1、. REPORT 1171EFFECT OF HORIZONTAL-TAIL SPAN AND VERTICAL LOCATION ON THE AERODYNAMICCHARACTERISTICS OF AN UNSWEPT TAIL ASSEMBLY IN SIDESLIP By DONALDR. RILEYSUMMARYAn investigation ha been condwcted in the Langley stu. this method yields a system of nequations in n unlmowns. solutions of Siml.lhtmeo
2、usequations of this type of high order are usually somewhatcumbersome to obtain by manually operated computingmachinea but are reasonably well adapted to the relay+pedigital computing machines.1SnpsrwiesNAOATN W, “EffectofHorizanb#-TailSpanandVartkalLocationontheAerodynmdoObnraaterktkOfan Umwept TaU
3、 AssmblyinS1desliby DonaldR. Riley,193.3.308G0G-ScAAo!:vE.c.=c1C;pClflC*Plift coefficient, L/vlongitudinal-force coefficient, XlVlateral-force coefficient, Y/q:2%724024s2m2:10.mo.139.!m.353.423.4m.Lw4.674.m4.6am.it y tunneL(b) 20-inch horizontal tail at location D.FmmzE 3.Continued.were obtnined fro
4、m reference 7. Thkse expressions show-that the desired veloci components are proportional to theproduct of the circulation nnd some function of the geometry.Tabulated values of the geometric contribution in terms ofthe vortex semispan are prented in references 8 and 9 forthe downwrwh where the contr
5、ol point is located in the same(a) 10-inch horizontal taif at location C.FIGURE3.Concluded.plane as the horseshoe vortex. All sidewnsh values werefound from the sidewash expressions given in reference 7 be-cause no sidewash tables were available.A correction for the difference between the actual and
6、 thetheoretical values of the section lift-curve slope must bo ILp-plied to the calculated results since the iinite-step method isbased on the theoretical thin-airfoil-secionlift-curve slope of2r/57.3. For the case where the actual section lift-curveslope cl=is about constant across the span and whe
7、re only nmmll correction is needed, it is su.flicientto reduce tho vnlues57.3clby the ratio . For more exacting compu tntions, the2Tposition of the control point may be shifted to account forvariation5 in c, (See ref. 10.)Calculations of the span loadings which were reducecl toforce and moment deriv
8、ative were performed with fourhorswhoe vortic6s distributed across the quarter-chord lineof the vertical tail from the fuselage center line to the tip soProvided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-AERODYNAMIC CHARACTERISTICS OF AN UNSWEWC TAIL AS
9、SEMJ3LY IN SIDEISLJP 355that each horseshoe vortex had a span of 5 inches. Usingthis vortex span resulted in an even distribution of horseshoevortices across each of the three horizontal tails. The choiceof four vortices to represent the vertical tail was obtainedfrom rLshort,analysis, given in appe
10、ndix A, of the effect of thenumber of horseshoe vortices. Sketches of several of thevortex configurations calculated, showing variations in spanand location of the horizontal tail, are presented in figure 4.The number of equations necessary to solve for each differentconfiguration could be reduced s
11、ince the loadings on the twopanels of the horizontal tail were of equal magnitude butopposite in sign. For the con.tignrationinvolving the 40-inchhorizontal tail, for emmple, the calculations involved asolution of only eight simultaneous equations. The solutionof eight equations by use of the Crout
12、method in conjunctionwith manually operated computing machines required about62 hours. Using rLrelay-type computer reduced the timerequired to about 3 hours.Several additional calculations for comparison purposeswere performed for the vertical tail $ combination with the10-inch horizontal tail by us
13、ing eight horseshoe vorticw dis-tributed across the vertical tail instead of four.RESULTSAND DISCUSSIONPRESENTATIONOF RESULTSThe experimental results for the fuselage alone, with thevertical tail, and with the vertical tail in combination witheach of the three horizontal tails at various vertical lo
14、cationsam presented in figures 5 and 6. In addition, the measure-ments obtained for the rolling moment of the horizontal tailabout its point of attachment to the vertical tail are presentedin figure 7. Experimental values for the static-lateral-Z however, the magnitude of thevalues of C. expected fo
15、r this fuselage is not within theaccuracy of the balance system used. The CZ8obtained forthe fuselagevertical-tail combination is probably due tothe (?% of tbe vertical tail alone and possibly ta a smallincrement due to interference e.ilects. Although the forceand moment center for the tests was loc
16、ated at the quarter-chord line of the vertical tail, both theoretical and experi-mental results for wings show a forwaxd shift in the aero-dynamic center with decreasing aspect ratio.The horizontal-tail-on data (fig. 6) indicate noticeablevariations in CYdand CZBfor changes in horizontal-tail spanan
17、d vertical locations. These results are summarized infigures 8 and 9 6L-12 + -8 - -4-Angle of sideslip, , detj.1 o-JI. .Cn o-.1 -16 -12 -8 -4 0 4 8 12 16Angle of .sideslip, , deg(a) Horizontal taile at location A.-E_ental aerodynamic oharacterietica of the unewept tail assembly, having horizontal ta
18、ile of 10-, 20-, and 4CMnohqmna, for thevarioue vertical Ioc.ationa of the horizontal tail relative to the vertical tail.Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-.AERODYNAMIC C!HARAOTERISTTCS OF AN UNSW3NW TAIL ASSEM8LY IN SJDESLIT 357.4 -.3 -
19、.2 -.1 -0 -cl-.1 -.2 -#3 -,4 -.4 -.3 -,2 -.1 -cl o _-. I -.2 -.3 -:,4-.8.6.4.2Cy 0-.2-.4-.6-,8-Lo-16 +2 -8 -4 0 4 .8 12 16Angle of sidesfip, , deg.1q o-.1.08 I I I i I I I I II 1.04c o-.04.1fm o-J.1c“ o-.1 -16 -12-8 -4” 048 1216(b) Horizontal taile at location B.l?mmw 6.CaMxmd.,8.6 I.4.2Cy ()-.2-.4-
20、.6-.8-1.0-16 -2 -8 -4 0 4 8 12 16Angle of sideslip, , deg “.1Qo-.1.C.04co-.04.10%-. I.1-.1 -16 -12 -8 -4 0 4 8 12 16Angle of sties-lip, , however, it was neglected in the plotting ofthe curves.The effect of horizontal-tail span on the rolling-momentdwivmtive of the horizontal tail CPEis evident at a
21、llhorizontal-tail locations and appears to become increasinglymore important as the horizontal tails are moved closer tothe tips of the vertical tail and as the horizontal-tail spanincreases. The effect under the latter condition oan beattributed mainly to the increased distance from the loadingto t
22、he axis of roll. A comparison of the valuea of ClandC,BHindicates that the magnitude of C,b=for the larger tailspans can provide a large contribution to the value of therolling-moment derivative of the complete tail assembly.An indication of the magnitude of the Cl=contribution canbe seen by noting,
23、 for example, that the 40-inch tail atlocations A and E produces a rolling moment of about 35 to40 percent of the rolling moment of the fuselage-vertical-tail combination. The unsymmetrical appearance of theClb= data for corresponding spans relative to 0+=0 fortail locations B and D can be attibuted
24、 to the influence ofthe stub fuselage. A direct indication of the magnitude ofthe fuselage titerference can be obtained at location C sincethe absence of a fuselage would result in a value of Clb=ofzero for all horizontal-tail spans. The effect of horizontal-tail span on Cdshows a very strong influe
25、nce at location E.For spans less than the vertical-tail span (20 inches), thedominating effect at location E appears to be the additionalvertical-tail loading and associated center-of-pressure shiftproduced by the end-plate effect. For horizontal-tail spansabove 20 inches, however, the rapid rate of
26、 change of Clflwith span is mainly the effect of CrH. At location D, theprincipal effect of span on C,Dcan also be traced to the C,flRcontribution, as can the variation appearing for locations Aand B for bE values above 20 inches. For values of bH lessthan 20 inches, the interaction of the rolling-m
27、oment deriva-tive of the vertical tail and C,P=provides only minor spaneffects at locations A, B, and C.ECT OFLOCATIONON STAEIIXPYDERIVATIVEThe experimental lateral-stability derivatives C.8JC,p,andClp=axe presented in figure 9 as functions of vertical locationof the horizontal tail. Horizontal-tail
28、 location appears toinfluence Crdirrespective of horizontal-tail span, the greatestvariation occurring for horizontal-tail locations near eitherextremity of the vertical tail. On CZP unpub-lished results obtained for some wings by the finite-stepmethod employing 10 or 20 discrete horseshoe vortices
29、showthat variations in c,= of this magnitude or less provide onlyminor changes in the span-load distribution as comparedwith the stricter solution involving a relocation of the con-.trol point at rLposition other than the three-quarter-chordpoint. The cl= of the horizontal tail was assumed to beequr
30、d to the c=of the vertical tail; this assumption, althoughnot strictly correct, would appear to be a reasonable estimateof the cl= of the horizontal tail. The rwults of figure 10,corrected for section-lifurve slope and including an estima-tion of the fuselage effect, are presented in figures 11 and
31、12and compared with cm-responding experimental data. Theresults are presented as a function of location in order thatthe cnlcuhted values for the 10-inch horizontal-tail com-binations, which have the most accurate estimation of thefuselage effect, can be directly compared with experimentaldata. The
32、comparison in figure 11 shows excellent agree-ment for CYPat all spans. Good agreement was also ob-tained for C,B, the largest variation between experimentaland calculated results occurring for the 10-inch horizontaJ-tail-on configuration.In iigge 12, the CBof the complete model has been brokendown
33、into its component parts, CIPHand CZ%; the fuselageis included in Cl however, it was considered negligible. As shown in figure14, in which the computations for four and eight horseshoovortices distributed across the vertical tail for the 10-inclIhorizontal-tail configurations are presented, a large
34、part ofthe effect of span may possibly be the result of the numberof horswhoe vortices employed. Figure 14 appears to indi-cate that, although the variation in A6/A due to span maypossibly be reduced by increasing the number of horseshoovortices deiin.ing the configuration, the values of A,/A1.6, 40
35、 Finite-step med-Reference 3I.4Aeii1.2 I.OA IB c DEWtbl b031bII of twizontol toilFIGURB13.Calculated values of AJA for the vertical tall by theilnite-step method ae a function of horizontal-tail location for varioushorizontal-tail spans compared with the synthesized rceults ofreference 3.Provided by
36、 IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-AERODYNAMIC CHARAC?ITJWS!MCF3OF AN UNSWOPT TAIL ASSEMBLY IN SIDESLR? 367Vertical Iocotkm of horizontal tailFIGURE14.Vah.Es of AJA for the vertical W determined by thefinite-step method as a funution of horizontal-
37、tail location for 10-Inoh horizontal tails only, using four and eight howhoe vorticesacross the vertioal-tail span (with one and two ho-hoe vortiw,respectively, on eaoh panel of the horizontal tail), oompared withthe synthesized reaulta of reference 3.,predicted by the finite-step method would be co
38、nsiderablysmaller than the synthesized results of reference 3 for hori-zontal-tail locations onequarter and three-quarters of thevertical-tail span above the base of the vertical tail (locitionsB and D). CONCLUSIONSThe results of the investigation to determine the effectsof horizontal-tail span and
39、vertical location on the aero-dynamic characteristics of an unswept tail assembly insideslip and to check the suitability of using the well-knownfinite+tep method to obtain the span loadings and, hence,the stability derivatives indicate the following conclusions:1. The use of the finite-step method
40、provided values ofthe static-lateral-stability derivatives that were in goodagreement with experimental results and, in general, appearsto provide a simple and effective means for investigatingspan loadings of intersecting surfaces.2. The addition of a horizontal tail to the stub-fuselageand vertica
41、l-tail combination produced the greatest increasein the magnitude of the lateral-force derivative CFP forhorizontal-tail locations near either tip of the vertical tail.For horizontal tails located at the top of the vertical tailan increase in horizontal-tail span provided a relativelylarge additiona
42、l increase in C=p only for horizontal tailshaving spans less than the span of the vertical tail. Spanslarger than the vertical-tail span provided only mmll addi-tional ineresse9.3. Horizontal tails of all spans located one-quarter andone-half of the vertical-tail span above the base of thevertical t
43、ail produced a decrease in the magnitude of thelateral-force derivative CYPof the horizontal-tail-off con-figuration.4. Variations in both horizontal-tail span and verticallocation produced a strong influence on the rolling-momentderivative of the horizontal tail. o5. The greatest variation of the r
44、olling-moment derivativeof the complete tail assembly with horizontal-tail spanoccurred for the horizontal tail located at the top of thevertical tail.6. The end-plate effect of the horizontal tail, calculatedby means of the finite-step method and expressed in termsof the ratio of effective aspect r
45、atio to geometric aspectratio, indicated a greater intluence of horizontal-tail spanthan that obtained from the synthesized results by usingthe minimum-induced-drag and lifting-line theories. Theend-plate effect indicated by the iinite-step method wassmaller for horizontal-tail locations near the qu
46、arter andthreequarter vertical-tail-span positions than the effectindicated by the synthesized results.IJANGLEy AERONAUTICAL LBORATORY,NATIONAL ADVISORY COMMITTEE FOR AERONUTICS,LANGLEY FIELD,VA.,December ,?!4,1962.APPENDIX AEFFECTOF NUMBEROF HORSESHOEVORTICESThe iinite-step method, as the name impl
47、ies, approximates .0sthe spamvise loadings of the vertical and horizontal tails bymeans of step lodinga with the number of steps, of course, 1 I I I I I 1 Icorresponding to the number of horseshoe vortices used to kUIOted by horhe vortic.07define the configuration. In order to obtain an indication a
48、s CL- Liftii-surface theory, to the number of horseshoe vortices that should be ued to +obtain reasonably accurate results, especially in regard to the GLa.06total load carried by the surface, calculations were performed for n rectrmgular-plan-form wing of aspect ratio 2 to deter- mine the tin.ite-s
49、panlift-curve elope CL=for various numbers .05 Lof horseshoe vortices distributed across the span. The cal- culationa were made for a wing rather than for a tail con.iigu-ration since the computations were much simpler and the .04results were more quickly obtained. The calculated results o 2 4 6 8 10 12Numb& of vorticmare presented in &me 15 and are compared with a lifting-surface theory. With four horseshoe vortices, n value of FIQUaEI1
