1、u)NATIONAL ADVISORY COMMITTEEFOR AERONAUTICSTECHNICAL NOTE 3123EFFECT OF VARIOUS ARRANGEMENTS OF TRIANGULARLEDGES ON THE PERl?ORkt the average height of the particles used was 0.10 inch. Theheight of most ledges was approximately one-tenth of the inlet boundsmy-layer thickness. wProvided by IHSNot f
2、or ResaleNo reproduction or networking permitted without license from IHS-,-,-NACA TN 3123The following(1)A band ofprocedure was used to install a typical rough ledge:cork particles, 1 tich wide, was cemented to the dif-fuser wall, transverse to the direction of flow, in the proper axiallocation.(2)
3、 The leading edge was buffed and faired to give the strip smapproxhnately triangular cross section with a trailing edge about 0.10 imchhigh. A view of the appro-te cross section of some typical roughledges appears in figure 2.The smooth ledges were made of balsa-wood strips of triangular crosssectio
4、n which were installed in the diffuser in such a manner as to havethe leading edge smoothly faired into-the diffuser wall. After install-ation, each balsa-wood strip was carefully filled and sanded to producea smooth ledge surface.The axial position and alphabetical designation of the ledges areshow
5、n in figure 1. Configuration a has a single l-inch-wide, O.l-inch-high roughness strip, designated ledge a, installed near the diffuserinlet (seefig. l). The first ledge and each succeeding ledge wasinstalled as indicated fi figure 1. Each configuration tested is iden-tified by a letter which denote
6、s the position of the last ledge installedfor that configuration. The following table gives a description of allthe ledge configurations investigated:Configuration:b-1b-2b-3cdd-1d-2efgNmberof roughledgesHeight,in.0.10.10.10.10.10.10.10.10.20.10.10.10.10Numberof smoothledges-111-3-Height,in.-0.10.1a7
7、120-.10-Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-6 NACA TN 3123Instrumentationand CalibrationA series of static-pressureorifices was installed along a singlegeneratrix running from the diffuser inlet to the tailpipe station inorder to measure
8、longitudinal static-pressuredistributions. At sta-tions 1, 2, and 3 (thediffuser inlet, diffuser exit, and tailpipe exit,respectively),wall-static-pressuremeasurements were made at six equallyspaced circumferentialpositions. All static-pressureorifices were con-nected to a multittie mancmeter and pr
9、essures were recorded photographi-cally. Total- and static-pressurestream surveyswere made at stations 2and 1, in that order, for all the ledge configurationsby using threeeqyally spaced, resnotelycontrolled, electrically driven pressure probes.A sketch of a pressure probe is included in figure 1.Th
10、e flow conditions at the diffuser inlet were determined by makingpressure-probe surveys at three equally distributed positions sround thecircwaference of the inlet. The diffuser inlet calibration is shown infigure 3, in which the inlet Wch number, the Reynolds number based oninlet dismeter, and-the
11、weight flow adusted for standard conditions of29.g2 inches of mercury smd 600 F are all plotted as functions of theinlet pressure ratio po* m ical velocity profiles at the inlet/$station 1 are shown in figure 4 for several values of PI 0. The inletboundary-lsyer thickness was of the order of 5 perce
12、nt of the inletdiameter.Accuracy of MeasurementsFor some of the configurations investigated, the occurrence of sepa-ration or asymmetrical flows or the presence of turbulent fluctuatingvelocities cast doubt on the accuracy of the results obtained. Comparisonsbetween inlet and exit weight-flow values
13、, shown for all configurations,give some indication of the inaccuracy resulting from these effects.CALC!UIATIONOFPERFOWCE PARAMETERSThe reference static pressure p. used in conjunction with the inletstatic pressure pl provided the required correlatingparsmet for cal-culating all performance characte
14、ristics.The volume-weightedmean loss in total pressure from the referencestation O to the station under considerationwas cmputed in thefollowingmanner:.0 “-Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-NACA TN 3123(1)The loss in mean total pressure
15、 was ccmputed for the diffuser by usingthe equation%,2 = (% -%) - (PO - %) (2)in which the inlet pressure ratio p. 5s used as a correlating parsmeter.E expressed ngndimensionally by dividihg by the act pressure of the inlet,. /the parameter AE 1 cam be defined as the total-pressure-loss coefficient.
16、 The rise in static pressure was ccmputed as the difference betweenthe arithmetic mean of the six wall-static-pressuremeasurements at sta-tion 1 and the arithmetic mean of the wall-static-pressuremeasurementsat station 2 or 3. The ideal static-pressure difference was determinedby assudng frictionles
17、s one-dimensional incompressible flow for theFssme values of pl o. The ratio of the measured static-pressurerise to/the ideal difference 4 Apidal is defined as the diffuser effectiveness. .RESULTS AND DISCUSSIONIn contrast to the flow in the diffuser with no ledges, which peri-odically shifted posit
18、ion and lacked axial symmetry, the flow in thediffuser with one or more ledges installed was steady and, in general, hadapproximately symmetrical velocity profiles at the diffuser exit for mostconfigurations. Because the flow was so unstable in the diffuser withoutthe inlet roughness strip, ledge a,
19、 accurate total-pressuremeasurementscould not be made at the diffuser exit./Therefore, in order to providevalues of W for comparison purposes, total-pressure surveys were1made at station 3 where the flow was steady. In subsequent sections,. comparisons are made between the values of / measured at st
20、ation 31for the diffuser with no ledges and the values of Z the ratio x/L is d-efinedas that portion of the diffuser length.over which the ledges were installed. At the diffuser exit, as the nmberof ledges installed was increased, the diffuser effectiveness decreasedprogressively until, with seven l
21、edges installed, a decrease in the effec-tiveness of about 7 percent was obtained. At the tailpipe exit the effectof the ledges was also detrimental, but on the order of a maximum of 2 per-cent. The net effect on the diffuser effectiveness of the ledge instal-lations indicates, therefore, that the l
22、edges were not effective diffuserboundary-layer control devices.Figure 7(b) demcmstrates the variation of b-2, O.1 inch high; and b-3, 0.20 inch high. Pressure measurements weremade at each ledge-height condition investigated,with all ledges down- a71stream of ledge b removed to avoid the ccunplicat
23、ionof titerrelated effects.The results of this investigation are shown in figure 13. Figure 13(a)/is a plot of ,2 qc as a function of inlet pressure ratio, withledge heights as the parameters. A-weight-flow check was made to deter-4mine whether erroneous values of Exit,uretmure orifice. station 1 st
24、ation 21h ilpilmexit,station 3stetion OFour GqullYE!iiliCcmfig- Distnncwration freminletdiffusejullctioa 1.J+9b.: 1:72ef :2E.Crma eactiomofdiffwer ohowinglongitud,tilPJaition of ledgeeFre 1.- Generalt *sced trkticorifices, .03 dh+m./ I .Og dhln,/ .:74PJPO u 0.95C9.6 . 6*= 3.75e = 0.89 g = 0.98E=4.4 200I 1 I I I.2 -.4 .6 .8 1.0 0 .2 .4 .6 .8 1.0L I I 1 IO a712 .4 .6 .8 1.0u/u(e) Conflguratimng.Figure 9.- Concluded.Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-
