REPORT No. 492 TESTS OF 16 RELATED AIRFOILS AT HIGH SPEEDS By Joun Stack and Aupenr E. vox Dosxnore SUMMARY In order to provide information that might lead to the development of better propeller sections, 18 related sym- ‘metrical airfoils hace been tested in the N.A.C.A. high- speed wind tunnel for a study of the effect of thickness form on the aerodynamic characteristics. ‘The thickness-form variables studied were the value of the macimum thickness, the position along the chord at swhich the maximum thickness occurs, and the value of the leading-edge radius. A system of equations was used to define the airfoil forms so that fair profiles having eystem- atic changes would be obtained. The basic thickness form ia very nearly the same as that chosen for the recent ‘investigation of a large number of related airfoils in the ariable-density wind tunnel (N.A.C.A. Technical Re- port No. 460). The tests wore conducted. through the low angle-of- attack range for speeds extending from 85 percent of that of sound to slightly in excess of the speed at which a.com- preseibility burble, or breakdown of flow, occurs. The corresponding Reynolds Number range ia $60,000 to 760,000. Because the Reynolds Number for the tests is somewhat lower than that at wohich most propellere oper- ate, and much lower than that at which airplane wings operate, the data are not directly applicable to many practical problems, but it is probable that some of the re- lations shown, particularly the relative effect of the shape changes as affected by compressibility, are valid at much ‘higher values of the Reynolds Number. The results ob- tained were applied to the design of three cambered airfoils swhich were tested as part of this investigation. ‘The principal factors affecting the choice of propeller sections are low drag at low and moderate lift coeficients and a late compressibility burble, that is, low drage at high apeeds. Considering these factors, the reaults indicate that the mazimur. thickness should be small and located at approximately 40 percent of the chord aft of the leading edge, Small eariations from the normal values for the leading-edge radius are shown to have mall effect on the aerodynamic characteristics. A comparison with similar tests of commonly used propeller sections indicates that at high speeds one of the cambered airfoila tested, the N.A.O.A. 2409-84, ia ouperior. ‘The results also in- dicate that some further improvement in airfoil shapes for high-speed applications may be expected. INTRODUCTION Experimental investigations of the relationship be- ‘tvoen airfoil shopes and airfoil nerodynamic character- istics have generally been made at some particular dynamic seale, or Reynolds Number, and usually at relatively low speeds. Becanso tho forces on an airfoil, aro affected by air compressibility, tho speed at which tests are made may become an important parameter in tho application of tho results. It has been shown that ‘the speed of flow exprossed in terms of the speed of wave propagetion, or the speed of sound, in the fluid is am index of the extent to which the flow is affected by com- pressibility, ‘Thus, the ratio of the flow velocity to the velocity of sound, V/V,, is a parameter indicative of flow pattern similarity in relation to compressibility effects just as the Reynolds Number is an index of the effects of viscosity. Thorofore, if the speed at which ‘the full-scale airfoil normally operates is greatly in ex- cess of the speed at which the model was tested, the test results may be subject to a correction for the effects of compressibility. ‘The importance of the compressibility effect cannot, be disregarded formany moder epplieations. Previous airfoil tests over wide speed ranges (see reference 1) indicated that for speeds in excess of 300 miles per hour the compressibility effect on tho airfoil characteristics may be largo. It is therefore necessary to investigate the relationship between airfoil shape and the aero- dynamic characteristics at high speeds for such appli- cations as the design of propellers, diving bombers, and high-speed racing airplanes. ‘The aerodynamic characteristics of airfoils may be considered as being dependent. on the thickness-to- chord ratio (hereinafter referred to as the “thickness”), the thickness distribution, and the mean-line shape. ‘Tho present investigation was mado to study the effects of changes in the thickness and tho thickness distribu- tion on the aerodynamic characteristics of sirfoils, particularly at high speeds, and to provide additional information for the study of compressibility phenom- ona, This information should lead to the design of better propeller sections. ‘The effects of these changes, were determined by tests over a wide speed rango of 13 symmetrical airfoils having systematic changes of three variables. ‘Theso variables are, for a fixed chord 339length, tho magnitude of the maximum thickness, the position of the maximum thickness, and the radius of, the leading edge. ‘Three cambered airfoils were also tested as a preliminary step in tho investigation of the offects of mean-lino shape on the aerodynamic charac teristics at high speeds. ‘The tests consisted of the measurement of the lift, drag, and moment about the quarter-chord axis of the models for a range of speeds extending from about 35 percent of the speed of sound to speeds slightly in excess of the speed at which the breakdown of flow corresponding to the compressibility burble occurred. ‘Tho tests were conducted in the N.A.C.A. high-speed wind tunnel during 1932-33. DESCRIPTION OF AIRFOILS ‘The variables herein considered as determining the thickness form are the maximum thickness, the posi- tion of tho maximum thickness, and the radius of the leading edge expressed in torms of tho chord. Theso REPORT NATIONAL ADVISORY COMMITTEE FOR AERONAUTICS DESIGN NUMBERS 008-83 9000-35, 013-83 9000-03, 009-83 0008-05. fo009-62 0000-2 000-64 000-34 (0000-85 2200-84 ‘000-88 2400-34 000-03 4400-34 In the design numbers given above, the first group of four digits gives tho camber and thickness designation and has the same signifieanco as tho airfoil dosign numbers given in reference 1; thot is, the first digit, indicates the mean camber in percont of the chord; ‘the second, the position of the camber in tenths of the chord aft of the leading edge; and the last two give the maximum thickness in percont of the chord. ‘The group of digits following the dash designate the thick- ness distribution. ‘The first digit designates the lond- ing-edgo radius and the second digit gives tho position ‘ovws 1--Dasi theses airborne gheped ind tunnel A.C. OTA) compared with bee tloknes desbaton fr N.A.O.A. fami ails (8A.0.A- 02 ress. parameters determining the thickness form so ex- pressed as ratios to tho chord will throughout this re- port be referred to simply as “thickness”, “position of maximum thickness”, and “leading-edge radius.” Arbitrary values of these three variables ware so choson as to provide systematic variations over the entire probable useful range of forms. ‘The resulting airfoil forms were defined by means of a system of equations to insure fairness of the profiles, and the coefficients of tho various terms of tho equations were determined from conditions imposed by the assumed values of the independent variables. ‘The basic form is shown in figure 1. On tho same figure the besic thickness distribution used in the investigation in reference 2 is, also shown. ‘The leading-edge radius, the slope at the tail, tho maximum thickness, the trailing-edge ordi- nate, and the position of tho maximum thickness were made the same as those of the basic form given in reference 2, which has been designated the “N.A.C.A. 0020” airfoil. Range of forms.—The range of forms investigated is shown by the airfoil design numbers in the following table: of the maximum thickness in tenths of the chord aft of tho leading edge. ‘The significance of the leading-edge radius dosigna- tion is given below: 0 designates sharp loading edge. 8 designates one-fourth normal lending-odgo radius. 6 designates normal leading-edge radius (the Tending-edge radius used in reference 2). 9 designates threo times normal leading-odgo radius or greater. ‘Tho leading-edge radius of the blunt-nosed airfoil used in this investigation is threo times tho normal value. ‘Thus, the N.A.C.A. 0009-64 is a 9 percent thick symmetrical airfoil having a normal leading-edge radius and its maximum thickness 40 percent of the chord aft of the leading edge. ‘The N.A.C.A. 2409-34 ir- foil has a maximum meen camber of 2 percent located at 40 percent of the chord, and is 9 percent thick. ‘The leading-edge radius is one-fourth of the normal ‘value and the maximum thickness is located at 40 por- cont of the chord aft of the lending edge.‘TESTS OF 16 RELATED AIRFOILS AT HIGH SPEEDS ‘Tho range of thickness ratios tosted in this investi- gation was small because it was considered necessary to show only how relations already found for the effect of thickness vatiation at low speeds are affected by compressibility, and also because the airfoils chosen for igh-speed application will of necessity be relatively thin, ‘The value 09 of the maximum thickness ratio of the airfoils designed for the study of variables other than the maximum thickness was chosen be- cause it is representative of the thickness range from which airfoil sections for propellers would probably be chosen, ‘The airfoil profiles are shown in figure 2, and the ordinates aro given in table I. Derivation of new thickness forms—The sirfoil forms tested in this investigation have beon defined by two equations. ‘Two equations were used rather than a single equation, because a single-equation aystom led to shape differences aft of the position of maximum thickness when the leading-edge radius was changed and also led to revorsals of curvature unless a very great number of terms wore used. If the chord is taken as the axis from 0 to 1, the ordinates y from tho leading edge to the position of maximum thickness aro given by an equation of the form aft aro given by an equation of the form tymdotd(l—2)+d,0—2)'+a(1—a> ‘The cooficients of the equation for the forward portion are determined from the following conditions: (0) Maximum thickness ‘m — y=0.1 (where m is the location of the ‘maximum thickness in terms of the chord measured from the leading edge) =0 (b) Leading-edge radius ‘The leading-edge radius is derived from eque- tion (1) and is $*, Values of a» chosen to give certain desired leading-edge radii aro shown in the following table: & 1 0008-63 012-63 | 0009-63, | 0009-62 | 0009-64 0009-65 0009-65 I 009-03 | 9009-33 | 0009-05 | 0009-35 | i | 2209-34, 209-34, (© Radius of curvature at tho point of maximum thickness Radius of curvature at 2=m, =m)? Re Tam) 05S : 4409-34, ama N.A0, 2 prolly eter tte NAO. iO ae NER342 (this value is derived from the equation for tho after portion of the airfoil and is the same for both equations at c=m). ‘The conditions that were taken to determine the coefficients for the equation for tho after portion are: (a) Maximum thickness dy, seo (c) Trailing-edge angle a grant) ‘The values of d; a8 a function of m, which were chosen to avoid reversals of curvature, are given in the follow- zl ing table: 20 abe Substitution of the coefficients derived from the foregoing conditions in equations (1) and (2) gives equations for symmetrical airfoils 20 percent thick. ‘Those coefficients are given in table TI. ‘The airfoil ordinates for any other thickness are determined by multiplying the ordinates derived from t the above equations by gf, where ti the airfoil thick- nest expressed as a faction of the chord. ‘The leading- dig radius for any thickness is given by 3] ai |” Derivation of cambered sirfoila—The three cam- bered airfoils were derived by combining one of the Dest thickmess forms, the 0009-34, with certain mean- Tine forms chosen from reference 2. The mean lines chosen have the 2200, 2400, and 4400 forms. The ‘methods of combining the thickness distribution with the meancline forms are given in detail in reference 2 APPARATUS AND METHOD Apparatus—A complete description of the high- speed wind tunnel and a detailed account of the method of conducting tests are given in reference 1. ‘Tho models used for this investigation wero of 2-inch chord and were mado of steel. ‘The method of con- structing the airfoils is described in reference 3. ‘Method of testing —The tests consisted of the meas- urement of the lift, drag, moment, end dynamic pres- sure for several speeds in the range extending from 35 percent of the speed of sound to speeds slightly in excess of that at which tho compressibility burble ocours. ‘Tho corresponding Reynolds Number range is from 350,000 to 750,000. ‘The angle-of-attack range for tho tests of the symmotrical airfoils extended from RRBPORT NATIONAL ADVISORY COMMITTEE FOR ABRONAUTICS —4° to 4°. ‘Tho additional tests required to determino the maximum lift coefficients would have unnecessarily prolonged the testing program because inferior airfoils for high-speed applications ean be detected by their carlior compressibility burble. ‘The tests on the three cambered airfoils were conducted through the low englo-of-attack rango, and one of tho throo which showed promise as a practical propeller section was chosen for tests throughout the complete angle-of- attack range. ‘The order of the tests was arranged, as far as practicable, so that the tests to determine the offects of a single variable wore made consecutively. RESULTS ‘Tho test results are presented graphically in figures 3 to 18. Each figure presents complete data for one airfoil for tho rango of angle of attack tested. Enoh curve shows the variation of one of the cosilicionts with V/V. for e given angle of attack. In the pre- sentation of the moment-coefiicient data the origin of the axes for each angle of attack has been raised above that for the previous angle of attack, so that the moment eurve for any angle may be easily distin- guished. ‘Phe data presented in figures 3 to 18 were cross plotted in figures 19 to 34 to show the aerodynamic characteristics of the airfoils in the usual form. Figures 35 to 88 show the offect of tho important shapo vari- ‘ables on the aerodynamic characteristics. A compari- son of the cambered airfoilsis given in figures 39 and 40. PRECISION ‘Tho various factors contributing to inaccuracy in those experiments may, in genoral, be classified under two divisions. ‘Tho first consists of systomatic and the second, of accidental errors. A detailed discussion of ‘the probable systematic errors is given in referenco 1. ‘Dho,accidental errors aro shown by tho scattering of the points on tho curves and by differences between original and check tests of the 0009-63 and 0008-65 sirfoils. Errors in the angle of attack may be as largo as 1°, owing partly to errors in mounting the airfoils and partly to dissymmetry of the symmetrical airfoils, ‘The balance and static-piate calibrations made bofore and aftar the tests checkod to within 1.5 porcent. Inne- curacies arising from other sources are within 0.005 for the lift coofficient, -£0.0008 for the drag coefficient, and £0,002 for the moment coefiicient. ‘The errors in tho results of the 0009-66 airfoil may be larger than the above-mentioned values because a special corree- tion was applied to these data to account for tho large dissymmetry of this airfoil. DISOUSSION ‘The data have been anelyzed to show primarily tho effects of shape changes on airfoil aerodynamio chernc- teristics at high speeds in order to provide informationTESTS OF 16 RELATED AIRFOILS AT HIGH SPEEDS 343, ele a Ansie of attack ease 35 4 oy a Ye 8 ot T Hy ; & = i "4 7 : Flower | § 5 1 ~ | “2 ao 8 : a oe eT a, wm a LUE | i sale of ottock fel i ese eu (| B38 g t i | {it fa | I & 1 { | igi 3 | re, 5 j eee | a a LA fi I of COC Pi ea | Lt i i LI os 28 79 rm tet of compre om he sro hr he NAO. li 4 ey ee y fobetort | AIS oo 3 : g aot {N13 8 i | = chodoobed lo tor oH + lo tar 5 T J TTT Eee Jorge lo ror 2 re To meDs a y a oe S28 ost ost 4 Phe e 12a 2 i eH bo s 7 } Coa i Bod He . é +e i bof SA pad |= | ad io : i i i cA i om Ta ttt fmaoty os ts wari deta ds na.o wwe 4 PAR | “caper oegs g : si ae 5. 4 g g ba 5 e FS § TREC THe 5 : & rT + g 5 im . a aS a mh HT 5! mae 4 = . re | ] a Ra re Beto peyote ets tt M.A a ; nl 4 ! deste ot s | mot ches 3 [oped %, i 12S «# : : : a P| 8 eT CS, LJ Re f a. 7 | tt 1| = § 1 1 meg sgt oat ate te wy mm ‘FooRE M-—teet a emprety on the ecodrnamls saat of tha N.A.O.A 8-8346 REPORT NATIONAL ADVISORY COMMITTEE FOR ABRONAUFICS “ = fn Angie, oF attook Fs . oe oF eee at pS oo 8, 4 z +2 ae s Lode 5 [| G 3 bor ; 2 POLE LIE? Sel beth |S g § & P/N ae a 3 bas 2 Blo he “oe Pt |e toe ow fal | 4 3 2- FZ ZF e ° We om 0 70 Tone S-Bte of ompriy eo tn eel hat th 6 we 4 a rl Angle, of attack |? 1 pee ee 3 § +e oe i : d & STA 829 aa > Somes i foal He ode Pet So} RoW Boers a 3 g IL] 8 lo tar de “T “Tens .- ap i 2] & Ras = ages oe aah oe a 8 ae To wr we owes 1h of onset eat wy are th AOA OS 12 4 eg! Angle of ottock Pile oF oes 7 0 oF 8 $ +2 oe & 8 3 s at £4 bu s g & o 2 gy ° + : 2 F i iy. om wn “@ ‘Fiooas Ett ef eomerntlty on the acodyamlecharstrs ofthe N.A..A, O9-S ll‘(TESTS OF 16 RELATED AIRFOILS ATT HIGH SPEEDS to bo used in developing better propeller sections. Bo- cause the Reynolds Numbor for the tests is somewhat lower than that at which most propellers operate, the data are not directly applicable to many propeller problems, but itis probable that some of the relations shown, particularly tho relative effect of the shape 347 creases uniformly with the thickness ratio of the air- foils at speeds below that of tho compressibility burble. Increasing tho thickness of an airfoil eauses the com- pressibility burble to occur at progressively lower speeds. The profle-drag coefficients for a lift coofi- cient of 0.4 show, in general, the same changes as the 6 10 a So ogi, of otto 3 oa ge { | = Q, MS I gfoaee Af g t Pert bod Seger ‘ aa 5 1 Begee | Pr 8 2 tor 22 Sac} =H er aad i : Bl Pesce el 2h, o se OF @ To 4 We A a {RoURE IS—Eaee of empreBty on te serdyaaml carci of be N.A.O.A 03 fl, changes as affected by compressibility, are valid at much higher values of the Reynolds Number. Tho affects of compressibility on profile drag are substantially in agreement, with the results shown in roference 1 and therefore are not discussed in de ‘Tho changes in tho drag coofficients aro small unt minimum profile-drag coefficients, excopt for a slight decrease in the profile-drag coefficient with increase of speed over the lower end of the speed rango. Figure 85 also shows that the speed at which tho rapid rise in the drag coefficient or the compressibility burble occurs, decreases as the lift is increased. Effect of maximum thickness position—Curves showing the variation of the minimum profle-drag co- 4 2 4 10) Angle of attack, ie ? xe OF CGE i Ye 8% = 3 & 5 u 3 Hie 8 $ 8 lore £9 Bor HL e.. a s LW Slo tor 2 -2| Eee] poset ion Pees ame en mere Ome eae foe os Ti We Ve wy & ° oons 18—Etet of compeusilty oa ho xryoamle charters of to N.ALC.A, 200-4 a speod corresponding to that of the compressibility burble is reached; the drag coefficient then repidly increases. Bffect of thickness.—The effects of thickness on the minimum profile-dreg coefficient and on the profile- drag coofficient for a lift coofficiant of 0.4 are shown in figure 35. ‘The minimum profle-drag coefficient in- officient with speed, for airfoils having various positions of the maximum thickness, are given in figure 35. ‘The N.A.C.A. 0009-64 hes the lowest minimum profile drag over the entire speed range, and elso has the highest speed for the compressibility burble. Airfoil having the position of maximum thickness forward or aft 0 the 40 percent location have progressively higher mini-15) 4 ast 8 Lift coefficient, G REPORT NATIONAL ADVISORY COMMITTEE FOR AERONAUTICS wal 9S Sib aie 7 Eodede 8 Profite-crag eoettilenh, Ge 8 ‘icone 1B of comproctlly on the aol careers of the N.A.0.A. 2109 eo ' so , Anais of ettock aa Pee nee | + ot vo Oe i Ss 4 £, | Sal We 1 3 : 2 I ‘ 5 dos J rt {b |5 d] | od osapeo ax pase FIA Flom a ; Glover ae | eT te r PGE oa ae Efepe be bel ¥_ fo tor nee ol . To fon tae BOB eg oe on a we ‘Moons Edo of sempreBity og the aeredyaumle carcass of the N.A.C.A Ah el,349 ‘TESTS OF 16 RELATED AIRFOILS AT HIGH SPENDS 70 © g Ss g ve c lo hy s+ £ @ 8 aa 00.609 %y yomryse0o ees oie Ubay on atbood non ls y “I E NS ug § jo 7 ny Saat: bis ay ‘Poca 2.—Aergaame charset fie NACA. OM ea. ‘roume 18—Asodyeala cir othe NAC. 8-8 ta 70 o 4 2 ‘g00.609 ‘20440 70 apbuy ‘yooyo 30 354y 9 wadiys00> “Botpropoud 2 Litt Coefficient C, ‘roves 22— serail characte be NAO. 00 al. a8 2 Litt eoofticiont ‘one 29—Asroynaal charter oth N-A.C-A 00 at.350 i 2 s s Chee, & 2 3s 3 2 Oe Litt Coefficient Provan 2\—Amodynanle curacao of tbe N.AO.A (00 fl — REPORT NATIONAL ADVISORY COMMOTTEE 70 FOR AERONAUTICS Angle of oftack, ‘degres 8 oostitiont by, 8 & 2 Mozent coon bau ca 02 #680 Litt Goefticient O, rooms 24—Aerodynamiesarcerae of the NACA 00-0 al, oe —" ee Soqreos, Angie of attack, Angla of attack, Sogcees, ef Meionh Cy 3 Profile-drog 2 eee Lith coefficient 6, ‘rovnn 2.—Aerodyamlo hanctrits ef the N.A.0.A. (O68 al fe 2 oe ost, Litt Coefficient, ‘a 25—Aerodyans carat fhe NA.C.A, 0-5 ar351 ‘TESTS OF 16 RELATED AIRFOILS AT HIGH SPEEDS 70 6 4 4 2 2 Litt Cootticient Litt Coetticiont, 6, 0 o = p 's0.60p ‘yoojs0 70 Buy 8 g % quepyso0> Bolpmasocg yiaion ‘icons 22—Aeroymum caret oie NACA, 00-48 ta web Say aoeauses eli May 2 ' i le 3 ‘ uf \ ft uu No ce i iL i p ‘seosta9 syoouie 40 abuy 3 g 8 "y quaoussoo* Pod 70 46 2 Litt Coctticiont ovnx 90-Asodynamle careers of tbo NA.C.A, O38 al.352 ‘mum profile drags and earlier compressibility burbles. Profile-drag coefficient curves for a lift coefficient of 0.4 aro also shown in figure 35. Differences in the profile-drag coefficient between the 20, 30, and 40 per- cent locations are very smell at the lower speeds. Air- foila having their maximum thicknesses located at 50 and 60 percent of the chord have tho highest drag for speeds up to approximately 65 percont of tho speed of sound. Athigher speeds the airfoil having the farthest forward location of the maximum thickness becomes the poorest, due to the earlier compressibility burble. Considering the speed rango as a whole, the 40 percent position seems to be tho optimum. REPORT NATIONAL ADVISORY COMMITEE FOR ABRONAUTICS Examination of the effects of variations of the load ing-edge radius for aivfoila having thoir maximum, thickness located at 0 percent of the chord shows 2 slight increase in tho minimam profile drag with {increase of tho leading-edgo radius. At higher lift coofficients, there is a very lange drag increase for the sharp leading edge. ‘The airfoils having tho leading- odge designations 3 and 6 show small differences. Bifect of leading-edge radius.—Figure 35 shows that | # the effects of changes in the loading-odgo radius on | $¥ mnjnimum profile drag are negligible excopt for very | 8 a BE gt a 7) | $e g Bex ef a, SE ° 4 { Ect L a a a 70. ltt Seefticiont ‘looue 3L—Aeodymuniechareat ofthe NA.O.A, oe srl Inge values of tho leading-edge radius. Airfoils hav- ing variations of the leading-edge radius from a sharp leading edge to the normal leading-edge radius have practically the same minimum profile drag over the entire speed range. An increase of the leading-edge radius to three times the normel value causes a rela- tively largo incresse in drag at the lower speeds, as wall as an earlier compressibility burblo, : ‘With incresso in lift, at lower speeds, the airfoil having a sharp leading edge hes the highest profile drag due to the rapid drag increase with angle of attack. ‘As the speed is incroased the drag of the N.A.C.A. 0009-98 becomes greater, because of its earlier com- pressibility burblo. Tho 0009-83 has the lowest profile drag. ‘Tho quarter normal leading-edge redius is thorefore the optimum value for the range of angle of attack tested. (Cpe ae eaeerel Litt Coetticiont 6. rovne #2—Aerodynaie cassettes of the N.A.O.A, 20-4 lol. ‘Tho lift cooflicionts for symmotrical airfoils in the usual working range can bo expressed in the following dar, C= Geta where 42 (the litteurve slope) depends on the shape of the airfoil and the flow parameters. For speods below that of the compressibility burble it, has boon shown theorotically in references 4 and 6 that, as first approximation, tho offect of compressibility on 1 lift is to increase 9! with the factor + ) ‘This factor has been substantiated experimentally for speeds below that at which the compressibility burble‘TESTS OF 16 RELATED AIRFOILS AT HIGH SPEEDS occurs, as shown in reference 1 and by figures 36, 37, and 38 of this report. At speeds in excess of thet at which tho compressibility burble occurs the value of 40: aecroasoe very repiily. de Eifect of thickness.—Tho effect of increasing thick- ness on {C+ i shown in figure 38. The ift-curve slope 4G. in gonerl, decreases with increase of Uhickmess, 0 re Angle of ottock, Gogrees, Lohr, gs g 0 2 4 8 tint costticienh. 6, ‘rauns 52—Aeroamal chart ofthe NALA. 240594 a result in ogreoment with previous low-speed tests. Figure 36 also shows that the speeds at which 22 attains its maximum value decrease progressively with increasing thickness. ‘Tho oarlier compressibility bur- ble for thicker airfoils shows clearly that the useful speed range decreases as the airfoil thickness increases. Effect of maximum thickness position—The effects of variation in the maximum-thickness position on 4Cé fro shown in figure 37. All the airfoils in this group 353 have the normal leading-edge radius and a maximum thickness ratio t/e of 0.09. At the lower speeds, it is, apparent that the value of {2 decronses almost uni- formly with rearward movement of the maximum thickness until the maximum thickness is located approximately at mid chord, Additional rearward movement tands to increase the valuo of 9C* stightiy. As the speeds are increased the airfoils having extreme rear locations of the maximum thickness are inferior because of an earlier compressibility burblo (fig. 37). ‘Movement of tho maximum thickness forward of the oo Angle of attack, Sogrees, 4 ier by Profile Ce cama Litt coefficient 6, ‘rooms 3—Auedyanmle erect of tm NAA. 0D ae 30 percent location tends to increase the magnitude of the compressibility effect as is shown by tho rapid riso at high spoods in the 9! curve for the 0009-62 airfoil. ‘The later compressibility burble shown by the 0009-63 and 0009-64 airfoils indicates that the airfoils having tho widest useful speed range will have their maximum thickness located between 30 to 40 percent of the chord aft of the leading edge. Eifect of leading-edge radius —The effects of varie- ations in the leading-edgo radius on tho lift-eurve slope for airfoils having tho maximum thickness located at 30 percent of the chord and also at 50 percent of the chord aft of the lending edge are shown in figure 38.354 REPORT NATIONAL ADVISORY COMMITTEE FOR AERONAUTICS | ahrindem drome op rote rag for 6,70. we I | \Voriotion with thickness| | [maca-oore-c3. of & -y & S [oT Westen boats : e & Bog ttt g PE i 3 = =| Besa I 3 & of Voriafion| wih reaeing-edge radius WAL 008" Te WALA MA.CA, 0009-93 en laca.c.a. 0009-03-- 3 ° 24 24 _ It 2 AoA, 00863-| Te \ vo 4G 3G Ec - ae ; i) . uc, | 2) 0009-63 ~| “2 ] 0009-65 Lyle WALA, o Mika, bees o. one meer omeere 2 Onur Prec 13 he ra 6 we We ‘che St—Rectoompreiyniesvelnn, Varn wth nes, Pax 5. of empreily on Meare ope. Valo with elon‘TESTS OF 16 RELATED AIRFOIS AT HIGH SPEEDS Considering first the airfoils having the maximum thickness located at 30 percont of the chord, it is apparent that the sharp leading edge is definitely bad. ‘Variations of the leading-edge radius, provided that the lending edge is rounded, heve apparently negligible effect on SC at lower speeds, in agrooment with the results of reference 2. Itis to be noted that the blunt- nosed aisfoll, the N.A.C.A. 0000-08, shows « very upid rsein 4 at high speeds, ‘This rapid risen UC fs also shown by the 0009-02 airfoil, which has its maximum thickness well forvard. It should also be noted that the compressibility burblo tends to occur progressively at lowor speeds as the leading-edge radius is ineroased. Similiarity of the effects of increasing lending-edge radius to the effects of inereusing thickness might have 24 355 trend of the effects of camber particularly at high speeds, a few airfoils heving certain camber variations and one of the best thickness forms have been tested. Selection of thickness form.—The choice of the best form for the thickness distribution was made prin- cipally on the basis of low drag and Inte compressibility burble. Within the lift-coefficiant range investigated, the tests of symmetrical airfoils indicate that, for an airfoil of medium thickness, the maximum thickness should be located at 40 percent of the chord aft of the leading edgo and the leading-edge radius should be one- quarter of the normal value. Thus, the 34-thickness distribution would seom to bo the best. ‘Tho N.A.C.A. (0009-34 and three cambered airfoils having this thick ness distribution were therefore built and tested. A comparison of the N.A.C.A. 0009-64 and N.A.C.A. (0009-34 airfoils (figs. 23 and 31) shows that, except for the slightly earlier compressibility burble of the Moximum thickness 0.3¢| redinah mkindas bse from leading 2ge 7rom jeading gage e 76 We 7 we louse Ee! of compresbilty ae cure slope Varlaton with nding alan been expected because of the higher induced velocities over the forward portion of the airfoil caused by such form changes. Results of the tests of airfoils with various leading-edge radii having their maximum thickness located at 50 percent of the chord aro in substantial agreement with the results obteined from the tests with the airfoils having their maximum thick- ness located at 30 percent of the chord, except that the compressibility burble shown by the airfoil having the normal leading-edge radius occurs at the higher speed. A dotailed study of camber variations has not been attempted as part of this investigation. However, in order to study the application of the date obtained from the symmetrical airfoil tests as well as to develop, if possible, by means of a few tests, # more efficient practical propeller section and to indicate the general N.A.C.A. 0009-84 airfoil, there is practically no difference in the minimum profile drag of these airfoils At moderato litt coofficionts the profile drag of the N.A.C.A. 0009-34 decroases slightly with increasing speed, wheroas the profilo drag of tho N.A.C.A. (0009-64 increases. Because of this difference, which may be attributed to the more gradual compressibility burble of the N.A.C.A. 0009-64, the N.A.C.A. 0009-84 has a lower profile drag over part of the speed range. As at minimum dreg, however, the N.A.C.A. 0009-84 hhas the earlier compressibility burble. From. this comparison it is apparent that the general superiority of either of these airfoils is difficult to establish. ‘This comparison shows, howover, that the shift of the maximum ordinate from the normal to the 40 percent location causes much greater improvement than ean be obtained from small changes in the leading-edge radius. ‘Future tests to study camber effects should probably356 include some tests of airfoils having both thickness forms. Camber form.—Tests of the cambered airfoils wore first conducted at low angles of attack. ‘The best of these airfoils on the besis of low profile drag and late compressibility burble was then chosen for testing at high angles of attack. ——_ sca, REPORT NATIONAL ADVISORY W/¥%20.70 5 Profite-drag soe icferth Exg os lint coe ‘owas 22 Campari Figure 39 gives a comparison of tho three eambered airfoils tested, the N.A.C.A. 2209-84, N.A.C.A. 2409- 34, and tho N.A.C.A, 4409-34. At low speeds the N.AC.A. 2409-34 hes the lowest minimum. profile Grag. The date thus far obtained indicate that the N.A.C.A. 4409-34 has the highest maximum lift co- efficient, but it has also the highest minimum profile COMMETTEE YOR AERONAUTICS drag. At high speeds, the N.A.C.A. 2409-84 becomes superior to both the other airfoils except for « narrow range of lift coofficients within which the N.A.C.A. 2200-84 has a smaller profile drag. ‘Tho results given in figures 31 to 34 showing tho effects of changes in the mean line on the aerodynamic characteristics are in general agreement with previous o 4 6 2 +4. 20 W/%e=040 Profiler SEHIoO W/%7083 48 Wiciert read camber al, 4 investigations at low speeds. ‘The changes in tho moment coofficient and in the angle of zero lift at low speed aro due principally to changes in tho mean lino, and are qualitatively in accord with thin-nirfoil theory. At high speeds larger differences aro apparent. These differences are due to the effects of camber changes on the speed at which the compressibility burble occurs.‘TESTS OF 16 RELATED AIRFOUS AT HIGH SPEEDS 357 Effects of compressibility —The effects of compres- | practically constant. ‘When the speed is increased sibility on tho lift and drag of the cambered airfoils | above that at which the compressibility burble occurs, are similar to those previously discussed for the sym- | the lift coefficiont decreases rapidly and the negative metrical airfoils. Some compressibility effects occur on | moment coefficient’ increases rapidly; consequently, cambered airfoils that are not shown by thesymmetrical | there is a large and rapid rearward movement of the airfoils. As the speed corresponding to the compres- | center of pressure. The magnitude of the change in sibility burble is exceeded, the angle of zero lift az, sud- | the moment coefficient over the low-speed part of the ° 4 4 0 4 2 08 Wen040 }-v/%=080 i oo i i] Ss i ieee —-— ans ! ess Ay a EY oe als Z AT. NACA. 2409-34 = 1 J 8 ge He we ° 20 weno} y%e=080 W%en08s i H Lie sf I f af ge i 1 ji Z $s $3 fr y / 123 Ba t if se 5 i & 08 Hy / 0 = a 4 "| ost oe ge SS TE “ay oO 4 a + o e 73? Pitt cos ricient t, nn Compu 904 30 N04 0a denly tends toward zero. An effective displacement of | range is sufficiently large to warrant full-scale studies the lift curve occurs. to obtain information for the design of wings for diving Over the lower portion of the speed range the nega- | bombers. tive moment coefficient increases with increnso of speed. | Comparison of cambered airfoils—A comparison of ‘The relative amount of the increase is approximately | the N.A.C.A. 2409-34 airfoil with the 309 and 3R9 the same as the increase in the lift coofficient. The | airfoils is given in figure 40. The date for the 3C9 and location of the center of pressure, therefore, remains | 3R9 have been interpolated from the results presented358 in reference 1. At tho lower speeds the N.A.C.A, 2409-34 has the lowest minimum drag, the lowest maximum lift, and therefore the smallest useful angular Tange. As the speed is increased above six-tenths of the speed of sourtd the N.A.C.A. 2409-84 airfoil becomes superior to both the 8C9 and the 3R9 airfoils, because of the larger compressibility offect on the C and R.air- foils. Itis probable that the low-speed maximum liftof the N.A.C.A. 2409-34 could be increased by increasing the leading-edge radius. Because the effect of the leading- edge radius on profile drag is small, it would seem that the 2409-64 airfoil section might ba better for applica- tions requiring a section to operate over a considerable range of the lift coefficient. In the development of cambered airfoils, of which the three tested form a preliminary step, the effects of shape variations on the maximum lift will be more thoroughly investigated, particularly at the lower speeds. concrustons ‘Tho principal factors affecting the choice of propeller sections are low drag at low and moderate lift cocffi- cients and a Jato compressibility burble; thet is, low drags at high speeds. Considoring these factors, these results indicate: 1, ‘The airfoil thickness should be small, 2. Tho best position for the maximum thickness is approximately 40 percent of the chord aft of the lending edge. REPORT NATIONAL ADVISORY COMMITTEE FOR ABRONAUTICS 8. Tho optimum values of tho leading-edge radius lio between 0.22 percent and 0.89 percent of the chord for airfoils of 9 percent thickness. 4. At high speeds the N.A.C.A, 2400-4 airfoil is superior to the commonly used propeller sections. ‘The restilts indicate that some further improvement may be expected. Lanorey Mzmoniat Annonavmica, Lanonarony, ‘Narrowat Apvisony Connrres ror AznoNaurics, Lanater Freup, Va., April £8, 1984. REFERENCES 4, Stack, John: ‘The N.A.C.A. High-Speed Wind Tunnel and “Tests of Six Propeller Sections. T.H. No. 403, N.A.C.A., 1938. 2, Jacobs, Eastman N,, Ward, Kenneth B., and Pinkerton, Robert M.: he Characteristion of 78 Related Alroil Sections from Tests in tho Variable-Density Wind Tunnel. T.R.No.460,N.A.C.A., 1933. 8. Jacobs, Eastman N., and Abbott, Tra H.: The N.A.C.A. ‘Variable-Density Wind Tunnel, 'T.R, No. 410, N.A.C.A., 1922, 4. Ackeret, J Uber Luftkrafte bel sehr grossen Geschwindl- ikelten insbesondere bel ebenen Strémungen, Helvetica Physica Acta, vol T, Faseiculus Quintus, 1028, pp. 301-822. 5. Glauort, H.: The Eaect of Compressibiity on tho Lift of an ‘Acrofoll. R. & M. No, 1186, British AR.C., 1928. TABLE T ORDINATES OF AIRFOILS Ben Roo Bou 3 |S |e | | | |e S| |= = eae a SS ee ee | = 28) | Ss || 8 || ea S| Ss || ea | ee ‘iim [ at | | ot | |e |e |e |e |e |e | || i 2S BEB | S| | 8] | | | | oo L bob bbb 2 vo | ae‘TESTS OF 16 RELATED AIRFOILS AT HIGH SPEEDS 359 ‘Desien 20, * l a | « « «fa a 4 are ‘.o02000 | 0. an9000 | 0.040625 | —a.orost2 28 011524