REG NASA-TN-D-7575-1974 Performance of an asymmetric short annular diffuser with a nondiverging inner wall using suction.pdf

1、ANDNASA TECHNICAL NOTE NASA TN 0-75759-m( ASA-Ti,-D-7575) PENFCO lA iCE OF AN .74-18922ASYMETEIC SHOz ANNULA DIFFUSE WITH AS ODIVEYGING INNE)E, ALL U3IiG SUCTIOCi(1-Ai) 41 p HC $3.25 CSCL 20D UnclasH1/12 32722PERFORMANCE OF AN ASYMMETRICSHORT ANNULAR DIFFUSERWITH A NONDIVERGINGINNER WALL USING SUCTI

2、ONby Albert J. JuhaszLewis Research CenterCleveland, Ohio 44135NATIONAL AERONAUTICS AND SPACE ADMINISTRATION * WASHINGTON, D. C. * MARCH 1974Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-1. Report No. 2. Government Accession No. 3. Recipients Catal

3、og No.NASA TN D-75754. Title and Subtitle 5. Report DatePERFORMANCE OF AN ASYMMETRIC SHORT March 1974ANNULAR DIFFUSER WITH A NONDIVERGING 6. Performing Organization CodeINNER WALL USING SUCTION7. Author(s) 8. Performing Organization Report No.Albert J. Juhasz E-761510. Work Unit No.9. Performing Org

4、anization Name and Address 501-24Lewis Research Center11. Contract or Grant No.National Aeronautics and Space AdministrationCleveland, Ohio 4413513. Type of Report and Period Covered12. Sponsoring Agency Name and Address Technical NoteNational Aeronautics and Space AdministrationWashington, D. C. 20

5、546 14. Sponsoring Agency Code15. Supplementary Notes16. AbstractThe performance of a short highly asymmetric annular diffuser equipped with wall bleed (suc-tion) capability was evaluated at nominal inlet Mach numbers of 0. 188, 0. 264, and 0. 324 withthe inlet pressure and temperature at near ambie

6、nt values. The diffuser had an area ratio of2. 75 and a length- to inlet-height ratio of 1. 6. Results show that the radial profiles of diffu-ser exit velocity could be controlled from a severely hub peaked to a slightly tip biased formby selective use of bleed. At the same time, other performance p

7、arameters were also im-proved. These results indicate the possible application of the diffuser bleed technique to con-trol flow profiles to gas turbine combustors.17. Key Words (Suggested by Author(s) 18. Distribution StatementCombustor flow control Unclassified - unlimitedDiffuser bleedCat. 1219. S

8、ecurity Classif. (of this report) 20. Security Classif. (of this page) 21. No. of Pages 22. Price*o SUnclassified Unclassified a I L $3. 9* For sale by the National Technical Information Service, Springfield, Virginia 22151NASA-C-168 (Rev. 6-71) /Provided by IHSNot for ResaleNo reproduction or netwo

9、rking permitted without license from IHS-,-,-PERFORMANCE OF AN ASYMMETRIC SHORT ANNULAR DIFFUSER WITH ANONDIVERGING INNER WALL USING SUCTIONby Albert J. JuhaszLewis Research CenterSUMMARYThe performance of a short, highly asymmetric annular diffuser equipped with wallsuction capability was evaluated

10、 at nominal inlet Mach numbers of 0. 188, 0. 264, and0. 324 with the inlet pressure and temperature held at near ambient values. The diffuserhad an area ratio of 2. 75 and a length- to inlet-height ratio of 1. 6. The straight, non-diverging inner diffuser wall was formed by a cylindrical section of

11、the same diameteras that of the inlet passage inner surface. This section was mounted downstream of theinlet passage in such a manner that a narrow circumferential gap was left between thetwo surfaces for the purpose of applying small amounts of inner wall suction. The outerdiffuser wall was shaped

12、in a form of a torus of quarter-circle cross section and it wasprovided with two stepped suction slots, continuous over the full circumference. Theperformance parameters that were determined included exit velocity profile shape, dif-fuser effectiveness, percent total pressure loss, and diffuser effi

13、ciency.Test results indicate that by selective use of suction, the exit velocity radial profilecould be altered from a severely hub peaked to a center symmetric or even a slightly tipbiased shape. At the same time, significant improvements in all the other performanceparameters were also obtained at

14、 suction rates of four to nine percent of diffuser flow.This capability of altering radial profiles of exit velocity and simultaneously im-proving diffuser performance, in general, suggests that the diffuser bleed technique maybe used to control inlet airflow distribution in gas turbine combustors.

15、The advantage ofa combustor equipped with diffuser wall bleed capability would be the possibility of per-formance optimization at each of several operating conditions. For example, the com-bustor efficiency at idle operation might be increased by establishing a more favorablefuel-air ratio (near sto

16、ichiometric) in the primary zone, because most of the airflowcould be directed to bypass the primary zone at the idling condition. Increasing primaryzone efficiency would decrease emissions of carbon monoxide and unburned hydrocarbons(an annoying problem around airports). The ability to control the

17、primary zone airflowdistribution would also lead to improved altitude relight capability, because of the re-duced velocity in the vicinity of the fuel nozzles and ignitors.Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-INTRODUCTIONThe purpose of thi

18、s investigation was to evaluate the use of diffuser wall bleed in aneffort to improve the performance of annular diffusers of the type used between the com-pressor and combustor of gas turbine aircraft engines. The primary function of suchdiffusers is to reduce the velocity of the air leaving the co

19、mpressor from a Mach num-ber range of approximately 0. 25 to 0. 40 down to a range of 0. 05 to 0. 10, in order to en-sure efficient combustion at a low total pressure loss. A second diffuser function, pro-posed in reference 1, would be to provide the combustor with a radial airflow distribu-tion whi

20、ch is optimum for the particular engine operating condition. However, becausegas turbine aircraft engines are required to operate at a wide range of conditions, afixed diffuser geometry would represent either a compromise between the various oper-ating conditions or the optimum design for a given co

21、ndition such as cruise. In theformer case, optimum combustor performance would not be obtained at any operatingcondition; in the latter case combustor performance would drop sharply at off-designconditions. Variable geometry diffusers could possibly provide the correct airflowdistribution at each en

22、gine operating condition. However, such diffusers may be quitecomplex, because of mechanical linkages and overlapping surfaces which would have tobe remotely operated.An alternate method of controlling combustor inlet airflow distribution which is es-pecially suited for advanced high-temperature eng

23、ines which require turbine cooling wasproposed in reference 1. This method employs an asymmetric diffuser with a graduallydiverging inner wall and a rapidly diverging outer wall as shown in figure 1. The diffu-ser is also provided with wall bleed (suction) capability. In the present report, the term

24、“bleed“ denotes a small fraction of the diffuser flow which is ducted to a region of lowerpressure as would be done in engine applications. The term “suction“ denotes ductingof this flow to a subatmospheric sink, as was done in this component study. At idle andaltitude relight conditions (fig. 1(a),

25、 no bleed would be used; consequently, the asym-metric geometry would cause the diffuser exit velocity profile to be hub peaked. A hubpeaked combustor inlet airflow distribution would be desirable at idle and altitude relightconditions. With a hub peaked inlet velocity profile most of the airflow wo

26、uld bypass theprimary zone of the combustor, thus raising the local fuel-air ratio and combustion effi-ciency at idle. Simultaneously, exhaust emissions of unburned hydrocarbon and carbonmonoxide would be reduced. While NOx levels would not be reduced, they are low enoughat the idling conditions to

27、be negligible. During altitude relight operation, the hubpeaked airflow profile would permit a low-velocity recirculation zone to be establishedin the region of the fuel nozzles and ignitors and thus createthe fuel-air ratio conditionsnecessary to improve the potential of low-pressure relight.At tak

28、eoff and cruise operation application of bleed flow on the outer wall of the dif-fuser would permit the combustor inlet airflow profile to be changed to a form that is2Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-considered optimum, such as a low-

29、curvature, center-peaked profile. Depending oncombustor geometry, a small amount of bleed from the inner wall may also be neededfor more precise profile control. Since the amount of bleed flow removed through thediffuser walls appears to be compatible with turbine cooling and auxiliary drive require

30、-ments (5 to 10 percent), there would be practically no net penalty on engine cycle effi-ciency.Most of the data of reference 1 were obtained with a short annular diffuser having asymmetric axial section of its flow passage. In an effort to obtain data trends withsimulated asymmetric passages some t

31、ests were also made with a cylindrical splitterring installed in the diffusing passage. This splitter divided the diffusing passage intotwo asymmetric half annuli, each having a cylindrical nondiverging wall and a rapidlydiverging opposite wall. Results indicated that diffuser exit velocity profile

32、could becontrolled by using wall suction.A better simulation of the proposed asymmetric annular diffuser with high diver-gence on its outer wall and low divergence on its inner wall, was tested in reference 2.The asymmetric flow passage was obtained by displacing the inner surface of the sym-metric

33、diffuser exit duct, used in reference 1, radially outward. Results showed goodcontrol of diffuser exit velocity profile when suction was applied on both the inner and theouter wall. With suction applied to the outer wall only, the diffuser exit velocity profilecould not be continuously controlled fr

34、om hub peaked to tip peaked forms. The reasonfor this reduction in velocity profile control capability was inner wall flow separation,which occurred at a critical outer wall suction rate of 2. 8 percent.The present investigation was conducted to determine whether such inner wall flowseparation could

35、 be prevented or delayed by changing the asymmetry of the diffuser.Performance data were obtained for a highly asymmetric diffuser, with all of the flowdeceleration occurring on its outer wall. The diffuser had an area ratio of 2. 75 and alength- to inlet-height ratio of 1. 6. The inlet passage flow

36、 area was 304 square centi-meters (47.12 in. 2). The straight, nondiverging inner diffuser wall was formed by acylindrical section of the same diameter as that of the inlet passage inner surface. Thissection was mounted downstream of the inlet passage in such a manner that a narrowcircumferential ga

37、p was left between the two surfaces for the purpose of applying smallamounts of inner wall suction. The outer diffuser wall was shaped in the form of a torusof quarter-circle cross section and it was provided with two stepped suction slots, con-tinuous over the full circumference. Velocity profiles,

38、 diffuser effectiveness, and dif-fuser pressure drop data were obtained for nominal diffuser inlet Mach numbers of0. 188, 0. 264, and 0. 324 at suction rates of zero to 9. 5 percent of total flow. Thegreater part of the diffuser performance data were obtained with suction applied to theouter wall on

39、ly. To check whether local separation was occurring on the cylindrical,nondiverging, inner wall, a limited number of measurements were made with suction3Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-also applied on the inner wall through the previo

40、usly mentioned narrow circumferentialgap at the diffuser throat. All testing was conducted with air at near ambient tempera-ture and pressure.SYMBOLSA areaAR diffuser area ratioB bleed flow fraction of total flow rateCp specific heat at constant pressuregc dimensional constantH diffuser inlet passag

41、e heighth specific enthalpyL diffuser lengthM average Mach number at an axial stationm mass flow rateP average pressure at an axial stationp local pressure at a radial positionR gas constant for airr wall contour radiusS entropyT temperatureV average velocity at an axial stationv local velocity at a

42、 radial positionX downstream distancey specific heat ratioE diffuser efficiency, eq. (5)77 diffuser effectiveness, eq. (3)p density4Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-Sub scripts:act actual conditionsatm atmospheric conditionb bleed or s

43、uction stationid ideal conditionsm maximumr local value at a given radial positions isentropic condition0 stagnation condition1 diffuser inlet station2 diffuser exit stationAPPARATUS AND INSTRUMENTATIONFlow SystemThe investigation was conducted in the test facility described in reference 2. Aschemat

44、ic of the facility flow system is shown in figure 2(a). Air, at a pressure of ap-proximately 100 newtons per square centimeter (145 psia) and at ambient temperatureis supplied to the facility by a remotely located compressor station. This air feeds thethree branches of the flow system.The center bra

45、nch (identified as “main air line“) is the source of airflow through thetest diffuser. The air flowing through this branch is metered by a square-edged orificeinstalled with flange taps according to ASME standards. The air is then throttled to nearatmospheric pressure by a flow control valve before

46、entering a mixing chamber fromwhich it flows through the test diffuser. The air discharging from the diffuser is ex-hausted to the atmosphere through a noise absorbing duct.The two other branches of the flow system supply the two air ejectors, which pro-duce the required vacuum for the inner and out

47、er wall diffuser bleed flows. The ejec-tors are designed for a supply air pressure of 68 newtons per square centimeter(100 psia) and are capable of producing absolute pressures down to 2. 38 newtons persquare centimeter (7. 0 in. Hg).The inner and outer diffuser wall bleed flows are also metered by

48、square-edged ori-fices. These orifices are also installed with flange taps according to ASME specifica-5Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-tions in the suction flow lines that connect the inner and outer diffuser wall bleed cham-bers to

49、their respective ejector vacuum sinks.Diffuser Test ApparatusThe apparatus used in this investigation was essentially that of reference 2, but fora few modifications. A cross-sectional sketch including pertinent dimensions is shownin figure 2(b). As in reference 2, the centerbody that forms the inner annular surface iscantilevered from support struts located 30 centimeters (12 in. ) upstrea