1、RESEARCH MEMORANDUM WIND-TUNNEL INVESTIGATION OF STALL CONTROL BY SUCTION THROUGK A POROUS LEADING EDGE ON A 37O SWEPTBACK WING OF ASPECT RATIO 6 AT REYNOLDS NUMBERS FROM 2.50 X 106 TO 8.10 X lo6 a By Robert R. Graham and William A. Jacques NATIONAL ADVISORY COMMITTEE FOR AERONAUTICS WASHINGTON Marc
2、h 11, 1953 Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-B NACA RJI L52L05 .-. NATIONAL ADVISORY COMMITTEE FOR ABRONAUTICS CONTROL BY SUCTION T:-IROUGH A POROUS UADINC- EDGE ON A 37 SiLWTBACK WING OF ASPECT RKTIO 6 AT R3Y!XOL;DS NU.B3ERS FROM 2.50
3、x lo6 TO 8.10 x lo6 By Robert R. Graham and Willtam A. Jacques The effects of suction through a porous leading-edge surface have Seen investigated in the Tangley 19-2oot pressure tunnel on a wing having of 0.5, and NACA 641-212 airfoil sections normal to the 27-percent-chord lice. Tae effects of ver
4、ylng the chordwise and spanwise extent of porous area were investigated on the wing without trailing-edge flags and the effects of one chorcwise and sganw5,se extent of porous area were iaves- tigated 011 Yae wing vith half-sgan split acd double slotted flcps. The tests covered e rage of Reynolds nm
5、ber from 2.50 x 106 to 8.10 x 106 and a range of Mach number from 0.08 to 0.26. k 37O sweepback or“ the leasng edge, an asqect ratio of 6, taper ratio L The results indicate that at Mach numbers of the order of 0.12 the outboard stall of the wiEg calz be delayed amd nose-down moments at maxi- mum li
6、ft can be produced about 8s effectively by bou-n the tip stsll by means of auxiliary devices sxch as leadlng-edge flaps, slats, or droop nose. (See, for instance, refs. 1 to 3. ) Nore recently, attention has been directed toward the possibility that stability at the stall might be obtained just as e
7、ffectively by mans of boundary-layer control. Sane data are available which demonstrate that longitudinal stability at the stall can be improved on sweptback wings by means of suction through leading-edge slots or porous area (refs. 4 to 6). An agpraisal of leading-edge suction as a stall-control Ce
8、vice on sweptback wings, however, can be =de ow if its effects can be directly cqared with the effects of auxiliary devices on the same wing. In order to make this comparison an for local sonic velocity xJ - Po Q% c; - Q P S S b - C C Y A 5 + MCOSA locel Mach number, 1/3.5 -5 (0.7-m2 + I) allgle of
9、attack of root chord, deg duct total pressure inside porous leading edge volme flow, at free-stream densfty, through porous surface locel stetic pressure total vipg area wing area aTfected by suction (See teble I) Xing span locel wing chcrd parallel to plane of syrmetry lateral coordinzte sveep of l
10、ea6hg edge Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-4 PO Jk VO PO P a free-stream static pressure free-stream total pressure free-stream velocity free-streau air density coefficient of viscosity speed of sound MODEL APSD APPARATUS The model us
11、ed in this investigation was a semispan wing mounted in the presence of a reflection plane as shm in figure 1. A photo- graph of the model. and reflection plane mounted in the tunnel is pre- sented as figure 2. Except for the modified leading edge, the wing was the sane one described in referencd 1.
12、 It hd an aspect ratio of 6, a - taper retio of 0.5, and 37.25O sweegback of the leading edge. The air- foil sections were of NACA 641-212 profile perpendicular to the 27-percent- chord line. The general plm form and some of the principal dinensions of the model are given in figure 3. For several te
13、sts the node1 was fitted with 0.50b/2 split flaps, 0.50b/2 double slotted flaps, and a fence at the 0.50b/2 station, details of which are presented in figure 4. The leading edge of the upper surface was constructed from a lam- inated skin attached to solid ribs. Two skins were tested, both of which
14、coosisted of 1/16-inch gerforated plate covered with a layer of 14 x 18 mesh bronze screen and an outer surface or 30 x 250 mesh, Dutch weave, Monel filter cloth. The filter cloth was rolled from its original thick- ness of 0.026 inch to O.Cl8 inch for one of the skins and to 0.016 inch for the othe
15、r skin to obtain the desired values of porosities and a smooth surface of tne skin. The porosity characteristics of the two skins as Installed on the mosel are shown in figure 5. The porosity of the skin with 0.018-inch filter cloth is designated as porosity A and that for the skin with 0.016-inch f
16、ilter cloth is designated as pqrosity B. A tMrd porosity was inadvertently tested in tne beginning of the test program when the porosity of the 0.016-inch filter cloth was reduced by the cor- rosive action of soldering flux whFch had been used only along the edge of the skin in the fabrication proce
17、ss but which apparently penetrated the entire ares of the skin by capillary action. This porosity is des- ignated as gorosity C and was used for only a few tests before the skin Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-NACA FiM L52LO5 5 - was
18、cleaned by means of hydrochloric acid, xater, and steam to increase the porosity to that desigpated as porosity 3. The 0.018-ih filter cloth was ceEented in place; hence, no corrosion problezn occilrred. c The solid ribs wbAch supgorted the porous skin dtvi-ded the leaGing edge into eight conpartmen
19、ts the dimensions of which are shorn in fig- ure 2. Each compartment was connected. to the main suction duct through en indiddud flow-aeasuring venturi aaCi flaw-control gate valve. Flow into the leadlng edge of the hing was obtained by connectirg the suctioo duct to the outside of the tunael when t
20、he air in the tunnel was compressed to about 21 atmospheres or to high-capaity vacuun pmps when the air in the tunnel was at atmospheric gressure. 3 The extent of the porous area was controlled by spraying the leading edge with a layer of nonporous stripgable plastfc and a layer of lczcquer sanded s
21、mooth End then strlpsing off only the area which wes to be porous. The porosity of the skin was snaintained by .oassing a cleaning agert such as acetone or cerbon tetrachloride through tk PO- yous area. - The leading edge of the -+ring was equipped fi-th surface orifices at 0, O.OOlc, O.OO3cY and 0.
22、005 at the spanwise midpoint of each conpert- ment to measure the De. Apparently increasing the Mach number above 0.14 offsets the effects of tne corresponding increase in Reynolds nmber above 3.46 x 10 6 . Canparison of. figures 7 and 10 shows that at a Reynolds number of abodt 4.40 x 106 the ouyoo
23、ard stall occws at an angle of attack of 17.30 at a Mach number of 0.08 and E.20 at 0.18. A corresponding reduction in the angle of attack for the outboard stall was brought about at a Reynolds nmber of about 5.40 x 106 when the Mach number was increased from 0.10 to 0.22 and at about 6.50 x 106 whe
24、n the Mach number w suction flow rate.- The data of figure 16 show tiit, at a Reynolds nlxLli)er of 6.80 x 106, reducing the flow rate from the maxjmum obtained with 0.015 chordvise extent and 50-percent span- wise extent of suction reduced the maxhwn lift coefficient frm 1.33 for a CQ of 0.00052 to
25、 1.19 for a CQ 05 o.aoo18. The pitching moments were stable at the stall for values of CQ of 0.00026 or greater but the unstable trend below Ck was more severe at the lower flow rates. Tests at higher flow retes and reduced Reynolds numbers indicate that increasing the flow coefficient did not compl
26、etely eliminate the unstable -trend prior to Ch. Thus the tests indicate that leading-edge suction delayed leading-edge separation, with the result t-hat considerable hprove- nent io stebility was obtained at c They also indicate that leeding- edge suction did not eliTd.nate trailing-edge segaration
27、 but did delay its Lma.x- Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-spread toward the leading edge, witn the result that the unstable trend below Ch was less severe with suction. It is possible that e mld- chord suction slot or suction area ope
28、ratfng in conjunction with leading- edge suction (similar to the two-dinensfond Errangenent in ref. 9) might delay the trailing-edge separetion so that tne pitching-mcment curve would be linear us to C- hax The effects of varying suction rates oz?. the leading-edge pressures are sham in figure 17. T
29、he ?nininum flow coeffictent tested (CQ = 0.00018) delayed the outboard leaafng-edge stall slightly, as shmn by a campar- ison of tle data of figure 17 with the data of figure 7 obtained at the seme Reynolds nunber with leadlng edge sealed. An increase in C was obteined with the minimum flax coeffic
30、ient (see fig. 161, but that amount of suctio9 did not maintei-n- enough lift over the outboard portion of the wing to cause any improvement in the pitching-mment characteristics. Increasing the suctioc rate to a CQ of 0.00040 delayed the stall to 8, higher -le of attack (a = 19.5O) end caused an in
31、crease in the outboard lift beyond tke stall which considerably bproved the pitchir-g-momert characteristics at . The lift that was maintained over the outboard sectiolls after the stall occurred w1Etnge occurred io the pitching-moment curve vas about constant st I .26 through the Remolds number rwe
32、 from 4.36 x 106 to 8.10 x 106. Below a CL of 1.26 the _Ditching-mment characteristics were similar throughout that Reylzolds nmber range. One test was nede at a Reynolds omber of 3.46 x 106 aad s CQ of 0.00036. Uoder these conditions the pitclzi-%-moment characteristics were shilar to those severe
33、change in the pitckdng-mmezzt curve occurred at a lower lift coefficiezt (1.21). . obtained e.t higher Reynolds nmkers with CQ = 0.00040 except tht the - Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-12 NACA RM L52L0-5 2eference 6 indicates that wi
34、th varying Reynolds number, aynamic shiiarity in the boundary layer will not oe obtained witn suction unless the product CQR/ is held co,n_stant. Figure 19 shows that the pitching- molnent characteristics and the values of C were essentially the sme through the Reynolds number range from 4.36 X 106
35、to 8.10 x 106 wlren CQRI/ was held constant or when CQ alone was held constant (see fig. 18). Wli differences in the pitching-noment characteristics are probably dGe to scale effects on the stalling characteristics of the sections inboard of the gorous part of the leading edge. LmaX The spanwise plo
36、ts of leading-edge pressure coefficients (figs. 20 and 21) show ollly slight variations in distribution over the outboard portion of the ving with suction as the Reynolds number WES varied whether the pro6uct CQ31/2 was held constant at approximately 0.00040 48.10 X 106 or CQ WES held constant at ap
37、proximatelg O.OOOk0. In either case, how- ever, when the stall occurred it covered a larger inboard (no suction) portion of the ving at the low Iieynolds number than at the high Reynolds nmbers . - Effects of leading-edge suction on wing characteristics at critical speeds.- Same of the effects of su
38、ction on the wing were determined at speeds at vUch critical or superzriticai oressure coefficients were xeasurec at the leading edge of tlie model. The results are not conclusive because only small flow coefficients were obtainable at those velocities. The results (figs. 8 end 22) indicate, however
39、, that the small flow rate used at M = 0.26 h that is, sone drag reductions were effected in the high lift rawe by delaying separation and the wing drag with suction was less ti the drag with leading-eiige flep (ref. 1). When the equivalent omp-power drag coefTicient IQ duct, pms, and exit losses) f
40、or porous-leaang-edge suction %ere calculates as follows: A wing loading of 50 pounds per square Toot and standard sea-level air density of 0.002378 slug per cubic foot were assurned in calculz%ing s, and Vo. The wicg area used was 306.1 squere feet, wnich correspouds to tpzt of a present-day fighte
41、r aircr the _nitching-moment data for the double slotted. flap configuration indicated nose-up xmments et Ch but nose-down nomects as the lift decreased in tze stall. At Mach numbers above 0.2, stelling occurred over the tip sections of the sealed-leading-eQe wing as locel sonic velocities were appr
42、o ched. Application of the highest suction-flow rates available /C = 0. OCXd) delayed the ti3 stall until local velocities of the order of Mz = 1.20 were attained but this delay in tip stalling was not sufficient to provide nose- of Aspect Ratio 8 With High- Lift and Stall-Control Devices at Reynold
43、s Numbers Frm 1,500,000 to 4,800,00. NACA RM L51JO4, 1952. 4. Cook, Woodrm- L., and Kelly, Hark W. : The Use of Area Suction for the mose of Celeying Separation of Air Flow at the Lea6ing Edge of a 630 Swept-Back Wing - Zffects of Controlling the Chordwise Dis- tributiors of Suction-Air Velocities.
44、NACA RM A51J24, 1952. 5. Pasmanick, Jerme: and Scallion, William I.: The Wfects of Suction Througn Porous Leading-Edge Surfaces on the Aerodynamic Character- istics of a 47S0 Ehnreptback Wing-Fuselage Conbillation at a Reynolds Number of 4.4 x 106. KACA R4 ElKl.5, 1952. 6. ?opgleton, E. D.: Wind Tun
45、nel Tests on a Swept Back Wing Y!hving Dis- tributed Saction on the Leading Edge. TN No. Aero 2081, British R.A.E., Nov. 1950. 7. Sivells, James C., end Deters, CMen J.: Jet-Boundary end Plan-Form Corrections for Fartial-Span Models With Reflection Plane, End Plate, or No End Plate in e. Closed Circ
46、ular Wind Tunnel. NACA Rep. 843, 1946. (Sugersedes NACA TN 1077. 8. Eisenstact, Bertram J.: Boundary-Induced Upwash for Yawed and Swept- Eack Wings in Closed Circular Kind Tmels . NACA TW 1265, 1947. 9. McC-dlough: George B., and Gault, Donald E.: An ExperinentEl Investi- gation of the NACA 631-12 A
47、irfoil Section With Leading-EQe and Midchord Suctiozl Slots. NACA TW 2041, 1950. Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-NACA 34 L52L05 - 19 10. Pratt, George L.: Effects of Twist and Camber on the Low-Speed Longitudinal Stability Characteris
48、tics of e 45O Sweptback Wing of Aspect Ratio 8 at Reynolds Nmbers From 1.5 X 106 to 4.8 X lo5 As Detemneci by lressure Distributions, Force Tests, md Calculations. NACA RM L52 J03a, 1952. 11. Lippisch, A., md Beuschamen, W.: Pressure Distribution Neasurements at Xigh Speed and Oblique Incidence of F
49、lox. NACA M 1115, 1947. 12. Edxards, George G., ad Boltz, Frecerick W.: An Analysis of the Forces and -Pressure Distribution OII aUizg With the Leaax Edge Swept Eack 37.25O. NACA W AgKOl, 1950. Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-20 - NACA FM W2L05 TAJ3JS I .- RATIO OF TOTAL WING AREA TO WING AREA AFFECT
