1、NASA TECHNICAL NOTE IMfZfiK Sonic inlet; Induction system;Pressure distribution; Potential flow; Bound-ary layer flow; Compressibility correction;STOL aircraft18. Distribution StatementUnclassified - unlimited19. Security Classif. (of this report)Unclassified20. Security Classif. (of this page)Uncla
2、ssified21. No. of Pages2222. Price*$3.00 For sale by the National Technical Information Service, Springfield, Virginia 22151Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-THEORETICAL AND EXPERIMENTAL INTERNAL FLOW CHARACTERISTICSOF A 13.97-CENTIMETE
3、R-DIAMETER INLET AT STOLTAKEOFF AND APPROACH CONDITIONSby James A. AlbersLewis Research CenterSUMMARYThe theoretical and experimental internal flow characteristics of a 13. 97-centimeter-diameter inlet with centerbody refracted and extended are presented at STOLtakeoff and approach operating conditi
4、ons. The theoretical results are obtained fromincompressible potential flow corrected for compressibility and boundary layer. Com-parisons between theoretical internal surface pressure distributions and experimentaldata are presented for free-stream velocities of 0, 24, 32, and 45 meters per secondf
5、or a range of inlet incidence angles from 0 to 50. Surface static-pressure distribu-tions are illustrated at circumferential locations of 0, 60, 120, and 180. SurfaceMach number distributions from stagnation point to diffuser exit are presented alongwith turbulent boundary layer shape factors.The re
6、sults indicate a large circumferential variation in surface static pressuresat the inlet highlight and throat at large incidence angles. Only small circumferentialvariations in surface static pressure occurred in the last 50 percent of the diffuser.The largest diffuser adverse pressure gradients occ
7、urred on the windward side of theinlet and at the highest incidence angle. A 45-per cent increase in local surface Machnumber (52-percent decrease in surface static pressure) was obtained as incidence anglewas increased from 0 to 40. Local theoretical surface Mach numbers as high as 1. 45were found
8、on the windward side of the inlet. Extending the centerbody of the inlet for-ward resulted in large regions of local sonic velocities in the throat of the inlet. Ingeneral, the theoretical and experimental surface static pressure distributions agreed.However, at the inlet highlight the theoretical s
9、tatic pressures were generally lower !than the experimental data.Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-INTRODUCTIONCurrently, there is much interest in the design and understanding of inlets for sub-sonic aircraft. The inlet must be designe
10、d to supply high pressure recovery and uni-form flow to the engine compressor during low-speed and cruise operation. Engineoperation can be adversely affected by circumferential distortions, occurring mainlywhen the airplane is operating at high angles of attack or yaw (ref. 1). Because of thehigh l
11、ift coefficients and low speed necessary for takeoff and landing operation of STOLaircraft, the engine inlet will be exposed to larger upwash angles than conventional air-craft (ref. 2). In addition, the engine inlet may be exposed to crosswinds as large as13 to 18 meters per second (ref. 3). These
12、large incidence angles result in large ad-verse pressure gradients over a large portion of the internal diffuser surface. Theselarge adverse pressure gradients may cause either laminar or turbulent separation. Ingeneral, the designer tries to avoid flow separation on the inlet surface in order toach
13、ieve high pressure recovery and uniform flow at the compressor face. Little exper-imental information on nacelle inlets at incidence angles other than zero is currentlyavailable. Also, accurate methods for estimating surface pressure distributions andboundary layer characteristics are needed to do d
14、etailed design studies of inlets forSTOL aircraft.1 This report presents theoretical and experimental internal flow characteristics ofa 13. 97-centimeter-diameter inlet at STOL takeoff and approach conditions. The theo-retical methods used in this investigation include both potential flow and bounda
15、ry-layerflow for axisymmetric inlets. The experimental data were obtained from wind tunneltests of the translating centerbody inlet reported in reference 4. The inlet configuration(fig. 1) is a representative geometry for STOL applications. Details of the inlet geom-etry are given in reference 4. Co
16、mparisons between internal surface pressure distri-butions and experimental data are presented for free stream velocities of 0, 24, 32,and 45 meters per second for a range of incidence angles from 0 to 50. Mass flowrates through the inlet ranged from 90 to 103 percent of design. The design corrected
17、flow rate was 2. 49 kilograms per second. Location of stagnation points on the inlet lipand surface Mach number distributions are presented along with turbulent boundary-layer shape factors. The effect of centerbody location on surface static pressures is. also illustrated.METHOD OF ANALYSISThe inco
18、mpressible potential flow solution for axisymmetric inlets is the basis ofan analytical tool to design inlets for STOL applications. The incompressible potentialProvided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-flow solution was obtained by the method
19、of reference 5. This solution yields velocityand static pressures on the surface of the inlet for any combination of free-streamvelocity, inlet incidence angle, and mass flow rate through the inlet. The method ofreference 5 is based on the Douglas potential flow method (ref. 6). The incompressiblepo
20、tential flow solution can be refined by including the effects of compressibility andboundary layer.Compressibility CorrectionHigh subsonic or supersonic flows exist on STOL engine inlets at takeoff and ap-proach operating conditions. Thus, a compressibility correction should be made to theincompress
21、ible velocity. A compressibility correction applicable to nacelle inlets isdiscussed in reference 7, which proposed an empirical relation between the local incompressible velocity V- and the local compressible velocity V_. It is expressed as1 txwherep. incompressible density, which is equal to stagn
22、ation densityp average compressible density across flow passageV- average incompressible velocity across flow passage at given stationThis relation was used in this investigation.Boundary-Layer FlowThe surface Mach number distributions obtained from the potential flow solutionwere used as an input t
23、o the boundary-layer analysis to determine its growth and sepa-ration (if any) on the inlet surface. The method of Herring and Mellor (ref. 8) was cho- sen to calculate both laminar and turbulent boundary-layer growth because of its accu-racy, physical soundness, and adaptability to the present prob
24、lem. The effective-viscosity hypothesis of Mellor and Herring should be applicable for high adverse gradientflows encountered for inlets during low-speed operation. A more detailed discussion ofthis hypothesis is given in reference 9.Provided by IHSNot for ResaleNo reproduction or networking permitt
25、ed without license from IHS-,-,- In order to compute a boundary-layer solution, it was necessary to prescribe thevelocity profile at the start of the calculation, namely, the stagnation point on the inlet.Because the flow in the immediate neighborhood of the forward stagnation point is bothlaminar a
26、nd incompressible, laminar similarity solutions were applicable. TheFalkner-Skan laminar wedge flow solution (ref. 10) for stagnation point flow was usedfor a starting profile. This initial profile was calculated in the computer program ofreference 8.In addition, the transition point from laminar to
27、 turbulent flow must be located.The location of the transition point was based on theoretical considerations because ofthe lack of empirical methods that were applicable to the present problem. According cto Schlichting (ref. 10) the point of transition in the range of Reynolds numbers from 107to 10
28、 almost coincides with the point of minimum pressure of the potential flow. Inthis report the transition was assumed to take place at the point of minimum pressure.Once the laminar and turbulent boundary-layer characteristics on the inlet lip anddiffuser are calculated, the location of the turbulent
29、 separation point (if any) can be de-termined. The criteria used for turbulent separation are as follows: (1) adverse pres-sure gradient; (2) skin friction coefficient of approximately zero; (3) increase in shapefactor H as the separation point is approached; (4) a value of H of 2. 6 or greater im-m
30、ediately ahead of the separation point.RESULTS AND DISCUSSIONSurface Static-Pressure DistributionsTheoretical surface static-pressure distributions are first compared with experi-mental data for conditions where the flow is generally well behaved (i.e., attached alongthe inlet surface). The two inle
31、t geometries considered are shown in figure 1. All datacomparisons are made for the inlet with centerbody retracted (fig. l(a) unless statedotherwise. The effects of compressibility and boundary-layer displacement thicknesscorrection on the incompressible potential flow solution are first discussed.
32、Effect of compressibility correction. - A comparison of the theoretical internalsurface static-pressure distributions with experimental data is shown in figure 2 forstatic conditions (V = 0). (Symbols are defined in the appendix.) The pressure distri-butions are presented from the inlet highlight (X
33、/L = 0) to the diffuser exit (X/L = 1- )-The incompressible potential flow solution (the dashed line) yields static pressures ap-proximately 12 percent higher than experimental data near the inlet highlight and approx-imately 5 percent higher in the aft portion of the diffuser. The incompressible po
34、tentialflow solution overestimates the static pressure because of the high surface Mach num-bers (up to 0. 9). The incompressible potential flow solution with the compressibilityProvided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-correction is shown in f
35、igure 2 (solid line). The theory with the compressibility correc-tion generally compares well with the experimental data in the first half of the inlet.However, in the aft portion of the diffuser the theoretical static pressures are up to3 percent higher than experimental data.Effect of boundary-lay
36、er displacement thickness correction. - A combined potentialflow and boundary-layer analysis improves the agreement between the experimental andtheoretical surface static-pressure distributions in the aft portion of the diffuser. Thisis illustrated for a static condition (V = 0) in figure 3(a) and f
37、or a free-stream velocityof 24 meters per second at an incidence angle of 40 in figure 3(b). Adding the theoreti-cal boundary-layer displacement thickness to the inlet surface results in a decreasedflow passage cross-sectional area and hence an increase in the local surface Mach num-ber. The boundar
38、y layer is thickest near the diffuser exit and hence affects the staticpressure most in this region. The increase in local surface Mach number yields a re-duced theoretical surface static pressure for the potential flow solution with a boundary-layer correction.With both the compressibility correcti
39、on and boundary-layer correction, the theo-retical and experimental static pressures are in good agreement over the entire lengthof the diffuser. The boundary-layer correction is also effective in improving the solu-tion at incidence angle (fig. 3(b) in spite of the fact the boundary layer is not un
40、iformlydistributed circumferentially in the aft portion of the diffuser.The boundary-layer correction is not included in the remaining theoretical staticpressures because of the large computation time required to make this correction.Hence, there are small discrepancies between the theory and experi
41、ment in the aft por-tion of the diffuser in the subsequent figures.Circumferential variation. - The variation of the theoretical internal surface staticpressure along the diffuser surface for several circumferential angles j/ and for inci-dence angles of 20 and 40 is presented in figure 4. The circu
42、mferential variation instatic pressure is greatest at the inlet highlight (X/L = 0) and decreases with increasingX/L. The circumferential variation in p/P, is small for X/L locations greater than0.5. The degree of the circumferential variation in static pressure increases as inci-dence angle is incr
43、eased. At the inlet throat a 9 percent circumferential variation inP/POO occurs at an incidence angle of 20 (fig. 4(a), and a 16-percent variation occursat an incidence angle of 40 (fig. 4(b).Both the theoretical and the experimental circumferential variation of static pres-sure at several stations
44、is shown in figure 5. The experimental static pressures gen-erally compare well with the theoretical pressures for all circumferential angles. Thelowest inlet lip static pressures and largest diffuser adverse pressure gradients occuron the windward side of the inlet (i/ =0). Thus, all remaining comp
45、arisons will bemade at circumferential angle of / = 0.Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-Effect of flow variables. - To illustrate the applicability of the prediction methodover a wide range of conditions, a. comparison with experimental
46、 data is made for vari-ous incidence angles, free-stream velocities, and mass flow rates. The effect of thethree flow variables on the theoretical static pressures is first discussed. This is fol-lowed by a discussion of the comparison between theory and experiment.The effect of incidence angle is i
47、llustrated in figure 6 by the theoretical static-pressure distributions (with experimental data) at a free-stream velocity of 32 metersper second and for incidence angles of 0, 20, and 40. The value of the static-pressure ratio on the inlet lip decreases considerably as the incidence angle increases
48、from 0 to 40. For example, the theoretical static-pressure ratio at the inlet highlightdecreases from 0. 57 at an incidence angle of 0 to 0. 30 at an incidence agle of 40.There is little change in the static-pressure prediction in the aft portion of the diffuser(X/L 0.4) as incidence angle is increa
49、sed from 0 to 40. The large decrease in static -pressure ratio as incidence angle increases is related to the shift in the stagnation pointon the outer cowl surface on the windward side of the inlet. As the incidence angle in-creases from 0 (fig. 7(a) to 40 (fig. 7(c), the stagnation point moves further under thelower cowl external surface. This results in higher velocities and correspond
