REPORT No. 465 DETERMINATION OF THE THEORETICAL PRESSURE DISTRIBUTION FOR TWENTY AIRFOILS By 1. B. Ganntox SUMMARY ‘Thia report gives the theoretical distribution of pressure at lift coeficiente of 0, 0.6, 1.0, and 1.5 for 20 airfoils, calculated on the basis of a rigorous potential theory of arbitrary airfoile. It also provides tables from which the characteristics of the airfoils for any angle of attack in 2-dimensional potential flow are readily calculable. The theoretical ealuce of the angles of zero lift the lift and moment coeficients, and the ideal angles of attack are listed and some comparisons with experiment are indi- cated. Some of the well-known characteristics and properties of airfoils are accounted for in terma of the theoretical preseure-distribution curces. Qualitative de- ductions are made concerning the causes of breakdown of potential flow and the efficiency of the airfoil in viscous flow. The results presented may be of value in pre- dicting structural loads and also in a correlation of theo- retieal pressure gradients with profile resistance. INTRODUCTION Until recently the theoretical distribution of pressure around airfoils could bo determined only for the so- called “theoretical” airfoils. Indeed, only in the par- ticular caso of tho Joukowsky airfoils is the calculation not unduly laborious. (See references 1 and 2.) ‘The theoretical airfoils, which are defined by special mathematical transformations, have, however, seldom been employed in practice. "Their use in e precise study of pressure distribution has in fact been due more to necessity than to dosiro. The distribution of pras- sure for mathematically “thin” airfoils (.., the airfoil is represented by the mean-cember lino) can be ob- tained, at least approximately, by the proceses given by Munk, Glauert, and Theodorsen (references, 3, 4, and 5). In another report (reference 6) Theodorsen doveloped a theory readily applicable to arbitrary airfoils. ‘This theory was extended by Theodorsen and Garrick in reference 7, in a report which gives a unified treatment of the 2-dimensional potential flow around airfoils of any shape. ‘The treatments given in references 6 and 7 avoid approximations in the anal ysis, and are referred to for details of the underlying ‘theory of the results of the present paper. ‘The differences exhibited by airfoils in potential flow, as well as the differences between the actual and ‘deal cases for a particular airfoil, can, of course, be critically studied only if tho ideal caso is known, ‘Furthermore, it is only on this besis that the assump- tions of the theory itself can be critically analyzed and modified. It is therefore believed that an oxisting gep in nerodynamical literature will be, to some extent, bridged by the publishing in the present paper of convenient tables and curves of the theoretical results for a number of commonly used and related sirfoils. ‘A knowledge of the theoretical distribution of pras- sure for an airfoil is, undoubtedly, & major factor in making it ultimately possible to predict accurately the behavior and efficiency of the airfoil under actual con- ditions, for the theoretical changes along the surface from pressure to velocity and from velocity to pressure are very significant in the determination of the drag characteristics, Knowledge of the theoretical results is of considerable value, too, for guiding experi- ‘mental work whenever the measurements are rather critical, and such information also directs attention to the significance and interpretation of differences between theory and experiment. ‘Unfortunately, beceuse of lack of sufficient accurate experimental data, comparison cannot be made direetly with wind-tunnel results except in a fow cases. In reference 7 an interesting comparison was given between, theory and experiment of the pressure distribution around the N.A.C.AM6 airfoil at 12 different angles, of attack. Reference § may be referred to for quali tative exporimentel results for five additional airfoi ‘A more accurate experimental study of pressure dis- ‘tributions is in progress at the present time at the N.A.C.A. Inboratories. ‘A part of the following work was undertaken at the request of the Bureau of Aeroneuties, Navy Depart ‘ment, for use in work on structural loads. 438434, In making the calculations the author was ably assisted by Miss Alyce V. Rudeen, of the Committee's Ce SUMOARY OF FORMULAS USED ‘Tho formulas used to obtain the results presented in the tables and curves are developed in referencos 6 and 7. A sample calculation for the N.A.C.A.-M6 airfoil cwith a comparison with experimental results, as well as explanatory figures ond diagrams illustrating the use of the formulas, is given in reference 7. The following list presents the symbols employed and their definitio sramol ppensini0N (). G, ¥) Seo discussion of the choice of axes in a following paragraph. @ 0 anton te where p=1~(2) (3) ©) ¥ deinbry=—p+ WPF Since y ia gon- erally small for airfoils, the following equation may be preferable: y : sinh¥~ >a Near the leading (or trailing) edge y is given epproxi- rately by o=/§ whore vis thozadius of curvature at the leading (or trail- ing *) edge. @ © = FH) oot 25H a See appendix? of reference 7 for method of evaluation, () ¢ Obtained graphically from the «, @ curve. @enoted $f in reference 7.) (©) ¥ Obtained graphically fromthe, # curve. a Denoted 55 in reference 7.) Mo $rbte ®) vo Aconstaint: vo “ES ¥@ ae (@) @ Angle of attack with respect to thez axis. (10) 6 The angle of zer0 lift, given by the value of efor 0=z, eh te) Gy kk TeaRy sm) CEVA Note: & is independent of the angle of attack, 0) % The ratio of the local velocity at the airfoil surface to the uniform stream Seago ames aaa ye re iets beers REPORT NATIONAL ADVISORY COMMITTED FOR AERONAUTICS po Hsin (at ¢) +sin(a+ A). ‘The ratio of the local superstream pres- sufe to the dynamic pressure (tho term “superstream pressure” is used to designate the difference of the local pressure and the static pressure in the ‘undisturbed uniform stream): Bai-(2) = Bat (5) and g= 57%, ‘The segment of the 2 axis intercepted by tho airfoil boundary. ‘Tho lift coefficient E82 inlat 6) 2 a3) a4) ¢ a5) OG, On: ecv? A point designated tho “focus” of tho air- foil. Wo may fitst define the complex constants o; and ¢ a8 eemelb= Ay +iBy (6) F LF 40) (om ovisin 9 a6 en Ati, ate ¥(G) (cos 2g-+4 sin 29) dy ‘Then writing : pow i tie wo have Bo (14 EERE a) 4 ABB eae z ‘Then the complex coordinate of F is ar=erindrm mers Eeorn (17) My The moment at F is constant for all angles of attack: Mym2e pb Vin 2y~ 8) ‘Tho moment coefficient referred to the point F: (18) Caer eee Oey Bt te 8 sin 2C4~ 0) (19) ay ‘The “ideal” angle of attack: _ ater ayn 25 where «y and ey denote, respectively, tho values of ¢ at the noso and tail; i.e, for 0 and 0==, respectively. ‘The ideal angle of attack for thin airfoils has beon defined by Theodorsen (reference 5) es that angle for which the front stagnation point is at the leading edgo.‘THEORETICAL PRESSURE DISTRIBUTION FOR TWENTY AIRFOILS At this angle of attack lange velocity gradients at the lending edge are avoided and the profile drag is at, or very near, its minimum value. ‘The definition ean bo naturally extended to actual airfoils to designate that angle of attack for which the front stagnation point is, at the foremost edge of the mean-camber line. How- ever, as pointed out by Theodorsen, the effective mean- camber line of a thick airfoil actually alters with change of angle of attack, and the ideal angle of attack for a thick airfoil represents an average of a rango of angles for which the profile drag is very near its minimum, PROCEDURE AND ACCURACY OP THE CALCULATIONS In order to evoid possible confusion it may be well to state beforehand that the term “chord” is used in this paper as synonymous with the segment.of the z axis intercepted by the airfoil. ‘The “standard chord” in terms of which tho airfoil is usually empirically dofined does not, in general, coincide with this above- defined chord. The angle between the z axis as chosen and the standard chord is designated » and is listed in table I. Gee also fig. 1.) In the procedure of the calculations, the axes of coordinates aro first chosen in a definite convenient way, sinco the easo and rapidity of convergence of further evaluations depend considerably on tho choice of axes (Feferences 6 and 7). This choice may bemade as follows: If the distanco botwoen the leading edge of the airfoil and the center of curvature of the lending edge is bisected at E (tho coordinates of Hare (2, 0)), and the sume is done for the trailing edge at Z” (the coordinates of 2’ are (—2, 0)), then the z axis should puss through EH’ and the origin bisects tho distance EE’, However, small variations from this particular choice of axis and origin do not noticeably influence the ease of caleulation* The quantities given in the headings of teble II are directly calculated in terms of z and y by means of the formulas previously listed. ‘Tho angle of attack corresponding to a given value of tho lift coefficient may be obtained from (15), in which cis Xy— Xx, where Xy and Xz denote the abscis- sas of the lending edge.and trailing edge, respectively. ‘The moment coefficient Cyr may be obtained from (18), in which the constants 8 and 7 are obtained from (16) by graphical integration of the ¥ sin , ¥ cos @, sin 24, and ¥ cos 26 curves. Tho ordinates of the airfoil are given empirically to hundredths of s percont of the standard chord for 16 stations’ of the upper and lower surfaces resp. tively. ‘Tho quantities 2, y, y, and ¢ are defined to tho sume degree of accuracy. ‘The y, @ curve is thus a faired curve through 32 points and ¢ (#) is estimated to bbe of the same order of accuracy ss ¥(0). ‘The deriva- {ay peng, tet mabe cba th re Eee See Se eee fst ort al reste ee "Rear area sober bo peperty Routed pee a encinrs cangcla Go Sagi Beige ae mente is a a paaton the a at 435 tives «’ and y’, being determined graphically, admit of 18 possible small error which, however, causes an error in k of probably less than 2 percent. ‘The angle of zero life, or the value of ¢ for #=x, may perhaps be inerror ‘as much as 16’, but the influence of this possible error ‘on the theoretical pressure-distribution curves for fixed values of O; is negligible. ‘Tho numerical data for the Clark Y airfoil are pre- sented in table IT, ‘The distribution of velocity and pressure for any angle of attack or at any lift coefli- ciont, as well as other theoretical characteristics, axe obtained with o minimum of effort from this table, Similar tables for the remaining airfoils are omitted hore for reasons of economy in printing and also be- cause itis mot known how genoral the interest in them will be. ‘They are available on request from the National Advisory Committee for Asronautics. DISCUSSION Although the airfoils chosen in this paper are mainly conventional airfoils (fg. 1) and not extremely radical types, itis nevertheless possible to isolate some of the individual effects of change of shape and compare these with experimental results. Tt is believed, how- fever, that future experimental work on radical and {ess ‘conventional shapes, for which the theoretical results are readily available (Gee, for example, reference 7, p. 31), will enable the isolation and analysis of effects which are probably masked and unemphasized in eonventional types. ‘We may first mako somo general comments regard- ing the curves of theoretical pressure distribution given in the following pages. In each figure the abscissa represents the location of a point of the airfoil surface in percent chord and tho ordinate gives the quantity Pla, the ratio of the local superstream pressure to the dynamic pressure q. It may be noted that negative ‘values of p/g are plotted upwards, ‘This is an arbitrary convention and is made because it is more readily associated with the upper surface of the airfoil, which {for ordinary angles of attack is the surface of suction or negative pressure. In figures 2 to 21, inclusive, it may be noted that the points of the curves above the zero, or normal pressuro, line represent sustion; that is, velocities, greator than V. Positive values of p/q denote pressures greater than normal static pressure; ie. 0> HRY) Re TER) EB] HR) la) ta) ch) agi’ Bite a eae ee es ces os ence sl pd mc lr aR