1、a71REPORT No. 530CHARACTERISTICS OF THE N. A. C. A. 23012 AIRFOIL FROM TESTSFULL-SCALE AND VAEL4BLEDENSITY TUNNELS. By EASTMAIJN. JACOBSand WILLIAM C. CLAYSUMMARYThis report giaes the results of tests in the N. A. C. A.f+scale and variable-dendy tunnei% of a new wingsection, the N. A. 0. A. %9019,wh
2、ich is m of the morepromising of an em%nded8erk8 of relC. A. A312”, was made at the Ninth Annual AircraftEngineering Research Conference in My 1934.At the subsequent request of the Bureau of Aero-nautic, Navy Department, a 6- by 36-foot model ofthe N. A. C. A. 23012 airfoil was tested in the N. A.C.
3、 A. full-scale tunnel to verify the aerodynamicoharaotaristics found for this airfoil in the variable-demitg tunnel. This test was made possible throughthe cooperation of the Chance Vought Corporation,who constructed the wing and supplied it to the Com-mittee for the purpose. The present report has
4、beenprepared to present and compare the results of thetm”tsof the N. A. C. A. 23012 motion made in theN. A. C. A. variable-density and fulkcale tunnels andto compare the results with those for well-knownsections.435Provided by IHSNot for ResaleNo reproduction or networking permitted without license
5、from IHS-,-,-cohmmrrm FoR tiONATJTTCS436 REPORT NATIONAL ADVISORY.CbOrct.442.0 .4018 .361.6 .32 QL4 28;2 I I i I I I Ic.pl I 1/IIlfkt I I I I .? -5 , , , ll 1 I- I I I I I I I I I I I I I i I wnI I 1. I 1, I I I ! II I I I I I I I I I r O?,.2 04ti-00-.2-.4t! thrttis, the mean line is stiaight from t
6、his point to thetrailing edge. The 230 mean line has its maximumcamber at a position 0.15c behind the leading edge.The camber is not exactly 2 percent but was deter-mined by the condition that the ideal angle of attackfor the mean line shouId correspond to a lift coefficientof 0.3, a value correspon
7、ding approximately to theusual conditions of high-speed or cruising flight. TheN. A. C. A. 23012 airfoil results born the combinationof the 230 mean line with the usual N. A. C. A. thick-ness distribution of 0.12c maximum thickness by themethod described in reference 1. The airfoil profile anda tabl
8、e of ordinates at standard stations are presentedin figure 1. In order to give a basis for the develop-ment of related airfoilE of diflerent thiclmwsw, theordirmtesy of the N. A. C. A. 230 mean line we givenas follows:Nose, from z=O to x=m!/=; W-3mo?+m2(3m)3Tail, from x=m to x=1where, for the 230 me
9、an line, m= O.2025and k= 16.957.VARL4ELE-DENSITY-TUNNEL TESTS AND RESULTSRoutine measurements of lift, drag, and pitchingmoment were originally made at n Reynolds Numberof approximately 3,000,000 to compare the vtiousairfoils of the forward-camber series under the con-ditions of a standard 20-atmosp
10、here test in thevmiabledensi tunnel. Later the N. A. C. A. 23012airfoil was reheatedm a pfut of a general invcdigationof scale effect. The data presented in this report weretaken from the latter twts which were made at severalvalues of the Reynolds Number between 42,400 and3,090,000.The test results
11、 obtained in connection with theforward-camber airfoil investigation, as well as thecomplete remits of the scale-effect investigation, areomitted from this report but both sets of results willappear subsequently in reports on the respective sub-jects. Complete results are given, however, gc.a -CDO(C
12、L=O)- :gC%(CL-l)- .- _:hors dnc)Zup IiIntern%.mm+aCtJmaL m.Mma.m+. 0310FIJLIACALE-TUNNEL TESTS AND RESULTSA description of the full-scale m“ndtunnel and equip-ment is given in reference 3. The N. A. C. A. 23012airfoil was mounted in the tunnel on two supportsFIGURE3.-The N. L O. A. !a312a.!rfoilmoun
13、tedin the fnIlaale wind tunnel.that attached to the one-quarter-chord point (fig. 3).The genemil arrangement was similar to that used intesting a seriw of Clark Y airfoils (reference 4).The airfoil had a chord of 6 feet and a span of 36feet. The frame was constructed of wood and cov-ered with sheet
14、aluminum. The surface was smoothand the section throughout was not in error by morethsn +0.06 of an inch from the speciiied ordinates.The lift, drag, and pitching moments were measuredthroughout a range of augles of attack from 8Provided by IHSNot for ResaleNo reproduction or networking permitted wi
15、thout license from IHS-,-,-438 REPORT NATIONAL ADVISORY COMMI!IWDE FOR AERONAUTICSto 25. These tests were made at 5 d.itlerent airspeeds between 30 and 75 miles per hour correspondingto values of the Reynolds Number between 1,600,000and 4,500,000. The maximum lift was not measuredat speeds above 75
16、miles per hour as the wing was notdesigned for the loads under these conditions. Addi-tional tests to determine the scale effect on minimumdrag were made at several speeds up to 120 miksper hour corresponding to a Reynolds Number of6,600,000.The interference of the airfoil supports upon the air-foil
17、 was determined by adding a duplicate supportingare given for the airfoil of infinite aspect ratio. Valuesof the pitching-moment coefficient about the aero-dynamic center, C.=.O.,are considered independent ofaspect ratio and are tabulatwd against 0 The loca-tion of the aerodynamic center (z, y) is g
18、iven as afraction of the chord ahead and above the quarter-chord point. A typical plot of the dnta from table VIis given in figure 4.Curves summarizing variations of these principalcharacteristics that change with Reynolds Number aregiven in figures 5 I%9. Curves obtainod from similarfull-scale-tunn
19、el tests on the Clark Y airfoil areord .13 52.12 48./ 4448 .10 4044 .09 T.%?Qy ebwall ef;ect. .4o-8-404 8 12 16 20 24 28 32I , , I t , , , I I 1 , N I Riiili (j-.2 -oo 4:-.3 -12E-.4$-16:6+:20 .2 .4 .6 .8 LO L2 14 16Angle of offock, ct degrees) Lift coefficient CLQTJIIE 4.TheN. A. 0. A. T301!2ahtoil.
20、 Rdl+walewindtunnel.strut at the center of the wing. This “dummy” sup-port was not connected to the airfoil or to the balanceand all change-sin the measured forces with the strutin place could be attributed to its interference. Dou-bling the effect of this single dummy support wasconsidered to accou
21、nt for the total interference of thetwo airfoil supports. All the data are corrected forwind-tunnel eflects and tares. The corrections arethe same as those used for the corresponding Clark Yairfoil (reference 4).The results of the full-tale-tunnel tesb of theN. A. C. A. 23012 airfoil are given in ta
22、bles IV to VIII.The values of C., a, C!=,LID, and c. p. me tabulatedfor the airfoil of aspect ratio 6 and values of and Cwpresented in these figures for purposes of comparison.These curves are presented in semilogaritlugic form toassistin extrapolation to higher valuea of the ReynoldsNumber. Figure
23、5 shows the variation of the maxi-mum lift coefficient for the two airfoils; the scale effecton the angle of attack at zero lift for the airfoil sectionis show in figure 6; figure 7 gives the effect of Rey- nolds Numb: on he sloe of the profile-lift curve;rmd figures 8 and 9 show, whereas the N. A.
24、C. A. 23012isunaffected by chaesin Reynolds Number. At zerolift a huge adverse gradient of pressure exists at theforward portion of the lower surface of the Clark Ythat probably results in an early disturbance of theI 1 I I I 1 I 1 I 1 1 1 I I 1 I I I 1 I IoI 2 4 6 8 10 20 xI06Reynolds NumberFIGVES7
25、.-uftQmv0 .dofk3.vdn titbmOm UMknomkm fUmfulkalo wfndtemd.flow at the leading edge (reference 4). This conditionof flow has a critical effect on the angle of zero lift andvaries considerably with Reynolds Number. TheN. A. C. A. 23012 airfoil has much less oamber thanthe Clark Y and the general profi
26、le, which is morenearly symmetrical, sets up a flow about the leadingProvided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-. _ _440 REPOILT NATIONAL ADVISORY COMMIT13EE FOR AJ?IRONAUTK!$3edge that is not critical; hence, the effects of male 01the angle of
27、zero lift should be small. This view %supported by the tests in the full-scale and variabledenti tunneIs.Figure 7 shows that the slope of the lift curve for thfN. A. C. A. 23012 airfoil is slightly higher than thalfor the Clark Y. Both sets of results indicate thalthe lift-curve slope increases slig
28、htly with ReynokbNumber.The curves of drag coefficient at zero lift (fig. 8)and minimum profihdrag coefficient (fig. 9) show thatthe drag of the N. A. C. A. 23012 airfoil is deii.niklylower than that of the Clark Y. These ures .012N“1 2 46 8 10 20 XIOSRe yriolds NumberFIGmE wherea!sin fLight thetwo
29、conditions occur at different air speeds. Thecomparative ratios indicate, however, that the speedrange of the new airfoil is much better than that ofthe Clark Y. As the result of the smaller mmber ofthe N. A. C. A. 23012 as compared with the Clark Y,C.On,thelift coefficient cmm.spendingto the minimu
30、m-proiih+d.rag coefficient, might be expected to be con-siderably less. Airfoils such as the N. A. C. A. 23012having the camber well forward tend, however, to havehigher optimum lift coefficients than airfoils with usualmean-line shapes. Actually, table I indicates that theoptimum lift coefficiemk f
31、or the two sections are nearlyequal.TABLE IFULL-SCALE WIND-TUNNEL TESTS COMPARINGN. A. C. A. 23012AND CLARK Y AIRFOILSAt R. N. = 4,6WYJcbamet81i9uo N. #. AcL-”-aJ+(d )- :dy, mrdegea)_. . .a- .101.%.- .-. - .-.- -.-. .MmCLWe- 1.19cq*,.-i.-i- -1. mI-.1.016Aerdynmrf ;center -n.JcOlerkY1.47-5. b.Q33.Wa1
32、.20-1.0761.0-26cf IA, .1-m” 161Cn.b - . .W9tiDmu -!250CLat(.u_-_-:g:doP:;:1No m$;+:z:=;: d “:- -The method of comparison employed utilizes theconcept of an efkctive Reynolds Number in order toallow for the effects of the turbulence present in thewind tunnels. This method, which was first proposedin
33、reference 5 and is discussed in the succeeding para-graphs, appears to be the best at present available forthe interpretation of wind-tunnel rwults as applied toflight.Marked scale effects, such as the ropid decrease ofdrag coe.tlicientwith Reynolds Number for the sphere,the rapid increase of the ma
34、ximum lift coefficient forsome airfoils, and the increase of drag coefficient forbut adequate data on mtium liftcoefficients arenot a,vaihtblefor making the comparisonbetween both the full-scale tunnel and the variable-density tunnel and flight by this method. A value ofthe factor of 2.4 was tentati
35、vely established betweenthe variable-density tunnel and the full-scale tunnelby a comparison of tests of Clark Y airfoils in bothtunnels. This value was employed im reference 5,assuming the factor for the full-scale tunnel to beunity (no turbulence).The assumption that the factor is unity for the fu
36、U-scale tunnel is approximately correct because dif-ferences in the turbulence between the full-scaletunnel and flight produce only small chaqges in thea:.2.-0.N.8, the actual values derived from spheretests are, however, dependent on the size of the spheresemployed.The results of the test at a give
37、n Reynolds Numbermight be directly applied at the higher effectiveReynolds Number; however, one change for which ap-proximate allowance may be made is to be expectid inpassing to the higher Reynolds Number. The part ofCOMMITTEE FOR AERONAUTICSthe drag associated with skin friction is known to de-cre
38、asewith the Reynolds Number. Therefore, althoughthe conditions as applying to the trrmsitionfrom lam-inar to turbulent flow may be considered as reproduc-ing those at the higher effective Reynolds Number, thevalue of the drag coefficient should be reduced in prlss-ing to the effective Reynolds Numbe
39、r. The actut-dvalue of this increment that should be subtracted issomewhat uncertain, but a value determined as sug-gested in reference 5 is used in this report for correctingthe variabledensity-tunnel results. The evaluation ofthe increment is based on the assumption that at thehigher values of the
40、 Reynolds Number encountered inflight, when the profle-drag coefficient is of importance,most of the profile drhg is due to skin friction from theturbulent boundary layer. The increment may thenredetermined from Prandtls analysis of the completelyturbulent skin-friction layer (reference 6) aa theamo
41、unt by which the skin-frictiondrag coefficientdecreases in the Reynolds Number range from thetest Reynolds Number to the effective ReynoldsNumber. Thus, when the standard airfoil test remdtafrom the variable-density tunnel at a test ReynoldsNumber of approximately 3,000,000 me applied toflight at th
42、e effective Reynolds Number of approxi-mately 8,000,000, the measured profile-drag coefficientsshould be corrected by deducting the increment 0.0011,It should be emphasized that the values employedin this report for both the turbulence factor and thedrag increment should be considered as only tentat
43、iveapproximation. The values may be revised as theresult of further twts now on the program at theCommittees laboratory. In particular, the fact thatthe skin-friction coefficient for airfoils tends to behigher than for flat plates (upon which the presentvalue of drag increment is based) agrees with
44、thepresent results in indicating that the drag incrementmay be too low.The comparison between the profile-drag resultsfrom the two tunnels may be made on the abovo-described basis by comparing the dotted curve infigure 2 with the profile-drag curve from the full-scale tunnel in iigure 4, although th
45、e values of theeffective Reynolds Number differ slightly. A bettercomparison is afforded by the curves in figures 10 and11 representing variations of certain characteristicswith the effective Reynolds Number. It will be notedthat the results from the full-scale tunnel indicatesomewhat lower profledr
46、ag coefficients but that thediilerences are smaller at zero lift where the resultsare more reliable owing to the absence of severnl moreor less uncertain corrections involved in deducing theprofile-drag coefficient when the airfoil is developinglift.The values of the maximum lift coefficient are com
47、-pared in figure 12 by means of curves representingvariations with the ,Reynolds Number. The agree-Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-CHARACTERISTICS OF THE N. A. C. A. 23012 AIRFOIL 943ment between the results from the two tunnels, con-
48、sidering the diihculties of mwumrement, is reasonablysatisfactory. The small discrepancy that remains mayindicate either that the value of the turbulence factorshould be modified or possibly that an incrementcorresponding to that used with the drag should beemployed.For the remaining characteristics, tabular valuesmay be directly compared. The results from bothtunnels agree in indicating that with
