1、-= and the resulting total-pressure andstatic-pressurelosses were less than those for a 10 diffuser withoutboundary-layer control. The air-flow separationwas eliminated in the 30and 50 diffusers with suction mass flows of 3 and 4 percent of the inletmass flows2 respectively.INTRODUCTIONThe results o
2、f experimental investigationshave shown that separationof the boundary layer occurs in diffusers that have total expansion anglesgreater than about 15. !J?hisair-flow separation results in total-pressureand static-pressurelossesj and in nonuniform velocity profiles at the exitof the diffuser (refs.
3、1 to 4). In order to improve these characteristicsinvestigationshave been conductedwhere the boundary layer in the diffuserhas been re-energizedwith vortex generators or blowing through slots orthe boundary layer in the diffuser has been removed by suction throughslots (refs. 5.to n). The results of
4、 some of these tests showed thatlarge improvementswere made; however, air-flow separation did not appearto be eliminated in the conical diffusers with total expansion anglesgreater than about 30.Since area suction (suctionapplied over a distributed area) waseffective in eliminatingair-flow separatio
5、n on a wing at high angles ofattack (ref. 12)j exploratory tests were initiated to determine whetherarea suction would eliminate air-flow separation in diffusers with largeexpansionangles. These exploratory tests were conductedusing a 30. . ._._ _ - - - - - .Provided by IHSNot for ResaleNo reproduct
6、ion or networking permitted without license from IHS-,-,-. ,.-2 NACA TN 3793and a “ conical diffuser with various extents of porous area. For com-parative purposes, a 10 conical diffuser with no porous surface was also .tested. Due to the exploratory nature of these tests, the tests were per-formed
7、with only one inlet boundary-layer condition and for a mean inlet 4Mach number of about 0.2.NOTATION_.speed of sound,cross-sectionalacceleration offt/secarea, sq ftgravity, ft/sec2local total pressure, lb/sq ftarithmetic average total pressure, lb/sq ft 4ratio of average total-pressure loss to theor
8、etical incompressible(% -E2)/z=value for an abrupt expsnsion,1-(AJA=)2length of diffuser,measured along center line,w lb/seemass flow of air, -, g ft/sec2-uaverage Mach numbers local static pressure, lb/sq ftarithmetic average static pressure, lb/sq ftP- 5=static-pressure coefficient,qdynamic pressu
9、re, H - p, lb/sqftradial distance in.radius, in.local veloci, ft/secvelocity outside of the boundary layer, ft/sec. -. _ ._ -.in. -Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-3*IWCA TN 3793 .u average velocity at a given cross section, ft/secw we
10、ight rate of flow, lb/seex longitudinal distance along center line of diffuser, measured frombginning of diffuser, =.Y distancegR= ratio ofR= ratio ofv. ratio ofvalue,!.J.2e12Psdiffuserinlet, 2exitfrom wall of diffuser, in.boundary-layer displacement thickness to inlet radiusboundary-layermomentumav
11、erage static-pressure(F2 - Fl)hz1- (A in fact, the values were less than those measured for the 10diffuser of the present test. The data for the diffusers without boundary-layer control were obtained from references 1, 2, 3, 4, and 10, and theyare presented as bands of data because of the difference
12、s in valuesreported. Values for the 30 and 50 diffusers with the porous areassealed could not be accurately measured with the apparatus used becauseof the-unsteady flow resulting from the separation of the boundary layer.However, the approximate values of loss factors measured for these dif-fusers w
13、ere within the bands of data shown in figure 2. With suctionapplied, the flow in the diffuser was very steady.The effect of area suction on the total-pressure,static-pressure,and velocity distributionsat the exit of the 30 and 50 diffusers (sta-tion 2) is shown in figure 3. These data for the 30 and
14、 50 diffuserswith suction are compared with those for the 10 diffuser without suctionin figure 4. This figure shows that area suction reduced the total-pressure losses to values less than those of the 10 diffuser, indicatingthat the air-flow separationnormally attendant with wide-angle diffuserswas
15、eliminated by the use of area suction. Jn figure 4, it is also seenthat the static-pressuredistribution at the exit of the diffuser withsuction became less uniform as the diffuser angle was increased from 10to 300 to 500. This type of pressure gradient would be expected in poten-tial flow ti a wide-
16、angle diffuser (see ref. 13).It should be pointed out that a comparison of the velocity profilesat the exit of the diffusers (fig. 4) does not provide a comparison ofthe boundary-layer profiles because of the previously noted nonuniformstatic-pressuredistribution. However, the boundary-layer thickne
17、ss atthe exit of the diffuser can be determined from the total-pressure dis- a71tributions presented in figure 4. A comparison of these distributionsshows that the boundary-layer thicknesses at the exit of the 30 and 50diffuserswith suction were less than that of the 10 diffuser. Further,the shape o
18、f the total-pressuredistributions (fig. 4) tidicates that theboundary layer in the diffusers with suction is at least as stable as thatin the 10 diffuser. Since there is little or no separationat the exitof the 10 diffuser, it can be concluded that area suction has eliminatedthe air-flow separation
19、that existed in the 30 and 50 diffuserswithoutboundary-layer control. A typical inlet boundary-layerprofile measuredin these tests is shown h figure 5. These measurements show that a thin,stable, turbulent boundary layer existed at the inlet of the diffusers.The longitudinal distributions of static-
20、pressurecoefficientalongthe walL of the 30 and 50 diffusers presented h figure 6 show that areasuction increased the static-pressurerecovery along the entire length ofthe diffusers. The longitudinal distributions of static-pressurecoeffi-cient for the 30 and 50 diffusers with suction are comparedwit
21、h that-. . - . _ _ - - . .- ._Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-6 IWK!ATN 3793for the 10 difftmer fi figure 7. The data of this figure indicate thatarea suction permits equivalent static-pressurerecovery to be obtainedwith a diffuser on
22、lya fraction of the length of the 10 diffuser.Suction RequirementsThe effect of suctionmass flow on the total-pressurerecovery ofthe 0 and 50 diffusers is shown h figure 8 for several lengths ofporous area. It csa be seen that a large increase in total-pressurerecovery is obtainedwith small.suction
23、flow ratios. It can be ascer-tained from the data presented in previous figures that air-flow separa-tion has been eliminated in the diffuser when the suction flow ratio issufficientto insure essentially complete total-pressurerecovery. Itcan also be seen in this figure that it iB not necessary, or
24、even desira-ble from a suction flow standpoint,to apply area suction down the enttielength of the diffuser. The suctionmass-flow ratios and the pumpingpressure coefficientsregyired for the 30 and 50 diffusers at the lowestsuction maas-flow ratio where separationwas indicated to be eliminated ware su
25、mmarized in the following table:The relatively large pumping pressure coefficientsrequired for the 50diffuser resulted because sufficient inflow velocities could be obtainedthrough the dense porous stainless steel only by providing a large pres-sure differential. It would be expected that these pump
26、ing pressurecoefficients could be reduced by the use of a porous material that had agreater porosity near the beginning of the diffuser. Based on otherapplications of area suction (e.g., ref. 12) it would be expected thatthe use of a material with a tapered porosity could also reduce thesuction mass
27、-flow ratios required to elidnate separation.CONCIJJSIONSExploratory tests were made with area suction alied to conicaldiffuserswith expansion angles of 30 and 50. These tests, made at amean inlet Mch number of approximately 0.2, indicated that the air-flowseparationwas eliminatedby use of area suct
28、ion, and that the resultingtotal-pressureand static-pressurelosses were less than those for a 10 , Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-NACA TN 3793diffuser without boundary-layer control. The air-flow separationwaseliminated in the 30 and
29、 50 diffuserswith suction mass flows of 3and 4 percent of the Met mass flows, respectively.7Ames Aeronautical LaboratoryNational Advisory Committee for AeronauticsJfoffettField, Calif., Jhne 27, 19561. Gibson, A. H.: On theREFERENCESResistance to Flow of Water Throu Pipes orPassages Having Divergent
30、 Boundaries. Royal Society o; EdaburghhS., VO1. 48, pt. 1, 19-19, pp. 97-116.2. Gibson, A. H.: Hydraulics and Its Application. Fourth cd., D. VanNostrand Co., 1930, pp. 92-93.3. Peters, H.: Conversion of Energy in Cross Sectional Divergences UnderDifferent Conditions of Inflow. NACATM 737, 1934.4. L
31、ittle, B. H., Jr., and Wilbur, Stafford W.: Perfoce and Boundary-Layer Data llrom12 and 23 Conical Diffusers of Area Ratio 2.o atMach Numbers U to Choking and Reynolds Numbers Up to 7.5x106.NACA Rep. 1201, 1954.5. Valentine, E. Floyd, and Carroll, Raymond B.: Effects of sever-Arrangements of Rectang
32、dar Vortex Generators on the Static-fiessureRise Through a Short 2:1 Diffuser. NACA RM L50104, 1951.6. Wood, Chsxles C.: Preliminary Iirrestigationof the Effects of Rectan-gular Vortex Generators on the Performance of a Short 1.9:1 Straight-WallAnnular Diffuser. NACARML51G09, 1951.7. Henry, JohnR.,
33、and Wilbur, Stafford W.: FYeliminary Investigation ofthe Flow in an Annular-Diffuser-TailpipeCombtition With an AbruptArea Expansion and Suction, hqjection,and Vortex-GeneratorFlowControls. NACARML53K30, 1954.8. Ackeret, J.: Removing Boundary Iayer by Suction. NACA TM 395, 1926.9. Biebel, William J.
34、: Iow-l?ressureBoundary-Layer Control in Diffusersand Bends. llACAIJRL-84, 1945. (FormerlyNACAARR L5C24)10. Gratzer, L. B., and Smith, R. H.: Boundary Layer Control for WideAngle Diffusers. Aero. Lab. Rep. 300, Univ.of Washington, Nov. 22,1948. . _ _ _ -. .Provided by IHSNot for ResaleNo reproductio
35、n or networking permitted without license from IHS-,-,-8 N/WA”TN 3793I-1. Manoni, L. R.: Wide Angle Diffusers Employing Boundary Iayer Control.Rep. R-95460-12, Research Dept., United Aircraft Corp., Mar. 19,1952. d1.2.Holzhauserj Curt A., and Martin, Robert K.: The Use of Leadfig-EdgeArea Suction to
36、 hcrease the Maximum Lift Coefficient of a 35uSwept-BackWing . NACARMA52G17, 1952.13. Rouse, Hunter: Elementary Mechanics of Fluids, Chapter II. John Wileyand Sons, hC., 1946.,u.1- . . . .Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-It.mII “-t4Tr
37、70”2011 dj1II,/JIsealed IWith suction I/ II1I/“/o .2 .4 ,6 .8 1.0uill(b) ZW =50, sealed sndwith suction (/m,= 0.(%, x/2 =0.15).Figure 3.- Concluded.Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-,II.I1.0.8.6.4.20; .2A.6.81.0o .2 .4 .4 .8 1.0=EI%o .2
38、 .4 .6 .8 1.0P:;xo .2 .4 .6 .8 1.0u=ugFigure k.- Caqarison of total-preaeure, Btatlc-pressure, and velocity distributions at exit of10 diffuser ti_th those of 30 and 50 diffusers with area suction, mB/ml = 0.04 and 0.06, Esrespectlvly; Ml = 0.2, El = 0.2, El = 0.2, A2/A1 = 2. . . - . -. . . _ _Provi
39、ded by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-=!-* 4 w-aw,*LA!-. 2 /1 I!kLt $Oro 29 -100.2-2 Q = SOO,m#q = 0.06, X/?UO.4 1 P.6- .8 .1.0-* 8 0 .8 1.6 2.4 3.2 4.o 4.8 5.6Ratio of longitudinal Wd.snce to Met radiua, x/+ Figure 7.- C_ison of the longitudina
40、l distributions of static-pressure coefficfsnlm along the10 diffuser, 300 diffuser tith suction, end the 50 diffuser with suction; 1 = !3.2,A=/Al = 2.Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-1.0.81# .7I .61+a715.4,ISiation 1/ Ml = 0.2, &/AX = 2.NACA .Iaey Field,%.,.Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-
