Document
?
.-,
‘,_,Y
COMMITTEE FOR NATIONAL ADVISORY AERONAUTICS —.. ....— .,—.
_..
WAIW’IME Iwwlr”
0R1GINALL% ISSUED Auguet 1944 ae Advance Confidential Report LkH.1 SCALEAND TuRBurJmcE EFFECTSOIV THE LIFT AND DRAG CHARACTERISTICS OF THE HACA 653-k~8, ~ = 1.0 AIRK)IL SECTIOH By John H. Quinn, Jr., and Werren A. Tucker Langley Memorial Aeronautical Laboratory Langley Field, Ta.
NACA WARTIME REPORTS are reprints of papers originally issued to m+ovide ra~id distribution of advance resezrch results to an authorized group requiring th-em for the war effort; They were pre- viously held under a security status but are now unclassified. Some of these reports were not tech- nically edited. All have been reproduced without change in order to expedite general distribution.
L- —... - . ,s , . .
31176 CM)1876706 .- .-
NACA ACR NO. 4Hll
NATIONAL ADVISORY COMMITTEE FOR AERONAUTICS ADVANCE CONFI12ENTIAL %PORT WALE AND TURBULENCE EFFECTS ON THE LIFT AND DRAG CHARACTERISTICS OF THE “NACA 65+8, a = 1.0 AD3FOIL SECTION By John H. Quinn, Jr. and Warren A. Tucker SUMMARY An investigation In two NACA wind tunnels has deter- mined the ef’feet“of .Re ynolds number and stream turbulence on the lift and dra Maracterlstlcs of’a low-drag air- foil, the NACA 653- II-8, . a = 1.0 section, particularly f at low Reynolds numbers, to give an Indication of the mrformunce of low-drag wings in low-scale tests. The results are correlated with nimilar data for the same airfoil section M the NACA two-dimensional low-turbulence pressure tunnel to provide da a over a range of Reynolds k number from 0.19 to 9.0 x 10 .
Large increases in mlnlmum drag coefficient were found as the Reynolds number decreased.
This effect was particularly marked at Reynolds numbers below 1.5 X 106.
At Reynolds numbers below 1.5 x 10 , stream turbulence had llttle effect on the drag characteristics of the NACA 653-418 airfoil section when compared on the basis of test Reynolds number but, at higher Reynolds numbers, stream turbulence hd a detrimental effect on drag.
Large decreases In maximum lift coefficient were found with decreasing Reynolds number; most of this decrease was encountered at Reynolds numbers above 2.0 x log. Marked differences in maxlmnm lift were avparent between the results obtained at high and low turbulence. When compared on the basis of effective Reynolds number, however, fair agreement was reached between the data obtained under both turbulence condi- tions. .
.— —.
2 NACA ACR NO. I&U Considerable variatlm of lift-curve slope with Reynolds number was found. Hesults at low snd hi~h turbulence differed as much as 6 percent but yielded the same value of lift-curve slope at a Reynolds number of’ ap~roximat ly 1;.0 X 106.
At Reynolds numbers higher z than ~.O X lCI , no scale effect on the lift-curve slope was observed over the rcnge tested.
In view of the large vai’~at~.ons in the lift and for the ~AgA 65 .~l~ a~rfo~l drag charactcrist~cs found section ove G range of Reynolds number from O.1~ ~ to 9.0 x 10 , it is thought that the use of low Reynolds number test data relatin~ to low-drag airfoils is unreliable either to estimate full-scale characteristics or to determine the reiatlve ?nerits of airfoil sf3ctiOrlS .
at b.i~her Reynolds numbers.
INTRODUCTION Tnvestlgctians of scale effect on the lift and drag characteri.stlcs of low-drag airfoil se~tions have egu- z larly been made at Reyzmlds nurbers abo’~e3.0 x 10 and at lcw stream turbulence in the NAgA twa-d.imensional low-turbulence pre:sure tunnel (designated TDT).
It iS well known that other investigations of low-drag-airfoil sharacterist:cs are carried olltIn t~nnels with higher tllrbulence levels St lx+wr ~e;molds numbers than the lnvestlgntians in the T?X. Pro~cr interpretation of these flata abtafaed et low Reynolds numbers and at various degrees of stream turbulence is difficult because of the unknown st~esm turbul.:nc? effect snd scale effect at low Vewolds numbers on the characteristics of low-drag airfoila.
Extrapalatian of thes.sdata to hi~her Re.~molds numbers and low turbulence (flight ccndit.ions) js unreliable fop this reasor..
The purpose of the present investlgatlm was to determine the effect of ~eynolds number and stream turbulence on the lift and drag ch~racteristdcs of a low-drag airfoil sect an tkro~h a r&nge of Re;mclds t number below 3.C x 10 . ?todals of the KACA 65z-@8, a=l.() airfoil section having shords of 6 and 2~ incti’s were tested In the NACA tv,o-ilimenslonal low-turbulence tunnel (dnsignqted LTT), which has a stream turbulence of only a NACA .4(JR NO. I@ll ~ fearhundredths of 1 percent . This turbulence is consider- ably below the level at which any change would be notice- able in the critical Reynolds number of a sphere. The tests covere a range of ReynoZds..number frcm 2.77 to 0.23 x 10 . Models of the same section having chords of 12.and 48 inches were tested in the LMAL 7- by 10-foot tunnel (desi nated 7 by 10 tunnel), which has a turbulence The test factor of 1.f as determined from sphere tests.
Reymlds numbers ranged. from 2.99 to 0.19 x 10 .
?1ODELS AND MZTI!KODS (lrdlnates for the NACA 653-418, a = 1.0 airfoil section arc presented In table I. The models having chords Gf 12, 2);, and )}b inches were of wooden construc- tion and were prepared for testin~ by the methods described in rei’crence 1.
The 6-inch-chovd model was butlt of solid aluminum alloy and was polish.ed by band to give an aero- clyna”ilic~lly rmooth surface.
The 24-inch-chord model was tested at tmnel pres- sures of”2, 3, and 4. atmospheres in theflTDT at Reynolds numbers of 2.77, 3.1, 6.1, and 9.0 x lG”. The same model WRS tested at atmospheric yessure in the LTT at ]+:: ;~~l’~ hi : numbers from 0.68 to 2.77 x 10L. !%J8 6-inc~~- C!lord ?;adel was sir~larly tested h the L?’1 f’or a range of l~ynolds number fron 0.23 to G.L6 x 10G and in the TDT for a range from 0.j8 to j+. o ~ 10( In tke TDT and LTT, drag was measured by the wake- SurVcy nethod and Mft was obtained by Integrating the pressures along the floor and ceil.ins of the tunnel test section.
??-ecause the TDT and UTT have test sections of the same size, the tunnel-wall corrections to Mft and drag for each model were the seine in both tunnels.
The tunnel-wall corrections for the 6-fnch-chord model were obtained from the same basic considerations that were used bo determine the corrections f’orthe 2)+-inch-chord model.
In the 7 by 10 tunnel, the models spanned the test section except for a small clearance at each end.
They were ri@dly attached to the balance frame by taque tubes — .—.
\~\ . NACA ACR NO. L&Hll
extending through the tunnel walls. This installation is thought to approximate closely two-dimensional flow and therefore to make It possible to obtain section charactgrlstlcs.
llft characteristics were ~ the 7 by 10 tunnel, obtained from force measurements on the tunnel balance s~stem. Drag characteristics were obtained by the wake- survey method. Lift coefficients have been corrected for effects of tunnel-wall interference by using the experimental correction explained M reference 2. The drag coefficients were corrected for tunnel-wall inter- ference by using the same considerations from which the were corrections obtalnad for the TDT and LTT data.
RESULTS AND DISCUSSION A comparison of lift data obtained in the LTT and TDT at a Reynolds numbsr R of 2.77 X 106 is presented in ~igure 1. Yhe LTT data were ohtaingd at atmospheric rressure and a Mach number of 0.19)+,whereas the TDT data .- were obtained at a tunnel pressure of 1< atmospheres and a ]~achn~ber of 0,1500 The curves are in good agreement both in respect to slope and maximum lift coefficient; it is therefore improbable that any Mach number effect on maximuw lift coefficient, which might have been expected from the results presented in reference 3, exists in the LTT data at this Reynolds number.
Lift data from the LTT and TDT are presented in ftgures 2 to 4 and from the 7 by 10 tunnel, in figure 5.
It ma be noted in figure ~ that tests of the 6-inch-chord and -inch-chord models in the LTT at Reynolds numbers J of 0.66 and 0.68 x 106, respectively, are in good agree- ment.
At values of the lift coefficient above 0.9, a jag in the lift curve (figs. 2 to 4) Is encountered.
This jog is due to a region of laminar separation on the upper surface just downstream of the leading edge.
The jcg becomes more marked as the Reynolds number decreases and, at the lowest Reynolds number, the jog in effect determines maximum lift. It may be seen in figure 5 tit no jog in the lift curve is found in the results from the 7 by 10 tunnel.
— — NACA ACR NO. ~Hll -’ 5 Tne absence of the jog in these curves indicates th~t, at the point on the airfoil where laminar separation occurs In the LTT, the flow is already turbulent In the 7 by 10 tunnel because of the high turbulence level.
A detailed investigation of this separation effect is reported in referenoe ~~.
Drag data are presented in figures 6 and 7. It may be noted in figure 6 that the extent of the low-drag rm~ge Increases progressively as the Reynolds number is decreased.
The high values of the drag coefficients at low Reynolds numbers appear to be connected with a re~ion of laminar separation just downstream of the point of minimum pres- sure.
Little is known of the laws governing the extent and quantitative effect of this local region of separated flow except that both the extent of the region and the dreg increase as the Reynolds number is decreased.
It may be noted in figure 7 that, for the higher test Re<ynolds numbers, minimum drag occurs In the 7 by 10 tunnel at a lift coefficient of about @.55 instead of at the design lift coefficient of 0.4.. llec~use of’tho difficulty of measuring drag by the wake-survey mathod in the 7 by 10 tunnel, drng data were obtained for only a lindtad range of lift coefficient.
Curves that show the scale effect on maxinum l~ft The test results cmfficlent nre Fre~ented in figure 8.
from the ~ by 10 tunnel are plotted against both test and effective WVnolds number. (Effective Reynolds number = Test ~evmolds number x Turbulence factor. ) The LTT and TUT results are plotted against the test Reynolds number which, of course, would be equal to the effective Reynolds number since the stresm turbulence is only a few hundr~{dths of 1 percent. Large decreases in maximum lift coefficient are apparent with decreasing Reynolds number, par lcularly above an effective Reynolds number k of 2.0 x 10 . Figure 8 indicates that,below a Reynolds number of 106, the data from the 7 by 10 tunnel are In fain a~rcement with the data from the TDT end LTT when plotted agal st test Reynolds number.
Above a Reynolds number of 10 the data from the 7 by 10 tunnel are in reemen~ with the data from the TDT and LTT”,when g~:de 3 against effective Reynolds number, It is seen ! hat the rate of Increase in maximum lift coefficlentfs6 reatest at a Reynolds number of approximately 3.0 X 10 .
$ or other low-dre airfoils neither the value of the Reynolds number ai! which th~s rapid increase takes plaoa nor its quantitative effect is known. It Is therefore thought that extra elation of low-scale data or data which do not determine tL s characteristic should be avoided.
—. .— .-. . . — NACA ACR NO. 4H11 Various curves of drag coefficient against Reynolds number are presented In figure 9. The results obtained for the NACA 653-418 section in the LTT and TDT show that for this a~rfoil the drag does not follow the law for the variation of either laminar or tnrbulcnt skin friction tlfI~hnm drag rosi~icient increases omr a flat plate.
progressively as Reymlds numbe”r de~l)anaes; this effect is parti~ularly marked at Reynolds nmbgrs balow 1.5 x 10 .
At Reynolds numbers below 1.5 x 106, LTT and TDT results are in fair a~reement witn results from the 7 by lC t:~.:~-l when ccmpared on the basis Cf test A curve of drag cogf~ictent at thg design lift coef- ficient for the NACA 0012 airfcil section is presented in ~am’esents che fiqure 9 fcr compx.~ison. This CK”7e test results in tb-eLTT. It may be average of’serera-l.
.
the 10W- noted that,~t Rey:lolds aw.bgrs below 1.5 x 106, m lunger shows a lower drag than the con- drag section ventional section.
Scale ef’%ct on lift-curve slope and on the angle of zero lift IS shown in figure 10. Data obtmined in the Li’T at ~e~clds numbers of 0.96 and 1.57 x 12 are not presents,? si.lce s~:f~cient data were not teken to define Althougk. the scale eff~ct on the the slo~~ &~~L~rfite~~.
angle of’zero lift is s:aall, considerable va~ietion of llft-curve slope with Re~olds number is found. In the ~~wolds nznber range from 0.20 to j.O x lG , there is at i’lrst a divergence and then a convergence of the data obtained under ths two turkulonce conditions; the maximum difference between the two curves is approximately 6 per- At Reynolds cent at Reynolds numbersof approximately 10 .
— . ,—.,.--, . - , . .,—. ,— ., , . , NACA ACR NO. 4~1 tll} 7 numbers above ~.O x 106, the slopes appear to be the same under the dlffe’rent turbulence conditions, and there seems to be no further scale effect for the range tested.
At a Reynolds number of approximately 106, it may be cbserved that the variation of’lift-curve slope with Reynolds number becomes small under the high-turbulence It seems reasonable to expect, however, that condition.
the Reynolds number above which the changes in lift- curve slope become unimportant depends considerably on the particular airfoil section and turbulence character- Istl. os of the air stream. The data presented in fig- ure 10 further emphasize the unreliability of using data at low Reynolds numbers to predict full-scale character- istics.
CONCLUDING REMARKS Large increases In minimum drag coefficient were found as the Reynolds number decreased; this effect wag particularly marked at Reynolds numbers below 1.5 X 106.
At Reynolds numbers below 1.5 x 106, stream turbulence had little effect on the drag characteristics of the = 1.0 airfoil section when compared on HACA 653-418, a th~ basis of test Ro-ynolds number but, at higher Reynolds numbers, stream turbulence had a detrimental effect on drag l Large decreases in maximum lift coefficient were founclwith decreasing Reynolds number; most of’this ~ecrease was encountered at Reynolds nunbers above 2.0 x 106.
Marked differences in maximum lift were apparent between the results obtained at high and low turbulence. When compared on the basis of effective however, fair agreement was reached Reynolds numbev, between the data obtained under both turbulence condi- tions.
Considerable variation of ltft-curve slope with Reynolds number was found. Results at low and high turbulence differed by as much as 6 percent but yielded the same value of lift-curve slope at a Reynolds number of approximately 4.0 x 106. At Reynolds nwnbers higher than )+.0 x 106, no scale effeot on the lift-curve slope was observed over the range tested, —.
8 NACA ACR No. 4Hll In VIQW of the la~ge variation In the lift and drag “ characteristics folmd for the NACA 655-~.18 airfoil section ver a range of Reynolds nllmber from 0.19 to ~.o x 10 it ts felt that the use of low Reynolds number test data’~leting to iow-drtag airfoils Is unreliable - either to est~mate full-scale characteristics or to det;rmine the relat~va merits of airfoil sections at higher !?eymol.ds numl?er~.
L:ln&leyNemor!al Aeronautical- Labore.tory Nat! onal Advisory Committee for Aeronautics Langley lZleld,Va.
RFmTmTcm 1. Jac~bs, Eastman E. , Abbott, Ira H. , and Davidson, Ydlton: Preliminary Low-Drag-Airfoil and Flap Data from Tects at L~rge Reynolds Numbers and lhw ~~~’llenc~, and Su,JPle%ent. NACA AC??, Harch lq)J2.
3. Stack, John, Fedziuk, Henry A., and glcary, Harold E.: “ Preliminary Investi Cation of the Effect cf Com- prcssibllity on the Maxinnnr Lift Coefficient.
llACA AG2, Ft?h. 19~.3.
4. von %cnhoff, Albert E. , and Tetervin, Nell: Investi- gation nf’the Variatim of Lift C~eff’icient wfth %ynolds Number at a Moderate Angle of Attack on a Low-Dr~. Airfoil.
NACA CI!,NOV. 19~2.
..— NACA ACR No. fiHll .
TABLZl I @RDfiATES FOR T.HENA~A 653-418, a = 1.0 AIREVIL SECTION ~11 stations and ordlnatss given In percent ohor~ .- -J NATIONAL ADVISQRY CCWMIT’2E13 FOR AERONAUTICS .
NACA ACR NO. L4H11 FIG. 1 a .0 Y Tunnel -1 pressure Maoh 1 .6 — number (ap) o L~ O.1* .
.
.4 .4 .0 mTmMl ~ ~m~ -1 .e -at -16 -0 0 B 16 a Sectionangle of attack, ~, deB Figurel .. Lift characteristics or 21pinoh-ohord model of 19AoA 653-U8, a = l.O airfoilsectionin MACA two-dlmenelonal low-turbulence tunnel (designated LTT) axl MA~ two-dimnmio~l low-turbulence pressure tunnel (designated TIYr).R = 2.77 x 106.
NACA ACR N(3. L4Hll FIG. 2a 1.8 o 0.23 E 106 + .37 x q :22 1.6 1.4 I 1.0 -+ “ \ l\ I \\ 1!
. .8 u / c Iix + : \ +$ c1 G ~,” : .6 + u .
r!
: .4 $ In .2 c -.
-.
NbTIONAL ADVISORY COMMITTEE FOR AERC+MUTISS.
I I
I
-. 6 .
-4 8 12 16 20 Section angle of attack, deg a.s (a) Model having 6-tnch chord in L1l’.
Plgure Z .- Lift characteristicsof the NACA 653-418 airfoil section.
I NACA ACR NO. L4H11 FIG. Zb 1.8 1.6 1.4 0- .8 !
0“ ~ .+ : : .6 + - d .-l ; l k v : .2 -. 2 -. 4 ml- ~ mm m ~1= I I -. 6 0 8 12 16 20 -8 lb -4 Sectiom ongle of mttaok, ~, deg” (b) Modal hmwlog 2&-lndi *oral 10 LTT.
m~ra 2.- Conttouod.
I -. .— —..
FIG. 2C NACA ACR NO. L4H11 1.8 1.6 1.4 1.2 1.0 .
0 .8 j ~ .6 ~ l-l g .4 * .2 -. 2 -. & -. 6 4 e 12 16 20 -4 4 8eotion angle of attaok, ~, dog s (o) undol hating 24-lnoh ohord in ~K munra z .- Oonoludad.
NACA ACR NO. L4H11 1.8 R I 1.6 1.4 1.2 1.0 .8 .6 .L .2 -. 2 -. & NATIONAL ADVISORY COMMITTEE FOR AERONAUTICS.
—— i -. 6 o -k 8 12 16 20 Angle of attack, ao, deg Figure 3 .- Llft characterlatlceof the HACA 655-I.118 airfoil sect~on; 6-f.nch-chord model In ~T.
.- NACA ACR NO. L4H11 1.8 1.6 /--!3.
1.4 1.2 0- . 1.0 b .4 .2 -o 2 1 -. + NATIONAL ADVISORY COMMITTEE FmAIROMAUTICS.
I -.
0 8 12 16 20 -8 -4 4 Angle of attack, deg a.~ =gure 3 .- Concluded.
NACA ACR NO. L4H11 FIG. 4 R Model Tunnel chord (in.)
0:2~ X 106 + 6 LTT ) x .68 LTT D 2.77
1.8 f
0 3.1 A 24 TDT ;:; v I 1.6 1.4 1.2 1.0 u
T
I .8
u---
.6 .4 .2
--1..-
-.
-.
NATIONAL ADVISORY COMMITTEE FIX AERONAUTICS.
I
I -. 6 I -4-o 4 8 12 16 Section angle of attack, deg aO, Figure 4 .- Lift characteristics of the NACA 653-L18 airfoil section through the entire range of Reynolds nmbers.
NACA ACR NO. L4H11 R Model 1.8 -1- 0.19 x 1( .3 .
A 12-inch chord “ .5z D .75 v .75 1.6 1.50 : 48-inch chord D % R M 1.2 1.0 \ .8 .6 .k .2 —_ -. 2 -.
NATIONAL ADVISORY COMMITTEE Fc4 AfROMAUTICS.
< .
I I -. 6 - o 4 12 SeCtiOn ~gle of attack, ao, deg .
Figures .- Lift characteristics of the l?llCA653-418 airfoil sectlan in the LMAL 7- by 10-foot tunnel (designated 7 by 10 tunnel).
—,,-, ,--,, ,,,.,.,,, .,,,, --,-,.,,.,,,,.,-- ..,, -., —, , , ,,,. ,.,--.—- NACA ACR NO. L4H11 FIG. 6 .032 .028 l 024 .CEO C3a .016 .012 G .008 .004 -.
8 -. 4 0 .4 .6 1.2 Section lift coefffclent, c1 r Figure 6.- Drag characteristicsof the NACA 61j3-1418 airfoil section in the LTT and lDT.
.
NACA ACR NO. L4H11 FIG. 7 ,, R’ Mode 1 i- 0.19 x 106 0 l 3 A 12-inch chord •1 ‘ :75 v 975 w b8-inch chord o ::$ Do / + .020 ~+ “ t I + .016 .012 .008 .004 o .2 .6 .8 .4 SectIonltft coefflclent, c1
‘-
W’lgure7 .. Dreg characterlatlcs of the HACA 6?5-418 l irfoil aectlon in the ? by 10 tunnel.
—— .-———————— — z R Tunnel Test . [ o LTT and TDT J.*O Effective + Tby10 I X Tby10 Teat / - , .
z / o 1.4 x 1.2 ~+ 1.0 I .8 NATIONAL ADVLSORY COMINTTEE K3RAERONAUTICS.
I 1 .6 :15 .2 .6 .8 1.0 .4 2.0 6.0 8.0 loxld 4.0 Reynolds number and effectiveReynolds number I Figure 8 .- Scale effect on maxhum lift coefficient of the ‘7 NACA 653-418 a~rfofl SectIon.
H o .
.
.030 .025 .020 Y z o .
.015 \ .Turbulent flat-plate drag coefficient T I I I \ .
.010 liACA 65=-418, 7 by 10 tunnel (test R) < \ \ ‘ .008 \ 1- ‘ .006 .
.
\ I I I . oo~ NAM 655-418, I-ITT and TDT
~’ I
l \ \ \ Laminar flat-plate drag o,oeffl clent .002 h .001 .1 .L L.O 2.0 .4 .6 .8 LO 6.0 8.O1OX1O6 NATIONAL ADVISORY COMMITTEE FOS AERONAUTICS Reynolds number Hgure 9.- Scale effect on drag coefficient at the design lift coefficient CD of the HACA 65 -418 airfoil section.
Model Tunnel z chord (In. ) z o .
-4 -3 -2 l 12 .11 o v _& .10 .2 .6 .8 1.0 2.0 4.0 .4 6.0 8.0 10X106 %1 NATIONAL ADVISORY Reynolds numbe~ H COIINITTEE fORAIROMAUTICS. ‘ Q .
Flgure10 .- Scale effect on lift-curve slope and angle of zero lif; .
P o