reproduction quality notice · recent developments in airfoil-testing methods and fundamental...

Reproduction Quality Notice

This document is part of the Air Technical Index [ATI] collection. The ATI collection is over 50 years old and was imaged from roll film. The collection has deteriorated over time and is in poor condition. DTIC has reproduced the best available copy utilizing the most current imaging technology. ATI documents that are partially legible have been included in the DTIC collection due to their historical value.

If you are dissatisfied with this document, please feel free to contact our Directorate of User Services at [703] 767-9066/9068 or DSN 427-9066/9068.

Do Not Return This Document To DTIC

Reproduced by

AIR DOCUMENTS DIVISION

N l

< L I • *

HEADQUARTERS AIR MATERIEL COMMAND

WRIGHT FIELD. DAYTON, OHO

*wob«, K. M,

-OH",

Asrodynanica (2) Wings and Airfoil* (6)

Airfoils, Laminar flow (08257)I Boundary layer - Calculation (12220)j Airfoil* - Design (08070)

Fralialnary raport on Laalnar-flow alrfolla and naw awthods adopted for airfoil and now methods adopted for airfoil and boundary-laysr investigations

* ACR-L-^5

«—• —r *-»«. te, JM-ta> toUojtai< D> o_

».s.

Mi Zrzvr.-.zls division, T-2 AR'I:, Wr> v.t r;;!d

;,>;;,(,';, .• ;<!e.

R<J-7*7 ~ WiZt

9o *7£ ACS June 1939

NATIONAL ADVISORY COMMITTEE FOR AERONAUTICS

WARTIME REPORT ORIGINALLY ISSUED

June 1939 as Advance Confidential Report

FKELIMIHARy. REPORT OH LAMIHAH-FLOW AIRFOILS AMD

HEW METHODS ADOPTED FOR AIRFOIL AND

BOUHDARY-LAYER INVESTIGATIOHS /

By Eastman II. Jacobs

/ Longley Memorial Aeronautical Laboratory /

Langley Held, 7a.

NACA WASHINGTON

/ NACA WARTIME REPORTS are reprints of papers originally Issued to provide rapid distribution of advance research results to an authorized gruu|i requiring them for the war effort. They were pre- viously lii^ld under a security status Lut are now unclassified. Scrr.e of these reports were not tech- nically i-iiited. All have been roppuuu-NfJ without change in or ier to expeditp peneral distribution.

L- 3»»5

FRELIMINA3Y HEPOHT ON.LAMINAR-FLOW AIRFOILS AND

NEW METHODS ADOPTED FOR AIRFOIL AED

BOUNDARY-LATER INVEST I&ATIONS

By Eastman N. Jacobs

SUMMARY

Recent developments in airfoil-testing methods and fundamental air-flow investigations, as applied to airfoils at the N.A.C.A. laboratory, are discussed. Prelim- inary test results, obtained under conditions relatively frec from stream turbulence and other disturbances, arc presented. Suitable airfoils and airfoil-design principles were developed to taice advantage of the unusually extensive laminar-boundary layers that may be maintained under the improved testing conditions.

For practical consideration, these preliminary results presented are of interest mainly in the lower Reynolds Number range below 6,000,000. Within this Reynolds Number range the new lacinar-flow airfoils and the new airfoil- design principles may be expected to yield drag coefficients on actual wings of a markedly smaller order than those heretofore obtained. For example, drag coefficients as low as 0.0022 and profile L/D values as high as 290 were measured.

IETR0DUCTI0E

During the past s ing conviction that la sible through the use properties of laminar ever, the turbulence p to so hasten transitio the Reynolds Number, t appeared so small that be expected from the 1 layers.

everal years there has been a grow- rge drag reductions should be pos- on actual airfoils of the low-drag boundary layers. In the past, how- resent in most wind tunnels tended n-, in the usual full-scale range of hat the extent of the laminar layer only slight drag roduc'tions could ow-drag properties of the laminar

I/.

"1 More

flight to tlona of s suggested more, test lence tend presence o ler sectio 3) and in ence 4). stream of have vanis Taylor's t conclusion boundary 1 even at mu rlth suita ly the tur In flight

recently, study the mall air-s that trans s made in ed to show f.laminar ns In the the loner The result the N.A.C. hingly sma heoretleal (referenc

ayers and ch larger Die turbul bulence-fr

however, tests occurrence of t tream turoulenc ition might occ tunnels having some drag redu

layers of appre lower Reynolds full-scale rang s of tests (ref A. smoke tuqnel 11 turbulence, considerations

e 6) that more consequently la Reynolds Number ence-free eondi ee atmosphere f

such as those made in ransitlon under condl- e (references 1 and 2) ur much later. Turther- moderately low turbu- ctlon owing to the ciable extent on propel- Number range (reference e for airfoils (refer- erence 5) in the air , which is known to as veil as some of G. I. , led the 'author to the extensive laminar

rger drag reductions s might be possible tions simulating close- reauently encountered

During this period, plans were st low-turbulence i?rge Fiynolds number a Dent. The first step was to eliminate of three-dimensional flows, thus reduc the two-dimensional flow about an airf type of airfoil testing equipment was to as a "two-dimensional flow tunnel.* ods of investigating airfoils extendin tively narrow test section were thus t airfoil section. In order to reduce t a level that its effect on transition ish, variations of the methods employe s.-noke tunnel v-ere coutezrolated.

arted for suitable irfoll testing equip- the complications

ing the problem to oil section. The new therefore referred The proposed meth-

g across a compara- rul;- tests of the he turbulence to such should tend to van- d in the K.A.3.A.

The next step was to verify the proposed methods of airfoil testing. A small model of the new equipment was considered, but in order to obtain conclusive results a tunnel sufficiently large to reach the lower range of flight Reynolds numbers was agreed Upon.

The first and most difficult problem with the new equipment was to reduce the turbulence to the desired level. The usual methods of measurement were not sufficiently sensitive. Recourse was therefore had to the direct comparison of actual transition effects on airfoils as observed in flight, in the new tunnel, and in other tunnels. Such comparisons indicated that the turbulence as affecting transition could be reduced below the level of other tun-

nels and, in moat eases, below tne level Inferred from many of tne flight tests. (Compare, for example, fig. 13 of reference 2. Values of Jones' N exceeding 6,500,000 have been obtained from some of the recent tests of airfoils in the new tunnel.) Such comparisons suggest that transition ras hastened in flight by other disturbing effects. In the tunnel, disturbances such as those due to surface roughness were carefully avoided and vibration effects vrere probably unimportant, at least at the lover air speeds. It remains improbable, nevertheless, that the desired effective zero turbulence (vanishing effect on transition) has yet been attained. The turbulence level vras considered sufficiently low, however, pending more relia- ble comparisons with flight, to justify the airfoil development and the transition work herein reported in preliminary form.

In many nays, the preliminary results o tigations have proved illuminating. It appe der these conditions of vanishing turbulence may be of a different character than in the The lnr.inar-boundary layers ahead of transit curately follow the laminar-boundary layer t pear to be free or nearly free from unsteadi tuations of the Dryden type. Thus the slcin- produced by these laminar lasers at the comp large Reynolds Numbers attainable with the n are no greater than the values predicted by boundary layer theory.

f these inves- ars that, un-

t ransition usual tunnel, ion often ac- heory and apt- ness or flue- friction drags aratively ew equipment the larainar-

The experimental airfoil investigations covered in a preliminary form in this report, moreover, are believed to be the first showing large drag reductions practically realizable through the design of airfoil sections to ben- efit from very extensive laminar-boundary layers. When airfoils are so designed that laminar separation is avoided, and particularly when falling pressures in the downstream direction are provided over a considerable portion of both upper and lower surfaces, laminar-boundary layers may be maintained up to Reynolds Kur.bers of 6,000,000 or more if sufficient care is oxercised to eliminate disturbances from air-stream turbulence, surface roughness, and vibration. Such methods are shown to yield, within this relatively low Reynolds Kumber range, unusual drag reductions.

DERIVATION 07 AIRTOIIS

The part of the Investiga development of the new section ing a brief chronological acco elates contributed to the proj ton. von Doenhoff, Abbott, Sta and Kiss Alice Rudeen, who mad trlbution calculations. Their tributions are acknowledged he definite references to the det each.

tion that resulted In the s is best described by giv- unt of the mri. Many asso- ect, In particular. Pinker- ck, Robinson, Allen, Bicknell, e many of the pressure-dis- genersl assistance and con-

re* for brevity in lieu of ailed parts contributed by

The project was first undertaken as the result of reasoning like that presented in reference 5, which suggested possible late transitions in the presence of favorable pressure variations. Airfoil shapes were therefore sought having the minimum pressure on both surfaces well back. Trial shapes were used and results were checked by means of calculations according to Theodorsen's method of reference 7. Pinkerton, in particular, rai successful In finding a shape (fig. 2) that was considered reasonably satisfactory for preliminary tests, although not as the basis of a family. Models having this section were con- structed for tests in the variable-density tunnel and in the new tunnel.

Some doubt was expressed as to possible drag reductions, owing to the severity of the trailing-edge shape. The development of a suitable family was thertfore not stressed, pending the completion of the new tunnel and the tests of this first section. In connection with Stack's project on propeller sections for high speeds, however, a special mean-line shape was derived by von Doenhoff and the author from thin airfoil theory to give a uniform chord- load distribution. When pressure-distribution calculations became available for sorce propeller sections having this mean line, it was apparent that its use, through adding a small constant velocity increment to the upper surface and deducting an equal increment from the lower surfe.ee, tended to leave both surface pressure distributions substantially unaltered. Hence it became necessary only to develop suitable thickness distributions for symmetrical airfoils giving the desired surface pressure variations.

In the meantime, the new airfoil testing equipment had been completed, and the first new airfoil (fig. 2) was tett-

/

-I

4 ed in June 1938. In comparison w.ith foil, tested under the same conditio showed very- extensive laminar-bounda and the unusually low minimum drag c Comparative tests of the same airfoi density tunnel, however, fai?ed to s coefficients. Two very important co resulted from these preliminary test hie to realize large drag reductions to promote extensive laminar-boundar designs lead to an abnormally abrupt ing-edge region of the airfoil. Sec must be investigated under condition from turbulence.

an N.A.C.A. 0012 air- ne, the new airfoil ry layers, as expected, oefficient of C.0030. 1 in the variable- bow unusually low dreg ncl'Jsions therefore s. First, it is feasi- by designing airfoils

y layers, even if such fairing in the trail-

ond, such airfoils s approaching freedom

A development program for this new therefore begun at once. The outstandir. the investigation were to determine a 11 the backward movement of the minimum pre airfoil surface and to investigate, in p degrees of favorable pressure gradient i laminar region. Suitable tnlckness dist rical airfoils) were therefore sought; a these shapes were to be combined with on selected to give the defsired pressure di c, = 0.2, a reasonable high-speed or cr cient.•

airfoil t;'pe W«B g objectives of miting extent of ssure point on the articular, various r. the forward or ribut ions (symmet- nd, to save time, ly one mean camber tribution at

uising lift coeffi-

The desired symmetrical airfoils were based on ones for which calculations had been made in connection with the high-speed airfoil investigations. One forked out by Sobinson, through.the process of pressure calculations following small empirical changes made to produce a nearly uniform pressure along the surface from a point near the leading edge to the 0.7c station was considered satisfactory as a member of the family having zero pressure gradient and was therefore designated N.A.C.A. 07.

Other airfoils of the same series »cre then derived to investigate the effects of a progressive backward movement

of the minimum pressure point. A compressing function was applied to the tail portion of the airfoil the function being so chosen as to leave the airfoil unaltered at maximum thickness where the two parts join. The function is given by

x' - 0 5 = ~- In /l + k (2x-l)| (x ? 0.5)

where x represents the original of unit chord, and x' represent resulting airfoil was subsequent 1 back to its normal chord length, backward movement of the maximum sure calculations for this group Ous positions of the maximum thic a series should be satisfactory, ly of airfoils therefore received follows:

station for the airfoil 8 the new station. The y stretched uniformly the final result being a thickness station. Prea- of airfoils having vari- Jcness indicated that such The members of this fami- designation numbers as'

-I

H.A.C.A. designation 16 18 19

position of maximum thickness

•

0. 5c 0. 6c 0.7c

Approximate position of minimum pressure 0.6c 0.6c 0.9c

Leadir.g-edge radius index (reference 3) 4 3 3

Thus the number 16 suggests the form of the thickness distribution and the complete designation number K.A.C.A. 16-209 for example, is formed by adding three more digits after the dash. The first digit increases with camber and refers to the lift coefficient, 0.2 in this case, for which the airfoil is designed. The last two digits refer to the thickness, 0.09c, in this example.

Finally, the test results for these airfoils and particularly for the modifications investigated with cusp- type extensions at the trailing edge to relieve the severity of the flow conditions in this vicinity, led to the development of a second series designated 27. This modified series, designated by the first digit 2, is much like the first, but the thickness distribution is modified to

produce a la also mo the lift-1 the entire cent and t tall. Ihi ble furthe sure gradi the trail! with sone foil profi figure 3.

tendency toward a cusp-type ta dified aligutly near the trail oad distribution instead of be

chord is constant only over t -ieii tapers off prOb'ressive_ly t 3 mean-line modification was c r to relieve the severity of t ents In the turbulent-boundary ng edge. This modified mean 1 of the airfoils of the first s les included in this- investiga

AIRFOIL OHDINATES

11. The mean line ins edge BO that ing constant along he forward 80 per- oward zero at the onsidered desira- he adverse pres- -layer region near ine >as also used eries. The alr- tion are shown in

The airfoil ordinates may be derived by combining the camber and the thickness forms in the usual way, as ex- plained In reference 6. The mean-line form may be found from the following general expression, worked out by pinkerton and Allen:

• y° - tsriW [ «==• fi (a-*)2 ln I*"*1 - - I (b-x)s In |b-x| + i (b-x)s - I (a-x)2} -

- x In x + g - hx I

b-a 12 (!• •a)a In(l-a) - J. (1- b)3 In(l-b) +

+ i (1-b)2 i(x-«)a} + e

There the chord is unity and the load is uniform from the leading edge (x - 0) to the chordwise position x ••• a, then tapers off uniformly to zero at x = b, and remains zero from this point to the trailing edge at x -- 1. For the K.A.C.A. 27-215 airfoil with 0.Sc trailing-edge exten- s ion, a > airfoils,

1>s •"jj— and

a = 0.8, 3' 1.

For the usual 27 group of

and two other airfoils desig-

nated N.A.C.A. 07, 8-209 and 16, 8-215 have this same mean line. 7or the rest of the airfoils having the uniform- load mean line, a=b=l, and the expression for the mean line reduces to the simple form originally derived toy von Doenhoff and the author:

yo =

dx

»[' (1-x) In (1-x) + x In x

IZc = £i 4n

in (1-x) - In .x 1

The value indicated b; :l is the "ideal" lift coefficient

for which the airfoil is designed, 0.2 for most of the present sections. All the mean-line ordinates nnd slopes p.t standard stations are given in table I.

Ordinates for the thickness forms (symmetrical airfoils of the one maximum thickness 0.12c) are given in table II. Various airfoils of the present families may thus be derived by combinations of suitable camber and thickness forms. The method, now employed by some manufacturers, of laying out full scale the thickness ordinates perpendicu- lar to the Dean line at the standard stations, is definitely recommended for practical users of airfoils of these new families.

TEST METHODS

The airfoil mod&ls tested were of 3-foot spun and usually of 5-foot chord. (See fig. 1.) They were of wood carefully machined to accurately la id-out and faired tem- plets. During the investigation the matter of surface finish received much attention. Slight raviness or roughness was found to hasten transition so that during the earlier tests, the lacquer surface finish was progressively improved by sanding and filling to reduce any unfairness and small-scale roughness.

The first model »?.s built by attaching a cover to a wooden- frame but the slight tendency toward dimpling at the points of attachment gave marked adverse effects on transition. In fact, it appears that no perceptible three- dimensional dimples of this type can be tolernted. Such composite methods of construction were therefore abandoned.

3 Ho additio realised, water clot remove all although 9 extreme no layer 1B V 10 mi 11 lor. measureraen pression o the estioa as a high

nal gain in however, hey h, working i appearances

oaie slight g ae portion o ery thin. X ths of an In t or. a typic f the surfac ta that the grade automo

surface smoothnes ond that obtained n the direction o of slight depres

a in may appear fr f the airfoil whe surface R.M.S. r

ch was obtained f al model. A bett e condition may p finish did not ne bile finish.

s on transition was by the use of 400

f the air flow to sions<or elevations, om polishing the re the boundary oughness reading of roa a "profilometer" er qualitative lm<- erhaps be had from ed to be as smooth

No attempt will be made to describe the tunnel and the detailed testing methods in this preliminary report. The air-flow uniformity in respect to both turbulence and distribution throughout the test section is Buch that departures from the desired conditions are extremely difficult to determine.

The Investigations were 'generally of an exploratory nature and followed no routine procedure. It was at first planned to use a balance to obtain some force measurements, but it later appeared that air-flow and wake-survey methods were giving all the information required for the preliminary tests. Consequently, a tunnel balance has not been installed.

The usual testing procedure was first to estimate the drag from the integral of total-pressure-defect measurements in the wake for several angles of attack near that of minimum drag to find the angle corresponding to the most favorable flow conditions on the airfoil. Later an "integrating" manometer connected with a survey "rake" was employed. This arrangement gave a direct indication of the drag by the depression in the general liquid level in the unaffected tubes, which are associated with the rake tubes that lie outside the wake. The method should be apparent from figure 4, which shows the wake from 0.1-inch-diameter tubes spaced on 0.2-inch centers and located in the wake 0.4c behind the trailing edge of the 5-foot-chord models. The wake in figure 4(a) is from an N.A.C.A. 0012 airfoil at zero lift and the wake in figure 4(b) is from one of the low-drag airfoils at approximately its design lift. The separate tubes at the left indicate the tunnel dynamic pressure and the wake static pressure.

The airfoil drags were thus estimated over a range of

I

10

tie boundary layers and the tran.it Ion point. .. • fune- • Y«„ «f the HevnoldB Number, to compare the experimental

ther discussed when the results are presented.

yinally. some oi the airfoils were tested in the varl- ahJe-Iensity tunnel in order to indicate the u.u.l -"£•" «irfoil characteristics and also the drag characteristics

f r.Sl 1«! Reynolds lumbers or other "-as »her, transition effects tend to be '»PP"'BeJ- "^ / "J '„. may also be employed to estimate the maximum 1 ift to be ex "cted in flight. The test, therefore include some in which split flaps are applied to the sections.

%

RESULTS

So attempt has been made to present these P"* *»*»"* result, in a complete or final form. Only the more .ignif- icant results are included and no correotlon. h.»« been applied exceut to the results from the variable-density tunnel! The el values given are simply the integrals from

the total-pressure-defect measurements. A 8mal^c°rr?';" tion will eventually be applied for the survey-tube "i« (effective centers not the geometric centers . *erhap. an improved approximation to the tr.e drag results would have been obtained by the use of the Jones formula, but no corrections of this type are being made pending the oo^P1"- tion of an investigation now in progress to determine the correct methods of drag measurement by the wake-survey method in a closed tunnel. In general, it appear, that the more exact methods will always result in corrections that will reduce the drag values presented. These corrections may, in some cases, be of the order of 15 percent.

Tunnel-wall corrections should also be applied te the results of pressure measurements on the airfoil surfaces. In tne future, this difficulty will probably be avoided by

•-J testing an airfoil somewhat thinner than the sec to represent. Tor example, the surface pressure the minimum-pressure point on an airfoil of 5-fo with a 15-percent-thick section in the tunnel is percent more than it would Be In free air. Such tion effects, of course, influence the lift resu the pressure determinations, although this error approximately removed by correction from some of suits for comparison with those from the variabl tunnel and presented in figures 28 to 33. A sma measurement error, of the type'that has sometime roneously referred to as "blocking," may also be tending further to reduce the coefficient values

11

tion it is; drop near

ot chard about 6 rest ric-

lts from has been the re- e-density 11 velocity- s been er- present

Transit ion pressures from flattened lower tive height of velocity Indies pressure and th thus indicated quently the loc the beginning o ity as the tunn transition poin servation.

was Judged from observati the inner mouse tube, whic wall against the wing sur

the tube was usually about ted by the difference betw e static pressure from the the surface velocity gradi al skin friction. Transit f a sudden and marked incr el speed was gradually inc t across the wing-surface

ons of impact h rested with its face. The effec- 0.008 inch. The een this impact mouse static tube

ent and conse— ion was judged as ease in this veloc- reased to move the station under ob-

Later, an improved method that gave more precise re-

sults was adopted. The function (q8)

1/3

there q is

the dynamic pressure indicated by the surface tube and q is the stream dynamic pressure, was plotted against -/~q • This procedure is substantially the equivalent of plotting against the Reynolds Number B a function of the surface velocity gradient:

d V^o __iZsLt

which tends to remain independent of the Reynolds Number as long as the surface tube remains relatively close to the surface in a truly laminar layer. Figure 5 shore the method applied to the determination of three transition

I

I

12

polnta on thef ir.A.C.A. 27-212. It will be noted that the function remains constant Indicating a truly laminar layer In the low-speed range and then rises abruptly in the transition region. The transition points were taken as the positions indicated by the arrows in figure 5 at the knee of each curve.

The condition of the boundary layer just prior to transition was investigated by the hot-wire method to study in greater detail the nature of transition, and to find an explanation for the tendency of the transition function to rise slightly before the appearance of marked transition effects. A fine hot wire was used with a high- gain d.e. amplifier and a cathode-ray oscilloscope. The results obtained, some of which are indicated in figure 5, are rather significant. Well ahead of the transition point the laminar-boundary layer was remarkably steady and appeared to be free, or nearly free, from unsteadiness or fluctuations of the Dryden type. Perhaps such steady laminar layers should have been expected under the test conditions of very low turbulence In the new tunnel, particularly after it had been demonstrated that the experimental and>theoretleal boundary layers agreed excellently, both with respect to total layer thickness and the velocity profile within the layer, but Sryden (reference 9) had found from experiments that some layers may become markedly unsteady while, at the same time, retaining laminar properties, at least much more nearly laminar than turbulent. The oscilloscope showed, however, a quite different behavior as the Reynolds Number was increased to bring about transition. Instead of fluctuations in the laminar layer, the observations indicated momentary transitions to skin- friction intensities comparable rith those of a full}- developed turbulent layer but of extremely short duration, perhaps less than 0.01 second and at first occurring only once every several seconds. These very short bursts of turbulence were much too fast to eppenr in the over-damped mouse measurements, which indicated only a mean result. The reason for the early gradual rise of the transition function is thus apparent. The total time duration of the turbulent type of flow was of the order of 1 percent when the "transition point" Indicated by the arror at H = 6,000,000 in figure 5 was reached. As t'.j> Reynolds Num- ber ras further increased, the frequency, a..d also the duration, of each of the turbulent bursts increased so that the relative total time in this condition increased as indicated by the percentage values given opposite the points in figure 5.

-*

a

-1

13

' J Pressure-distributIon results, experimental, when extensive mouse vatlons Fere made, are presented to 21. In some cases, the theoretical directly from calculations o;~ Iheod ence 7) and, in others, by Allen's of velocity-Increment addition to basic symmetrical section to allow tribution This methci is very sin applied to the prediction of pressr. critical speeds for other derived a ilies. For such purposes, tie theo distributions for the symmetrical s figure 22 and in tables III to Til.

both theoretical and static-pressure obser- gethar in figures 6 to pressures rere obtained

orsen's method (refer- method (reference 10) he velocities about the for the lift-load dis- ple and may be readily re distributions and irfoils of the sew fam- retical basic pressure actions are given in

Certain additional important data are also included in figures 6 to 21, in addition to the arrows indicating the measured transition-point positions, and the corresponding wing Reynolds Numbers indicated in millions near each ar- row. Separation of the flow is also indicated as Judged from the mouse measurements. Included also are the angle of attack, the corresponding measured minimum drag coefficient, and the Reynolds dumber at rhich it occurred. The theoretical compressibility-burble speeds, expressed as the ratio Mc of the critical speed to the speed of sound obtained both from the measured and the theoretical peak negative pressure coefficients by the method of reference 11. are also included in each figure.

Some other experimental data are presented Trith the discussion. Data dealing with further details of scale- effect calculations, Bkin-friction distribution, and boundary-layer studies in comparison with theory, the analysis of the transition .data, the extension of these airfoil developments to higher Reynolds numbers and speeds, studies of the relative tunnel turbulence, and check tests in other tunnels and in flight will be separately presented and discussed by various authors in subsequent papers.

nsetjssiaii

Best Airfoil, the Optimum Reynolds Number Range

This discussion will first consider the experimental data on the various airfoil forms almost without regard to the Reynolds Number, considering mainly the minimum-drag

14 '„

results in the range where the gains from extensive laminar flows were largest. Applications at higher Reynolds Numbers will he considered later. The emphasis on the minimum-drag results may appear inconsistent with earlier N.A.C.A. airfoil reports, but several factors have altered the point of view. A commendable trend has begun, led particularly by the Army Air Corps, away from an arbitrary landing-speed requirement that tends to fix the wing area in relation to C- Changes in CT therefore no

•"max •"max longer necessarily produce correeponding changee in minimum wing drag through the process of forcing a change of wing area. High maximum lift coefficients, moreover, were often associated with large and abrupt lift losses beyond the maximum, a combination of perhape less than no value except through the possibility of thus circumventing the arbitrary landing-speed requirements to effect area and drag reductions Finally, the drag reductions possible through the realization of extensive laminar layers are relatively so large that variations tend to become

"max relatively Unimportant between airfoils of slightly different shape, particularly when used with a high-lift device, as in figure 31, for example

The data should thus be considered first in relation to the more important factors affecting tne minimum drag. As pointed out earlier, one of the objects of the investigation was to determine the limiting extent of the favorable press-are, or laminar flow run over the forward part of the airfoil and also the effects of variations in the degree of the pressure gradient within this range. As re- gards the limiting extent of the laminar layer, the drag results (figs. 23 to 27) show that it tends to increase as the airfoil thickness decreases. The lowest drag coefficient obtained was 0.0023 with the N.A.C.A. 18-204 section at a Reynolds Number of approximately 4,000,000. Figure 6 Indicates that the extent of the laminar run was then more than 0.80c. The determinations indicated that this run could not be greatly increased. If an increase was attempted through a reduction of the Reynolds Number, wake fluctuations and. drag increases were nncountered. The adverse effects were evidently associated with transition occurring momentarily so far behind laminar separation that an incomplete closing in of the turbulent layer with con- sequent pressure drag was encountered. (See reference 12.; On the other hand, adverse effects also appeared when a further backward shift of transition was attempted through a backward movement of the minimum pressure point, as in

15

changing to the N,A.C.A. 19-304 section (fig. 7). The minimum drag coefficient obtained was 0.0039; this value indicates that such an extreme backward movement of the minimum-pressure point results in too Blunt a trailing edge even for a 4-percent-thick airfoil. Tne pressure recovery and adverse gradients are evidently too severe to "oe over- come even by the most favorably situated turbulent layer.

Turning now to the 9-pe tne 18 series is still found ficient (0.0026), but the min Reynolds Number. This change a lower boundary-layer Heynol the increasing airfoil thickn increased favorable pressure portion of the surface. Thus at a given Bg , the given va 5 on the thicker wing. It i drag of the N.A.C.A. 37-209 c value and shots a cpnsiderabl higher Reynolds Numbers than cusp-type trailing edge was a 10T in an attempt to roduce t edge conditions. The wake su determinations on this airfoi ratic that the cusp was cons tablish the circulation and 1 aents could be made.

cent-thick ai to have the 1 infant now occu

may be attri ds Number Be ess because o gradients ove

if transit! lue will bo r s noteworthy irfoil reache y less marked the 18 airfoi dded to the 1 he severity o rveys and the 1 without the dered almost ift so that a

rfoils (fig. 34). owest drag coef- rs at a higher buted partly to associated with

f the resulting r the laminar on tends to occur sacked at a higher also that the s nearly as lor- a rise at the

1. In fact, the 8 airfoil (fig. f the trailing- surface-preseure cusp were so er-

necessary to es- ccurate measure-

A consideration or the pressure diagram in figure 10 suggests another reason for the use of the cusp. With blunt trailing edges like those of the 18 series, a marked pressure drag may be associated with the failure of the actual surface pressures to rise toward the stagnation pressure at the trailing edge as the theory indicates they should when the form drag is zero. As shown by figure 10, the theory does not require for the cusp tail this unattain- able type of pressure rise.

Associated poor pressure recoveries and marked form drags for tho blunt airfoil are apparent in the results, particularly at Reynolds Numbers higher than that for minimum drag. Eere the conditions in the turbulent layer become progressively less favorable for a pressure recovery as the transition point moves forward on the airfoil. The increased turbulent skin friction occurring ahead of the minimum pressure point thickens the turbulent layer at the

16

beginning of the pressure recovery and hence makes the given adverse gradient relatively more severe. Under extreme conditions, turbulent separation may be expected.

IT.A. the ery pect for ly f mark 27-2 seri the the fact a lth

Sues c C.A. 27 sever it region ed redu the min or the ed favo 15 sect es were other a airfoil ory as ough in

ons ide serie

y of t behind ct ion imum i 12-per rable ion ( f revis

irfoil that

a nemb eluded

rations led s (fig. 11) he flow cond

the minimum in drag at R s shown in f cent-thick s result is no ig. 27), but ed betreen t s were const sag tested a er of the 27

should the

to the deve which was de itions in th pressure po

eynolds Numb lgure 25 and ection in fi t indicated

the ordinat he time the ructed. The re not norr c family. Th

refore be di

opmen signe e pre int. ers a

more gure for t es fo 15 ai ordi

or.sid e t es scoun

t of the d to relieve ssure-recov-

The ex- bove that part icular-

26. The he N.A.C.A.- r the 27 rfoll and istes of ered sat is- t results, ted.

Another unanticipated result of changing from the 18 to the 27 series was the shift of the minimum drag to lower Reynolds Numbers. The same result is again indicated in figure 25 in changing the minimum pressure farther forward from 0.7c to 0.6c in going from the 27 to the 16 series. The opposite shift of the minimum drag to higher Reynolds Numbers was expected or.-ing to the lorer local Reynolds Number at the minimum pressure station at a given wing Reynolds Number. The explanation is that minimum drag with these airfoils doss not occur when transition is near the minimum pressure point, or ever, forward of the laminar-separation point. (See figs. 12 and 18.) These experimental data do not conflict with the laninar-separation theory (reference 13), rhich places the laminar separation point very near the minimum pressure point after the layer has become thickened by its long run over the forward portion of the airfoil. When the miaicum pressure point is not well back, minimum drag occurs at a Reynolds Number so low that moderately extensive larr.ir.cir separation is actually present. The transition occurs soon enough to close in and permit the pressure recovery but not soon enough at minimum drag to produce excessive turbulent skin frictions. In the separated or adverse pressure rr>nge, however, this transition tends to occur at a reduced Reynolds Number.

Figure 25 also throws some light or. the question of how steep the favorable pressv.re gradient should be over the forward part of the airfoil. A comparison of the

^ 1?

N.A.C.A. 07-209 and the 3.A.C.A. 27-20° airfoils shows a higher minimum drag and an earlier rise with Increasing Reynolds Number for the airfoil with the flat pressure distribution.

A tentative explanation can be given as the result of further fundamental boundary layer and transition studies not included in this report. The difficulty with the flat pressure distribution is not primarily that transition necessarily occurs at a much smaller value of Hg in the absence of a favorable pressure-field, although there may be some slight tendency in this direction, but tliat very small disturbances such as-slight imperfections in the model, slight departures from the design angle of attack, or slight flow fluctuations may produce regions of local adverse pressure gradient; This condition tends to produce regions of excessive boundary-layer thickness (or even local separation), which tend to grow three dimtnsionally in the absence of a favorable pressure gradient impelling the low-energy air.along in the normal flow direction. Hence, excessive values of He may appear locally leading, in turn, to a premature transition. The.optimum magnitude of the favorable pressure gradient for these airfoils therefore becomes largely a matter of practical compromise. Small gradients require extreme care in the elimination of disturbances, whereas large gradients cause excessive skin friction, excessive form drag due to the more severe pressure recoveries, and low critical speeds due to the excessive peak negative pressures.

It thus appear considered, the H.A approximation to th the minimum drag oc particular design r optimum lift may be for the N.A.C.A. 27 designed for an opt an extreme procedur principles too far, may find some appli airfoil section for planes, gliders des and propeller blade the maximum profile

Applicat ions

s that, within the Heyno C.A. 27 series represen

e best compromise. The curs may be varied .at 11 equirements. The extent increased is suggested

-2012 in figure 16. Thi lmum lift coefficient of e probably pushes the pr but high-lift airfoils

cation. The ultimate pe application such as- Ion

igne<1 for small glidin s, guide vanes, etc., is

Xi/D for the section.

Ids Number range ts a reasonable lift at which berty to meet to which the

by the results s airfoil was 2.C. Such esent design of this type rformar.ee of an g-range air- angles, blower measured by ?ith these new

18

airfoil-design principles, low-drag coefficients nay toe attained at rather high lift coefficients. Tith the older type of flapped airfoil, for example, the pressure recovery realizable over the uppe- surface of the flap was re- stricted toy the excessive thickness of the turbulent layer in this vicinity. Owing to the possibility of maintaining laminar flows over the forward portion of the new airfoils, the turbulent layer at the after part of the airfoil may toe relatively thin with the result that relatively abrupt pressure recoveries are attainable. Although the boundary- layer studies on the N.A.C.A. 27-2012 indicated that the lift in this case had probably been pushed too high, a maximum profile L/D of over 290 was attained. For airfoils similarly designed tout with slightly lower optimum lifts, the turbulent separation that occurs near the trailing edge may toe sufficiently reduced to produce even higher L/D ratios.

By a suitable choice of the camber to give the desired optimum lift, the lift range of low drag (figs. 28 to 32) will toe sufficient for many practical applications. Out- side the low-drag range, the variable-density tunnel result* suggest that the airfoil drag will not toe excessive. The results presented herein are applicable only within the lower Reynolds Number range and therefore appear most naturally suited for application to small aircraft and gliders. It should not toe overlooked that they may have much wider application to special designs in which it is feasible toy reduction of wing chord or density at high al- titudes to achieve the proper Reynolds Number. In application to airplane wing design the camber will probably be selected so that the optimum lift will occur near the cruis- ing speed. An airfoil somewhere between the N.A.C.A. 27-112 and N.A.C.A. 27-512 will thus probably be employed. The advantage of the new sections will then appear through increased curising speeds and in more economical operation within this speed range.

It should toe emphasised, however, that the gains will not toe marked unless suitable applications are selected. It may toe desirable to employ unusually large aspect ratios in order to reduce the induced drag and to reduce the chord sufficiently to obtain a suitably low Reynolds Number. The wing surface must, of course, toe- fair and smooth over the forward 80 percent. Vibration should toe avoided and, in all probability, the propeller slipstream on the wing must toe eliminated. Pusher propellers are therefore to toe recommended pending an experimental demonstration that the disturbing effects of the tractor propeller can be toler- ated. Disturbances arising forward of the wing along the fuselage will affect only small portions of the wing ad-

19

Jacent to the fuselage. Only that part of the wing Inside a line extending from a point at the leading edge just outside the fuselage-boundary layer backward toward the trailing edge and outward with the flow direction at an angle probably less than 8° need necessarily be subjected to the usual high turbulent skin friction.

Host important of all in any application, however, is the reduction of fuselage, tail-surface, and parasite drags to a reasonable einlnr."i. High parasite drags may easily mark any marked gain from a large reduction in wing-section drag. One private-orner type of airplane tested in the N.A.C.A. full-scale tunnel showed for example, a drag coefficient of approximately 0.0600. A reduction of wing drag from 0.0060 to 0.0030 would consequently have reduced the over all drag of the r-irp'lar.e only in the ratio 55/60 The resulting speed increase ironld thus represent an almost inappreciable gain. On the other hand, if the airplane to wnich the new wing is applied is so clean that the wing-profile drag represents e large part of the entire drag, the performance gains will be very large. The higher speeds attainable, in turn, reduce the induced power, and often improve the propeller efficiency. Particularly in bucking a head wind, the time saving and tae economy expressed in miles per gallon, a matter of vi- tal importance to the private flyer, should thus be improved to a very marked extent by the application of the new wing sections.

Applications at Reynolds Numbers Above the Optimum

Littlv will be said regarding the application of these data at the highor Reynolds Numbers because further investigations outside the scope of this report are now in progress to develop methods of maintaining these same low- drag properties at very high Reynolds Numbers. It appears, however, that comparatively small gains of this same type may be readily realised at the higher Reynolds Numbers by maintaining the laminar layers over only a comparatively small portion of the forward part of the airfoil. In fact, full-scale tunnel tests of the N.A.C.A. 23012 airfoil (reference 4), and of the N.A.C.A. symmetrical airfoils (reference 14), as well as tests of the N.A.C.A. 23012 airfoil to study roughness effects in the 8-foot high-speed tunnel, indicated that some gains of this type would be possible on existing airplanes if sufficient attention were given to the surface condition on the forward part of the wing.

20

Actua owing suits heren po ss i t iona metr i pep.k f ice. tends t ion.

lly the gel to tunnel- On the o

tly ur.suite bly (>t ext r 1" type of eal airfoil to occur ve

(See refe to lead to

ns might "be turbulence ther hand, d to the de emely large lift distri , for examp ry near the rences 15 a p r e ra t u r e

noticeably lar effects present these types of sired flow oond Reynolds Humbe

"bution associat le, causes n mi leading edge o

nd 10.) This c laminar separat

ger in flight- in the test re

air foil are in- iti ons, except r s. The "addi- ed with the syn- ni;.: jn-p res sure r. t he upper sur- ond it ion always ion or transi-

An obvious improvement ir. the medium Reynolds Number range is possible with an airfoil like the N.A.C.A. 2412-34 from the family of reference 3. This type of airfoil has a better lift-load distribution and a thickness distribution that does not produce a minimum-pressure peak exces- sively far forward. The K.A.C.A. 2412-34 and H.A.C.A. 1412-34 airfoils are therefore to be tested in the r.ew tunnel and will be separately reported when the results are avaliable.

It should be urged, however, that snap Judgments based on boundary-layer calculations along the lines suggested by reasoning similar to that presented in the preceding para- graphs be withheld pending further experimental investigations. Some cf the test results (figs. 15 and 26, for example) show large drag increases associated with comparatively small forward movoments of the transition point. The cause of this rather peculiar behavior of the drag was found, as the result of a supplementary investigation to be separately reported, to be associated with the very high sfcin-frlet ion intensities usually present at the onset of turbulence in the boundary layer. The adverse effects of the high friction intensities are moderated, however, rhen the transition occurs in a region of pressure recovery as It does on the best sections in the optimum operating condition. In fact, the type of flow leading to a relatively high intensity skin friction is then actually desirable in order to avoid separation. It thus appears that it may always be desirable to effect some pressure recovery in the neignborhood of the transition point, not only because of the immediate saving in skin friction and lower losses associated with recovery but also because the turbulent layer is left to run over the remainder of the airfoil in a thicker, and hence lower, drag condition.

The same conclusion was reached in a different "ay.

which astua sion on the than the OF employed, t ed when the ward part o exposed to relat ively part where ties are a 24 cannot b they be sal and the res vey rake, t sidered str withheld pe less, that large forwa to result f

21

lly led to the design of the 0.5c-chord exten- N.A.C.A. 27-215 airfoil. If chords longer

timum, that is higher Heynolds Numbors must_ be he least adverse drag effects should be expect-

best possible section is chosen for the for- f the airfoil and the remainder, which ir.ust be turbulent skin friction anyway, is added as a thin extension lying in the wake of the forward the velocities and turbulont-friotipn intensi- mlnimum. Although the test results in figure e 3aid to substantiate these views, neither can d to disprove them. Owing to the larger chord ulting different relative position of the sur- ho results for this airfoil should not be con- ictly comparable, and conclusions should be nding further tests. It is apparent, neverthe- drag gains will be much less marked if any rd movement of the transition point is allowed rom increasing values of the Reynolds Sumbor.

CONCLUSION

For airplane wing design and for other airfoil and streamline body applications in the lower Reynolds Number range the new laminar-flow airfoils and the general design principles deduced from the present investigations may be expected to yield actual wing-drag coefficients markedly smaller than those heretofore possible.

Airfoil and flow investigations of the type considered must be made under tunn-el-flor conditions approaching freedom from turbulence. Under tiiese suitable conditions, truly laminar-boundary layers may be maintained to unusually high values of the Reynolds ITumber. Transition appears to be sensitive to very small disturbances of various kinds including surface roughness and air-stream turbulence and, in the absence of such disturbances, appears to be of a different character from that usually observed in wind-tunnel testing.

Langley Memorial Aeronautical Laboratory, National Advisory Co.amittoo for Aeronautics,

Langley Field, Va., April 25, 1939.

22

REFERENCES

1. Stuper, J.: Investigation of Boundary Layers on an Airplane Wing in Free Flight. T.lt. No. 751, N.A.C.A., 1934.

2.

3.

6.

9.

Jones. Melvill: Layer. Jour, pp. 81-94.

Flight Experiments on the Boundary Aero. Sci. . vol. 5, no. 3, Jan. 1938,

Stack, John, and von Doenhoff, Albert E.: Tests of 16 Selated Airfoils at High Scecds. T.R. No. 492, N.A.C.A., 1934.

4. Jacobs, Eastman N. , and Clay, William C: Characteris- tics of the N.A.C.A. 23012 Airfoil from Tests in the Full-Scale and Variable-Density Tunnels. T.R. No. 530, N.A.C.A., 1935.

5. von Doenhoff, Albert E.: A Preliminary Investigation of Boundary-Layer Transition along a Flat Plate with Adverse Pressure Gradient. T.N. No. 539, N.A.C.A., 1938.

Jacobs, Eastman N.: Laminar and Turbulent Boundary Layers as Affecting Practical Aerodynamics. S.A.E. Jour., vol. 41, no. 4, Oct. 1937, pp. 468-472.

Theodorsen, Theodore: Theory of Wing Sections of Ar- bitrary Shapes. T.R. No. 411. N.A.C.A., 1931.

Jacobs, Eastman N., Ward, Kenneth E., and Finkerton, Robert M.: The Characteristics of 78 Related Air- foil Sections from Tests in the Variable-Density Wind Tunnel. T.R. No. 460, N.A.C.A., 1933.

Dryden, Hugh L. Jour. Aero. 85-100.

: Turbulence and the Boundary Layer. Sci.. vol. 6, no. 3, Jan. 1939, pp.

10. Allen, H. Julian: A Simplified Method for Calculation of Airfoil Pressure Distribution. T.M. No. 708, N.A.C.A.. 1939.

11. Jacobs. Eastman N.: Methods Employed in America for the Experimental Investigation of Aerodynamic Phe- nomena at High Speeds. Misc. Paper No. 42, N.A.C.A., 1936.

-!

£ 23

12. von Doenhoff, Albert E. , ar.d Jacobs, Eastman B.: Tran- sition as It Occurs Associated with and Following Laminar Separation. (Paper Presented before Fifth International Congress for Applied Mechanics, Cam- bridge, Mass., Sept. 12-16, 1S38.)

13. von Doenhoff, Albert E.: A Method of Rapidly Zsttreat- ing the Fositlon of the Laminar Separation Foint. T.B. Bo. 671. B.A.C.A., 1938.

14. Silverstein, Abe, and Becker, John V.: Determination of Boundary-Layer Transition on Three Symmetrical Airfoils in the B.A.C.A. Full-Scale Wind Tunnel. T.B. Bo. 637, B.A.C.A.. 1939.

15. Jacobs. Sastman B., and Rhode. R. V.: Airfoil Section Characteristics as Applied to the prediction of Air Forces and Their Distribution on Wings. T.R. Bo. 631, B.A.C.A., 1938.

•

'„ N.A.C.A -r»»i»» i, a

4 0

S u E

VI I o

Sal »>»0iiO9-S83Sio5«ii>inoii>ei22£,,'^>t,a«?«5_ (s.

^.^niniojn^^^iQiairiiqicjifiijtdfiijinXfr^iriinvi--

5§ ^5

i- •* « *-*> >• << a IT, Q !2 "> Q "> * "*> S — <S • -"« *> o>"> o 1 * Coitf-S: vj m K^fl rOi*-Q M ssr> tooioqiQ vJCVsa s»^»> NdOOO *)

31 36 < •< oi oi Hi x ^ V *« *» "i *ri •*>' ^ ^ >•> wi KS «J- NJ c« •«• **'

" < < *i "i •o V ^ «o W *«i "«i >rj *i »r> Jo ,n*o^v'f>*J«j'^"^",: "

"Si*.

Q •*"• ^ 9 K t? P1 h* "* K « N » » N K» _. <a |8|lf *R2$!5 S S 8 5 1? * S o R

"- •< ^ ><i K i; 5 cv «* n "i» «• si "•> \s -J I- >- «10 o> m * <n <n <n o Q

-J1

u

Swisses •-,•,,•

a i i ' i i Jit

f"» Q o S • FO ») >* «o ~- »|<i> •* >• ^ <• •« » o A. 10 >o w -> « s o o c o

•la *

i' • r ij i

r\ !2 2 * a •• * ° • * t1* S 9 * *» *» •* » a 5 * *> • • £i «5 is. 9 o

—is I I I I I I I I I I I I '

QQv2:»^,i)>aisqaooo'sOoofl)<N4^*KHBC(^ Jo <y

4 N.AC. A . To»lc» 3.4.S.&

HL35!$£*fl$MBltt«

1 s

V)

N

o

^ • ^5

If

T2

IO in w- KI r- rs. r«\ KI /s. ^ u o

O-4 «l«« « r^ at M «l «9 C« W W «< Q 0) Q

!«

s

Srr ty r>j Q ». w) ~ ^- 1 ^ ^ ss <v W>

! 2 "S % & & 9 N 8 °o £ »•»*»• «» ' < >J •< "< "i* •*»' "*» < •<"< •<•< "<

issmii 5>°

!55SS5SS?8S888!8S j

1 3

^5

8 3 2 < -a

So

m r* Q •• A 2" 5* *» Ui w» m «t v* = W In -^ -^ «n *» K J- M -i j K f <n *» K

0)O

'VJ TAILS VII

Basle Fres9<ire Distribution

Station percent

c

Values °* pressure co&fficient " S"

N.A.C.A. 19-004

N.A.C.A. iC—00 6

E . A . C . A. 19-009

N.A.C.A. 19-012

N.A.C.A. 19-C15

N.A.C.A. l'J-OIR

0 1 .25 2.5 5.0 7.5

10 15 20 30 40 50 60 70 60 90 95

100

0 1.009 1.027 1.044 1.050 1.053 1.056 1.C57 1.0 60 1.064 1.070 1.079 1.095 1.129 1.172 1.C91 0

0 1.01C 1.039 1.065 1.075 1.0S0 1.Geft 1.0B7 1.C91 1.097 1 .10 ft 1.118 1.141 1.195 1.259 1.135 0

0 l.COO 1.056 1.097 1.113 1.121 1 .12 6 1.130 1.138 1 .146 1.159 1 .177 1.210 1.297 1.Z12 1.195 0

0 o.s<e9 1 . C 7 5 1.129 1.151 1.1 (-1 1. 170 1.17 5 1.1 fc c 1.201 A • *- i. C

1 .240 1.26 5 1 . 40 4 1 • i. 1 5 1.239 0

0 C . 072 1 .C&2 l.i rr, 1.190 1.202 1.214 1 ,:J2£ 1.242 1.2T-1 1.2S5 1.311 1.3'-5 1 .518 l.r-25 1.2 C 3 0

0 0.9 55 1.10'' 1.197 1 .2 ?.Q 1.24 C 1 .2 »!0 ..272 I . 300 1.323 1. 3 rtO 1.395 1.455 l.fPl 1.70 9 1 .2'.;4

i

'-I

^1

V.A.C.A. rig*. 1,3.4.

NAC* •• 20*

•-- - 11-204

NAC A 27 £OS N*C* £7 ,'ia

NACA Z7-2IB 5C ' * '

NACA 2' 2012

Figure 3.-K.A.C.A. laminar-flow airfoil profile*. NACA Ml*

ra) N.A.C.A. 001* airton at aero Ufl IbJ Laminar-flow airfoil at optinum lift. Figure 4.-Wake meaaureuenta by integrating nanometer.

»

J

NACA Fig. 2

«•»•"" -—

Upper

\

s

r "Lower s \\

i

1 \\

4 \ 1 1

50 100

Figure 2.- Preliminary form of laminar-flow airfoil, (NACA 25 B 09-46).

i

Fig.- 5

•

\ "l If

Li_ J. NACA 27-2i; ;

1.6 / <—>^

/

~-»

1.2 /

/ u< 5Cjt

in .8

.. / 2SJ&

,

7 /

/J y { 10*

/ >——•*"* \ s * -T^

.4

s

Pi

ho

4 5 6?'' Reynolds number, millions

pare 5.- Method of transition measurement. Variation of the "transition function" and correlation with

t-v/ire Etudies.

j

! i

1 i

i I

_ i

•

•1

in

-7

IJACA *i«f. 6

l.fi

11

1.2

4 .5

-; r" 2

.8 r

4.9

.4 -

C i.n

Theory 0' Experiment 0*

ci | M, C<-»n

0.84 _~

.83 0.00^2 at R=4.i!x106

100

Figure 6.- NACA le-J;C4 airfoil.

1

-a- NACA Fig. 7

r~

1.6

1.2 ; r

• 5.2

IIX

.4

00

Theory Experiment l/2°

0.2 0.81 ,82

C<1„,, aan 100

0.0029 at R=4.6xl06

Figure 7.- NACA 19-204 airfoil.

HACA Fig. e

• -a

1 fi

X~ \ ljr

/7 Kl /

—if— 6.1

a ( 1 \ 4 'i A

SO

Theory Experiment

0° 1/2°

c\ 100

0.2 0.71 **mii

0.0039 at B=4.6xl06

Figure 8.- KACA li*-209 airfoil.

I SAGA fi*. 9

1 6

6 .0 f .8 4

i .&*

f r

1 2

1

a f Ml

4

*see FiiSUJ e 11.

SO

Theory Experiment

0° 1/4°

ci Mj.

0.2 0.77 .76

c*mir.

0.0026 at H=S.3xl06


ICO

i 4

HACA ?lg.lO

wi

1.6

1.2

.R

,|

5 .8 5 .6

7

'•

I- & .0*

r "N

\ f

•—r~"- V 'i > 6. 1 5 e

\ :

\

*sso«> Pl^ur. 11.

r: '" < 5C

-lam

It 0

Theor; Experi a.-i.t

C r

--> o

O.S ' 0.7f .74 o.ooi .« fct r =S..3K1 o*

Figure l'J.- HACA. lfc-20'j airfoil with 3iri o:iip extension.

t

i

NACA »ig. 11

1.6

1.2 •-

1C:0

Theory Experiment 0.002' at H---1.6xl0 Arrows indicate locntiori of transition corres to the Reynolds number ir.iicatei in millions.

S rnr.dir


i

* NACA Fig. 12

ICC

Theory Experiment 0° - ,7fi 0.0031 at K=T.8X106

Arrows indicate location of tras.sition corresponding to the Reynolds number in millions.

Figure 12.- IIACA 16-205? airfoil.

1

•4

NACA ?ig. 13

r — .—

J J. ,1 H.fl , ,

i r * J < V

.... ^ k "

N 1

X,

T - 1

9

6. 8 ! 3."

'

1

i j

I I

! 1 1 ! • ' 1 1

so a I n i «c '-,!..

Tneory j 0L [0.2 0.""!» Experimental!l/2°j - | . vg GO.'I at H*3.3xlOr

Figure 13.- NACA 07,8-209 airfoil.

100

-(

u KACA Fig. 14 -«» flri

—J

•

7.3 1.6

r* 3.8 x^-~^"

X |

1.2

»'•

( i 6.5 \

ftum —

.8 r

'

\

V

\

4

J :

0

5C

! 100 •

c 1 CX Mc c,inin Theory ( J° 0.2 0.73 _ Experimei.t ( 3° .72 0.0033 ;xt K=b.3nl06

•

Figure 14.- N>WA 18-212 airfoil. i

i • i

! i •*

HACA Fig. 15

1.6 i.Z

6.0

1.2

I ' '.,.-•

.6

100 I a c\ , Kc

Theory | 0 Experiment 11/2*

c<irr 0.2 |0.?4

19 i .7? j 0.0029 at R^.PJUO"

Figure 15..- NACA 87-212 airfoil.

HACA Fig. 16

100

Figure 16.- NACA 27.2012 airfoil.

:M

/

MA'CA >

i -4

««. 17

| 1 ^

1 fi 7-2 4.6

: t * j .5 1 [ 2.4

1.2 t *

.8

^TV X 1

.4 l

0 50

Theory Experiment

ct I "o 00 o 0.74 0«> 0 .731

cd, min

0.0071 at R=4.6XA06

Pigure 17.- HACfc 0012 airfoil.

1

c-t itfACA

Fig. 16

50

Theory Experiment 1/2°

0.2 .18

0.72 .70

lrr.in

0.0041 at E=4.6xl06

Figure 18.- NACA 1«.8-215 airfoil.

100

._./

NACA ng. so

1.6

1.2

Theory Experiment O.OQJb at R=4.6xl06

100

Figure SO.- NACA 27-21b airfoil.

i

NACA

Wf. 21

w

l^e 21- NACA £V_£15 atrfoU wUh >{. •JC extensic

NACA Tig. 22

1.6

1.2

.8

.4

>'

\

HA :A OOI i NACA J7-012

1.2

a f \ (

0

;:A !A lli- na \

NACA 27-012

1.2 ~N "A f 4

HA: A 13- )12 \

NACA 9-012

1 .. .. 1 50 100 ftO

Figure 22.- Theoretical pressure dietrlbutions for the basic symmetrical sections.

100

31 2 NACA

*ig. 23

.0120

Reynolds number, millions

Figure 23.- Mini.•,, lrag Coefficient8

4-percent-thick airfoils.

NACA Jiff. 24

.0120

.0010

.0060

cd, min

.0350

.0040

.0020

A airfoil a, it, . NA 194209 184209

27i215

209 (i in cisp)0

1/8| 1/4

Turbulent friction

.~c 6Xt.

;Jcin

1 4 e

Reynolds number, millions 10

Figure 24.- Minimum drag c^effici ants.

i

NACA Jig. 25

.0120

.0100

.0080

5<W

.ooeo

.0040

.0020

NACA tirf oi L

27- 07,8-

16- 30 a

509 30&

o 1 •icti

ir.ar 1 iir. •ictio I

0

irfrulcht skih

a. <1

1/2

1/2

slcin

4 6 Htynolls iiui:.ber, liiilllons

10

Figure 20.- Minimum -it as- coefficients of 9-percent-thicK airfoils.

i NACA /ig. 26

Reynolds man', tr, millions

Figure 2G.- Minim-un dra£ coefficients of 12-perce:it-tniclc airfjile.

NACA Fig. 27

4 6 Reynolds nuuter, millions

"Figure 27,- Minimum drag coefficients of 15-percent-thicK airfoils.

I

N.A.C.A. Fig. 28

CO

OlfBy

N.A.C.A. Tig. ZS

(saa-ibap) °ja 'yzof+ojo a/buy 2i jr to oo *> <\j ^

\ 4u30J3d

~~ ^—-1.-4

(pjOUD JO PUS pjOMJOJ WOJ/)P-JOLjD

Of //// *° °!l°U

N.A.C.A. Fig. 30

°!t°y

N.A.C.A. Fig:. 31

to 9) ci

•; <\j nj nj nj "*i "i * ^ "^ 'J1 ^ ^ tf ^" > tj <n <\j -:

iP i\| io o -- in o «yr^i\|«5-ioio<o^t*\jri«>fficsto

(pjOLjC> JO puS P-JOMJOJ. LUOJj) ''

fl/7 '6/3Jp o; /y// yo Oj/oy

N.A.C.A. Fiff. 32

N.A.C.A. Fig. 33

(saa-ibap) °jo ' i/ao^o yo a/buy

5 8 <0 <\l

-v.

*o'jt/a/3/j/aOD bo-jp-a/?/o-/d uoifoas

CO Q <\j Jf to ^ ^ "5 "5 ~

'a'fuaiaij/aoa •&•».

% to 00 o on' a/7 'boJp Of //// ;o cy/cy

10 to 0)

isorafonra (otcifl) Jacobs, E. N.

AUTHOR(S)

DIVISION: Aerodynamics (2) SECTION: VJings and Airfoils (6) CROSS DEFERENCES: Airfoils, Laminar flow (OS257); Boundary layer - Calculation (18220); Airfoils - Design (08070)

ATO- 20^71* ORIG. AGENCY NUMBER

ACR-L-3U5

REVISION

AMER. TITLE; Preliminary report on Laminar-flow airfoils and new methods adopted for airfoil and boundary-layer investigations

FORG'N. TITLE,

ORIGINATING AGENCY: National Advisory Committee for Aeronautics, Washington, D. C. TRANSLATION:

COUNTRY U.S. Eng.

LANGUAGE FORG'NruSS U. S.CLASS Uhclass.

DATE Jun'lj9

PAGES 63

IU.US. FEATURES photo, tables, graphs

ASStfOACT Recent developments in airfoil-testing methods and fundamental air-flow investigations,

as applied to airfoils, are discussed. Preliminary test results, obtained under conditions relatively free from stream turbulence and other disturbances, are presented. Suitable airfoils and airfoil-design principles were developed to take advantage of the unusually extensive laminar boundary layers that may be maintained under the improved testing conditions. The results are of interest mainly in range of below 6,000,000.

NOTE: Requests for copies of this report must be addressed to: N.A.C.A., Washington, D. C.

T-2. HO. AIR MATERIEL COMMAND AlR "DECHNICAL ONDEX WRIGHT FIELD. OHIO. USAAF

mmrimm (DMRKT)

Jacobs, E. N.

AUTHOR(S)

DIVISION, Aerodynamics (2) SECTION: Wings and Airfoils (6) CROSS REFERENCES: Airfoils, Laminar flow (08257); Boundary layer - Calculation (18220)) Airfoils - Design (08070)

ATI- ZOUTh ORIG AGENCY NUMBER

ACR-L-j!*5 REVISION

r^oiT AME«. TITLE: Preliminary report on Laminar-flow airfoils and new methods adopted fo

FORG'N. TITLE and boundary-layer investigations

ORIGINATING AGENCY: National Advisory Committee for Aeronautics, Washington, D. C. TRANSLATION:

COUNTRY

U.S. LANGUAGE [FORG'NrUSS

Eng. U. {.CLASS.

Unclass. DATE

Jun'lfl PAGES

63 (LLUS. FEATURES

photo, tables, graphs ABSTRACT

Recent developments in airfoil-testing methods and fundamental air-flow investigations, as applied to airfoils, are discussed. Preliminary test results, obtained under conditions relatively free from stream turbulence and other disturbances, are presented* Suitable airfoils and airfoil-design principles were developed to take advantage of the unusually extensive laminar boundary layers that may be maintained under the improved testing conditions. The results are of interest mainly in range of below 6,000,000.

NOTE: Requests for copies of this report must be addressed to: N.A.C.A., Washington, D. C.

T-2, HO., AIR MATERIEL COMMAND TGTT ECHNI; AL INDEX WRIGHT FIELD. OHIO, USAAF W*-0-2l aUI tl .

reproduction quality notice · recent developments in airfoil-testing methods and fundamental...

Documents