Document
..- .
1 - .
* -. * *
N A S A C O N T R A C T O R
~- R E P O R T
INLET-TO-INLET SHOCK
INTERFERENCETESTS
by D. Lo Motycku uzld J. B. Mzlrphy
Prepared under Contract No. NAS 2-2077 by UNITED AIRCRAFT CORPORATION Pratt & Whitney Aircraft Division East Hartford, Conn.
f o r N A T I O N A L AERONAUTICS AND S P A C E ADMINISTRATION WASHINGTON, D. c. S E P T E M B E R 1965 TECH LIBRARY KAFB, NM NASA CR- 264 INLET-TO-INLET SHOCK INTERFERENCE TESTS ByD. L. Motycka and J. B. Murphy Distribution of t h i s r e p o r t is provided in the interest of informationexchange.Responsibilityforthecontents resides in the author or organization that prepared it.
Prepared under Contract No. NAS 2-2079 by UNITED AIRCRAFT CORPORATION Pratt & Whitney Aircraft Division East Hartford, Conn.
f o r NATIONAL AERONAUTICS ANDSPACE ADMINISTRATION .
Forsale b y the Clearinghouse for Federal Scientific and Technical Information Springfield,Virginia 22151 - P r i c e $3.00 FOREWORD This final report describes the work which w a s accomplished by P r a t t & Whitney Aircraft for NASA AmesResearchCenterinaccordance withtherequirementsdefined by Contract NAS2- 2079 entitled "Inlet-to-Inlet Shock Interference Tests", dated June 30, 1964. This report has been prepared to fulfill the require- ments of Article IX (C) of the subject contract.
iii TABLE OF CONTENTS Page Foreword iii vi Illustrations Tables vi i ix List of Symbols x i Abstract Summary and Conclusions 1 Introduction Description of Models 4 Model Geometry Instrumentation 9 Description of Test Equipment 17 1. GeneralDescription of 17 InchSupersonic Wind Tunnel 17 2. Special Equipment 21 Data Analysis 2 2 1. Calibration of theBasicInlet 22 2. Determination of PerformancePenalties 27 3. Establishment of the No InterferencePenalty Line 3 0 4 . TheEffect of theInterferencePlate 37 5. "Buzz" Analysis 4 0 Recommendations 44 References 4 5 V ILLUSTRATIONS Figure Page 1 Variation of Wind Tunnel Reynolds Number 3 With Free Stream Mach Number 2 Model Installation in Supersonic Wind Tunnel 3 Inlet Coordinates With Centerbody at Mach 3. 0 Design Position 4 Ratio of Inlet Internal Passage Area to Inlet Capture Area Versus Centerbody Axial Distance Variation of Inlet Throat Area With Centerbody Position Parameter 6 Inlet-to-Inlet Shock Interference Test Model Components 7 Schematic Diagram of Inlet Airflow Passages 8 Schematic Drawing of Inlets With Phase I & I1 Instrumentation Model Installation in Supersonic Wind Tunnel 10 Interference Plate and Support Strut 11 17 Inch by 17 Inch Supersonic Wind Tunnel 12 Stagnation Pressure Versus Run Time 13 Reynolds Number Versus Mach Number and Altitude for Intermittent Flow Wind Tunnels 14 Variation of Total Pressure Recovery with 2 3 F r e e S t r e a m Mach Number 15 Variation of Mass Flow Ratio With Free Stream Mach Number ILLUSTRATIONS (Cont'd) P a g e Figure 16 Variation of Centerbody Bleed With Free 25 Stream Mach Number 17 Contraction Ratio Versus Free 2 5 Stream Mach Number 18 Theoretical Mass Flow Ratio as a Function 2 6 of Centerbody Position 19 Downstream Inlet Pressure Recovery When Operating With Upstream Inlet Shock Expelled 20 Test Results Showing Stability and Expelled Shock Positions and Shapes 2 1 Selected Schlieren Photos Showing the Various Shock Shapes Encountered During l l B ~ ~ ~ " Cycles 22 Selected Frames from High Speed Schlieren 36 Movies Showing Expelled Shock Motion Schlieren Photographs Showing the 23 38 Effectiveness of the Interference Plate 24 Selected Frames from High Speed Schlieren Movies Showing Effects of Interference Plate on Downstream Inlet Stability Oscilloscope and Oscillograph Tracings Showing 4 2 "Buzz" Frequency and Pres sure Amplitude 26 Oscillograph Tracings Showing "Buzz" Frequency 4 3 and Pressure Amplitude TABLES Table 3 3 1 Tabulated Summary of Test Points on Figure 20 vi i LIST OF SYMBOLS Flow area Physical cowl-lip capture area Capture flow a r e a at free stream conditions Internal pas sage area
Gravitational constant = 32. 2 f t lb
lb sec2 O v e r a l l m a s s flow ratio Contraction ratio = A3/A1 M Mach number 1 1 2 W a G A Mass flow function,
-
m fi value at Mach number 1. 0, equals 0. 5318 Boundary layer bleed flow Mass flow ratio at subsonic diffuser exit Upstream inlet throttle position Pitot pressure, local Static pre 6 s u r e Total pressure, average Flow coefficient Radius measured from model axis ix
G a s constant for air = 53. 3 f t lb/lb O R
Reynolds number based on inlet cowl lip diameter Balance throttle position (used for downstream inlet) Total temperature Actual airflow Axial distance measured from tip of centerbody Axial distance measured from tip of centerbody to cowl lip leading edge Distance from cowl lip Ratio of specific heats for air = 1.4 Angle in degrees measured from the plane of the expelled shock cowl lip to the closest point on the adjacent inlet cowl lip (see Figure 30) Centerbody initial half angle Conical shock half angle Radial distance between pods measured in units of inlet diameters Subscripts Apply to the Above Symbols 0 Free stream conditions Capture cowl lip station 3 Inlet throat 7 Compressor face station 8 Plenum at aft end of model Balance throttle throat station Downstream inlet D o r DSI U o r US1 Upstream inlet X 1 1 1 1 I 1 . 1 I ABSTRACT A s flight speeds increase, supersonic aircraft powerplant installations may encounter complex aerodynamic problems due to shock reflection and the possibility of inlet shock instability created by pod-to-pod inter- ference. This report describes the experimental program conducted and presents wind tunnel test data obtained during a comprehensive study of adjacent inlet interference. This test program established, for a typical Mach 3 . 0 design mixed compression inlet, the dividing line between the regions of expelled shock interference and completely stable operation. A study was also conducted to determine the effect of placing a plate between pods.
xi SUMMARY AND CONCLUSIONS The results of this experimental program can be summarized as follow 8 : 1. It is possible to operate a mixed compression inlet in the region of influence of another inlet which is "buzzing", but it must b e operated at a reduced contraction and hence a reduced total pres- sure recovery.
The pressure recovery penalties encountered were large for steady 2.
state operation behind the "buzzing" inlet, but the penalty required to maintain stability'during the initial transient of the onset of IlbuzzlI was so severe that it cannot be considered practical.
3. The line of demarcation between the regions of stable and unstable operation was found to approximately coincide with the shape of the expelled "buzzing" shock in its maximum forward position.
4. The pressure recovery penalties encountered while operating in the unstable region behind the expelled shock were reduced as the dis- tance between the inlet cowl lip and the source of the expelled shock was made very large. The penalties were also reduced as the Mach number was reduced.
5. The use of a plate as an interference shield was very effective in eliminating the penalty imposed by the expelled shock.
6 . The magnitude of the initial pressure pulse at the onset of "buzz" is the same as it is for subsequent cycles.
The "buzz" frequency increased as model airflow was reduced 7.
(throttled) but the pressure impulse amplitude remained approxi- mately constant.
8 . The basic "buzz" frequency was dependent upon the volume in the model between the cowl lip and the choked throttle rather than the volume from the cowl lip to the throat of the inlet.
INTRODUCTION Inlet development for supersonic aircraft presents many problems to both airframe and engine manufacturers. A thorough investigation of all performance penalties associated with the various types of air induction systems must be undertaken before the proper engine spacing can be selected. One of these penalties could come from the safety factors needed to overcome inlet-to-inlet interference effects.
For many applications the podded powerplant concept rather than the airframe integrated system has many favorable features. However, individually podded engine configurations on a supersonic aircraft can experience shock interference between adjacent inlets. Inlet instability may result which could prove severe enough to create a critical flight condition. Oscillations of the inlet expelled shock wave could be created which may cause a flameout or damage the engine. Furthermore the oscillating shock may adversely affect an adjacent inlet. Safety factors incorporated for the purpose of preventing inlet-to-inlet interference result in a thrust loss and an increase in the specific fuel consumption.
Economics demand that any penalties be eliminated or minimized. The program undertaken was designed to establish the pressure recovery penalties for operating a supersonic inlet in the flow field of an adjacent subcritical inlet, and to establish the non-interference envelope.
A very preliminary program, reported in Pratt & Whitney Aircraft TDM-1753, entitled "The Effect of an Interference Shock on the Per- formance of a Mach 3. 0 Axisymmetric Moveable Centerbody Inlet", provided some insight into inlet performance penalties which can occur when podded powerplants a r e not properly aerodynamically positioned.
A more comprehensive program which better defines these inlet-to- inlet shock interference effects has been studied under the subject contract.
The parameters investigated include the total pressure recovery at the compressor face, inlet mass flow, and inlet contraction ratio for various pod-to-pod axial and lateral spacings. In addition, fre- quency and pressure amplitude were measured. The possibility of using a plate placed between adjacent pods as a means of reducing shock interference effects was also investigated. Testing was accom- plished by using two inlets operating simultaneously in the United Aircraft Corporation Research Laboratories 17 inch x 17 inch super- sonic wind tunnel facility between Mach numbers 2 . 0 and 3. 0. The Reynolds numbers based on the inlet capture diam'eter, ReD1, in the wind tunnel test section is shown in Figure 1.
FREE STREAM MACH NUMBER-Mo
Figure 1 . Variation of Wind TunnelReynoldsNumberwith F r e e S t r e a m Mach Number DESCRIPTION O F MODELS Model Geometry Two axisymmetric mixed compression inlet models designed for Mach 3.0 cruise operation were selected for this experimental pro- g r a m , ( s e e F i g u r e 2). These inlets were scaled down versions of a previously calibrated axisymmetric mixed compression inlet having an airflow schedule compatible with the demands of a typical turbojet airflow schedule, They were 2.4 inches in cowl lip diameter with initial half cone angles of 12 degrees followed by a gradual amount of contraction to the throat. Figure 2 is a photograph of the inlets mounted in the wind tunnel with the cowls removed to show the construction de- tails. A centerbody ram type scoop was used on both inlets for removal of the boundary layer from the cone surface. The centerbody bleed flow was collected internally and discharged through four struts directly to t h e f r e e s t r e a m air as seen in Figure 2. This was followed by a n e a r l y constant throat area passage and a divergent subsonic diffuser passage.
Figure 3 shows a schematic drawing of the model with tabulated inlet coordinates. The cowl of the downstream inlet model had 15 rows of perforations slanted at an angle of 30 degrees to provide for the removal of boundary layer from the internal surface of the cowl. There were never more than 10 rows utilized at any one time during the program.
In the interest of economy, the upstream inlet cowl did not have any bleed perforations as it was felt that the "expelled shock" performance of this inlet would not be materially different with or without the cowl bleed. The performance of the downstream inlet was the most critical for this type of program. With the exception of the cowl perforation bleed the two inlets were identical. Figure 4 shows the ratio of internal passage area to inlet capture area, Ap/A1, in the Mach 3.0 design PO- sition. The variation of throat area ratio, A3/A1, as a functionof centerbody position, R1/Xc is shown in Figure 5.
The inlet centerbodies were movable and were actuated by Acme drive s c r e w s w h i c h i n t u r n w e r e d r i v e nby 50 volt miniature motors. The motor movements were controlled by remote servo power positioners through the feedback part of dual helical potentiometers. The indicator half of the potentiometers were connected to the Bristol charts for visual record and then to the encoder for a punch card output. This in turn was fed to the data reduction program.
Figure 2 . Model InstallationinSupersonic Wind Tunnel - M a c h 3 . 0
CenterbodyPositionwithCowlsRemoved - $ = 2 . 0
Diameters - 8 = 5 6 . 5
. 032 in. Diam.
Holes \/
I I
CENTERBODYCOORDINATES X ( I N . ) R (IN. ) COWL PERFORATIONS COWLCOORDINATES LOCATED AT 0 0 X (IN.) R (IN.)
STRAIGHT’ LINE Z (IN.)
3.947 .839 1. 189 2 . 8 0 1 L.E.
4.043 .85 3 1.161 1. 193 2 . 9 8 6 4.122 .860 1.267 1. 194 3 . 144 4.228 .860 1.373 1. 193 3.303 4 . 3 3 3 . 8 5 8 1 . 4 7 9 1. 183 3.5 14 4 . 4 3 9 . 8 5 1 1.585 1. 160 3.779 4 . 4 9 2 . 8 4 1 1 . 6 9 1 STRAIGHT LINE 4.465 .849 1.797 1.018 5 . 0 2 0 4.465 B. L. Scoop . 8 6 9 1 . 9 0 2 5 . 2 5 8 . 9 8 9 4.545 .867 2.008 . 9 7 6 5.417 4. 677 .852 2.
5. 681 . 9 6 4 4.809 Throat . 8 3 1 2 . 2 1 9 . 9 5 9 5 . 9 4 5
. 8 0 8 . 2.325
4.941 6.474 . 9 7 0 5 . 0 7 3 . 7 8 6 2 . 4 3 1 1 . 0 1 1 7 . 0 0 2 5 . 3 3 7 . 7 3 6 2 . 5 3 7
5 . 6 0 1 . 692 2.642
5 . 8 6 6 . 653
STRAIGHT LINE 6 . 8 0 0 Compressor. 52 1 Face Sta.
Figure 3. InletCoordinates With Centerbody at Mach 3.0 DesignPosition MACH 3 . 0 DESIGN POSITION 62 6.6 5.4 58 4.2 4.6 5.0 3.0 3.4 38 Figure 4. Ratio of Inlet Internal Passage Area to Inlet Captur.e Area Versus Centerbody Axial Distance M a c h 3.0 Design Position STATION 0 CONICAL SHOCK I 3 COWL
f
STREAM WVABLE
TUBE "i
CENTERBODY \ - M O c I
i
I I
CENTERBODY POSITION PARAMETER N R I / X ~
Figure 5. Variation of Inlet Throat Area With Centerbody Position P a r a m e t e r F i g u r e 6 shows a photograph of the inlet model components utilized during this program.
The downstream inlet model was attached to a 4 . 7 inch diameter tunnel shaft by a rotating connection. This type of arrangement allowed the downstream inlet to be throttled normally by the tunnel's movable plug, Figure 7 . Rotation of the downstream inlet caused the spacing between the inlet models to vary (see Figure 8 ) . The upstream inlet was at- tached to the tunnel floor by means of a base plate and track combina- tion. Translation along this track enabled the axial distance between the inlets to be varied. The combinations of the downstream and up- s t r e a m mounting configurations allowed for an infinite number of po- sitions to be attained for testing the models. An independent method of throttling the upstream inlet was employed. A sleeve arrangement, which slid over the outer surface of the cowl, was driven by a motor encased in the rear section of the upstream model (see Figures 8 and 9').
T h e r e w e r e six internal air passage discharge ports that were used in conjunction with this throttling device.
A plate w a s mounted on the upstream inlet to evaluate the feasibility of shielding as shownin Figures 9 and 10. This would enable the down- stream inlet to operate normally when a shock was expelled from the adjacent upstream inlet.
Instrumentation Figures 7 and 8 show the various types of instrumentation employed during the complete experimental program.
The downstream inlet airflow was measured bya calibrated 3 square inch movable plug throttle. The airflow on the upstream model w a s throttled independentlybutwasnotmeasured.Massflowsforthismodelwere estimated from known centerbody positions and centerbody bleed flows.
A pitot rake with three probes and a static tap was installed on the down- stream inlet at the compressor face (station 7 ) . Static taps were also placed in both model plenum. sections (station 8). A single pitot probe was placed in the centerbody bleed discharge passage of the downstream inlet and in the upstream inlet model at the simulated compressor face.
The pitot probe placed in the centerbody bleed discharge strut was used to calculate the amount of centerbody bleed flow which was bled off by the ram scoop. This was accomplished by using the following equation: By continuity, The bleed flow is assumed to be choked at the discharge exit. PTb1, P T ~ , Ab1, and A1 are measured quantities and % is a function of Mach number. Qbl was estimated at 0.80 based on previous experience with a similar model.
The pitot rake, consisting of three probes, and the static tap located at station 7 of the downstream inlet was used to calculate the average total pressure recovery. Two methods of evaluating the total pressure recovery were used: (1) continuity average total pressure and ( 2 ) a r e a weighted average integration. The former method was determined from measurements, at station 7, of the static pressure, of the area, of the airflow, and of the total temperature.
The following relationship illustrates the method of calculating the continuity average total pressure recovery: The average Mach number at station 7, M7, must be evaluated before P T ~ can beevaluated.Therefore, Then m7 = A7 p s 7 and m7 determines M7 which, in turn, determines PT7/PS7.
1 0 I Then The latter method, area weighted average integration, was accomplished by integrating the total pressure profile across the passage area.
It was the original intention to use the static pressure in the plenum, station 8 , as a m e a s u r e of total pressure recovery. However, appar- ently because of higher than expected velocities and a distorted flow profile in the plenum, the total pressure recoveries of both downstream and upstream models were found to be erroneous. Therefore, the pitot probe in the upstream inlet at station 7 was added (serving pri- marily to determine when the normal shock was expelled) and the rake in the downstream inlet at station 7 was used to determine total pres- s u r e r e c o v e r y .
A static tap at downstream inlet station 7 (different circumferential location than that used to calculate the continuity average total pressure) was used to measure"buzz" frequency and pressure amplitude of the ex- pelled shock by means of a close coupled connection to a differential transducer.
I 111 I I INTERNAL AIRFLOW
PASSAGE 7
CENTERBOOY STREAM IWLET THROTT (PARTIALLY CLOSED) THROTTLE ACTUATOR ARMS’ UPSTREAM INLET BASEPLATE AND MOUNTING BRACKET Figure 6. Inlet-To-Inlet Shock InterferenceTestModelComponents ALL DIMENSIONS IN INCHES AIRFLOW EXIT TO ATMOSPHERE Figure 7 . SchematicDiagram of Inlet Airflow Passages
"I- " I
Figure 8. Schematic Drawing of Inlets With Phase I & I1 Instrumentation
Figure 9. Model InstallationinSupersonic Wind Tunnel - Mach 3.0
Centerbody Position-Interference Plate in Aft Position -
$'= 2 . 0 Diameter - 8 = 5 6 . 5 "
A?
f
2.500 A
'f
0.120 ALL DIMENSIONS IN INCHES
#
SECTION A - A 0 . 5 0 0 * t = I . 5 5 0 -+t-"0.500 Figure 10. InterferencePlateandSupportStrut DESCRIPTION O F TEST EQUIPMENT 1. GeneralDescription of 17 InchSupersonic Wind Tunnel The tests were conducted in the United Aircraft Laboratories 17 inch x 17 inch intermittent-flow (pressure blowdown) supersonic wind tunnel facility. This tunnel provides a uniform flow at Mach numbers ranging from 1.5 to 5.0. The nozzle, Figure 11, in this tunnel consists of a pair of flexible plate whose contours are adjusted by means of hydrau- lic jacks. Dry air is supplied to this tunnel at a total temperature of approximately 80" F and at stagnation pressures up to 400 psig. Curves showing the available run time as a function of tunnel stagnation pres- sure and Mach number are presented in Figure 12. The corresponding Reynolds number envelope i s p r e s e n t e d i n F i g u r e 13.
An inlet mount shaft having an outside diameter of 4. 70 inches was utilized. The shaft was mounted on a variable angle-of-attach sector and airflow was ducted through the inlet model and sector cavity to a constant-area section just upstream of a choked flow-measuring bell- mouth whose area is varied by a hydraulically operated plug throttle (see Figure 7). Twelve channels of digital information are recorded continuously on strip charts and punched simultaneously into IBM c a r d s a t the r a t e of 100 cards per minute. A complete description of the equipment and procedures available in the supersonic wind tunnel is described fully in reference 6. A single-pass schlieren system, which incorporates a 20 inch parabolic mirror and a schlieren viewing screen, was used for observing and photographing flow characteristics.
Figure 11. 17 x 17 InchSupersonic Wind Tunnel - Side Wall Removed
to Show Flexible Walls and Actuating Mechanisms S U P E R S O N I C TUNNEL-289 SO. I N .
STAGNATION TEMPERATURE-BOOF I N I T I A L RESERVOIR T E M P E R A T U R E - 8 0 ° F I N I T I A L RESERVOIR PRESSURE- 415 PSIA 40C X- MINIMUM STARTING PRESSURES AT VARIOUS MACH NUMBERS CONSTANT MACH NUMBER
- 30C
" CONSTANT REYNOLDS NUMBER m a u I w U cn
2 200
a a z F 150 a z a I- o 100 0 20 40 60 80 100 I 2 0 140 RUN TIME-T-(SECONDS) Figure 12. Stagnation Pressure Versus Run Time MAXIMUM TOTAL H E A D " 7 3 . 5 P S l A 2.5 MINIMUM RUN OF 2 0 SECONDS.
2.0 1.5 I. 0 MINIMUM PROBABLE OPERATING PRESSURE 0.5 MINIMUM OPERATING TOTAL HEAD 0 2 4 6 0 I O MACH NUMBER Figure 13. ReynoldsNumberVersusMachNumberandAltitudefor Intermittent Flow Wind Tunnels . " . ". .. ... . .. . - .. .
2. Special Equipment Several types of special equipment were utilized during the test pro- gram to record the "buzz" frequencyandpressureamplitude of the inlet model. A Tektronix type oscilloscope was connected to a differential pressure transducer mounted at the compressor face, station 7. This unit was completely self contained and required no external equipment other than the transducer. Excitation voltage for the strain gage type transducer was provided by a plug-in unit. The transducer consisted of a four strain gage bridge having a range 50 psi differential and a f r e - quency response of 5 . 1 KC. The duration of the oscilloscope trace was limited for the length of trace required. A complete trace from throttling through the "buzz" sequence was desirable. Therefore, sub- sequent runs were made on various types of oscillographs, (i. e.
Sanborn and Visicorder). The Sanborn type oscillograph was a mechani- cal galvanometer direct writing recorder which provided a continuous t r a c e of the test run. The Visicorder was a multichannel, general- purpose oscillograph direct recording instrument. It is commonly referred to as a light beam interrupter-type identifier.
A high speed rotating prism camera (Fastax) was used on several selec- ted configurations to record the actual physical movements of the ex- pelled shock through the schlieren system. This is the type of c a m e r a where the film and the image are moving at the same speeds due to the rotation of a prism within the camera. The film speed was approxi- mately 1500 frames per second. A Polaroid still c a m e r a w a s a l s o c o n - nected to the Schlieren visualization system to record the shock struc- ture. This was accomplished by connecting the shutter system of the camera to the spark source allowing the spark to be tripped whenever the shutter was opened. The exposure time of this camera was approx- imately four microseconds.
I DATA ANALYSIS The purpose of this test program was to establish the effects of operat- ing a high performance external-internal (mixed) compression inlet in a region of influence of an adjacent inlet which is operating with its nor- mal shock expelled and "buzzing". The inlets used were axisymmetric, movable centerbody inlets, capable of demonstrating the high perform- ance characteristics which were desired.
The program was conducted in two phases (I and II) between which slight modificationsweremadetothemodels.Fordiscussion,thecomplete experimental program can be divided into five main subjects. These will be discussed in detail as follows: ( 1 ) calibration of the basic inlets with no inlet interference effects, ( 2 ) determination of the steady state per- formance penalties of an operating inlet in the presence of an unstarted ("buzzing") inlet, ( 3 ) establishment of the line of demarcation between inlet to inlet interference and no inlet to inlet interference, (4) the effect of placing a plate between inlets, and (5) the measurement of "buzz" frequency and pressure amplitude as well as shock motion during "buzz" cycle.
Items 1 and 3 were determined during both Phase I and Phase 11. Item 2 was determined during Phase I of the test program and the remainder was determined during Phase 1 1 .
1. Calibration of theBasicInlet The inlet configuration was selected from previous tests because of its generally good performance over the Mach number range from Mo = 2 . 0 to Mo = 3.0. Figures14, 15, 1 6 and 17 a r e p r e s e n t e d showingthepres- s u r e r e c o v e r y , m a s s flow ratio, centerbody bleed and contraction ratio of the reference inlet. Also shown with the reference inlet is t h e p e r - formance of the smaller scale inlet used in this test program. In P h a s e I, the cowl boundary layer was bled from the aft 10 rows of cowl per- forations. In P h a s e 11, the front 10 rows of cowl perforations were used.
In scaling the 4. 5 inch diameter reference inlet to the 2.4 inch diameter inlets used in this program, the perforation diameter could not be scaled because of practical machining problems. During the Phase I testing, it was noted that large amounts of airflow spillage were present at Mach 3 . 0 a s the peak recovery was approached during the throttling of the inlet airflow. It was further noted that the large decrease in mass flow ratio
Mo * 3.0 DESIGN
BC = 12.
-. .
2 . 0 2.5
FREE STREAM MACH NUMBER-M, M . . ‘3.0 RADIPL DISTANCE FROM CENTERBODY-R- IN.
Figure 14. Variation of Total Pressure Recovery With Free Stream MachNumber and Compressor Face Pitot Pressure Profiles was occurring as the normal shock approached the throat region and forced larger amounts of flow through the cowl perforations. This in- creased bleed produced a subsequent reduction in the overall aerodynamic contraction ratio. The testing in Phase I1 was accomplished utilizing the forward 10 rows of cowl perforation bleed by moving the last few rows out of the throat region to avoid this excess spillage. This caused a reduction in bleed, and an increase in contraction ratio, however very little change in pressure recovery resulted. Although the contraction ratio was low while using the downstream series of bleed perforations, the increased cowl bleed tended to compensate, probably because of a n improved radial profile as shown in Figure 14. Apparently, the com- bination of the cowl bleed scale factor and the lower test Reynolds number for the 2.4 inch diameter inlets caused the slight reduction in the per- formance from the reference inlet. However, the performance was sufficiently high to establish realistic interference effects, since both inlets were operating with large amounts of internal contraction.
0 187-SC86-085
2.5 .
FREE STREAM MACH NUMBER -Mo
Figure 15. Variation of Mass Flow Ratio With Free Stream Mach
Number - Mach 3 . 0 Design - 8, = 12"
0 0.08 W W A
m
*IO L # > E .0.04
s,- K
O 187-SC86-085
W 0 k
A PHASE I
0 PHASE IL
W ~ " 0 0 2 .Q 2.5 3.0
FREE STREAM MACH NUMBER-Mo
Figure 16. Variation of Centerbody Bleed With Free Stream Mach
Number - Mach 3 . 0 Design - 8, = 12 O
2 .o 2.5 3 .O 3.5
FREE STREAM MACH NUMBER "Mo
Figure 17. Contraction Ratio Versus Free Stream Mach Number The data shown in Figures 14, 15 and 16 are for the downstream inlet only. The contraction ratio of the upstream inlet is shown in F i g u r e 17 where it can be seen that only slight differences were present between the upstream and downstream inlets. The contraction ratio for the down- stream inlet was determined from the measured mass flow ratio passing t h e c o m p r e s s o r f a c e s t a t i o n ( s e e F i g u r e 1 5 ) a n d t h e t h r o a t area ratio associated with a measured centerbody position (See Figure 5). It therefore does not take into account any mass flow bled off between the throat and the compressor face. The Phase I inlet configuration had s o m e cowl bleed perforations between the throat and the compressor face and therefore the contraction ratios shown in Figure 17 f o r P h a s e I are somewhat misleading. Since the cowl bleed was not metered, it i s not possible to make an allowance for this discrepancy, however; it will probably account for most of the difference between the Phase I and I1 contraction ratios shown in Figure 17. The contraction ratio of the upstream inlet was determined by using the measured centerbody position to obtain ( 1 ) o v e r a l l m a s s flow ratio from theoretical curves (see Figure 18), and ( 2 ) t h e t h r o a t a r e a r a t i o f r o m F i g u r e 5 . T h e t h r o a t m a s s flow was determined by assuming that the centerbody bleed (Figure 16) was the same for both inlets, and by subtracting this quan- tity from the overall mass flow. The cowl bleed was equal to zero.
The following centerbody positions were used for both inlets as well as for the reference inlet.
MASS FLOW RATIO -A, /A, F i g u r e 18. Theoretical Mass Flow Ratio as a Function of Centerbody Position Centerbody Position ( R 1 / X c ) Downstream 187-SC86-085 Phase I Phase I1 Upstream .336 .316 .330 Mo = 2 . 0 .328
. 378
Mo = 2 . 5 .353 .362 .360 Mo = 3.0 .441 .423 .424 .413 A s a point of interest, test points were attempted for Mo = 1.7 1 and 1.88 but the blockage and total pressure losses created by this relatively large frontal area per square inch of tunnel flow a r e a would not allow the tunnel to remain started. Consequently, no valid test points were recorded below Mo = 2.0.
2. Determination of PerformancePenalties The intent of this experimental program was to map out the penalties for operating an inlet at several locations in the region of influence of an unstable inlet. A t several Mach numbers, various combinations of pod spacings were tested to determine the penalty loss which occurs when operating in this condition. Figure 19 i s a plot of the tested com- binations and shows the performance loss as a function of the displace- ment distance parameter from the inlet to the expelled shock, G/D1.
The displacement distance parameter, G/D1, is defined as the axial distance measured from the inboard leading edge of the downstream inlet cowl lip to the intersection of the "no-penalty'' line at the specific
podspacing, # . This"no-penalty''lineisshowninFigure20for
various free stream Mach numbers and is the line which represents the maximum forward travel position of the expelled shock. A more descriptive analysis of this "no-penalty" line is given in the succeeding section entitled, "Establishment of the No Interference Penalty Line. I ' The procedure used to establish the pressure recovery penalties shown in Figure 19 was to, first.cause the upstream inlet shock to become ex- pelled by throttling the inlet airflow (throttle P), then re-establish a new maximum contraction ratio for the downstream inlet. The pressure recovery and mass flow ratio for the downstream inlet was then obtained withtheupstreaminlet'sshock"buzzing".Thisprocedurewasrepeated for variations in pod spacing from 1. 0 inlet diameter to 3 . 0 inlet dia- meters, and for the Mach number range Mo=2.0 to 3.0.
1 . 0
a '
\ $0,
& '
W
> <
0 0.8 W
a
-
0.7 v) v) W
a
e
I -
J 0 . 6
I- .
Q50 0.5 I .o I .5 2 . 0 2.5 3 . 0 35
RELATIVE POSITION OF EXPELLED SHOCK AND -AM INLET-GID, Figure 19. Downstream Inlet Pressure Recovery When Operating With Upstream Inlet Shock Expelled An examination of the data of F i g u r e 19 shows several trends. The greater the distance away from the source of the expelled shock, the l e s s thepenalty.AtMach3.0,forinstance,themaximumpenalty at $ ' = 1.0 diameter was approximately 2 2 percent relative to the no
interferenceperformance.Anincreaseinspacing to + = 3.0 d i a m e t e r s
reduces this penalty to approximately 7 percent. It is evident that the strength of the expelled shock plays an important role in the determina- tion of the magnitude of the penalty. A similar trend is also noted at the other Mach numbers.
As the distance away from the expelled shock is increased, by increas- ing the axial spacing for a given lateral pod spacing, the penalty gets initially more severe until it reaches a maximum value. However, as the distance is further increased, the penalty is lessened, until at v e r y great distances, it is apparent that no penalty will be incurred. The trends noted agree very well with the preliminary data recorded in P r a t t & Whitney Aircraft report TDM 1753 (reference 1).
The results of Figure 19 further show that as t h e f r e e s t r e a m M a c h number is reduced, the maximum penalty incurred is also lessened.
The trends noted lead to the conclusion that the weaker the interference shock, either from lower free stream Mach numbers or greater distances from the source, the less the operating penalty. Such a conclusion is not contrary to the normal trend in supersonic flow.
An interesting aspect of the above test results is that the mass flow ratio of the downstream inlet did not vary more than 1 to 2 percent from the values obtained with no interference, even though the center- body position parameter, R1/Xc, and the pressure recovery did vary.
R1/Xc is tabulated as follows for the various points shown in Figure 19.
Mo J.
2 . 0 1.0 .319 2 . 0 3.0 .306 2 . 5 1.0 .364 2 . 5 .324
3 . 0 1 . 0 . 364
3.0 3. c .397
It should be emphasized that the results of Figure 19 indicate the per- formance levels which can be maintained by the downstream inlet with the upstream inlet "buzzing", after the initial shock expulsion of the upstream inlet. This would represent a condition on an aircraft where restarting of the upstream inlet was not possible through a control or mechanical malfunction.
The transient effect of unstarting the upstream inlet was found to impose a much more severe operating penalty. The upstream inlet was restarted while the downstream inlet remained operating at the contraction ratio previously determined with the upstream inlet I'buzzing". However, as soon as the upstream inlet shock was expelled, the downstream inlet shock was also expelled. High performance inlets appear to be much more sensitive to a step change in pressure than to a continually fluctuat- ing p r e s s u r e (i. e. , "buzzing"). This indicated that a much lower operat- ing pressure recovery level would be required to allow for the most severe transient. It is possible that the amount of contraction in the downstream inlet could be reduced enough to prevent the shock from becoming expelled when operating in the unstable region, but, the resulting pressure recovery would most likely be too low to w a r r a n t locating inlets where this penalty must be imposed. It was, therefore, decided to forego further investigationsof the penalties imposed in order to maintain operation of the downstream inlet at the instant of the shock expulsion of the upstream inlet, and to concentrate on the determination of the line of no interference penalty.
3 . Establishment of the No Interference Penalty Line Probably the most important information obtained in this experimental program was the establishment of the line of no interference, which distinguishes the region of stable and unstable operation of an inlet which is positioned near an adjacent inlet whose shock is expelled. The location of this operating line limit is needed in or.der to position inlet pods on supersonic aircraft without a penalty for interference.
The procedure used to establish this line was as follows: With the upstream inlet started (shock swallowed), the downstream inlet was operated at its maximum contraction ratio and throttled by the balance throttle,S,toapproximately its peak pressure recovery. (These values were previously determined from the model calibration proce- dure, described fully in the model calibration section.) The upstream inlet shock was then expelled by throttling the airflow with throttle, P, resulting in the downstream inlet either remaining stable or going into a "buzzing" state. Due to the lack of a control system it was not pos- sible to maintain exactly the peak pressure recovery on the downstream inlet and the normal sequence of running was to allow this inlet to run one to three percent supercritical. It was determined during this test program that this amount of supercritical operation did not have any noticeable effect on the stability of the inlet or the position of the no interference penalty line.
The tests were conducted at Mach 2. 0, 2. 5 and 3.0 for various lateral
pod spacings, J. , (1. 0 to 3. 0 inlet diameters) and axial distance
p a r a m e t e r s , 8, (-5" to t70").
The angle 8 is measured to the inboard edge of the cowl lip illustrated in phantom on Figure 20 at 6 8 " . F i g u r e 20 also shows the shape of the line of demarcation between inlet to inlet interference and no inlet to inlet interference. The test points are shown for both the stable and unstable operating points.
The no interference line continues to sweep back for constant Mach number until it approaches approximately a Mach line at the larger spacings.Notethatalthoughtheshock is veryweak at 9 = 3.0 d i a m e t e r s , its effect on inlet stability was still noted.
An interesting feature of these no interference lines is their similarity inshapeforthevariousMachnumberstestedThedisplacement difference from Mach number to Mach number is approximately the s a m e a s the centerbody extension. This simplifies the requirements for the re-establishment of the line in the event that the model geometry is changed or the environmental flow conditions are different.
The shape of the no interference line approximates the most forward t r a v e l of the "buzzing" expelled shock. Figure 21 shows several shock positions which were photographed with a still camera through the Schlieren viewing system. A high speed light source stopped the buzz- ing shock in these various positions. It is interesting to note that the shock apparently does not move in and out in the same uniform manner during each "buzz" cycle. Only by coincidence would the still schlieren photos record the shock in its most forward position. Even the high speed movies, from which selected frames are shown for one "buzz" cycle in Figure 22, w e r e not useable for getting a n accurate location of this no interference line. The sensitivity of the inlet itself made the best indicator. The resulting shape when compared to the still photos and the high speed movies, tends to substantiate that it repre- sents the envelope of the most forward positions encountered for the "buzzing" shock.
EXPELLEDSHOCKUPSTREAMPOSITIONSANDSHAPES (LINESOF NO INTERFERENCEPENALTY)
w
e =oo IO" 20" 30° 40°
""" $ (DIAMETER) =3.0(
2 .o
EXPELLED SHOCK PHASE I 0 Mo = 2.0 STABLE Mo = 2.0 WSTABLE
0 M O = 25 STABLE Ma = 2.5 UNSTABLE
0 M i = 30 STABLE M i = 3.0 UNSTABLE
I . FLAG SYMBOLS -PHASE II TEST POINTS (i.e.b,b,o)
2. "PHASE 1, INTERFERENCEPLATE FWD POSITION 3. <-PHASE IT, INTERFERENCE PLATE AFT POSITION (SEE TABLE I ) Figure 20. Test Results Showing Stability and Expelled Shock Positions and Shapes " TABLE 1 Tabulated Summary of Test Points on Figure 20 fi = 1 . 0 Diameter Conditions Phase M o
-
I1 stable -4 O 2.0 11 stable 0" 2.0 I1 marginal 5" 2.0 I unstable 54 0 2.0 unstable 2.5 1 4 . 5 O I stab le 3. 0 0" I1 stable 12 O 11 3. 0 22.5' 11 stable 3. 0 unstable 3. 0 28 O I1 unstable 3.0 32 O I1 unstable 3. 0 54 O I J. = 2. 0 Diameters Conditions Phase
e Mo
stable 2.0 3 0 O I stable 2.0 36. 5" I marginal 41.5' I 2.0 unstable I 2.0 48.5 O
stable - plate aft
I1 2.0 4 9 " unstable 2.5 49.5 O I unstable 2.5 52.5 O I
stable - plate aft
2.5 55 O 11
stable - plate forward I1
2.5 55 O unstable 2.5 68 O 1 .
stable 3. 0 4 2 O stab le 3. 0 44 O marginal 3. 0 4 5 O unstable 3. 0 55"
I
Conditions
e Phase
M o
-
-
-
stable - plate aft
5 5 " I1
3. 0
stable - plate forward
3. 0 5 5 11
+ = 3. 0 Diameters
e Phase Conditions
M o
- -
stable 1 0 " I 2.0 11 stable 2.0 2 9 O 4 8 . 5 " II stable 2.0 11 marginal 2.0 5 1 " unstable 5 1 . 5 " I1 2.0 I unstable 2 . 0 5 9 . 5 " 0" stable 3. 0 I 10" I stable 3. 0 stable 3. 0 56 O I1 58 O I1 unstable 3. 0 unstable 3. 0 I1 59 O I unstable 3. 0 59 O Figure 21. Selected Schlieren Photos Showing the Various Shock w Shapes Encountered During "BUZZ" Cycles ul w J. =1.0 c) = 2 8 " t = 0 Sec. t = ,000645 Sec t = .00129 Sec. t = .00194 Sec. t = . 0 0 3 2 3 Sec.
t = ,00452 Sec t = .00516 Sec. t = .00645 Sec. t = .00774 S e c . t = , 0 0 8 3 9 Sec.
L ""4
L3
t = .01419 Sec. t = .01742 Sec. t = .02580 Sec.
t = .00903 Sea t = .01225 Sec.
t = .06450 Sec. t = . 10707 Sec.
t = .OS612 Sec.
t = .03612 Sec. t = . 0 4 6 4 4 Sec.
Figure 2 2 . Selected Frames From High Speed Schlieren Movies Showing Expelled Shock Motion 4. TheEffect of theInterferencePlate Following the establishment of the unstable region of operation, an attempt was made to shield the downstream inlet from the expelled shock of the upstream inlet by using an interference plate. Figure 23 illustrates the effectiveness of this plate. Without the plate, inter- ference was present which unstarted the downstream inlet when it was operated at maximum contraction. When the plate was positioned be- tween the inlets as a shield, the downstream inlet was allowed to operate normally. It was discovered that continued throttling of the upstream inlet (throttle P) after shock expulsion, caused the expelled shock to spill out around the leading edge of the plate. When this occurred, the downstream inlet became unstable. An extension of the interference plate to a further upstream position, allowed for additional throttling before the downstream inlet became unstable. Apparently, this plate would have to be tailored in size and spacing from the cowl in order to assure the best results. Figure 2 4 illustrates, with selected f r a m e s f r o m a high speed schlieren movie, the shock motion for both a partially throttled condition and for the condition which eventually triggered the downstream inlet. The effect of the plate was tested at both MO = 2 . 5 and 3 . 0 and found to have essentially the Same charac- teristic s.
At both Mach numbers, the centerbody shock reflected from the plate and back to the cowl well behind the cowl lip. No change in contraction ratio of the upstream inlet w a s noted. In the event a plate is used as a shield, care must be taken to prevent the centerbody shock from reflecting back into the inlet as this will surely affect the started performance of the inlet.
" Mo = 3 . 0
4 = 2 . 0 Dia. 8 = 50. 5O $ = 2 . 0 Dia. 8 = 55O
N o Plate Plate in Aft Position ( Unstable) (Stable) M o = 2 . 5 # = 2 . 0 Dia. 0= 52. 5O d = 2 . 0 Dia. 0 = 55O N o Plate Plate in Forward Position (Unstable) (Stable) Figure 23. SchlierenPhotographs Showing the Effectiveness ofthe Interference Plate Upstream Inlet Shock Becomes Expelled, Downstream Inlet Stable """"""""~""""""""""~""""""""""~ Further Reduction of Upstream Inlet A i r f l o w Downstream Inlet Shock Is Finally Expelled Figure 24. Selected Frames From High Speed Schlieren Movies Showing Effects of Interference Plate on Downstream Inlet Stability 5. "Buzz Analysis'' During the Phase I1 testing, "buzz" frequency, pressure amplitude, and shock motion were recorded for the downstream inlet between Mach 2.0 to 3.0. The l%uzd'frequency and pressure amplitude were recorded by oscilloscopes and oscillographs. The shock motion was also re- corded by a high speed camera described in the test equipment section.
The pressure variations were measured with a differential transducer close coupled to the model to eliminate lag.
It was initially assumed that the "buzz" frequency and pressure amplitude would be dependent on the volume between the cowl lip and the inlet throat and therefore be the same for both models. Analysis of the "buzz" frequencies showed that this was not the case. By observing the inlet expelled shock motion in the high speed schlieren movies the "buzz"fre- quency was seen to cycle at a much faster rate for the upstream inlet than for the downstream inlet. Because the next flow restriction which is encountered in the model takes place at the choked mass flow throttle, it was deduced that the frequency was dependent upon the volume between the cowl lip and this throttle. This volume was much less for the up- stream inlet than for the downstream inlet and therefore probably ac- counts for the fact that the upstream inlet frequency was approximately twenty times the frequency of the downstream inlet.
In the early stages of the "buzz" measurement phase it was found that the oscilloscope would not completely record the pressure trace. This was due to the scale factor which was available on the oscilloscope. There- fore in order to determine the complete "buzz" frequency and amplitude for the initial transient as well as subsequent cycles, the oscilloscope was replaced by an oscillograph for the remaining runs.
The "buzz" frequency and pressures are shown for the downstream inlet at Mach 2.0 and Mach 3.0 in Figure 25. It can be seen from this figure that the "buzz" frequencies are approximately the same for both Mach numbers and that the only variation is in the pressure amplitude. This might be explained by the difference in the contraction ratios or the total pressure recoveries between these two Mach numbers. The am- plitude at Mo = 2. 0 is approximately 28 percent of the free stream total p r e s s u r e a n d at Mo = 3 . 0 it is approximately 12 percent of the total p r e s s u r e . F i g u r e 26 shows that the amplitude at Mo = 2 . 5 is approxi- mately the same as the Mo = 2.0 amplitude. The peak total pressure recoveries at Mach 2.0 and 2.5 are also approximately equal (See Figure 14).
Figure 26 also shows the effect of throttling the inlet airflow on the "buzz" frequency. During this run the frequency increased from six c. p. s. to approximately nine c. p. s. It should be noted that although the frequency changed during throttling, the amplitude remained approximately constant. This trend indicates that although the "buzz1f frequency on the upstream inlet was higher than the downstream inlet, it can be assumed that the pressure amplitude remained approximately the same. Duetothelack of instrumentation on the upstream model, this assumption could not be verified.
Figure 22 shows selected frames from the high speed schlieren movies with the time indicated for each frame. Note that the shock from the downstream inlet is not expelled by the upstream expelled shock until a finite elapsed time has occurred (i. e. frame 4 to f r a m e 6). Note a l s o that the shock remains at its inner most position for a significantly longer time than it does at its maximum upstream position. The oscillo- graph trace in Figure 2 6 tends to verify this fact. It is also interesting to note in the pressure trace that the magnitude of the minimum pressure obtained in the initial transient pressure drop is the same as the pres- sure in subsequent cycles. The amplitude for all cycles is also a con- s tant .
I
OSCILLOSCOPE TRACE Pto = 138.9IN. HgA
Mo= 2.0 BARO = 30 -25 IN. HgA
INCREASING T I M E -
80r I
t
' VISICORDER TRACE Pto = 159.71N. HgA
Mo= 3.0 BARO = 30.001N. HgA
Figure 25. Oscilloscope and OscillographTracings Showing "BUZZ" Frequency and Pressure Amplitude OSCILLOGRAPH TRACINGS Mo.2.5 Pto= 149.5 IN.HgA , O N S E T OF BUZZ BARO; 30.18 IN.HgA ..-. .-. .... n . . 1 .
. . . . . . . . . . . . . .
IC--t=ISEC-j INCREASING TIME- .... ..-.
CREASING T I M E - Figure 26. OscillographTracings Showing "BUZZ" Frequency and P r e s s u r e Amplitude Recommendations 1. Based on the magnitude of the penalties which are encountered in the region of expelled shock interference, it is recommended that the pods of a supersonic aircraft be located so that no interference will be experienced while the inlet operates with internal compression.
The best arrangement for satisfying this condition is to place the pods with the cowl lips in the same plane and spaced at least 1.0 diameters apart. Probably 1. 5 diameters would be better to allow for a margin of safety.
In order to position pods in the interference region so that the penalty would be a minimum the spacing required would be s o great that they would probably be farther apart than the aircraft structure would allow.
2. In testing the plate which was used as an interference shield it was evident that this plate should be tailored to f i t the specific applica- tion. It is recommended that further work be done to investigate the minimum plate size and optimum plate location which might be po s sible.
3. Becausethemaximumupstreamposition of theexpellednormal shock defines the line of no interference, it is recommended that this information be recorded in future tests of all inlets with internal contraction. One very simple means of obtaining information of this type could be to take a time exposure on a single film plate during the 'buzz" operation. The envelope of the blurred shock image should show the maximum forward position of the expelled shock.
REFERENCES 1. Motycka, D. L. : TheEffect of an Interference Shock on the Performance of a Mach 3.0 Axisymmetric Movable Centerbody Inlet. Pratt & Whitney.Aircraft Report TDM- 1753, January, 1962.
2. T r i m p i , R o b e r t L: AirAnalysis of BuzzinginSupersonicRam Jets by a Modified One-Dimensional Nonstationary Wave Theory.
NACA TechnicalNote3695,July,1956.
3. Mirels,Harold:AcousticAnalysis of RamJetBuzz. NACA Lewis Flight Propulsion Laboratory, NACA Technical Note 3574, November, 1955.
4. Sterbentz, W. H. andDavids, J.: Amplitude of SupersonicDiffuser FlowPulsations. NACA TechnicalNote3572,October,1955.
5. Sterbentz, W. H. andEvvard, J. C . : CriterionsforPrediction and Control of Ram- Jet Flow Pulsations. NACA Technical Note 3506, August, 1955.
6. McLafferty, G. H. andVergara, R . D. : Description of Equipment and Techniques Used for Inlet Tests in UAC Research Department 17-inchBlowdownTunnels. UAC Research Laboratories Report R-2000-30, July, 1957.
NASA-Langley, 1965 CR-264 I .. ~ . -~ - .