https://ntrs.nasa.gov/search.jsp?R=19750019932 2019-04-12T06:25:35+00:00Z NASA TECHNICAL NOTE 14 PROCEDURES FOR THE DESIGN \ OF LOW-PITCHING-MOMENT AIRFOILS \ / , Raymond L. Burger Langley Reseurcb Center Humpton, Va. 23665 . 4. i t NATIONAL AERONAUTICS AND SPACE ADMlN&TRATlON WASHINGTON, D. C. ,AUGUSH975 2.- f ii TECH LIBRARY MFB. NM I 1. Report No. 2. Government Accession No. 3. Recipient's Catalog No. NASA TN D-7982 4. Title and Subtitle 5. Report Date PROCEDURES FOR THE DESIGN OF LOW-PITCHING- Aueust 1975 MOMENT AIRFOILS 6. Performing Organization Code i 7. Author(s) 8. Performing Organization Report No. L- 1011 4 Raymond L. Barger P 10. Work Unit No. 9. Performing Organization Name and Address 505-06-31-02 Y NASA Langley Research Center Hampton, Va. 23665 13. Type of Report and Period Covered 12. Sponsoring Agency Name and Address Technical Note I National Aeronautics and Space Administration 14. Spon~~ soring Agency Code I Washington, D.C. 20546 15. Supplementary Notes 16. Abstract Three approaches to the design of low-pitching-moment airfoils are treated. The first method decreases the pitching moment of a given airfoil by specifying appropriate modifica- I tions to its pressure distribution. The second procedure designs an airfoil of desired pitch- ing moment by prescribing parameters in a special formula for the Theodorsen +function. I The third method involves appropriate camber-line design with superposition of a thickness distribution and subsequent tailoring. Advantages and disadvantages of the three methods II are discussed. I I x rp 17. Key-Words (Suggested by Authoris) 1 18. Distribution Statement Airfoils Unclassified - Unlimited Design Low -pitching-moment airfoil Subi ect Cate gorv 0 1 I I 19. Security Classif. (of this report) 20. Security Classif. (of this page) 21. NO. of Pages 22. Price' Unclassified Unclassified 22 $3.25 I E PROCEDURES FOR THE DESIGN OF LOW-PITCHING-MOMENT AIRFOILS Raymond L. Barger Langley Research Center SUMMARY Three approaches to the design of low-pitching-moment airfoils are treated. The first method decreases the pitching moment of a given airfoil by specifying appropriate modifications to its pressure distribution. The second procedure designs an airfoil of desired pitching moment by prescribing parameters in a special formula for the Theodor sen €-function. The third method involves appropriate camber-line design with superposition of a thickness distribution and subsequent tailoring. Advantages and disadvantages of the three methods are discussed. INTRODUCTION Low-pitching-moment airfoils find application primarily as helicopter rotor blades; but more recently some attention has been given to the advantages of low -pitching-moment sections for a "span-loader" vehicle. For such applications a symmetric airfoil could con- ceivably be employed, but cambered airfoils can offer significant advantages. The usual difficulty that is encountered in the design process stems from the fact that the airfoil shape and performance are sensitive to the parameters that control the pitching moment. For example, an airfoil with zero pitching moment but with moderately small positive lift at zero angle of attack deviates significantly from a symmetric section. Similarly if one attempts to modify the lower surface of an airfoil to reduce the pitching moment, while retaining the upper- surface shape and pressure distribution, he generally finds that sub- stantial modification is required even for small reductions in pitching moment. In this paper three approaches to the problem of designing low-pitching-moment air- foils are treated. Generally these methods utilize equations or procedures that are already in the literature but that have apparently not been heretofore applied in a systematic man- ner, with the required modifications, to the specific problem of designing low-pitching- moment airfoils. SYMBOLS Ao,Al,A2 real coefficients a , b2 modulus of a complex quantity (see eq. (7)) C chord length section lift coefficient cZ C section lift coefficient at zero angle of attack 270 Cm pitching- moment coefficient about the quarter- chord point Cm, ac pitching-moment coefficient about the aerodynamic center airfoil pressure coefficient cP cl/R and c2/R2 are coefficients in the complex Fourier expansion c17c2 of *(@) !2 complex quantity (see eq. (6)) M local Mach number M, free-stream Mach number basic lift distribution pb R radius of circle into which an airfoil is mapped by the Theodorsen transformation t maximum thickness X, Y Cartesian coordinates mean-line ordinate yb a section angle of attack P negative of the angle of zero lift 2 I 2Y amplitude of a complex quantity (see eq. (7)) phase angle E function relating angular coordinates of near-circle and exact-circle airfoil transformations I rl =Y - P transformation variable (see eq. (10)) angular coordinate for points mapped from airfoil surface onto a circle function relating radial coordinates of near-circle and exact-circle airfoil transformations +, L12T0(@) average value of d@ 2n 0 GENERAL CONSIDERATIONS Inasmuch as three distinct approaches to the low-pitching-moment airfoil design problem are discussed in this paper, an initial comparison is perhaps appropriate to pro- vide orientation and avoid possible confusion. The first method is an application to this I specific design problem of the design technique that has been described in reference 1. This technique is applicable both to subcritical and supercritical airfoils. The design is effected by modifying an initial airfoil and providing an analysis of the modification on each iteration. The second and third methods represent the full-thick-airfoil theory and the thin- airfoil superposition theory, respectively, applied systematically to the low-pitching- moment design problems. They are both essentially incompressible, and in both cases i the design is initiated by specifying a set of parameters that determine certain airfoil characteristics. The pressure distributions are obtained by an independent analysis k program. All three design procedures are inviscid, but in each case an allowance for boundary- layer effects can be made. This problem has been discussed for the first procedure in 1 reference 1. For the other two methods, a rough estimate of the displacement thickness I i 3 ! effect can be obtained by a judicious use of the thickness distribution controls in the design process. However, if the primary boundary-layer considerations are loss of lift and increase in pitching moment, it is a simple matter to overestimate the lift and underesti- mate the pitching moment in specifying the design parameters. DE SIGN BY PRESCRIBING PRESSURE DISTRIBUTION VARIATION The first method to be described is applicable to airfoils that have approximately the desired characteristics but require a reduction in the magnitude of the pitching moment. The designer prescribes a change in the known pressure distribution of the original air- foil in such a way that the pitching moment will be changed in the desired manner without destroying the favorable characteristics of the airfoil. The usual procedure is to shift some of the loading from the rear forward by prescribing changes to the pressure distribu- tion along the lower surface and possibly to the rear of the upper surface. This method has the advantage that by working directly with the pressure distribu- tion the designer can avoid those adverse forms of pressure distribution that are condu- cive, say, to flow separation or to shock formation. Furthermore, he can indirectly con- trol the results of a specified change in the pressure distribution; that is, whether it will decrease the pitching moment, the lift, etc. However, the effects of a prescribed change on Cm, t/c, etc. are more difficult to control in a precise manner; and consequently cl, a number of zttempts may be required to obtain closely specified values for these param- eters. If the thickness is altered slightly in the design process, it can be adjusted by the incompressible method of reference 2. Figure 1 shows two variations of a basic airfoil that were obtained by this procedure with the use of the design technique of reference 1. The lower-surface pressure distribu- tion was altered so as to unload the airfoil near the rear. To compensate for the conse- quent loss of lift the loading was increased in the middle part of the lower surface. In the first variation (fig. l(b)), the pitching-moment coefficient remained constant at -0.025 while the lift was increased by more than 25 percent. In the second variation (fig. l(c)), the lift remained virtually constant while the pitching moment was reduced in magnitude to the more acceptable value of -0.010. In this example, the decrease in thickness ratio represents a significant alteration to the geometry of the original airfoil. It is generally true that moderate changes in pitch- ing moment are associated with relatively large changes in airfoil shape especially if the lift coefficient is held constant. For this reason, it is often preferable to start with a design that has near zero, or even positive, pitching moment, and then, if necessary, tailor that design. Such a procedure is discussed in the next section. 4 J DESIGN BY SPECIFYING €-FUNCTION PARAMETERS A second procedure for the design of low-pitching-moment airfoils is based on a formula used in reference 3 for the E-function of a class of airfoils. Basically, this formula E(@) = A1 sin ($I - 61) + A2 sin (2$I - 62) (1) represents a simplified €-function with only two Fourier components specified in terms of the amplitudes and phase angles. For this E-function, the angle of zero lift -p is approx- imately determined from the relation p = E(T) = A1 sin 61 - A2 sin 62 The conjugate function to E(@) (see ref. 3, eqs. 11 and 12, where the notation is slightly different) is +. where +o is the average value of In order to compute the pitching moment, two complex numbers are needed: c1= -RJ:7 r +(@) ei@d @= AIR cos 61 + iAlR sin 61 (3) 7r and I:'+(@) "," c2 = - e2i@d @= A2R2 cos 62 + iA2R2 sin 62 (4) where R is the radius of the circle into which the airfoil is mapped by the Theodorsen transformation. Now the real number a is related to R by i R = ae*o L and the complex quantity g defined by n c2 5 is represented in polar form as in accordance with the procedure of reference 3. Then the pitching moment about the aerodynamic center is given by (see ref. 3, eq. 51) P) It is clear that the value of Cm,ac depends on the angle q = y - P; specifically, cm,ac = 0 when q = 0. Now one can express q = y - /3 from equations (2) and (7) in terms of AI, A2, 61, 62, and +o by means of equations (3) to (6). Thus a unique airfoil can be determined by specifying the five parameters AI, A2, 61, +o, and q in the equation y - P - q = O (9) and solving it for This highly nonlinear equation is solved by interval halving. 62. Varying each parameter produces a class or family, of airfoils. The value of q chosen controls the pitching moment according to equation (8). The selection of the other parameters requires some care. Although varying any one of these parameters influences to some extent all the airfoil characteristics, each individual parameter has a dominant influence on a particular property of the airfoil. The value of A1 provides the basic thickness distribution, which is then modified by the choice of A2. The effect of varying A1 can be seen in the example shown in fig- ure 2(a). (For all the airfoils shown in fig. 2 the pitching moment about the aerodynamic center is essentially zero.) Very small values of A1 yield a shape very much like an ellipse, whereas large values produce negative thickness near the trailing edge. When A2 is varied, the distribution of thickness is modified, as shown in fig- ure 2(b). The magnitude of A2 also influences the extent to which the second term in equation (1) affects the airfoil performance. Since this term is the one that involves 62, j equation (9) may not be solvable for 62 if A2 is too small. On the other hand, large values of A2 (relative to Ai) tend to produce impractical distorted airfoil shapes. This i effect is seen in figure 2(b) where for A1 = 0.1 and A2 = 0.06, the airfoil becomes too thin in the 75-percent chord region. The parameter +o affords a control over the max- imum thickness (ref. 2), as is seen in figure 2(c). The parameter 61 primarily controls the lift, as indicated by figure 2(d), where varying 61 from 0.1 to 0.9 radian has the effect of changing p from 0.0027 to 6 0.03 radian. Notice, however, in figure 2(b), that the variation of A2 has very little effect on the lift. Since five parameters can be varied in this design procedure, it appears that a wide variety of shapes and characteristics is attainable. However, the fact that the €-function is represented by only two Fourier components is a significant restriction. Furthermore the availability of numerous parameters is in one sense a disadvantage in that the designer might spend a considerable time "toying" with the parameters in an effort to obtain exactly some desired design characteristic. These difficulties can usually be circumvented in actual practice. For example, the airfoil shown in figure 3(a), which was designed by this method, was too thick near the trailing edge. Its other properties - lift, pitching moment, and maximum thickness - were satisfactory. Therefore a smooth analytic fairing was made, starting at the 0.60 chord station and proceeding to the trailing edge, so as to reduce the thickness in this region while maintaining the same mean line. The resulting airfoil is shown in figure 3(b), together with its pressure distribution. (The viscous pressure distributions in figs. 3 to 5 were computed by the method of ref. 4.) The lift, pitching moment, and maximum thick- ness are essentially unchanged, but the trailing-edge angle and consequently the pressure distribution near the trailing edge are improved. Of course, not every arbitrary combination of parameters yields a solution of equa- tion (9). Furthermore, as has been seen, even those combinations that yield a solution do not necessarily correspond to a practical airfoil shape. DE SIGN BY GEOMETRIC SUPERPOSITION Perhaps the simplest approach to the design of airfoils is to design the mean line and then superimpose a thickness distribution on it. In reference 5 it is shown that, if the e* variable is defined by the relation x =-C( I - COS e*) (10) 2 then the basic lift distribution (that which is dependent only on the mean-line shape and not on the angle of attack) can be represented by a Fourier sine series G 2 Pb = 4 An sin (ne*) (11) n= 1 Then reference 5 also shows that the distribution of slope of the mean line dyb(8*)/dx at the ideal angle of attack is the conjugate of Pb(67/4 provided that both functions are extended to the interval (7~,2n)w ith dy/dx symmetric about 7~ and Pb antisymmetric 7 about T. The situation is similar to that in thick-airfoil theory where the €-function can + be prescribed and its conjugate - t,bo can then be calculated to determine the airfoil geometry. Here, a basic lift distribution can be prescribed and the corresponding mean line calculated. For a lift distribution expressed as a sine series as in equation (1l ), the conjugate of ~b/4 is dyb = An COS (ne*) dx n= 1 Naturally some experience would normally be required to design a lift distribution that provided the desired lift and pitching moment as well as a reasonable mean-line shape. However, a simpler, more direct approach is available. From reference 6, equa- tions (4.7) and (4.8), it is seen that the lift coefficient at zero angle of attack is simply where and the pitching-moment coefficient about the quarter-chord point is Here A1 and A2 are the first two coefficients in the Fourier series of equation (11). Thus, in the design of a mean line, the lift coefficient can be controlled by specifying the value of A1 and the pitching-moment coefficient is proportional to the difference between A2 and Ai. Specifically, A2 = A1 gives a pitching-moment coefficient of zero. Families of mean lines can be derived by specifying various values of A1 and A2 in a simple 2-component lift distribution. However, it should be noted that large values of A2 yield impractical distorted mean lines; consequently, large values of lift cannot be specified if the pitching moment is required to be near zero or positive. 1 For each mean line so derived, a family of airfoils can be obtained by specifying var- I ious thickness distributions. It is in this phase of the design that the superposition proce- dure of this airfoil theory displays its limitations. These limitations appear whenever the assumptions of thin-airfoil theory are violated; specifically over the entire airfoil if it is sufficiently thick and near the leading edge for any airfoil. The former problem is not as troublesome as the latter.