NASA NACA-WR-L-489-1940 Drag analysis of single-engine military airplanes tested in the NACA full-scale wind tunnel《在NACA全比例风洞中试验的单发动机军用飞机的阻力分析》.pdf

1、ACR Oct. 1940NATIONAL ADVISORY COMMITTEE FOR AERONAUTICSltrlRTIMl REP()RTORIGINALLY ISSUEDOctober 1940 asAdvance Confidential ReportDRAG ANALYSIS OF SINGLE-_NGINE MILITARY AIRPLANESTES%_DIN THE NACA FULL-SCALE WIND TUNNELBy C. H. Dearborn and Abe SilversteinLangley Memorial Aeronautical LaboratoryLa

2、ngley Field, Va.COP_“i!iiilii_, ii_iii!iiiiii_:.:.:_._:WASHINGTONNACA WARTIME REPORTS are reprints ofpapers originallyissued to provide rapid distributionofadvance research results toan authorized group requiring them for the war effort. They were pre-viously held under a security statusbut are now

3、unclassified. Some ofthese reports were not tech-nlcally edited. All have been reproduced without change in order toexpedite general distribution.L - 489Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-DRAG A_TALYSiS 0_- oIN_,_LE-E_R_.!I_S MILITARY J.

4、 )n of s lar imeffi-cio:?.t design feat_.res un m,.%ny of the _irp!s.nes in_.cated_“-_.n.ee._irability_ of _,_“,l,Y.ing.,_an_ combining all of ther_,su.ts into _-.single pape. r for distrioution“ to des “_:_,n,;.rsoThe data for the v:_ricus a,irpi_nes are not consistent insco:oe since t,l,: extent o

5、f th_, tests d.el)ended on the possi-b.!ity o_ ma,cinc_ alte:ations to ti-o “,_,-_ ,_._.o_ cular rpls, nono_ i,l_e time _v_iiabie for _e tests.The discre:_;_ncie_ bet_eon the compn t,ed hish _“b_ “ I- cec_sfo“ idee_! a_rpla, ne _ _ -,-_: _e _,rtect“ “,vo_,sb:_ndard m.il_t:,z, typc. s ;_re well known

6、, anditis l,._z_._.,“_n,:.:purpose of tb_-is paper to indico, te theretirees Of _“ “ ;“ “: _11 ,_,CSeC4!_._eIOIIQS, Q com_:)rem.isos invoiv-.8.i! bie similarto those shown in figure 17 were tried in an ef!ort tOmaintain a good external shape and at the sc_me time to pro-vide sufficient air flew. It

7、was f0und that the ion,g-nosecowlings with air flowing through them showed no decreasein dr_g over that of the Ccnvcntion_,l NACA cowling, indi-cating thc,t some peculis,r internal or external flow phe-nomena existed to nullify the gains which apparently shouldProvided by IHSNot for ResaleNo reprodu

8、ction or networking permitted without license from IHS-,-,-14be realized from the improved external shape. This in-vestigation was of a preliminary nature and more detailedinvestigations are now in progress _t the laboratory.For conventional NACA cowling installations, it hasbeen found that the best

9、 not efficiency and the minimumnegative pressures are realized for cowling C, which w_sdeveloped from tests in the NACA high-speed tunnel andreported in reference I0.In a further attempt toward improving the blunt shapeof the NACA co_vling, tests were made with spinners of var-ious sizes attached to

10、 the propeller (fig. 18). Thesespinners vthis resulted in a decrease of drag coefficient of 0.0005for the airplane with the NACA cowling without cooling _ir.For the airplane with the solid streamline nose the dragwas the same with or without the lengthened afterbody, Afurther small change was made b

11、y enlarging the tail of thecockpit canopy to decroase the divergent air-flow angle.Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-15This had no measurable effect in the case of the NACAcowling; however_ the change increased the drag of theairplane b

12、y 0.0006 in the case of the streamline noseinstallation.)I Some air-cooled engine airplanes when viewed fromthe top show a distinct necking-in of the fuselage aftof the cowling. On airplane 5 the fuselage was modifiedso as to eliminate this necklng-in feature, as shown infigure 19(c). The stralght-l

13、ine fuselage elements ex-tend from the front of the fuselage to points of tangencyaft on the fuselage. This change reduced the drag coef-ficient of the airplan by 0.0009. A similar change wasmade on airplane 6 _(fig.19(a)which reduced it.sdragcoefficient by 0.0006.Air inlets.-The rules for the desig

14、n of duct inletsare not so well established as those for the design of theoutlets. The principles are known, however, and have beenverified by experiments. It is a primary requirement ofa duct inlet that it recover the full total pressure cor-responding to the flight speed of the airplane. If thetot

15、al pressure at the inlet is lessthan Ho there willbe a power loss calculable by means of equation (i). Theopening should therefore be located at an existing stag-nation point such as the wing leading edgeor the nose ofthe fuselage, or at an artificial stagnation point createdby means of a scoop. The

16、 use of_scoops is discouraged,however, by the requirement that the flow into and aroundduct inlets should not create local gradients in the pres-sure distribution over the body or increase the values ofthe negative pressures above those of the body without theinlet. A well-designed opening at the no

17、se of a wing orfuselage will in fact tend to reduce the negative pressuresover the body near an opening since a part of the air isbypassed through the duct and the external velocities arelower (fig. 20). ,Large adverse pressure gradients (negative to posi-tive) cause a transition from laminar to tur

18、bulent flow,and tend to precipitate flow separation. Large negativepressures on a body further lead to compressibility effectsat low critical speeds, and require that the afterbodybe long to reduce the adverse pressure _gradients. Whileawaiting a theory for specifying the shape required foropenings

19、of different size and alr-flow quantity the ex-periments of referncs 5 may serve as a guide. By properlyproportioning the opening, inlet velocity ratios vi/vomay be varied over a wide range without increasing theProvided by IHSNot for ResaleNo reproduction or networking permitted without license fro

20、m IHS-,-,-16external drag. _ghen the internal duct p_sss,_es cannot bedesigned to expand the air efficiently it may be desirableto provide low inlet velocity ratio to reduce the ductlosses.The corners and sides of rectan_ular duct inletsshould be Carefully rounded and faired into the body. Ifan opti

21、m_._.m hil(c) w_re usedin the _in_%-fuselage fi.lets of _irplane 9 which o.ddcd0.0019 to the dr_g coefficient of the _irplane. In figure_l(c) the tufts show the large e_tont of the flow disturb-.,_eance on the airplane caused by _,;_.esescoops._uft operation in airpl_ne I0 showed that a satisfac-tor

22、y flow existed over the -_rburctor- scoop,which was lo-cated in _n_ nose of the co,_ling (fig. 21(d) for thepower-off condition; however, _,_ith the propcl.er opero_ting,Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-17a distinct flow separation was

23、 observed on .the. downstreamside of t!_e scoop o,Jing t0 the _slipstream rotation. Toelininate .this undesirable flow, the sides of the carbu-retor sc00P were faired out more gradually into the cowl-ing line, as indicated by the section line, on ifigure 21(d).This fairing decreaseirplt tiio roar of

24、 the TACA covzling (fig. 22(C:). Forthis in st_ll_tion a clr_og increment of O.0007 was me6tsurod,which is not considered excessive for the external insto.1-lation. It will b.o noted that this Scoop has a well-,_ stro_:mline shape.f or;._.The oil-cooler scoop on airpl_.no 4 was placetion obtainedby

25、removing t!_o scoop and so_ling the outlet _:as 0.0007.This ,_,_asreduced to 0.0003 by refairing the scoop, as shownProvided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-18by the section lines in the figure. An cxtremely inef-ficient oil-cooler inst_llatio

26、n was used in airplane 8(fig. 22(d). It consisted of a sharp-edge scoop ioc._t-ed on the bottom of the fusel_ge which diverted air at arather sharp _ngle up into the oil-cooierl ducts locatedin %he fusclngc. The _ir then was discharged _t cn angleof about 60 to the fuselage axis. This oil-cooler in-

27、stallation faile to supp!y sufficient air flow for oilcooling _nd in s_ddition increased the airplane drag coef-ficient by an increment of 0.0017. Since it w_s impossi-ble to modify this installation without m_jor changes tothc mirpl_ne structurc, an underslung radiator installa-tion :was designed t

28、o be mttached to the bottom of the NACAcowling (fig. i0). When the required quantity of airflow passed through the cooler the drag coefficient was0.0009. To determfne what part of the dr_g was due to theprotuberance and what part due to the air flow, the oil-cooler duct was faired over at the nose a

29、nd tail so _s toprevent air flow, and an increment in _rng coeffic:_ont of0.0004 w_s measured.As an example of _:n extremely poor insto_l!ation o_ndar_ illt_.stration of its h_rmful effects on the s_irplo.nedreaD, results o,re presented for the temporary oil-coolerinstallcotion which was insta_lec o

30、n alrpl_ne 9, c_s shownin _ fi_u_e 22(0). This lo_r_e scoop increased the air pianodrs_L_ coefficient by an increment of 0.0040, whic!l corre-sponecd to approximately 25 percent of the entire airplanedrag. This installation was le.ter _h_,_,_._.o_into a rela-tivol_ inefficient wing duct in which loc

31、:_tion it in-creased the drag coefficient by 0.0011. A win_ duct oil-cooler instn!l:_tion was _Iso used in L%irpl,o_ne l!, asshown in figure 22(g). The duct p_ssages through bothwings wore bent sharply to _void interference _ith the!anding-geo.r struts and a considerable loss in internalefficiency r

32、esulted. The drag coefficient of the airplanew_s increased by 0.0006 because of the wing ducts. It isbelioed bhmt with an efficient internal duct the dragcoefficient wouldhave boon increased by no more than0.0004 for this installation. The oil coolers for air-plane 10 were lees, ted in strcennline d

33、ucts on the lowersurf_ces of the wings outboard of the fuselage. The oilcoolers were approximately n_._if submer_Tec_ into the wings(fi_. 22(f). These oil-cooler installations increased theairpl:_ne dro_ coefficient by an increment of 0.0008. Ass_ check on the added external skin friction drag due t

34、othose ducts, streamline noses an( t_ils were o.dded to theunits eond o_ dr_g coefficient increment of only 0.0001Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-19measured, This substanbiates other ds_ta showing thatstreamline b!isters loc_ted at no

35、ncritical positions onthe airple.ne do not add large increments.The l_rgest scoops added to the airplanes were thoseprovided for the Prestene radiator ins.tallations on air-planes r? and l!, On airpl_ne 7in its original conditionthe Pres_tone radiator was located under the Allison engineand below th

36、e normal fuselage llno, The S._irwas taken in-to:the ro.diator by means of a ls_rgo, scoop which fs sketchedon flguro 23(_). This installation increaslod the drag co-efficient-of the airolane by an increment of 0.0034. In“ trio Prestone radiator in-o.n o.ttcm_0t to rcduce the drag of “_stallation, t

37、he re,diator _,lasrs_i._ed so s_S to p!ace it with-in .thedriginai _ineS of the fuso!a,o_Cnosc,_as shown infi,gurcs23(b) and _(C). For this arr_.ngemont it ,_,_illbe_ _ _3 .“ . . . , .noted r,n._tthe i_:let did_not protrude be_ow the normalfusole.go_line. ._iTh-odrg.g coefficient of i.the modified i

38、n-st_llation _,_.s_0,00117 or npp.rOximntelz: one-half that ofthe origin_-Ll inst_ll_tion for the _same air“ flow quantity.Other scoop nrr_.,n,_emonts similar to the moclif.ied scoopsused on .o.irplane 7 _.cre investigate,.% on _irp.lane il.2_gain the Pres%onc r_.cliators were ins t6%.llod.within th

39、eo.ri,f._ino.lf_.ire.dbontour of th.e_fuselage; however, thescoop i.nlet protruded: S_i_.t!_ _ bolow the original fuselageline (fig. 13). Owing7 to the _ici _-_,_ en_ internal flow m_,depossible through the sx_c_u_ expansion _of its internalduct, c dro,_.!_coefficient incremen $ of only 0.0011 _zasm

40、eas,ared:for th_s airpl_tnc. A. s_m_l _ underslung scoopne_r t_. tr_.i!l._s .odgO.of the wing(fig 14) For this “ceJse with the coo.lin_ c_ir flow asfor the .f.orwo.rd unde:rsi_ng arr,-_ngeme_t _-_ _“cient increment was 0.00i0. Attlention is _c:o.l.ledin bothof those cases tO it.hofact “_“_t, with _

41、_,zell-designedscoop cve_ nf !_.r_._ si s the.so g_ ze ucn. as just d-cscribcd, ex-cesszve _ag_ w.Cre :not .ob.t,?.i-ned . T .i_ules .for the desi_-_._,of Scoops b_se_1 on_h.o, e.xperi-enc.o gaine_l _,_-it.ht_ho airpi_.nes arc _s. follows= _I. Provide., a nose ra.d : “ius on“the, lipSOf, the scoopsi

42、milar to that at the noso of e_n airfoil.-“ _“ _ “r“-e dge Scoop. .; .“ . “ . . “ “ _- . Ji “ “- “2. _rovide sufficientcc.mber l.n the scoop contourso- aS to m_tch _e streamlines of the flow. ,. _Sc.oops With low inlet velocities require moreProvided by IHSNot for ResaleNo reproduction or networking

43、 permitted without license from IHS-,-,-2Ocamber. (See fig. 24.) When possible, meas-ure the pressure distribution over the scoopwith correct air flow through the opening.3. Until more detailed data are available, designthe scoop inlet are_ to provide an inlet ve-locity of from one-hslf to two-third

44、s of thestream velocity at the high-speed condition.If the scoop inlet is not made adjustable,the inlet velocity ratio will necessarily berequired to be lower and the camber in thescoop greater. (See rule 2.)4. Provide a well-shaped afterbody behind the maxi-mum scoop section with sufficient length

45、toavoid flow sopmration. Four times the scoopheight will generally suffice, olthough anafterbody too short will be much more harmfulthan one too long.5. When the scoop is located in a cross flow suchas a propeller slipstream, fair the sides ofthe scoop gradually and smoothly into the bodyzdj_cent to

46、 it (fig. 24). The sides of thescoop for this case correspond to the after-body in a straight flow.6. If a scoop is located in a thick boundary layer,considerable difficulty will be experienced inobtainin_ high efficiency. The inlet areashould be exactly proportioned to avoid flowseparation in the b

47、oundary layer ahead of theinlet, and vane_ used in the duct to obtain amore uniform velocity distribution._xhaust stacks and turbosuperc_har_er.- The require-ments for the recovery of thrust from exhaust stacks byrearward discharge of the heated gases have already beendiscussed _. However, it is des

48、irable tO further considerthe extern_l drag due to protruding exhaust stacks on thefuselage. Tabulated results on the drag due to the vari-ous exhaust stacks are given in table III.The exhaust stacks listed are for air-cooled engineswith the exception of these for airplanes 7 and Ii. Thetwin stacks on the air-cooled engines protruded from theengine cowling at right angles except those for airplane 5which were directed to rear at an angle of approximatelyProvided by IHSNot for