C.P. No. 1032

MINISTRY OF TECHNOLOGY

AERONAUTKAL R E S E A R C H COUNCILh I* v , 1. ’

Flight and Tunnel Measurementsof Pressure Fluctuations on the

Upper Surface of the Wingof a Venom Aircraft witha Sharpened Leading-Edge

byR. Rose, M.Sc. and 0. P. Nicholas, &SC. (Eng.)

LONDON: HER MAJESTY’S STATIONERY OFFICE

EIGHT SHILLINGS NET

,

U.D.C. 533.6.071.5 : 533.693 : 533.694.25 : 533.6.054 : 533.6.071

C.P. 1032*November 1967

FLIGHT AND TUNN!ZL MEASUUXENTS OF PRESSURE FLUCTUATIONSON THE UPPER SURFACE OF THE VtING OF A 'LENOM AIRCRAFT

WITH A SHAX'ENED LXADIXG-EDGE

by

R. Rose, &SC.

0. P. Nicholas, B.Sc.(Eng.)

SUMMARY

Part span sharpened leading-edges were attached tcr the wmg to produceregions of separated flow. Flight measurement:. were made of both the spectraand root-mean-square intensity of the pressure fluctuations at orifices at1% and 6% of the local chord. Corresponding measurements on a model in alow speed wnd tunnel covered a ,vider range of uxidence and frequency thanthe flight tests. The flight test results are In good agreement with the tunnelresults. The root-mean-square value of the pressure fluctuations, p, reached amaximum value of s/q = 0.072 for the 6% orifice at an Incidence of IO’. Theincidence for buffet onset was close to that at which the break in the :/q vs. acurves for both orifices occurred.

.

* Replaces R.A.E. Technical Report 67282 - A.R.C. 30135.

CONTENTS

INTRODUCTIONDESCRIPTION OF AII(CRAFl! AND INSTRUMENTATION2.1 The aircraft2.2 InstrumentationCALIBRATION3FLIGHT TESTSANALYSIS OF THE MAGNETIC TAPE RECORDSRESULTS AND DISCUSSION

6.1 Flow visualisation6.2 Icing effects6.3 Power spectra of pressure fluotuations6.4 Root-mean-square pressure fluctuations6.5 BuffetCONCLUSIONS

Table Venom N.F. 3 - principal dimensionsReferencesIllustrationsDetachable abstract cards

=3

3

37

46

6

7a8

91 0II

13

1 3

1516

Figures i-i8

3

1 INTRODUCTION

Aerofoils have become progressively thinner as axcraft speeds haveincreased. !l’h13 has led to an increase in the curvature of the nose SeCtlonof the aerofoils, and Indeed 3ome research aircraft have been deslgned withsharpened leating-edges. The adverse pressure gradient3 associated n?thsuch leaillng-edges will cause the upper surface flow to separate; the flowre-attaches at some downstream point on the aerofoll'. A separation 'bubble'13 thus formed and flow unsteadiness associated with this, produces quitelarge pressure fluctuations on the aerofoll surface. These pressure fluctua-tions may have significant effects on the buffet characteristics of aircraftand the fatigue of wing panels.

Technique3 had been developed to lieasure the total 1ntenslt.y andspectral function of the pressure fluctuations on wind tunnel models* and inflight. Since there was only very meagre inforqatlon on the character ofpressure fluctuations in full-scale flight, a simple flight experiment usinga Venom aircraft with part-span sharpened leading-edges was made. The fllshttests were made in two phases, the first when only the root-mean-square ofthe fluctuating pressures was measured: the second when a time hlstory ofthe pressure fluotuatlons was recorded on a tape recorder. For comparativepurposes , measurements were also made on 2/7 scale model of the Venom at aspeed of 150 feet per second in a wind tunnel at R.A.E. Farnborough. Theresults of these test3 were unpublished, but were made In assoclatlon withoverall force and moment tests 3 .

The work was lnltlated by Aero Flight Dlvlsion, and all the flight testswere made at the Royal Aircraft Establishment, Bedford. However, a large partof the experimental work and lnitlal analysis was done by S.C. Roberts whilstat the College of Aeronautic3 and was used as material for his thesis.

The experimental work was completed in 1960, but, In view of thecurrent interest in separated flow phenomena, publloation is warranted.

2 DESCRIF'TION OF AIRCRAFT AND INSTRUKENTATION

2.1 The aircraft

.The aircraft used for the tests 1s shown 111 Flg.1. It 13 a two-seater

Venom Mk.3, powered by a Ghost Kc.104 engine, developing 4850 lb of thrustat sea level. The prlnolpal dimensions of the alroraft are given in Table I.

i

4

The leeding-edge of the wing, between the engine intake and the boundarylayur fenoe was sharpened by the addition of wooden f&rings attached withAraldlte, as shown in Figs.2 and 3. To preserve synnetry, both wings weremodified. In the first place only a small 12-inch spanwise section betweenAA ad BB of each wing was modified; as flight experience was gained andit was found that the modification did not have too large an adverse effecton the aircraft handling characteristics, the sharpened leading-edge wasgradually extended to the limits of the intake aid the fence. Tne aircraftwas limited to an equivalent airspeed of 250 knots to avoid over-stressingthe Pairing attachment. Fig.2 also shows a section of the modified wing attwo spanwise stations, AA and BB; Fig.3a shows the fairing attached tothe leading edge of the port wing. The type of profile was chosen to ensurethat a separation occurred at low incidence. Pressure orifices, of j/16 inchdiameter, were installed midway between stations AA and BB, at approxi-mately 1% and 6C$ of the chord on the upper surface of the starboard wing.For flow visualisation, tufts were attached to the upper surface of the star-board wing, Fig.jb.

2.2 Instrumentation

The pressures at the two orifices on the starboard wing were measuredby temperature compensated unbonded strain gauge miniature Statham pressuretransducers of range t2.5 lb/in2 and natural frequency 5.5 kc/s. Thesetransducers measure differential pressures and since only the fluctuatingcomponent of the pressure 1s of interest, one side of the transducer wasoowoted directly to the orifice by a short length of $ inch diameter tube,and the other side to a settling chamber, and thence by 1 mm capillary tubingto the orifice. This arrangement 1s shown in Fig.4 and the high lag of thecapillary tube ensures that only the fluctuating component of the pressure issensed by the transducers. The natural frequency of the transducers is wellabove that of the range of the fluctuating pressures being measured, andtheir acceleration sensitivity 1s low.

The output from the transducers was passed through a zero-adJustcircuit to remove any drift which occurred. This drift, whilst not large interms of the full scale range of the transducers, was significant, as thefluctuating pressures being measured were typically only about 1% of therange of the transduoer. The output was then amplified. The frequency res-ponse of the amplifier is shown in Fig.5, and is flat up to 80 c/s, thenfalls rapidly, due to a low pass filter.

The amplified outputs of the transducer3 could either be recorded astime histories using a frequency modulated signal on a magnetlo tape recorderor a3 a root-mean-square (rms) value of the fluctuating pressure. Thedifferent recordings were made in two series of tests.

The magnetic tape recorder used was sn eight channel, $-inch taperecorder, manufactured by the Data Recorder Company. The tape capacity is900 feet, which, at a tape speed of 74 inches per second gives a recordingtime of 24 minutes. The wow and flutter of the tape is less than 0.15% rmsunder laboratory conditions. A constant amplitude and freq&enoy signal wasalso recorded, to check the behaviour of the magnetic tape recorder duringthe high normal accelerations and vlbratlons encountered in the flight tests.A fourth track was used to record the pilot'3 and observer's speech.

The rms value of the fluotuatlng pressure was measured by a vaouo-junction*, which consists of a resistance oontalned wlthln a vacuum chamberand a thermocouple to measure the temperature of the resistance. The outputof the pressure transducer3 was applied, after ampllficatlon, to the resistance,and consequently Its temperature r~as proportional to the mean-square of thefluotuatlng pressure. The vacua-Junctions were shielded from extraneoustemperature effect3 by insulating material. The output3 of the vacua-Junctionswere recorded by a Hussenot A.22 recorder.

Normal acceleration, elevator angle, and incidence measured by avanemounted on a no3e boom of the pitot-stntlc system, were recorded on anotherA.22 recorder.

In addition, the following quantities were recorded on an automaticobserver.-

Indicated airspeed.

Altitude.

Elevator tab angle.

Jet pipe total pressure.

Time.

l Footnote a3 a suitable tape recorder was not immediately available thevacua-Junction tests were made first.(1959) the tsonnique wa7

At the time the test3 were plannedrelatively untried in flight. Some of the limitation3

(see section 6.4) of using a vacua-junction] to measure the rms of a signalcontaining 3pw1ou3 noxe were not fully rpprealated then.

6

3 CALIERATIONS

Steady state calibrations using the two alternative recording systemswere made.

The dynamic response of the transducers and associated pneumaticsystems was obtained by placing them behxd a baffle plate in the wall of awind tunnel, which was run at different wind speeds, and comparing their out-put with previously calibrated transducers. The response of the systems isapproximately flat up to 100 c/s and 300 o/s for the 1% and 6% orificesrespectively. Above these frequencies, the response increases rapidly to aresonant peak at the first fundamental frequency of the plping. The effectof these resonances was attenuated by the low pass filter in the amplifier.This effectively limits the useful spectral analysis of the time historyrecords to 100 c;/s, although the rms measurements may oontrun spuriouseffects due to the resonances in the pneumatic systems at higher frequencies.

The dynamic response of the vaouo-Junctions is sluggish, having a timeoonstant of about IO seconds.

4 FLIGHT TESTS

Using the magnetic tape recorder, time histories of fluctuating pressurewere obtained at two nominally constant speeds at several values of normal

.acoeleration penetrating well beyond the buffet boundary. The reoordingtime was 90 seconds , giving a ler?gth of magnetic tape of about 60 feet, whioh

was the minimum length of tape that could be used on the ground analysisequipment to analyse accurately the results down to a frequency of 2 c/s (seesection 5). Records were also taken with the orifices sealed off, to checkthe vibration sensitivity of the whole system. During these tests the tuftson the starboard wing were photographed by the observer, using a hand-heldcamera.

The rms values of the fluctuating pressures were measured over a some-what tider range of oonditions,at both ICOXI feet and JWJO feet. Fig.6 showsthe flight envelope covered;21 x 106.

the Reynolds numbers quoted are mean values

It was necessary to hold conditions steady for about 30 secondsprior to the recordmgs, to allow the observer to adjust the zero circuits .and give the vaouo-junctions time to stabilise. The reoordings were not soaccurate at the higher values of normal acceleration and more severe buffet .intensity, as the zero adjusting was difficult under these conditions. To

7

extend the range of Reynolds number and Mach number, scme tests were made inlevel flight at various altitudes up to 37000 feet. The aircraft buffetlevels were also determined, based on pilots' assessments.

5 ANALYSIS OF MGNETIC TAPE RECORDS

The method of analysis of the tape records and the equipment used isessentially similar to that used by Owen* and thus only sn outline is givenhere.

The root-mean-square value of the pressure fluctuations ever thespectrum is p, and this is made non-dimensional by dividing it by the dynamicpressure, q (=&pv*).

A non-dimensional frequency parameter is defined as

fLn = -,V

where f = frequency in cycles per second,8 = a representative length, taken as the local chord of the wing at

the measuring station,v = free stream velocity.

A spectral function F(n) is defined, such that F(n) dn is thecontribution to (s/q)* in the frequency range n to n + dn.

Thus

= F(n) dn = n F(n) d (log n) .

The results cf the spectral analyses can therefore be plotted in theform of F(n) against n, or n F(n) against log n; and, in either case,integration over any range of frequency gives the corresponding mean-squareintensity of the fluctuation. In partioular, the total area under eithercurve is equal to the total intensity for

00F(n) dn =

In F(n) d (log n) .

0 0

The spectra measured generally covered a wide range of frequency andthus it is more convenient to present results in the form ofn F(n) vs. log n.

8

The spectra were analysed to give the spectral function F(n) using *Muirhead-Pametrada D-489 analyser which covers the frequency range 20 to2oooo C/S. The low frequency range was extended to include the range 2 to20 c/s by introducing a low frequency modulator. In fact, the records shouldnot contain any significant content above 100 c/s because of the low-passfilter in the amplifier used to augment the signals from the pressure trans-dUCXl-S. The magnetic tape records were formed into loops for the analysis.Experience in the analysis of wind tunnel results2 has shown that at least150 cycles of the lowest frequency to be snalyssd must be contained withinthe loop; inthe present test the minimum length of the loops is 60 feet,which contains about 183 cycles of the lowest frequency that could beanalysed.

6 RJERJLTS AND DISCUSSION

6.1 Flow visualisation

Fig.7 shows a photcgraph obtained frcm the hand held camera of thesurface t$ts on the starboard wind under typical test conditions. Outboardof the boundary layer fence the flow is, as expected, attached. However,inboard. of the fence behind. the sharpened leading-edge, the flow has separatedover a considerable portion of the chord; large regions of reverse flow areshown by the tufts pointing upstream.

This type of bubble separation for aerofoils with sharp leading edgeshad been observed in wind tunnels2 and has since been studied extensively4.The flow re-attaches behind the bubble; the point of re-attachment dependson the Incidence and Reynolds number since these effect the point in thelastinar shear layer where transition occurs. Unsteadiness of the bubble canoccur giving rise to fluctuating pressures on the wing surface.

The behaviour of the surface tufts has been used to define the position

of the re-attachment line. Because of the simplicity of the method the linecannot be defined to better than 5% local chord and at the highest liftcoefficients tested much less accurately, due to the large fluctuations inthe bubble length. Fig.8 shows a typical contour of the separation bubblewhich starts near the intake lip and terminates at the boundary layer fence.The flow in the outboard part of the bubble is approximately two-dimensional,but near the intake three-dimensional effects sre present.

9

Fig.9 shows the variation of bubble length, at the spanwise transducerstation, with aircraft lift ooeffioient for various flight conditions. Thebubble varies in length from 20% chord at the lowest lift coefficienttested, CL = 0.2, to a full chord separation at lift coefficients of theorder 0.6. There may be a systematic variation, for a given CL, of thebubble length with Reynolds number, but in view of the difficulty of definingthe xx-attachment point at high lift coefficients and the relatively smallrange of Reynolds number covered 111 the tests no definite conclusions can bedrawn. Also it should be noted that the lift coefficient is the mean for thewing, which is not necessarily the ssme as the local value for the sharpenedleading-edge section. Fig.9 also shows results deduced from original tuftfilm records taken during tunnel tests3 and these are in reasonable agreementwith the flight results showing that scale effects are small, although theseparation bubble reaches the trailing-edge at a slightly lower incidence inflight,

6.2 Icing effects

On one of the flights the aircraft was climbing slowly through 4000feetat an indioated airspeed of l&.0 knots. The tufts on the starboard wing werepointing forwards showing that the flow on the modified wing section wasfully separated, the buffet level was severe, and a stick force to trim ofapproximately 30 lb pull was required. Suddenly a very large trim changeooourred requiring a push force of 30 lb to trim, the buffet intensity wasreduced considerably, and the tufts indicated a flow re-attachment at about7% chord. When the power was reduced and the aircraft allowed to descendslowly through 4.000 feet the reverse occurred. This process was repeatedseveral times and the behaviour was found to be most consistent.

On the day of this test the icing level was at 4000feet and the weatherwas good unth little cloud. Although no icing was observed a possibleexplanation of the occurrence is the formation of a small amount of ice onthe leading edge, which would increase its radius and thus reduce the sizeof the bubble. Another explanation could be that ice crystals in the sirpromoted an earlier transition than usual in the laminar shear layer, thusreducing the bubble size. This seems less likely sinoe the data in thisReport show that quite large changes in Reynolds number have little effecton the flow.

10

6.3 Power spectra of pressure fluctuations

For the flight results the low pass filter begins to cut off at 80 c/sand thus the flight spectra have not been snalysed beyond about 100 o/s.Hence the more extensive wind tunnel spectra are discussed first flowed by acomparison with flight results.

Fig.10 shows the spectra of the pressure fluctuations at the 1%orifice measured inthe wind tunnel for incidences from 0' to 14'. Asincidence is increased from zero the spectra show a primary peak in n F(n)at gradually reducing values of n. The maximum value of this primary peakis O.OOl7 snd occurs at an lnoidence of 3' when the separation bubble length(Fig.9) 1s about 15% of the local chord. As the Incidence 1s Increased furtherup to 7' the maxImum value of the primary peak decreases and the energy ofthe spectra is reduced. At 8' incidence a secondarv peak develops at aboutn = 0.35 snd the primary peak disappears. At IO', when Fig.9 shows that theseparation bubble is approaching the trailing edge, the secondary peak atabout n = 0.1 increases rapldly and the spectral energy increases rapidly.With further increase of incidence the flow no longer re-attaches to the wingsurface, and this causes a reduction xx the secondary peak until at anincidence of 14' the spectrum is quite flat.

Fig.11 shows spectra for the 6O$ orifice for the tunnel tests froma = 00 to 14O. As incidence is increased from zero a primary peak developswith the same characteristics as noted for the 10% orifice although the peakvalues are much larger. The primary peak reaches a maximum value of about0.0125 for an incidence of II' at about n = 0.5; this primary peak persistsuntil at least 12'. A secondary peak first appears at 8' at about n = 0.15,this secondary peak reaches a maximum at an incidence of IO'. At 14' boththe primary and secondary peaks persist, but are much smaller.

Comparisons of the spectra measured at the iO$ and 6% orifices showcertainsimilarities. As incidence is increased a primary peak develops atgradually reducing values of n. The timurn value of this peak is approxi-mately proportional to the orifice distance from the leading edge and occurswhen the orifice position is about two-thirds of the separation bubble length.As the sepsraration bubble approaches the trailing edge a secondary peakdevelops rapidly, this has the same value for either orifice and occurs ata reduced frequency of just greater than 0.1. As the incidence is increased

11

further and the flow no longer re-attaches to the wing tile energy in thespectra is reduced rapidly.

The power spectra of the pressure fluctuations measured in flight forthe IO,% and 60% orifices are shown in Figs.12 and 13 at various liftcoefficients. As explained previously the spectra have not been analysedat high frequencies and thus only a limited comparison with the tunnel datais possible. For the IQ% orifice the primary peak is cut off, but thesecondary peak develops in a iranner siti1s.r to that indicated by the tunnelresults. In the case of the 60% orifice the spectra measured in flightcover both primary and secondary peaks end the results are similar to thoseof the tunnel tests.

The good agreement between the flight and tunnel results for a sharpleading edge suggests that at least at up to subsonic speeds of the orderM = 0.7 and Reynolds numbers of about 20 x 106, satisfactory scaling lawsfrom low speed wind tunnel tests can be established and the effects ofReynolds number, Mach number, tunnel turbulence and aircraft flexibilityare small.

6.4 Root-mean-square pressure fluctuations

The spectra of the pressure fluctuations have been integrated to obtainthe root-mean-square (rms) level of the pressure fluctuations p, and thesehave been non-dimensionalissd in Figs.14 and 15 for the 1% and 60% orificerespectively.

The flight results have been integrated over the maximum availablerange of the reduoed frequency parameter, n, 0.05 to 1.0. For comparativepurposes the tunnel results have been integrated over the same range as theflight results and over the 'full' range 0.05 to about 30; integration tolower reduced frequency values was not possible due to limitations of theanalysis equipment at low frequencies. The spectral shapes suggest thatterminating the full range integration at an n of about 30 does notexclude any appreciable energy.

The full integration of the tunnel results for the 1% orifice, Fig.14,shows that i/q increases from zero at 0' incidence to a peak of 0.016 at 3'and another peak of 0.037 at II'. These two peaks correspond to the primrryand seoordary peaks noted in the spectra, Fig.10. At 14' incidence i/q hasfallen to 0.005. The integration of the tunnel results over the truncated

12

range n = 0.05 to 1.0 is broadly similar except that the maximum value ofthe first peak is lower due to the out-off at n = 1.0 eliminating the major

oontent of energy of the primary peak. The flight results are only availablefor a small range of incidence, but these are in quite good agreement withthe oomparable tunnel results.

For the 60% orifice the full integration of the tunnel results, Fig.15,shows that i/q increases from sero at a = 0' to e single peak value of0.072 at 11' falling to 0.020 at 14'. The integrated results of the tunnelspeotra over the truncated range n = 0.05 to 1.0 are very similar becauselittle energy exists above n = 1.0. Here the flight results cover most ofthe tunnel incidence range and are in very good agreement with the tunneltests.

It should be noted that the maximum values of ;/q are much larger thanthe normdly accepted value of about O.OO6q 516 for attached turbulentboundary-layer flow.

The rms pressure fluctuations ulere also measured by the vacua-Junctionsand the results expressed in terms of s/q are shown in Flgs.16 and 17 forthe 1% and 60% orifices respectively. These results are compared with theintegrated spectra in Figs.14 and 15. It is apparent that whilst the vacuo-junction and integrated. results show similar trends with incdenoe, thevacua-junction results are much larger. The results from the vacua-Junctionsare almost certainly spurious, and the most likely reason is background noisein the system in flight. The spectral analysis of the tape records waslimited to frequencies where the response was flat, i.e. below 100 c/s andnoise at any frequency aDove this would not have been noticed, even if itwas on the tape. Resonances at 300 o/s and 600 c/s associated with the 1%and 6% orifices respectively may have contributed to the signals, despitethe nominal cut-off of the amplifiers at 100 C/S. The vacua-Junction has a

very high frequency response ad could also have responded to any noise beyondthe upper limit of the tape deck (about 2 kc/s). Thus the vacua-Junctionresult? are regarded as unreliable snd show that great care must be taken whenmeasuring the overall level of pressure fluctuations. If such an automaticintegrating syste'n is used it 1s essential to filter the signals first sothat they are confined to the frequency range of interest, and to make aspectral snalysls to confirm that the filter effectively suppresses anysignals including resonances or noise outsde this frequency range.

13

In spite of the limitations of the vacua-Junction tests they were ofsome use. They showed that a significant pressure fluctuation built-up withincidence and thus it was worthwhile proceeding with tie more complex spectrameasurements.

6.5 Buffet

Pilots' assessments of the aircraft's buffet levels are shown in Fig.18,together with the boundary for appreciable buffet on the original aircraft.For the modified aircraft buffet onset is at a CL of about 0.35 (an 5.5')and varies little mith Mach number. Thus below M = 0.65 the addition of thepart-span sharpened leading-edge roughly halves the CL for appreoiablebuffet, and almost com$etely suppresses the dependence of buffet level uponMach number.

It is interesting to note that, omitting the initial peak et low inoi-dence in the tunnel results for the 1% orifice, the onset of increasing;/q derived from any of the sets of data in Figs.14 or 15 is at an incidencebetnreen 6' and 7'. Thus ;/q vs. a from any of the sets of tunnel or flightdata, ]ncluding the suspect vacua-Junctions, would provide an adequate indioa-tion of buffet onset on the modified eiroraft.

However it should be noted that for the Venom, the wing fundamentalbending frequency is about 8 o/s, and the frequency corresponding to thesecondary peak in the spectrum is at 4.5 to 9.0 c/s for Mach numbers from0.3 to 0.6. Thus the break in the s/q vs. a curves, which corresponds tothe growth of the secondary peak, shows the onset of excitation at near Rstructural frequency in the present tests. Caution should be used in apply-ing tnis result in cases where the aerodynamic and structural frequenciesare not closely matched.

For a similar range of flight conditions the primary speotral peakoccurs in the frequency ro.li;r betbeen 20 end 200 c/s and thus could affectpanel fatigue.

7 CONCLUSIONS

The part span sharpened leading-edge fitted to the Venom wing produceda separation of the flow on the upper surface. This separation re-attachedbefore the trailing edge at low incidences, but the re-attachment pointmoved aft as incidence increased until eventually at about an incidence of10' the flow failed to re-attach to the wing.

1 4

The spectra of' pressure fluctuations at two orifices at 1% and 60%of the local chord within the separated floa agreed well III the flight andlow-speed wind tunnel tests. This suggests that at least for a sharp unsweptleading edge up to a Mach number of 0.7 and a Reynolds number of 20 x 106satisfactory scaling laws can be established from low-speed tunnel tests andthat the effects of Reynolds number, Mach number, tunnel turbulence andaircraft flexibility are small. The characteristic variation of the spectrawith incidence and chord posltion were established, but they are probablyspecific to the particular conflguration tested. Integration of the spectragave root-mean-square values of the pressure fluctuations i/q that reachedmaxImum values of 0.037 and 0.072 for the 10% and 6% orifices respectivelyat an lncldence of 11' as the sepsratlon bubble reached the trailing edge.Results of an attempt to measure root-mean-square pressure fluctuationsdirectly using a vacua-Junction were not reliable, probably due to backgroundnoise in the system in flight. The incidence for buffet onset was close to ,those at which the break in the i/q vs. c curves for both orifices in tunneland flight, lncludlng the suspect vacua-Junctions, occurred.

On one flight considerable ohanges of longitudinal trim and buffet,associated with changes m sxe of the separation bubble, were experiencedas the alrcrsft climbed or descended through the icing level. These effectsmay have been caused by small amounts of ice forming on the sharp leading-edge, although none was observed.

i

.

Table

=OY N.F. 3 - PlUtiUPAL DIIIENSION

B?sArea 279 feet2

Setting relative to fuselage datum 0 degrees

Mhearal 3 degrees

Leading-edge sreep 17.1 degreea

span ld feet

Aerofoil section (unmodified): symmetrical section, madmumthickness/chord 0.1, at 0.4 ohord.

Td.l~l=e (including elevator)

Area

Mean weight during tests

52.3 feet2

11~ lb

i

1 6

plo. Author

1 P.R. GwenL. Klanfer

2 T.B. Owen

REFTRENCES

ntle. etc.

On the laminar boundary layer separation from theleading edge of a thin aerofoil.A.R.C. C.P. 220

(1953)

Techniques of pressure fluctuation measurementsemployed in the R.A.E. low speed mind tunnels.AGARD Report 172,(A.R.C. 20780) (1958)

3 M.B. Guyett Low speed wind tunnel tests on the De HavillandSea Venom with a part-span sharp vmng leading-edgeextension.R.A.E. Technical Note A.STC 2644 (A.R.C. 21823) (1959)

4 Y. Gaster The structure and behaviour of laminar separationbubbles.N.P.L. Aero Report 1181, (A.R.C. 26226) (1966)

5 D.R.3. ?7ebb Surface pressure and structural strains resultingA.R. Keeler from fluctuations in the turbulent boundary layer ofG.R. Allen a Fau-ey Delta 2 axcraft.

A.R.C. C.P. 638 (1962)

6 E.J. Richards Boundary layer noise research in the U.S.A. andM.K. Bull Canada - A crItica review.J.L. WillIs U.S.A. A. Report 131, (1966)

10% Orif ice

60%0rif ice

II

A t SeCtiOn A - A x = I *88in

At section B-B x = I *80 inLocal chord at measuring station = 96 in

Fig. 2 Details of sharpened leading-edge

r surf win ts

.

Surface ofwing

Fig.4 Schematic diagram of transducer installation

3 I-0

iJ io.9z tE ri 0.06-i

0.7

0.6 II 2 4 6 8 IO 20 40 6 0 80 100 200

Frequency - cycles per second

Fig. 5 Amplifier frequency response

.

E

Measuringstation3//ZZ

-- -

t

rFig. 8 Typical contour of the separation bubble

at CL- O-5 in flight

60

100 _

Bubble length% local chord

80

+il X+X0a x 80

+ 0+Tunnai Ref 3

0

Reynolds No2*95xIO~

on local chord t

Flight /Reynolds No ~10’~

v IO.0+ 12.00 14-oX 16-OA IS.0El 20.0

Incidance (deg) I

CL

Fig.9 Separation bubble length at the spanwise transducer stationversus lift coefficient

. 4.

-

-

s

3-

=r--A

?rL3-.---=

-

-

?

<-------

5-

-=.=

-

-

-

=

2--.--r--.s.-_7-

-

-

-_--

=

-

-

-

-

-

-

=--

02

5

12-L- .Oo--

Fig. IO Power spectra of pressure fluctuationsat IO”/0 orifice at various incidrnces, wind tunnel

l2XIO-3

loxlo-3

8 x lo-3

6 x IO-’

nF(n)

4 x lo-3

2 xlo-3

0

--

-

-

-

-

1I7

7

.T

1 IIR e m o l d s N o = 2.2~10~

I0.4 0.6 04 1.0 2.0

on

IOrd

‘d) 1.

f r e q u e n c y n *

Fig. II Power spectra of pressure fluctuationsat 6Oo/o orif ice at various incidences, wind tunnel

-

-

I

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-

-04 0.6 Od I O 20 4 0 6.0 0.0 IO 20 3 0

Reduced frequency n

Fig. 12 Power spectra of pressure fluctuationsat lOo/o orifice at various incidenceqflight

12x16

lOXid

8 Xld

3

.3I

f

1Reynold9 No = 14.4 and 19-l wlOi -

I (based on local chord)

II

5.4 0.34- 5.6 0.35

6.1 0.396.7 0.437-L 0.467,2 0.47

‘\\ 7.8 7.7

\o.sa o.EAl

-4 9-l a-4 0.57 0 53

\\ 9 .4 0.59

9.5 0.59

R.N. xlOc

IS.1 -19.1 - - -19.1 -19-l - - -19.1 -I91 - - -144 -19-l - - -14.4 -191 - - -144 -M-4 - - -144 -14.4 - - -

105 0.63 1 14.4 -

-

-

-

-

-

-004 0.06 0.08 0.1 02 0 4 0.6 081.0 2.0 40 6.0 8.0 I O 20 30

Reduced frequency n l

Fig. 13, Power spectra of pressure fluctuationsat 60/o orifice at various incidences, flight

.

00 Flight results O-05< h 4 I.0Flight results O-05< h 4 I.0

X Tunnel results 0*05* h< I-0X Tunnel results 0*05* h<I-OYY Tunnel results 0.05 <n-C 00Tunnel results 0.05 <n-C 00

0.06 .-Flight results vacua-jenctiol-Flight results vacua-jenctiol

0.080.08

0.06

0.04

0.02

0

0.04 /YY

x xx x

0.02YY

YY 00xx

xx y 0y 0 YaYaYY xx

s6' * 0s6' * 0 L oxL oxOXY **

00 22 4

4

66 88 IOIO Ii!Ii! I4I4Incldenco - degree+

Fig.14 comparison of t from integrotion withvacua - junction results IO O/O orifice

O-08I I

0Y

Flight results 0.054114 I.0Y

X Tunnel results 0.05 <nd 1.0

Tx

/0 I Y Tunnel results On05 4 nL 00

0.06 x -4?

Fhght results vacua-junctioe

,:0

Y

PFig.15 Comporison of q from integration with

vacua - junction results 60’10 orifice

. .

O-06

“,s Maan Rap(based on

0XA

4;alncldence

2.0 3.4 6.2 7.80- 9;+ (dq)O-I O-2 0 3 0.4 0.5 O-6 C, 0.7

Fig .I6 $ versus C L for 10% orif ice-vacua -junction

O-08

0.06

FT

o-04

0.02

o 210 3

olds Wolocal chord)14*0 x IO16.0 x IO6IS.0 x IO' -

Means Reynolds No(based on local chord)

x 16-O x IO6 -

49 6-Z

A 18.0 x IO6

I I lncidancc7:s s;4 (deq)

0.1 o-2 o-3 o-4 05 0.6 C, 0.7

Fig. 17FTJ versus CL for 600/0 orifice- vacua -junction

0.8

0

Present tasts

0 Moderateb Moderate to heavy

10

10-t

Io-3

I0.4

Io-5

IO-6

I I I0.7 0.0 0.9

Mach number

Fig.18 Buffet boundaries of originaland modified aircraft

. . l ,

. .

A.R.C. C-P. 1032NoveUdZr1%7

553.6.071.5 :533.693 :533.69L25 :--. .~ _

Rose, R .Nicholas. O.P.

g3.ym$ :. .

FLIGNTANDluNNELflEIG~ OF PUEWJTS FLXFXUATIOMm THE UPPER SURFACE W TKS WING OF A ‘iIWX4 AIRCRAFTWITII A SHARPENEO lEANNO-EDOE

Rose* Il. 533.6.054 :Nicholas. O.P. 53&m

F L I G H T AND lvNNEL WNIS OF PRFSSURE FUJcRIATIWiONWWPFRSURF~E(rTHEWINC~AA~~~WIT,, A SHllllpENED lSXDlNDED(E

part span sharpened IeadlnS-zdses were attaChed to the Wing to produce pal-t spa” sharpened leadingedges were attached to the wing to produceregions of segarated nm. night q ea~u~ene nts were made of both the reg1a”s o r s e p a r a t e d Il.%. Fl1l#t mewts ware made of both thespectra and root-mean-sq”eJX lntenslty Of the prepSsITe flUCt”atl~S at spectra and rootear,-ScParm lntenslty of the PreSauP flUCtllllti”“S atorlflces at lC% and 6C% 02 the local Cbod. CDnaspDndlnS meaSUlPme”ts crlllces a t 1Og a n d 6c% o f the l o c a l chord. Con-esp@ndi”S mealauementson a model I” a 1s~ speed alnd tun”el covered a rider range of incidence on a model ln a loi speed wind W”“el cover-4 a wider Ia”Se of lncldenceand frequency that the flleht tests . me n i g h t t e s t I-eSults al-3 I” g o o d and IreqWncy that the fllgbt t e s t s . The flight test resllts are I” wxlagreement Wlrn me tu”“~ results. me mtlnea”-square~tiue Of m e amement with m e tu”ne1 results. Tile r-oOt-ax4n-~ v-a&e or m epressure lluctllatlo”S. p, reached a maXlmLn TalUe of p/q = 0.072 loI- pressre nuCtuatim3. 6. reached a marlnmm value 01 iilq = 0 .072 far

(OP=-l tOVd

A.R.C. C.P. 1032Novenber 1967

533.6.071.5 :s3.6g3 :53.694.25 :

A . R . C . C . P . 1 0 3 2 E33.6.07I.5 :NOPember 1967 53.693 :

I R0se.RNh.hol.¶S. O.P. 53.6-W

i

Y

C.P. No. 1032

0 Crown copyrzght 1969Published by

HER M AJESTY’S STATIONERY OPPICB

To be purchased from49 Htgh Holbom, London w c 1

13~ Castle Street, Edmburgh 2109 St Mary Street, Card,ff crl IJW

Brw.e”nose street, Manchester 250 Farfax Street. Br,stol BS1 3DE258 Broad Street, Blrmmgham 1

7 Lmenhall Street, Belfast BTZ gnuor through any bookseller

C.P. No. 1032S.O. CODE No. 23-9018-32