https://ntrs.nasa.gov/search.jsp?R=19800024837 2019-04-11T21:19:38+00:00Z i.IASA Technical Memorandum 81 221 Experimental Unsteady Aerodynamics of Conventional and Supercritical Airfoils Sanford S. Davis and Gerald N. Malcolm August 1980 Nationai Aeronautics ar7d Space Administration NASA Technical Memorandum 81221 Experimental Unsteady Aerodynamics of Conventional and Supercritical Airfoils Sanford S. Davis and Gerald N. Malcolnl Ames Research Center, NASA, Moffett Fiela, California Natlan,?lA ei~~auialccd~ Space R3rn1rl~strali,1n Ames Research Cmter Moffett F~eidC a~tfnrrl~9a4 035 TABLE OF CONTENTS . . . . . . . . . . . . . . . . . . . . NOMENCLATURE v . . . . . . . . . . . . . . . . . . . 1 INTRODUCTION 1 . . . . . . . . . . . . . . . . . . 2 TESTHARLWARE 2 . . . . . . . . . . 11- by 11-Foot Trapsonic Wind Tunnel 3 . . . . . . . . . . . . . . . . . Splitter Plates 3 . . . . . . . . . . . . . . Wings and Push-pull Rods 4 . . . . . . . . . . . . . . . . Motion Generators 6 . . . . . . . . Pretest Verification of System Components 7 . . . . . . . . . . . . . . . . 3 DATA ACQUISITION SYSTEM 7 . . . . . . . . . . . . . . Dynamic Data Acquisition 8 . . . . . . . . . . . . . . Static Data Acquisition 9 . . . . . . . . . . . . . . . . . . . 4 TEST PROGRAM 9 . . . . . . . . . . . . . 5 DATA REDUCTION AND PRESENTATION 10 . . . . . . . . . . Static-Pressure Coefficients 10 . . . . . . . . . . . . . Integrated Static Pressures 11 . . . . . . . . . . Dynamic Pressure Complex Amplitudes 11 . . . . . . . . . . . . Integrated Dynamic Pressures 12 . . . . . . . . . . . . . . . . 6 SUMMARY OF RSSULTS 12 APPENDIX A: METHODS FOR INTEGRATING EXPERIMENTAL PRESSURE . . . . . . . . . . . . . . DISTRIBUTI3NS 14 . . . . . . . . . . . . . . . . . . . APPENDIX B 16 . . . . . . . . . . . . . . . . . . . . REFERENCES 17 . . . . . . . . . . . . . . . . . . . . . TABLES 19 . . . . . . . . . . . . . . . . . . . . . FIGURES 44 iii NCMENCLATURE complex amplitude of the unsteady airfoil motion: for pitching, A = oscillatory angle of attack in radians; for plunging, A = displacement normalized by half-chord. The physical motion is ~ e ( ~ e ~ ~ ~ ) ALPHA mean angle of attack, deg C chord of airfoil, m + CL mean lift coefficient, up + CL,A normalized unsteady lift coefficient, up + CM mean moment coefficient at leading edge, nose up normalized unsteady moment coefficient at leading edge, + nose up CPU ,A (CPL ,A) complex amplitude of the unsteady upper (lower) surface CPU (CPL) mean value of upper (lower) surface pressure coefficient, PU (PL) - PINF QINF + exp(iwt) cos wt i sin ut 1 f FREQ frequency, Hz, -T IU,A(Q)(IL,A(Q)) Qth moment of the complex amplitude of the unsteady upper (lower) surface pressure coefficient IU(Q) (IL(Q)) Qth moment of the mean value of upper (lower) surface pressure coefficient wc - k,K reduced frequency, 2U free-stream Mach number Mm complex amplitude of the unstead pressurr ; the physical pressure = Re(Pelbt) PINF fr ee-stream static pressure, ~ / r n ~ PL,P U mean value of surfa ce pressure, N/m2 PTOT total essure, ~ / m ~ QINF dynamic pressure, N/m2 Re, RE chord Reynolds number period of the motion, sec time, sec free-stream velocity, m/sec distance along airfoil, m complex amplitude of unsteady angle of attack, deg mean angle of attack, deg instantaneous angle of attack, deg Complex notation: imaginary part of [ 1 magnitude of [ 1 phase of [ 1, deg real part of [ 1 EXPERWENTAL UNSTEADY AERODYNAMICS OF COKVENTIONAL AND SUPERCRITICAL AIRFOILS Sanford S. Davis and Gerald N. Malcolm h e s Research Center SUMMARY Experimental data on the unsteady aerodynamics of oscillating airfoils - in transonic flcw are presented. Two 0.5-m-chord airfoil models an NACA 648010 and an NLR 7301- were tested in the NASA-Ames 11- by 11-Foot Transonic Wind Tunnel a t Mach numbers to 0.85, at chord Reynolds numbers to 12x106, and a t mean angles of attack to 4'. The airfoils were subjected to both pitching and plungicg motions a t reduced frequencies to 0.3 (physical frequencies to 53 Hz). The new hardware znd the extensive use of computer-experiment integra- tion developed for this test are described. The geometrical configuration of the model and the test arrangement are described in detail. Mean and first harmonic data are presented in both tabular ana graphical form to aid in com- parisons with other data and with numerical computations. 1. INTRODUCTION The unsteady aerodynamics of both fixed- and rotary-wing airfoil sections must be thoroughly understood in order to provide safe margins for flutter, buffett, and other undesirable aerodynanic phenomena. This need is most apparent i n the critical transonic speed regime where these detrimental effects are most prevalent. Recent developments in numerical simulations of transonic unsteady aerodynamics have also highlighted the need for new experimental activity in this area. In response to these needs, an extensive test program was developed at Ames Research Center to measure the unsteady aerodynamics of both a conventional and a supercritical airfoil under a wide range of flow conditions. The objective of the test was to measure unsteady pressure distributions at higher Reynolds numbers over a more extensive range of test conditions than had heretofore been attempted. This report presents, in graphical and tabular form, the mean and fundamental frequency data from that test. The data were obtained in the 11- by 11-Foot Transonic Wind Tunnel at Ames Research Center. Over 200 data sets, representing various combinations of airfoil geometry, Mach number, Reynolds number, mean angle of attack, motion mode, motion amplitude, and frequency arc reported. For each data set both the mean and first harmonic loads are tabulated, and the pressure distributions are presented in both tabular and graphical form. Section 2 describes the important features of the test apparatus in detail, including the wind tunnel, model installation, motion generators, model construction, and model geometry. (Some of the hardware was also described in ref. 1.) A discussion of the computerized data system, devei- oped especially for this test, is provided in ssction 3. The software was written such that on-line comparisons could be made between the current data set and theoretical predictions. The measuring system is also described in references 1 and 2. Section 4 outlines the test program and section 5 pre- sents the data. The method used to ir-tegrate the chordwise pressure distri- bution is described in appendix A, and the tabulated first harmonic pressure data, enclosed in microfiche form, is designated appendix B. Some of the data have already been analyzed and can be found in refer- ences 3-6. A small subset cf the data has been selected by AGAFJ) for inclu- sion in its "Standards for Aeroelastic Application "; it is cited in section 4. 2. TEST HARDWARE The arrangement of the apparatus and the special two-dimensional flow channel installed in the 11- by 11-Foot Transonic Wind Tunnel were based on the choice of an acceptable ratio of wind-tunnel height to wing chord (greater than 6). A chord of 0.5 m was c2ioser?, resulting in the ratio (height)/(chord) = 6.8. Lowest hardware cost and minimum overall tunnel blockage could be obtained with a model spanning the tunnel, but construction of a full-span 0.5-m-chord model was impractical because first priority was assigned to obtaining high frequencies with minimal aeroelastic effects. An acceptable span-to-chord ratio of approximately 3 (1.35-m span) dictated the use of the splitter-plate arrangement shown in figure 1. Although previous investigators have successfully used splitter plates, a pilot test of the concept was nonetheless conducted in the Ames 2- by 2-Foot Tran,onic Wind Tunnel (ref. 7). : is test demonstrated that good quality transonic flow could L,e obtained with the chosen splitter plate arrangement. Figure 1 shows the general arrangement of the ving/splitter-plate/ actuator system as installed in the vind-tunnel test section. The normal 3.35 m x 3.35 m test section was segmented with two steel splitter plates, 3.35 m high by 2.8 m long. To minimize blockage, the thickness was the minimum necessary to accommodate the push-pull drive rods. To prevent exces- sive deflections of the splitter plates, side struts were installed for lat- eral support. The splitters extended illto the tunnel's plenum area at the top and bottom; there they were bolted to I-beam anchors. Access panels for instrumentation cables a1.d clearance for the push-pull rods were included in the splitter plate design. The wi.1g model was instrumented near its midspan station and attached to independently controlled hydraulic actuators throu2h the push-pull rods. Thus, the wing was free to pitch and plunge in response to the actuator's command signal. The wing was restrained in the fore-aft direction by a pair cf carbon-epoxy drag rods, and in the lateral, roll, and yaw directions by sliding cover plates, which moved with the wing on the inner surface of the splitter plates. The hydraulic actuators, located in the lower plenum area, were supported by flexures; they bore directly into a massive concrete foun- dation through the four support columns. With this design, the tunnel pres- sure shell does not have to support the oscillatory reaction loads induced by the actuator's motion. The capabilities of the test apparatus include sinusoidal pitching I oscillations over a frequency range of 0 to 60 Hz, with the maximum oscilla- tion varying fro^ ?2O at low frequencies to f0.8' at 60 Hz abmt any chord- wise axis, and a vertical plunging motion up to 25 cm (2 in.). The various components that make up the system w i l l be described i.n more detail since the basic performance requirements dictated state-of-the-art designs in many cases. Many of the components are shown in the installation photograph in figure 2 and the pre-test setup in figure 3. In the following description it may be helpful to refer to these photographs to visualize the interrelationship among the various components. 11- by 11-Foot Transonic Wind Tunnel The 11- by 11-Foot Transonic Wind Tunnel is a closed-return, variable density facility with a 3.35 x 3.35 x 6.7 m (11 x 11 x 22 ft) test sectiol, enclosed in a 6-m (20-ft) diameter cylindrical pressnre cell. The air is driven by a th:se-stage, axial-flow c9mpressor powered by four induction motors with a maximum continuous combined output of 135 MW (160,000 hp). The Mach number can be varied continuously from 0.4 to 1.4 with the stagna- tion pressure variable from 50 k ~ / mto~ 2 25 k~/m*( 0.5 to 2.25 atm) resulting in Reynolds numbers rrom 6x10/~m to 31x106/m. Maximum Mach and Reynolds numbers for this test were 0.85 and 25x106/m, respectively, The ventilated wall of the 11-Foo: Transonic Wind Tunnel has a baffled slot arrangement (fig. 4). Six slots - 1.78 cm (0.7 in.) wide - between the splitter plates yield an effective open area ratio of approximately 8%. A resistive baffle fabricated from 0.16 cm (1116 in.) sheet stock was inserted in each slot. The baffle is flush with the flcor and ceiling, extends 5.72 cn (2.25 in.) into the slot, and has a "wavelength" of 3.43 cm (1.35 in Splitter Plates Vertical splitter plates with trailing-edge flaps and horizontal side struts form the support structure for the wing. They each h&ve a sharp lead- ing edge and a movable trailing-edge flap which is manually adjustable betwsen 22' from the plane of the splitter plate. All testing was done with the flaps set at 0'. Horizontal side struts attach to the outside of the splitter plates just below the horizontal plane of symmetry 2nd protrude through the test section into the exterior structure. They provide stabilization and eliminate excessive lateral deflection from the aerodynami-c loads. The splitter plates were installed with a 0.1' divergence angle from tunnel centerline to account for boundary-layer growth. The thickness cf the splitter plates varies in the streamwise direction in the following manner: following the sharp leading edge the next immediate section is 3.2 cm (1.25 in.) thi-k; it is followed by a 5-cm (2-in.) thick s e c t i ~ ni n the center to accommodate the push-pull rods. The trailing-edge section is 4.4 cm (1.75 in.) thick and tapers to a sharp trailing edge. The inside surface of the splitter plate is straight with a l l thickness variations tak- ing place on the outer surface. Openings in the splitter plate (figs. 5, 6) permit the wing to be attached to the top of the push-pull rods, which are centered in four channels cut into the lower portion of the splitter plates. When the wing is oscil- lating, sliding covers (figs. 7, 8) attached to the wing seal the openings. The covers are made of graphite epoxy to reduce weight and are Teflon-lined for free sliding. The splitter plates contain a total of 125 static-pressure orifices distributed over the inside and outside surfaces of both plates. The inside orifices were utilized to select the proper channel Mach number and, in con- junction with the outer taps, were used to monitor the loading on the splitter plates. While testing, accelerometers on the trailing-edge flaps were used to sense any large or potentially destructive f k t t e r motions such as might be produced from the oscillating flow behind the wing or naturally induced from the channel air flow. Wings and Push-Pull Rods Model geometry- Two airfoil sections were chosen for this test program -- one a conventional airfoil (an NACA 64A010) and the other a supercritical air- foil (the NLR 7301). The two wing mcdels - span 1.35 m (53.2 in.), chord 0.5 m (19.685 in.) -were designed to withstand accelerations of 2 . 3 ~ 1 0m~/s ec2 (230 g) and aerodynamic loads of 44,000 N (10,000 lb). Both airfoils were subsequently chosen for inclusion in the AGARD standard series of test cases for aeroelastic applications (refs. 8, 9). Photographs of the models installed in the wind tunnel are prese;.ted in figures 7 and 8. Due to expansion of the molds in fabricating the models, the artual airfoil sections were slightly thicker than their theoretical counterparts. To expedite - numerical simulations, three sets of ordinates are presented the measured ordinates, smoothed versions of the measured ordinates from Olsen's computa- tions (ref. 8), and the theoretical ordinates. Because the measured ordinates contain large variations in the higher derivatives that adversely affected sore trial solutions, it is recommended that either the smoothed or theoreti- cal ordinates be used for computing. Computations using the theoretical ordinates were satisfactory for the flow conditions attempted. The measured and theoretical airfoil sections are shown in figure 9. In each case -he measurements correspond to the thicker section. Data for the NACA 64A010 aad NLR 7301 airfcils are presented in tables 1 and 2, respectively. Model instrumentation- The wing is instrumented with static pressure taps and dynamic pressure transducers, all of which are located at approxinately midspan. The dynamic pressure transducers communicate to the wing surface via a small orifice with a small volume cavity. Locations of the static and dynamic orifices in both wings are shown in tables 3 and 4. It should be noted that dynamic transducers were not inst~lledi n the lower surface of the NLR 7301 airfoil. The lower surface unsteady pressures were sacrifi-ced on that airfoil for the sake of increased resolution on the upper surface. Static pressure tubes are routed from the end of the wing through a cavity in the splitter plate to the tunnel plenum chamber below, and out an access port to scanivalve-transducer units exterior to the tunnel shell. Dynamic transducers are mounted in the wing by inserting the transducer (2.36 mm diameter) in the end of a long plastic sleeve, which is, in turn, inserted into a cylindrical channel molded into the interior of the wing. The sleeve terminates a t the center of the wing at the orifice communicating to tile wing surface. The lead wires are then routed out the opposite end of the sleeve in the wing (fig. 6) through the splitter plates and out through the tunnel walls to che data acquisition equipment in the tunnel control room. A single reference pressure tube from each dynamic transducer is also inserted into the plastic sleeve and routed thru~ght he splitter plate to the scanivalve- transducer assembly outside the tunnel. The transducer reference pressure can be selected to be the static pressure of the adjacent static orifice on the wing or any other selected pressure (such as the tunnel static pressure). Six accelerometers were mounted inside the wing, one at each of the attachment points of the four push-pull rods near the corners of the wing and two at mid- span near the leading and trailing edges. The actual motion of the wing can be cietsrmiced from the accelerometer output and compared with the output of t1.e motion transducers located in the actuator piston rods. These data sl.owed t'lat the wing motions were faithfully recorded by the motion transducers. Model sclpport system- The wing model, mounted between the splitter plates, is connected to the push-pull rods through ~.>eciafll exure bearings. The push- pull rods are, in turn, screwed directly into tho actuator pistons. Both the wing and the push-pull rods are fabricated from a lightweigh2 graphite-epoxy material. A short discussion of the fabrication of the rods and wings is given later in this section. The push-pull rods, 0.0412 m (1.625 in.) in diameter, are each capable of a 22,000 N (5,000 lb) tension load. The flex- ures located between the push-pull. rods and the wing are also desiqned for a 22,000 N (5,000 lb) load. A pair of graphite-epoxy rods mounted t 9 itle wing with a flexure support and attached to the splitter plates forward cf the wing provide a means of counteracting the drag loads (see fig. 5); each rqd . can withstand 6,700 N (1,500 Ib) Model fabrication- The fabrication of both the wing models and the pu~h- pull rods required an extensive development effort by the Ames Model Develop- ment Branch. The requirements for maximum strength, stiffness, and light weight suggested the w e of composite fiber materials. The problem of con- structing the wing was compounded by the requirenent for internal mounting of the pressure transducers. The following description w i l l illustrate briefly the steps used to fabricate the wing models.