1、NASA Technical Memorandum 85760Wind-Tunnel Investigation ofan Advanced General AviaJZJRREFERENCECanard Configuration R , .Joseph R. Chambem, Long P. Yip,and Thomas M. MoulAPRIL 1984Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS . . _ ._e-NASA Technical
2、Memorandum 85760Wind-Tunnel Investigation ofan Advanced General AviationCanard ConfigurationJoseph R. Chambers, Long P. Yip,and Thomas M. MoulLangley Resetwcb CenterHamton, VirginiaNASANationalAeronauticsandSpaceAdministrationScientific and TechnicalInformation Office1984Provided by IHSNot for Resal
3、e-,-,-SUMMARYWind-tunnel tests of a model of an advanced general aviation canard configura-tion were conducted in the Langley 30- by 60-Foot Tunnel. The objective of the testswas to determine the aerodynamic stability and control characteristicsof the config- uration for a large range of angles of a
4、ttack and sideslip for several powerconditions.For forward center-of-gravity locations, the model did not exhibit any stabilityand control characteristicswhich would be viewed as unsafe. The results also indi-cate that the configurationwould be extremely stall resistant. his highly desir-able stall-
5、resistancecharacteristic resulted from the fact that the canard wasdesigned to stall prior to the wing. Stalling of the canard resulted in increasedlongitudinalstability and decreased elevator effectiveness; both effects limited themaximum obtainable trim angle of attack to values below those requir
6、ed for wing stallfor all power conditions tested.For aft center-of-gravity locations and high-power, low-speed conditions, thecombined effects of nose-up trim changes due to power and reduced longitudinalsta-bility overpowered the stall resistance provided by the canard. Large nose-up ele-vator cont
7、rol inputs in this condition could result in stalling of the wing. Wingstall results in longitudinal instability and large nose-up moments which would tendto increase angle of attack to a high-angle-of-attack,deep-stall trim condition.The configurationhad insufficient elevator effectiveness for reco
8、very from the high-power deep-stall condition. F!otha reduction in power and use of nose-down elevatorwere required for recovery.Lateral-directionalstability and control characteristicswere degraded at wing-stall and post-stall angles of attack. In particular, the dihedral effect becameunstable at s
9、tall, large directional trim changes occurred at high power settings,and the rudder and aileron effectiveness became negligible at angles of attack asso-ciated with the deep-stall condition.The wind-tunnel results also indicate a marked reduction in longitudinal sta-bility at negative angles of atta
10、ck because of increased aerodynamic interferencebetween the canard and the wing. Although the elevator remained effective for thiscondition, the loss of longitudinal stability (particularlyfor aft center-of-gravitylocations)is undesirable.INTRODUCTION:-.Wind-tunnel tests of a l/3-scalemodel of an ad
11、vanced canard-configuredgeneralaviation airplane were conducted at the NASA Langley Research Center. An extensive? test program was accomplished for a large range of angles of attack, angles of side-slip, and power conditions. Flow-visualizationtests were also conducted to aid inthe interpretationan
12、d analysis of aerodynamic characteristics. The informationpresented herein is a summary of the more pertinent results and conclusions obtainedduring the tests.Provided by IHSNot for Resale-,-,-SYMBOLSAll longitudinalforces and moments are referred to the wind-axis system, andall lateral-directionalf
13、orces and moments are referred to the body-axis system.Moment data are presented for a forward centerfravity position of fuselage sta-tion 23.3 in. (-73percent of the reference mean aerodynamic chord) and for an “aft”center-of-gravityposition of fuselage station 24.8 in. (-63percent of the reference
14、mean aerodynamic chord). The center of gravity was located on the thrust axis toeliminate any moments due to the thrust moment arm. Dimensionalquantities are pre-sented in U.S. Customary Units. bCLCL,CCL, Wc1cPcmCNCnc$%EFcPFwq=s2wing span, ft .Liftconfiguration lift coefficient, qmslift coefficient
15、of canard, Canard liftq=sclift coefficient of wing, Wing liftqmsrolling-momentcoefficient, Rolling momentqmSbac=appitching-momentcoefficient, Pitching momentqmsenormal-force coefficient, Normal forceqmsyawing-momentcoefficient, Yawing momentqSbac$thrust coefficient, Thrustq=smean aerodynamic chord,
16、in.normal force of canard, lbfnormal force of propeller, lbfnormal force of wing, lbffree-stream dynamic pressure, lbf/ft2referencewing area, ft2,-.”4Provided by IHSNot for Resale-,-,-Sc exposed planform area of canard, f2a angle of attack, degP angle of sideslip, deg6 deflection angle of elevator,
17、positive for trailing edge down, degeAbbreviations:#BL butt line, in. C.g. center of gravityFS fuselage station, in.L.E. leading edgeWL water line, in.DESCRIPTION OF MODELA three-view sketch of the l/3-scalemodel is presented in figure 1, photographsof the model are shown in figure 2, and geometric
18、characteristicsof the model arelisted in table I. !Ihedesign incorporated a close-coupled, fixed canard and an aft-mounted wing of relatively low sweep. A single-slottedflap (referredto herein asthe elevator) on the canard provided pitch control, inboard wing-mounted aileronsprovided roll control, a
19、nd a conventional rudder provided yaw control.The model was constructed primarily of wood with a fiberglass outer skin. Powerfor the propeller was provided by a tip-turbine air motor driven by compressed air.Aerodynamiccharacteristicsof the complete model were measured with a conventionalsix-compone
20、nt strain-gage balance that was internally mounted. addition, auxil-iary balances were used to measure the individual aerodynamic contributions of thecanard and of the outer right wing panel. he canard spar and the carry-throughstructure were mounted directly to a strain-gage balance in the fuselage
21、 nose sec-tion. The right wing was constructed of separate inner and outer panels, and theouter panel was mounted to a strain-gage balance located within the inner wing-panelstructure. The gap between the inner and outer wing panels was sealed with flexibletape.The tests were conducted in the Langle
22、y 30- by 60-Foot Tunnel. AS shown infigure 3, the model and its internal strain-gage balances were mounted to a motorizedsting assembly which was remotely actuated to travel along a curved strut for varia-tions in the model angle of attack. The variations in angle of sideslip were pro-vided by a sec
23、ond remotely actuated motor which rotated the base of the curved strut? about a vertical axis. As shown in figure 3, compressed air for the air motor wasprovided by flexible plastic hoses, which trailed behind the sting assembly duringtests.The tests were conducted for a range of angles of attack of
24、 -28 to 92 and fora range of angles of sideslip of *I5S Besides longitudinal and lateral-directionalforce and moment tests, control effectiveness tests and component build-up tests (toidentify aerodynamic contributions of individual airframe components and aerodynamic3Provided by IHSNot for ResaleNo
25、 reproduction or networking permitted without license from IHSinterferenceeffects) were conducted. In addition, wool tufts were used in flow-visualizatioritests to define airflow characteristicsover the model.The test program was conducted at a wind-tunnel airspeed of 69 ft/see, whichresulted in a d
26、ynamic pressure of 5.6 lbf/ft2 and a Reynolds number of 0.55 x 106based on the mean aerodynamic chord of the wing. M view of the relatively low valueof test Reynolds number, the reader is cautioned that the aerodynamic characteristicsof a full-scaleairplane may be different than those of the present
27、 model because ofReynolds number effects. All aerodynamic data have been based on the geometric char-acteristics of the wing. bSTALL CHARACTERISTICSOF CANARD CONFIGURATIONS .he results of the wind-tunnel test indicate that the stability and controlcharacteristicsof the model were generally satisfact
28、oryfor the low angles of attackrepresentativeof cruise conditions. However, the stall and post-stall characteris-tics of the configurationvaried from highly desirable to undesirable, depending oncenter-of-gravitylocationand power condition. Prior to discussion of theseresults, a brief review of some
29、 fundamentalprinciples of design for satisfactorystall characteristicsof canard airplanes will provide background to aid in interpre-tation of the data and discussion.Shown in figure 4 are wind-tunnel data (ref. 1) measured in the Langley 30- by60-Foot Tunnel for a pusher canard-airplanedesign known
30、 to be very stall resistanton the basis of flight experience. In figure 4(a), the variations of lift coeffi-cient CL and pitching-momentcoefficient Cm with angle of attack a are pre-sented for the elevator fixed at a maximum nose-up deflection angle. The lift curveshows two distinct breaks. The firs
31、t break, which occurs near a = 11, resultedfrom stalling of the canard surface, which was designed to stall prior to the wing.The second lift break occurs near a = 24 and is indicative of wing stall.The inherent angle-of-attack-limitingcharacteristicof the foregoing stallsequence is illustrated by t
32、he pitching-momentdata. The configurationis longitudi-nally stable for angles of attack from 0 to 10, since the slope dCm/da is nega-tive. As expected, the maximum elevator deflectionproduces large nose-up valuesof cm at low angles of attack; however, as angle of attack is increased to 11, theprevio
33、usly mentioned canard stall is encountered, resulting in an incremental loss ofcanard lift with further increases in angle of attack. he stabilizing lift contri-bution of the unstalled wing then dominates, and therefore the configurationexperi-ences a marked increase in stability, as shown by the pr
34、onounced increase in negativeslope of Cm near a = 14“ in figure 4(a). The maximum obtainable trim angle ofattack is limited to about 16, well below the value of 24 required for wing stall.In addition to the increase in stability provided by canard stall, the phenome-non also results in decreased ele
35、vator effectiveness,since stalled flow also exists .)on the canardounted elevator. Therefore,as shown in figure 4(b), the elevatordeflection required for trim at high angles of attack increases significantly,and a71wing stall cannot be induced for maximum elevator input.The effectivenessof this high
36、ly desirable stall-resistancecharacteristicpro-vided by the canard configurationconcept can be influenced by many design variables,including airfoils and relative geometry of the canard and wing, propeller location,and center-of-gravitylocation. The effects of these variables must be accounted forin
37、 order to ensure that the wing cannot be stalled; in addition, the airplane must be4Provided by IHSNot for Resale-,-,-recoverable from excursions at high angles of attack generatedsuch as tail-slide maneuvers or zoom stalls to zero airspeed.RESULTS OF FLOW-VISUALIZATIONTESTSFlow-visualizationand for
38、ce tests made with the model ofby special maneuversthis investigationindicated that the configuration complied with the basic principle of canard airplanedesign in that the canard stalled before the wing. Results of wool tuft flow-visualization tests conducted to analyze stall behavior of the canard
39、 and wing sur-faces are presented in figure 5. The photographs, which were taken from a rear over-head position, illustrate the flow over the model for neutral controls with thepropeller windmilli.ng. The photographs are presented for a range of angles of attackfrom 0 to 28.For a = 0 (fig. 5(a),whic
40、h corresponds to cruise conditions, the flow wasattached over the canard and wing surfaces. When the angle of attack was increasedto 6 (fig. 5(b), flow separation occurred at the canard-fuselage juncture. Theseparated-flowregion increased in a spanwise direction for a = 10 (fig. 5(c).When the angle
41、of attack was increased to 12 (fig. 5(d), the flow over the leftcanard surface stalled abruptly, followed by a similar abrupt stall of the rightcanard surface at a = 14 (fig. 5(e). Also apparent at a = 14 was the onset oftrailing-edge“separationon the wing. At a = 16 (fig. 5(f), the wing trailing-ed
42、ge separation increased, and at a = 18 (fig. 5(g), the outer wing panels ofthe wing stalled abruptly. For a = 22 (fig. 5(h), the outer wing panels werestalled, as was the canard. The tufts indicated attached flow on the canard elevatoras a result of flow through the slotted elevator. me downwash fro
43、m the canardresulted in a significant reduction in local angle of attack on the inner wing panelsand, therefore, the flow on the inner wing panels remained attached up to high anglesof attack.Shown in figure 6 are photographs which illustrate the effects of power on stallpatterns at a = 28. Figure 6
44、(a) shows flow over the model for the windmillingpropeller condition, indicating stalled wing and canard surfaces with small areas ofattached flow on the inboard leading edge of the wing and slot flow over the canardelevator. !Iheeffects of power on the flow patterns are illustrated by conditionsfor
45、 a thrust coefficient CT of 0.4, which is a value that corresponds to a high-power,low-speedcondition. For CT = 0.4 (fig. 6(b), the slipstream of the trac-tor propeller significantlyaffected the flow over the right inboard canard and wingsurfaces. The previously noted separated flow at the canard-fu
46、selage juncture becameattached, and the attached flow area on the inner right wing panel was increased.The left canard and wing showed little effect of power, suggesting that the propellerslipstream swirl may have caused the asymmetry effects by decreasing the local angleof attack on the inboard rig
47、ht side of the model.Flow-visualization tests made for the elevator deflections other than 0 andanalysis of force and moment data indicated an effect of elevator angle on canardstall characteristics. For example, as elevator deflection was increased to themaximum value of 35, the stall angle of atta
48、ck of the canard decreased by about 5and the canard stall was more abrupt.As discussed subsequently, the relative angles of attack for onset of stall forthe canard and the wing, the relatively abrupt stall of both surfaces, and the effectof power on the stall progression all had significant effects
49、on the stall resistanceof the configuration.!3Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHSLONGITUDINALCHARACTERISTICSFOR FORWARD CENTER-OF-GRAVITYLOCATIONSThe overall static longitudinalstability and control characteristicsof themodel for forward center-of-gravitylocationswere satisfactory. k addition, theresults indicate a high degree of stall resistance, in ac
