NASA NACA-TR-1109-1952 Experimental investigation of base pressure on blunt-trailing-edge wings at supersonic velocities《在超音速下 钝后缘机翼基准压力的实验性研究》.pdf

1、REPORT 1109EXPERIMENTAL INVESTIGATION OF BASE PRESSURE ONBLUNT-TRAILING-EDGE WINGS ATSUPERSONIC VELOCITIES By DEAN R. CEAPM.W, lVII.LIAM R. WIMBBOW, and ROBERT H. KESTEESUMMARYMeawrement of ha-se prewwre are presented jor W blunt-trailing+dge wings huring an apect ralio of 3.0 and cariowaigtx”l proj

2、ile8. The different profiles comprised thicknewratios between O.OJ and 0.10, hoattail ange between ATand 20, and ratio8 oftrailing-edge thickne8s to airfoil thicknes8between 0.2 and 1.0. The tests were conducted at Mach number8( 125, 1.5, $.0, and 3.1. For each .fach number, the Reynoid8number and a

3、ngle ofattackwere mied, The lmce+dReynoldsnumber investigated uxM 03 X 1P and the highad was 3.5 X 1(P.Uea.wrement8 on each uing were obtained eparately withturbulent flow and Luninar no in the boundary layer. Span-wie mrreys of the base pressure were conducted on sereralu*ing*.The rad?a ui.th turbu

4、lent bounday4ayer jlow 8hWed onlywall e.fect8 on ba8e pressure ofuwiuiions in Reynold8 number,airfoil prole shape, boattail angle, and angle of attach-.Theprincipal twiable ajecting the baae pre8eure jor turbulent Jowvw the Mach number. At the high.wi Mach number inreA-gated (3.1), the ratio of boun

5、dary-hqw thknew to trailing-e(ige thickne88 ako afected the base pressure significantly.The redt.s obtained with laminar hmdary-fayer JOE tothe trailing edge showed that the effect. of Reynode number onba+v pre%ure Wz3 iarge. In all but a fe exceptional ca8e8the t#ect8 on base pressure of m-iations

6、in angle of attack andin pro$e shape upstream of tb base were appretible thoughnot large. The prinm.pal oariable aecting the base pre8mreJw luminar JOW was tb ratio of boundaydayer thuknes8 totrailing-edge thiekrtew+.F,r a few exceptional ca8es inrolning laminar Jow to thetrailing edge, the e$ects o

7、n bme pre8mre of rariation8 in pro$led.ape, boaituil angle, and angle of atiack uwe found to beunueualy urge. In mch cases the miution of he pre8wirewith ange of atiack wm discontinuous and exhibited a h.yster-twi. Stroboscopic schlieren ob8ermtiWL8 at a Mach numbertLf 1.6 indicated that these appar

8、e nty 8pecial phenomena uwra.wwciated with a rortex trail of relatively high frequeney.INTRODUCTIONIn comparison to the numerous base pressure investiga-tions conducted in the past on bodies of revolution, therehave been relatively few such investigations conducted ontwodimensional airfoik Some meas

9、urements of base pres-sure on vredge+ype profdes have been reported in referencesIauw8Weah-ACA Th- MU, “Ex.srbnentd In=srigsthn d Bzse Pre=ure on Blumc-lkgUlng.Edge lVLngs at SupersonicVekwitk” by Dean E. CbsprruU WIJlfsm R. Wfm.brow, and EcMrt H. Keater. 1352.2724*54 ._81, 2, and 3. These exist dat

10、a, however, are inadequatefor engineering purposes. Without considerable expwi-mentrd information on base pre.wme, the base drag cannotbe estimated for a given airfoil profde at given flight con-ditions.Recently interest has developed in bhmt-trailing-edge air-foils because of certain structural and

11、 aerodamic advan-tages at high flight velocities. In particular, it has beenfound that. tit supersonic velocities a properly designedblut-trailkyg-edge airfoil can have less drag and a greaterIift-curve slope than a sharp-traihng-edge airfoiI havkg thesame strength or stifhwss. A method of determini

12、ng theairfoil profiIe having the Ieast possible pre=ure drag hasbeen developed in reference 4, bu this method requires aknowledge of the base pressure for any given set of designlIight conditions. Siice the available base pressure data wemeager, the purpose of the present instigation was to obtainin

13、formation on the effects of Mach number, ReynoIds num-ber, type of boundary-Iayer flow-,and a.irfoiIprofile shape onthe base pressure of blunt-trailing-edge -wings. Quantitativeinformation on these effects is particulady important at lowand moderate supersonic elocities because the base drag atthese

14、 -reIocit.iescan contrl%ute the major portion of the totalprofile drag. The base drag of a $percent-thkk wedge air-foiI at a Mach number of 1.5, for example, amounts toapproximately thre+fourths of the totaI profiIe drag.NOTATIONairfoil chordTortes frequencytraiIiug-edge thicknessstatic pressureMach

15、 numberReynolds numbermaximum airfoiI thickness-reIocityangle of attackboattail angleboundary-layer thicknesstrailing+dge beveI angle, measured between traiHng-edge pIane and plane normal to chordSUBSCRIPTSbasefree streamSPECIAL NOTATIONrounded ridge lines when added either after the identifi-cation

16、 number of a wing or after a symboI in a figurelegend1145Provided by IHS Not for ResaleNo reproduction or networking permitted without license from IHS-,-,-1146 REPORT 1108-NATIONAL ADVISORY COITTEE FOR AERONAUTICSAPPARATUS AND TEST METHODSWIND TUNNEMThe experimental investigation was conducted in t

17、heAmes 1- by 3-foot supw-sonicwind tunncls No, 1 rmd No. 2.The ho. 1 wind tunnel is of thu closed-circuit, continuous-operation typo and is oquippwl with a flexiblo-plnte nozzlethat provides a vmiation of Mach uumbcr from 1.2 to 2.2.The total pressure in the tunnel can be varied to providoReynolds n

18、umbers from 0.2 to 1.7 million based on tho3-iich chord of the models omploycd in this investigation.The No. 2 wind tunnel is of thti nonreturn, intermitt.ont-operation type and is also cquippwl with a flexible-platenozzle that provides a variation of Mach number fromabout 1.2 to 3.8. The roscrvoir

19、pressure can he varied toprovide a variation in Reynolds number.Tho water content of the air in both the 1- by 3-footwind tunnels is maintained at loss than 0.0003 pound ofwater per pound of dry air; consequently, the oflcct ofhumidity on the flow is negligible.MODELSFifty-fivo wings with rectangula

20、r plan forms and blunLtraiIing edges were employed in this investigation. Dataare presc.ntcdfor 29 of those wings; the others exhibited thesame propmties as the wings for which datti are presentdAll these wings were mado of steel with a sptin of 9 inchesand a chord of 3 inches. Originally each had a

21、n orificolocated in tho blunt traiIing edge 3j4 inches inboard fromone wing tip for measuring the base pressure. During thocourse of the investigation it was found to be desirable torelocate each orfice to a position 2X inclms inboar(i fromthe wing tip (approximate center of exposed semispan).The fi

22、st orifice position iuvestigatwl is referred to as tho%nboard” orifice position, and t-he relocated position ismferred to as the “center” orifice position.Most of the wings may be divided into two groLlps ac-cording to the purpose for which they were intonclcd. Osmgroup was employed to investigate t

23、he effects of airfoilthickness ratio t/c and trailing-edge thickness ratio h/tonthe baa8 pressure. The profilos, dimensions, and tlmmethodof identifying these wings am shown in pmt A of table I.Thq am hereafter referred to as the “thickness group.”The ridge lines on threo of tht+sewings were rounded

24、 duringthe course of the investigation. In the figures, wings withrounded ridge Iinos are designated by (R)” aftw the wingidentification number.Tho sccoud group of wings was employed to investigatethe variation of base pressure with the boat tail angle I?.The profhl dinmnsions, nnd identifying symbo

25、ls of wingsin this group are shown in part B of table I. Tlmy will bereferred to as the %oattail group.”TIw surfaces of all tho wings were originally grouncl findpolished to approximately a 10-nticroinch root-mean-squaresurface. However, during the course of tbo investigationthd wings became scratch

26、ed from handling and from smallforeign particles in the wind tunnels. In addition, al thewings were modified rstleast once during the investigation.From time to time various wiugs were poliakl h restorethe surface finish ta approximately its original smoothes.However, it was obvious that all the tes

27、ts were not madeon wings with tho sanmclegrccof surface finish. Consequentlynear tho end of the invdgation thu surfuce rougln.wssofall the wings wus measured. Selectcd scgmrmts of thoresulting tra.cc records arc shown in Ilguro 1. lb traceshown in figure 1 (a) is typical of most of tlm snrfaco of al

28、ltho wings. That Wown in figure 1 (b) is tho roughest localsegment of surface found on any wing. The traco shown infigure 1 (c) is typical of tho random wratihos thtit wcrofound on many of the wings. *TEST METHODSWing supports,During the course of the investigatio,three types of wing support were em

29、ployed. The supportadopted during the initial stages of t.hc invcstigation WQStho sting-type support. shown in furc 2 (a), This supportwas designed from the viewpoint of minimum intwfercncc,but it proved to be too weak for tho starting loads in thoNo. 2 wind tunnel. A stronger support was Lhen adoph

30、ulwhic utilized a 25-caliber ogive-cylinder body, This bodywas provided with two bt crchimgmblo nose sccL.ions ofdiffe.rent lengths so that Lhc effect of the position of L11Obow wave relative to iho wing COUMLrcobserved. Theshorter length support is termed tho “short body No, 1“(. 2 (b), and the lon

31、ger length support is termed thoqong body No. 1.” The diameter of each body yas0.75“inch,and the nose was located 5% and 12 diameters, rcspcc-tively, upst.reamof the wing leading edge. Unfortunni dy,Table IDimensions of the Wings EmployedIn the investigationA The Thickness GroupL=-J Wing l- o?y far

32、all wingsIn this groupika,l M+1 Whg F Ma -M, Ru-1.lXIOfi.FKWEE 4.-Typical ohina-olayphotograph of wings with wire trip%self-synchronizing stroboscopic schlieren unit similar tothat described in reference 7. This Iatt,er unit will makemy periodic flucuat.ionsin the flow field covered by thesc.hlicre.

33、napparatus appear stdionary if the frequency isless than about 1,600 cycles pm second. A photoelectriccdl pick-up contained in this unit responds to fluctuationsup LO80,000 cycles per second, nnd, therefore, an oscilloscopewas employed in conjunction with this unit so that fre-quetwiw above. 1,600 c

34、ycks could be measured, although not“stopped” on the sc.hliercnviewing screen.RESULTS AND DISCUSSIONSince prcviom meawnwmentson boclies of revolution haveshown a mmkcd cliffercucc between the. base pressureRu-L7x1OLFmum 7.-Span!rSse vq!atfon of kasepressurefor turbtdrnt boundary-layerflow.with a tri

35、p at Re= 1.7X 10 also apply at least up to thehighest Reynolds numbers of the present tests. In addition,it is seen from figure 9 (b that the correlation curve forill= =2.o applies with fair accuracy to the data obtainedwithout a trip a a Re-noIds number of 3.5X106. In thiscase natural transition ev

36、idently occurs somewhere alongthe smooth surface upstream of the trailing edge.The small difference between the measurements taken atRti=3.5 x 105 with tind without a boundary-Iayer trip, andrdso the small siopes of the correlation curves for Ma= 1.5and ilfm = 2.o, indicate that the effect of Reynol

37、ds numberon base pressure is small for tmlmIent.boundary-layer flow.Even though the slope of the correlation curve for M-=3.1is sizabIe, the effect of Reynolds number is relatively smalIsince the abscissa involes the fifth root of the ReynoldsCOMME FOR AERONAUTICS.8.6 Re-1.7xloL(c) M- -S.1; Re-2.6X1

38、0J.QQmma prwsum mmmrremcntson the thicknessgroup of inss with turhlcntEow.8 - FOked curve of fq8 (a);1 wire trip, I?e =1.7x10G/f.6A IG ,4 u m o -#Lfv.lo i!5 5.2M= 0,25 0 all?) c(R)0,50 a75 m =)(4 0.75 0 0 a71LOOA AAoLo I I.8 -Falred we of fig8 (b)I wire fri R = 1.7x10*/I.6Rs-3.OX1O$.(b) .Wm-2.0; Re-

39、a.sxw.EhGUEE9.-CorJIsorI of tie Dreawre nICWrrCMent8on tho thicknessgroup d whsswith turbulent flow at dffcrent Reds numhv%Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-B.+SE PRESSURE OX BLUKT-TB-JJLR-GEDGE _“GS AT SUPERSONIC YELO.CITIES 1153 _ _nu

40、mber. The fact that at a given 31ach number one cor-relation curve appliee to aII wings of the thickness groupindicates that for a given boundary-layx thickness andtrailing-edge thickness the base pressure is insensitive tomoderate chtmges in profle shape upstream of the tra=edge.Effect of boattail

41、angle for turbulent flow.+ince t-hevariations in profiIe shape between the diflerent wingg ofthe thichms group did not invohre Imge variations in boat-tail angie, it was thought desirabIe to measure the basepressure on a separate group of wings. The boattail groupof wings was used for this purpose M

42、 this group containsthree sets of profiles with a fixed trailing-edge thickn- butffEEH2 O 2 4 6 8 10 12 14 K 18 20Boattail angle, , degW .I= -1 P4- lJxIC#; h/c-O.0125,hl .f= -2. e=O.05.FIGUEE fO.FXwt Of bosltaII eL oa ?JUSSpmsn.re wfth tnrbdent flow.with boattaiI angles ranging from 0 to 20. A plot

43、ofbtwe pressure against boattail angle is shown in figure 10.Included in this figure we se-iemd measurements from thethickness group pIotted at their respective boat taiI angks.The effect of boat taiI angIe on base pressure for the twcases shown, nameIy, L/c=0.05 and h/c= O.0125, is seen tohe small

44、for turbukmt boundaryayer flNV. This result81s0applies to the intermediatee case, h/c= 0.025, not shownin figure 10.Effect of angle of attack for turbulent flow.-.llI measurements described up to this point vmre taken m-ith the wi%wsset at zero angIe of attack. .* plot of the base pressureagainst an

45、gIe of attack for a number of wings of the thich=group k presented in figure 11. fi-!L3Xl.FIGUM IL-Eflect of .M - 1.S-1.0.6/Pb .6 wingN .tfm -xl, Rc-zoxlor.xmmts taken at this Nfach number ahow a considerable in-crease in btise pressure as 8/h increases.At ilf =2.o measurements were taken on all win

46、gs of thethickness group in the range of Reynolds numbers .betwcen0.2 x 106and 1.7X106. The resuhs for the three thicknessratios investigated are plottd in figure 14. They show thatin till cases pB/pminc.rcaseswith increasing lh irrespectiveof .irfoil thickness, t,raiIing-edgethickness, Reynolds num

47、-lwr, or boattail angle (within the limited range 2.9.7 ,.,6 L =) 1 , ,5 7 1A_ Re-O.5X10$.I.Msepressure to be more prevalent at low supersonic hlachnlmbersthan at K =1,5 or higher. This same trend wasfound with wing 5-0.50 (R) on which nonconforming basepressures (Pbh?m = 0.5 at d Re2 0.4 X106) were

48、 measedat M. =1.25 for all angles of attack up to the maximuminvestigated (a=5). As is evident in figura 24 (a), wfng10-0.25, alhough having the same trailing-edge thickness =wing 5-0.50 (R), did not albit the UnSXPCCY10w bmepressures at ill. =1.25. This may be due to tie effect ofboattail angle; on

49、 wing 10-0.25 the boattail angle is 5,whereas on wing 5-0.50 (R) it is 2.15.SeveraI of the e.fkts described previously are deo evidentOn a pOt Of pb/p. Temus the parameter d(e)l% shown in figure 25. This figure illustrates the conditionsunder which the base pressuremeasurements on wing 5-0.25(R) correlated with the main body of measurements atJ1- =1.5. These conditions are: either (1) sticiently lowReynolds numbe