARAND.ANNIMi'c ';EAII EXPEND)EII AT' iY)L KHrT
 A spectrum for this aircraft at 182 knots with landing gear down is
plottpd in Figure 36. As with the comrarison for the Aero Ccmniander Shrike,
landing gear noise as calculated by the drag element method was predicted to
occur only at low frequencies. This method underestimated the measured noi~se
by Lbout 10 dB between 1000 and 2500 Hz and about 5 dB over most other fre-
quencies. The noise component method gave the general level of the data,
which oscillated roughly -5 dB relative to this prediction. Worst agreement
occurred near the peak which was attributed *to engine noise.

Measured spectra. are shown in Figure 37 for the JetStar at 170 knots,
with trai~ling! edge flaps down and at 158 knots with landing gear and flaps
down. There was no systematic Cifference between the tiio spectra; the
dlecrease of' flap noise caused by reduced velocity waas approximately matched
by the added landing gear noise. These spectra, as With all others for
diifferent confij-urations of this aircraft, contain a peak from 1000 to 2000
liz frequency. The dragr element method predicts a large amount of trailing
edire flap noise caused by high profile drag at 500 flap deflection. Airframe
noise is:, generally overestimated about 10 dB for this configuration.
Decreac-ingý the airspeed and lowering the landing gear is predicted to cause
about, P dB nioise reduIction, contrary to the lack of' change in measured levels.
"In cornt~rnst, the noise componlent. method closely predicts these spectra except
Cor, thle peak of' apparent. eng7ine noise. This peak occurs, at. frequenciesý which
are highYly woir~hted in predticting annoyance-weighted noise levels, so the drag,
element method more closely predicts dB(A) and PNL. Spectra calculated by the
nnise component, method for the two configurations and airspeeds intersect each
o hor, inoagrreement with the data.

',).6 Boe in,,

ry 74

A fl~y-ver peak nol.se :-pectrtim for the clean configuration at. 233- knot~s
iV c-ompared with predict ions in Figure 38. Both the drag elemenft method and
t he a ocomponen. mnet hodI predIi.0 the genieral I level of data uip to '')00 Hiz
(,ont er t'rouenqcry . A ove tý-0 liz the data have higher- level.; as, would be
o~xrwtIfed jOlterrvin rio; Ie . Thisi port.i on of the sipectrum aiý,ree.'; vit~ tiL
noi: c ve .1 pr oýd ctcd by the t1otal ai rc raft method . TheW airClaft, usetd illi
lii>
11:. 1. W~,; n eiry Boeirr 1 .ý '01Y-110) with thin-tip inilet:; eqCuipedwiI
iW-w-ini doot)S . TurL ul eic e genericted by thi ; type of iuilet ant] convected into
- tit' lu 5 ~knrwii to criuse etn nt.ltiIrnois;e. Current Boeilw 74(1;
.i.l
ti.IV. AlpilietwithoilL Lbiowin door'.;, rand have I '?5; propuil.i vt-systAem

7
is l r ii i ~'; I''. Thtr i r t s s aI
! a'' lig L piSV- it oi

Loll
V ~ii ir 1 ri 1 ertrrr iso it i lthf
,nt lie K i r
lk' ( I' it, WNss i'ýLlll.1
material in the inlet, fan exit duct, and core exit duct to suppress flight-
idle engine noise. Only after those changes was it possible to measure the
clean-airframe noise. Thu,, it is likely that spectrE, from thc unmodified
larger, louder Boeing 747 were dominated by installed-engine noise at frequen-
cies where such noise is importanit.

A spectrum measured for the Boeing 747 at 233 knots with landing gear
extended is plotted in Figure 39. This ai'rcraft has four four-wheel main
landing gea adatw-henoelding gear. As with other aircraft, the
spectrum predicted by Týhfe drag element method matoches that predictedi for the
clean airframe oxcept at low frequencies. Measured one-third octave :3PL' s are
underestimated by about 10 dB betueen 100 and 500 Hz frequencies and more than
15 dB in the apparent eng'ine-noise peak. The noise component method is 2 to
3 dB high below 500 lHz, about 8 dB low for most of the higher-frequency peak,
and in general agreement above 4000 Hz. Another predicted curve, obtained
from the component prediction meathcd described within the NASP. ANOPP airframe
noise document (Ref. 14), is plotl-ted as a dot-dash curve. This curve was,
obtained from the predicted spectra plotted in Figures 55 and 57a of Ref. 4
for this aircraft with landing gear extended and flap~s down to ottai~n landing-
gear noise and'clean-aircraft noise at 177 knicts. Amplitude was scaled with
velocity to the sixth power and i~nversely with altitude squared; frequency was
scaled directly with velocity,. The resulting predicted 3pcctrum closely
swatches that shown for the noise component me~thod. How~ever, the nioise Campo-,
nent prediction is dlominaued by calculated landine gear noise. Landding, f:ear
noise had been calculated by the component pre-dictionl method of Ref. 14to be
about 1ý dB below the clean wing adhorizontal. taill noise a~t the frequencies
-hown. Landint,- gear noi se ass given by 11hat mrethod appears only as cavity
noise at, very low frequencies. Thus the spectrum shown for the componient pre--
diction method realily applies, to clean airfrar~es. Comparing, with the clean-
za1irframe spect~rum shoi~n ill Figure 38, it would match the j-enerol11 level of thAe
enginie-attributed noýi se above 800 Hz but, 40Ud be mere .han 10 dB higher than
inns, of the lower -f'requency data . Tne at reetment shown 'iure
Fin j3ý) between
data an il t.ho componeicnt, predticet ion method for' the al r(raI't w it.i~i gear diown

!pen t ra meOa:,;ured. wit.h the Po g/(a_ approx Imu tel y I9 im .s wit.h

k&t
tra i i re Ipex ni a(id with f?_aps, lind 1and'i ogt, got,,r extenidecý, aýre
p1 ot..~t.e n iF' ti~re Wi. Extendting, th 1ardin' rig gAr o'dde ", to )4 Ili at ae-
cio: ies Ir 1,)c1.[ ( to ')00O H.Spectra pred ic t~el by thc drat, .,letrielit. method
for, hi- o('atirlOia4 io with i~'lenIet ed iFlarps are ab1100 diýJi t41li0 g bet~Wee(N
;~Ir HA ult in onil H dBII too I(j 1)ebtwneen 900 nt ()iof( IL. c-t-ier Irnqnjen-
. Th(r
b nrrnoa'ed noi-, op inind byr innlirig rear exteyi:ýi -n lit 1'requtrienci., ul
fio- :-, I ~ ol Ct
0r "he NA! A-,-pcam ored c'rn1ponot.it prve iti'(7 I tltkel from
~gne I QoPIlP2setnt oepnrt elno
emi u 1"lr I Ic Winl L
jir !11( IId I It bIII eore o (1 ; o f Ii a
rI() e I Igt. TiI p oJi
!()I i -ýC
a, to ii low f'cr mol I' "he frequelsny 1-trine t'iv the revion of
 100

NASA ANOPP TOTAL

> ~AIP.CRAFT METHOD
Lii

uj 90

Lii
cr 00DAT A, RUN 5

0 0~
0
0 ~ 0 0

<r 0

AANOISI-HO

50 1001 ?00 5001 1000 20(00 )OOo 10, 000

1 A C T AV I B AN 1 C t N I H L OU I MQC H,

LI (U{(K
-* ThMPRI ; 4FA8TP)
IM"0! AIJTh CAl !ILATEý) l`,Yo'flKYP NO I: UK ;I'FCTKIA 1,01
1K1
1NC I4N P1A~
CN Y", IRATH t AT KNOT: AlIR!'EEP

AjI 1
 100

NASA ANOPP COMPONENT PREDICTION MET HOD (CLEAN AIRFRAME

GEAR NOISE NEGLIGIBLE)
IFAA NOISE COMPONENT METHOD
LU 90
LU C)0 0 0 0
U2 DAAFU0
0J 0 0
0 8
DA RUN 6
NA, P
V) 0 0
LUj 80 0

0N\-NAS)A AN(IPP DR~,

ftLF-M NT ME IRUI0)

o
(3 60 -

!10 1 0 200 J0(I I 2000) W2OO1

(00 10 000

i, (3CIAVi ANI) (A Nil H I lit ()(t) NC:Y, H,,

FI t RE C VP4'A R I :',()N ( , . ,i, A • , l Y.' I,•+'t.Nk i . I4TV

buE 114iýd;
'i'•+ r 1TII I1A N[ I N(
.iq :FA k FX''iT,,,, It, . ... A' 1% , .,,q
11.'iFFll
+'

1'4
 100

IAA NOISF COMPONE NT METHOD

cn F LAPS AND GEAR DOWN

Li 90

CL
-.

21
:DNASA AI\OPP DRAG EL FV EN-iNWMTHOD
cn
Cl tAPS AND) (;f AR IjO(N
o 70-~
<l I Afs D(wVN

NAS;A AN()PP COMPONf N I I'RF fic: TIUIN

M\ SýMRL DIN I;D)NI(,IFA I I(N

0 31 AI'S I)DVVN
A 2 t Ar'; AN?) (J All DDNN

!0 100 2001 WIO 1000 2000 M()(

O0N()
13
i (I AV[ "AND IA
I I [H E Ht OtJENCY H

~
YMAI ~ ~
MVI 1A'K
1' I~ Y( \'Ek
!L:T~ N[ :1 *1PEI,'T A OP1
'VP;:? Ply
-N 1132I T411:1 A. Fl.
 possible engine noise. However, this method incorrectly predicts too low a
level of landing gear noise. The noise component method developed under this
study closely predicts the flaps-down spectrum and generally predicts the
spectrum measured with both flaps and gear down. It is about 3 dB below the
data for the highily weighted range between 800 and 2500 Hz, but the closest
other p-ediction (component prediction method) is about 7 dB too low. Both
the noi,,e component method and component prediction method gave a close
approximation to the general spectrum shape. For aircraft with deflected
trailing edge flaps, the drag element method gives a more sharply peaked
3pectrum which has the nofmalized shape as;sociated with clean-airframe noise
and an amplitude set by deflected-flap profile drag.

.1 Convair F-lOB

Airframe xioii:e spectvumn and directivity measurements are available from

a study (Reit3 2.3 and 24i) of in-flight jet noise suppres;si.on. In those te:;ts,
conducted by NASA Lewis; Re.,earch Center, an FiO6B turbojet-powered supersonic
delta-wing interceptor was, used as a test airframe. Two s-mall afterburnint,
turbojets were attached beneath the aircraft'. Various; noise :;uppre(s,.,ion
devices, were mounted on the.se turbojet::, and changes in noise amplitude and
s;pectrmn were determined at. supersonic trans:port climb-out conditions. Becau;,e
jet noi:.e has; (onsidlerabie variation with di rection i-itie, the flyover noise
ins-trumnentation and data reduction was; chosen to permit accurate measnurement.
over a considerable range of aircraft position. A Iframe noise was, determined
as backtround riotise that would have to be s;ubtracted from the total noise
spectrum at eauch direetion angle. The spectrum of measurel maximum airf'rame
noifce without the 55t,11 erinine:; nt, center frequenciei, from 200 toisI,'O Hs,, is-
s;hovwn in Figure )I1 along with s"everal predictions. The portion of the :spec-
trum below ý)00 Ifs: wa:s lomizruit. ed by jel noi.;e from the miain turbi Jei at
f1 i Iht
- idIe t. rusI ; 1he h i h- Pro quency pocrl i on w, 0ýmmina ted by compr s::-(,I
noi:,e I'rt m ho! enrgine. :;pe I rca nire given in 1 wýiqtre:s 7 !nlt 11 of iRet. :'P ani
l.'irrrre I" (of' '4f Pr
?ti r'fiurde1-tline (6irection anl:from -0t t,(o$)() Al:;so,
the normaliz]ed :pectt.rum nhtspe g;iven in .'igure 0 (i" Ret'. oi id ,en 1'fitt 1,)
thi:; airt'ramf e roi ne portion a)I' 1he s-pec rum 1.(o a] law (A'iliatian
at' maximum
(A.; P1,

'The tolti aircraft. noi ;e method overpredic t.- the meii:;ured :spect' rim 1,y
Pt) to l(),(W.A:; noted on p 31 of Ref. l , Ihi; meaisureneni wa:: ::pt'iinly
excIL ded fri 'in 11We ea.; t, -~;(18re:; data correlat io ut i i zed ti (ievelopirn,' *,ha
h1leth,11(. lT i1:: Po It_, jj( • ere tim:it t- al:;]s
1P vocc r:; (iot, here) it' Ilt,
h;h)wl,
t 4. 111 a i'crrni I I 01 ; met h- I 'I a: pi iedi t.() preO i ,, l Y i, tUr'!
,'tý!II T11 ,i tr' m tII,
Z'(
II"itiey Li,' lip-il'' I
Ii:;jL'. Ilow tw i I i -win I, :'i t'c
I RII. ).

S77'7
 The drag element method predicts a spectrum that generally agrees with
the data. A comparison 'between these data and a calculation by this meth(xi
had been shown in Figure 11 of Ref. 14, the major presentation of the drag
element method. The calculated curve shown in that figure is less wavy,
somewhat. larger in amplitude, and in closer agreement with the data. Differ-
ences between thesie two spectra calculated by the same method reflect differ-
ences in aerodynamic coefficients and wing trailing edge velocity ratio esti,-
mated for the same aircraft in these two studies and small differences in the
Lockheed-California and NARA ANOPP computer programs.

Two curves are shown for the noise component method. The method as
proeýrammed uses a relatively simple analytical expression for wing trailing
e(ire noi.-e spectrum. This prediction matches the general level of the data
Cor iw t'requencie:; but. uives too small a peak frequency and too rapid a
[lon.y. linwovwr', that. result had found (Figure 14)to underestimate the hig-h-
t'rei.ýn(cy noi..e "it wine taper ratoins less than 1/14. The spectrum therefore
wi:ý: reccmpiitid for :'er'o taper ratio. The resulting curve would be about 1 dB
Lhwer n-,or peak frequency -ind, as,-, shown in Fig.ure 41, does not give as bad an
'r..ore: ima e (o. dl rather than 1),t0 at 12Y,0 Hz frequency. The ncdi:'e c ompo-
nent methi c] early tgive:ý a wort-e prellietion than the drag element methodi for
h ,iper':oni.-tr,,n:-,port type of highly tapered low aspect. ratio wing.

By ,.;e ot' a rnormulizei i:peetrtm fitted to thes;e data fur different

flyec,,ion ang4ie:;, the variat.ion of ()A.;PL with ,direction ang;le had been
,lBh tined "tnd wu:; givt-n in Fjigure l)a of Ref. Pý. Thi.; varialion is repr4)-
tuced in ii 4urt
*L'. Maximum UAJPI, arid nil, occurred near 70() from the approach
ýir,-ct: d,.. Al-) ;hL)Wn are directivitie:: calcalated by the ,hre,- prediction
mt-!hod.; . Both the total aircraft method ari. Irat, element methotd as pruogrammed
t, fAk\
nave tiu ,i rec ti vi ty cf a lift iptole wi th c4nvective ampi fi cat ion.
1,r 'li L I 1 11 Ttch xiul () ' ?,t X d .1. t he cnal cul ntedi di rect i vi Iy is_ rel niv-.ly
.'l. :; l, L ['W t ,t ' i~t• i 151",• up.t Iream.n lit- fi c "I--. ctmpuoielth me Wi ,d, which,
u .'.. th,' Li r ktaliVI.y ,1 t1,1- ii l' edge rii,;e Wi h.1, u;! .",V(ecttive am41J1 fli'i t i on,
t, t .' , r (l i' " ,v,.; lan .. t'vel.; ian thi; angtle range. The Iatat ag,ree
'•it.t, h.; t'e l te t, ren,l. At larter the calmco,,a te liratliitu, te I' *r a
",iV' diptJ tfe cay.; inre rapi tly ;han t fthato•
r 'ra l rig ede P.
:'h- :'I! l'h. ,.ho ' rnpi ,i leo' y (1.;:e I ill tie 1I( i.: tlr'lIxrl'ell mo.th t I. 'll(-
1riirni r 'vtth-!iintioi(, l d(I'rlrNctivity for the drag, element meth•,i, Eq. (() T
H.-'. 1,', . w. t atu t',,r a liift. Iipole withoutt oenivectiwe ampi ii .ata n. lria ¶
no-,, i v ,' V(C.l I to. n xi l,,ail a1t (40 ' anrigl' anil :;.y1 nctzr (-nl fn w( it 'h111 ah Iv, t,
I oo !1- 11 ,it1 If"l~l .
 cn
:D

LLJ SONS ~f~ W

CI. 40 AI0 1 0 ~ 1N1MISI

/()7
L i,

/ "I
/ 0 0000O00OW 1001

o-W-1h, : l, , , ý
Ck w~k F xH 1,lJk ý MW
 120 - _ _ _ _ _

NiA'A ANOF'P IOTAL AIH(HAf I MU IHO[)

0

cr
U100

LIj
Ir DAIA

00 C
0~ 0

f AA NOISI COMPONI NI ME IMMJI

Lu ~ N;AA AN'WFP O)HM, f I F Mf N I Mf IHOD!

040 80 120 160

Po I AH AN(;I i Ifit IARD)H

A'~ I ME. DUG

ri-/ ' Ar ~IR :f VI

 p.8 Add]i ionali A ircr aft

To farther evaluate this noisie component method, flyover spectra were

als o calculated for several other airframes for which tabulated data were
available. Measured spectra had been given in Ref. 9 for the Dougls DM-3
and Convair 24v0 twin -enigine- propeller -driven transports in the clean config-
uration with engines- off and propellers feathered. Calculations were conduc-
ted by the ncoi:;e component method and the NAA ANQPP total aircraft metnod for
these two aircraft at their highes-t airspeeds tested, for 500C ft alt lIude.
Spectra measured with the Convair 990 jet transport, with 3(.1 trail inn, edg~e
flap deflection, anid the landing gear retracted or ext~ended, usnwere com-
pared with predictions; by the noisýe component. methodi.

Spectrat Fur the Douglas DC-3 ait, 148 knot:; and the Convul r PUG4at .!:)
knots are compared -witLb predictions in Figpure:; ¾-i and 44 . BOth1 aircra ft had
local spectrum peaks. centered at 125 Hz which caus-ed OASPL of the actual1
s~pectra to be about 2 d[B larger than that for smoot)(hed spectra. S)pectra coil
cuLated from the total aircraft met~hod tienerally wereý about 3 dB l-arger thaIu
tho~se from the noise componient, method. They matchedi the actual OASPL but,
ove rpredlc ted the measured spectra, parti cularly above 1.000 lIz where largec
contribo Li ott; to rioi,:;C annoyance occc r . The nois,-e componenit met~hod under -
predlic tel the is La bel ow 200 Hz frequency hal. generally agreed with the dii La
for highe r Irequenci e:;.

Wit-h
that, the two sýpectra :hewn in FT gore 14t6 f'or teCony;'li r ',1)1 have
s eiti1lythe s-ame level:-, oc] ow 10001 Hz cef:nter tresiijenci es bit, diI'D r at,
hitt.er frequencies. S1pectra, for es .t. runs '5, 0, and 7 of Ref, 9) for ti
p al ri ane 1'-tWn n tesmedy hv more rapi d hi ,it -fre quency 4wcay t. 11,
t~hose Ž"( t~e:'l rol:; I-ti f)Y' the saime ýiiirpin
'iEand same rangýe of axsel it

1 'lo hli e r ~'r :wei t'il. . Ph~i:: dtfference h,,:; prov en ite be i mportant.beo
'lhe mettboil for pre];1 cli (iiA' peak ifrequency liev-I opel i rin Ief% ), ¶n~

in the %A:SA AL,01E? t otd a-irren t. met Sod of' He': iri , wa.; si Yrtdl<ly
'rIVt'1 '(eli I by thle:'e It x.i it-:k freqwency wasý- Icrtermi xil'rmspcram(x;tr
witlii hi't Aero) (otmmritner :;tiiikt', t ,lIa:I?
,rnd he ti rst
PoiaIr.'i two,
itj rplara?.; have wing, thii ekrxez:'5 raio i(s of'I .' and 1?.1;their peak frejixen-
ei1: cO00011 hxivi been seca led( by us-e of' cither w4ig cho(rd or win, n-t x i fiim 'Coh-k
lie:'_ it 1C111 th parameter for ;t r()uhfAJ number. "lhe Convair :)tiu hit.; j H 'U'/"
thickxc:sratl so !. ea Inný the peak frequency inlv er::ely fwI iitht~ muxmm lhicoh-
ties:- reduce,: 'hat frequtency by about, one -t hi1ý idoctave. S:10.Žct~rz nt~ituýýre 'dfr
cte: rirn:; i-4 of. t~he y'onvu r 21;ho cani b~e Jihmn Lt) mliIcit ilose tou' tilt'e ii
wiý '.ý piun,it;"Ih Ale ,cile witLb a'2t.COUIRI ixurri.. r bas;ed exn willig churl It v
c'Vii', 11lb utit 1' thr e lest Tao. f.i the
Lii?' 'n ria r .SUGk wo~iI i.httii pp' t".
Ii :l oad'i .o ýw ji `,t.Je)unal nunitter. (ire sf 21riiai iuflhýitt ls'11 *nwiI4
A tiN iXt.;. stoas; the two ;t ,: ;ri li :ý'
;irt'ira for)? xi it
1111', t t "t

~
I. 1 'l ~ cthigt
its1 tlt no' tilx p)h- tI n t: o t ' Soi:Vt 1 ;an o lt lvt itlý,1.

'1c1rt l4v
 80 - _ _ _ _ _ _ _ _

W , NASA ANOPP TOTAL AIRCRAFT METHOD

UJ 70 c)

(r-

uJ 60

z ~COMPONENT METHOD)

WU 50 -
> ~Sy V3801 HUN V, K T
0 P, 148R

40 _ __ _ _ _ _ _

50d 0 2 00 500 1000 2000) 5000 10000

1 '3i OCTAVE BAND CE-NTE E?f RE WE1NCY, H/

I ~ ~ ~ 1U1t:I4:8
A~I1 PP i:~
I -TY:1) , 1 Yý VFR NO I ;F :;EE:[EA xi
A2 '-' IN ('!ArAN 'N Al
wvp:!AIH m4()'!., A! R;P:Ei)
 80

tj NASA ANOPP TOTAL

~ U AIFI(,'AF T ML THOU
uj
w70 Ll-
-J N.
u-i ~
ocr A~
L,*4 ("o

0 C-

z A
i:A

InSYMBOL Rt)N V.K f

LkJ

50 100 2U()LW 1000 201)000 130000

1./3()(. TAV[1 1AUVi')CI N IFIH R-U

Ri- ENCY

Y ,I \' - '!(i,"

('1
o :A
Ni (;llAT~ N V' I K t)l, Aý ,ý IT-3
 90 - 1
-J
LU 189KTFLAPS
UJ
-J
LU 80

ci)
C/)
LU

160 Kr (,FARANDFLAPS A
0
z 70 -
D
0
SYMLIOL V Kr CON& (URATION
C
z 0 189 jIG FLAPS
4
Lfl
LU GO A 16(1 GEAR \ND FL PS

4
I--
I'

__ ___iL il
100 '200 Y)0 1000 20.30 !)oou 10000
1 3 0CIAV [3AN[) (NH Hf HI cliii N(>' II:

[ .I*A1*' I iiA:bo;j A1l IRFAic'I'F' Fl Y 'f Ni:;F

r'i,ii * '.i:ii }•}l 13; F:IK;; 11.A1 i.lI':r'Ii:l Al ]¾( KI''I'V AIxPLK Ar;:
Ar; IN r f*:AH 1111 I>''Y* Al I I': A I } 11

U ill
rF.

caused by a change of atmospheric humidity and therefore atmospheric

attenuation that day. This difference has resulted in use of wing maximum
thickness as the Strolihal number eference length for predicting clean-air-
frame spectra by some current mc , ds.

A comparison of measured spectra for the Convair 990 with predictions by

the noise component method is given in Figure 45. The peaks near 500 and 1250
Hz correspond to peak frequencies in the noise ,pectrum for the engines at
idle power (Figure 4 of Ref. 26). The data for tie flight with trailing edge
flaps extended oscillated +3 dB about the prediction below 1250 Hz frequency.
However, rapid decay of spectrum amplitude at high frequencies was predicted
to occur about one or two one-third octave bands too low in frequency. The
measured spectrunm above 500 Hz frequency was underpredicted about' 5 dB by the
noise component method. Decreasing the airspeed, and exýendiln the landring
gear in addition to the flaps, caused about 3 dB noise reduction at midfre-
quencies and no reduction at low or high frequencies. The noise component,
method correctly predicted the size of noise reduction and its variation obove
500 Hz frequency. As with the flaps-extended configuration, noise at impor-
tant frequencies was underestimated by about 5 dB. Any modification to the
trailinr, edge flap noise prediction method that would improve agreement for
this aircraft, would worsen agreement for the Lockheed JetStar. Additional
data are needed on trailing edge flap noi-ie, including, dependence of the
spectrum ,shape on chord of each flap s;eiment rather than overaPli flap chord.

J •,I ' t,M, K. C. I. 1-., an t ti t iim, To 'rt 1Aird ury

P'V
M a,a.;tir,.ifwnll kii Aicr•'al" Airfr-tn,, 1N I~ie< WiliI 1,h(- 'iA *A CV - Ai•
JJAýýA TIM X-'% ,lun. Pi ;t,
 STABIE I - FLYOVER NOISE TEST CASES

Aircraft Clean Gear Flaps Both

Sailplane Prue-2 Runs 21-26

Ref. 9

Light Twin Shrike RUns 14-20 Runs 6,7 Run 4

Ref. 9 Ref. 17 Ref. 17

Business Jet JetStar Runs 2,5 Run 9 Run 7A Run 10

Ref. 11 Ref. ii Ref. 11 Ref. 11

Jumbo Jet. Boeing 747 Run 5 Run 3 Run 6 RuN 2

Ref. 11 Ref. 1i Ref. 11 Ref. 11

Delta F-IOGB Ref. 23

F|L

L•j
 TABLE 2 - COMPARISON OF MFASURED AND FREDICTED NAXiMUN NOISE

Configuration, Method OASPL dcB(A) PNdB

Prue-2 Sailplane
61 kt Actual Spectra 52.1 .35.7 +3.7
Smoothed Spectra 35.4 33,3 40.5
Total Aircraft 47.8 42.4 5.1.8
Drag Element 35-5 34.0 42.8
Noise Component 42.8 35.3 43.1
82 kt Actual Spectra 49.5 4o.8 51.0
Smoothed Spectra 43.5 40.8 58..
Total Aircraft 53.1 48.9 58-3
DraL Element 42.9 42,.1 52.7
1Noise Component. 48.2 L4: 7 52.1
l(00 kt Actual Spectra 58.5 )8.(- 6r.5
SmoeLhed Spectra 47.5 45.3
Total Aircraft, 56.2 ,9.9 42.1
Drag Element 47.8 477.i4 58.5
Noise Component 52.9 48.6 '8.5

Aero Commanler Shrike

Clean 1118 kt,
Actual Spectra (T.6 :'.5 72.9
Smoothed Spe'tra (.1 (0.1.
Tot ý 1
-- ] rc'at. 69.3 ('(.7 ''7.7
1)raj El ement 59J') ',•.
No i:e Component. (411.0 5 ) .) 74 .H
(;ear IDown, '112 kt.
Ac.tl alc ,('t.r14 7'] • t 0.
rr F lotrrie . ''.1,
N,)i:ie C:omponent.•' <•4 •7 .
(s'ear andi ,']ap;: Down, II ktV
Actiial ',peetra 0),.3 ,4. ,
1r-14,. Element.,.'( , .,-.2 '/
NoJ :j; L'ompw Ieent. .t 04 . '4 .4
' • _I i uxi , 1 4, kt
It - .. A t 1~a
,;I)Cttr) 7? .7 th .4 72.7
,iithu
>mc45u L;p,,' tx'a ,47,'.',
'7(.2 (( '
Tuttul Ai rcruaft 7. 72' .
1 i aii ' 1l ,merit. . , * ."
S.{ 't- Co In nI~a!(.II#Il '() . 472
,' '' '
Configuration, Method OASPL dB(A) PNdB

Gear Down, 153 kt

Actual Spectra 76.5 68.7 81.7
Drag Element 75.3 66.1 78.0
Noise Component 76.7' 73.1 83.5

Lockheed JetStar
Clean, 2147 kt
Actual Spectra 81.5 80.5 92.0
Total Aircraft 94-5 92.3 103.9
Drag Element 81.0 79.5 91.3
Noise Component 75.5 72.8 83.4
Clean, 358 kt
Actual Spectra 88.0 87.6 98.8
Total Aircraft 102.5 JOl.6 113.1
Drag Element 92.6 91.7 104.O
Noise Component 82.2 80.9 91.9
Gear Down, 182 kt
Actual Spectra 86.6 85.0 97.0
Drag Element 88.0 76.8 88.4
Noise Component 85.8 83.1 93.?
Flaps Down, 170 kt
Actual Spectra 84.3 81.6 93.3
Drag Element 91.3 89,3 99.Q
Noise Component 82.8 78.8 90.)
Gear and l'1aps Down , I jG kt
Actual Spectra 83.8 8,].' 93.8
Prig Element 89.q 8,7.2 97;{
Nc'i ,ue CThmporierit 82 .9) 78'{.'[}. .•

hot
iag 7t P(i
A' to, l Si ;Ip (c'+ r, t .7 .! i1 .
Tota.t Airertift. ifl.] .11i.2
DJrl, Element W7..,".7 " .
14•,i •;, ( ,mtozil t ', .1]4. t, • ,.
(G'in' lh wr , J1']' kt
Act,± pp.il trt '-. '., .'

ND i:;*' E Cat
e.'t O n) rt, ] .t '4-4

NOI....
 Configuration, Method OASPL dB(A) PNdB

Flaps Down, 198 kt

Actual Spectra 97.0 94.0 1-07.4
Drag Element 97.4 93,4 104.4
Noise Component 95.1 93.3 107.1
Gear and Flaps Down, 197 kt
Actual Spectra 98.8 96.4 109.4
Drag Element 99.9 93.4 104.4
Noise Component 98.2 94.9 108.3

F-106B, Mach Number 0.4

Actual Spectra 83 - 99
Total Aircraft 107.5 i06.2 117.5
Drag Element 81.4 79.8 93.0
Noise Component 80.2 75. 86.3

[!

Jr
 6.0 POrENTIAL CONCEPTS FOR AIRFRAME NOISE REDUCTION

6.1 Basic Geometry

Changes of the airframe basic geometry have relatively negligible effects

on calculated airframe noise. Variations of wing taper ratio between 0.25
and 1.0 at constant wing area cause less than 1 dB change in one-third octave
sound pressure levels. Reduction of wing aspect ratio at constant wing area
would, in concept, lower the peak frequency of broadband noise from the clean
wing, leading edge slats, and trailing edge flaps. This change would tend to
reduce annoyance-weighted noise levels. However, reduced aspect ratio is
accompanied by reduced lift coefficient and therefore higher airspeed during
climb-out and approach. The increased noise amplitude and increased peak
frequency associated with higher airspeed generally dominates the calculated
chantge in annoyance. increased trailing edge sweepback will reduce noise
inuensikie" by cosine squared of the sweep angle. Increased sweep also
causes i.norcased wing structural weight, decreased maximum lift coefficient,
tad increased minimum flirult speed. Neglecting those adverse effects, th.
m.aximum noise redution attainable by increasing the trailing edge sweepback
Lro•r Ph to An" would be only .[.2 dB.

As: nn treme change of basic geometry, the wing area could be greaLly
incrva .I to allow reC -ti on of air speed during takeoff and approach.
Cft in
i Vr he eI.>,t of loubi ing the wing urea at constant wing p1anform and
w up tech o]A,-,> .
f.li
a Wing weighibt of cormmercia transports is about 81%of
ct's:' wghth. .IJ 1lint 'pted ia pr )porti =nal to the sqt:are rotot of weir1,•
dlivi dcl by il, t)r()luct. of I i CtC coefficienL and Wiht' area. At cunst•n lilft
t'•C,'
I'i , t , Itulb i rig
tn' t lit,
wilur areta tlhen wol] d al l ow rluci ni ai rspeed to
( . ", .i
met,', , thet r v i W rcro
I'the Io:;ic . (The ulditLi n.l Iftu
l
',t'i tl. Ot t' d 1t ) 11lu ilC:t'ill fi t ,:afi ti ' range and Itayjl ad with the larger Wiat' "
pr.Iile I•r' 1i nit, Itr'' ted.) This.::'mua lI r at r; tele , ctr iined witl I
th ,.l'I'e
" cIt.
it •ii {( w n' ar ,tan II rip a ,eo, W i t id cr,'.b:g;' th
te clean -w
ni , i
lit ,

r't';I.teiv i i 'v. Al I IL ltt ,t:e pAta! itl iiis' redurct'i, s:: are it i,,w'lir'ibly
r'nIl , I
W ' I, Iiitw•l
1 1 tilt ar- P, :i mnlat :!Ac'r ta "C:'t
NI' i rirrill tIn'I l
1
i, ! , 'p 't'-rt i : - '" I'::I . 'lie Pat)t I
J( j ni t f gt' t re:'p i:;ev, n ntid re l Iu t' iromr air kIrm
nit

of ride ualIily, imlMt ly wouaic 1 bew unacceptable. 'Tberet cr,, 'ii zI'rumit
n , n 'T
r,' . ' ) lli
i('11 :'r be,. achi vve, bty , etempin: to mc.ilty each aoni e r i'.arceo :,
,' , a W ~ l' [ 'I l ' • It t l •i
 6,2 Trailing Edge Flaps

The strongest, single airframe iloise component of comnercial transports

during landing approach is the trailing edge flap system. Noise from trailing
edge flaps seems to be lift fluctuation noise caused by the wing turbulent
wake convected past the flaps. One method for achieving small reductions of
perceived noise is the avoidance of small-chord vanes within a multiple-
slotted flap. Rapid decay of the flap noise spectrum then begins at a lower
frequency. Use of the smallest flap deflection consistent with flight safety
will also cause small noise reductions. For example, OASPL of the Vickers
VC 10 airframe (clean except for deflected flaps) was reported in Ref. 12 to
decrease 2.2 dB when flap deflection was decreased from 450 to 350. The
method developed herein predicts 1.8 dB noise reduction for this decreased
deflection. Clearly, changes of flap geometry alone are inadequate to cause
much noise reduction.

Experimental studies have been conducted (e.g., Refs. 27 through 33) of

shape arid surface modifications to reduce noise of airfoils with incident
turbulence and of externally blown flaps. As described by Hayden in Ref. 28,
these modifications usually can be classed as (1) changes of edge impedance,
(2) changes of surface impedance, and (3) changes of flow mean and fluctuating
properties. The first category includes serrated and slotted leading and

2
'Potter, R. C.:
An Experiment to Examine the Effect of Porous Trailing
Edges on the Sound Generated by BLades in an Airflow. )NASA CR-66565, March
1968.
28Hayden, R. E., Kadman,
Y., and Chanaud, R. C.: A Study of the Variable
Impedance Surface Concept as a Means for Reducing Noise from Jet Interaction
with Deployed Lift-Augmenting Flaps. NASA CR-L12166, July 1972.
29
Hayden, R. E., et al. : A Preliminary Evaluation of Noise Reduction
Potential for the Upper Surface Blown Flap. NASA CR-1122(46, 1972.
30
McKinzie, D. J. and Burns, R. J.:Externally Blown Flap Trailing Edge
Noise Reduction by Slot Blowing - Preliminary Study. AIAA Paper 73-2145,
Jan. 1973.
31
Scharton, T. D., et al.: A Study of Trailinr Edge Blowing as a Means
of Reducing N,-Ise Generated by the Interaction of Flow With a Surface. NASA
CR-132270, Sept. 1973.
32
Pennock, A. P., Swift, G., c3d Marbert., J. A.: Static and Wind Tunnel
Testts fc)r the D)ewve opmenl of Exte rnn]l y 13 own Fla p Noi se Red ic i ion
Techniq tie. NA 2A CR-14()75 eb. 1)7'.7
17,
33 P. : investigation of'
Hersh, A. Hayden,
H., _. E., and Soderman,
Acoustic Effects of Leading-Edge -,errations on Airfoils. J. Aircraft. Vol. i],
No. )4 , Apri I 1971 , pp 1 -( )2.
 trailing edges (Refs. 27 and 32) to provide spanwise variation of edge
location, perforated or porous surfaces near the edges (Refs. 27, 28, 29, 32,
and 34) to provide a gradual change of impedance with distance, and compliant
flexible surfaces near the edges (Ref. 32) for the same purpose. Serrated
leading edges tested at low Reynolds numbers and low turbulence (Ref. 33)
have caused transition of a- airfoil laminar boundary layer and therefore
eliminated airfoil laminar tone noise (Ref. 10). Except for that one case at
a low Reynolds number, serrations and slots at leading and trailing edges have
not reduced noise and sometimes (Ref. 27) increased noise. Porous material
with a relatively large (40%) open volume, and perforated thin sheet surfaces
with or without acoustic bulk-absorbing backing, sometimes (Refs. 28, 29, and
314) were reported to give 6 to 10 dB noise reduction. In contrast, other
studies (Refs. 27 and 32) achieved a maximum of 2 to 3 dB reduction with
iimilar materials and geometry. Flexible surfaces (Ref. 32) had no effect on
noise radiation, as might be excpected from the relatively high frequencies at
which acoustic radiation is important and the relatively large inertia of
practical surface materials. Distributed surface impedance by use of com-
pliant surfaces (Ref. 32) nad no effect on noise.

,Many of the above tests were conducted at small scale (chords of s~veral
inches). Tests with an 18-in. chord uncambered airfoil in grid-generated
turbulence were reported in Ref. 13. Porous leading and trailing edge regions,
interchangýeable with hard surfaces, had perforated sheet metal surfaces backed
by a bulk acoustic absorber. This nonlifting airfoil represented an engine
internal strut. Lifting airfoils, representing trailing edge flap segments,
would also need an impervious central plate within the bulk absorber to pre-
vent. airflow from the lower to the upper surface. Porous trailing edges were
found to cause no reduction of incidence fluctuation noise. Porous leading
edges, however, caused 6 to 10 dB noise reduction above a frequency that
increased with increasing airspeed. The level and extent of this noise reduc-
tion was difficult, to evaluate because the measured noise spectrum rapidrly
jecayed into tunnel background noise, Reductions of' at least 6 dE were
achieved at IOi Hz center frequency for 150 knots (80 m/sec) airspeed and
l',i9) liz center frequency for 243 k ots (125 m/sec) airspeed. Use of practi-
cal-construction porous leading edge regions of wing trailing edge flaps
should therefore yield at least 6 dB noise reduction at approach airspeeds for
Crequencies which significantly affect perceived noise level. Perforated
metal sheets with acoustically absorbing backing material, capable of with-
,itanding strong acoustic pressure fluctuations without fatigue failure, are
current,]i used in turbofan engine nacelles. Such material is somewhat
heavier and more expens:ive than the conventional aluminum skins of wing trail-
ing edge flaps, for the same :structural reliability.

U34 iayden, R. E.:

USB Noise Reduction by Nozzle and Flap Modifications.
SPowered-Lift Aeiudynamics and Acoustics, iMUA SP-406, May 197T, pp 283-305.

N NII(J
 Another possible method for reducing trailing edge flap noise is the use
of screens (Refs. 28 and 13) extended below the wing lower surface ahead of
the flaps. The screens generate turbulence which has a small integral length
scale. If this length scale dominates the wing wake and does not couple
effectively with the flap chord, the flap lift force f)u.tuations and there-
fore flap noise would be reduced, The screens themselves would generate high-
frequency noise, but such noise would be attenruted by the atmosphere and has
low annoyance. Noise reductions due to screens have generally been less than
those from porous surfaces. However, retractable screens should impose
negligible weight penalty.

6.3 Landing Gear

As,,, ,how by Heller (Ref. 18), landing gear noise is primarily bluff-body
vortex sheddingý noise caused by the struts, dampers, axles, and other strut-
mounted hardware. Noise from the wheels themselves was found to be small.
Isolated landing gear cavities were noisy, but the presence of strut,; pro-
truding from the open regions inhibited the organized cavity oscillation modes.
Bluff-body wake fluctuations can be suppressed (Ref. 35) by use of a splitter
plate extending about 2 diameters downstream. This concept has been used for
reducing, hydrodlynamic dragý of underwater cables. The splitter plat.e prevents
formation of a Karman vortex sti'eet in the strut wake. It is not clear
whether the altered strut wake would continue to reduce cavity noise. Cavi ty
noise was shown in Ref. 36 to be eliminated or reduced by air injection with-
in the cavity base. As the injected air is convected out 01 the cavity, it,
shields he cavity downritream edge from flow disturbances; within the shear
layer. Flow disturbances otherwise would impinge agairst the down-tream edge
to produce inflow to the cavity which generates alditional periodic shear-
layer disturbances. With flow injection, these oscillations of the shear
layer and cavity internal flow are uppressed. A cavity tone which protruded
about 25 dB above background noise was shown in Figure 10 of Ref. 36 to he
eliminated by air injection. The required flow rate for arbitrarily chos;en
uniform i.n.jectJi on was about 1/8 the product of cavity plant'orm area arid free
stream density arid velocity. otier concepts:; for decreasing cavity noise were

S3511allett, D. W.: Splitter Plate for Prevention of Vortex STheddintg

Behind Finite Circular Cylinders in Uniform Cross Flow. Naval OrdnRane D,ab.
NOTR t'9-il, July 1967.
g' ;arohia, V. and NMrsier, P. F.: Control of Cavity Noise. AIAA PoIter
7(-',2R, July 197().

4I,
 discussed in Ref. 37. Cavity noise generally occurs at such low frequencies
that iti contribution to perceived noise is small. High Reynolds number
S---bluff-body noiose, which is broadband a.nd decays slowly 8t moderate and high
frequencies, should be greatly reduced by use of lightweight inexpensive
m ~ splitter plates.

Km 6.4 Clean Aerodynamic Surfaces and Leading Edge Devices

Airframe noise from slats, leading edge flaps, and clean aerodynamic
surfaces has been shown to be trailing edge aoise caused by the turbulent
boundary layer. One concept for decreasing trailing edge noise, described in
Ref. 38, is the use of sawtooth trailing edges to take advantage of the edge
sweepback effect. Depending on whether sweepback is assumed to affect the
spanwise correlation length, the ratio of acoustic energy is predicted to vary
between cosine and cosine squared of edge sweep angle. Model data were pre-
sented in Ref. 38 for two different sized sawtooth trailing edges, both having
600 sweep. Noise reductions ranged from about 3 to 6 dB, in agreement wi .1
prediction. Effects of a sawtooth trailing edge on weight and aerodynamic
performance are not known.

Porous trailing edge regions have been extensively tested (e.g., ReFs.
28, 29, and 34) for decreasing trailing edge noise from upper-surface-bL-wing
"externally blown flaps. Tests have indicated up to 10 dB maximua reductions
and about, 5 dB reduction over a fairly wide frequency range. Methods for
tailoring the perforated material acoustic resistance and streamwise extent
for the required noise attenuation properties were given in Ref. 39. As with
porous leading edges for wing flaps, a central impervious structure would be
needed to sustain aerodynamic lift. A very small increase of wing profile

L drag would be caused by the small chordwise extent of perforated material.

This passive noise reduction concept seems to have no other disadvantages
except the need to provide drains for rainwater collected within the acoustic
absorber.

3
V Heller, it. H. and Bli;3, ).
B.: F]ow-lnduced Presure F]uctuations in
Cavitie.; and Concepts for Their Suppression. Aeroacous;tics: 2TO1 Noise;
Airframe and Airfoil Noise, Vol. 45, Progress in Aeronautics and Astronautics,
M.I.T. Press, Cambridge, Mass., pp 281-2%6. Also, AIA, Paper 75-491, Mar. 19'(5.
3ýi ý'iller, L.: Swept Edge to Reduce the Noise Generated by Turbulent Flow
Over the Edge. . Acoust. Soc. Am., Vol. 59, No. 3, Mar. PP 697-499.
39
Bohn, A. J.: Edge Noise Attenuation by Porous-Edge Extension.-_ AIAA
Paper 76-80, Jan. 1976.

91j
 Use of trailing edge blowing to interpose a layer of low-turbulence air
between a trailing edge and a turbulent flow has also been examined (Refs. 27
through 31, and 3h). This concept has not worked consistently for externally
blown flaps, possibly because the blown air must have a very low turbulence
level. The need to obtain engine bleed air, and provide ducts and control
valves, causes increased weight, increased cost, and decreased reliability.
Use of a passive device such as porous trailing edges appears preferable.

Data are not openly available for the noise radiated by a clean airfoil
or wing in low-turbulence airflow at Reynolds numbers large enough to provide
turbulent boundary layers on both surfaces. This noise level of trailing edge
noise caused by the wing turbulent boundary layer is less than acoustic open
jet or wind tunnel background noise. Trailing edge noise of clean airfoils
has been measured in a small number of unpublished tests with directional
microphones rather than conventional microphones. A directional microphone
uses a single microphone placed in a physical reflectcr, or an array of con-
- l'e ei(a••nnic signal addition at appropriate time
s...........................
delays. The resulting output contains an enhanced acoustic signal from the
geometric region being scanned and an attenuated signal from other regions
which produce background noise. Such microphones had originally been developed
for studies of jet exhaust noise source location. Measurement of airframe
noise from clean wing surfaces and leading edge devices is within the current
state of the art for direoional microphones. Accurate measurement of large
reductions in such noise for much of the frequency range of interest may be
beyond current state of the art.

-- J
 7.0 CONCLUSIONS

1. The noise component method developed in this report correctly

predicts the amplitudes and spectrum shapes of noise due to extended landing
gear and trailing edge flaps. Measured noise of clean jet aircraft is
correctly predicted at relatively low frequencies but is underpredicted
at
higher frequencies where engine noise probably occurs. Measured noise of
clean propeller-driven aircraft and a sailplane was correctly predicted.

2. The NASA-recommended drag element method correctly predicts the

general level but not the spectrum shape of trailing edge flap noise at
small
deflections, and overestimates the overall lpvel at large flap deflections.
It poorly predicts noise from extended landing gear but correctly predicts
clean-airframe noise.

3. The NASA-recommended total aircraft noise method for clean aircraft

gave poor predictions of spectra and annoyance-weighted noise levels.

4. It is likely that trailing edge flap noise, at frequencies which

are highly weighted for annoyance, carl be reduced 8 dB by porous leading
edge
regions on all flap segments.

* 5. Landing-gear noise can be reduced by use of splitter plates down-

stream of struts and axles to inhibit bluff-body vortex shedding. Additional
devices such as cavity bleed may then be needed to reduce landing-gear
cavity
noise, which generally is decreased by the strut flow field.
6. Trailing-edge noise of clean wing and horizontal tail surfaces can

be reduced up to 6 dB by porous trailing edges.

7. Directional microphones will be needed to measure basic trailing-

edge noise levels of clan wing models in acoustic wind tunnels.
 8.0 REFERENCES

1. Brines, G. L.: Studies for Determining the Optimum Propulsive System

Characteristics for Use in a Long Range Transport Aircraft. NASA CR-
120950, July 1972.

2. Sofrin, T. G. and Riloff, N. Jr.: 'No-Stage, Low-Noise Advanced

Technology Fan. V. Acoustic Final Report. NASA CR-134831, Sept. 1975.

3. Morgan, H. G. and Hardin, J. C.: Airframe Noise - The Next Aircraft

Noise Barrier. J. of Aircraft, Vol. 12, No. 7, July 1975, pp 622-624.

4. Hardin, J. C., Fratello, D. J., Hayden, R. E., Kadman, Y., and

Africk, S.: Prediction of Airframe Noise. NASA TN D-7821, Feb. 1975.

5. Fink, M. R.: Approximate Prediction of idrframe Noise. J. of Aircraft,

Vol. 13, No. 11, Nov. 1976, pp 833-834. Also, AIAA Paper 76-526,

July 1976.

6. Ffowcs Williams, J. and Hall. L. H.: Aerodynamic Sound Generation by

Turbulent Flow in the Vicinity of a Scattering Half Plane. J. of Fluid
Mechanics, Vol. 40, Part 4, Mar. 1970, pp 657-670.

7. Chase, D. M.: Sound Radiated by Turbulent Flow off a Rigid Half Plane
as Obtained From a Wavevector Spectrum of Hydrodynamic Pressure. J. of
the Acoustical Society of America, Vol. 52, No. 3, Part 2, Sept. 1972,
pp 1Oli-10 2 3.

8. Fink, M. R.: Experimental Evaluation of Theories for Trailing Edge and

Incidence Fluctuation Noise. AIAA Journal, Vol. 13, No. 11, Nov. 1975.

9. Healy, G. J.: Measurement and Analysis of Aircraft Far-Field Aerodynamic

Noise. NASA CR-2377, Dec. 1974.

10. Fink, M. R.: Prediction of Airfoil Tone Frequei, ie.;. J. of Aircraft,

Vol. 12, No. 2, Feb. 1975, pp 118-120.

11. Putnam, T. W., Lasagna, P. L., and White, K. C.: Measurements and
Analyses, of Aircraft Airframe Noise. Aeroacoustics: STOL Noise;
Airframe and Airfoil Noise, Vol. 45, Progress in Aeronautics and
Astronuatics, M.I.T. Press, Cambridge, Mass., 10(y6, pp 363-378. Also,
AIAA Paper 75-510, Mar. 1975.

97
12. Fethney, P.: An Experimental Study of Airframe Self-Noise. Aero-
acoustics: STOL Noise; Airframe and Airfoil Noise, Vol. 45, Progress in
Aeronautics and Astronautics, M.I.T. Press, Cambridge, Mass., 1976,
pp 379-403. Also, AIAA Paper 75-511, Mar. 1975.

13. Fink, M. R.: Prediction of Externally Blown Flap Noise and

Turbomachinery Strut Noise. NASA CR-134883, Aug. 1975.

14. Revell, J. D., Healy, G. J., and Gibson, J. S. : Methods for the Pre-
diction of Airframe Aerodynamic Noise. Aeroacoustics: Acoustic Wake Pro-
pagation; Aircraft Noise Prediction; Aeroacoustics Instrumentation, Vol.
46, Progress in Aeronautics and Astronautics, M.I.T. Press, Cambridge,
Mass., 1976, pp 139-154. Also, AIAA Paper 75-539, Mar. 1975.

25. Munson, A. G.: A Modeling Approach to Nonpropulsive Noise. AIAA Paper

76-525, July 1976.

16. Pendley, R. E.: Recent Advances in the Technology of Aircraft Noise

Control. J. of Aircraft, lol. 13, No. 7, July 1976, pp 513-519.

17. Lasagna, P. L. and Putnam, T. W.: Preliminary Measurements of Aircraft

Aerodynamic Noise. ATAA Paper 74-572, June 1974.

18. Heller, H. H. and Dobrzyný;ki, W. M.: Sound Radiation From Aircraft

Wheel-Well/Landing Gear Configurations. AIAA Paper 76-552, July 1976.

19. Shearin, J. G., Fratello, D. J., Bohn, A. J., and Burggraf, W. D.: Model
and Full-Scale Large Transport Airframe Noise. AIAA Paper 76-550,
July 1976.

20. Perkins, C. D. and lage, R. E. : Aircraft Performance, Stability and

Control. John Wiley & Sons, Inc., New York, 1949.

21. Abbott, 1. H. and von Doenhoff, A. E.: Theory of Wing Sections,

Including a Summary of Airfoil Data. Dover Publications, Inc.,
New York, 1959.

22. Hersh, A. S., Putnam, T. W., Lasagna, P. L., and Burcham, F. W., Jr.:
Semi-Empirical Airframe Noise Prediction Model. AIAA Paper 76-527,
July 1976.

23. Burley, R. R.: PrelimLnary Measurement of the Airframe Noise From an

F-106B Delta Wing AirFaft at. Low Flyover Speeds. NASA TM X-71527,
Mar. 1974.
 24. Barley, R. R.: Suppressor Nozzle and Airframe Noise Measurements During
Flyover of a Modified F-1O6B Aircraft With Underwing Nacelles. ASME
Paper 74-WA/Aero-l, Nov. 1974.

25. Gibson, J. S.: Nonengine Aerodynamic Noise Investigation of a Large

Aircraft. NASA CR-2378, Oct. 19714.

26. White, K. C., Lasagna, P. L., and Putnam, T. C.: Preliminary Measure-
ments of Aircraft Airframe Noise With the NASA CTV-990 Aircraft. NASA
TM X-73116, Jan. 1976.

27. Potter, R. C.: An Experiment to Examine the Effect of Porous Trailing

Edges on the Sound Generated by Blades in an Airflow. NASA CR-66565,
Mar. 1968.

28. Hayden, R. E., Kadman, Y., and Chanaud, R. C.: A Study of the Variable
Impedance Surface Concept as a Means for Reducing Noise from Jet
Interaction with Deployed Lift-Augmenting Flaps. NASA CR-112166,
July 1972.

29. Hayden, R. E., et al.: A Preliminary Evaluation of Noise Reduction

Potential for the Upper Surface Blown Flap. NASA CR-112246, 1972.

30. McKinzie, D. J. and Burns, R. J.: Externally Blown Flap Trailing Edge
Noise Reducticn by Slot Blowing - Preliminary Study. AIAA Paper 73-245,
Jan. 1973.

31. Scharton, T. D., et al.: A Study of Trailing Edge Blowing as a Means of

Reducing Noise Cenerated by the interaction of Flow With a Surface.
NASA CR-132270, Sept. 1973.

32. Penmnock, A. P., Swift, G., and Marbert, J. A.: Static and Wind Tunnel
Tests for the Development of Externally Blown Flap Noise Reduction
Techniques. NASA CR-134675. Feb. 1975.

33. Hersh, A. S., Hayden, R. E., and Soderman, P.: Investigation of Acoustic
Effects of Leading-Edge Serrations on Airfoils. J. Aircraft, Vol. 11,
No. 4, April 1974, pp 197-202.

34. Hayden, R. E.: UISB Noi-se Reduction by Nuzzle and Flop Modifications.
Powered-Lift Aerodynamics and Acoustics, NASA SP-406, May 1976, pp 283-
305.

)ii
 35. Sallett, D. W.: Splitter Plate for Prevention of Vortex Shedding Behind
Finite Circular Cylinders in Uniform Crosa Flow. Naval Ordnance Lab.
NOLTR 69-31, July 1967.

36. Sarohia, V. and Massier, P. F.: Control of Cavity Noise. AIAA Paper
76-528, July 19,76.

37. Heller, H. H. and Bliss, D, B.: Flow-Induced Pressure Fluctuations in

Cavities and Concepts for Their Suppression. Aeroacoustics: STOL Noise;
Airframe and Airfoil Noise, Vol. 45, Progress in Aeronautics and
Astronautics, M.I.T. Press, Cambridge, Mass., 1976, pp 281-296. Also,
AIAA Paper 75-491, Mar. 1975.

38. Filler, L.: Swept Edge to Reduce the Noise Generated by Turbulent Flow
Over the Edge. J. Acoust. Sec. Am., Vol. 59, No. 3, Mar. 1976, pp 697-
699.

39. Bohn, A. J.: Edge Noise Attenuation by Porous-Edge Extensions. AIAA

Paper 76-80, Jan. 1976.

iU(

i||lj
 9.0 APPENDIX I: MATHEMATICAL CONVENTIONS FOR COMPUTER PR IRAM

As with the NASA ANOPP method for airframe noise, the acoustic field is
represented in spherical coordinates r, 0, 0. The polar angle 0 is measured
from the aircraft forward dire-';ion, which for calculation of airframe noise
can be taken as the forward horizontal direction. The azimuth angle 0 is
measured from a reference plane containing the flight direction and the verti-
cal direction. It is assumed that. the distance from the airframe is large
enough so that acoustic pressure varies as the inverse of the radius.

The Fourier transform of the acoustic pressure is

P": DP(fle0 d t
O- (A-1)

so that the mean-squared acoustic pressure is

2
:2f o j 2 du (A-2)

The integral in Eq. (A-2) may be expressed as a sum of inLegrals; over all
frequency band.-. For one-third octave bands, the mean-squared acoustic
pressure in any band is

Lf
| ]<: () IPW
_2#"o,flu,
2dc A'

where and ca are the lower and upper limils of the frequency band.

Since the acoiustic press.-ure vwrie.- inversely with rtlius, the i nten:sity
I may be uiven as

i A
, (r,&,'P):
-I ,
r-c (
IQIIIIF 4)(Ab
 where A is a representative area of the noise source. The acoustic power
which is within this frequency band and is radiated through a solid angle da
is

P(el)): r2I A (A-5)

Neglecting acoustic absorption, the acoustic power is thus independent of'

radius and is a convenient quantity to represent each acoustic source.
Equation (A-5) may be written in terms of dimenLionless ratios by dividing
the acoustic pressure by Paca to obtain

P(3C - A
((O
(8,)
Pf
Poc 02/
\(A -6)

The term on the left of Eq. (A--6) is a dimensionless acoustic power which is
equal t(, the 1quare of the dimensionless pressure on the right of the equa-
tion, That i s,

in dimen,'ionle:;s notation. As with the NASA ANOP? noise prediction programs,

this computer program is,written entirely in terms of dimensionless variables
so that. they will be valid for any system of un, t,.

Al.1 s;ubro0Lt,1ne.! follow the :me

a, computation ,iequence. The dinens ionile;ss
total a(cousti c power for each rioi se s:ource is: enIcLc./Uted firs-t. The power
P(O (,) rmliateki in any direction is then found by mulTiplying the total power
by a directiviLy factor i)( 0,)/4ar where I)( (4,) his been normalized by the
C oafl~ it t, ~i I

jo Jsn 0 D(0,4)dkd 0 -4r- (A-8)

00

L11 other D(0,4b) is def'ied such that, i Us aVeraet val te is equal to one.
 The power within any one-third octave band may be found by multiplying
the power by a spectrum function S(fn) which satisfies the normalization
condition

Go S(f):I
(A-9)
Zl SO:

Thus the total acoustic power produced by a noise source in a one-third

octave band i s

P(fn ): PS(fn) (A-10)

and the power within a band and radiated in a given direction is

P(O,-,fn): P (-(e-o)- S(fn , ) (A-i )

The mean-squared acoustic pressure at a given distance in a given direction

is found from the power by multiplying by the factor A/r 2 ; that is,

p2 (r,O,q6): P(A/r?) S(fn,8e,)

S (A-12)

These mathematical conventions are identical to those of the NASA ANOPP

computer program for airframe noise. Differences occur in the specific noise
mechanisms assumed for noise produced by varicus airframe components. These
mechanis•ms can have directivity functions D074), convective amplification
function (not described in tne preceding analysis), and spectrum functions
S(fn) which differ from those used in the NASA ANOPP method for airframe
noise.

AlI airframe noise raliation wqs repre3ented in the NASA ANOPP method as
being, produced by lift dipoles-. In a coordinate system moving with the air-
frame, the directivity fa',Ior would be proportional to siin2cos*. Motion of
this ac.oustic field past a fixed obs-erver causes the frequency and amplitude
measured, by the fixed obs-erver to differ from those of the moving coordinate
system. A frequency f in the rground-fixed sy.;teem corresponds to a frequency

i03
(I-M cosO)f in the moving system. This Doppler shift occurs for all types of
noide sources. For a lift dipole, the directivity factor in ground-fixed
coordinates is

DL (84)= sin 2 8 COS2 /LA(M)(I-M Cose)4] (A-13)

Here, the function A(M) is an average convective amplification factor caused

by compressibility. It is defined by

A(M) 2wfV sine9 sin 2ecos 2¢ 4 d d0: 1 2

4- (o
Vo
(I - MCOSe) d (I-M- (A-TL)

Convective amplification for wing and horizontal-tail trailing edge

noise and leading edge slat noise was taken as

DE(0,0): COS2 (e/2)cos20/1A(M)(1-Mcos 9)4J (A-15)

Edge noise from the vertical tail is rotated in orientation from that of the
wing and horizontal tail. Its directivity function is

Dv(O9#) -Cos 2 (e/2) sin 2 9ý / I[A(M)(i- McosO)ý] (A-16)

However, measured airframe noise directivity for clean airframes has been
found (Ref. 15, and Figs. 5 and 6 herein) to be more closely given by
"neglectintgl the Mach number dependence in Eq. (A-15).

Landing-gear noi:se has been found experimentally (Ref. 18) to be

approximately independent of azimuth angle. Variation of directivity with
polar angle was not specified but is arbitrarily assumed to be that for a lift
dipole as in Ref. 4. Sound rauiation from the wheel-well cavity would be
expected (Ref. 4, p 44) to be that for a monopole source in an infinite
baffle. However, cavity rnoise is not specifically represented in the method
developed here. The directivity function for landing-gear noise then i;;

loa
 Dz(,):j8[A -Mo0)4] (A-17)

where

i f0 0 '_(H s ddO - 2 A(M)

do (A-18)

The equations used in the method developed herein for calculating noise
radiation from various airframe components are listed below.

Convective
Airframe Component Directivity Amplification

Wing (A-15),M:O A(M)=1/3

Horizontal Tail (A-15),M=O A(MN)=/3
Vertical Tail (A-16),M-o A(M)=i/3
Trailing Edge Flaps (A-14) (A-14)
Leading Edge Slats, Flaps (A-15),M=O A(M)=l/3
Landing Gear (A-17) (A-18)

Although the mathematical conventions used in this noise predictLion

method are the same as those of the NASA ANOPP method for airframe noise,
there is a difference in viewpoint. The NASA ANOPP method starts with a
basic acoustic quantity (sound power) for a basic well-understood noise
source (lift dipole). Calculated sound power is distributed in space by use
of readily available solutions for directivity in compressible flow.
Empiricism occurs only in choosing equation2; for maximum amplitude and
normalized spectrum shape. In contrast, calculation of noise from landing
gears and trailing-edge flaps as given herein has started with empirical
correlations of acoustic-pressure spectra. These spectra were integrated
numerically to obtain maximum OASPL. Directivity functions had to be assumed
arbitrarily, and compressibility effects caused by relative motion had to be
picked, to allow back-calculation of acoustic power. That is, acoustic
power is the most fundamenital quantity from an analytical viewpoint but the
least readily obtained quantity from empirical correlations of airframe noise
data.

10')
 10.0 APPENDIX B: COMPFT•h PROGRAM FOR CALCULATI-NG AIRFRAME NOISE

10.1 General Description

This digital computer program, written in FORTRAN4 IV, predicts airframe

noise that would be measured by microphones mounted on posts or tri.rods abnve
the ground, in lines parallel to the flight path, at input-designated azimuth
and sideline angles. If the atmospheric properties are not specified as
input but allowed to remain at their default values, and the input integer
UNITS is' specified as 1, the aircraft altitude and linear dimensions should
have the dimensions of feet, areas square feet, and airspeed knots. If UNITS
is not equal to 1, and the atmospheric properties are not specified, airspeed
should be input as feet per second. Input values of atmospheric properties
for use with input dimensions and airspeed in the metric system (meters,
square meters, meters per second) are given in the first set of comments
within the program ]isting. Comment statements are placed throughout the
program listing to descr:.be the purpose of each portion of the program and to
define the program variables.

The program consists of a main control program, four subroutines which

calculate noise radiated from seven airframe components, a subroutine NOYS
"which calculates perceived noise level (PNL) of the combined noise spectrum,
and an output subroutine OIUTP which organizes the printout. Subroutine
OFT123 calculates trailing edge noise from the wing, horizontal tail, and
vertical tail, respectively. Subroutine OPT4 calculates trailing edge flap
noise and subroutine OPT5 caiculate- leading edge slat noise. Subroutine
OPT67 calculates landing gear noise from the main landing gear and nose land-
ing gear, respectively. Each of tiese last four subroutines follows the
same flow path. Normalized acoustic power is computed first. This quantity
is multiplied by the appropriate directivity factor, and by the ratio of
reference area to far-field distance squared, to obtain the ratio of overall
mean square acoustic pressure to reference pressure squared. Finally, the one-
third octave spectrum for that noise component is computed. The main program
calculate:; the resulting OASPL and one-third octave band SPL for each conipo-
nent. It also adds the acoustic pressure ratios to obtain the sum of noise
from all designated components. These airframe acoustic preýsure ratios are
utilized to calculate complete airframe OASPL, one-third octave band SPL,
A-weighted noise level dB(A), and perceived noise level (PL) at each desigi~a-
tedI position.

The following, -s a lisit and definition of the input variables and their
liefaul,. values. Wherever po:ss-ible, the programi synmbol:; arnd default values are
equal to those of the NASA ANOPP airframe noise digital computer prog~ram
JCI(1AR[,'M. However, numerical default, values, correspond to use of EngIi sh-sys -
tem unit:i rather than metric-system unitf.

IIU
..... lut
 Program
Symbol Definition Default Value

U CA Atmospheric speed of sound 1116.44

RHOA Atmospheric mass density 0.002377
* NUA Atmospheric kinematic viscosity 1.576E-4
PREF Acoustic reference pressure 4.1773E-7
FL Lower limit for one-third octave 50.
center frequency, Hz
FU Upper limit for one-third octave 10000.
center frequency, Hz
THL Lower limit for polar angle, deg 10.
* THU Upper limit for polar angle, deg 170.
DELTH Increment between successive polar 10.
angles, deg
PHIL Lower limit for azimuth (sideline) -80.
angle, deg
PHITT Upper limit for azimuth angle, deg 80.
DELPHI Increment between successive azimuth 10.
angles, deg
H Altitude 3.281
UNITS Equal to I if velocity is in knots and 0
any other integer if ft/sec (integer)
V Velocity 100.
AW Wing area 10.765
BW Wing span 3.281
AT Horizontal tail area 10.765
BT Horizontal tail span 3.281
AV Vertical tail area 10.765
BV Vertial tail ,;pan 3.281
NF Number of trailing edge flap chordwise 2
segments (integer)
AF Trailing edge flap area 1.615
CF Trailing cdge flap total chord 0.82
DELF Trailing Pdge flap aft-segment deflection, 30.
deg
NS1 Number oj main landing gear units 2
(integer)
NS2 Number of nose landing gear unitst 1
(integer)
ND Equal to 0. if wing is aerodynamically 0.
very clean or . for typical low subsonic
speed aircraft, and for extendod leading
edge slats or flap,

_.10
 Program
Symbol Definition Default Value

N1 Number of main landing gear wheels per 2

unit (integer)
N2 Number of nose landing gear wheels per 1
unit (integer)
D1 Diameter of main landing gear wheels 1.
D2 Diameter of nose landing gear 1.

LI Ratio of main landing gear strut length 2.94

to wheel diameter
12 Ratio of nose landing gear strut length 2.94
to wheel diameter

All input lengths (altitude, span, chord, diameter) and areas must have the
dimensions ft and ft 2 if the first four quantities on this list are kept in
3
English units (rt/"ec, slugs/ft , ft 2 /sec, and lb/ft 2 ). The default values
correspond to flight in sea-level standard atmosphere. lnput quantities can
be given in metric units (m and m2 , with velocity m/sec) if these first four
quantities are supplied as metric-system input. For sea-level standard
atmosphere, these are CAý340.3, RHOA-1.225, NUA=l.464E-5, and the metric-
system acoustic. reference pressure PREFý2.E-5.

This program has been run on a UNIVAC 1110 large digital computer but
should operate on all machines of generally similar capability. Each
execution and output cycle (one aircraft configuration and flight condition,
for tb.e full range of polar angles at one sideline angle) requires only
several tens of seconds of central processing unit time.

10.2 Program Input Format

The input for each casie comprises (1) a title card which provides an
identification label in the printed output, (2) the calculation frequency
range, angular positions, flight condition, and.1 airframe geometry, (3) the
specific program options (noise components) to be used, and (4) an indication
that all required program options have been called, so that noise of the
complete confityuration can be determined. Data input i s done using the
NAMELIST format. A sample input is given on the next. page for calculating
airframe flyover noise of the Lockheed Jet3tar aircraft with trailing edge
flaps extended and landing gear retracted (NASA Flight Research Center Flight.
7A). Because the vertical tail does not radiate nois,,e directly beneath the
aircraft, subroutine OPT3 and its input quantities for vertical tail noise are
not. used. Each line denotes a separate input card. The range of polar angle
TIt has been decreased from the default range bUt. 100 increment.:s have been

nmm08
 retained. Specifying zero sideline angle for both PHIL and PHIU causes
calculations to be held to the flyover line. Default values for frequency
range and atmosphere properties are retained. UNITS is input as 1 so that
velocity can be input in Ikots.

JETSTAR, FLAPS DOWN, GEAR UP, RUN 7A

$INPUT

THL=30., THU=90., PHIL=O., PHIU=O.,

AW=542.5, BW=53.67, AT=149., BT=24.75, ND=O.,

AF=53.67, CF=2.2C, DELF=50.,

H=500., UNITS=l, V=170., IOPT=I,

$END

IOPT=2

$END

$INPUT

ToPT-r

$END

$INPUT

IEND=i

$END

This sequence of card:; can be followed by another title card and

calculation case or by a termination card. *11 of the above cards start. in
column 2. Output for this case is given after the program listing.

imlm10n
 * K -j 4z01-11t IO.- 1

-t~ I-4c 0'

LD 0 9 QN a 34
M IZ
"~ c - .J l c - icl 0 A&
* -V h.
c. TL~ z >. -J -1, Ci 0 - cLi e- 0L
0 -xL LJ a. "K A..ýL 4K Z lý -I-

*4 .. -.. E I-- ::j~ o Cr -C:

I -- > 1-- 0 La Cý I- .. 4- '4d- .i

r- L 0i (Ni ft
4 6- -J0.. 1.- .ýL.j Z L~jaOI--i Q.0 LAu C. 4-4ý- r-
4 tn 1- LA o >.>- C C) %A -I J G Z V1 $-- -. 1 1-
-
* ZiJ_j-1Qcc 4: ZLA W Li CL U. Z- Z-N L...JQ _j V.
* ý 41 LAfl
L- e ki 0i LZ " , -. L- C.;~LA *-
* - L. - fl -4 C -4 9 .- 4L.j 0 0f J-j C)Q 0O3 'D- f*
4 LiALW . - I.- Lý. I .3- ".I.- -)U #-- ,iJL-j -L4 - 7I--Oý

L* C 4 a)43 - t L JI. I'- L11

Zf. z4 C3 _j_ C t.kJ
-* *-L-4 3i~ "-- M-U*-~~ t- - ý t- "16 4
Li
L 'DVl< -4 L 4cm
U "-.z0-- _j Lo XI "i. .L
* *; k rn .4 Q LJ C
JQ0,-" LAJ L~J I-- . ~
1.1 LLt.LJ -- l z .J Z4~
A c I-~/f UZ 0

* 2 Z It -: -4
01-- CAL.j L.ji/
V, M- 4b 0~.
a i490
* LJ¾i J~3- l . "4 3 4Q -* U- I L.AQ..i-i -ZC C* xp
1
* 3-D 6n 61
:D .( . v X- CL0 w)i.~
CL 14 ILA zoo,:

( 1. 0 0+d~ 1'.
- U-/ -IA .. ,V 2 .
1-- ~
U-J~;,JC0C) LI L, -* *411
n l uJ-At Z.-4L v Uoi- Oi: in 11, Ll; t X Z -L -j L- ..- '. ** *4 * eci

LA-ZUi LJ-L :1 L JA~

01>L., C-) Ca -H -(,0
.- jL Vb-C)
Z -~ '- .4f L P-- , -l -j .- kJQ~.4..w '
Q-LAiU-I C . . 1 1L< L. "0CJli. .4 "a . -E . - 0 1-1 JA 46A -6 M 0

*,
cn-JL --. -1j2 F- Ct 2 .d' 3-U I-- e - - 4 1('J r- C: ti-X'
(1 W * LJ b-
- #--*- ,U~ C3LAi - -) *-t -Q n a, AA. Z 5C - V
.--
O C Ir-i ~-A U -L , p-iZ I Li
L c-i,LA ýL - l _ -1 .A -z * ;a we.ial
I--j i-0J :L I,- CL (A. C4 0~. 0 0.4- 0i- 0 01 JO

-,.- LJ -I Ll :- - r- .j4) 0~
*l _ 'J
L~ Z.J *i C2Li .,- L- 2-( :a ; . w \ k J. "4 * 04d1J L ' Z..J.
*a r ~'- C Z1 C 44 C v. 7.t.- .. j4-K
C-L..L i( - iZ aa. .
MCL

4 llu -1 L:,2 CA: - P:6 -l r:

D --. :ý J-- 4. 0..L~... '- .", Z- X- VL > LA os.
L ,_ _j -f 1. L: e1 (Ný Ci Z I- i, L - ,, .. 7L * -#- -uI J IA. JUt- U I -A)
-,) 1,i1f~ (1Z 4 L -4 CLI Lý b ( -- i.iC 2' r - A U. ~LM
I< L.' If 0.4"i VWD L, OI

C"~~~ ~ v, I32 IW Z L 48.v.--J(L

L. *
k% Z. Q/.TI.i-
-:~ilL. P--
! OI Z-1 *t4UO~. IP-04L
*j 0 I CC>- 0 - L .0 V) -- I- V)w W - .- ,..a -T -4 fL~)
C 1 *a
-- 4 - 0L'
4 ~~ ~ ~ 4( ' -7L b-4L.
-4 rýI A- Z V)L..C
0.5 I" '
fl./< - - 4 OA. _L)- J*i J -% -Z 0 2. jC L 4
L'D- l' LII Q..4t .UC: c I..W
CU -o ~ kAJU.~ r- j~%a .J
0 M4 -i U~ Z(L .Ij2 .61l 4 4P4q6 l- .1
aJJi-..L I04&00.JD fllm
Jba OCZL. -EL b.-c-ir 0 T. U r fiLDU" - 3.0a a a L* 2
* o -i04 :3
:;.~ 1:rhX ' D- AJ 4.4
X 1.L .L1 P.-. LL .a. qx I.. li . ft a *)
T-( I6.1 d d'L -- LJ' C: LJU) -4Il U ý
2.L- AZ* "40 0C
V) b-ut V I.- I d 0 L=V 4- -,j1 I (-j y.Z XIa,.Z j
1 z11 > tJ. 4 - Z I JU u 4&lL0 LLT .41 '.3t -1 x .1
6 P.- -
L- j on L~hJ
(0bluCX4. Z1Lij Z- LA- A-. -C z4w 11j rW

Or2*4C-ja MII< 20- >AlL) 1. X0 23.40 LZ L a 4. 1 r-or. If

1 * 4- T 0.3- 4A
J -' L.)i-) 'L4 .1 Z)s -1 .L.
W L1
* L d 'D jV 04 11a
-UI,
aE -44-

***n 4*4Z* "* 0* at* 4

1
P-44ut~-J) C
L._ .-'21 V1 ýilf .i-3 Z 0-. ý4- 13'1 L)3 3 3J1,' 0 z l-.-

o~H
-C 11 L -4oMm2'-'~:j
. a)P
_rc~ > I yI VL' "

110W
 * *D

44)

* *J
* U4

*Lai

z~ co

$I
40 4

4 *
jp 6 49

Ift 4 V) X4L

Z( 0;*1w4YV ) *

04 *-a 44
*0 z Li. *W o

I: ý vODt.@Ilt* ZO 0 4wo

z
o
1
'-4-
4C41c
ililiII
o' n 0 o t)
14
D
to
*
*I -
-D 0.-
-
-
b
M
,
0 4
i Z4 A
0
of&41
1104

9'-4 0o *-I
044X-04
c:Qo~~~*
L4.W~~~~~ ~ w LlZnDO
~ O4DODD -M O a * Wet O40ý u L.I * 1)6 ýU IO~l .&
-
at
-, LL.
.C316I11
I U-4-* "0* Jr

4~i 4n * WW~
II
Z7 rT t-xL, c- , fo's r--CUE0- E -4-,. m- 1.- :j-)I .D r+0- . U - 0r(
L- x 4: a-j
,I',C W..WIA 9J.

c*1,c ;:)~a o- 00 010 o~ 440 .40 C00 U. 00 010 0

 q *r

* LL)

* 16.5*
3c*

C-

Ii. -j. rM
ju:
co1

0 0- *

(Y** t *o
Lsi~ *c 00-

0
Wi: Li 4 d
.J * ;N 0

.~~*0 0
40 0--

efl*. 0. 9 4w 1.0 * Lo

0
c; U 0 . 1 OD *uD%
x n~ C0 0 0xc cr x.*-
4 4*0 X- I I- CL 0ý ~ D~ ~..cZ
0 *3 -C 4 4 j CI
0- 0 ) C 0 z * J -40.
W -
IX*0
j r0*0P-ao
dD w o ww -ao 0co0
0-4 0 wID* 4P 0 Lo *0-fv) x-4 X
-~C Lo B.-
0-0 0
-
0L
~ 0 ~ * Ili -C CL I- P- z=
00 _0 &, .A .0J .0 * ZZ.J 1.0

2 - - Z~Z*.J
.J 0- 04- t'-~ ..- - zN - 11- 0. z6-4. Z I 0e

,*- 0 *. 4 X#4 -0- M.-III - iLL0

I Z
(k L0 ~*
w40 s-Q!0-0 0--0B #,Al0-.0&..- M AL.Z) X0

I A*Uz04I4 U - 4 I
oci; ~ 0*
obOn e.
0
MV-
4A o

a~a..
.- ~--.e .4.. . 1--
.6. - - ,.4 -4 -4.-- -, -- -~--
-- -4

112
 MbI

I"ic wz CIO

* 0.

oa I C

-r x
* *o.

-4 - -4 j

11-_4 a

* 1J7 C0 0 M.~*
* 40 *i q j. L
*A . -0- 944

w ýo
,- a. Ilal
6 a..- 7
;~Q
i.-
II
-*11 6-l1- 0- 00 0 0-4

0::
x M a.
U.1 t `,-*"

-- onmQ. a.NIDy y P
4T-
L4J&A CI ED QD 0 Zo: 110 & C n

:.0 -41-4 Ia-Ia- DO~b .31 z-

IC %.0 oo IL0 !0 N AC
0* *- I.ZI
C3-~S 4I 0 +1-P -C 0,I)o 0o fi Ii
0--
*040*L&..I& 6ý * * - N. 0Il -A
zjb.
LDzl.- Z* 1 -4
bill n o 1 -4 I I s.
Z-Ji L.Jj U* (Y r4 IN PQI-
4 z

0**- 00 G 0 0-~2

f,~ r It 31Ln Do~

r-0.~a~' a.o' ~ N P- -1J1Z F- CCK
9I
z*~~~~~~~~ .- *-h
l D 'I- Y111'
,1
IL~r
1
41 0 4~0
O\c, 0j r, 04
j-- ir,, NI 08ý
N0
l
IIr,. ari I4

*27 ~888
lii J..gBCODY1-
'a
-z -q.-- '-~ 4 113.U .. IL. fA4N rj'
 IN AJ ~ N Sl

kw"LOO LOO*

41- ft . J 0 M m
&A 0 *-II &jwX 0 0
coD owI- -2 'nn co
CL 0 -j : z L

6.4 ow zw LDa&A

n- -4iO
"o CL 1.-
Zw 2 IO
"-oh w 0 *z

cr z mmo &.Cm 0 9- C
* Cr co 2 0 OOV) ~X
zwýC CL o-

w 1- w max
-t. xxt- j J z
DL OL& M MD L C

CL a 0UJ-,* 1-e-CXZ ZZLd. 22. I.

Wa 6-40--0 0-MO 0 -ad)
1CX 0O.91- iAJ LJ FJJ m X

IL CrQ0 Dz co conw I-Z Id

z 0c
O& 4Kmg w a ="
En
0- mf m Mx -X X-.1
V n~b - 0-64"o 00no"Lov

.40~E ...J .4..) j -j4.) i

0 5 m m X u-49)J4
Z Lflj. 0x

l&.X X -z

-K -3 t AD r -0 4- -A - 9 - -9 t9 .06 "0
Z. Cr 3E9-x,-b Xxxxb zwxxvz I.- I-O . 2j
.

(z rK tv(zcrCOXk ".JzZ CC
Z 4Z 0: (XrC IC r(Xcc r cZccr
0.100 000 OC00000 0000000OO 0w-a' 0 ).o0- a

0 C 04 pv r- O , "f *~AD r- C0i 0

N 114
 *0

*~L $ e n
01 &f:lNtt

In * 0 0 z 41
on-
0 lM.4 1
In N N_ ?n 0 cc .0cI
*N in 0n - 0~ ql 0f$ N~ ON N -
in w %eo C)o -_ I IL Px N
_j 00 0 V'DC; I &
-2:. No N 04 mC0 x
04 0 1.4 * 0 * .1 go *. *o M - 4 - frý .i-

P- N- P- f- - -

_5. -ft -4n r -na

NZ 0, N GoI
-o r40 X~$

ozIN NP_4 N 0 *&N IL^ .

*IA 0 ft) 0.. . IC WN m
_.jZ 9*n LAD~ ** I 9C4 lr do -4 C~ *) -0
%A W*
0

."-T -IbuZ 41* c) ain

v -0 1(o0~ _jC C)
41 _ .0a 0m 4 a+

: * E Pj I0

V..Jt'j a0 :u *.0Qv .'On C *4 00 a - P-CC3 C3 -.. "-ED

~
0.0!' l * r oi 17 It ~-3 . ý XD 0

z1P ]6. N -0% n 0 n-.~ N-4 A in I 1*C r 0 -b 0--C J -N 4 -ýn i .- u

0nC3 .4 11* 1( _j N C3.-Jj ...JNM--41L'ýC -3xI* OCU

in.Po 4-4 C3
-i
4-K~-3 . -. u, -

_q
*r -9: ow. c :~-a', *,a 4.tr 0*'
D" *rLL N*Inqftlla -a 4 . 4. * *,a0 0 * t o0 n 6ta ON ) 0,CLNL L4 O #.n
A u. LA - Z * L 0w
C3N~u 0j 4' ry*
9$ C* C *CJ
0000-4~ *~t 04 ý 9:V)

p4-4~.' ~4iO! ***N~51P~-- . gp ~.-. ~~ a.x~--1CK

1~
e)N N 0 10- '10s seO
m C)C'Og . C-1 Q )
C~

OZN 4 ~bN~i~ .Q~c)0 N-" ~ 10 10 )odDD .'

~K9 -4Es--
$- 094 C-0, . C)0-0
I i 1,0
_J w~ k)k--Z 0,[ZS
~~,-4..J&...JEU- U OIL.~a
C_9,c-LcK
'-U43 r-- '-J.Oh.q0L..
'J( '' L. O ~ L.

-. t' 4 4 .4 .4
 i .. ..Si+..... . ... . .. ..
S: . I / l. l.l . . i. I. . i
• I ] / • 1 • el • . e • •. el • e e
• _* 1 / / * ! * •z *t ' * • • • * *• *
"' "• ! | / se | ek •e[ •l i • * , • 411 * J*l-- ,4t
• * | ! ! * ! * elm- • • * •' * el • I•• e

• * I " 1 / .z / • e!,. • I • • • • • • •.u ++

+e • I • ! / *• / • +11o * I • e • * e • /..= ++
• . •(,•'-o. .! / e-1 ! * 41! • ! I * • •eJ • le 4t
• * I m+< / * • * *•" * ! • * • * +• * I*•u ,
• * I•u-• I / ••, * ,IiU e I ,4 . . • .ll .e le"- 41
• * I •z"r i ! e.•Ql: • el• • i e * , • ei•• I•,,,• *
• . p,• ,,,m I I • W • elC• • • e • . ca-. I*• ++
• * •z'• I I en":l: * ,No • i +1 * "• ,ll•e leo *
• .- * • "UZ J I CLOG+ • ,lln, • 41 . ,Lu• ,Jlu•• /•• el
•.- -* •ll•.-'•. ! I _*"• * ,I *, +I * ''* •,.,* l*z .

• •u . p-..+,= 1 ••o •z+Su * i q..a•, v+• ,n ++•. *o

• H e •,,n•,.. ; I •• •Qet,., * I +fzeNeo• .-+ el . •• e!
, *o * • ,'•z I / • •t •.--+e•- +l / •.,[•.-•u• • +it.'* *',.. •11
• :z e !'+'-"+-" i I ••rZ else,., * ! •e •",.e•* • e•'* ,*.-+ •
.+,., .* •<_.•': I *_o• •0+4'. i I -++"'•"'* * • ++'•* •• e

.•.._-"•* •" "+ I I e',,•l. .•.a:. e I •fl.oe.-.,q. oe .- melz.a. ,.._+

e• . •,,+N=r I | • .•I *÷+I• • • •@ • .. v•. e °•<*o ,*-J *

• .• . I•.+n *-:3 ', [ •t. • • o• • I o+•.-+•*e.,(• •*e! • +.]•.•,+,++

.•no:. 141.,J...J ,. + * °l I •'•' I ,• 0•0.•• I o'r.-*e ••.)*,,•* ,.•v•.o. -. •NZSI

e• t p•,.,• ; A',..1' -- e up •c3el Z• !NO.'-"R•e•..•.Ro:• .o--• *+(•e• •
e•, * [email protected] .-.. p •e•e*Z• *eq*.•.J.I.• e• *..-,*=-• *a0e•.e.-.c•e<z3•
• ...Joe p,•. v• +,.),-p -. • •- • .. • ++k-• • j -'. o-J 4#o: • • e o• ,.-.--• • •* NC•• Z O.
• •.J•• I.#i2. e.-... I el::):3•n•Xl•.*Xe:;31.--• Iu --- e,,n.,-..,--•• *•.P'-*,"••O •,•Oen•
• J * I-.• I ua•z.,.,d) **•o • t*+',+••..*,.•• •)oo*,,,*o,.,.z•.

.•,e *.•q•:•m I o•ue• ..- ,q •• ] -. e*e •-,,.e-J. -- .. l.il.-l.•..•i•.oJel

"+*•J-J*-•t•Q---J• I •el •. • *f"e•,• l*•te..Je•e,•. C •.•"+'--el,.t•-.J• ...l.m

• *-•,*-+"P l u. [•z•o,+z••l.an+• !.•,.+ ,.Roe--.•e ,.,, •o ++zer•ak.,,,•o.

•.oj. -•" • "-,-,•uez.•oeze,•,•. I•ZO*e•,.,Ne *'.. .... ,+ .nd•. •.

::3.•"•0. •'•I• ZUi[ •t• .. el:::)* .dZO*•i.J•. •!.SI ...leVI, I..0 el'.•* =.•* 7•elZ
O* .',,+• C• •"-. OlI •v)•:•Z..O• me. •--.l: •,--+e ,-•i[zeO _t e,•ze..-+,-ke,-+•,n•Q:
°+*'•"•*-J•: sp•'*• o!•***"*•"-J* •llu• **to• *• *.-.+ewe..-* o* ,. e•,+e--•ev+•e•

• • 1 / * 1 • el . +i . . ,m. t ,e * ne ,•
• •e* J | *e 4'** e•*** P •ee +;•. ++e* !***•1

; • +r-')" [{+Tr-I•,
• ; '-} +"')r++.... ... ( +•" +l' ' J' '+ e{'* " ..... I . +" , + ' r

S.•!--+ ... . • • r,•r •jl• , F -'!'•" "-', ,+•L + ( ,'r.r ( • .j+., ,

 *-ZK 2

L* w I LI

az . au. I ftf *-f

*,a ft. *4 onx*f 1. f
ftj ftl- itS ft ftf

ft- 1-)
ot ft. *I ft wA ftA

Ct Lt-aA.
61t cf fl
"0f 01 f

U. ft 4I* 4*
uAf ft-4 ft
'A a
ft ft .4 t- ft4 0. I*#f Z0
u0. Lo* I. " I* ft *f
-y* 0

*0 ft* 40. ;* .)t ft 0 A.

4 0
ftT ft &n*. a. 00 "4t
ADCAft ft&1.1 W.f caf ft 0Ja j W
z4' z, xr .0* ft .1 -j 4- X ft
&t O.Z =.: ) t

ft~Di..j ~ o2 ;a o.af t x - -' qIA ft:

ft&..Uf 3.-) ft0 ft N it1.0 &Az.f
.o

ftI- f _jA .40- 1-.o 0*A ft N0i 0~ Uja Na

ftGa f L'. z I u V4i 4 * 0f KJ 0J ~ f
ftir~f fl* 09 ft - ft -uC~ ftaf I. 0-L
L"" -ý K 0o -j t- CA.0*

ftJ Z) z4 Oft
* ~ ;, 06 0- N 04

or~.I* a0LL -V I-fa -* . - Q9qg. 1--

0-'4 .

co* C.; -Ju ItAf ANP.- j. j

03 04* 9-j ft O-0f A- 0- 1 IA4 U0'

.- '0

ftZ') P"*ft -iCf LD 0o.um :0.-ft*

3142 IA 40-t 0;f 46VI 49f1-1 ft-
fI ft .A "*l f.401.. a) *a I.- 0IA...ft
-af e... f t a -0-81-1"aa zcf
0,).1f
Xf b- (L -CmA
'4ZIf k*1~..'-4 N 9-4 4-0- Loo

ft' OW
If, 0' b.Jf ft a...'
fA I
ft aiV, a-r
Ul~b I94
ft f al flt 1f *0 Ll.O~a ~44 C~4D* .4r
*L
ftifl....*~~~~~~ *-L* 4 .0aA lta- 011
J~L'- -f

ft'0Lf4 pg-.cof 4
 * 41* * * 0.11

* *L * * * 41;) * * 41
*0 z * * * * n* i* I
411j * 41 **f4 k*0
*, -4u *v IL**L~4 *I4
*4LL * *4A1 * * * 4
*K0 U f-4 *d * f9-* 11u
0~* " 1 41* * 6-1*1 *f
&*wz~- *U. * I.* 41
4A- so x0 *r * ** V*0* *
0-4 * a- * * **
*0 * &_jG1 *n*
D1a4 LI
* A to1 0+-u * *,A*k* 14
C(3p*4 ou.IA *L 0 *:* 4Z1
41j'f)* b*4.a* 41D4 e*" A c*'/4 *Cf*
t CAC-441 CA*
4-1( *a-4 0 0 1 -* *
fl41 C4
* 0*
. 0.0- 4Q* *n z 1& * 4 0iJ
414ZI4* ýP4:J. I a 1,-Jy* * co 0
Om*LJ4 mOtj. 1. * 41J 41 * 41 1
*443
XJ- 41o4 0 *Z*
* 4 *-T6 +6~ 41J 41* G*0.d

IIJhi
LOD041 m4 *0* U1*-4 1-
*Z*!JI.4 mzN4- *6i* *1 714 S1-

*" -* , o0,= a ~*AK * * W0 *04

*II*
41Z* .J4
LA 41.1* 41 46Z1U* )

-jj
41 z *h4 0. *z 4r* -4e 1 444
*CIO*~" %A41-c~ 41 41* , V

_j6O4 49 4 1* *6641 c 49 - c
4!!:

~~~~
*iaO41~ (~- - 1 :~ **%a 46-
I44 *41 * *1

iw.i
41'fl~~U 41 * 6 *(. *0* o *4 ,*0 .. *0 .1 -j
_1 la Zi#)C* 64Z ,n31 O.410A w* 044-4

490 -. * In41 4~ *Zj* £44~(~ .- J"

*i * .ij4 da ."4%0
0.h* _j4 641. to1~ P"44 441.4 *
41- o)( 140 6 * *14 4 *.44 J*04 _j*-41

to.wj 0n 61. I* xN *44 *41% '--4411

%.1 - * *.69
6fl'L*' ... j& Cc * 41u*
LZo. ~.00*41
$-,CA
.-. . oiIi' 41 lwLJ *&4
CIJ LA *X*Z4 l X .. Z 6.f)It

0 1,-Z41 %.411 I-' 1

cr"*,l-41 -4 O O 0-40
. 4.1 I 41 $*44 1

04 ot 6 01 .0 -j14 664 66.0 -a -"z*

tL041J
4A I mo1 4 It 66
oil A *1.1 410 66 I6 41 j* L"J4
(JD* -Z*4 66c Il*. ** Cf 41(or a.~ z14Z

&A* P.-441,Qr I6OI ux ) 400. 49 c -0.ol 41A b-D*

46-416-4* V 4LDJ 6n A

Q% I U6' QLWu-1- UL)t

.-. 'J(J-.'3 ~L~-

'.j k~ ~O'- ~i J~.s~' ,j Cf.0- LY-) ( 1'' 1-7 'Lir- W O.UC ' )'
~~~~~~~~~~
.4'- iýrr (r-ý~~~J \trJ(~r '' ý( I.r~ ý ., ' 17 77 - 1
L., L,'L)Lj L.~
C--1K L I Li r-) k~L L4 -3
j L)LA j. L-i U jILJ U L) C (L L-) j r LJW 4 1 L L4~C

BEST
A"i. nOY
 .1p

lot
AAl ca
Li
w 0 QL
*Z U *0 i
w OIL *u
NCOX a:
o I,. -a I *W 00 0 CA
31 -Z 2: "1
It* N ýgm 1 0 cr
T.4* ZULA- w
6j* O.UZ *0
" U 0. W *
_j O.U_ 0: * w
*-Uj :3 cc
OAA V.4 U w to 0
("m4a in -U. un LA
fZ OU P" uj 4 * w
cr 0
ZN" IL
dox:w

u.-J* zood.-1
04L x

ýqo 0
'D u
04ý- O*U u *002.* U)I 'VI
0*

0 49 * Oj at 0 * Uat 0-*
wu* ýw +_ z CA 'm -W-4 4
#.A MA& LLJ D
"tA* X *4.A 0 0 *wk
0 C446-4 *uj u 10- ry
*Z C) .40
C4 oo,-, 0*0
ix -J* 0
Ax 0. * coz Pn * * w 4AC 4A I*
UJIX * I .:) w Pn *U 4CE .7 * W) it * I *
Oý4* '. .:a wm -*" n pn*:) Ln z -4 kD .4 *
Uj* -KNý* * * x cy **-K *- 0 4x_.ý
-0 LIAD* X e4A- N*m V) * * L) _JN CA co*w I
N* z .ýBzx N* Ln
00 &MI CA *C N-*
0-4kn , Z * L'i -9 O*W N. A -M w
Z-K* I'-,A 0, *3D W DA +
rv -a x:
.j 1*0 0
C03 _JL&- Pn V) 0* Ur#
c4w* Li-i ". u _j 0
WJ* M" I U

Ln Cl. It 00* In TL Cr
Z j ki "U- V) u CL uj uj
D) M49 . * 0 0*0 *0 - u V e-n
rN uv)*N Q. -- K -4 OD *U 0- ** ki dx * ->" L!) *
04n 0 7 L-) Lb *4c * r3 4A *.w oo* _* tA * ww 14 N.
_j D Z --r I... * * 0 * * (%j3* w z 1 6) LL. OCD * t,4
U-4* " &_ a- MR 0- 0*0 r-4 0*0 . týj * -. x
c XNM
Not U-mg 00 Z W LJCt "4uj C-) '4
:Z.' 'W*4.Jl
"Wh m 0. -A L.. *1- 0 ',,j _jo -0 r+j -a-*
ZO.*Z >W a LO u qK J* -a 1% u) 0* x ,
_% - * I x * 2:" -*_j ; * 0: M a: Lfiý
wo- L&j cr x 0 ol X M _j 0 0 00* C)
0 U. zr (V X: x 1- LD 0ý.o pl* U)
04,A* ým L)ZU w wzýo*or *K) t-- 'T *(k -0 ZZ*Lý-- &A* ct a
* (XCD*-4 "Co *04 1 Lx: LAJ6 -4 * 0 *on-H 7- .9 * 0 ZCL " * " " "Ci a 0: a, Lo)l- * CLCL
* CID=)* _J co .. ?-. 0 CNat OL.Z ev * * z Wn 10 -2:z ol- * 14- -Lj CJ6* wz * Do
* Z)46A *9 0 U3 OD 0 * I- w *
*T *%ýf eYll-;*#l) *#.A ini =1 1A. L6 1: 4
xt. I I -1 11 *o-e Lo o 0-- -41 11 CL 0 * 0
Ln-3 _j 1 14 ti 461411 0: vlu *-t of 11Zo L!)* L.Ja * I.-LD Lv*
ml$
0-4(1 *-a 114 to 49 - Z LD 4.0 * cr or LO LO cr cx m 4 - V)* ZX.b UAO
XZ*w "ir gy c -7LA. ou u * _j4D 0, Qý-a V)* kA4 Lno. L._J*OL)QrLA. Ck*wO.*OCror*
"U OCr 4- 3C JK LýML Ia. OACDQO 012, ou. *00. cjý*
P,

1 0
e4
u uuu u 0 t-) L) W OU4.00 Li u 4-4 kj

'7 10 -0 f U_ 0' _7 L. LX,0 L*- I I ýY ý ,u r., k.L:,i 1 .7 Lf) C)r- 'A

CIL C"CIL-) Li I IC, L, ---- (" (,(,(, ,
C) ýý'LJL. I L-) i jlr
L- L, L_ "Lý L, L.A- L4- C, Lit-)t-) :y Y'Z i"
L.)C-)t
 40*

*r*
F ox 4116

Ii. 40 -aCP

-
cyV - * "0Z *
ll ) 0 * # 0 0

4m LP 0
4,0 .fýK *.00

*CD* -0
&jj (a 4f U* 4*
30 - *&M . ID
4" "40 X *A -0*ý
ef> *), *jD * NOW~0
z 0 *%* *&40

O V 0* *jj * N .4

ca : kJ * 2 1,.. j0 -- )~v zi

U~S- L/) zz0. La *

0% ~ -J0 -J 31*~
'-4 -C2- 44. (L -9
c a* It*.r- 0-f

.i*P -04*r- nc t =- )
00 i *0 * Nr.D -t4 0-4 . j ion

vzI.* CAL*
Lnl
co *1 (L a
* 00 i114A CO
4 -in
.b s0 ~--40 Q
4*l WIN ' y-
-
**45.5.5. ME~
t. I 0 Z_ -4 X ~- VgoO u*D =

'4ZA it0**.J- NN~b I. 4-4 IX* b4

A& d* *" go*C * ý .. -. 4 &.J- "OLo! - #'-6.

LJ-4Z* Xfl D*4 ka* J*Z Z Z-0 04C ZZ 1 D Z
UJI) -- *0 * i )fl J 0-4LA. -M -- -- . 0 . . - In

a.oy it 04 .. N

c1. -*'qI rZC.*4 mJ *0*r 00-

O ~ f)
O~ia ~111* - C--~
Cl C_)-r-4); ~ O *-
i
l
* :
a
- 0.r
~K O 6
c *0ON1
C C:*- .*.0 * -. 4

In~ -~"J*0C.~ LA O40P. ef120P r. io * c4 u-e-5 0

 V) w
-JA o
w

u0-

LI a

W-0

fill i

Li

a. . La

+- 0 Lo
'- - z z
a- Co-a 0
I'~~ s. ~9&1 ~

ifLi

ýw ; 4 "0 F--B
j*
z-4% W0
0 O-it-4r.

m LLiU o 0-

J.-Q 4A 3117 0 -. 1

az0- I; L 1.4 0Q

* ~ ~**N 0
*
0a. No,*
BL~
'D t~LD#4c 0
0 tg 4 " WMOF a . c% ;
0 a,: . %L
rIý. '%.

-- 14~ 44- I----P-

c w p-P-cO0~
I Si. &A..Li0-c 10 0 OOZ
AP-V-1-11111XIA -4'A i k6
01:1 0 (3)0 00
C 0 10 OD t~ CD'*

r jý-j l,- tr C* j ); j, r c

4r t r ý '-
S~COPY

10.4 Sample Pr.ogram Output

Program output for the sample input given in the preceding section is
shown here. The output table for each of the three airframe noise components
is preceded by a table of all the NAMELIST variables and their numerical
values, not reproduced here.

JETSIARIFLAPS DOWN,RUN7A

CLEAN k,ING NOISE

AZIMUTH ANGLE, PHI r. DEGREES

TABLE OF VALUFS OF SOUND PPESSURE LEVEL (DECIBELS) FOR

CONSTAI*.T ALTITUDE FLIGHT

1/2 OL DIRECTIVITY ANGLE (DEGREES)

CTR FPFQ:
(HEPTi) c
32. 4. 5P. 6b. 70. 80. 90.

5f.05C39.78 42.18 43.92 4 .17 46.LC 46.43 4b.4ý

63.00 42.8P 45.24 46.93 46.13 4b.Yr 49.26 49.22
bO.0O0 3.35.77 48.947 49.70 s5.93 51.32 51.851 51.86
412O.00 546.10 50.31 55.52 56.37 53.57 53.77 53.56
12s,0c. 51.73 5,.72 w3.71 54.67 55.1F 55.26 54397
51.64
63f160.00 53.75 53.39 53.26 56347 56.40 5529.
""GO.0O 524.19 S.62 55.93 56,72 572.2 56.90 56,03
250.00
315-00 53.C54
53.03 47.65
S14.94 56.52
56.10 56.92
56 *69 57.79
56 .63 56.94
56e54 56.33
55.8r,
4bC0.00
506.00 43.90
51.76 45.59
5_3.57 46.0
c4.61 56.83
55.08 56.C9 55.73 54.97
55.CP 54.67 53.86
639.00 50.63 52 .40 3 . 39 53 . P2 53 .77 53. 32 52.47
RLo90o 49.19 50.93 FI.F9 52.28 52.iC 51.71 50.83
1O60.nO 47,65 49.36 50.3C 50.67 50.57 50.05 49e15
125ý1.00 4r,.95 47.65 146.57 4P,92 4,8.79 48,26 47,3S
16LO.OO 43.99C 45,59 146*50 46.83 46.7r 146,,15 45o22
200C.00 41.92 43.60 44.49 44,P2 44.07 44.12 43.18
2500.00 39o79 41.46 42.35 42.67 42.ol 41.95 4j1.01
3150.00 37.40 39.07 39.96 40.27 40.11 39.54 38.60
4O66.66 344.70 36.36 37.25 37.r5 37.39 36.82 35.87
5Q e00 3 1 .f3 3 3 .S4 34o42 34.73 34,56 33.99 33.04
6300.00 28.55 30. ? !1.o09 31.39 31.22 30.65 29.69
800O.CO 24.49 26.15 27.03 27.33 27.16 26.59 25.63
10000.00 19.91 2-1.57 22.o45 22.75 22.5F 22.00 21.05
OVERALL 62.27 b4.21 65.42 66.r9 66.32 66.14 65.58

122
 JETSTAR,FLAPS DOWN,RUNA

HORIZCNTAL. TAIL NOISE

AZIMUTH AN'GLE, PHI-- . GpGPF5

TABLE OF VALUES OF SOUND PPESSURE LEVEL (DECIBELS) FOR

CONSTANT ALTITUDE FLIGHT

li3 0b DIRECTIVITY ANGLE (DFGPFES)

CTR FREQ:
(HERTfl 30. '4[.'5. 60. 7 nO. 9C

50.06 2E.50 30.95 32.7' 34.07 34.9P 35.50 35.bFl

(,3.00 31.99 34.41 36.18 37.47 38.34 38.A1 38.88
EF .O0 35.339 37.78 39.51 40.75 41.57 41.98 441.99
1[,0.00 38.32 40.67 42.35 43.54 44 . 35r 44.65 44.59
105.90 43.2b
4O.9b 44.89 46.00 46.bo9 46.96 46.82
1600.0 43.47 45.70 47.25 48.28 48.F6 49.04 48.80
200.00 45.29 47.46 48.92 49.P6 5C.36 50.43 50.09
250.00 46.64 48.74 5C.12 50.97 51.36 51.34 50.9fn
315.00 47.53 49.56 50.85 51.60 51.69 51.77 51.23
400.0C 47.90 49.96 51.07 51.73 51.92 51.70 51.07
5C0,00 47.79 49.67 50.82 51."0 51.51 51.21 50.51
630.00 47.23 49.07 5C.15 50.66 c,0.71 5''. 34 4 9.58
800.00 46.27 48.07 49.1I 49.c5 49.54 49.12 48.31
100C.00 45.08 46.84 47.83 48.25 48.20 47.73 46o88
1250.06 43.65 45.38 46.34 46.72 46.6L 46.15 45.27
1600.0 L 41.82 43.S3 44.47 44.83 44.72 44.20 43.30
2000.00 39.98 41.67 42.59 42.93 42.61 42.27 41.36
25i00.00 37.95 39.A4 40.54 40.P7 40C.74 40.19 39.26
31507.00 35.6S 37.33 38.22 38.-4 38.39 37.84 3b.9 n
40C0.00 33.00 34.67 35.56 3ý.F8 35.72 35.16 34.Z2
5OU0.00 30.22 31.89 32.77 33.08 32 .92 32.36 31.41
6300.00 26.91 28.58 29.46 29.77 29.61 29.04 28.09
8000.00 22.88 24.54 25.42 25.73 25.56 24.99 24.04
10000.00 16.32 19.98 20.86 21.16 20.99 20.42 19.46
OVERALL 57.11 59.C5 60.26 60.03 61.16 60.98 60.42

123
 JE TSTARFLAPS DOWNRUN7 A

TRAILING EDGE FLAP NOISE

AZIMUTH ANGLE, PHI C,. DEGPEES

TABLE OF VALUES OF SOUND PPESSURF LEVEL (EICIBELS) FOR

CCNSTANT ALTITUDE FLIGHT

1/3 OB DIRECTIVITY ANGLE ICEGPFES)

CTP FREQ:
(HERTZ) 30. 40. 50. 60. 70. 80. 90.

50.00 62.42 t4.59 6S.82 66.29 6b.L9 65.23 63.64

63.00 63,42 45v60 f6.82 67.29 67L.L9 b66a?3 64.64
80.00 64.446 66.63 67.86 6.e.33 66.13 67.27 65.68
100.00 65.43 67et,, 66.83 69.30 69.10 68.24 66.6F
125.0G 66.40 b8.c7 69.R0 70.?7 70.L7 b9.21 67.62
160.00 67.47 69.64 70.87 71.34 71.14 70.28 b8.6Q
20000 6P.29 70.61 71.84 72.31 72.11 71.25 69.6f-
250OO 67.71 70.11 71.60 72.37 72.49 71.95 70.63
315.00 67.11 69.51 71.00 71.77 71.t:9 71.35 70.0F
4CC.CO 6ý.49 68.P9 70.3R 71.15 71.27 70,73 69.46
500.O0 6c.SO t8.3u 69.80 70.57 76.69 70.15 68.88
630.00 65.30 67.70 69A19 69°o6 70.08 69.55 68.27
8..0.00 64.6P 67.0 68.57 69o!4 69.46 68.92 67.65
1000.00 64.10 66.50 67.99 68.76 68.88 68.34 67.07
1250.00 63.52 65.92 67.41 68.18 68.30 67.76 66.49
1600.00 62.87 65.27 66.77 67.54 67.65 67.12 65.85
2000.00 62.07 64.69 66,18 66.95 67.G7 66.54 65.26
2500.00 59.16 61.90 63.eO 65.o02 65.61 65.56 64.68
3150.OU 56.15 58.49 60.79 62.00 62.60 62.54 61.75
40L)0.00 53.04 55.78 57.67 58.F9 59,49 59.43 58.64
50O0.00 50.13 52.87 54.77 5.96 56.58 56.52 55.73
6300.00 47.12 49.86 r,1.75 52.97 53,57 53.51 52.72
800.O00 44.01 46.75 48.64 49. A6 5C.45 50 . 0 4 49.61
10000.00 '41.10 43.84 45.73 46.95 "7.55 47.49 46.70
OVERALL 78.35 80.66 P2.05 82.71 82.71 82.05 80.66

-i,

J 1?
 SU4MMA PY

TAB'LE OF VALUES OF SOUNL' PPESSURF LEVLL 4DFLIrtELS) FOP

CONSTANT ALTITUDE FLTGI4T

1/30Of, IRECTIVITY ANCLE iOEGPEES)

CTR FREQ:
(HE RTZ) 3 . 40.& f-
C,~ 6 C,. 70. 80U. 9.
1.*25 .00 .2 00 .rU .0 0L)
.0o 0c
1 .6G .0c 11 .0 .00 Ono u .Ou or0
2.0cc .C
.0
0~ .o .2 .00 .0
2 .5G, .0 c~ c00 ,00 0 . .0 .0c
.0
'.15 .~]l .0 ., L[ 0 .00
447w no Cro .c .00
540 L 00 .00 0L0C0 .00 .00
6.0.00 .O0 n 5.L
.r0 0 .00 .00
6.30 0 U .000 .00 r.0 .00' .00 .0o
1fl
8*5 CcO
.0' .0
co e.G00
40 .000
10.00 *GO . 0 .010 CIO L:0 .00 .00
1..r0 co .u .rcjao .n Ln . 3 . r
;ýc0.0 .0.0o .r . L ? r

31 .5C0 Cc0 * 0 CD .n0 Or

.0 000 .00
40.C0 .00 .OL .00 r.10 $0 00 000 G0r
50.00 62.4S 64.62 65.85 66.32 66.13 65.29 63.73
63.00 63.46 65.64 6b.67 67.!5 67.16 b6.33 64.78
80.00 64.E'? 66.70 67.9! 68*41 68.23 67.41 65.87
--
100.00 65w. 5? 67.69 6b.92
'&
69.41 69 .23 68.41 6b6.b6P
125.00 66.51 6A.-6 8 69.91 7n.4 0 70.23 b9 .41 67.88
160.00 67.60 69.77 71.S0 71.48 71.'71 7C.49 68.96
200.00 68.43 70.74 71.97 72.45 72 .27 71.a44 69.91
250.00 67.89 70.*27 71.76 72.53 72 .65 72. 13 70.83
"315.*00 67.32 b9.70 71 .18 71 P4 72 .6L 711.54 70. 30
400.00 66. 72 69 9 '9 7 C,5 7 71.e33 71 e 45 70 ,9 2 69.67
500.00 66.13 68.51 69.9A 7n.74 70.FE 70ý32 6'9.07
6370.00 65.C51 67.89 69.36 70.12 7L..2 69.70 68*44~
8 00.00 64.86 b7.9"4 6b.71 69.47 69.58 69.05 67.79
1000.00 64.25 t6.63 68.10 68.86 68.98 68*44 67.18
1250.00 63.64 66.02 67.50 68.26 668.38 67.84 6f.v5 7
1600.00 62.96 65.35 66.83 67.60 67.71 67.11 65.91
20LC.00 62.14 64.75 66.213 67.00 67.11 66.58 65.31
2500.00 59.24 b1.97 63.85 65.C6 65.64 65.59 64.711
31S0.00 56.?5 58.*96 ý 0 e4 205 62 . C' 62.58 61. 78
40G0.00 53.14 55seb 57,74 SP*94 59,53 59.47 58.67
50L0.00 5C.24 52.95 54.83 56.04 56.C2 56.56 55.77
6.500.00 47.22 49,Q4 51.82 53.02 '>3.61 53.55 52.75
8000.00 44.CQ Ub.41 '4S.69 49.90 50.4q 50.43 49.63

*10000.00 41s16 43.P8 45.77 46.98 47.57 47.51 46.72

2 4
OVERALL 76,4P8 0*7? 82.17 F2.P3 8 .6 62.20 80.83
CP (A) 73,91 76. 13 77.83 7F .62 78. 7P 7F8.29 77.90 P
PtN r)F 85.(,7 88.23 F 9.79 99l.63 QQ0,.94 90 . r9 89,47
P'
 11.0 APPENDIX III: TABULATED AIRFRAME FLYOVER NOISE SPECTRA

Flyover noise one-third octave spectra, at the measurement time that

produced maximum OASPL, were plotted in Refs. 11, 17, and 26 for tests con-
ducted at NASA Dryden Flight Research Center. The aircraft tested were the
Aero Commander Shrike, Lockheed JetStar, Convair 990, and Boeing 7),7. Values
of OASPL, flitght speed (m/sec), and mass (Newtons) were given in '±.ule 2 of
Ref, 22 for ten flights of the Lockheed JetStar, four flights of the Convair
990, and five flights of the Boeing 747. The aircraft configurations comprise
the clean aircraft, trailing edge flaps extended to one deflection for each
aircraft, landing gear extended and flaps retracted (for the Lockheed JetStar
and Boeing 747), and both the landing gear and flaps extended. Spectra were
[ supplied for the Aero Commander Shrike at two airspeeds with landing gear
extended and flaps retracted, and at one airspeed with landing gear and
trailing edge flaps extended.

These data were measured using a microphone installed flush with a large
ground board. The resulting one-third octave band sound pressure levels (SPL)
should be regarded as 6 dB above free field because of in-phase reflection of
acoustic waves at the microphone. In contrast, noise certification measure-
ments generally are obtained with microphones mounted 4 ft above a hard sur-
face. The resulting measured sum of directly radiated and randomly-.phased
reflected acoustic waves is 3 d b above free field at greater than several
hundred Hz center frequencies.

Measured flyover noise spectra were corrected by NASA for differences

between measured and 500 ft flight altitude. Two adju.stments were applied.
Amplitude at all frequencies was adjusted for the altitude difference by use
of an inverse squared variation with altitude. The spectra were further
corrected for the atmospheric attenuation over a path erength equal to the
difference between 500 ft and the actual altitude. This frequency-dependent
correction was culculateC for the attenuation properties of standard--atmo-
sphere temperature and humidity. Therefore the tabulated spectra correspond
'o the direction for peak OASPL of a flyover at 500 ft altitude. They include
the effects of atmospheric attenuation between that altitude and the ground.
Both OASPL and preceived noise level (PNL) of the resulting adjusted spectra
were generally calculated by NASA. These spectra and _ntegrated levels for
the four aircraft types are listed in Tables 3 throuvh 6 herein. Velocity
(knots) and grcss weight (1b) were taken from NASA-supplied data sheets for
most runs, For other run,-,, they were calculatea fr( m Table 2 of Ref. P"2
which provides flight information in metric-sy-stem units.

---- 1 26 -
 Spectra for the Vickers VC 10 in the clean configuration and with several
individual components deflected were obtained by the Royal Aircraft
Establishment (RAE) and reported in Ref.12. These flyover peak-OASFI spectra
also were measured with a flush-mounted microphone and are 6 dB above free
field. All of the eight spectra supplied by RAE were obtained at 160 knots
nominal airspeed and 600 ft nominal altitude. These spectra, which include
whatever atmospheric attenuation was present, are given in TaOle 7. Values of
OASPL for these spectra were given in Ref. 12 only for the frequency range
from 40 to 1600 Hz because engine tone noise dominated the spectra at higher
frequencies. Those values are not tabulated herein.

Airframe noise spectra were measured by Lockheed-California Co. for the

Douglas DC-3, Convair 240, Aero Commander Shrike, Prue-2, and Cessna 150.
These measurements used microphones mounted on tripods 4 ft above the ground.
labulated spectra are readily available from Table A-I of Ref. 9 and are not
repeated herein.

These spectra are being published to facilitate the evaluation of other

methods for predicting airframe noise. Geometric properties are given in
Table 8 for the five airframes for which spectra are listed in Tables 3
through 7.

12'(
 TABLEl 3
AERO COMANIDER SHRIKE, 500 FT ALTI'IJDE

1/3 Run 4, gear down, Run 6, gear down, Run 7, gear down,
Octave flaps down, 113 kt, flaps up, 112 kt, flaps up, 153 kt,
freq., iz 6566 ib, UPL, dD 6475 ib, SPL, d1i 644o ib, SPLd
d13
50 63
63 62
66 68.9
66.5 70.
80 61
100 60 69.9
65.5 69.4
125 59.5 63.5 69.4
160 59
200 61 66.4
59.5 60.')
250
i: 59.5
66.9
315 59.5 58.-
58 664.9
400 60 53 66.4
500 60.5 57.5 65.8
630 5-8.5 6•.
800 (,4 59 66.6
1000 59.5 586.
1250 57.5 56 64.)
1600 54 52
2000 60.,'
2000 52"ý,
53.5 6o., i
2500 51.5 60.4
50 48.)
"3150 58.0
46
4ooo 43.5
4654.54.
5000 40. 4 4.
6300 34 44.4
8000 34.3
10000 3
A4.P1, 72.3
7.
7.

12!
 TABLE 4

LOCKIIEID JI•"i'AR , 500 FT ALT'i'I-IDE

Run 3, Run 4 Run 4-a, km 5,

Run 2,
clean, clean, clean,
1/3 clean, cleaij,
297 kt, 342 kt, 350 kt, 358 kt,
Octave 247 kt,
34,986 lb 33,966 ib 33,336 lb
freq., 1Iz. 36,156 lb 35,586 lb

63.8 65.0 63.8

50 58.0 61.1
66.5 65.8 63.4
63 65.3 64.1
72.0 69.9 71.1
80 66.9 69.3
73.2 71.9 72.7
100 69.2 70.9
74.3 72.9 73.5
125 70.6 71.7
75.5 75.2 74.9
160 68.5 71.5
77. )J 75.2 T) . 4
200 69.5 72.3
77.b 7o•.7 76.1
250 21.8 73.5
78.7 77.2 77.2
315 72.3 74i.6
7W.2 75.3 '0.3
4o00 70.6 72.,",
76.4 74.9 '(0.1
500 69.7 71.,1"
78.2 7_.3 79.0
630 72.1 714.6
'8.6 7(. 1 (0.2
800 73.0 76.2
81.1 82.3
60.7
1000 '(5.0 78.2
8>0.) 2.3 (2.9
1250 75.3 7V.
82 .*
7T .6 'i.( 80.5
1600
714.3 (5.2 78.3 79.5 7o.8
2000
76.7 '0..I '(9.5
22500 71.9 74.0
73.0 74.7 P(5.
"3150 95.6 70.7
(( .[ 71.3 72.9
I4000 h1. [( ';6.6
t5.7 67.7 01).')
?. 4
5000 58.
2 . •_j 650,o.)
.6
8- WOO ',14. ;4 61 0
' ,A.'.'( ,,1,.,
2,.
"/000 "6300"1 p2.3

14.I,1 .; (3.>
9)0. 1.3"
1.0t
[ 10000
OA; 1'1, •14
58..)
. 17.•r/ NJ.)2

_N]L 5,0 17. . 101.< 11.

1219
 TABLE 4 (Concluded)

LOCKHIIEED JETSI'AR , 500 FT ALTIDITDE

Run 7A, Run 8, Run 9, Run 10, Run ii,

flaps 500, flaps up, flaps up, flaps 50°, flaps up,
gear up, gear down, gear down, gear down, gear down,
1/3 Octave 170 kt, 194 kt, 182 kt, 158 kt, 204 kt,
freq., lz 31,539, lb 31,011 lb 30,386 ib 29,686 ib 29,486 ib

50 71.6 68.8 65.7 70.8 66.1

63 72.6 69.7 6(8. 5 72.5 67.8
80 75.0 72.6 71.5 72.3 70.7
100 76.4 714.4 74.1 75.0 72.8
125 76. 4 73.8 75.9 74.6 73.5
160 73.5 75.9 77.1 73.0 75.4
200 73.6 73.7 77.7 74.7 70.3
250 74. ) 75.0 7. 8 74.6 76.1
315 75.0 75.5 7Lu. 75.3 75.7
400 71.7 74.J4 75.0 73.0 74.7
500 71.( 72.2 75.2 72.o (2.5
630 75.6 75.7 77.5 72.0 70.1
3 76.) (8.o 75.0 77.14
1000 7,. 1 (8.4 T0.8
77.8 79.1
1250 7(.9 77.8 80.5 (0.2 78.8
I1O0 6.6 (8.04 80. in.2 7. 7
2000 73.7 76.2 77.0 73.4 77.4
2500 78.7
(0.6 (3.71 68.7
6.
31"A) 5'0.3(4 08* .1I0
40oo0 •I.5 02.7 5 .8 (6. ,. 0
WO0 O. t.,. )M.( ,0*. )
6 $()t 8 .2 25.1 7 ',2. 55. * 5(.6
8000 - - - ,1 . .2
10000 - - - 714.;1;4.
8-,( ,
 TABLE 5

CONVAIR 990, 500 FT ALTIJTDE

Run 11, R ui O0,JI

Run 3, Run 17, Flaps 360, Flaps 360,

Clean, Clean, gear up, gear down,
312 kt 187 kt, 189 1t, 160 kt,
i/3 Octave
freq, Hz 182000 .b 157,000 lb 166,UOO lb 168.6OO lb

68.5 68.1 77.7 76. i

50
63 69.1 66.5 y77.6 '76.
80 74.1 70.3 79.7 79.3
77.2 74.7 82.2 81.0
100
125 80.1 7 .o 82.7 81.7
160 80.0 73.9 8;- .2 81.2
200 79.4 71.5 81.8 80.2
79.3 71.6 80.7 80.9
250
315 79.4 '(.3 81.5 8o. 9
400 78.2 72.3 83.3 82.)
79.2 73.0 84.4 82.4
500
6,,• 79.4 l74.q 84.6 81.6
80.2 74. o 82.5 79.9
800
o000 80.5 7(3.7 ,.3 79. 3
1250 81.3 1(.u 8o.1i 79.0
1600 875 .7 Ho.' '(7.8
2000 82.6 73.1 '('7.7 (6.1
2500 81.8 72.9 75.1
3150 81.1 6s.d 72.0 70.0
4000 79.8 6 7.o 68.6 (7.8
79. 63.5 65.2 (4I.0
5000
6300 77.3 60.1, 61.0 00.0
8000 75.2 5(,. 3 58.2 58.'5
10000 (1.i 51.3 0(,. 1. 3
9.,).19 9(..1
OAS PL
P 10. 5 ,.02.210.

•LjI
 TABLE 6

BOEING 747, 500 FT ALTIT'UDE

Run 2 Run 3, Run 6,

Flaps 250, Flaps 250, Run 4, Run 5, Flaps up,
gear down, gear up, clean, clean, gear down,

1/3 Octave 197 kt 198 kt, 26(4 kb, 233 kt, 215 kt,
51000 lb 507,000 lb 5022000
9 b 501,000 lb 497,000 lb
freqi z

50 85.5 81.)1 Y5.1 7.1 87. 3

8-.4 85. '77.0 75.3 82.2
63
80 88.1 88.0 8]..l 79.0 83.0
100 90.5 90.6 80.5 77.8 84,4
125 88.2 88.0 80.7 78.2 85.8
160 83.1 87.1 80.3 83.3 88.1
86.9 79.4 85.2 88.9
200 88.8
250 88.8 8'.2 79.0 78.5 87.8
89.8 87.2 79.59 79.2 88.7
315
4oo 89.1 87.3 80.3 '('(.9 87.7'
89.4 87.6 81.9 78.1 88.3
500
630 89.8 87.2 83.8 79.0 89.9
800 9W.5 86.0 83.6 83.6 90.9
1000 90.1 87.1 85.,5 83.8 90.5
89.5 86.o 86.5 80.8 90.0
1250
1600 88.6 87. o 82.7 80.9 89.8
2000 88.!; 86.6 84.1 81.8 88.8
88.)j 80.4 81.1 80.2 88.1
-2500
3150 85.8 84.1 81. 5 78.7 85.7
I1000 83.7 82.0 80.4 '7.3 841.2
5000 80.8 'o -.6 78.2 {6.u 81.2
6300o (7.2 76.5 o(3.0
(01.0 '5.9
8000 71.4 71.o3 - 6y.)1 6().2
10000 - 59.1 - G,3.n 5().8
101.8 100.0 95.3 93 .'7( L. 3
()
OASPL
S1p.1030425 11-.9

13
 JAJLE

VICKERS VC 10 FLYOVER SPECTRA

160 knots airspeed. 600 ft altitude, 195,000 1b

1/3 Octave Freq. Clean 2PL, Dirty SPL, Gear Down, Gear Down,
Hz dB3 dB Doors ,Inui1 Doors Open

40 70.0 84.5 7i.•5 84.o

50 71.5 83.0 75.0 82.0
63 70.C 83.0 7 1 .5 80.5
80 71.0 82,. 0 75.0 81. o
100 71•.5 83 .0 77.0 82.0
1- r72 .5 8".5 77.0 83.5
16o 72.5 83.)o 77.5 83.0
200 73.5 83.5 80.0 82. o
250 7).0 84.o 81.5 82.5
315 7T3.5 82.0 80.0 81.0
hOO 7).0 81.0 77.5 79.0
500 74.5 80.0 76.5 78.5
630 73.5 81.0 77.0 79.0
800 7)1.0 81.5 77.0 79.5
1000 7) 5 79.5 76.5 8o.0
1250 74.0 78.5 76.0 '78.0
16oo 73.0 78.o 7V1.5 76.5
2000 78.0 79.0 75.5 79.5
2500 8u•5 83.0 83.0 M-.5
3150 75.0 81.5 77.0 78.0
4000 70.0 76.5 7W. 5 73.5
5000 68. 5 73.0 70.5 72.5
6300 (8.0 7). 0 70.0 72.5
8000 67.-0 9.5 08.5 70.5
10000 60.5 -8.
5 67.0 68.0

133
 TABLE 7 (Concluded)

VICKERS VC LO FLYOVER SPECTRA

160 knots airspeed, 600 ft altitude, 195,000 lb

1/3 Octave Freq. Slats Out Flaps Extended, Angle =

11z "PL. dB3 ?0O, dBB 145 0 d

40 73.0 72,5 79.3 84.6

50 74-5 74,o 80.o 83.1
63 73.5 74.6 8o. 3 82.1
80 75.5 75.5 8o.6 81.8
100 76.o 77.7 80.5 82.0
77.0 77.3 80.3 S125 81.0
160 77.5 77.8 79.6 80.8
"200 78.5 77.0 79.6 81.o
250 79.5 77.), 79.2 8:. 0
315 78.0 76.1 77.8 80.3
)o00 77.5 76.3 77.7 80.0
500 78.0 76.3 77.2 79.0
630 77.5 76.0 78.3 79.6
800 77.5 77.)h 77.' 80.3
1000 78. f, 76.1 76.5 78.-
1250 76.5 76.1 76.5 76.5
1600 7)1.5 7'.1 75.7 76.),
P000 76.0 75.5 75.0 78.6
S2500
83.5 83.5 81.7 83.5
3150 77.5 77.3 78.5 80.2
II000 72.5 72.5 73.1 75.0
5000 71.5 70.7 70.7 73.8
6,OO 71.5 (o.0 69.3 71.7
R0o0 70.5 67.7 68.2 69.8
10000 70.0 66.8 67.2 68.4

S34
 TABLE 8

AIRFRAME GEOMETRIC PROPERTIES

Airframe Property Aero Commander Lockheed Convair Boeing Vickers

Shrike JetStar 990 747 VC 10

Wing area, ft 2 255. 542.5 987. 5500. 2932.

Wing span, ft 49.0 53.67 95. 195.7 146.2
Horizontal tail
area, ft 2 33.1 149. 185. 1470. 508.
Horizontal tail
slan, ft 16.75 24.75 25.67 72.75 t3.83
Flap area, ft 2 21-2 62.6 337,5 847. 508.
Flap chord, ft 1.3 2.28 4.7 7.5 6,41L
Flap angle, deg 22. 50. 36. 25. 20,35,45
Main gear/aircraft 2 2 2 4 2
Wheels/main gear 1 2 2 )4 4
Main wheel diameter,
ft 2.22 2.17 3.25 3.83- 4.17
Wheels/nose gear 1 2 2 2 2
Nose wheel diameter,
Sft 1.46 1.5 2.42 3.83 3.25

135