1、- .- No* 775 AXALPSIS OF W IiXl-3?UXBLL DATA O2J DIRECTIOiTAL STd3I and the fuseln$e to directional stabilitg. This paper does not attempt to establish criterions for directional stability and control; rather, the empha- sis is placed on providing some basis for desin to spec- ified riteions,. An ox
2、ample applyin3 the design methods has been included, IMTRODU CTION As a part of a general investigation directed toward developing a rational system of tail design, a study has been ma.de of available wind-tunnel data on directional stability and control. The nain emphasis has been placed on a study
3、 of the a.er0dynapj.c characteristics of the ver- tical tail surfaces and their contribution to the static stability and control characteristics of airplanes. Data on the characteristics of yawed fuselnqes, hulls, win%s, wins-fuselage combinations, and wing-hull combinations have also been collected
4、. The purpose of this study has been not to establish tha stability and control criterions for satisfactory fliqht handling characteristics but rath- er to provide methods for desin to specified criterions. Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS
5、-,-,-2 . . NACB Technics1 Xote lio. 775 Rud4er-effectiveness data were available for 4 air- planes aild 28 models, including two 35-f oot-span models of nultiongine airplanes. The contribution of the verti- cal tnil to st,c,bility, that is, ysving noments for both tail-attached and tail-rcmo-red con
6、ditions, was available for eight of these nodels. Yawing-moment data for fuse- lages and hulls were available for 17 nodels, For 4 of the 17 models, yaw tests had also beeen made of the wing alone and of the wing-fuselage combination. The study of the forces on the vertical tail is an extension of t
7、he work of references 1 and. 2, which concern the horizontal tail, slid considerabls use has been made of the letlods that thoy present. Analyses vere thus di- racted tovard ths determination of tho characteristics of the isolcted tail surface and the affective velocity and the direction of the air
8、flow at the til, 1Lnalyses of the yavini; mocents of tho rvisg-fuselags com“oinations were, in gsncrnl, much less stisfctory, owing to the inadequacy of methods for evslv2ting zither the contribution of the fusclzgc nnd the wing or of the lnrgc wing-fuselage inter- ference effects. Two-view drawings
9、 of the 4 airplanes and the 28 mo6els arc givcn in figure 3. Many diverse types are rep- rasentcd, most of then of recent design. The geometric charzctcristics zre listed in table I. WoCels 1 and 2 and zirplanes 3 to 6 vere tested in the KLCA full-scale wind tunnel; nodals 7 to 10, in the XACA 20-fo
10、ot wind tunnel; models 11 to 16, in tho BACA 7- by 10-foot win2 tunnel; and models 16 to 32, in the Wright Field 5-foot wind tunnel, AIRFOIL APPL 122, TO THE 9EBT I CiLL TAIL Considerable uncertainty attends the application of the usual airfoil thcory to tbc design of vertical tails, owing to their
11、low aspect rntio, the nccessnrily arbitrary methods of defining tSe area, and the large aerodynamic effects of the fuselage ans the horizontal tnil. Further- more, the cir flow in the region of the vertical tail may Provided by IHSNot for ResaleNo reproduction or networking permitted without license
12、 from IHS-,-,-EACA Technical Bote Wo. 775 3 be very irregular, parti-culrrly when the airplane is yawed, because of the .ow velocities in the wakes of the wing and the fuselage and the vprticity in the air flow due to the trailing-vortex system. These fcctors are sep- arately discussed with the purp
13、ose of developing consist- ent methods of taking them into consideration. Symbols d aspect rotio b span L fu-sdlnge length Z distance from center of gravity of model to the rudder hinge line. S area Fi faselage-wing interference factor q cffective iiliynamic pressure at tail q/Yo ratio of effective
14、dynamic pressure at tail to free-stream dynamic pressure p density dCn /dsr rudder effectiveness T relative rudder effectiveness ( y2/%) - c mean chord CN normal-force coefficient Cn pawing-momcnt coef f icient (wind axes) I effective thrust Tc thrust coefficient p V2D2 Provided by IHSNot for Resale
15、No reproduction or networking permitted without license from IHS-,-,-BACA Technical Tote iJo. 775 i* II propeller diametcr a angle of attack, degrees 1 angle of yaw, degrees (wind axes) G local sidewash augla neasurcd from the wind axis, ncgative when it increases the angle of attack of thc rawcd vc
16、rtical tail, dcgrecs S dsflcction of movable surface, dcgrecs Ch hinge-moment cocff icieat / hinge moncnt 1 i Yf cross-wind force of fuselage Cyfl cross-wind forco coefficient of fuselnge ,I. q volzi3 ) u, v coefficients of Cn+ and 6, in the hinge- moment eguntion Subscripts: t vorticl tail r rudder
17、, excluding balance b balance f fuselage FI wing A airplane Def initions of Gcome tric Characteris tics The usual vcrtical tc.il surfacos fall into five fairly well-defined groups. An example of each is shown in figure 2, which also d.efinos the span. Type I, corro- Provided by IHSNot for ResaleNo r
18、eproduction or networking permitted without license from IHS-,-,-MACA Technj.cal Note Bo. 775 5 sponding to the twin-tail construction, is mast nearly a normal airfoil and its span and area are defined in the usunl manner. Type I1 is attached to a fuselage that tapers to a point 2,t the rear. Tlle s
19、pnn and the area are both mecxsured to the horizontal tail, which assumes the part of an end plate. Types 111 and V are found on fu- selzges that Gaper, not to 3 point, but to a vcrticcl knife edge at the rear, The span Is measured to the hori- zontal tail, and thc area is tzkcn as the sun of the fi
20、n area, measured to the horizontal tail, and the total mov- abic area. For type IV with the horizontal toil mounted oil the vertical tail, thc span is measured to the upper surf2cc or to the extended upper surfacc of thc fuselage and the arca is the sun of tho fin area, measured to thc upper surfacc
21、 of the fuselage, and tho total movable arca. T:iese definitions mcy appear rather arbitrary and are perhaps no better thzn others that could. be chosen; yet thc results obtained with them were generally consistent. Aerodynamic Charncteristics of the 1s.ola.tod Vcrtic2.1 Tzil ormnl-force characteris
22、tics.- The slops of the nor- mnl-force curve, dCm,-/data is prinnrily a function of as- k pect rctio. It must bc natcd, hovevar, that the horizon- tal tail acts as an and plate for the verticnl tail, which cnuscs thc effective aspect rztio of thc vertical tcil to axcced its geometric vnlus. A thcore
23、tic,l sna,lysis made by membcrs of ths full-scale-tunnel stzff has shown that for thc usu2.1 ratios of vertical-tnil spnn to hrizontl- tail spa, tho increase in nspect ratio wiii bc ?“bout 55 percent. Tcsts of model 7 with two different, horizontal tails fndicated that the spn ratio is not n criticc
24、l fac- tor. In thc ,zbscncc of tho horizontal tail, the fscl2ge itself probably exerts a considcrnblo end-plate effect, Such nil effect is not rendily calcu-lable although some of the tests indic2“ted it to be quite large. Thc varintion of dCgt/dat with aspect ratio is shot.ri2 in the curve of figir
25、rc 3, vhich summarizes the re- sults of i-efcrcnce 3 for q*spect ratios snzller than 3 and those of rcfcrcncc I for aspcct ratios l?*rgar thnn 3, Thc curve rcprcsets only an avcragc of cxpcrimcntl results and, under certain conditicns, nay bc soncwht inaccurate. For cxample, the value of dC/da nny b
26、e incrcascd 5 to LO percent by a sacled gap bctvecn thc fin and thc Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-6 ITACA Technical Mote !loa 775 rudder (reference 4) or may be ioveret! an equal amount or more b: a bad gap or bg an irregular plan f
27、orm. The -alue of the relative rudder effectiveness T as a function of the relative rud-der and balance areas is plotted in figure 4, whicli repzoduces the curves of figure 19 of referciice 2. Berc c.gain certain deviations fron t?ie curvcs may be expcctd under variou-s conditions, for the ruddcr cf
28、fcctivencss will also depend on the spanviso distribution of tlic ruddcr arca and on tho na- ture of thc gap bctwcon thd fin and thc ruddcr (rcfcrcnco 4). Scalirc; the gap may incrcasc T by as much as 15 percent. 1iiii:o-momcnt cfinract oris t ics. - Thc hiiigc-noclzt cocf- ficicilt of rudder mag be
29、 exprosscd (rcfcrcnco 5) as a function of thc normal-force coofficicnt of the tail and anglc of ruddcr dcflcction Thc pramctors u and v nay bc convcniciitly defined from thc cquation in the following forms: Binga-nom2nt datn on isolated tail surfaces without Valance aiid rsith off set-hinge balance
30、were available in references 2, 4, and 6. From these data values of u and v, at smal.1 angles of attack and rudder deflectioas, were detorminad. Thc results are surnmerizcd in figuTcs 5 aiid 6 vrlicrein u and v arc givon as functions of sr/st and S%/S,. The hingc moments, for a givcn increzse in nor
31、mal force, may bc appreciably Less than icdizatcd by tlicsc curvcs if .the gap betwccn the Tin and tlzc ?udder is scaled, but mag bc somcwhat grcatcr if thc ruddcr nose is very oLu-at, Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-MACA Tzchnicnl Bo
32、te No. 775 7 Dynamic pressure at tk2 tail.- The lower part of the singXe vertical tail is generally in a region of diminished dynamic prosnure cnuscd by the fuselage boundz.ry layer and perhaps also by thc wake from the wing-fuselage junctures. Pronounced downwash, such as will exist when parlizl-sp
33、an flaps are dsflected may, hol:ever, lowor this wake and change the average dyncmic pressure over the tail. Some surveys of the air flow slightly ahead of the vertical tail of airplane 4 are shown in figure 7. The boundary layer is seen to have a considerable thickness and doubtless is even thicker
34、 farther back where it passes around the base of the vertical tail. The average dynamic pressure, as determined from such surveys, is generally slightly higher than the effective dynamic pressure acting on the tail because of the influence of the adjacent un- disturbed air stream, (See rcfereace 2.)
35、 On tlie basis of thcse surveys and the rcsults of reference 2, the ef- fective dynamic presrurc at a single vertical tail is estimated to be, on thc .-,vcrage, for propollcr-removed conditions, ?-bout 0.90 qo. This factor aay be low for a flap-down coadition or for some types of flying boats h2v- i
36、ng hulls that curvc upward toward the rcm-. At angles of attack approcching the stall, tha fctor may 2-ccrease owing to the effect of the thickened fuselage 311d the wing wake (reference 7). Twin tails are somewhat more favorably located than single tails as the wing and the nacelle iiakes appear to
37、 be less detrimental to the dynamic pressure at the tail than the fuselage boundary layer. A value of q/qo =l.00 was used in calculating the rudder effectiveness for the models with twin tails (nodels 3, 8, 9, 12, and 13) and gave good agreement with the experimental valucs. This factor should prob2
38、bly be reduced if thc tails arc located directly in thc wake of large nacelles. At high thrust coefficients, as in take-off or climb, the slipstream will appreciably increase thc average dy- namic pressure st the vcrt,ical tail. (f, fig. 7.) In this rcgcrd, the results of rcforence 2 indiccte that t
39、he corresponding increase in rudder eff ectivencss d-Cn/dr may be only hclf as much as would correspond to the in- crezsc in average dynamic pressurc, Direction of air flow at the tail,- The air velocity in the region of the vertical tail of a yawed airplane will, in gsncrnl, possess a sideward comp
40、onent. Accordingly, the Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-8 NACA Technical Boto Xo. 775 effective anglc of attack of the vortical tail will not be equal to the cnglc of yaw, , but will bi? - G), where a is the sidewash angle. The sidcvn
41、sh angle, which mzy be quite large, is associztcd vith the trailing vortex systam behind yawed wings and wing-fuselage combi- nations. Rn analysis of some recent tests at the ifBCA 7- by 10-foot wind tunnel (references8 and 9) indicates that thc sidcwash angle probab1.y consists of several com- pono
42、nts, thc tentative 5hcorg for which is given in the fol1oi;in; paragraphs. Ihc order of presentation corre- sponds to the order of importance (as indicated by calcu- lations). A yawed fusclcgc (or airship) cxpzricnccs a cross- wind force, ansocizted with which there is a vortex sys- tem similar to t
43、hat of an airfoil (refarcnce LO). A fuse- lage with a low wing is comparable, in this respect, with an airfoil with an end plate, and thc trailing vortcx sys- tem for positive angls of yaw (nose right) wi1.1 be such that: 1. Thc fuselgc vakc and the air basldc it flow to the left (destabilizing side
44、wash, comparable with the usual destabilizing downwash). 2. The air above the fuselnge wake flows to the right (stabilizing sidesh). 3. The air bclow the intersection of the wing and fuselage wakes has prccticolly no sidewash. The vertical tail surface will. thus, for a low-wing air- plcnc, bc msin1
45、.y in thc region of stabilizing sidewash. For a high-wing airplana, however, the vertical tail will 5e partly in the region of destabilizing sidevcsh nncl partly in thc region of no sidcwash. Ti12 vortices shed behind z* lifting wing rotate in such a direction that the air moves inboard above the wc
46、kc (or the trailing rortcx shccf) and outboard bclow it. If the trailing vortex shoet is assumed to be unnltered by yawing the airplane, the vsrtic*al tail of a yawed nirplnnc will be in pun invard moving strenm if it is abovo thc wzke and in cn outward noviqg stream if it is below the wake, The eff
47、ect should increase with lift coeff icicnt and. de- crease with nspcct ratio, and it should be cspccially pro- nouncod for wings vith partial-span flaps deflected, Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-NACA Techuical Note No. 775 9 For a wi
48、ng 3;rith dihedral s change j.n .if t at the center occus whon thz wing is yawed. The vortex shed from this point rotates in sxch a direction as to in- duce outflori above tho wing wake and inflow below the wing wake. Calcilztions indicate that this effect will be rcla.t;ivelg small. From thc foregoing discu-ssion it will be clear that, as rcgnrds the direction of the air flow at the tail, a low-wing design is much mors favorable than o high-wing design. Moment equationg.- In conformity with the preceding discxssion and analysis of the forces on -the vertical tail surface, thc equations for
