REPORT 1274

SECOND-ORDER SUBSONIC AIRFOIL THEORY INCLUDING EDGE EFFECTS

By MILTOND. VAN Drm

Recently Hayes (ref. 17), improving on a result of hai (ref. 18), has given a second-order similarity rule for surface f3everal recent advanctx in plane 8ub80nic jhw theory are pressure that implies a second-order extension of the Prandtlcombined into a un@ed second-orb theoryfor airfoil 8eciionJ3 Glauert rule (ref. 19). This remfikable result was overlooked qf arbitrary dupe. ~ sobuhbn i8 reached in three 8tep8: by earlier investigators because they did not calculate surby ini!egraiion, it ti wnme inwmpres8ibt?erewdt i8 found face preesums, but were content with finding surface speeds, verted into tti conwponding subsonic wmprawibk redt by for which the second-order compressibility rule is more means of the 8ec07uLorder wmprtxwibili-tyrule, and it is rencomplicated. These rules reduce the second-order problem dered un@rmly redid w -aiion poim% by f& ti. of subsonic compressible flow past airfoils to, the correspondSolutiw for a number of airfoib are given and are compared ing incompressible problem. with the re.sula of other theoria and of expmm.ent. A 8traigtiHowever, the solution by successive approximations forward wmputing scheme is &lined for ca.kdating the 8u.rbreaks down near leading and trailing edges if there are face velocities and premuaw on any aiq6i.1 at any angle of stagnation points. The result is therefore merely a fo~rd atid. series expansion, which fails to converge near the edges. INTRODUCTION In first-order theory spurious singularities arise at stagnation Thin-airfoil theory provides a useful first approximation edges, but it is known how they can be taken into account, to the incompressible flow past two-dimensional airfoils, since they are integrable. In the second approximation, and the results can be immediately extended to subsonic however, these singularities are intensified, so that at round compressible flow by the Prandtl-Glauert rule. It is natural edgea they are no longer integrable. In any case, the calcuto attempt to improve this simple theory by auceessive lated speeds and pressures are incorrect near such edges, approximations so as to increase its acxmracy for thicker and more so in the second approximation than the first. airfoils and higher subsonic Mach numbers. There results Moreover, in subsonic compressible flow the second approximation may be incorrect everywhere as a consequence a swiss exTansion of the flow quantities in powers (suppleof the defects in the first approximation. mented by logarithms in the fourth and higher approximatFor round edges in incompressible flow, previous inions) of the airfoil thickness ratio, camber ratio, and angle ve.atigators have shown how these defects can be corrected. of attack. l?or incompressible flow, the higher-order theory haa been Riegels (ref. 2) gave a simple rule that renders the firstorder thin-airfoil solution valid near the edge. LighthiU studied by various writers, in particular Riegels and Wittich (ref. 5) gave an equivalent rule for the second approxima(mfs. 1 and 2) and Keune (ref. 3). A less straightforward series of approximations was developed by Goldstein (ref. tion. Recently, corresponding rules have been developed for higher approximations, for sharp as well as round edges, 4). Perhaps the most concise exposition of higher-order and for subsonic compressible flow (ref. 20). incompressible thin-airfoil theory is given by Lighthill (ref. 5). It is the aim of this paper to combine these recent advanoes There results a uniform second apFor subsonic compressible flow, the corresponding analysis into a unified theory. proximation to subsonic flow past any profile at angle of was first undertaken by Gtktler (ref. 6), followed by Hantzsche and Wendt (refs. 7 and 8), Schmieden and Kawalki (ref. 9), attack, expressed in terms of integrals that can, if necessary, It may be noted that, except be evaluated numeriedly. Kaplan (refs. 10 and 11), Imai and Oyama (refs. 12 and 13) and others. These investigators treated only specitic simple possibly for certain particular shapes at isolated Mach shapes by rather laborious analysis. Later, it was discovered numbers, the resulting solution is now generally believed that particular integrals of the second-order iteration equato be valid only belo~ the critical Mach number-that is, Although only flow quantities tion can be expressed in terms of the iirst approximation for purely subsonic flows. (refs. 14 and 15). This permits the second-order subsonic at the airfoil surface are considered here in detail, the entire solution for any profile to be given in terms of integrals flow field could be treated in the same way. (refs, 16 and 16). For numerical computation, the most useful method However, the resulting solutions are sometimes incorrect everphere for airfoils with stagnation appears to be that initiated by Riegels and Wlttich (refs. points, for reasons to be discussed later. 1 and 2) and independently by Germain (ref. 21), with
1SnpersedfsNAOA TN 3W, Slender-IJ&iy Thcary, 1%54. and portfom of NAOA TN 3S4S, Snbwn& Edgesfn Thfn-Wng nnd %umd-Order Snbonfo Alrfoll-Sectfom Thmry and its Rmtf&l AppMatfm, 1E55,

SUMMARY

541

. . 542

.-

MMITIEE FOR AERONAUTICS REPORT 1274NATIONAL ADVISORY CO The corresponding horizontal velocity disturbance on tlm chord line, which is required for calculating the surfnco pressure, consists of a term associated ~ti tile ~irfoil thiclmess, and another associated with its camber and For the thickness angle of attack. (3) and for the camber and angle of attack

extensions by Watson (ref. 22), Thwaites (ref. 23), and Weber (ref. 24 and 25). It requires a lmowledgo only of the airfoil ordinates at a specified set of point~. In this report n straightforward scheme, based on an extension of this method, is given for computing the second+rder subsonic solution for any airfoil. The reader interested only in calculating a specific case, without necessarily understanding the details of the theory, can turn directly to the section PRACTICAL NUMERICAL COMPUTATION. The author is indebted to R. T. Jones for many helpful discussions throughout the course of this work. THEORY From the preceding remarks it is clear that the solution First, the formal second+rder is reached in three steps. Second,. incompressible solution is found by integration. this is converted into the corresponding subsonic compressible solution by means of the second+rder compressibility rule. Third, this is modiiied near stagnation points by the appropriate ruhw for round or sharp edges. These three steps will be considered successively.
FORMAL INCOMPR~LE SOLUTION

%=(2%Y2[+%%Y-1 The latter result is was first given by values are indicated The surface speed

4)

due to Munk (ref. 26) and the formor Pistolesi (ref. 27). Cauchy principal in each integral. is then given to a first approximation by (5)

The expansion of the velocity components in a formal series of powers of tb e airfoil thickness ratio, camber ratio, and angle of attack has been discussed in detail by Lighthill (ref. 5). It will suflice here to summarize his results for the second approximation. We mainly follow his notation except to make it more mnemonic, and to suppress his parameter c characteristic of the airfoil thickness, which is only convenient in the detailed analysis. Accordingly, consider an airfoil of moderate thiclmess and cnmber at a moderate angle of attack to a uniform subsonic stream @g. 1) ~ It is essential that the z axis be chosen to pass through both the leading and trailing edges. Let the upper and lower surfaces o{ the airfoil be described by y= Y(z) = c(z)+ T(z) (1)

Seoond-order solution. In the second approximation, the tangency condition is transferred from the airfoil surfaco to the chord line by Taylor series expansion. The condition on the second-order increment in vertical velocity is thus found to be

T,(z) =c,(z)+ $..0

where

(60)

(Oh)

(We depart here from Lighthills notation in order to emphasize that the functions Cl and TZ me effectively th~ camber and thickness for some fictitious airfoil.) Tho problem is identical with that in first-order theory except for the condition at infinity, which is readily disposed of. Thus, corresponding to Tz is the increment in horizontrd velocity (7) and corresponding to 02 (8)

l?mmm l.Notation for nirfoiL where C@) describes the mean camber line and Z(z) the thickness. The airfoil extends over the intemd X.<X <~bj which is usually conveniently taken to be either 1 <x <1 or O<z<l. All symbols are defied in Appendm A. First-order solution.-In the fit approximation of thinairfoil theory, tbe condition of tangent flow at the airfoil surface is imposed on the two sides of the chord line y= O rather than at the surface, and requires that . WI ~ ,=
u-

The velocity components on the surface of the airfoil include also terms arising from the trader from the chord line to the surface, which is again effected by Taylor swim expansion. Hence the surface speed is given to a second approximation by

and the surface pressure coefficient by

Y-(z) =C(z)+

z(z)

(2j
-2(%

)-(%%)

(lo)

SECOND-ORDER

SUBSONIC

AIBFOIL ~ORY
SECOND-ORDER COMPR=~ILITY RULE

543

Higher-order solutions can be found by continuing this process, Lighthill gives explicit formulas for the third-order solution. Mrfoil integrals.-The incompressible solution to second order (or, indeed, to any order) is thus reduced to a succession of airfoil integrals typfied by equations (3), (4), (7), and (8). Goldstein (ref. 28) emphasizes that in fi.rm%rder theory those integrals can be evaluated analytically for practically wow profile for which formulas have ever bees proposed. In second-order theory this appears to true to a somewhat lcswr extent, although the labor of calculation becomes great Often the integrals are most except for simple shtipes. readily evaluated by guessing (uio) as a function of the complex variable (z+iy) that has the required behavior on the chord line. A short table of airfoil integrals useful for finding second-order solutions is given in Appendix B. Other can be found in references 28, 29, and 30. For complicded proiiles, exact analytic evaluation of the integrals may be impossible or excessively laborious. Then numerical integration may be resorted to, or the profile can be approximated by a simpler shape that can be treated analytically. The most useful numerical procedure is apparently that originated independently by Riegels and Germain and simplified and extanded by Watson, Thwaites, and Weber. In this method the airfoil ordinatea are approximated by the trigonometric polynomial y=~+~ N1
r=l

The second-order counterpart of the Prandtl-Glauert compressibility rule is implicit in an extension of transonic similitude that was initiated by Imai (ref. 18) and carried to completion by Hayes (ref. 17). Inmi sought to improve the transonic similarity rule by retaining in its derivation all terms proportional to the square of the airfoil thickness except one appearing in the condition of t,amgent flow at the surface. The correlation of experimental data was not appreciably improved, which led him to suggest that the neglected second-power term should also be included. This probably wmnot be done for the whole flow field. However, in attempting merely to reproduce Imais result as amounted before publication, Hayes actually included that term in a second-order rule for surface pressure. Hayes result is that for two-dimensional subsonic or supersonic flow the ratio of the second-order to first-order pressure term on the surfam is proportional to the parameter (11) where ~ is some measure of the thickness, camber, or tingle of attack. Now at subsonic speeds the first-order pressure term is related to its value in incompressible flow by the Prandtl-Glauert rule. Combining these two results yields the second+rder compressibility rule (ref. 19). In incompressible flow the second-order surface-pressure coefficimt has the form C%(z) =Cq(z) +@@ (12a)

(k, Cos ti+~

Sill re)+kN 00S NO

where the firs&order term CPl contains linear terms in thickness, camber, and angle of attack, and the second-order incxement AC% contains their squares and products. Then for the same airfoil in subsonic compressible flow, according to the compressibility rule, the pressure coefficient is
CPM=.KG1+K2(ACPJ

(12b)

where

K,= ~~=; 1
(12C) =,=(7+ l) M+4& @

FIGURE

2.Parametrioangle 0.

The corresponding compredbility rule for surface speed is readily found from the above rule for pressure by considering the small-disturbance series form of Bernoullis equation for compressible flow. Thus it is found that if the surface speed ratio in incompressible flow is (13a) where Aql containa linear terms in thiclmess, camber, and angle of attack, and Aq, their squares tmd products, then at subsonic speeds

where 0 is the am.gle indicated in figure 2. The coefficients k, (for camber) and& (for thickness) are chosen to give the actual ordinatea at the 2N points for which O=mm/N. h this way it is found that the airfoil integrals can be expressed approximately as sums of the airfoil ordinates ah certain pivotal points multiplied by stwndmd iduence coefficimts. The details of this method, as adapted to second-order thin-airfoil theory, are given in Appendix C. The nueti~ mmputing procedure is outlined in the last section of this paper.
. 43m375-574u

. ..

54-4
with ==~2

REPORT 1274NATIONAL ADVISORY COMMITTEE FOR AERONAUTICS Lheory. It is enough to take account of this shortcoming in even the crudest fashion. Thus, if the region of integration is extended an intlnitesimal distance beyond the edges to include the pulses (Dirac delta functions) of the thin-oirfoil approximation, Harder and Kluukers expression yields w solution that is correct to second order except in the viciniLy of the edges. Keune has discovered an alternative particular integrnl containing the stream function rather than the vdociLy potentkd, and so has obtained another expression for the second-order solution (ref. 15). Because the tangmcy condition is one degree smoother for tbe stream function then the velocity potential, his expression yields tho correct result (except near stagnation edges) for sharp-edged shapes. It fails, however, for round-edged shapes, so that his solution for subsonic flow past an ellipse is incorrect everywhere. Both these expressions can be manipulated by partial integration so as to be correct except nenr stagnation edges, However, the result is simply that obtained by applying the sewnd-order compressibility rule to the expressions for second-order incompressible flow. Eence these more serious difficulties are of no further concern here. They do serve, however, to warn of the danger of false second-order solutions in more complicated problems. The role of edge corrections.-Thin-airfoil theory fails near stagnation edges because there the basic assumption of small disturbances is violated. It might be feared that uniformly valid solutions could be found only by abandoning that assumption, which leads to such great mathematical simplification. Fortunately, however, Riegels and Llghb hill have shown that for round edges in incompressible flow all the results of small-disturbance theory can be salvaged. They have given simple rules to be applied to the formal thin-airfoil solution that render it uniformly valid near tho edge. For present purposes, corresponding rules are required for subsonic as well as incompressible flows, for sharp as well as round edges, and for cambered round edges, for which Lighthills rule is correct only to first order. It is believed that neither Riegels nor Lighthills technique can be extended to these casw. Instead, a different techniquo is used here, which is particularly suited to the study of edges, It consists in comparing the exact and thin-airfoil solutions for simple shapes that approximate the airfoil in the vicinity of the edge. This technique was &t applied in reference 20 to the surface velocity on airfoil sections, three-dimensional wings, and bodies of revolution. It is reproduced lmre insofar as it appliea to second-order theory for airfoil sections, and is extended to treat surface pressures as well as velocities Iiotmd edges in incompressible flow.-Most, subsonic airfoil sections have finite leading-edge radius, and many aro actually analytic (except at the training edge). This meons that all derivatives are continuous, so that, with s Lho abscissa measured from the edge into the airfoil, the upper and lower surfaces are described by y= +T#~+C,s+

(~+w$+J@ simplicity

(13C) of the

This rule is seen to lack the fundamental rule for pressure. EDQECOREECrIONS

Thin-airfoil theory is known to fail near leading and trailing edges if there is a stagnation point. The flow is actually brought to rest, but thin-airfoil theory predicts . infinite speeds instead. If r is the distance horn the edge, the velocity contains powers of r-I/z for a round edge and for any leading edge with flow around it (associated with angle of attack), and powe~ of in r for a sharp edge. Firs& order theory cent sins &et powexs of these singularities, second+rder theory their squares and products, and so on, so that the formal thin-airfoil series diverges in some neighborhood of the edge. Not only are the velocities and pressure incorrect near stagnation edges, but nonintegrable singularities appear in the higher-order expressions for aerodynamic forces on round edg~. False subsonic solutions.-Even more serious difllculties may arise in subsonic compressible flow, where the infection spreads in some cases so that the formal second-order solution is incorrect not only near the edges but over the entire airfoil surface. Thus, using the particular integral of reference 14, Harder and Klunker gave an expression for the second-order solution for any symmetric airfoil at zero angle of attack (ref. 16). However, they noted that their expression does not apply to round-edged airfoils, for which it contains divergent integrals. A more deceptive defect appenrs if their expression is applied to a sharp-edged airfoil such as a biconvex section; then the predicted surface speeds are fhite (except near the edges) but incorrect everywhere by a term proportional to J&. This defect arises from the fact that near sharp edges the ilrshorder source distribution is not approximately the airfoil slope, as is assumed in thinairfoil theory. The second-order solution involves the derivative of the source strength which, as indicated in figure 3, has iniinite peaks that are missed by thin-airfoil Source \

Actual volues
FIGURE

Thin airfo]l approximot ian

3.-Source strength for biconvex airfoil.

T,2n+Cti+

. . .

(14)

SECOND*RDER SUBSONIC JWRFOILTRDORY Here the T. and Cw are coefficients for the thickness and camber, respectively (fig. 1). Thus TO determines the nose radius, 01 the initial camber, and so on. Thin-airfoil theory bredcs down close to the edge where s is of the order of the leading-edge radius P. Now P is proportional to the square of the thickness ratio r, so that thinnirfoil theory fails where s= O(#). In this small region the airfoil is described to second order by the first two terms of equotion (14). In terms of the leading-edge radius p and initial slopo A of the camber line, the airfoil is given by y= kcfl+b This is the equation of an inclined rotated coor&ates by ~=+= parabola, (15a) described in

545

md, as it must be, this is the second-order thin-airfoil solution for a parabola. It is clear from this formal expantion how the singuhui~ arises at the leading edge. The ratio of the exact speed on the parabola to its formal 3erie9 expansion serves as a multiplicative correction factor for other shapes having the same nose radius. Thus the second-order thin-airfoil solution Q for any airfoil of leadingedge radius p is. converted into am appro.simation ~ that is nniforrnily valid near the e~~e by

(w%)(l%),* ~= n
, 1 **-&

(18)

(15b)

where the two origins of coordinates are separated by a negligible distance, and the difference between p and j can also be ignored.

Simplifying this insofar as possible without destroying its validity near the edge, and retaining only second-order terms, gives the rule

W+TJ(W

(19)

It might be supposed that since the airfoil nose was fitted to second order, this rule yields a solution that is uniformly valid to second order. Comparison with various exact solutions indicates that this is true, in the sense that the velocity disturbance and hence the pressure coefficient are correc~ to second order everywhere (&cept near the trailing edge, where additional modification is required). This will be indicated by replacing 7j by& The oblique abscissa ~ am be expressed in terms of the original abscissas, since on the surface of the parabola (20) Hence the rule becomes finally (21) In tho special case X=O this reduces to Lighthills rule (ref. 5). The corresponding rule for surface pressure coefficient is found by proceeding in the same way with the exact pressure on the parabola, which is found from equation (16) using Bernoullis equation. Again expanding formally for smal p/s and taking the ratio as a multiplicative correction factor yields the rule

FIGURE

4.Flow p.wt inolined parabola.

The exact velocity on this parabola in a uniform stream is found, from conformd mapping or otherwise, to be given by (16) where the signs refer to the upper amd lower surfaces. a has the physical interpretation that the stagnation Here point

(22)
Airfoils of the NACA four- and five-digit series are not T(z) analytic at the nose. Their thiclmess distribution consists initially of the ordinates of a parabola minus those of a wedge, so that the airfoil is descrl%ed by

lies at s=(z amd v=(z6 (fig. 4). ~~ the flow Pmt the parabola is related to that near an airfoil nose, a is some moderate mtitiple of the single of attack measured from the ~iderd angl~the angle at which the stagnation point coincides with the vertex The factor of proportionality ~epends on the entire airfoil shape (and the trailing-edge condition), but its value is not required here. Expading this expression for small p/3 yields, to second order, the formal series (17)

These airfoils are fitted only to* order by an inclined parabola, and it follows that the preceding rules render the thinairfoil solution uniformly valid only to tit order near the

 -.

546

REPORT 1274NATIONAL ADVISORY COMMITTEE FOR AERONAUTICS The second+rder rules can be manipulated into tho form of Riegels rule (ref .20), but the slight advantages of accuracy and simplicity then scarcely offset their defects. Tho same is true of the rules for compressible flow given later. Hmce Riegels rules are recommended only for tit-order incompressible-flow theory. Round edges in subsonic flow.-The previous rules con be extended to subsonic speeds simply by considering subsonic rather than incompressible flow past an inclined parabola. Although no exact solution of this problom is known, existing approximate solutions by the JnnzonRayleigh method are probably sufficiently accurate, at lmst at the ideaI angle of attack, and could, in principle be refined one could use experimentrtl indefinitely. Alternatively, measurements on a parabola. The problem is defined by p and a and the free-stream (The adiabatic exponent T nlso rmteml Mach number M. but is assumed fixed at 7/5.) Therefore, dimensional reaaoning shows that the surface speed is given by

edge, though it remains correct to second order elsewhere. h appropriate second-order rule would require finding the conformal mapping for a parabola minus a w-edge. First-order theory; Riegels rules.-k the first-order theory, the terms in A can be neglected, so that the rule for presmre, equation (22), simpliihs -to

(24)
The last term in equation (21) ~ also of second order, so that the rule for speed becomes 11:

()8
~2
8 )5

B gl

(25)

This is not precisely Riegels rule.

However, for a parabola

()

8+ P/2

=Cos q Hence

(26)
an alternative

where ~ is the angle of the surface. form of equation (25) is

(29) (27)
where, as usual, the upper and lower signs apply to tk e upper and lower surfaces. (Choosing p/2 rather than p as the reference length leads to later simplification.) Expanding this function formally for z large compared with p and az would yield the thin-airfoil series, which is, to second order

and this is Riegels rule (ref. 2). preswre is

The corresponding

rule for (28)

U,l= (Cos T)*C%

These alternative forms sufFer the defect of falsely implying that the nose exerts an influence even at remote points if the local airfoil slope is appreciable (for example, at the trailing edge). To this extent they fail to render the solution uniformly valid. However, it happens that they are exact for ellipses as well as parabolas, and are accordingly much more accurate for most airfoils, as indicated by the example of figure 5. 1.3 ; t I , .. 1.2

(30)
where

K1and

(12c).

KS are the compressibility factors of equation Again the ratio of Q to its series expansion serves as

a multiplicative correction factor to be applied to any roundnosed airfoil. Simplifying as before, and replacing 3 by s according to equation (2o), gives the rule 2

G ..-

~ u Here a where ql/ U is the first-order thin-airfoil solution. must be identified as the coeiiicient of + K1/fi in the secondorder solution, as is clear from equation (3o). (The physical interpretation of a shown in fig. 4 is valid only at zero MaclI number.) The corresponding rule for pressure is found by considering the exact pressure coefficient for the parabola, which must have the form

Lo
. ----Exact ~ml (ref, 37) first-onjer theory

.96

First order, Light hills rule 0 First order, Riegels rule I I ,4 .6 .2 Froction of chord

(32)

.8

Lo

Fmwm 5.Comparison of Lightbills and Riegds ndss for incompressible flow past ATACA0012 airfoil. .

~As M+ this rein~ not to the rule gtvenprevlotrdyfor brcornprcdbIoflow, but tom aSt8rnstive thet b qulvslont up to terms of mwrrdorder. The dtfferenca srlscs from tbo faot thet Snthe lnmmprcsdhlemsa the dependenceof g/U upon a SeglvonoxpUoltlyby tbo factor (How) tn equetkm (16), whloh k nsed to cancel a mrmqwndtng twrn In tho dormmlnstorof eqrrstlon(18).

SECOND-ORDER SUBSONIC AIRFOJII THEORY Its series expansion would give the thin-airfoil result
1.2

547

() g c=-;+=~
Hence the rule for any airfoil is

(33)
1.0

I P3r!

-order

+hin!oirfoil

(34a)

b
k \

1
I

I ~1 st -order thin-oirfoil . . . . . L. .,.... . . . . . . . . . -----------

,-

. ~ 2nd-o rder thin-oirf oil \// 0 r.. I, Jonren-Royletgh: :$-%~p-ib

(34b)

In this case a is the coefficient of + 2KJ@ in the secondorder pressure coefficient. Inmi has recently calculated the Janzen-Rayleigh solution for a parabola at the ideal angle of attack including terms in Mi (ref. 31). Thus, his rem.dts give

Q ~$J 0,~~)=go+M2g*+M4gi+o(Me)

(35)

where the g. are increasingly complicated functions of 3/(P/2). In reference 20 an attempt was made to increme the rwcuracy of this approximation by modifying it so that for large ~/(p/2) it tends exactly to the secondader thinairfoil solution of equation (3o). However, the third-order thin-airfoil solution has since been calculated by Kaplan (ref. 32): Q=l o

I
vP
approximations for velooit.y on parabola

)
at i~~= 0.6.

~IXJRE 6.Various

-ww=[(iY+

The corresponding table of the function appearing in equation (34) is:

P[g/(p/2),

O, Ml

Ww%)l+%a
where

(5JM)
W.&9
\
M

10
0.Uo91

(It is a matter of taste whether the logarithmic term included here is regarded as being of third or fourth order.) The comptin with this result shown in figure 6 suggests that the modification was detrimental, and that Imais solution is adequate for practical purposee.s A short table of the function Q[t7/(p/2), O, M calculated tiom equation (35) is given below: Q

o .4 .6 :: .8 .8s .9

Ilflsw .9015 .SYn .m5 .76s7 .65ia .W4 .s427 .!2256 :% .11622 .Im5 .ema . 795s

I
m
am-l

.9402 .Sw .M173 .0517 .4166 .2774 . 14s7

($

O,M

For other angles of attack, the function Q to order M can be extracted by a limiting process from Kaplans JanzenRayleigh solution (ref. 33) for an inclined ellipse. This gives, with 3/(p/2] =X, a/~=A

w (lA*)@-A(X+&)+ Q(x,A,w=@+A =x 2(1+-W{

(1-x+2A@

tan-1~1]

(36)

--.

..

.. .

548
The corresponding

REPORT 1274NATIoNAL ADVISORY COHTTEE approximation for the pree.sure coefficient is

FOR AERONAUTICS

II(X,A,M)=

I 2A3~-A2
1i-x

lvl -

(1+X)

(l+A~)(~~+A)
1+x K

@+~fl~A)

]n
%x+

(370)

(1X+2A@)

tan--@

IJ

and for the function P, according to equation (34b)

P(X,A$J)=
._%l

lI(x#$M) 1A X &K~ x () ,R

(37b)

though this relation is not uniformly valid near s= O. Hence, the thin-airfoil series for flow in an angle is, to second order.
3=C

l+!

hl 8+5

The error involved in retaining only terms of order M2 c-an be estimated from figure 6 in the special case a=O. At other angles of attack the error maybe greater; in particular, neither the Janzen-Rayleigh expansion nor the thin-airfoil espansion is believed to converge if the local Mach number exceeds unity. Sharp edges.The corrections for sharp edges can be found by considering flow in an angle. At a trailing edge with Kutta condition enforced, or a leading edge at ideal angle of attack, the surface and the dividing streamline meet at slightly less than a straight angle, as indicated on the

hl 8+:h128

(40)
)1

and it is clear how the spurions logarithmic singularities arise. Comparing these eqressions gives a rule that rendws tho seeondarder thin-airfoil solution for any sharp-nosed profile uniformly valid:

ij=8&a ~: [

~ hl 8$

1118:h128

)1

(41)

For a leading edge not at the ideal angle of attack, the flow includes a circulatory component, as indhmtecl at the left of figure 7. For incompressible flow, conformal mopping gives the surface speed associated with this component as

(42)
and its thin-airfoil expansion is \

--

(43)
Comparing component: / \ fiQURE

[q \ / / /

these yields a rule for correcting

the circulatory

?282(.:0

&&&I18
(

(44)

7.FIow near sharp edges.

right of figure 7. For incomprewible flow, the surface speed is found from conformal mapping to be given by

(38)
where 6 is the seroivertm angle, ands is again measured into the edge. In fig the constant c, the diilicnlty that in the angle flow the velocity increases indefinitely upstream can be circumvent ed by requiring that at any point the velocity must approach that of the free stream as the angle 6 tends to zero. Thus it is seen that c is unity except for terms of order & The connection with thin-airfoil theory follows from the - fact that for small e
s=l+e~ 8+; & 10<8+. . . (39)

The second-order thin-airfoil solution can be treated by splitting off the terms that are singular at least as 8-lfl near the leading edge, applying equation (44) to this circulatory component, applying equation (41) to the remainder, and recombining. These rules could be extended to subsonic speeds, in tlm case of ideal angle of attack, by calculating the JanzenRayleigh solution for flow in an angle. However, at othor angles the Janzen-Rayleigh approximation certainly fails, beeause it would predict infinite speeds that am tolmablo only in an incompressible fluid. In any case, the correction is negligible for most practical purposea, because it is significant in only a minute neighborhood of the edge, and, furthermore, sharp edges are usually trailing edges, in which ease the details of the flow are masked by viscous dfects. For these reasons, no correction for sharp edges is included in the computing scheme given later. Combined edges.Airfoils with two sta=mntion edges are treated by applying the appropriate corrections in turn at

SECOND-ORDER SUBSONIC AIRFOIL THFIORY each edge by shifting the with round leading and located at z=% and ~=~,, camber angles L and ~@ origin. Thus, considar trailing edges of radii respectively, and initial Assume that the Kutta an airfoil ~ and p~ tmd final condition

549

is imposed at the trailing edge. Applying equation (31) twice, identifying s successively with ZG and x*z, and then simplifying to keep no more than second-order terms giwa

(45)
(The simpler form of this equation for incompressible flow is given as equation (24) of reference 20.) The corresponding rule for pressure coefficient is valid second approximation by applying equation (21) twice in succession, or using the combined rule of equation (24) of reference 20, which gives

(46)
Similar rules cm be found for combinations of a round and a sharp edge, or two sharp edges. For example, for incompressible flow past two sharp edges of equal angle, both with Kutta condition imposed (as for a symmetrical biconwm airfoil at zero angle of attack) and located at cc= +1, the combined rule has the form of equation (41) with s replaced by (1 d). EXAMPLES : COMPARISON WITH EXPERIMENT AND OTHER THEORIES INCOMHWSSIBLE FLOW It has been seen that the solution for subsonic flow depends on that for incompressible flow. It is therefore pertinent to test the second~rder theory in the case of incompressible flow, where it can be checked against the exact results of conformal mapping. Ellipse,-Consider an ellipse of thickness ratio ~ with the interval 1 <x< 1 ns chord line. It is demxibed by y=*
TJ=7,

l.:

o - ---

I<xsl

(47)

Exoct Modified second-order theory Formol second-order theory Formol first-order theory

Suppose that the Kutta condition is satiaiied-the rear stagnation point coincides with the end of the major axis. Then the fit-order solution for surface speed is found, from equations (3), (4), and (5), together with Appendix B, to be . (48a) Proceeding with equations (6) to (9) gives tbe formal secondorder result
, 1

.L4

?G

..

i)

Froction of chord

This can be checked by expanding the axact result, which is ;=(1+T) J=& 00s %t(lz) ~~z sin a (49)

FIGURE 8.-Speed

on 18-percent-thick

ellipse at zero

angle of attaok in

incompressibleflow. These approximations are compared in figure 8 with the exact solution for an 18-percent-thick ellipse (which has nearly the same nose radius as an ATACA 0012 airfoil) at zero angle of attack. The precipitate descent of the formal

The formal second-order the ends of the ellipse.

solution clearly breaks down near It is converted into a uniformly

-.

550

MMIITEE FOR AERONAUTICS REPORT 1274NATIONAL ADVISORY CO g=

second-order solution toward negative iniinity is just discernible near the nose and is eliminated by the edge correction. Symmetrical Joukowski airfoil.-To second as well as tit order a symmetrical Joukowdri airfoil of thickness ratio r is described by y=+ T,(lz)E, I<z<l

(=): -{
23@ T T
=(

1+~T[2(l_h)

(l+~)(l~)h(l-~)]+

l&)-3(l+z)ln(l

+z)-3(1-x)ln(l-z)

-!-

-()[

* (l+3z)(l+z)ln(

l+z)~

(13z)(lz)ln2

(1x)+

4 =3.@

T= O.7698T

By the foregoing procedure, is found to be Q=l+T1(l2Z)+

the formal seumd+rder

(50)

;(lzw(l+z)ln

(1z)

l-i}&l+z)*{l(63b) }

solution

~[2(l+z)ln(l+z)

-(l+2z)ln(l-z)-4]

lz . d l+Z

1 * lz (1+2X)+ 2 l+Z (51a) \

\ \ \ -
0

where the fit three terms give the first-order solution. Correcting this by means of equation (21) with 8=1 +Z and p=4r1s (and A=O) gives the uniformly valid second nppro.ximation

Exact (ref. 34) Second-order theory, formol or modified Formol first -arder theory

@-2,zhEw

(51b)

In figure 9 these approximations are compared with the emict solution (ref. 34) for a 12-percent-thick seotion at The effect of the edge correction on zero angle of attack. the second-order result is not discernible to this scale.
Biconvex &foil.-To second order a symmetrical blconvex airfoil of thickness ratio r bounded by either circular or parabolic arcs is described by ?J= + T(l &), The formal second+rder 1<2<1

(52)

solution is found to be

(53a) In deducing from this a uniformly valid second approximation, the terms independent of a are treated by the rule for combined equal shaxp edges that was described just after equation (46), with ~=2T. The terms in a are modified according to equation (44) with s= 1+z. The result is

-1
Froction;f chord

FIGURE 9.4peed on 12-percen&thick symmetrical at zero angle of attack in incompressible

Joukowski flow.

airfoil

-1

SECOND4RD13R SUBSONIC AIRFOIZ TEU30RY

I

551

With the aid of Appendix to be

B the first-order solution is found

(3b,+2&)z4b.# I
/ /

1-+=
+

(55n)

/ / / /

in agreement with the result given by Goldstein (ref. 28). Applying Riegels rule (eq. (27)) renders this a uniformly valid first approximation except very near the trailing edge. The second-order terms in thickness, in addition to being very complicated, involve integrals that apparently camot be evaluated in terms of tabulated functions. Accordingly, the second-order terms have been calculated using the Riegels-Germain numerical method discussed in Appendix C, with IV= 16. The accuracy of this approximation is nswred by the fact that cruder approximations modify the numerhxd results only slightly, as will be seen in a later example. The formal second-order solution for surface speed therefore has the form g=l+TQr+ lx --+r2Qm-&TaQ,a-5 1$ (55b)

<

Exoct Second-order theory, formal or modified Formal first-order theor b

where values of Q, from equation numerical values of Q~ and Q,= are

(55a) and approximate

.&d

9G of chord

Froction

itii~

*F

FICNTBE10.-Speed

on lS-percent thick biconvex airfoil at zero angle of attack in incompmsible flow.

Them approximations are compared in figure 10 with the erect solution (ref. 35) for a circular-arc airfoil 18 percent thick at zero angle of attack. Although the vertex angles are large in this example, the edge correction is appreciable in such a small neighborhood of the edge that it would be invisible even on a much larger plot. NACA OOXX airfoils.-S~etrical airfoils of the NACA 00XX family (such as the NACA 0012) are naturally defined for the interval O<z <1. The airfoil of thickness ratio r is described by (ref. 36) y=* where b,= 1.48450 Z@)= +@lfi+bM+w+w+w), o<~sl

(54)

Applying equation (21) with P=1.10187 P, 6=1.16925 ~ (and ~=0) yields a uniformly valid approximation. However, as discussed previously, the curvature of the proiile does not vary continuously near its nose, so the result is only a first approximation there, though a second approximation elsewhere. The various approximations are compared in figure 11 with the result of a long and elaborate calculation by conformal mapping for the NACA 0012 airfoil that is given by Goldstein (ref. 37). Again the effect of modifying the second-order solution is indiscernible. Also shown is the exact solution tabulated in reference 38. The agreement between the fit-order solution with Riegels rule, the second-order solution, and Goldsteins calculation leaves little doubt that his is the more accurate of the two exact soluti ens. COMPRR9SIBLE FLOW When extended to subsonic compressible flow, the preceding examplea can all be compared with other theories or with

b,= O. 63000 b,= 1.75800 bo= 1.42150 bs=O. 50750

..___.

552
I..

REPORT 1274NAIIONa

MMITIEE FOR AERONAUTICS ADVISORY CO in agreement with the result of Hantzsche and Wendt (ref. 7) The corresponding third-order result has been given by Hantzsche (ref. 8); his expression can be simplified to

.. o
A

Exact (ref. 37) Exact (ref. 38) Second-order theory First-order theory, Riegels Formol first-order theory

rule

1. where (58b) The terms in r% T are found correct together with those in 73in the second iteration; it is a matter of taste whether they are regarded as being of third or fourth order. The solution to # has subsequently been given by Hasimoto (ref. 39), and although typographical errors unfortunately appear in his equations, his numerical results agree with those calculated from equation (58a). Values of the maximum speed ratio calculated from thcso and other approximations for a 10-percent-thick dlipm at zero angle of attack (whose critical lMach number is about 0.80) are:
I I

al &

I df-O.70 I MOO.75 I M-O.KI I

FfmtaderthemY (or I%ndti-olauertmle epplld to=tfnwmprmlble valaeofl.l)-..= . . ...=. ~-Ttim tie---.2--------., ................. -ndatiwfiwm ------------------------------Third-orderthex% 1.140 1<149 1.14.9 1.151 1.lm 1.10s 1.107 1, 1s4 1.1s6

hdtigra--:-:----. . . . . . . . . . . . . . . . . . . . . . . . . . . 1.lml 1.ml 1.107 1.149 htitigfib---------------------------------1.104 1.lea

u Fractit

of chard

FIQUaFJ il.-Speed

on NACA 0012 airfoil at zero angle of attack in incompressibleflow.

experiment. As before, the comparisons will, for simplicity, be made only for zero angle of attack. Ellipse, -Applying the second+rder compressibility rule of equations (13) to the incompressible solution of equation (48b) gives as the formal second-order solution for the speed on an elliptic cylinder

(56)
For zero angle of attack the maximum midchord, is given by speed, occurring at

gl u=

=I+KIT+;

(Krl)T

(57)

(Here the K&rrn6n-Tsien rule has been applied to the pressure coefficient for incompressible flow, and the speed ratio then calculated from Bernoullis equation.) It might have been anticipated that the second-order theory is more accurat o than any of the compressibility correction formulas such as the K&rm4n-Tsien rule, because it allows for a dependonco on the particular airfoil shape and on the value of ~. However, in this example the results of second-order theory and the Kfwmhn-Tsien rule are practically identical. In the same way the second-order solutions ore readily calculated for the Joukowski and the biconvex airfoils, and are found to agree with the results that Hantzsche rmd Wendt obtained by laborious analysis. NACA 0012 airfoil.-lle formal first- and second-ordm solutions for NACA 002CX airfoils in subsonic flow cm easily obtained from equations (13) and (55). The socondorder solution can then be rendered uniformly valid nmr tho nose using equation (31), although again the modification is sign.iiicamt in only a very small region of the nose. For the NACA 0012 airfoil at zero angle of attack, llmmons has calculated the flow field at Mach numb em of O, 0.70, and 0.75 using the numerical relaxation method (ref. 40). The last of these Mach numbers is supercritical, so that the flow contains shock waves, and is beyond the scopo The pressure distribution calculated of the present theory.

.(,

<

SIJC&-ORDER

SUDSONICACRFOIL THEORY -1.

Formal t%st-ander theory - Formal second-order theory -- Prandtl-Glauert rule K6rm(n-Tsien rule Experiment (ref. 41) 0 Upper surface K Lowersurface

553

b;pi$, f- ,..,yq~.,\
0/
-d

~,

L
f

~\>\
K. .

M=O
Cp Cp

.. . --. -. --

First-order .Second-order Relaxation

theory theory method rule (ref. 40)

Pmndtl-Glauert K6rmdn-Tsien Exact,

rule

inmmpressible

Fmctian

of chord

FIQI.WI 12.Prt?fWUre Coefficient On NACA 0012 8kfOil at zWU W@e attaok in subsonio flow.

Of

by the relaxation method for 434=0.70 is compared in figure 12 with the results of first- and second-mder theory and various other approximations. The relnation solution for incompressible flow is also shown in comparison with Goldsteins react solution, and is seen to be inaccurate near the nose. The solution for 34=0.70 probably contains similar inaccuracies; however, just as for the ellipse the pressure coefficients calculated by second-order theory maybe slightly lCESnegative than the true values near the minimum. Experiments on NACA 0015 airfoil.-Experimental pressure distributions in two-dimomiomd flow over the NACA 0015 airfoil at high subsonic speeds are reported in reference 41, For zero angle of attack, the critical hfach number is The mwwurements at this Mach approximately 0.70. number me compared in figure 13 with the results of firstand second-order theory and of the two common compressibility correction formulas applied to the incompressible Unfortunately, flow values tabulated in reference 38. the model was imperfectly constructed, and the ordinates were inaccurate nem the nose and midchord. Otherwise, the measured pressures are in satisfactory accord with either secondarder theory or the results of the Kfmnhn-Tsien rule. !I!omotika and Tamadas ai.rfoil,-Using the hodograph method, Tomotika and Tamada have calculated the flom past n certain family of symmetric airfoils (ref. 42). As usual in hodograph solutions, the airfoil shape varies somewhat with free-stream Mach number. The critical Mach

)
Fmctionof chord

FIGURE 13.Comparieon dietributionrs on NACA attack.

of thearetioal 0015 airfoil

and experimental at .ii=o.70, zero

pressure angle of

number is 0.717, and the corresponding shape is shown in figure 14 together with the surface speeds predicted by various theories. For mathematical simplicity, Tomotika and Tamada have adopted a hypothetical gas, which is fitted at Mach numbers zero and unity to a polytropic gas having ~= 7/5. At any intermediate Mach number, however, the hypothetical gas corresponds to a polytropic gas whose Y is greater than 7/5, reaching a maximum value of 1.91 at i14=0.78. To seeond order, any such hypothetical gas is equivalent to a polytropic gas having the veke of -r corresponding to the free-stream flOTV,gh%ll by ~=1+ $% () -_

where p is the density and c the speed of sound in the hypothetical gas. For Tomotika and Tamadas gas with -ii= 0.717, that value is 1.82. Actually, the second-order solution depends so alightly upon the value of -r that the change from y=7/5 to Y= 1.82 increases the maximum value of gJUby only two parts in a thousand. However, the nonpolytropic

.~

___

554

REPORT 1274NATIONAL ADVISORY COMMITTEE FOR AERONAIJIMCS foil at any angle of attack. It requires a knowledge only of the airfoil ordinates at 7 (or 15) points along the chord. It is based on the foregoing theory together with the nunmrical method of Riegels and Germain that is discussed in Appendk C. It includes the correction for a round leading edge, but not for sharp edges. COMPUTfNG PROCEDURE WITH~.8 (OR1S) (1) Tabulate the ordinates Y and I, of the upper and lower surfaces at the 7 (15) pivotal points Z. listed in table I. (The z axis must pass thro~mh the leading and trailing edges with the leading edge at z=O and trailing edge at z= 1.) (2) Calculate the corresponding valuea of

~
Hodograph

method, Tomotib

and Tornado

(ref. 42)

1.4

1.2

7+-.-YJ,
Lo

C+-.+YJ

(59)

(3) Using the influence coefficients of tibles II, III, and IV, calculate

u~~ 7 15) . h CUT, u 8-1

%?

U,C* 716 = 3 dUC, u 8-1 c=7g f. c, (60)

T=7~ e.T, 8-1 TN=7~ gmTs e-l

FIGURE 14.--6peed

on Tomotik*Tamada

airfoil

at M=

O.717.

IY=7~ hMC8 8-1 1

nature of the hypothetical gas must be considered in converting Tomotika and Tamadas values of surface speed (which are referred to the critical speed) to the form gjU. At the free-stream Mach number of 0.717, the ratio of freestremn to critical speeds is 0.769 for the hypothetical gas, compared with 0.748 for a polytropic gas having -y=7/5 (and 0.774 for a polytropic gas having y= 1.82)S Thus, as Tomotika and Tamada emphasize, their resndta should be regarded as exact for the hypothetical gas; and the present method should be r&ed as giving the true second approximation for the same gas (which to second order is equivalent to a polytropic gas with ~= 1.82). The agreement betwean the hodograph method and second-order theory is satisfactory. Third- and higher-order terms are seen to increase the surface speed near its mtium and reduce it slightly elsewhere, which is the effect of third-order terms calculated by Asaka for a biconvex airfoil (ref, 43). As shown in figure 14, the results of second-order theory and the Kfirmfi-Tsien rule nearly coincide except over the forward portion, where the second~rder theory is apparently more accurate. The second~rder solution calculated with the present computing scheme is in close agreement with an unpublished second-order solution carried out by Naruse. PItACIICAL NUMERICAL COMPUTATION The following computing procedure yields the second-order subsonic solution for the surface speed or pressure on any airh fndebbl b Mr. E. Naruw af Tokyo Metro@tan UnlvemftYfw hafig ~The natfm Pointedaut theseId% same of whfch stem fa tarn from W mllmgms, L DIM and H. Tn!u3mL

(4)Using table I (with a in radians), CaIcuhte

?qC u~c* . ~u+ lx x (61]

<

(5)Calculate
T,=@

T+

C,

~z=~

c+%

(62)

(6)Using the influence coefficients of table II, calculate

%,71 E cWTti-; T 8.1

d,

lLZc 7 6 = E dR,C>

8-1

(63)

(7)Thing the compressibility

factors of table V, calculate

2(C+ T)(C+T)+(+T +)++

GwYl

Mb)

The & signs refer to the upper and lower surfacea of the airfoil.

SECOND-ORDER SUBSONIC AIRFOIII THEORY Oorreotion for round leading edge.If the airfoil has a round leading edge, continue as follows: (8) Using the influence coefficients of table I, calculate (65)

555

(66)

(68) =4% (10) At each pivotal point on the upper and lower surfaces calculate ~n=x.%h~ p/2 and then calculate (69)

4. Seven pivotal points yield sufficient accuracy for most purposes. If conditions near the nose of a thin airfoil are of interest it may be necessary to repent the computation using 15 pivotal points. 5. The above scheme is designed for calculating a single case. If the same airfoil is to be calculated at more than two anglw of attack, it is economkd to subdivide the computing scheme to separate terms in a and ~. Similarly, the scheme should be subdivided if more than two thickn= or camber ratios are to be calculated for the same family of airfoils. 6. For NACA airfoils T is the basic thickness and C the camber line. To second order it is immaterial that the tb.ickneas is added normal to the camber line rather than to the chord line. 7. k step (10), if A=O the values of Q or P can be taken horn the tables following equation (35).
EXAMPLE

Q(x,,j4, M)

or

P(X.,&4,

M)

(70)

The following table gives the complete computing form for calculating the fitand second+rder increments in surface speed for an NACA 00XX airfoil (of unit thickness ratio) at zero angle of attack and zero Mach number:
n z

from equation (36) or (37) (11) Using the compressibility at each point

factors of table V, calculata

w 43 77-7

1 ao3a@3 2 .14045 3 .mE9 4 bawl . lm34 5

6

C123S16 L8Q30 .44Za L 0168 .4WW L!NM .44W1 .M

.XS43

[ (71a) or

. 161W .W .MuW .044w t

. m24 . W51

.4939

54. m CL#i& -8.923 3. 638 . ff1517 L ~ .3W79 .16142 L 0$8 L 0163 L 674 .01336 LW L m . m
x crJ14 .88.52 . 0225 . 5446 -. We3

zm=Pc,2
REMARKS

(71b)

The accuracy of this solution with only 7 pivotal points is indicated by comparison with the following valu~, which were obtained analytically for AgllU and with 15 pivotal points for Agzl~

H
1 ,
_: g I . 4239

T,

1. The summations of steps (3), (6), and (8) are ccmveniently carried out by tabulating C and Tin columns that can bo matched with successive columns of tables II and III while cumulative multiplication is carried out on a desk calculating machine. 2. If the airfoil slope and second derivative can be found more directly, step (3) can be simplified by omitting the calculation of T, C, T, and C. 3. If the ledng-edge radius P and initial camber angle A me known, omit their calculation from step (8).

&14046 .bXm .&Ws

1.615!3 .Mm .072s

It is seen that the solution using 7 pivotal points yields ample accuracy. AMES AERONAUTICAL LABORATORY ADVISORY COMMITTEEFOR AERONAUTICS NATIONAL MOFFETT FIELD, CALIF., March 1%, 1966

556

REPORT 1274NATIONAL ADVISORY CO MMTPIEE FOR AERONAUTICS

APPE.NDIX

PRINCIPAL SYMBOLS

A a b. c(x) c,(x)
c, c,, AC*

a/@I factor proportional to angle of attack measured from ideal angle coefficient of xfl in series for airfoil ordinate camber of airfoil camber of fictitious airfoil in second-order solution surface-pressure coefficient iirstmrder surface-pressure coefficient seccmd~rder increment in surface-pressure 00eficient influence coefficients for calculating velocity, slope, and second derivative of airfoil ordinate influence coefficient for oalcukting a analytic function of complex variable j.., intluence coefficient for calculating initial camber angle imaginary part off(z) on unit cirole first-order eompressibfity secondarder faotor, ~ factor,

u
u

v x x Y(z) Y., Y I/

coefficient in trigonometric polynomial approximation to T free-stream speed velocity perturbation parallel to chord lino velocity perturbation normal to chord lino abscissa

8/(p/2)
ordinate of airfoil ordinates of upper and lower surfaces airfoil, respectively ordinate complex variable angle of attack of

z ;
P,
7

c-l d,mje-, LS,gn$, L, }

:/2) f. I K, K,

-fP

6
e :

compressibility

JJp II P 7 TI

(7+l)ikf+4fY 4P k, M N P
coefficient in trigonometric polynomial approximation to C free-stream Mach number number of subdivisions of chord line in numeriwd integration ratio of pressure coefficient on parabola in subsonic flow to second-order thin-airfoil value surface speed ratio on parabola in subsonic flow (See eq. (55b).) flow speed on surface of airfoil fit- and secondader increments in g real part of}(z) on unit cirole influence coefficient for oahndating edge radius abscissae. measured from edge positive into airfoil distanoe from round edge measured along initial tangent to oamberline tbickne9s of airfoil thickness of fictitious airfoil in second-order solution

; ( )= ( )b (). ( )M ( )rn ( )x ( )Xl ( )s ( ( ( ( [; )t )0 )1 );

Q,> Qm> Qm L? I% Ms

R
r
8

T(x)
T,(z)

m coefficient in numerical calculation of u adiabatic exponent of gas coefficient in numerical calculation of Y semivertex angle of sharp edge polar angle angle of airfoil surface to chord line terminal angle of camber line to chord lino coefficient in numerical calculation of Y pressure coefficient on parabola in subsonic flow radius of round edge airfoil thicknw ratio 4 T for Joukowski airfoil 3& perturbation velocity potential perturbation strwun function value at leading edge value at trailing edge component associated with camber and angle of attack value at Mach number ~ value at mth pivotal point counted from trding edge value at n.th pivotal point countod from leading edge index of summation, counted from trailing edge index of summation, counted from loading edge component associated with thickmms value at zero Maoh number that-order &pproximation second-order approximation derivative uniformly vd.id approximation

SECOND-ORDER SUBSONIC AIRFOIL THEORY APPENDIX B

AIRFOIL INTEGRALS The following me the Crmchy principal values ford< 1: l+L

6.

f -1 ~

1 (zf)

@=()

558

RDPORT 1274NATIONAL ADVISORY CO&f3H!ITEEFOR AERONAUTICS APPENDIX

c
ShlC13

THE RIEGELS-GERMA.IN METHOD The numerical procedure introduced by Riegels (ref. 2) and Germain (ref. 21) can be adapted to give approximately the thin-airfoil velocities on any profile in terms of its ordinates at certain tied points. Thwaites has applied Germains procedure to thin-pirfoil theory (ref. 23), and Weber has systematized the calculation of the slope and surface velocity (refs. 24 and 25). Here we must also find the second derivative. In applying the edge correction at a round nose we also require the edge radius. It is convenient to derive all these results from Watsons analysis (ref. 22). Let j(z) be regular within the unit circle, and on the unit circle have the form j(e~ =R(0) +fl(0) (cl) Now /&_,=
2N-1 X I%y.+p= p-o

19,and Y,= Yx=O,

N1 ~lBP-myP+ N1 ~ll%+.yw-. ((26)

Symmetrical airfoila,-1or Y,= T,. Then according (C6)

a symmetric airfoil Yw-,= to equations (C4), (C5), and

~ )

N-1
cmPgin

u. =P~cmpT,,

(BP-m-l% +.)

(C7)

This form is convenient for calculating the cm,. easily shown, using trigonometric identities, that

It is also

(Our R and I are Watsons # and c.) Then following Germain, Watson approximates to R by the trigonometric polynomial N1

Lv a. I
% { [ Q

pi-m=O p*m=even, not O

I
} (C!8)
J

R(o) .=k~+~(k,
r-1

cos

Te+G

SiII @

+k~

cos

Ire

(C2) iV(COS

4 sin

0, e,y km=odd

ILCOS

which can be made to coincide with R at the 2iV equally spaced pivotal points o=o.=m~/iV. Thus he derives appro.simate forndas for 1 (tiide from a constant), R, 1, in terms of the values of R at the pivotal JR, and fl points times fixed influenm coefficients. In thin-airfoil theory the cumplex perturbation potential P+it is regular outside the unit circle in the absence of circulation. Inversion shows that this involves a change in sign of either the real or imaginaxy part, since j (e-~ =RiI. Hence (w,#) or (#,P) maybe identied with (R,.1). theory the tangency condition on the perIn thin-airfoil turbation stream function # is ~= Y, where Y is the airfoil ordinate. Therefore, in order to obtain a solution in terms of the airfoil ordinates we identify (Y,p) with (R,l). It is assumed throughout that the z axis passea through the lending and trailing edges.
STREAMV?TSE VRLOCITT INCREME NT FOR a=O

which is Webers result (ref. 24). Corresponding forma can, if desired, be found for the other intluence coefficients discussed later. Antisymmetrio airfoils,--lor a cambered airfoil of zero thiclmess Yw-,= Yp= Cp. Equation (C6) gives

(C f))
This expression represents the velocity on the unit circle into which the airfoil is mapped. , The Kutta condition will be violated at the trailing edge of the airfoil unless the expression happens to vanish for m=O. Adding rLcomponent of circulatory flow changes the velocity on the circle by a Hence the Kutta condition is enforced by subconstant. tracting from the expression of equation (C9) its value at m=O, so that

Let z run from O at the leading edge of the airfoil to 1 at the trailing edge, and z=; (l+COS e) velocity 2 G ap To

(C3)
on the airfoil is

13ence, according

to equations (C4) and (C5) L*=& @,-m+13,+m-Z9P) (cl 1)

Then the streamwise perturbation given by u ?)(0 _ . appo u-ax W=

(CM)

SLOPE OFAIRFOIL The airfoil slope is given by

Now according to Watsons equations (10), (24), and (27), in the absenca of circulation the values of Z)@O at the points . em are given by

y=ig..z
Now according to Watsons

$V,
aP % ) ZN1 ~ = ,~o BFY.+P} P,=

smOd8

(C12)

p=o (31), and (34) equations (29),

p=even, ~p=odd not O i J

o,
1 N(lcose,)

(C5)

SECOND-ORDER SUBSONIC AIBFOm THEORY or, since -yw_P= YP

so that

559 (C23)

Symmetrio airfoi.ls.-Using gives for symmetric airfoils

NI

the symmetry

conditions

again

Setting m=~ and O in equation (C14) and using the symmetry properties of the thiclmess and camber distributions gives

T.= ~
Antisymmetic

hpTp, &p= -*
airfoils.+%nilarly,

(7,-.+7,+.) (C15)

for camber lines (C16)

~= 2 217N-PTP J_

N1

~ ~:

Y4P

(C24)

(Yp+m-Yp-J CL=3J.POPJ .fmp=& m

SECOND DEfZfVATIVE

at the leading and trailing edgea, respectiv,ply.

SLOPE OF CAMEER LINE AT END

The second derivative

of the airfoil ordinata is given by

Equation (C16) is indeterminate at the ends of the airfoiL Applying LHospitals rule to equation (C12) gives at O= O

Z=sie

O?Y

(PY

T=-wt % 7

(C17)

0=-2 w
N1 Ye= ~ f,%

OEY

(C25)
Yp= Cp

An approximation for @Y/d8z, which is required here, is found by extending Watsons analysis for the first derivative, as he suggests. l?ollowing closely his section 2.4 gives, after some computation,

Then using equation (C18) and the fact that YN-2= for a camber line gives

f,=jop=-4(-ob ~_:oa

(C26)

m=%pym+pwp=
ANGLE-OF-AITACK PARAMETER a

Symmetrio airfoils .By the foregoing found that for symmetric airfoils

procedure,

it is

The parameter a is the coeflkient of the square-root leading-edge singularity in the surface speed. There is a firstarder contribution from uJU (eq. (4)), and second+rder contributions from wJU (eq. (8)) and (C+ T) (C + T) + l/2(C+ T) (eq. (9)). For zero angle of attack, equations (C4), (C5), and (C1O) give

Y.=p$l g.p Yp,

/j [(1%-.-PP+m@t@t 9mp=ah:
m

WJP..+7P+*)I

(C19)

Near the leading edge

sin O-@ (C28)

Antisymmetrio

airfoils,-similary

for camber lines

G%=NJ ll?#,Yp,

lhp=&m K%-m+h+m)-cot
EDGE RADII

(C29)
Wp-m-y,+m)] (cm

I&Ming the contribution (4) gives for the fkstarder al=a .

of angle of attack from equation approximation to a dpCp dp=!2(/3N_p-/?p) (C30)

If the airfoil has round leading and trailing edges, their radii rtre given by p= lim 2XY9,

(@-N~~ )I

p= lim 2(1+) Y*
-1

(C21)

For a round nose described by equation (15a), the terms T) of equation (9) contribute. (c+ T) (C+ T)+l/2(C+

Near the edges, to a first approximation dx =Z () ~_z= @ ()do 2 (C22) Hence the second~rder increment in a is (C31)

I Tho oxprmskmglvm for % in NAOA TN 33W M Income& dMdng from that of the prmcnt equntbo (018) by lfZ(-1).N. TIMerrormncab out fn the o-* but the v81umof b., tabulotd them amaffwtd. Howover, tbo mrrmtbn hm Utile effed (oscept to siruPlfIY thotibl~)tim tbeappmxlmatlou merely bewm- exact for the flint 8 FmWmIUJI@ rothrx tbnn the first 7 h the mso N-8).

. ..

. ..

560

REPORT 1274NATIONAL ADVISORY CO~

TAB-

E FOR AERONAUTICS

The various influence coefficients have been calculated for iV=8 and N=16. The valuea have been checked by applying them to a number of simple shapes-ellipse, parabolic arc camber line, etc.for which the approximation of the airfoil by a trigonometric polynomial is exact. They were also checked against Webers values (refs. 24 and 25) where The vahms are believed tm be accurate to within possible. one unit in the laat place. For convenience of computation, the coefficients have helm renumbered so that the pivotal points are counted ilom the leading to the trailing edge. This renumbering is indicated by using indices (n, s) instead of (m, p). The above results are formally unchanged, except that the roles of leading and trailing edge are interchanged in equation (C24), the sign of equation (C26) is reversed, and the sign of dz is reversed (eq. (C30)). The renumbered influenee eoefficienta are given in tables I to IV. REFERENCES 1. Riegele, F., and V7itiloh,H.: Zur Berechmmg der Druckverteilung von Profilen. Jan. 24, 1942. Aerodynarrdsche Versuchsanstalt Aerodynamic. Gtittkgen E. V. Institut ffirtheoretische
Benoht 41/1/15. 2. Riegele, F.: Das UrostrOmungsproblem bei inkomp=iblen Poten~t~Mstr6mungen. Ing.-Archiv., Bd. XVI, 1948, pp. 373-376, and Bd. XVH, 1949, pp. 94-106. 3. Keune, F.: lMtriige zur Profilfomohung. VI. Zweite N5herung zur Berecbnung der Gesohwimiigkeitsverteilung naoh dem Singulsrit5tenverfahren. Luftfahrtforsohung, Bd. 20, Lfg. 6, June 1943, pp. 196-206. 4. Goldstein, Sydney: Low-drag and Suction Airfoils. Jour. Aero. Soi., vol. 15, no. 4, Apr. 194S, pp. 1S9-214 5. Llghthill, M. J.: A New Approach to Thin Aerofoil Theory. Aero

16. Harder, Keith C., and Klunker, E. B.: On a Sourca-Sfnk Method for the Solution of the Prandtl-Bus-emann Iteration Equations in Two-Dimensional Compressible Flow. NACA TN 2263, 1950. 17. Hayes, Wallaca D.: Second-Order Premuro Law for Two-Dimensional Gomprwwible Flow. Jour. Aero. Soi., vol. 22, no. 4, April 1955, pp. 28+286. (Originally publfshed as: SeoondOrder Two-Dimensional Flow Theory and Imais Similitude. Britiih A. R. C. 15,722, F. M. 1877, 1963.) 18. Imai, Isao: Extension of von K4rrmins Transonio Simjlmfty Rule. Jour. Phye. Sot. Japan, vol. 9, no. 1, Jan.-Feb. 1064, pp. 103-108. 19. Van Dyke, Milton D.: The t%cond-Order Compre.mibjlity Rulo for Airfoils. Jour. Acre. Sci., vol. 21, no. 9, Sept. 1964, pp. 647-648. 20. Van Dyke, Milton D.: Subsonic Edges in Thin-Wing and SlendorBody Theory. NACA TN 3343, 19S4. 21. Gormain, P.: Sur le calcul num6rique de certains opt$mteurs lim%res. Comptes Rendes Aoa. Soi. Park, vol. 220, 1945, pp. 765-768. 22. Watson, E. J.: Formulae for the Computation of tho Funotions Employed for Caloulatiig the Veloci~ Distribution about a Given Aerofoil. R. & M. No. 2176, British A. R. C., 1946. 23. Tlrwaites, B.: A Method of Aerofoil Design. Part I-Symmetrical Aerofoils. R. & M. No. 2166, Brftish A. R. C., 1046. 24. Webm, J.: The Calculation of the Prewsuro Distribution over the Surface of Two-Dimensional and Swept Wings wfth Symmotrfoal .Aerofoil Sections. R. L E. Rep. Aero. 2497, 1963. 25. Weber, J.: The Calculation of the Proemm Distribution on the

26. 27. 28.

Quart., vol. 3, pt. 3, Nov. 1951, pp. 193-210. 6. G&tler, H.: Gasstr6mnngenmit ~ergang von Untemchall- m Z. a. M. M., Bd. 20, Heft 5, ~ersohallgeschtidigkeiten. Oct. 1940, pp. 254262. 7. Hantmche, W., and Wendt, H.: Der Kompresaibilitfitseirkks fti dOrmewenig geWlmnte Profile bei Unterschallgeschwindigkeit. Z. a. hf. M., Bd. 22, Nr. 2, Apr. 1942, pp. 72-86. S. Hantzsche, W.: Die Prandtl-GlauertsoheNaherung ale Grundlage fiir ein iterationsverfahren zur Berechnnng kompremibler Untersohaflstr5mungen. Z. a. hl. M., Bd. 23, Heft 4, Aug. 1943, pp. 185-199. 9. Sohmieden,C., and Kavmlki, K. H.: Beitrage nun Umstr6mungsproblem bei hohen C%chwindigkeiten. Bericht S13, Teil 1, L. G. L., 1942, pp. 40-68. (Available in translation as h-ACA
Thl 1233, 1949.) 10. Kaplan, Carl: The l?low of a Compressible Fluid Past a Curved

29.

Surface of Thick Cambered Wigs and the Design of Wfngs with Given Pressure Distributions. British R. A. E. Rept. Aero. 2548, June, 1955. Munk, Max M.: Elements of the Wing %otlon Theory and of the iVingTheory. NACA fip. 191, 1924. Pistolesi, E.: Sulla Teorfa deLIeAli %ttili. Aota Pent. Acrid. Sci., VOL1, 1937,pp. 57-72. Goldstein, Sydney: Approximate Two-Dimensional Aerofoil Theory. Part I. Veloci@ Distributionsfor SymmetricalAerofolls. C. P. 6S, British A. R. C., 1~2. Jones, Robert T., and Cohen, Doria: Aerodynwnies of Wings at High Speeds. &ection A of vol. VII, Aerodynamic Components of Aircraft at High Speeds, High41peodAerodynamioe and Jot
Propulsion. McGmw-Hill, Princeton Univ. Prom, 1957. 1%54.

30.ErdQyi, A., ed.: Tables of Integral Tmnsforms, vol. 2, ch, 16.

of the Mz-Ex~ansion Method to tho Subsonio Flow of a Goxnpressible Fluid Past a Pambolio Cyllnder. Proc. 1st Japan Nat. Ccmg. Appl. Mech., 1052, pp. 349352. 32. Kaplan, Carl: On the Snmll-Dieturbanco Iteration Method for the Flow of a Compressible Fluid With Application to a Pnrabolic Cylinder. NACA TN 3318, 1955. 33. Kaplan, Carl: On the Uee of Residue Theory for Treating tho Subsonio l?low of a Compressible Fluid. NACA Rep. 728,1942.

31.Imai, Isao: Application

Surface. NACA Rep. 768, 1943. 11. Kaplan, Carl: The l?low of a Comp=ible

Fluid Past a Circular

Arc Profle. NACA Rep. 794, 1944 12. Imai, Isao: Tvw-Dimensional Aerofoil Theory for Compressible Fluids. Rep. hTo. 294 (vol. 21, no. 9), Aero. R~. Inst., Tokyo Imperial Univ., hlay 1944, pp. 283-331. (Japwusse text with English abstract.) 13. In@ Imo, and Oyam~ i%ichi: The Third Approximation of the Thin-Wing Expansion Method for Gompresible Fluids. Repts. Inst. of Sci. and Tech, Univ. TOLTO,VOL2, nos. 3+ hhr.-Apr. 1948, pp. 33-44. (Japanese text) 14. Van Dyke, hlilton D.: A Study of &cond-Order Supersonic Flow Theory. NACA ~p. 1081, 1952. 16. Keune, Friedrich: Daa Tragfl~el-Profil in der ebenen inkompressiblen und kompresiblen Str6mung. Rapport FI 12, FI 13, Flygtekniska Institutionen, Kungl. Tekniska H6gskolan, Stockhol~ Feb.-hfay 1951.

34.Allen, H. Julian: General Theory of Airfoil %otions Having Arbitrary Shape or Pressure Diatrfbution. NACA tip. 833, 1945. Second 35. MiIne-Thomson, L. hf.: Theoretical Hydrodynamics.
ecL, Maemlllan Co., 1949.

36.Jacobs, Eastman N., Ward, Kenneth E., and Pinkerton, Robert

hf.: The Characteristics of 78 Related Airfoil Seotions from Tests in the Variabk-Deneity Wind Tunnel. NACA Rep. 400, 1933. 37. Goldstein, S.: Notes on Ganeral Theory of Airfoil %ctions Having Arbitrary Shape or Pressure Distrfbutlon, by H. Julian Allen. Britiih A R. C. Rep. No. 7142, F. M. 624, 1943.

38.Abbott, Ira H., Von Doenhoff, Albert E., and Stivere, Louis S.,
Jr.: Surornary of Airfoil Data. NACA Rep. 824, 1945,

39.Hasimoto, H.: Application of the Thin-Wfug Expansion Method to the Compressible Flow Past an Elliptio Cylinder. Jour. Phys. Sec. Japan, VQL7, no. 3, 1952, pp. 322-328.

SECOND-ORDER

SUESONIC

AIRFOIL

TREORY

561

40. Emmons, Howard W.: Flow of a Compressible Fluid Past a

Symmetrical Airfoil in a Wind Tunnel and in Free Air. NACA TN 1746, 1948. 41, Graham, Donald J., NitrJwg, Gerald E., and Olsen, Robert N.: A Sy~matio Investigation of Pmseu< Distributions at High Smede ovm Five Remw.entative NACA IAMV-DraEand Conv;ntional Airfoil %ct~ons. NACA Rep. s.32, 1045. -

TABLE I.PIVOTAL POINTS, INCLINED FLAT PLATE SOLUTION, AND INFLUENCE COEFFICIENTS FOR P, ~ a

I
. . aowa hem .14646 2 U42
.2Q9M .W .69134 .s6366 .96104 1.4fw3 L6CKII .WQ .4142 . 1%9

1
N-S
N=16 ($.) .

I I

42. Tomotika, S., and Tamad~ K: Studies on Two-Dbnensional Transonio Flows of Comp=ible Fluid. Part III. Quart. Appl. Math, VO1. 9, July 1951, pp. 129-147. 43. Asak~ Sabur8: Application of the Thin-Wing-Expansion Method the Flow of a Cempreseible Fluid Pasta Sym~etncal Ciroular DD. Arc. Aerofoil. Jour. Phvs. Soo. Jamn. vol. 10. ,.. ., , no. 6. 1955. 482-492.

to

.
1 2 3 4 5 6 :

1r. r.

(for% ([2A) Co%)

-:4423 . 6735 0 . m ~. 0507 0 .Ca7 o .m o .673s

E.0273
1.4066

62.54%9 3.
& 4707

M43

2 U42

13. S@

. 2242 0

l. cwo

-4. aMo

.W32

2S726

.2242

. 4142

23431 L ml

II 11 12

. 1S32

3. 11H3

;] 15

&4423

TABLE H.INFLUENCE
% (velwlty due to thickness) 1 2 a 4 6 6

COEFFICIENTS FOR VELOCITY (a) N=8

& 7 1 2 (veloeltr due to camber) 3

I
o o

20,mm

-4. 071a lL S137 -4.3W3 o -. m o . 0719

o
-3.3692

o. 2242 0 3. 1s43 acuo6 -3.1643 0 . 2242

0 . 2377 0

-0. ona o -. m o -AS&x IL 3137 -4.0719

0 . Km o . mm o 7. 6246 m. w

1 2 3 4 5 6 7

2L E.S41 4. S4Qa 9. w .W 1. S307 2.116s . mm 17.1644 lL3137

-7.6240 0 -, Ss.5s o . 1329

0

8. SW 3. U42 o -.2377 0

-x U42
a S191
3. 3s92 o

N= 16
cu (veledty due to thlebHs) \n 1 2 3 4 5 6

T
o.%18 -&m 4. s90 o 9.C$C4 3. 02S5 .3767 .3W 0 . m 3.s4$s 7.1097 11 . 0.151S . m .1642 . 26.27 . 1W3 .U42 .2516 . 9741 7: %% 19.919!3 7. !MB3 1.73U L 07W I& S4w

0. mw o 3. 0#6 s.aoo 2 @43 o hs?m

o.2S13 -u 6719 . S790 .3914 -% elms 9.Fx.s -3. 7K16 7.lm7

u ml . 6120 .04K1 1. 6307 z lla5 9. 6406 Slwu

. 24Fd

-z 7bi9 11.3137 4 e719

10

12

13

14

16

Y
.520133 ;? O&4 -29, SW -1: Ants ; W-rl I T-l -iwl . 6467 0 . 4616 :. 2bw o . 19W -.1242 n ... 0 ( -pl&I -.6620 0 . 04fd -.0234 0 0 . 0134

I
o -.0310 0 -.0747 0 -. 1s67 0 -. we .-! 14a3 -1! 43M E o i% Ii 0614 0.0134 0 . 6451 0 . C9al o . 19ss o . 4616 m o lJ&~ ! 0 . o . 0 . 0 . o . i ea34 0620 1242 m 64s7 3#a

It 2032 23,7692
-11 Awl-l

+ Csd2 k WM

o . 7653

-$1363 . m

-6
1469

22 @274 -:0622

!. ma 19.M
: . o . 0 . 0 . o 7274 SIW 2834 1162 M37

&w 17.Wa
-: f@76

o . Iwl o . 6743 -l M -s. o . 0 . E433 3135 dew 6919 2172

-pm . m 0 . 6736 ;~ ~ -6.4423 0 . 67%5 o -. 2W9 0 . Cw37

o . 0764 o . 2172 0 . 6919 -: Ws lIL 3125 -y3 . 6743 o . 1831 o

-. w16 o

-. W3 o . X12 n -. .... I&m 0 . 1324 -.6747 0 o . 0s57 -. IJS1O o 0 -. Oln

. 7345
:. m . 0w2 o . 0m2

:.0764

o 0 . 6437 0 . 6%2 . 1109 o 0 . 2331 . 2324 0 0 . 7346 n . mm -i 9676 J. 7374 17.31H -$ a54.5 19.24W 7. m o . m . 7053 o . 1363 0
-0.6262

-o. Olrl
:. CL&97 . lW! 0 . 3o12 0 . !m3 : 0522 226274 : W4 . 6512

25!s9 S20133

d. (veloelty dne to mnk) -la m 41.Slm 1; 377s -L 0240 0 -. 4edo o . 196s o Lo1e3 o x 6015 2656s L 2@2 -. 2z& -9. E!17 o -. SM7 o . 66W n . 4%9 o L8@37 11 lm o. lFJS L07S9 .1642 ~; &w -7: 6-m7 . me . 9741 : in : E L 7341 . 2370 lh64m -pa . 7683 -: W45 17.31sa ; Bs9 . 6349 0 . ml o L3!?37 1: Es27 o. Em -o. 12=37 -.3934 o .1392 . 2251 . 23s2 0 . le3a -.0492 -& iQ2J5 lii 6267 ;! ~ -6,6972 . 3s7 -6.25.93 : g 0 . 2474 . 3&34 0 L SUM 1.Ells . 2403 13.Ws Ii 7610

-g. 00?a
. W42 o . 1%s o . 6447 4.7701 17.3m -:5193 .m 1} 6103

-$0317 ;64U . US16 o . KIM o . 72S3 : 6E25 ZL6274 7. 321S o 16.!3!3s

0.2272 : ~ -. 301s :i% -: E -1: E% 1.0125 11. w M% n. 41w

-0.0134 0 . 01s2 o . 0317 0 . 0e&3 o . Kk37 o . 3740 0 13. Wa 4LSIMI 16.CKd4

0.6469 : ~ . 7617 %-RI L0TZ3 ;. ~ i :% 3% Hi 14a 704s 2377

___.

562

REPORT

1274NATIONAL

ADVTSORY

COHTYI!EE

FOR

AERONAUTICS

TtiLE

111.INFLUENCE COEFFICIENTS FOR SLOPE (a) N=8

I%

(slope due to thkhmss) 4 6 0.6340 -L4046 26131 -h m 6 7 1 1 2 8 4 6 6 7 6.36M %2Z2Z -3. @w 21648 L ~ 2 9. 2202 L 4142 & 1648 2 w

f. I -o. &336
1.4142 z 3978 4m L @24 -2. 206a 3.69M -h 6W 3.3218

(slope duo to -mber) 3 4 6 1.mo7 1. m 6

. 7

a 1 2 3 4 6 6 7

1 -6.3WI

2 -4m2

I sI
L6307 -4. 7E3 . 44E3 h6MS -% 6131 L4046 0

-o. &s4 2m -48284

3. IE16s 2164s & 164s .4423 5.2262 182?4 5. 2262 0 6.!ZW 2 8284 2.1648

1. 2202 1.4142

1.Iw4 1.2DR 1.6207 -2.1048 8,m56 -Q. 2202 -a ?u%l

17.0479 -L 4142 -&9218 &w -3. m 22W9 -L W & 0347 -4m 23978 -L 4142 .6536

2.6131 l. ES62 -h 2262 2.8?%4

&3234 -x m .82S4

.4483 a 0647 L 71E3 L 6W7

1.E3.62 2 6131 1.33s2 -1. &307

-. 44s3 -0.1048 e% 1648 -1.4142 3. 6966 9.2262

L 4142 17. 0479 48339 6.m

1.2262 1. 4142 1. a324 Lm

(b) N-16
& (SIOP3

due to thl@km@ 9 10 11

=FF
-o. m

16 -: ~

2 0102 -17.9172 6. am 3L 4204 213m$ -_< ~ -17.0479 20.48s3 IL W 13.rd94 -l&w a 9213 7.41a3 a 0M7 h41&l h 6432 : g -44 :% -3.6!463 2 3976 L8626 :% ; g L 442 L 0213 k E . WI! : ag .M .Sm9 . S376

-Ii %%

-am

i w 47V35 -1;= . 44s3 -m. 0423 . 2023 1; E% IL 3181 la 4526 ~ ~ 180M7 h 1269 -4&2d 26Hl L 9134 :%% ;%% L 4046 L3364 :%% . @374 : %x . w .3978 . 33%3 .X@3 f.

:% kl%% ki%

-aW78 i=

aw . @74 L 1025 1. E&3a i %%

-a mm
: L4616 ;:% .3.6748 Uh 3% .44!33 lcl 4112 -4 7M 26%9 1. .5307 . ml

!%%

-1:%% 1.4142
1. w i E : g

L 3576 . 7%48 -: z 216SJI -2 m

-1:M41 i %

+3: E%!

1: H! 10:z -47V3U 23162 L W L07&3 . m

-$$% ; ~ : Pm la 84W -IL ~ -l& 51M 17.047!3 L 4142 XL 46%5 1213a66 -31. 4m4 1%%% -4m -4.7a?6 1?% 20162

(SIOWdm to camber) L W8 L 36L9 L 4419 L W 2 E $ %% 11 . lx 7. -L 1849 :% L 4142 L 6S39 k H% i k % 3.1423 -4 45M 7.24KI 16. a3Zl -2 w 2L 6427 13. 3944 -k E 26466 -3.3439 ;: ~ L 11X36 L 1394 -: %x L 4419 L6472 1.0100 -L06W 1.lmo 1. 1349 L 2018 -L4C4KI

-it %

2fM3s le a320 7. mm t %% 24064 L 94W L 6472 L4419 -: E

:%%

%% % & 1648 Ea3a _~ ;W&J 1232 2WI 16.Cram 9. m -H% 7.m

-3& 14m -% m

SECOND+RDER TABLE IV.INFLUENCE

SUDSONIC AIRFOIL =ORY FOR SECOND DERIVATIVE

u. (wend

derhtlve 3

due to thlckn~) 4
6

n
. 1 2 a 4 E 6 7

a -
106.262

12 693 624X! -97. w

.
3.s37 IL 314 5). 4@

-414292

1. 873 &m

69.478 -l@.m 4&m -46 ES2 23.314 -23. m 11,314 7&m6 -11.314

lLa14

34.066 84 m 314 Ea.4@

34M.9 07. w

!2019 -lL 0 -. m

3..W3 lL a14 L 373 a. s37

6241M -lea. m -E 6s5

11
COEFFICIENTS (a) N=8
h. (mmnd dwivative due to camk)

\ .

-o. m o

IL a14

7W. m %W

144152 24 m

23 082

2 a

184 030
93.446

9.373

-6. O@

& 476

la m

z 019 a14

a 314

73. Wl W6M

I @m2

70.M

16. m

&S32

a m

l& 476

34.627 m

l& m

lL 627

M. m

-IL

-46. Ea2 41L 627 w. 478

104 m

m. a14 -a

a 314 S@

-104 Cml

23.13M

7e% 523

-24 m

lL $37 la w am

54.627 -m.

93.446

7X M

6 7

la 479

-8. m

Ill m

m 481 -18LWI

Z3&m

106.233 -414.392

-16. w

& 476

-tow

9.37a

24 m

144132 71m.C@3

10 I

11

13

14

lb

ah 17 ma B 44216 -m al % x -Fi R :; o -. .E . 16

A% % E 2m. 21
-397.33 % k! 2216 IL 31 -:8 233 -L 37 .es h- (uemndddvatlw due to camk)
I

-0.36

.6s 1. B 210 x 60 1! : E % 276.74 -4&121 aki 31 7a 01 3L 16 12 16

-IIWL E3 42$L69 -10{9. 16 4s345 -fi ; -:% g

Ea 14 -144.16 47L77 -m: 93. 46 4L 58 ~ CEJ IL 92 $ ~

lL 93 a E t % 11.37 1.2 m 2 E 210.19 844 m 21a 19 m. ea E i 38 7.H -IL 48 :

&16 t 49 :: lL W I&10 ~ IxJ -;;

-;

7
--: ~

17. 10 ~

m. 47 1OL 83 1=04 21L 17

~.

-4Lg

m. 14 %: w 48

-4ti CH W4.40 ?la 24 %% 944C8

m% % 14$& m

%%E

&69

-m. 45 -% -%% _E# %

% 5 m 01 337. 78 214. aa 54 53 2h a7 ;: ~

Q 37 10.33 -lL 93 14.m -~ ; -N E &Cl z

2. E % E +!$ :
144 15 w. m % m 4m.45 1649. 15 4231. m 116% m

-i

4s &16

5. g a.m

-%w CO 47L 77 In % 144 M 82a 21 S&14 2s066

TABLE V.COMPRESSIBILITY

FACTORS (7==7/5)

M
0

K,
LW131 L IM13 LCWJ3 L 0114 Lm L03i?J :% L 11R3 L 1232 L 1329 Llaw L 1472 L 1647 L 1626 L 1707 L 17%4 L IE.31 L 1974

Krl
0
.0Lr2m . O1OM .a2a34 :E .10477 . lali9 .21224 .2aa30 .31161 . W77 .2$315 . 3%96 .m .42Ma .4S333 . 4bl18 . 5m .54634

0
. a)lm .Cnw9 .01167 . 021w :E% . 07bsa . 1CW2 . 14!?W . lIXXI . M&w . Imw . 18m3 :%% .fZEJ1 .24109 :%

o. a

:!fl ,16
,2U .2a .SJ1 .3.5 .40 .43 .46 ,47 .48 ,40 .m ,51
.32 .E3 .64 .3.5

.37
..59 .59

am .maa @3112 .
.mbm .75224 .fum :%%! i%% LLW2 L2126 i%l L 5146 LIW35 L 7716 L !3218 :%% -

Slkk
II 29KM . alou

X2-1 2 L 2447 L 3345 L 5M9 L M71 L E3711 20438 2 Z@31 2m 2 ml? 3.33M a.wtl 446m h 2512 am 7.3s36 1! %! 16 m H%

a76

.60 .61 .62 .a .64 :: .67 .53 .@ .m .71 :; .76

.SmE4 .352%s .37617 . 4ol&3 .42919 .4aw9 .46128 .m .E64e%l := . 7U213 . 757X2 .81827 .mawl . ml H%

:Z .79 .Ea :E .a .84 .83 .W .87 .E$ .m .W .91 .02 .E2 :2

- . . . . ...