1、TECENICAL NOTESNATIONAL ADVISORY COMMITTEE.FOR AEONAU!TICS ._.=. . -. . . . . _-_. .No, 646 . .- .- _. .,WZ.?.?I)-TU$W3L TESTS OF A 2-ENGINE AIRPTJJW3 lfODiL “-” :“-,., ;:JAS A PR13LIMIXARY STUDY OF 1%IiIT COW ITIQFS “ “:“-:-,;-:-:-:=:“”-i-.-,_=,.,. .“- -,.- :-_.-:- -L.-ARISING ON TE1 FAILURE 03 GNE
2、 WGI%! -“By Edwin P. HartnanMemorial Aeronautical Laboratory .- .=. . . .hlwl”dtqi - “-WashingtonApril 1938.I-,. .- ., . . . - -.-.-Provided by IHS Not for ResaleNo reproduction or networking permitted without license from IHS-,-,- IlllllllMmMllllIilll!“-”- T“- -:”:-_;1;6.“ :-.;-.-.-.tProvided by IH
3、SNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-. . .s2 N.A.C.A. Technical Note No. 646rth,p.n tho mode in which the anglo of sides lip”was sQvera!_acgroo$.INTRODUCTIONThe failure of an engine in flight has always pre-sented a serious situation, especially in the
4、case of air-planes having %ut one engine; under such conditions imue-diate descent is necessary, regadlese of the terrain be-107. In the. case of multiengine, airplans , this situa-tion may be somewhat reliev, for the gliding angle maybe nade much smaller or horizontal flight may even be con-tiiucd
5、with one engine dead.Althou,:h aodern multiengined airplanes are designed tofly with one engine dead, occasional crashes caused by en-gine failure do occur. The obvious necessity of providingenough power in the remaining engines, after one hao failed,to continue flight iS by no means the only questi
6、on to.baconsidered. A.sid and (2) a consideration-of the conditions offlight after equilibrium has been breached, involving thvarious nodes of flight in which equtlibred three moreor less basic attfltudes in which a 2-cmgino airplane canmaiuta,ia oquil.ibrium with one ongino dead. Ono of the ,basic
7、modes involves considerable sidesliy and appearsless favorable than tho others from drag considerations.The othar tlo basic attitudes may be described rLS follows:If it is .5ssunod, for example, that tho left-hand onginoof :1conventional 2-engine transyort airplane has ftiilcd,equilibrium may be oht
8、aincd %y: (1) Applying enough rightrudder, with zero ,angle of-bank, not only to halanco theya:ring monont of the active engine but nlso to causosov-orv.1 degraes yaw (sid.eslip) to th left; (2) Ayplyingright rudder sufficient to balance the yawing noment. lhosido forces in this method are lmlncod b
9、y the weight com-ponont caused by a small ,ngle Of,bank rather than by thoside wind force on the fuselage iq the first method. Inthis nttitudo, the airplane ho zdro .nnglo of yaw. Thoctia,gr,.mshown in figure 1 illustrates these two attitudes,one of which is essentially an attitudo of yaw and thooth
10、er, an attitude of hank.In concoction T;ith the equil$%rium phnse of tho proh-len, the added drag due “to the two basic modes of flightwith oae engine dead was deter.niried. The tests were lim-ited to one”model airplane, which was considered fairlyreprosontative of the conventional 2-engine lor=wiug
11、 trans-port airplane. Information was also obtained concerningthe angles, ncj:ents,and forces for equilibrium for thLsparticular model.sli?Jst this arraagomcnt gave the model fr,eedomin pitch, yaw, an3 roll. The movoaents of the model wor6restrained. %y fin steel Hires attached to the upper sur-face
12、s of the wings and the rear part of the fuselage a-dconnector?. to balances upon which the pitching, the yawing,and the rolling moments were measured. L5ft, drag, afid .lateral :Orce mere measured with the regular wind-tunnelbalances. The parts of ihe mounting causing are Provided by IHSNot for Resa
13、leNo reproduction or networking permitted without license from IHS-,-,-6.-. *U.A.C.A. Techni-cal oto No. 646Instruments for measuring tho propeller revolutionspocd and tlzepower supplied to the notors woro loctntedon the tect-chamber floor.SYMBOLSA list of th symbols used in this report follows:lift
14、 coefficient.lift, 1%.wing .arca, Sq. ft. :dynamic pressure, lb. per sq. ft.dymnmic pressure in the slipstream.air density, slugs pe cu. ft.resultant-force coefficient (along Uiad axis).resultant force along mind ,axis, lb. (Dra ispositive. )2TrQno effective drtig,coefficient.qSJD sum of tho torquo
15、of all engines operating,ft.-lb.propulsive efficiency taken from propellertests made at zero angle of attack.d = V/nD.T-, air speed, fp.s.n, propeller revolution smed, r.p.s.D, propeller diaueter, ft.Cn =M/qsc, pitching-nonent coefficient.M, pitching nonent, ft.-lb.a71Provided by IHSNot for ResaleNo
16、 reproduction or networking permitted without license from IHS-,-,-.c,Cn = lT/qslI ,N ,v,6,a,6.r6.,c,X.A.C.A* Technic .e study of the equilibrium conditions of.flight-a-f- ?ter one eniae has failed. lhe dead-engine ,Qopeller na5locked ith the,propeller vertical in all cases. A con- -.+.ciderable nun
17、ber of plots and “cross plots of data were nec-essary to obte.in the final data representing equi.lihriuconitions. A few of the final plots showing .interestiand useful relations betee the yarf.ous”factois invOIVOare given- in this report. These figures relate largelyto the yawed condition of flit o
18、n one engine4 as”preti- -.ously doscr%be; _in all cases the loft=hand engine Ha% “stopycd. The fi=gurcs roprosent lvcl-fligh% conditions at -soa level. -7 .“ .-I?igures 11, 12, azzd 13 show the variation of yawtng-monent coefficient and lateral-force coofficien”t ith Si?.o-slip anlo for various rutl
19、d.orangles and e,nglo”sof ataclz.The ruiiLcr angles and yaf anlcs giving- z“6royariing momctand zero lateral force he shown as fi=gures 11(c), 12(c),and 13(c). The intcrsoction of the curvcs df Cy and Cndetermines the sideslip and rudaer angles for oquibriuufor each particular anglo of attack. It aj
20、pears from-thefi,;rs that the tin rudders od %e insu,ffCient tOmaintain cquilirium at angles of attack higher than 13.Si.;ure 14 shows the varia,tfon of pi.tching-monent co-efficient with sideslip angle for the thee angles 6fattack tested. Rolling monent”s were also easy-iem %utthey were .SO erratic
21、 as tc be useless-for showing “trends;howeer, average values did show the iolIing nom-efitto besmall finall cases. Thadiheflral angle of the model flight nith one (of two) engines dead, .-/(Afa )% . (i.hP Afp = .0.4 Af .md .,.Afa= 0.6 Af;therefme” Afp = 0.4 x 0.30 f = thQrefore (Afa ) = Afa X k X (4
22、80/34).clim%If k is assumed to he squ-a to 1.1, (Afa )C=im% =0.18 f x 11 (480/340) = 0.28 f. The total increase ofequivalent. yarasite area in full-throttle climb at sealevel will be Afm + (Afa)climb =0.12 f + 0.28 f = 0.40 f.GIn normal 2“-engine fliht ha thrust horsepower re-quired for level flight
23、 at 93 mils per hour is 294. A.tiouthalf the poer (147 t.hp,) is used to OVeZC.Dme parasite *drag and the other half to ovorcoc induced drg. Thethrust horsepower required to ovor-come the parasito dragof the airplane climbing on onc engine atisealevel will-bC?147 X 1.40 = 206 tiihP.Tho tote.1 power
24、required will be206 + 147 = 353 tih.480 - 353 = 127 tohp,and, as the airplane weigs 13,600 pounds, f-ully loaded,its initir.1 rate of climb lith that load will be(127 x 33,000)/23,600 = 308 ft. pnr min.Tho Va.luo of (_f+ Af)/f for high syeed at sea lcvolwill he fipproximaticly equal to the ratio (t*
25、hp.l/(t.hP.)a.H-Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-H.A.C.A. Technical Note No., 646 17nhere (t.hp. )l is the thrusthorsepower required at highspeed on one engine and (t.hp. )a is the thrust horsepowerrequired in normal 2-engine flight at
26、 the same “air speed.:For the pre.sen% airplane,(f + Af)/f = 590/496 = 1.19tihich gives a value of 0.19 for Af/f at high speed onone engine.-In c summarization of the case for the present air-plane, the approximated values of Af/f, which represeriian average for the yarred and banked modes of flight
27、 onone engine, are as follows: At ceiling, 0.30.I?ull-throttle climb at- sea level, 0.40.High speed, 0.19.Anp3.icntion.- It should be pointed out that thesevalues of Af/f are the result of some fairly loose ap-proximations and that they apply strictly only to the air-plane represented by the model.
28、There are a great manyvariables in ,the Lesfgn of airplanes that will have amarked influence on the ratio Af/fi Among these varia-”bles are: propeller diam”ter, num%er of “blades, bladefan la, engiqe spacing, and rudder design. The values ofAff will also vary directly with the ratio “of enginepower
29、to f. Also , most modern 2-engine transport air-planes use 3-hla”de propellers instead of .2-blade ofissuch as used in the present tests. For this reason alone,the values of .Af/f for modern airplanes may possibly %e25 percent greater than the values given for”the presentairylane model, It is beyond
30、. the scope of this reportto give consideration to the quantitative effects Of allthe factors that affect the flight of airplanes with deadengines. The lest source of information on this subject”at the present time is found in “rofererice 1, which shouldhe consulted in all performance calculations o
31、f this iype _.,.-.-JEuilibriurn a.tti.tudea Some of the equilibrium atti-tudes and control angles nessary for flight on one en-gine can be examined by the aid. of figure 19. In thtsfigure are shown the equilibrium angles o sideslip andbank and the equilibrium rud.dor angles for the twO basicProvided
32、 by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-. .18 N.A.C.A. Technical Note No. 646,modcs of flight on one engine. In the banked mode offlight, the sideslip angle is ZOI$O and the angle oflbankis about 2, increasing slightly .wlth increaso in airspeed, The
33、 rudder anglo increasti with a decrease in a,irspeed hut is much lowr than the angle necessary for theyaed mode of flight. In the yawed mode of flighk, therucider is inadequate to maintain equilibrium below an airs,poed of about 72miles per hour, In this mode of flight,the an 10 of bank is zero and
34、the angle of sideslip iS un-“Edor 10 , decreasing rather rapidly with increasing airspeed. ,The fP.ct that the rudder SJIglQ for tie banked mode offlight is considerably less than for the yavad mode offliGht is very important, because there is always a etrongpossibility that the rudder will be inade
35、quate and anymeans OS increasing its effectiveness will be otialue,For this reason, as well as for the purpose of securing a “;”lower drag, the banked mode of flight is much superior tothe yawed mode.Dein-engine or a multiengined airp?.ane, .the designer should PEWparticular attention to the rudder
36、design in addition, ofcourse , to the obvious necessity of maintaining tho max-imum excess oowor available, with one engine dead, ovar.that required for level flight, The rudder (and finshould be somewhat larger than the size required for ordi-nary flight purposes and it shoula have as great an aspe
37、ctrntio as is structurally feasible for the plrpose of main-taining high values of L/D, for the vertical tail sur-faces, at large rudder angles.The full-feaherfng controllallo propeller that can baset to 90 after engine failure is an important asset inflight with one engine dead. It is of especial v
38、alue ininstallations where the normal blade angle is 10W$ for thedrag of a dead propeller increases rapidly with decreasein blade angle. A full-feathering propeller will not onlyreduce the drag of the dead propeller to a large degreobut it will also relieve the load, as well as the drag$ onthe alrea
39、dy heavily loaded rudder.a71Slipstream Survey and Incidental Fowor ffectsXl_ stream surve.- The results of the slipstream sur- .Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-.K.A.C.A. Technical Note No. 646 19vey arc shown in figure 20 as contours
40、of .qsfs the ratioof the dynamic pressure in %hc slipstream to t“hti“ayn thedestabilizing effect of adding pofver, however, is clearlysho7n*Iigure 22 gives an interesting comparison of thepitching momonts with the twin- aad single-rudder tr.ilgwith which the airplane model was alternately equipped.A
41、lthough the horizontal urfnces Of tho twi.n-ruddar tailTVOI?C of somowhat smaller area th thOSG of-the single-rudder tail, tho twin-rudder tail is obsorved to ho moroeffective. This quality of thQ tin.rudder tail has boonknuwn for- sone time and is usually a.ttributod to the end- .plate e:fcct of th
42、e twin rudders, which tend to prewmttip 10ss. All the MOI_ZOntS in this rport are given with “rospect to tho pivot point, which was at 32 pcrcont of thoMean aorodyn,amic chord (seo fig. “2) except in the case offfre$ 21 and 22 where the moments wero transferred to CLpntnt,5 porccnt of the me,n aorod
43、ynanic chord ond md 12porcont of the moan aerodynamic chordal)ove the pivotpoint. The point about which the pitching moments in fi- .uros 21 and 22 are given 3S at 27 porcont of the moan aoro-dynanit? chord. ,= Figure 23 shows %hat the addftion of”power increasesthe effectiveness of tiie elevator co
44、ntrol. Although thecurves appear to be somewhat erratic, the effect of poweris clearly. shown. The curves in figure 24 for level-flightpowr conditions show the increase in Uag due to elevatordeflection.Tho yawing moments plotted against yaw angla for themodel airplane equipped with the single- and t
45、he tirin-.rudder tails. are given in figuro 25, which shows the twin-rudder tail to 3e losefffe than the single-ruddrtail in regard to yawing moments, though both have the samevertical tail area. Figures 26 and 27 show the rudder ef-fectiveness, on this type .of airplane, to be very littlealtered by
46、 the application of power.CONCLUSIONSThe following conclusions refer, in general., to 2-enine airplanes similar to the modal tested.Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-.?J.A. C.A. Technical Note I?o. 646 21, . . . -1. Toeing-out the engin
47、es or toeing-in the twin rud-ders anfi fins mere impractical methods of reducing theyawing moment after engne failure on the.low-ming 2-engineairplane that the model represented. c2. The use of power i.n the present tests producad-adestabilizing effect on the pitching moment of the air-plane model.3
48、. The yawed. mode of flight on one engine after en- .gfne failure increased the drag of the airplane somewhatmore than the banked mode of flight .-4* The increae in the parasite drag of the airplanein three conditions of flight with,one engine.de.ad was p-roximately as follotis: ceiling, 30 percent; full-throttle.clmh at sea level, 40 percent; high speed, 19 percent. The tests indicate that a powerful and efficient “,(high50L/D) rudder would %e necessary
