Scribd Logo
Upload

Welcome to Scribd!

  • NACA Report 267 - Drag of Wings With End Plates

    Uploaded by

    Gabriel Rodrigues Felix
    96 views13 pages
    NACA Report 267 - Drag of Wings With End Plates

    Copyright

    © © All Rights Reserved

    Available Formats

    PDF or read online from Scribd

    Did you find this document useful?

    Is this content inappropriate?

    Report this Document
    NACA Report 267 - Drag of Wings With End Plates

    Copyright:

    © All Rights Reserved
    Download as PDF or read online from Scribd
    Flag for inappropriate content
    Download as pdf
    96 views13 pages

    NACA Report 267 - Drag of Wings With End Plates

    Uploaded by

    Gabriel Rodrigues Felix
    NACA Report 267 - Drag of Wings With End Plates

    Copyright:

    © All Rights Reserved
    Download as PDF or read online from Scribd
    Flag for inappropriate content
    Download as pdf
    of 13
    REPORT No. 267 DRAG OF WINGS WITH END PLATES By PAUL E. HEMKE Langley Memorial Aeronautical Laboratory REPORT No. 267 DRAG OF WINGS WITH END PLATES By Pact E. Heusr SUMMARY Jn this report a formuta for caleutating the induced drag of multiplanes with end plates is derived. The frictional drag of the end plates is also calculated approvimately. It is shown that the reduction of the induced drag, when end plates are used, is sufficiently large to increase the ficiency of the wing, Curces showing the reduction of drag for monoplanes and biplanes are constructed; the influence of gap-chord ratio, aspect ratio, and height of end plate are determined for typical cases. The method of obtaining the reduetion of drag for a multiplane is described. Comparisons are made of calculated and experimental resulte obtained in wind tunnel tests swith airfoils of carious aspect ratios and end plates of rarious sizes. ‘The agreement beticeen cal- eulated and experimental results és good. Analyois of the experimental results shows that the shape and section of the end plates are important. ante@puction ‘The end plates which are dealt with in this report are fins or shields which aro attached to the tips of airfoils; the plane of the fin is perpendicular to the span of the airfoil, The purpose of this davice is to serve as bertier to the flow along the span and around the tips of the airfoil. ‘This low in a vertical plane containing the span is usually called the transverse fow. By obstructing this transverse flow its kinetic energy is diminished and es a consequence the induced drag of the airfoil is thereby reduced. 5 From wind tumnel tests made et Gattingen (Reference 1) and at the Langley Memorial Aeronautical Laboratory (Reference 2), it hes been learned thet the induced drag may be materially reduced by the use of end plates. The reduction of the induced drag exceeds the additionel frictional drag of the end plates which were used in these tests for all but small values of the lift coefficients. ‘The German tests were anelyzed but the method of anslysis gives no direct, information on the value of the frictional drag of the end plates. ‘The American tests have not been pre= viously anelyzed. Tn this report a formula for caleulating the induced drag of multipianes with end plates is derived. The induced drag is calculated by finding the kinetic entergy of the transverse Slow using the method of conformal transformation. ‘The induced drags of the monoplane and biplane with end plates are found as special cases of this general formula. ‘The formula for the monoplane is used to ealeujate tho reductions of drag for the airfoils tested at+Gottingen end the Langley Memorial Aeronautical Laboratory. Comparison of celeulated and experimental results for airfoils of various aspect ratios with end plates of various sizes are made, Curves illustrating the total reduction of drag for monoplanes and biplanes are constructed; the influence of such major factors as gap-chord ratio, aspect ratio, ond height of end plates is determined. In all oases the frictional drag of the end plates is considered in calculating the total drag. 258 254 REPORT NATIONAL ADVISORY COMMITTEE FOR AERONAUTICS INDUCED DRAG OF A MULTIPLANE WITH END PLATES ‘The longitudinal projection of the multiplane consists of equal, parallel straight lines. The same projection of the end plates will also be considered as straight lines whose direetions are at right’ angles to the projection of the multiplane elements. ‘The induced angle of attack and consequently the induced drag depend upon the induced downwash. It is sufficient then for the purpose in hand to consider the transverse flow in @ vertical plano at right angles to the air stream. ‘This two-dimensional transverse flow may be determined in seversl ways. Tn this report it has been obtained by using the method of conformal transformation. The rectilinear polygon bounding the longitudinal projection of the wings and end plates is transformed into a single straight line, The flow around such a line is well Inown, Let tho plane of the lines with end pieces be the plane of a complex variable 2 as shown in Figure 1. ‘The mid-point of the system of lines is at the origin and the direction of the lines coincides with that of the axis of imagi- naries. The end pieces CE and C’Ly ™ are then perallel to the real axis and are each of length 2k. The bounding ree- tangle BFF’B’ has dimensions 26/x2H/. Itis assumed that ASH’. ecie ro .,~ The transformation to-an auxiliary cere Tie “* tplane is effected by means of the pote theorem usually called the Schwarz | fe Christoffel theorem. According to this i theorem, if 2=2+iy and t=f+in then any polygon bounded by straight, lines jin the e-plane ean be transformed into the axis of £, points inside the polygon corresponding to points on one side of the axis of & ‘The transformation is Fr, t--Conll eesti of ile with ad plates a 2 ex Grate + doa) @ where a, a, -~-~- eq are internal angles of the polygon in the z-plane and &, &, --+ & are the points on the axis of & thet correspond to the angular points of the polygon in the zplane and A is a real constant. In our epplication we shall consider the polygon ~ ©A BC D EF += in the zplane as transformed into the upper half of the plane. By the theory if the point ¢= = be taken to correspond with one comer of the polygon the corresponding factor in the expression for dz/dt is omitted, Also we may choose three values of £ arbitrarily since a polygon similar to a given polygon of m sides may be constructed by means of n—3 relations between the sides. ‘We shall accordingly choose the three points es follows: D F o 2 ib N+ oo to fe Then to the points E and G will correspond the values Ec and = respectively. The transformation is then found in the form . de (f=) @ where «is a real constant. DRAG OF WINGS WITH END PLATES 255 ‘We may integrate (2) by substituting ¢=snw (mod. ik). sno denotes an elliptic function and from the relations given in textbooks on these funetions we find, eof] Ze)-2(1-F-4e)]+0 ® Here Z (w) is Jacobi’s zeta-funetion. ‘The correspondence between the various points is exhibited in the following table: A B c D EB F G ° WH, ht, |, Ae, WEY, OM, ° 4-Uk, -1, -¢, eo 1, 1k, 2 -K+ik', -K, ~o, 0 wy E, Evik’, ik E and E aro the eomplat allipti integrals of the Gust tad cecond kind, K’ being the eomplete elliptic integral of the first kind for the comodulus i" (X?=1—I). Since cen ay we may now use the relation 1-H ent =Z" (w). This relation is also found in textbooks on elliptic functions, respect to the variable a. We then have ‘The prime accent indicates differentiation with 2= [-o2! (@) +Z (o)] +8, a and 6 constants. wo Using (4) and the correspondence given in the table we have the following equations: Flee L! (oo) +Z Co) +6=h+i0" @) FIRKZ Wt 6=V +i @) fl- BHR) 2 (oo Eh v8 @ Solving these we have: B= ELE 2! (a); io” W=- E12 (wll vag lee wt de] BF ad o42! () BL (alt ge ® B= EZ (od) Oe eee ® It ban 2 dar et oRy (20) a _ 2b’ (K+ E'y) $-B EEE) any (12) 256 RBPORT NATIONAL ADVISORY COMMIPTEE FOR AERONAUTICS The transverse flow around the lines with end pieces is parallel to the real axis at infinity in the splane, the direction of flow at infinity being in the direction of the positivez-axis. ‘The potential funetion for this flow is more easily expressed in terms of f. It is = e,tiys— a Ve where Vis the velocity of the flow in the «-plane at a great distance from the origin. If we superimpose on this flow one whose potential function is =~ 'z we shall have the potential function for the flow set up by the motion of the lines in‘a fluid otherwise at rest. ‘The lines will move in the direction of the positive reel exis in the zplene with a constant velocity To find the kinetic energy we evaluate the integral defied 32 as (Reference 8, p. £4). ‘The path of intogeation, Z, isthe complete boundary of the wing, » is the real partof (a Ft— T's), 2€'5 the normal velocity component slong the boundary and de is the element of length of the boundary, We have Ta 9p Feat E 5 (or) -«] By using again the relation 1-2 —Met= 2’ (a) as well as equations (10) and (11), we have A, PU aE ht Key 5] (3) ‘This may be written T=5 9 V'Ay where at RE RT, Ep fo EEE Ek tte a4) Prandtl (Reference 4, p. 48) hes shown that, if itis assumed that all air particles within an ait stream of crosa sectional area 4,7 are uniformly deflected downward by the wing and all outside are not defected, then the correct lift and the correct work due to the induced drag are obtained by application of the impulse theorem and the energy theorem. Doctor Zahni has suggested that yy be called the arca of the equicalent air stream and this term will be used throughout the report. As the minimum induced drag of the wing system is known to be Demge4- where Lis the lift end g the dynemie pressure, it may be computed by using equation (14). THE MONOPLANE WITH END PLATES The area of the equivalent cir stream of the monoplane with end plates is found from (14) by setting 7=0. ae 2k (2E_ yi) ) If the span is 3 we have Aan RUDE (18) This reduces when A’ =h=0 to w= which is kmown to be comet. ‘The ratio of the area of the equivalent air stream of the monoplene to thet of the mono- plane with end plates is R-—ee (18) BE ye se) DRAG OF WINGS WITH END PLATES 237 Values of R were calculated in the following manner: When y=0, 2’ (x) =0, and ey was then found from a table of elliptic functions (Reference 5) after the modulus & had been arbitrarily assigned. is usually expressed es the sine of e modular angle, @ ‘Then 2h/b and R were caleu- lated using equetions (8) and (16). ‘Table T contains the values of R for various values of 24/6. The quantity J as given by ‘Nogel (Referenco 2) is related to R by the equation By ‘The value of B obtained by using the approximate formule . 2h A s=Eross15(F) is also given in Table I. Let us consider now an airfoil whose span is b and chord c. We shall assume that the longi- ‘tudinal projection of the airfoil is a straight line. ‘The width of the end plete is taken as a chord length where no statement to the contrary is made. ‘The coefficient of minimum induced drag is then Using (16) this may be written Co= EBS. an The use of end plates introduces en additional frictional drag, ‘This frictional dreg may be designated Dr and is given by foe ‘The end plates act es airfoils so that Cy is a drag coefficient and S’ is the area of the two end See ee oe eee nee Cop, is Co~ Oop EFS + OF Cus. ‘The same expression for a wing without end plates is “2°. ‘To determine the actual gain with a given set of end plates we subtract these two expressions. If we call this difference AC, we have Os Mom Si UB) Cry as) If the width of the end plate is a chord length and the height 2h, then S’_4ch_4h sb e ‘Then a9) It we use we have eo 258, REPORT NATIONAL ADVISORY COMMITTEE FOR AERONAUTICS 7 From this the reduction of drag for a given end plate may be easily found. Figure 2 shows parabolas abtained from (20) by using a fixed value of B/S, Cr, and different sizes of end plates, For smaller values of B/S the slopes of the parsboles ineroase but the intersection points on the vertical avis are not altered. ‘There is consequently # greater average reduction of drag for the smaller values of 54/8 over the usual range of values of the lift coefficient:— ANALYSIS OF WIND TUNNEL RESULTS ‘The Gottingen tests (Reference 1) were made with single airfoils. The end plates were flat and of various shapes and sizes. ‘Two aspect ratios were used, 8/3 and 4/3. In the analysis made of these tests the frictional drag of the end plates was not actually calculated. In this report this quantity hes been calculated approximately. Figures 3, 4, and 5 show the results of the Géttingen tests, for aspect ratio 8/2, in the form of polar curves. In each case except for end plate 3, Figure 4, the reduction of drag was computed from formula (19) using the appropriate value of aspect ratio, ratio of length of end plate to span and area of end plates. In the case of end plate 3, since the end of the wing was only partially shielded, an average height of the plate, over the chord, was used. The value of Op which gave the best agreement for the frictional drag of the end plates is 0.0140. ‘This calculated reduction of drag was then i Jaid off on each diagram starting from the polar od LL obiained when using the wing without end plates. Unfortunately, no tabular values of drag and lft coefficients were given in the o Géttingen report. Consequently, no accurate | (/4/) comparisons of ealculated and observed redue- 2 ; t tion of drag could be made. ac | \| A On the whole, the agreement between cal- a rt culated and observed reductions of drag is | | | good for that portion of the polar curve where q = the angles of attack are smaller than the Sree burbling angle. ‘The use of a partial end plate -0)- et such as end plate 3 is notsatisfactory. In this | | instance the disturbance of flow around the end ~o044 | of the wing apparently increases the frictional drag at low values of C; and the reduction of on wid eo plates, Cased duction of deg, Pag is notras great at the higher values of Ci. seit i rates aes ai. a-hole ot "The American tests (Reference 2) were cafe tee aati Drewes O-C ade at the Langley Memorial Aeronautical Laboratory. Single airfoils, wing section N. A. C. A. 73, with aspect ratio @ were used in these tests.” Only two plates, one circular, the other trapezoidal, wore tried. ‘These plates were streamlined and consequently the value of Cr to be used in making the calculations is different. The value 0.008 (Reference 1) proved satisfactory. ‘The results of these tests are shown graphically in Figures 6 and 7. ‘The reduction of drag was calculated as before, Op only being changed in value. Since tabular values are available much closer comparison could be made. The celculated and observed reductions of drag are shown in tabular form in Table II, and graphically in Figures 6, 7, and 8. ‘The agreement is good, considering that we have only’ an approximation to the acutal’reduetion in eonsequence of the simplifying assumptions made in the valeulations.. Comparing the two series of tests we seo that theory and experiment show a Jarger redue+ tion of drag for smaller aspect ratios. ‘The use of end plates with good aerodynamical profile sections is preferable since this reduces the drag for all values of Oz. This is illustrated by the two series of tests studied; in one the coefficient is 0.024, and in the other 0.008. ‘The agreement between caloulated and observed reductions of drag in the American tests suggests using an end plate of given aren in the best possible way. Since the use of the entire height of the trapezoidal end plate (ig. 6) in the formula gave calculated results agreving very 0 Gz ad as as 10 12 8 DRAG OF WINGS WITH END PLATES 259 — a a do 40} endoltet ch ms) oe s 6 4 6 3 +x 4 LL 3 f al rng reduton of dg in. Petr ine ent ‘d rpetpsel et gle tt re] = &} | 4 banana L [eararen?o| ° g TT | a 7 si Z aN ae as (a ee ae a ae Pa cave of ontaum wt merog dn ng‘. 7A ing of Slo _ Saas tgerates ober SALES ae Sy 24 ou | i Z I Tepesotiel end Aotes Dah end pietes zo oe. Ll | od PEE ia femal deserved! O Observed O+ fee d Sen 2 Seem, 2 Pe San t _ . Tolle £ aa oft 7 | Hr gepote}| #8 : eee & B14 | tear" 2 & a 2 3 G2 04 06 Of 0 G2 04 G6 B8 10 . - f a & ie, Soar caret of Gotingea tate showing cegusion ot Fis. §~Colelaled snd eberved dicnces of drug of wings wth dle! en plat rope ood pistes Gand Aspe atl, 3 RG a. manel, aspera 260 REPORT NATIONAL ADVISORY COMMITTEE FOR AERONAUTICS t = \Comadted poler| : {jt | eer || fi oat a wit | oa al ‘“ = | {jit 4 + HE & orvictet| | | ZI rere | - bi | | i F FLL a lnleneaheateg | 291 = i Boer cunye oF Pet | | Beis wibening eerie f Sree rereoseu by Lt 4 fl ore pendpioies =F 4 i | Ui i i L ol 4 G2 04 05 OF 011416 ya [oe [ae fs [eae @ ‘io. sL—Bipines wither pass. Coealeted reduce ofa, Co pleted send fering wing aca" aga i cotati ero als of. hebeihe of nd plate. 20 ‘plebrdos.On Dee este, Como Fo,0—Cleltedrtustin of dg ong ik ene Tea acount eqn en fad plates NAO. A, Magee ato, gap. Spanier Val _ asl Lp. | od | A 05+ aa 1 fe | . od " att 4 | a\ | Ace } | { | | | | J | 16 o1| 4 4 C poe re = | { { i 1 of + + | pos et T | | Zc 4 -0/ = | EES : TH. 7 EEL segchae as as Bs Tee i Fic, 12.—Biplnacs with end plates. Calculated reduetion of drag, &Co, {eh tipa w plt_ Chd eaeg, atnd Piles anne ser t acinomae eee es re fare ae eae eae ome ose eee se eeaenl cee en renee well with observed results, it appears that an end plate of a given area, and therefore having a fixed frictional drag, may have a greater effect on the drag in one shape than in another. In particular, the tests and calculations with this end plate make it scem plausible that. the greater part of the restriction of the transverse flow occurs over the forward portion of the span and that a shape which has greater height in the portion nearer the leading edge is edvan- tageous to use. oe Nagel (Reference 1) has stated that the reduction of drag when using end plates, doe’ not equal the corresponding reduction obtained when the.span is increased by an amount equal to the sua of the heights of the end plates. ‘The reason for adding to the span an amount. equal to the total height of end plates is not quite evident. ‘The addition to the span might also be made, with as good reason, by adding to the wing area the sum of the areas of the two end plates. Such a comparison is shown graphically in Figure 9. The polar curve of an airfoil is shown eompared with the computed poler curves when end plates arb added to the wing and DRAG OF WINGS WITE END PLATES 261 when the wing area is inereased by an amount equal to the total area of the end plates. The ‘comparison was made using the same drag coefficient for the end plates as for the additional wing area. ‘The results do not differ much. CALCULATED REDUCTION OF DRAG FOR MULTIPLANES WITH END PLATES ‘The method will be used to calculate first the reduction of dreg for a biplane. The area of the equivalent air stream for a biplane with end plates is given by equation (14). The end plates must have heights equal to or greater then the gap of the biplane, for otherwise the flow sould change to such an extent thet (14) would no longer hold. ‘This formula reduces when A=W, thatis, when the end plate hes # height equal to the gap of the biplane, to one which has been previously derived (Reference 6). It is evident from the correspondence of values between s,tand a that when = RaW, oe E, Z (wo)=0 and Z! (w,) 1B From (10) we find then For a span 6 this reduces to which is the value previously found. : ‘The expression for reduction of drag is derived in a manner quite similar to thet used for the monoplane. We find ee 2h) aco FE ay, oF). b=span of biplane. c=chord. Oy =lift coefficient. -Ayarea of the equivalent air stream of a biplane without end plates. Ay, area of the equivelent eir stream of a biplane with end plates. Cp= drag coefficient of end plates. 2h=height of end plate (width assumed to be c). It wo iatroduco Prandi’s ratios 2 and « similar ratio en qipcwe heve C022 cen) 0r(%) 22) ‘The value of x, depends on both the gap/chord ratio and the ration of height of end plate to gap. ‘The former ratio is 7 and has been chosen in each calculation. The calculations were made in the following menner: ‘ x (6) Z (a) was found from Z' (ws) =p FH! Alter 7 hes beon arbitrarily assigned. (b) From dnte,= +2 (o), ay was found using the table of elliptic functions (Reference 4). (©) The value of FZ od —s47' (8) wis calculated next. Tet EZ oo) (@) Ayn EEE EY [2 Ewe_2 2” (y)| was evaluated then quite exsily. Figures 10, 11, and 12 show the calculated reductions of drag for span/chord of 6 and gap/chords of 0.6, 1.0; and 14. In each case the following values of h/h’ were used; hi’ = 1, 2,4, 6 and, as in Figure 2, the value Qp~.01 was used as it represents an average value of Cr. 262 REPORT NATIONAL ADVISORY COMMITTEE FOR AERONAUTICS ‘The curves illustrate how properties of end plates affect biplanes of various proportions. ‘The incroase of drag at lower values of Cz is considerably less than it is for larger values of gap/chord while the reduetion of drag, especially when larger plates are used, is relatively large. ‘The reduetion of drag persists at the higher values of C, for plates which ere quite large, so large as to be useless in practice. ‘The method used in ealoulating the effect of end plates for biplanes may be used for multi- planes, Formula (22) would differ only as far asx is concerned. This quantity may be cal- culated, however (Reference 7), and the reductions of drag for multiplanes may be calculated in precisely the same manner as that just described and carried out for biplanes. CONCLUSIONS. Caleulations show that the induced drag of monoplanes and multiplanes may be decreased. by attaching end plates to the ends of the wing. The frictional drag of the end plates may be calculated approximately. ‘The reduction of the induced drag exceeds the additional frictions! drag due to the end plates at ell but small values of the lift. For given dimensions of wings and end plates the reduction of drag less the frictional drag of the end plates varies direetiy as the square of the absolute lift coefficient. ‘The average reduction of drag increnses as the aspect ratio decreases. Calculations and experiments agree quite satisfactorily for single wings equipped with end plates. ‘Wind tunnel tests show that the coefficient used. in calculating the frictional drag of the end plates may be reduced materially by fairing the end plates. ‘The shape of the end plate determines to some extent the reduction of induced, drag. Further tests would be necessary to find end plates of good shape and having at the same time a low drag coefficient. Recent. experiments have shown that much higher lift coefficients can be obtained then has been the case up to now with the conventional airfoils. Since the reduction of drag due to end plates is much greater at the higher values of the lift coefficient, the use of end plates together with the use of means for increasing the lift will result in a material improvement of the efficiency of the present-day airplane, Laneury Memonsat Annowavzican Lasonarory, Narronat Apvisony Comrise ror AERONAUTICS, Lanier Piezo, Va., January 20, 1927. ‘REPERENCES 1, Naoze, Fz Plage! mit seltiichen Schelben. “Vorlanfige Mittellungen der Aerodynamischen Versuchsan- salt 20 Gottingen,” No. 2, July, 1924 42. Reso, B. G.: The Eifect of Shielding the Tips of Airfols. N. A. C. A. Technical Report No. 201, 1925, 8. Lan, E.: Hydrodynamies—Fitth edition, 1924, 4. Praxori, Lt Applications of modern Hydrodynamics to Aeronautics, N. A, C. A, Technics) Report No. 116, 1028. 5. Hiprisuzy, R. Le: Smitheonian Mathematical Formulae and Tebles of Eiptie Functions, Smithsontan Miscellaneous Collections, Volume 74, No. 1, 1922, 6, Tragiligel Theorie, zweite Mittelung. (Nachrichten der Koniglichen Gesellschaft der Wissenschaften 2u Gottingen. Mathematische-Physikallsche, 1919.) 7. Prani, La: Induced Drag of Multiplanes. | N. A. C, A. Techateal Note No, 182, 1924. ‘DRAG OF WINGS WITH END PLATES 263 ‘TABLE I Values of and 2/9 for given values of ‘area of the equivelent alr stream of & monoplane)/(ares of the equivalent air stream of a monoplane with end plates). (nelght af end plste)/(span of wing). i met cd ie | co | oon a8 | %B mt 3 ih 1 3 | dos 8 | za 3 EB i 2 | ae 6 | a8 | 8 ' 8 | 3 ' ae if iB a | en (3 Hl 80 | +152 8 a i$ 2% | I canal TABLE I OBSERVED AND CALCULATED REDUCTION OF DRAG (Tests made at Langley Memorial Aeronautical Laboratory. Mozoplane, 4 by 24 inches; win ‘ meal section, NAG. A. a) one 4 by * j (x end plates) [pinerease of Go (rapctotdat end plate) | pizerene of obseeved and} lobeerved andl eateutated | salevsted ce reduction ae seduction as Cateiated | "percent of dine fo +390 mo18 joa oa =te \ 12 Eo = | 3

    You might also like