1、J 2 - and from 4.0 x 106 to 8.0 x 106 forcruising. Tunnel power, balance limitations and model geometrylimited the Reynolds numbers of the WSU test to a rangebetween 2.2 x 106 and 2.9 x 106“ This is a reasonable rangefor development of the flap system. Tests at Reynolds numbersabove 3.0 x 106 to eva
2、luate flap nested high speed cruisingperformance were carried out by NASA in the low turbulencepressure tunnel at the Langley Research Center (Ref.3).Force Measurements and Comparisons With Theory29% c Flap Model. - Results of the lift and pitchingmoment measurements for the 29% flap model with comp
3、uterProvided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-designed flap slot geometry for various flap settings areshown in Figures 7 and 8, along with theoretical computerpredicted results. Agreement between experiment and theoryis quite good except for t
4、he flap nested and 40 flapdeflection cases. The discrepancy between theory andexperiment for the flap nested case is disturbing, in thatthe progressive loss of lift prior to stall indicatespremature boundary layer thickening, possibly with separation.It is to be expected that such a trend would resu
5、lt in highdrag. Flow visualization studies confirm this separation,but NASA tests reveal that separation is greatly delayed athigher Reynolds numbers (Ref. 3). Comparison of the resultsof the present tests with airfoils of similar thickness atcomparable Reynolds numbers (Ref. 12) reveals that non-li
6、nearity of the lift curve is the rule rather than theexception. Reducing Reynolds number from 6 x 106 to3 x 106 leads up to substantial non-linearity for cI valuesgreater than 0.4 for virtually all airfoils having thicknessto chord ratios of 15% or more. Since the ATLIT airplaneordinarily cruises at
7、 Reynolds numbers in the range of4 x 106 to 8 x 106 , this boundary layer thickening phenomenonat lower Reynolds numbers is not viewed as a serious short-coming.Comparisons of theoretical and experimental pitchingmoment data reveal the same trends observed with the liftdata, i.e. excellent agreement
8、 except for the flap nestedand 40 flap cases. The airfoil with flap nested has a sub-stantial zero lift pitching moment, as would be anticipatedfor a configuration with rather large camber near the trail-ing edge.Experimental drag data for the computer developed flapsettings are given in Figure 9. N
9、o comparisons betweentheoretical and experimental drag data are provided, sincethe computing routine in its present form does not have thecapability of drag prediction. (See Ref. 6 for a discussionof this limitation.)Results of force tests to determine slot geometry forhighest Clm _ with 35 and 40 f
10、lap deflections are shown inFigures I0,_i, and 12. These data show substantial improve-ments in Clmax compared to the computer developed slotgeometry. The changes in slot gap are quite small, however,Maximum lift coefficients for 35 and 40 flap deflectionsare essentially equal for this configuration
11、.Data for flap deflections of 50 and 60 are included inthese same figures. These runs with the higher deflections7Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-were made to provide information as to the feasibility ofgenerating additional drag with
12、 flaps for approach pathcontrol. For these settings, the flap was simply rotatedabout the wind tunnel flap fixture pivot point, withoutchanging the track settings from the optimum 40 flappositions. Consequently, the performance presented for 50 and 60 settings cannot be considered as optimum. A more
13、detailed discussion of optimization is given in a latersection of this report.Examination of the lift, drag and moment data shows thatdeflecting the flap from 35 to 40 results in very littlechange, while rotating the flap from 40 to 50 results in alarge drag change with essentially no change in lift
14、 or moment.Rotation of the flap from 50 to 60 results in a severeloss of Olmax as well as a large drag increase. Thus flaprotation 5etween 35 and 50 might be utilized to change dragwithout changing lift for airplane path control.30% c Flap Model. - Results of the force tests of thismodel are shown i
15、n Figures 13, 14, and 15 for flap deflectionsof 0 to 40 . The data for 35 and 40 flap deflectionsrepresent gap and overlap settings optimized for highest Clmax.The gap and overlap for the 20 through 30 flap deflectionswere selected as intermediate values to give a constant gapof 2.5%, and nearly lin
16、ear overlap adjustment.Data for flap deflections of 50 to 60 are shown inFigues 16, 17, and 18, along with the 35 and 40 settingsfor comparison. As before, the 50 and 60 data were obtainedby simple rotation from the 40 flap pivot position. Again,the gaps are not necessarily optimum for these cases.
17、Trendsare very similar to those observed for the 29% flap. Rotationsabove 40 result in a modest change in Clmax an_ pitchingmoment, with large changes in drag.Optimization of Flap Settings - ObjectivesConsiderable test time in the present program was devotedto optimization of flap settings. Determin
18、ation of any “optimum“must be related to airplane flight conditions. For a typicallight twin, the desired flap system performance characteristicsmay be identified as follows:Takeoff. - The requirement for takeoff is to attain asatisfactory Clmax at an angle of attack within the landinggear capabilit
19、y of the configuration, i.e. an angle of attackthat can be achieved by rotation about the main gear withoutaft fuselage contact with the runway.8Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-Climb. - For twin-engine aircraft, single engine rate ofc
20、limb _-i-s-_rdinarily a crucial performance parameter. It isdesirable to have the maximum airplane lift-drag ratio occurat climb lift coefficient or higher in order to avoid thedifficulties associated with flight operations in the “regionof reversed command“ or “back side of the power curve“.While t
21、he maximum airplane lift-drag ratio must, of course,be evaluated including fuselage and nacelle drag as well asthe three-dimensional wing induced drag, it is imperativethat the airfoil section have the minimum possible drag atthe climb lift coefficient.Another consideration for total airplane climb
22、performanceis the adverse influence of angle of attack on fuselage andnacelle drag. To minimize fuselage and nacelle drag it isdesirable to operate at near zero fuselage incidence. Thus thedesired airfoil characteristics for climb performance may besummarized as follows:a) attainment of the lowest p
23、ossible sectioncd value (or highest possible i/d value) atbest climb c I.b) attainment of best climb c I at an angle nearthe cruising angle (zero d_grees in the present case).Landing. - Low approach speeds are required for shortlandings. This requirement dictates high maximum liftcoefficient. Fairly
24、 high drag levels are permissible inthis flight regime.Summary of Objectives. - The desired performance goalsoutlined above may be summarized into two optimization objectivesas follows:(I) attainment of a high value for Clmax.(2) attainment of maximum i/d for agiven cI .Optimization of Flap Settings
25、 - ResultsMaximum Lift Coefficient. - Results of the optimizationtests-ZY6_ C. ma x are presented in Figure 19 for the 30% c model.These data _how the locus of slot gap and flap-airfoil overlapfor constant values of Clmax, for flap deflections of 35 and40 . These figures show a Clma x value of 3.8 f
26、or 40 flapdeflection with 2.7 % gap and a negative 0.7 % overlap. (FlapProvided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-nose aft of wing trailing edge.) The attainment of a Clmaxvalue of 3.8 is considered to be a significant achievementfor a single-sl
27、otted Fowler flap at these low Reynolds numberswithout leading edge devices and without blowing or suctionboundary layer control. While it is recognized that optimumslot geometry is influenced to some degree by flap and air-foil contours, some generalization of these data should bepossible. In fac_t
28、he optimum slot gap and overlap contourspresented here are quite similar to those presented inReference 13 based upon tests of a Clark Y airfoil conductedin 1932, even though the maximum cI values from the presenttests are substantially higher.Maximum (l/d) for a Given cl. - The optimums for (i/d)ma
29、xwere determined from consolidated plots of cd versus cl forall flap settings. (Figs. 9, 12, 15, and 18). The envelopesof these curves represent the desired optimums. Cross-plotsof these envelopes are presented in Figures 20 and 21. Flapdeflection, gap and overlap required to achieve optimumperforma
30、nce are also presented.Runs were made for several overlap and gap positionswith flap deflections of 5 and i0 , in an attempt to identifya configuration which would substantially reduce drag in theclimb condition (cI = 1.0) to a level below the flap nestedvalue. This search was unsuccessful in that n
31、o deflectedflap setting Could be found which reduced the drag at climbc 1. These tests did provide, however, a slot configurationwhich results in minimum drag for the i0 flap deflection.Furthermore, total airplane considerations, such as fuselageand nacelle drag and the sensitivity of these items to
32、 angleof attack may lead to a situation in which some modest flapdeflection such as 5 or i0 will provide minimum total air-plane drag at the climb condition. Thus the optimum air_laneconfiguration during climb may be with some flap deflection,even though the optimum isolated airfoil configuration wo
33、uldbe with flap nested. It should be noted that detailed opti-mizations of slot gap were carried out only for i0 , 35 , and40 flap deflections. Therefore, slightly better performancemight be found for the intermediate flap deflections ifadditional slot gap and overlap variations were studied.However
34、, the data presented here represent achievable per-formance for the settings specified, and are probably nearthe “absolute“ optimum in every case.It can be argued that the gaps for highest Clmax with35“ and 40 flap deflections are not necessarily the gapsettings for maximum (l/d). Examination of the
35、 consolidatedc d vs. c I plots reveals that the c I range for which any flapi0Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-deflection produces minimum drag is fairly narrow. Further-more, the 35 and 40 flap settings provide minimum cd onlyfor c I
36、values very close to Clma x Thus any change in slotgeometry which reduces Clma x will almost certainly result inhigher cd values. These arguments are also consistent withthe premise that the function of the slot is to reduceseparation, and that reducing separation will reduce dragand increase Clmax.
37、Comparative Performance of the 29% and 30% Flap ModelsComparison of Figures 20 and 21 illustrates that the twoflap configurations are nearly identical in (i/d)ma x performance.The 30% c flap with slightly larger effective chord, provides aslightly higher Clmax. The performance differences betweenthe
38、 two configurations are so small that a choice between themshould probably be dictated by non-aerodynamic factors such asstrength, stiffness or ease of manufacturing. Close examinationof the main airfoil trailing edge geometry indicates that the30% c flap model with the finite trailing edge thicknes
39、s mainairfoil is probably the easier to manufacture.Pressure DistributionsFlap Nested. - Pressure coefficient data for the GA(W)-Iairfoil, flap nested are shown in Figures 22 and 23, for 0and 6 angles of attack. Computer generated theoretical dataare also shown for these cases. Agreement between the
40、ory andexperiment is excellent. The kink in the lower surface Cpdata at the 70% chord location is a result of the lower surfacenotch on the model in the flap nested configuration. Thisnotch was not represented on the computer program. The 0angle of attack case illustrates the pressure distributionat
41、 design cruise lift coefficient, characterized by a flatregion over the forward portion and a concave region ofpressure rise over the aft portion of the airfoil.The Cp distributions at higher angles (Fig. 24) showsubstantial flat regions over the aft portion indicating theeffects of upper surface se
42、paration. Separation locationsare marked, based upon flow visualization studies. Notheoretical curves are available for these cases, since thecomputing routine is not presently capable of predictingpressure distributions for cases with partial separation.Computer integrations of the Cp data provide
43、cI and cmvalues for comparison with the direct force measurements.iiProvided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-A comparison of this type is shown in Figure 25, and theagreement is seen to be quite good. Similar agreementexists for flap extended
44、cases, although no direct com-parisons of this type are shown.Flap Extended. - Flap extended pressure distributionsare given in Figures 26 through 43. Integrated normal forcecoefficient values for flap and main airfoil are tabulatedfor convenience. This series is for the 29% c flap, andthe flap posi
45、tions shown here are the “computer designsettings“. These settings were established in the preliminarydevelopment phase of the project based upon computer designstudies. For angles of attack below separation, theoreticalCp values are also shown for comparison. In general, thecomputer program provide
46、s an excellent modeling of the flow.The pressure distributions over the forward portion of theairfoil and over the entire flap consistently show onlyminute differences between experiment and theory. Significantdifferences are present on both upper and lower surfaces nearthe trailing edge of the main
47、 airfoil. Evidently some improve-ment in modeling of the slot flow and its influence on uppersurface pressures is required. Nevertheless, the high orderof agreement shown here between theory and experiment establishesthe credibility of the computer program as an airfoil/flapdesign tool.An unusual pr
48、essure distribution is shown by the datafor 40 flap deflection and e = i0 in Figure 41. Pressuresalong the upper surface show a high degree of scatter forthis case. The implication is that a “long“ laminar separationbubble is present and that the reattachment point is not fixed,resulting in wide var
49、iations in the surface suction developed.At lower angles, the pressure distributions are quite stable.At 15 angle of attack (Fig. 42) the pressure distribution isagain quite stable. In this case, an extensive “flat“ Cpregion appears from about 40% chord aft. This is indicativeof a turbulent separation in this region. Similar
