1、-* .NATIONAL ADVISORY COMMITTEEFOR AERONAUTICS .lV1/Iy2 1947TECHNICAU M3TENo. 1270EXPERIMENTAL AND CALCULATED CHARACTERISTICS OF SEVERALNACA 44-SERIES WINGS WITH ASPECT RATIOS OF 8, 10,.AND 12 AND TAPER RATIOS OF 2.5 AND 3.5tBy Robert H. Neely, Thomas V. Bollech,Gertrude C. Westrick, and Robert R. G
2、rahamLangley Memorial Aeronautical Laboratory;lhowever, the method of cctiaton whichallowed the use of nonlinear lift curves gave better agreement atanglas of attack. The two methods of cumulation gave differentmanwise lift Ustributions at maxiuzm lift. Comparisons made athigheusl values of Reynolds
3、 nuuiberindicate that the-valuee of Lhe maximumlift-drag ratio (L/D)= of “khesmooth wings _ticre%sEuZJLMJJincreasinqaspect z%dzlthrdughout the renge investigated in spite of theincreased drag of the thicker root sections associated with the hightir .aspect ratios. The values of (L/D)E for the wings
4、of taper ratio 35with leading-edge rouness indicated the came trend;however thevalues for the wings of taper ratio 2.5 with leading-edge roughnessProvided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-2 v . cA YO. 1270showed no gain when the aspect ratio wa
5、s increased from 10 to 12,apparentlybecause of the larger increment of profile drag due toroughness on the thicker root sections of the wing of aspectratio J.2. The decrement in (L/D)= due ta roughness was considerablylarger than the incrementdue to changing e a6pect mtio.The maximum lift coefficien
6、tsdecreased with increasing aspect ratio,mainly because ef the associated increase in root thickness-chordratio.INTRODUCTIONElementary aerodynamic cormiderations indicate that wings ofhigh aspect ratio are essential for efficient long-r-e airplanes.Structural considerationsfor such wings favor relat
7、ively thtck rootsections and higjhtaper ratios. Sections with large thickness-chordratios have high proflledrags, and high tar ratios usually resultin impaired stalMng characteristics. The aerodynamic advantagesof high aspect ratio are thus partly offset by a desn necessaryto satisfy the structural
8、requlroments. Although the main aero-dynamic effects of the desl variables are readily calculated bylifting-line theory from section characteristics,considerable doubtshave at times been expressed as to the absolute accuracy of the theoryfor determining an optimm combination of aspect ratio, per rat
9、io,and root thickneee-chordratto.An investigationhas accordingly been made in order (1) todemonstrate the corrdaticm of wing characteristicsobtalnedbycalculation and by wind-tunnel te%s and (2) to show s- of e,effects of aspect ratio, taper ratio, and root thickness-chordratioon aerodynamic characte
10、ristics. Seven unswept wings having NACA -seriessections, aspect ratios of 8, 10, and 12, and taper ratios of S!.5and 3.5 were studied. For six of the wings, tie ratio of span to root .thickness was held consmt at 35 so that the root thickness-chordratio increasedwith increasing aspect ratio and dec
11、reased withincreasing taper ratio. The seventh wing combined the lowest aspectratio and taper ratio with the highest root thickness-chordratio of tieother wings. The wing characbristics were calculated by en applicationof the lifting-line theory which allows the use of the nonlinear ectionlift curve
12、s as well.as by the umze,lapplication of the theory whichassumes linem lift curves.a71EU-M130LS% Mft coefficient (L/qS)r.Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-NACA TN HO a71 1270 3M.b.a71a15bDoMsc. .cYbwaa.section.liftmeffcient (Z/qc)drag c
13、oefficient (D/qO)profile-drag coeffelent (Do/qspitching-moment a?$Lent (M/qSc)Reynolds nurdxw f9vQM)Mach number (v/a)free-streaO dynamic pressure ()%V2mass density of airvelocity of air in free streamliftsection lifttotal drag ofprofile dragwingpitching mmnentwtng area vabout quarter-chord(Jb/2mean
14、aerod.mic chord =2b.local chorddistance fro!nplane of symmetvwing spancoefficien of viscosityvelocity of soundangle of attack of ting root chord,line)vdegrees.Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-4 NACA TN NO. 1270 “b% root chord.a71Ct . c
15、onstruction tip chord .,. . ,Et twistat construction tipSubscripts:min minimumMax maximum(L = 0) at zero 13.ftWINGSSeven wings of NACA 44-serik sections with apect ratios of 8, *10, and 12 end taper ratioa Qi2.5 and 3.5 were investigated. The .wings had straight tapered plan forms with parabolic tip
16、s extendingover the outer 5 percent of the semlspfin. !IMre was neltxr dihedral 4norswmgp; that is, the qwtbr-chsrd Uxm was perpendicular to the .pleme-”osymmetry. A typ5;c” ytiut is-sham In figure 1.Six of the wings were constructed to have a ratio of span toroot thickness of 35 with the root thict
17、iess-chordratio varyingbetween 0.147 and 0.2hj the seventh wing had a ratio of span to rootthickness of 23.3 with a root thicknesg-chordratio of 0.24. Thetip thickness-chordratio was 0.12for aJ.1wingo. meneional MMfor the wings are summarized in table I. The dedgnati.onfor the wingsIs formed from nm
18、ibers representing, consecutively,the taper ratio,aspect ratio, NACA airfoil series, and root thickness in percent chcrd.For example, in the designation 2.-8-44,16, the first number “2.5”represents the taper ratio, the numbe following the first dash “8”represent-sthe aspect ratio (approx.)jthe nucib
19、erfollcwhg the seccmldmh “44”represents the NACA airfoil series, andthe final number “16”represent-ethe root thiclmess in percentichor.The wings were twisted to improve the stalling characteristics.a71 For the wings of taper ratio 2,5, twistiwas introduced to give acl-margin of approximately0.1 at t
20、he 0.7 semispm station when Cz=was reached at some inboard wing section. (Seereferences 1 ma 2.)For the wings of taper ratio 3.5t calculations indicated thatithewashout necessary t.oprovide this cz-margin would cause excessiveinduced drag. The twist was therefore limited to s0 for this groupof wings
21、, .4fProvided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-NACA TN NOc 10 5The wings were constructed of laminated mahogany end werefinished with lacquer. Two surface conditions were provided for “testing. For the smoth-del condition, the wtigs were sanded
22、to an aerodynamically smooth finish. In order to eimulate a rough-model condition, a leading-edge roughness similar to that esti%lishedby the Langleytwo-dimensional.low-turbulencepressure tunnel wasused. The roughness was obtained%y application of No. 60 (0.011-inchdiameter) Carborundum grains to a
23、thin layer of.shellac along thecoete span over a surface length of 8 percent chord measuredfrom the leading edge on bath upper and lower surfaces. The grainswere intended to cover to 10 percent of the affected area. “Somedifficulty was encountered, however, in obtaining the same densityof the grains
24、 for all wings. The roughness on the 2.5-8-44,24wing was lighter tha.pon the other wings and the aerodynamic charac-teristics of this wing are believed to be somewhatbetter than wouldbe obtained with the desired roughness. METEOIH -. , ,Te8ksThe tests were conducted in,the Langley lg-foot.pressure t
25、unnelwith the “tingsmounted as ehown “infigure 2. For all tests the airin the tunnel was compressed a denaty,of approximately 0.0055 dugper cubic foot. The”tests were made ht several values of Reymolds aATutinumber between 1.5 x 106 and 7.0 x 106. The Mach number range wasfrom O.06 to 0.25. The rela
26、tion of Mach nuuiberto Remolds numberis given in figure 3. The relation of Mach nuuiberto Reynoldsnumber varied from wing to wing because the change in aspect ratiowas accomplished by changing the chord while the span constsnt washeld constant.Measurements of lift, drag, end pitching moment were mad
27、e overan angle-of-attack raq.ge from -4 through the angle of stall. Profile-drag measurements were made by wake surveys at 24 spanwise stationsat several angles of attack covering a lift-coefficient rsnge from Oto 1.0. Flow separation on the smooth wings was studiedby meansof wool tufts placed at 20
28、, 40, 60, , and 90 percent of the chordand spaced 6 inches spanwise on the upper surface of the wing. Nostudies were made of the flow se”paratiorion the rough wing.,.Corrections for support tare and interferece have W?n appliedto all force-test data. Jet-bodary and air-flow-tisalinementcorrections h
29、ave been applled to the angle of attack end d!$agcoefficient:Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-6 NACA TN Iio. l!270 *Calculations .*The characteristic of the wings were calculated from two-dimensional airfoil data by the lift-line theor
30、y. The requiredairfoil section characteristicsatopriate Reynolds numbers wereobtained f.rtiunTubliehed data from the Lszey t-wodirneni-mallow-imrbulencepressure tunnel. These section data were obtifned at aMach nu?ibernot exceeding0.17, HO that compressibilityeffucts arebelieved to be negligible. Th
31、e section dab” for the rough conditionswere obtained for two sizes of Carborundum grains so that the effectof the variation of relative grain size across the span of the taperedwings could be taken into account in the cal.oulatims for the wingswith leading-edgeroughness. Lift induced drag aracterfst
32、icswere deta?ninedby a generalized ayplicatio”nof tho lifting-linetheory which allows the use of-nonlinear section lift curves end bythe usual.applicationwhich assumes linear lift curves. A correctionto the lifting-line theory for the effect of chord of a ftiite-span wing was made by app3.yingthe ed
33、ge-velocityfactor given inrf3ference3. The profile-drag and”pitihing-moment coefficientswereobtained by using section coefficients at the corresponding sectionlift coefficientsand integrating the loadings across the span,The procedure by which the wing characteristicswere computed isgiven in detail
34、In reference 4. For the sake of brevity, the twoapplications of the theory mentioned previously me hereinafterreferred to as the !genera3.izedmethod;and the “linearizedmethod”.Aerodynamic characteristicsfor the wings of taper ratfo 2.5were calculatedby both the generalized and linearizedmethods fort
35、he smoath-model condition and by the generalized method for therough-mtiel conditi. For the wings of taper-ratio 3.5, thecharacteristicswere calculated only for the smooth-model condttionby the generalizedmethod.RESULTS AN33DISCUSSIONComparison of Exyerimmtal andCalculated CharacteristicsThe experim
36、ental.end calculated Ilft, drag, and pitching-momentcharacteristicsfor the wings of taper ratio2.and 3.5 are presentedin fives 4 to 10 for the smooth-tiel condltlon. The experj.menland calculated lift and drag characteristicsfor the wings of taperratio 2.5 are given in fures 11 to 14 for the rougk-m
37、odel condition.Some of the importantresults of the comparisms are summarized in,.9,Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-.mm m Jio, 1270 7. .tables II and 111. For better accuracy, the exrimental valuesof meximu Mf t-dragratio (L/D)- given
38、in these tables wereread from faired curves of (L/D)_ against Rewelds number.Typical calculated spsnwise distributions of section lift coefficients”at the predicted maximqn lift, for estimating stall characteristics,are given in figure 15. Experimental stall characteristicsderived,fromtuft studies a
39、ro shown in figures 16 ard 17 for all smooth w-s.In the linear lift-curve range, the characteristics calculatedbyeither the generalized or the liriearizeamcthodwouldbe expecbdto be the same. Differences in lift-curve slope and induced-dragcoefficientswere obtained,however, and ere attributed to inac
40、curacies.incomputing that erose In reading, fairing, end integrating plottedcurves.prw.-A comyarisonof the calculated.and expcrimen%al total-drag curves for the snioothwings (figs. h,to 10) shows that goodagreement was obtainod at low lift coefficients. Less satisfactoryagreement was obtained at hig
41、her,lift coej,cients where the calculatedaragwas generally lower then the experimental drag. This effect wasmost pronounced for Wings.of aspect ratio 8. As would be expected.,the sauieresults are shown in a ccmparieon oftho calculated ana .experimental grotile-drag coefficients. (Force-test profLie-
42、dragcoefficients were deherminod by subtrating the induced drag coef-ficients obtained by calculation frcm the total drag coefficientsmaswed by fm-ce tests.) The test values detmned by wake surveys,however, are in excellent agreent with the calculated values. Possiblereasons for discrepancy between
43、force-test profile drag and calculatedand welceysurveydrag are (1) errors in corrections for support tare,interference, and stream misalinement, (2) inaccuracies in calculatinginduced drag and (3) inaccuracies in evaluating the drag at the wingtip from section data or wake surveys.Generally speaking
44、, the agreement between calculated end experi-mental drag for the rough cauditim (figs. U. to 14) was about the sameas for the smooth conditionbut waa less consistent. In addition to thesources of errors menticmed before errors “inprofile drag for the rough.condition can easily arise from (l inaccur
45、ate simulation of desireroughness in the wi tests and (2) inaccuracies in accounting for grainsize in the calculatj.cme.These errors would also influence thellft characteristics.I?orwhgs of the type investigated, the value of (L/D)- is apredominant factor in determining the odmum design. Aa indj.cat
46、eIn tables 11 andIII, the calculated values of (L/D) were, for thecase giving the greatest descrepancq within 7 percent of the experimentalProvided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-valuee. l!%omthe preceding Mscuesion Qf possible errors in thed
47、ehmmination of drag, even this larest clifferenoe %etwean calculated a71and experimental (L/D)ti appears reasonable.Lift.-The “differencesbetween “tievalues of maximum liftcoeffi,cientvCL obtatned from tests and frcm calculationby themaxgeneralizemethod (tablesII and HI) mnged however, a ntore quant
48、itative discussion of the agreement is not.a71a15.Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-NACA TN No. 1270 9,.it maybe noted, howovorj that the roughness was somewhat extreme. Generally,there was little difference in (L/D)_ for correspondingwfngs of_taper ratio 2.5 gind3.5 in the smooth conditionbut in the roughcondition the values of (L/D)-Y for the wings of taper ratio 3.5 wore-consi.stentlyhigher than these for the wings of taper ratio 2.5. Thisdifference was probably due to tiie larger ef
