1、TECHNICAL NOTESNATIONAL ADVISORY COMMITTEE FOR AERONAUTICS=WIND-TUNNEL INVESTI_:T_O_ OF GROUND EFFECTON WINGS WITS FLAPSBy Isldore G. RecantLangley Memorial Aeronautical LaboratoryFILE COPyTebt_ _l_ fj_ ofh I_onat,d_r7 Com_t_Wash !ngt o nMay 19._9, _ duced -“ : : ;. NATIONAL TECHNICALINFORMATION SER
2、VICE ;_:; _:_ _ _ i _ _ i _Springfield, Ve. 2215l _ F._G “Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-NATIONAL ADVISORY COMMITTEE FOR AERONAUTICSTEC
3、HNICAL NOTE NO, 705WIND-TUNNEL INVESTIGATION OF GROUND EFFECTON WINGS WITH FLAPSBy Isidore G. RecantSUMMARYAn investigation was conducted in the N.A.C.A. 7- bylO-foot wind tunnel to determine the effect of ground prox-imity on the aerodynamic characteristics of wings equippedwith hlgh-lift devices.
4、A rectangular and a tapered wingwere tested without flaps, with a split flap, and with aslotted flap. The ground was represented by a flat plate,completely spanning the tunnel and extending a considerabledistance ahead and back of the model. The position of theplate was varied from one-half to three
5、 chord lengths belowthe wing.The results are presented in the form of curves ofabsolute coefficients, showing the effect of the ground oneach wing arrangement. The effect of the ground on llft,drag, and pitching moment is discussed. An appendix givesequations for calculating tunnel-wall corrections
6、to beapplied to ground-effect tests conducted in rectangulartunnels when a plate is used to represent the ground.The tests indicated that the ground effect on wingswith flaps is a marked decrease in drag, a decrease indiving moment, and a substantial reduction in maximum lift.INTRODUCTIONThe phenome
7、non commonly called “ground effect,“ where-by the aerodynamic characteristics of plain wings undergomarked changes in the presence of the _round, has been sub-ject to considerable investigation (references 1 and 2).Both theory (reference _) and experiment (references 4, 8,and 6) indicate that the pr
8、oximity of the ground decreasesthe drag and increases the slope of the llft curve in thesame manner as an increase in aspect ratio would affect theProvided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-2 N.A.C.A. Technical Note No. 705same factors. In fact,
9、 it is customary to predict the ef-fect of the ground on the basis of an apparent increase inaspect ratio (references _ and 5). Although a wing oper-ating in the vicinity of the ground is subjected to an in-crease in lift over the free-air value at any given angleof attack, it does not necessarily f
10、ollow that the maxi-mum lift coefficient is increased. Available evidence in-dicates that, for wing heights decreasing from infinityto one-half chord length, the maximum lift is unaffected(references _ and 5) or slightly decreased (references 4,7, and 8).Much less study has been devoted to the effec
11、t of theground on airfoils equipped with lift-lncreasing devices.Viaud (reference 9) found the customary increase in theslope of the lift curve and decrease in the drag for wingswith various types of flap. The maximum lift for splitand plain trailing-edge flaps increased as the wing ap-proached the
12、ground. For the slotted flap, the maximumlift remained nearly constant; whereas, for the multipleslotted flap, the maximum lift decreased considerably asthe ground was approached. For split flaps of the Zaptype, SerebriJsky (reference 10) found a decrease in themaximum lift as the ground was approac
13、hed.Since almost all present-day airplanes are providedwith flaps of one type or another and since the flap ef_fect is of particular importance in the immediate vicinityof the ground, the necessity for further study of the prob-lem is obvious.The present investigation was made in the N.A.C.A. 7-by 1
14、0-foot wind tunnel to study the effect of a simulatedground area on a rectangular and a tapered wing, eachequipped successively with full-span split and slottedflaps. It may be pointed out that the tests were run at acomparatively small scale, and the method of groun_ simu-lation is not exactly repr
15、esentative of actual flight con-ditions. Nevertheless, the results are believed to be in-dicative of the comparative effects on the various devices,but flight tests are required to determine the applicabil-ity of the wlnd-tunnel results.Provided by IHSNot for ResaleNo reproduction or networking perm
16、itted without license from IHS-,-,-7fN.A.C.A. Technical Note No. 705 3APPARATUS AND TESTSModelsThe wing models used have the N.A.C.A. 2_012 profileand are made of laminated mahogany. They have a span of60 inches, a geometric aspect ratio of 6, and an averagechord of l0 inches (fig. 1). They had been
17、 used in a pre-vious investigation (reference ll) and were available forthe present tests. “The tapered wings (fig. 2) have a root chord of 16.67inches and are tapered 5:1. The maximum ordlnates of allsections on the upper surface are in a horizontal planeand the plan form is symmetrical about a lln
18、e perpendicu-lar to the root chord at its .50_percent point.The split-flap models are shown in figures l(b) and2(b). The flaps are full span; their chords are 20 per-cent of the wing chord; and they are located at 80 per-cent of the wing chord. The flaps were set at 60 , whichis the deflection neces
19、sary for maximum lift. On the ta-pered wing the flap also has a taper of 5:1.The slotted-flap models are shown in figures l(c) and2(c). This flap is designated 2-h in reference 12, whichgives the slot shape, the flap profile, and the path ofthe flap nose for various deflections. The flap chord is25.
20、56 percent of the wing chord and the deflection is 40 ,which is the angle necessar_ for maximum lift. On thetapered wing the flap also h_s a taper of 5:1. _-Wind TunnelThe tests were ma_e in the N.A.CoA. 7- by lO-footclosed-throat wind tunnel described in reference 12. Themodel was mounted on the re
21、gular 6-component balance (ref-erence l_) that measures the aerodynamic forces and momentsindependently and slmultane_usly with respect to the windaxes of the model.Ground RepresentationThe most common methods of ground representation arethe flat-plate and the reflection methods. These methodsare co
22、mpared by Raymond (reference 6) and Cowley and LockProvided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-4 N.A.C.A. Technical Note No. 705(reference 14); both references show a discrepancy betweenthe two methods. Cowley and Lock impute the discrepancyto a
23、_hift in the angle of zero llft that is due to thedeflection of the air stream by the plate, but Raymondstests also show a shift in zero lift with the reflectionmethod.In the present tests the ground was simulated by aflat plate. The most obvious objection to the plate isthe fact that the air moves
24、with respect to it, creatinga boundary layer; such a condition does not exist in actualflight. A survey of the boundary layer over the plate, how-ever, showed the maximum thickness under the trailing edgeof the wing to be about 1 inch. Since the models were al-ways at least 2 inches from the plate,
25、it is unlikely thatthe results were greatly affected. The present tests,moreover, are comparative and, under these circumstances,the plate method is considered valid.The plate is made of 3/8-inch plywood, is 7 feet longand i0 feet wide, completely spanning the tunnel width.The leading edge of the pl
26、ate has a faired nosepiece 4inches long and 2 inches thick. The plate was fastened toa steel frame; vertical steel rods passed through the plateand the frame at each corner and were rigidly attached tothe tunnel floor and roof. The plate was free to slide onthe rods in order to vary the distance fro
27、m it to the wing,which was mounted on the tunnel center llne. The platewas held in any desired position by set screws that clampedit to the vertical rods. Two vertical rods under the plateat its longitudinal center llne kept it from sagging. Themodel was mounted about four chord lengths back of thel
28、eading edge of the plate. Figure 3 shows the plate andthe method o_ mounting it in the tunnel.TestsDynamic-pressure surveys at the location of the modelwere made for each positron of the ground plate. The dy-namic pressure was maintained constant throughout thetests at 16.,_7 pounds per square foot,
29、 corresponding to anair speed of 80 miles per hour at standard sea-level con-ditions. The average test Reynolds Number was 609,000based on a mean wing chord of l0 inches. The effectiveReynolds Number due to the turbulence of the tunnel wasapproximately 974,000. A survey of the boundary layer oveTthe
30、 plate at the trailing e_ge of the wing was made.Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-N.A.C.A. Technical Note No. 705 5The rectangular and the tapered wings were testedplain and then successively with full-span split flaps andfull-span slo
31、tted flaps. Each arrangement was tested inthe clear tunnel and then with the ground plate a:t one-half, one, two, and three chord lengths below the wing.Distances were measured from the quarter-chord point of themodel to the ground plate. Lift, drag, and pitching mo-ments were measured for an angle-
32、of-attack range from -6 to the stall in 2 increments.RESULTS AND DISCUSSIONCoefficients and SymbolsThe results are given in absolute nondimenslonal coef-ficient form.CL, lift coefficientCD ,Cm(a.c.)owhere(L/qS).drag coefficient (D/qS).pitching-moment coefficient about aerodynamiccenter of plain wing
33、 (M(a.c.)o/qCw S )“L is lift.D, drag.M(a.C.)o pitching moment about aerodynamic center ofplain wing.S, wing area.C w , mean geometric chord of airfoil with flap fullyretracted.q, dynamic pressure ( p V S).and8f,h,is angle of attack.flap deflection.distance of quarter-chord point from ground.Provided
34、 by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-6 N.A.C.A. Technical Note.No. 705Ground DistanceThe distance of. the wing from the ground is expressedas a ratio h/Cw, where h is the distance from theground to the quarter-chord :polnt of the wing. The choiceo
35、f the referenBe point from which to measure the grounddistance is somewhat arbitrary and varies with differentinvestigators. The nose of the wing was chosen in refer-ence 9; the quarter-chord point was used in references 6and 7; the half-chord point was used in references 3 and8; and D_twyler, whose
36、 work is summarized in references 1and 2, chose the trailing edge. Regardless of the pointselected, the ground distance will evidently vary as theangle of attack is changed unless the wing is rotated aboutthat point in changing the angle. For long ground distances,the choice of reference point is no
37、t likely to make any ap-preciable difference in theresults. For short ground dis-tances, however, different results may be expected foreach reference point chosen. Since the lift caused by thechange in the angle of attack acts at approximately thequarter-chord point and the lift that is due to the w
38、ingcurvature acts at about the half-chord point (reference15), the quarter-chord point seems convenient as a refer-ence. When the quarter-chord point is used as a referencepoint, the ground distance to the point of action of thelift that is due to the angle of attack will not changeand the ground di
39、stance to the point of action of the liftthat is due to the curvature will change only slightlywith a change in the angle of attack. No substantiatedtheory indicates any one reference point to be preferable,but it is well to keep in mind the reference points usedwhen test results are compared.Wind-T
40、unnel CorrectionsThe tests run without the ground plate in the tunnelwere corrected for tunnel effect to aspect ratio 6 in freeair. The normal Jet-boundary corrections were applied(reference ll). The tests run with the ground plate inthe v_rious positions below the wing were not correctedbecause the
41、 tunnel-wall interference factors calculatedfor these conditions were small enough to be disregarded.The method of calculating the interference factors for theground-board installation is given in the appendix.Provided by IHSNot for ResaleNo reproduction or networking permitted without license from
42、IHS-,-,-N,A.C.A. Technical Note No. 705 7PrecisionExperimental errors in the results presented in thisreport are :believed to be Within the following limits:. 0.i CL 0,001CL 0.005maxCD 0.001. = 0o ooo5 (_f )CDmi n0,001Cm(a. c. )o8f I ,0h 1/16 inchThe _round plate was parallel to the tunnel axis with
43、in0.i uNo tests were made to determine the effect of theflap fittings. Because the tests are comparative, the ef_fect of the fittings would probably not materially changethe results,Aerodynamic Effects of the Proximity of the GroundLift-curve_!_R2.- The effect of the ground on theslope of the llft c
44、urve is shown in figures 4 to 9, Forthe plain wings (figs. 4 and 5) the slope increases asthe ground distance decreases. In general, this effect isin agreement with the Wieselsberger theory (reference _).The increase, especially _or the shorter ground distances,is numerically greater than predicted
45、by the theory. Thisdiscrepancy may be due to the fact that the theory is basedonly on the effect of the trailing-vortex system and neg-lects the effect of streamline curvature due to the ground.(See references 7 and 8.) This effect will tend to in-crease the slope further. Reference _ gilves theoret
46、icalequations that indicate an increase in the: angle of zerolift as the :ground :iS_approached; this increase is due tothe thickness of thewing_ The present test_ _how no suchshift, _“ “ Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-8 N.A.C.A. Tec
47、hnical Note No. 705For the wings Nith the slotted and the split flaps(figs. 6 to 9), the lift curves in general appear to beparallel but displaced from each other as the ground dis-tance decreases. These tests, however, were not run belowa lift coefficient of 0.6; and, if no shift in the angleof zer
48、o lift (as is indicated by the plain-wing tests) isassumed, the slopes below CL = 0.6 would increase with adecrease in ground distance.DEag.- The effect of the ground on drag is shown infigures 4 to 9. As indicated by theory, the drag for allthe wing arrangements was substantially reduced as theground distance decreased. The reduction in drag of theplain wings is somewhat greater than the theoretical esti-mate that is based solely on an apparent increase in theaspect ratio; whereas, the reduction for the wings withflaps averages 40 percent greater than Wieselsbergerstheory indicates. The re
