1、Lopy NO. Ab * . . RM No. A6K15 I RESEARCH AAEMORANDUlw AN INVESTIGATION OF THE LOW-SPEED STABILITY AND CONTROL CEAR.4r:TERISTICS OF SWEPT-FORWARD AND SWEPT-BACK WINGS IN THE AMES 40- BY SO-FOOT WIND TUNNEL -c- BY Gerald M. McCorrsack and Victor I. Stevens, Jr. Ames Aeronautical Laboratory Moffett Fi
2、eld, Calif. NATIONAL ADVISORY COMMITTEE FOR AERONAUTICS WASH!NQTOM June 10, 1947 Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-NalionaI AefonB and , Soace Admmtralton . Lutplmy Re6earoh Cmter FROM: alJEJscT1 Diatr ibution lSO however r UUtil IMtarf
3、d ia re-marked by lining through the classification and annotating with the followin9:statempnt, hmust continue to ba prottated aa it olaslficdr .* . . n Declassified by r, be attained beflre serious. compressibility cffccts arc enc%?ntercd; Theory and experiment also show that -sing sweep introduce
4、s a number of I . -. . .- .- _ i stability and. cqntrQ1 groBloms, .thc scrixsncss ?3f which -d%c 4 bocs accentuated a,-t.low flight speeds. _ t- $lti11-EW.lf2e tC6t6 haVC PCilltCi? out the genera,1 nature of thcso problems and indic ted those which must be overcome if the high-speed bsncfito ctf swe
5、ep are Gq be realized. They r have ah3 s?zggste-d thtkt boundary-l.q-cr flop and, hence, -. ,_ _ ._.- Xficynqlds number has a qrofQund ce 3n measured charac- teristics and that the va1u.e of, small-scale tests remain srlmelarhat doubtful until the extsnt of this influen.ce is under- stood. and the r
6、esultaarereportcd herein. It is believed that Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-. pr these data.trill QCI far towards establishing the datum required t3 est5nP.t.e the effects qf scale m highly swept wing ;?lan f3rizs. With this Lmqwled
7、ge at hand it is evident that the value qf future snaLl-scale tests w3ll be cmsidcrably inc-eased. This repmA discusses a smmry qf the basic results and cmpares them 74th simple swept+%ng thcqries and, r=rherc pmsible, with existLng mall-scalzdata (rcfcrences I, 2, and 3). T? nake the basic data avF
8、A.lzdod as an appcndLx t? this report. L l l The five nodels 3ested were cmpgsed of wfng panels from an available airpUzxwhfck wzm given-the desJ.red plan fxm and swceg by Individually fabricated tips and center eectims. Tke resulting angles of swceg were 03, 30“, aind configuratfons six-component f
9、orce and moment data were obtained through an angle-of-attack range at each of several angles of sideslip. The data were obtained at dynamic gressures- which range from 5 ti 75 pounds per square foqt (R = 2.9 x 106 tr, R = lG.0 x 106ja; most of the data were obtained at dynamic pressures of 10 to 20
10、 pounds per a square foot (R = 4.0 x lo6 and R = 5*3 x 106, respectively). . The basic data obtained from SIG wind-tunnel tests ?f the five swept wings are dcscribcd in the appendix. AI-s? included in the appendix is E description of tne corrections and tares an?lied to the data. ix3 cuss I OF In th
11、is dfscussion an,evaluation is made of the effect of wing sweep on the more important aarondynamic parameters and of the consequent effect on airglene gcrf3rmance and stability. a-These Reynolds numbers are based upon the 1i.is.C. as a refer- ence lonqth and are the mini.mum and maximum Hnits of the
12、 variation fncluding tIic change in chord length with sweep. Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-$ NAGA FUrI No. A wheieaa for swept-forward wings, basing the aspect ratio on the conventional span gives the better agreement. It is believe
13、d that neither of these aspect ratio concepts gives a correct picture of the induction effects of the vortex pattern on swept wings. It can be shown that If a wing is swept back, the induction influences of the trailing vortices on the wing ehould be reduced, and conversely, if a wing is swept forwa
14、rd, the induction influences on the wing should be increased. That is, the effective aspect ratio increases with sweepback and decreases with sweepforward for wings of constant geometric aspect The lift-curve slopes for the been estimated using the method of rati (b/S). Wings Of this repmt have FaLI
15、mer (reference 5) whioh “It is recagnized that a further aspect ratio, correction, namely, Jones edge-velonity correction should be used. The effect is small, however, comnared to the errors resulting from the use of simple -sweep theory, It has been omitted, therefore, in an offort t-o indicate Cle
16、arly the adequacy of simple sweep theory in indicating the lift-curve slope of highly swept wings. Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-Ni4CA RM No. APip*cs 4 and 5 shows that values of lift-ourve slope determined fr39 tigl.-scsle tests sh
17、ow no better or poorer agreement with simple theory than values from small-scale tests. It appears th:. t the principal disa- greement between theQry and experiment lies in failure qf the theory to properly account for the induction effects 3n swept wings and that in comparison the effects of scale
18、are rela- tfvely small. Examination of the nonljnear portion qf the lift curves and comparison with small-scale data shows that no consistent effects exist which cauld be attributed directly to scale effect. ThQse differences which dg exist are emall and erratic in nature and probably result from di
19、fferences in plan f9rm, wing section,and local wing roughness. Flap effectiveness.- According to simple sweep theory, flap effectiveness decreases as csA* An additional correctkqn to account for induction sfi“ects must be applied Provided by IHSNot for ResaleNo reproduction or networking permitted w
20、ithout license from IHS-,-,-,XACA F3$ No. 615 13 when comparing flap effectiveness on wings of different aspect ratio. The cormients previxzsly made regarding the effects of induction on lift+urve slope a?pLy equally well to flap effectiveness. In fact +?hen a, Chen prediotions are made in terms of
21、lift-curve 4Note that in correcting for aspeqt ratio, the aspect ratio was based on, the-span. As in the caee of %g if aspeot. $atios were baeod on the len better agreement in flap lif P h of the quarter-chord line increment would have been obtained for swept-back wings and poorer agreement for swep
22、t-forward wings. Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-14 NACA RH NO. 615 slopes, the agreement with experiment is almost exact. Tim s the control which Ck has on flap effectiveness emphasizes the importance of fully understanding the effec
23、t qn CL, of the many factors involved; this is especially significant when flap effectiveness is considered in terms of airplane control and performance. Since the flap lift increment is dependent upon lift- curve slope, the conclusions concerning the effect of scale on lift-curve slope apply equall
24、y well to flap lift inorement. In general, it can be said that sweep introduces no new scale effect on flap lift increments measured +-low angles of attack. Haximum lift.- The effeat of sweep on maximum lift of the wing without flaps, with 0.623 span flaps deflected 600, and with full-span split fla
25、ps deflected 603 is shown in figure 23. Attention is called to the fact that the-wings tested. were composed largely of production wing pnels with normal rqughnesa and irregularities such 8B caused by access plates. As a result qf the roughness, maximum lifts measured on these wings may be somewhat
26、lower than those measured on smooth wingsc However, since the measured values on the unswept wing appear to be reasonably high for the particular airfoil section8, it is believed that the roughness was not sufficient to seriously reduce the maximum lift measured. As shown in figure 8, sweep in wing
27、plan form produces serious losses in maximum lift. KowevcTJ-FP -all but one of Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-the wing configurations the measured 1x2.mum lift was equal to or greater than simple theory would indicate, that is, CLmax
28、 did not in general decreage pr9portioral:to c?s“wIng tne whereas on the swept vlngs tne.gains in maximum lift coefficient are, Fqmewhat iesTfth sweep are disan- .poi.nting.but n% surprIeing, since ricer stall the air flaw is separated 3n the ?utboard section qf s-wept-back wings and the inboard sec
29、tion pf s!?ept-fvward wInga. Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-1.6 IJaC whereas the increases for the swept wings fall far short of what would be anticipated from experi- ence on straight wings. These data indicate that l,?rge-scale tes
30、ts show a much more rapid decrease of cr, nx with sweep than do model tests. This seems true whether flaps are deflected or not. Since large-scale results tend to approach small-scale results at large angles of sweep, considerable care should be taken in trying to estimate large-scale ntilane perfor
31、mance from s%;ept-wfng model tests. Expectations of irnprovkg kas commensurate with that experienced at zero sweep are not likely to be fulfilled. The importance o*f this problem would indicate a pressing need for aw-ept-v$ng tests of-a-number of-given models throughout the full Reynolds number rang
32、e. Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-It should be note whereas the L/D ratio8 of the swept-forward wings show a rapid decreaee. Longitudinal CharactorJ.stics The effects of sweep-on the pitching-moment characteristics of the plain wing,
33、 wing-with partia L-span flaps,and with full- span flaps are-shown in figures ll.to 13. The renarks which follow are based upon the data obtained on the plain wing (fig. 11) but in gcneralap;,ly also to the wings with partial- or full-span flaps (figs. 12 and 13). Fcr lift coefficients less than.0.5
34、, the pitching-moment coefficients vary almost linearly with lift coefficient and indicate that forward sweep moves the aerodynamic centor forward (4 percent K.A.C. at This chart is reproduced herein E-G figure 14. further- more, the chart applies to swept-forward as well as swept-back wings. Insofa
35、r as the.over-all shape of the.pitchfng-moment curve is concerned large-scale tests agree generally With small- scale results with the exception of minor diff erencea. Again It should be noted that these compa,risons have been made from exam- ination of results of investigations on wings using conve
36、ntional sections. The validity of the statements regarding theee cam- parisons is as yet unsubstantiated in cases where lamin-flow sections are involved. Later8.l. Characteristics Dihedral effect.- The variation with lift ooefffcient of the rolllfng moment due to eideelip is shown in-figI.q?e 15 for
37、 the plain win and in figure 16 for the wings :tith flaps. The pow- erful influence of sweep on the dihedral effect is immediately apparent. (A scale of effective dihedral for the unswept wing has been shown on the figures to .J.low convenierft comparisons.) Within limits, the dihedral effect due to
38、 sweep increases in proportion to lift coefficient. Both.the 30 and 45 plain swept-back wings reached a maximum value of CI, P of -0.0034 (170 effective dihedral) at lift coefficients of 1.15 and O., respectively. In the case of the swept-forward wing the maximum value of OX B increased with angle,
39、of sweep, being 0.0014 for the -30“ swept wing and O.OOZQ for the -45 swept WiiIg. These maxinm values for the swept-forward wings occurred in both cases new a lift c0ef- ficient of 0.9. It should be noted that Chile the swept-baok wings show much greater dihedral effect than do the swent- forwpxd w
40、ings, this is due largely to the dihedral effect of Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-. NACA RI4 Mo. A Even with adequate lateral control it is felt that a pilot would have difficulty in reacting sufficfently fast to prevent reaching ex
41、cessive angles of bank. For the case where lift fs changed by changing angle of attack (flap deflection constant), simple sweep theory gives the following relation for the paremeter aczp/acL : 6 oczp/w,), = (Wz B/wL)*=o - 114 -g-y+ J 9 6It is recognized that bqth qf the terns on the right side of th
42、is equation should be modified furt.her by a correction involving aspect rati and edge velocity, Simple theory shows that, where asynnetrical lift exists, the corrections wmld be the form A/(AZ+y simil?r cher=cteristice. . ThEt-t is, the-value of BczpL “asproximates thet predicted by theory, with P.
43、 maximum velue of CIB being reached prior to the stall, end followed by a rcducti?n in czB 7.8 the stall is qprocohed. iC3tc that t:kesc znd the followfog considerations regi.r swept-bdck wings tke measured value wp.6 86 nuch af3 14 percent more tkn Provided by IHSNot for ResaleNo reproduction or ne
44、tworking permitted without license from IHS-,-,-24 NACA RM No. X and tp.pcr ratios less than 1.0 tend t7 shift the load center towsrds the root. A comalctc undcrstsnding of this action cannot bc had until more thorough Studies are made of the effc:bts of swcr) e.nd taper on ILwZ! ccntcr. A first app
45、roximation (prob2bl.y cn ovcrcorrcction) of the ansvcr can be rcachcd, however, by simply adjusting the load ccntcr to correspond TV the percsnt by simple theory; that a closer approximation can be had - probably within 10 percent - if the centroid of l?r,d is nasumcd to lie on the centrqid of area,
46、 It is belfcvcd thnt the effects of scale fall p:ithin this latter error snd probably zrc ?f the same gencralmagnitudc as the effects of section 3r tip shnge. No +ta could be found to aid fn a quontitativc evrlustion of thcsc effects. Yith regsrd to the maximum values 3f Cl e likely tq be cncQuntcre
47、d with EL highly spzpt wing, it r;:?e,ars ingosoiblc to . . . Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-conclude more than the feet t-t a nlax decreases, the sexinum value of Cl %fk?eas cz kmx f7r 8 30“ swept-beck T%I: 3ccurEl at O.$O Cbaxr 0.8
48、2 CT %8x and 0.91 CL, fqr the same data, respectively. Since the .phenqmenon : . the choice can be considered conservative, bq-:t for the present, . wfnd-tu,nnel tests must be relied upon to give the exact answer. Certs.inly this prgblen is or?hy of hdditlonal study. Until tl;E; governing factors are more clearly defined it renaiz Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-impossible to determin
