AD-756 895

A COANDA INLET/JET FLAP DIFFUSER EJECTOR

Morton Alperin

Flight Dynamics Research Corporation

Prepared for:

Air Force Flight Dynamics Laboratory

August 1972

DISTRIBUTED BY:

Naflonal Technical Information ServiceU. S. DEPARTMENT OF COMMERCE5285 Port Royal Road, Springfield Va. 22151

Y.. m '__ __rl= l II i

AFFDL-TR-72.106

4 • A COANDA INLET/JET FLAP DIFFUSER EJECTOR

MORTON ALPERIN

FLIGHT D YNAMICS RESEA RCII CORPORATION

BURBANK, CALIFORNIA

TECHNICAL REPORT AFFDL-TR-72-106

h6p..d.i(d byNATIONAL TECHNICALINFORMATION SERVICE

U $ D*Ot-•,1ln of ComweltoStn.h., d VA 21 < ,

AUGUST 1972

i Approved for public release; distribution unlimited.

AiR FORCE FLIGHT DYNAMICS LABORATORYAIR FORCE SYSTEMS COIMMAND

z WRIGHT-PAITiERSON AIR FORCE BASE, OHIO

NOTICE

SWhan Gtovernment drawings, specifications, or other data are used for any purposeother than In connection with a definitely related Government procurement operation,

the United Mates Government thereby incurs no responsibility nor any obligation

whatsoever; and the fact that the government may have formulated, furnushed, or in

I any way supplied the sald drawings, specifications, or other data, Is not to be regardedby implication or otherwise as In any manner licensing the holder or any other personor corporr.!ion, or conveying any rights or permission to manAfacturej use, or sell any

patented Invention that may In any wAy be related thereto.

Copies of tWis report should not be returned unless return is required by securityoraldorationt contractual obligations, or notice on a specific document.

S~~AIR FORCF-/56730/7 DOO~MW~ 1972 -- ISO

I UNCLASSIFIED

Securtty CIM__fiction

DOCUMENT CONTROEW) ArA. & DrSt.rstj, etha. ,raUion at f tile. b4'N 64 Abstaet ýWj 64•,z#*.. W,*,Xwtm P •"o i• & en ted Whe.. thta-" tall ,rpa. i. Classified,

1. OftIZINATING. Ac TIVITY (CotpoatoI 4.AV") 1ASPA CCUPITY CLASIIVICATIO,4

FLIGHT DYNAM.ICS RESEAKC-1 CORPO•TZ• N UNCLASSIFIED2115 KENMEPE AVE. 2b. CROUP DNABURBANK, CALIFORNIh 91504

.• RZgPONr riI..C

A COANDA INLEt/JET FLAP DIFFUSER EJECTOR

F INAL

MAT1ORTON ALPERIN

61 "POAT OATC F9. TOTX!. NO. OF PAOCS or her&,,.*o• oFEBRUARY 1972 • 6e 7 768. CONTtACT OR GRANT KO•. 90. OHI'1NATOR'S NCPORAT NUUKOLRIS)

F33615-70-C-1656b. PROJCCY NO, TR-72-02-01-106

1366( 1 6. 0164CR ICP*ftT 6-It

AFFDL-TR-72-10 6

IC, 0ISTRIUUTION STATUChNT

APPROVED FOR PUBLIC RELEASE; DISTRIBUTION UNLIMITED

11 SUPPLICUCNTARY NOTYC I'II, *PONSOMINZ SILITANY ACIIVITV

_tIR FORCE FLIGHT DYNAMICS LABORATORYNONE "RIGHT PATTERSON AIR FORCE BASE

I1.1_10_ 45433

The combination of a Coanda inlet and jet flap diffusion, for the.ichievoment of high performance, low volume, thrust augmentation, hasIoan investigated in a two-dimensional experiment.

The Coanda/JFD Ejector has demonstratvd a performance superiority,leuivalent to 25% in thrust per horsepower, with a volume of approximatel'.5% of the volume of the highest performance solid diffuser ejector, with.onvontional inlet.

The use of jet flap diffusion also provides a mechanism for theichievemnnt of large control forces, through the use of variable jet flapangles, or through the application of indremental power to the jet flap.

Vectoring of the total thrust force is achievable through the use6f the fluidic effect, which can be utilized to detach thf, flow from

ine side of the jet flap, by decreasing the plenum pressure of that side[elative to the plenum pressure of the other side.

DD INO6.47UNCLASSIFIE,Iq Svcuit~y Classitcuon

FOREWORD

The research reported upon in this document wasconducted under Air'Force Contract No. F33615-70-C-1656,"Jet Flap Diffuser", Project No. 1366, "AeromechanicsTechnology for Military Aerospace Vehicles", TaskNo. 136617, "Aerodynamic Analysis of Advanced MilitaryV/STOL Aircraft". by Fiight Dynamics Research Corp.,Burbank, California.

Sponsorship was by the Air Force Flight DynamicsLaboratory, Air Force Systems Command, Wright-PattersonAir Force Base, Ohio.

Work was initiated in June 1970, and completed inAugust 1972, Mr. R.L. Clark, AFFDL Project Engineer,was the technical monitor for the Air Force.

During various phases of the effort, the author wasassisted in the experimental work by Dr. Gary L. Marlotte,and by Mr. G. Gonda-Bonardi.

This tachnical report has been zeviewed and isapproved.

AntonatosCh f, Flight Mechanics DivisionAi• Force Flight Dynamics Laboratory

ii.

NIP"

ABSTRACT

The combinatioii of a Coanda inlet and jet flap diffusionfor the achievement of high performance, low volume, thrustaugmentation, has been investigated in a two-dimensionalexperiment.

The Coanda/JFD Ejector has demonstrated a performancesuperiority equivalent to at least 10% in thrust per horse-power, with a volume of approximately 25% of the volume ofthe solid diffuser ejector with conventional inlet reportedin Relerence 3.

The use of jet flap diffusion also provides a mechanismfor the achievement of large control forces, through theuse of variable jet flap angles, or through the applicationof incremental power to the jet flaps.

'Vectoring of the total thrust force is achievablethrough the use of the fluidic effect, which can be utilizedto detach the flow from one side of the jet flap, bydecreasing the plenum pressure of that side relative tothe plenum pressure of the other side, or by differentialoperation of the jet flaps.

lil

TABLE OF CONTENTS

Section Title Page

I Introduction...,.......... .... .... ...... 1

Ii Summary............... ............. 2

III Apparatus ..... .. .... o. o... .... .... o....... 6

1. Ejector Rig... .... ... o.o ..... o ...... 6

a. Primary Nozzles.................. 6b. Coanda Turning Vanes.... ......... 6c. Solid Diffuser................... 6d. Jet Flap Diffuser................ 6e. Boundary Layer Control..,........ 6

2. Air Supply................. . . ...... , 7

3. Instrumentation. ...... ............... 7

IV Test Procedures ..... ..... ,..0 . . ... , 11

V Test Results .... ... 000000400..0..06 ...... 13

1. Performance Parameters............... 13

a. Optimization Parameter (Vl)...... 13b. Fixed mass flow thrust

augmentation (0 ).0.. ........... 14c. Fixed nozzle ext area thrust

augmentation (Op)9...... ........ 14

2. Performance ....... ..4 ... . . . .. . ..... . 15

a. Effect of Inlet Slotr .... ........ 15b. Effect of Solid Diffusion ....... o 16

VI Discussion., ........ ... . .... ,.. ,..... . 22

1. Performance .... .. ,., ... ,.,........ 22

a. Physical interpretation ofperformance paramcters........... 22

(1) Thrust augmentation atfixed area ( .2......... 2

(2) Thrust augmentation atfixed mass flow (0 )..,....00 23

(3) Optimization parai6ter J(.00 24

v

ilreceding pagre blank

TABLE OF CONTENTS (Continued)

Section Title PageVI 2. Comparison of C/JFD Ejector with

conventional ejectors...,..,,,.,.,, 25

3. C/JFD Ejector as a control device... 26

a. Thrust vectoring................ 26b. Thrust magnitude control........ 26

(1) Power increment response.,, 26

(2) Flap angle response......... 26

VII Conclusions....,,,, .... ,.,,,,, 27

Appendix I Determination of 'EjectorPerformance,.,,,,....,,, .,,,.,,, 28

Appendix II Test Data,,,,,,,,,,,,, ... ,, 36

Appendix III Performance Data,,,.... ,,,,, 49

References......., . .,,,,,, 64

vi

I

LIST OF ILLUSTRATIONS

Figure Title Page

1. Jet ?lap and Uolid Diffuser EjectorsEqual Total Peter. ....................... 4

2. Schematic of Ejector .. 7

3. Schematic of Coanda/JFD Ejector........... 8

4. Schematic of Air Supply System............ 10

5. Coanda/JFD Ejector PerformanceA/a-16, s/a-3,1/ 3130 deg., I 17

6. Coanda/JFD -jactor PerformanceSA/a-16,. s/awB,/3-40fi50 deg., 5oi.12...... 18

:7. Coanda/JFD Ejector PerformanceA/a-16, s/a3,3 ,4-20 deg., 5-1.12.....0.... 19

8. Coanda/JFD Ejector PerformanceA/a-16, s/&=.A1, 20 deg.,G-i. l2....... *se.e 20

9. Coanda/JFD E,"ector PerformanceA/a.16, s/au3,/Sm30 deq., -l.30........ 21

vii

LIST OF TABLES

Table Title Page

I Overall Ejector Geometriesand Performance...........................

II Comparison of Ejector Characteriotics.....

viii

LIST OF SYMBOLS

A - Minimum Channel Width

a - Primary Jet Exit Area

E - Power

F - Thrust

h - Height in Y-direction

K - Orifice Calibration Constant

n- Mass Flew

p - Pressnre

R - Gas Constant

s - Jet Flap Exit Area

ST - Absolute Temperature (deg.Rankine)

V - Velocity

x - Coordinate in Thrust Direction

y - Coordinate Normal to Thrust in Plane of Symmetry

z - Coordinate Normal to Thrust and Plane of Symmetry

oc- Inlet Area Ratio (A/a)

S - Jet Flap Half Angle

Y - Ratio of Specific Heats (C /Cp v

S7 - Solid Diffuser Area Ratio (A,/A)

- Thrust Augmentation

- Total Thrust x Vj/Total Power

Subscripts

So - Stagnation 1 - Primary T - Totalp - Fixed Power 2 - Jet Flap

- Free Stream RP- Ref. jet related to OP

Other symbols and subscripts defined in text as utsed.

ix

SECTION I

INTRODUCTION

The Coanda/JFD Ejector, is a very compact, highperformance device, capable of producing superior thrustaugmentation, compared to that of the highest performancesolid diffuser, conventional inlet ejectors known.

The inlet, which is comprised of two opposed, two-'' dimensional nozzles, utilizes the Coanda Effect, to rotatethe efflux from these nozzles through ninety degrees, intothe thrust plane.

During this rotation, the primary fluid mixes with theambient air, thereby eliminating the requirement foradditional mixing length in the ejector duct.

The Coanda turning produces a velocity distributionat the ejector inlet which assists in delaying separationin the solid diffuser (which is located at the ejector'sentrance), thereby making it possible to use short, wideangle solid diffusers.

The jet flap orifice at the downstream end of thesolid difiuser also provides a mechanism for delayingseparation in the solid diffuser, and when properlydesigned, provides a mechanism for achieving additionaldiffusion of the core flow, without solid surfaces.

These effects, in combination, result in an ejectorhaving a volume which is only 25% of the volume, andperformance which is at least 10% higher, in terms ofthrust per horsepower, than that of the conventionalejector reported in Reference 3.

The jet flap diffusion process also provides a meansfor achievement of large control forces at constant power,by increase of the jet flap angle, or with power incrementapplied to the jet flap.

The total thrust force can be directionally controlled,without vanes in the ejector efflux, by means of the "fluidic"effect, which requires only a differential change in the jetflap plenum pressure, or in the jet flap angle.

1

SECTION II

SUMMARY

The Coanda/JFD Ejector, whose configuration andperformance are described in this document, was developedin three cI• In the first phase, emphasis was placed only upon theinlet, which consisted of two primary nozzles, ejectingfluid in exactly opposite directions. These jet sheetswere rotated through ninety degrees by a pair of Coandasurfaces which were slotted to permit the induction ofI additional air mass. The optimization of these Coandainlet surfaces, for maximization of the total thrust, wasperformed without consideration of the effect of diffusion,and is reported upon in Reference 1.

During the second phase, both solid and jet flapdiffusers were added to the inlet, to determine theoverall performance of the system. In addition, a theorywas developed to predict the performance of a jet flapdiffuser ejector. Comparison of the measured performancewith the theory, indicated large discrepancies, and invest-igation of the flow field indicated serious degradationof the performance as a result of an interference betweenthe core flow and the jet flap flow. If the jet flap diff-user were to perform as desired, it was shown to be essentialthat the core flow have a uniform velocity distribution atthe jet flap exit slot cross-section. This investigationis reported upon ir Reference 2.

Recognition of the necessity for avoiding the presenceof low energy regions near the walls, at the jet flap exitcross-section, and the desirability of some solid diffusion,to achieve maximum performance with minimal ejector size,encouraged the initiation of the third phase, which isreported upon in the present document.

Since the critical region at the jet flap slot isstrongly influenced by the diffuser boundary layer, andby the slots at the inlet, the third phase investigationwas launched to determine whether the overall performancewould be seriously affected by closure of the slots andwhether solid diffusers could be utilized without adverseeffects upon the performance of the jet flap diffuser.

As a result of these investigations, it was learnedthat closure of the slots, and thereby removal of theregion of low energy fluid at the wall, made it possibleto utilize short, wide angle solid diffusers, since the

2

prohlem of separation in the diffuser was somewhat delayedby the presence of the high energy fluid at the wall, whichemanated from the primary jet and remained near the wallas a result of the Coanda turning. In addition, the mixingiof the primary and induced flows occurred during the Coandaturning, and the flow at the entrance to the ejector was

4 already completely mixed, thus eliminating the requirementfor long mixing lengths in the ejector.

Utilization of these phenomena, produced an ejectorof remarkably small size and high performance. The overallgeometry and performance are presented in Table I, andcompared with other high performance, solid diffuser ejectors.

TABLE I

OVERALL EJECTOR GEOMETRIESAND

PERFORMANCE

Amax( LENGTH Thrust x VEJECTOR RATIO Power

Ajet (2)

Coanda/JFD 9 3.5 3.87

Fancher (Ref.3) 38 5.1 3.56Drummond (Ref.4) 28.5 7.0 2.60

Scott (Ref.5) 43.0 7.0 3.12

(1) Diffuser exit area/total jet area(2) Overall length/minimum channel width

The performance comparison presented in the lastcolumn of Table I, represents the published data for thesolid diffusez devices, but the comparison is weightedin their favor since, as shown in Section VI, the Coanda/JFD Ejector used for comparison, has smaller inlet anddiffuser area ratios than those of the solid diffuserejectors. Despite these limitations, the Coanda/JFDEjector demonstrated superior performance.

To permit easier visualization of the comparativesize of the two types of ejectors, the Coanda/JFD Ejectorwas scaled up, to the identical power utilization as thatof the Fancher Ejector, and schematically ,compared onFigure 1.

3

04

L

e.41

0

~~~ .

.. ...._•.. _. _,.

00

o 00

N'ri..

144

0

44 M

-- --- -

..-i i

The large reduction of ejector volume, attributable

to the Coanda inlet and to jet flap diffusion, ar, readilyobservable on Figure 1. As a result of the reduced size ofthe ejector (and therefore reduced skin friction), theoverall performance has been greatly improved.

In addition, the use of jet flap diffusion providesa flexibility which is useful in the achievement of largecontrol forces.

A symmetrical change of the flap angle, can producechanges in the total thrust of ten percent, without anaccompanying change in power requirement, or with anincrement of power, the thrust force can be changed byapproximately two to four times that which can be achievedby a solid diffuser ejector (or by an unaugmented jet),without geometrical changes.

As a result of the "fluidic" effect the total thrustforce can be vectored, without the requirement for vanesin the ejector or jet efflux, or without gimballing of thethrustor.

.I

SECTION III

APPARATUS

1. EJECTOR RIG

A schematic of the ejector rig is presented onFigure 2, with the major components labeled. The ejectorgeometry is quasi-two-dimensional, with a basic channelwidth of two inches, and end plates separated by eighteeninches. Boundary layer control jets are used on each endplate to compensate for the end plate friction.

The major components of the ejector rig are describedin detail below.

a. Primary Nozzles

The location of the primary nozzles relative to theejector channel is illustrated nn Figure 3. The slot widtha/2, was maintained at 1/16th irn'h during all tests with thejet flap diffuser

b. Coanda Turning Vanes

The slotted Coanda Vanes used as the b;Asic turningmechanism at the inlet of the ejector are illustratedon Figure 3, and the nomenclature for the slots and othergeometrical parameters is also described on this figure.

c. Solid Diffuser

The solid diffuser was formed by sealing Slot No. 4as illustrated by the dashed line of Figure 3.

d. Jet Flap Diffuser

The Jet Flap Diffuser position and geometry are alsoillustrated on Figure 3. The jet flap slot width s/2, wasvaried from 1/16th to 3/16ths of an inch to obtain dataat s/a-1.0 and 3.0. This was done by variation of thethickness of the spacer between the jet flap plenum andthe outer lip. The jets are emitted tangentially to thecylindrical plenums and are rotated by the Coanda Effectuntil they are prevented from fther rotation by theflaps.

e. Boundary Layer Control

Blowing jets were installed in the position shownon Figure 3, to compensate for the end plate friction and

6

to permit measurement of the total thrust at the rakestation, thirty three inches downstream of the jet flapnozzles, without penalty for the end plate skin friction.

To produce a fairly uniform flow at the rake station,it was found necessary to supply air at the boundarylayer control orifice having mass flows varying from 5% ofthe total mass flow, at high values of E, to as much as 15%of the total mass flow at small values of B. This mass flowwas not taken into account in the performance of the devicesince it was not required for the operation of the ejector,and was used merely as a convenient means to correct for theskin friction on the end plates downstream of the diffuserjets (see Fig. 2).

2. AIR SUPPLY

The system air was supplied by a Roots-ConnersvilleModel RAS 60, 50 h.p. compressor, having a vol,,a flowrate of 1500 cfm @ 7 inches of Mercury gauge pressure. Aschematic of the air supply system is presented on Figure 4,F with its major components identified.

The mass flow to the primary nozzles was controlled byvalves in each leg, and measured with standard 3.0 in.orifice plates as illustrated on Figure 4.

The mass flow to the jet flap diffuser nozzles wasalso controlled by valves in each leg, and measured withstandard 2.19 in. orifice plates as illustrated on Figure 4.

The dump line was used to control the system pressure,by means of a valve, as shown on Figure 4. The mass flowthrough this line was measured with a standard 2.182 in.orifice plate.

3. INSTRUMENTATION

The measurement of pressure acro3s the orifice platesand .in the plenum of each nozzle was made with manometers,and the mass flow in each leg of the system was thendetermined from the orifice pressure drop as described inSection IV.

The total thrust of the ejector was determined froma total head survey at a position 33 inches downstream ofthe jet flap nozzles, with a 37 tube averaging, total headrake, which was moved across the flow with a parallelogramleadscrew traversing mechanism. The pressure at this stationwas measured by an inclined oil manometer, connected to the"rake through a large settll ig chamber.

7

00

IX .

I0

/ u

00

N

/\/ .c4

\ 4s

44 4

a o3

00

"14E-4

B~oundary'Layer

•.L a/2 Control

Slot 13Slot #4

-A3

(A 3-A 2 )/A

0 0.0625 0.187510 0.2627 0.387720 0.5100 0.635030 0.7969 0.921940 1.1147 1.2397

Figure 3. Schematic of Coanda/JPD Ejector

9

DUMP ORIFICE-# COMPIZESSOR T0

P

PRIMARY

I~ I

JFD ORIFICE - -PD ORIFICE

Figure 4. Schematic of Air Supply System

10

SECTION IV

F TEST PROCEDURES

To determine the ejector performance, it is essentialto know the performance of the input jets, and the totaloutput of the ejector. Therefore, the mass flow, temperatureand pressure at the nozzle entrance and eult, aw at theejector exit were measured. This infornztkm weas then usedto determine the thrust and power of each nozzle, and thetotal thrust output of the ejector as fllows.

The thrust of a nozzle is equal to its mass flow m,multiplied by its average exit velocity,

F -in xV 1

The mass flow through each nozzle was determined withthe aid of the standard orifices located in the ductsupplying that nozle from the relationship,

rm K (P)(Ap) (2)

where p - static pressure (psia)T - temperature (deg. R)

SK - orifice calibration constant (KI=.029, K2 =.0175)I The average exit velocity of each nozzle was determinedfrom a calibration of the jet flow, to correct for theviscous effects of the nozzle walls and the spacers usedto maintain structural integrity of the slot width. Surveysat the nozzle exit to determine the distribution ofstagnation pressure were used to determine the nozzleefficiency v , where,

1 f 'Po, - p] dy dzS= (-PO (3)(po - p)

where Po - plenum pressure

Pi = total pressure measured by survey

A - nozzle exit area

y = coordinate across slot width'

z - coordinate along slot length11

The nozzle efficiencies determined in this manner, had

the following values:

Nozzle Efficiency

Primary (1/16th in slot) 0.971

Diffuser (1/16th tih. slot) 0.965

Diffuser (3.16th in. slot) 0.767

Then knowing the nozzle efficiencies, the average exitvelocity V, could ba determined from the pressure drop andstagnation pressure and temperature from the expression,

S- p

which derives from compressible flow theory, and the defin-ition of nozzle efficiency.

The nozzle thrust is then determined from Equation 1.

Since the rake survey was made at a large distancedownstream from the nozzle exits, the flow at the surveyplane was at a sufficiently low Mach No. to permit use ofincompressible theory without the introduction of asignificant error. Therefore the thrust was determined fromthe equation

- - p) dy dz (5)FT 0

where the integral with respect to y was accomplished bythe averaging rake, and the integral with respect to z wasperformed by numerical integration of the survey data.

These quantities, together with the geometricalparameters, are then used to determine the performance ofthe ejector in terms of the parameters described in thenext section.

12

SECTION V

TEST RESULTS

Since the C/JFD Ejector utilizes two sourcesof energy (primary jet and diffuser jet), which in generalhave different stagnation pressures and temperatures, adescription of its performance requires the use of morecomplex parameters than those utilized for reporting theperformance of conventional ejectors.

In addition, to provide a basis for

a. Optimization of the system

b. Comparison of the system with other, moreconventional ejector systems, and,

c. Utilization of the reported performance inpractical vehicle designs; it appears desirable to usethree distinctly different parameters to describe the overallperformance of the system. These parameters are brieflydefined below and discussed in more detail in Section V!nnd in Appendix I.

1. PERFORMANCE PARAMETERS

Animportantgoal of any propulsion system is toachieve the maximum thrust available from a given rate offuel consumption. Other factors such as controllabilitythrust/weight ratio, size etc., are also to be considered inthe selection of a propulsion system but these considerationsare beyond the scope of the present report and will bediscussed only briefly in Section VI.

For purposes of this document, the optimal systemis considered to be that configuration which produces amaximum thrust per unit power input to the system.

Since thrust/power is related to the thermodynamicscale of the device or more precisely, to the stagnationconditions in the driving jets, it is more appropriate todescribe its performance in terms of a non-dimensional parameterwhich is independent of its size and power level.

a. Optimization parameter I(L)

The ratio of thrust/power can be made non-dimensionalby multiplication by a velocity, and in the interest ofsimplicity, the particular system velocity used was theaverage primary jet velocity V,. Further discussion of thischoice is presented in AppendiAjI, but for purposes of discus-sion of the test relults here, the parametdr is defined as

Total Thrust X Primary Jet Velocity (6)Total Power Required

and plots ofrIvs. E, where

E = E2 /(E 1 + E2 ) (7)

is the ratio of diffuser jet power to the total powerrequired by the ejector, are utilized to indicate theeffect of the various geometrical or configurationalchanges as well as the influence of changes in diffuserjet power.

To permit comparison of the C/JFD Ejector with othertypes of ejectors, whose performance is generally reportedin terms of thrust augmentation, the performance of theC/JFD Ejector is also presented in terms of two differentthrust augmentation parameters defined as follows.

b.) Fixed mass flow thrust augmentation (0m)

Comparison of the C/JFD Ejector's thrust outputwith the thrust output of a free jet which has the samemass flow and power requirement as the ejector jets providesa parameter 0 , which is useful for determination of theC/JFD Ejector~merformance when the source of power iscomprised of a single gas generator having one uniformstagnation pressure, supplying the energized drivingfluid to both the primary and the diffuser jets. Thenozzle exit area of this free jet is related to the exitareas of the ejector jets s and a, through a relationshipinvolving the stagnation pressures of the various jets andother ejector parameters. These relationships are discussedin Appendix I in somb detail, but for purposes of discussionof the test results here, the parameter 0 mis defined simply as

Total Ejector Thrust (8)Thrust of free jet with same mass flow and power

To permit comparison of the C/JFD Ejectors performancewith that of a gas generator whose nozzle exit area hasa fixed relationship to those of the ejector's jets, athird parameter 0 was developed. This parameter isdescribed brieflylbelow.

c. Fixed nozzle exit area thrust augmentation (0 p)

This parameter 0 , describes the thrust output of aC/JFD Ejector relativf to that of a jet whgse nozzle exit areais equal to that of the ejector's primary jet (a), andwhose total jet power is equal to that of the ejector's jets.

14

This parameter is defined as

0= Total Ejector Thv:ust (9)P Thrust of a free jet having E - ET and a Rp a

The test results measured experimentally with variousC/JFD Ejector configurations are discussed below andcompared with theoretically derived performanceý, in termsof the three performance parameters described above.

2. PERFORmANCE

Initial testing of the Coanda/JFD Ejector withconventionally designed solid diff-i.sers and a slottedCoanda inlet (developed and optimized for use in a soliddiffuser ejector),(Reference 1), demonstrated a seriousinterference between the diffuser boundary layer and thejet flap diffuser. This interference appeared to be ofmore serious consequence as a result of the presence of thelow energy air induced through the inlet slots. In thatconfiguration, the performance of the device, departedseriously from the theoretical, as indicated by the resultsof testing reported in Reference 2.

Further testing, with systematic sealing of theinlet slots, and with short, wide angle, diffusers, whichappeared to perform satisfactorily, as a result of theinteraction with the jet flap, is reported in this document.

Using a basic configuration illustrated schematicallyon Figure 3, and with various slots sealed, the overallperformance was measured and the results in terms of 0 , X.and 0 and compared to the theory, as described below.

a. Effects of Inlet Slots

As indicated by the dashed line on Figare 3, sealingSlot #4 produced a solid diffuser having an area ratio of 1.12,and a half angle of 7.12 degrees. With this configuration,the perforziance was measured, over a :ange of flap anglesand with systemitic sealing of the slots. Results ofthese tests at s/a = 3.0 &re illustrated on Figures 5,6, and 7.

Figure 5, with f = 30 degrees, illustrates the performanceimprovement resulting from sealing Slot #4,compared to theperformance with all slots open. Sealing this slot,produced performance gains of approximately 15%. SealingSlots #2 and 3 produced vsry little additional change, overmost of the range of jet flap power ratio.

15

Sealing of the inlet slots produced noticeableimprovement in the steadiness of the flow. This was anindication of the ability of the flow to remain attachedto the solid diffuser and to the jet flap.

Increasing the jet flap angle to 40 degrees produtcedonly slight gains of performance, and required increasedjet flap power to achieve maximum performance, as illus-trated on Figure 6. Further increase of the flap anglecaused the flow to become unsteady, and produced no furtherperformance improvement. The performance a•= 50 degrees,is also illustrated on Figure 6.

For completeness, the performance of this configurationwith the flap set at 20 degrees has been measured and theresults presented on Figure 7. The maximum values of 0 a0 and JtL, are smaller than those at -30, 40, and 50d~grees, as expected, but with slots 2, 3, and 4 sealedthe flow became extremely steady at - 20 degrees.

The influence of the inlet slots with s/a = 1.0, isillustrated on Figure 8. Comparison of performance withall slots open and all slots sealed, indicated gainsof 15% to 20% in the value of the performance parametersas a result of sealing the slots.

b. Effect of Solid Diffusion

The effect of further increase of the diffuser halfangle was determined by increasing this angle to 12 degreesand trstutgy the configuration with all slots sealed, andwith only Slot #1 sealed. The results of these measurementsare presanted on Figure 9o

As indicated, the performance at small values of E isvery close to theoretical, as E increases however, thepower of the diffuser jet at first appears ineffectivecausing a decrease in the value of each parameter. Thisappears to be due to the fact that there is insufficientdiffuser jet power to overcome separation from the flapresulting from its adverse pressure gradient. However,as E is inciaased above approximately 5%, the performanceparameters increase with increasing values of E. At aboutE = 0.20, the measured performance becomes very close to thetheoretical, indicating that the diffuser jet has sufficientpower to overcome its adverse pressure gradient and to remainattached to the jet flap, at angles up to 30 degrees.

16

0-- Theory --1.124.0

3.0

3.0 L-_--•-• • • •' +Experiments G-.12 0

Theory 1, =u1.12

2.0 + y

SExperime~nts S-1.12

1.0 y Configuration'0 All slots openV Slot #4 sealed+ Slots i2,3,&4 sealed

0 I I I I I1I I .O .10 .20 .30 .40 .50

Jet Flap Power/Total Power

Figure 5. Coanda/JPD Ejector PerformanceA/a-16 s/a-3.0 /3 -30 degrees G !lol2

17

XL. Theory G;-1.12 15-40 deg.

4.0

k\2Z X Experiment 9-1. 12

Op Theory 9-1.12 f-40 deg.

2.0 O

+>- O-•p ,Experiment !-1.12 _

i.0 Symbol Configuration

V Slot 14 sealed :-400Slot #4 sealed •-50

+ Slots 12,3,4 sealed,8S 00

0 .10 .20 .30 .40 .50

Jet Flap Power/Total Power

Figure 6. Coanda/JFD Ejector PerformanceA/a-16 s/aa/3.0 -40 & 50 deg. 5-l.12

18

:4.0 __. Theory S-1.12

+ + ++ +3.0 " Experiment SI.12

p Theory 5l.12

2 . 0* i -I - + 4 -

-Experiment S- 1. X2

1. nSymbol Configuration

+ Slots 12,3,4 sealed 6200

0 i ii I I.I I I0 .10 .20 .30 .40 .50

Jet Flap Power/Total Power

Figure 7. Coanda/JFD Ejector PerformanceA/a-16 s/a-3.0 /3-20 degrees 5-1.12

19'

4.0

SL Theory 5 -1.12

3.0-0

u- Experiment 5-1. 12

2.0 -/- p Theory 5-1.12

I0

-- • Experiment 5-1.12

1.0 Symbol Configuration

O All slots open0 All slots sealed0 1 1 -. . I -1 I I,

0 .10 .20 .30 .40 .50

Jet Flap Power/Total Power

Figure 8. Coanda/JFD Ejector PerformanceA/au16 s/a-l.0 /0 =20 degrees '9=1.12

20

1 5.0

I)- Theory =1. 30

4.0

0

4 q rL Experiment 5-1.30

3.0 -

-p Theory 5=1.30

2.0

A do-p Experiment 5-1.30

SSymbol Configuration

Slot 01 sealed

D3 All slots sealed

0 .10 .20 .30 .40 .50

Jet Flap Power/Total Power

Figure 9. Coanda/JFD Ejector PerformanceA/a-16, s/a=3, 13-30 deg., 5-1.30

21

mA-m •

SECTION VI

DISCUSSION

1. PERFORMANCE

The performance of the Coanda/JFD Ejector has beendetermined experimentally, and the information has beencompared with theory, and presented in terms of 0 , 0andfl, as functions of the jet flap power ratio A SeVtion V.

Since these parameters are somewhat unconventionalfor the description of ejector performance, a briefdiscussion of the significance of these parameters, andtheir relationship to the thrust augmentation parameter 0,used for the description of the performance of soliddiffuser ejectors, appears to be appropriate.

High performance solid diffuser ejectors deriveapproximately 15% to 20% of their thrust as a result ofincreased mass flow (and therefore increased thrust),through their primary nozzle, due to the presence of theejector shroud.

The increased mass flow throuqh the primary nozzlerequires increased pnw'ir Io drive the primary jet at thesame stagnation pressicz,. This requirement for additionalpower is seldom discussed in the literature.

Although the Coanda/JFD Ejector requires thisadditional power only for the jet flap nozzle, it isaccounted for in the definition of the parameters fl, &0mThus these parameters represent a more accurate descripionof the ejector performance than can be obtained from aknowledge of the conventional parameter 0, commonly usedto describe the performance of solid diffuser ejectors.

a. Physical Interpretation of Performance Parameters

(1) Thrust Augmentation at fixed area (0 )

The quantity 0 is the ratio of total ejectorthrust to the thrust Bf a reference free jet having thesame nozzle exit area as the ejector's primary jet andthe same power requirement as that of all ejector's jets.

The variation of 0 with E (as illustrated onFigures 5-9) is an indica~ion of the relative merits ofthe application of power to the ejectors diffuser jet

22

compared to the application of the same percentage ofpower increment to the free jet without change of itsnozzle exit area.

An increase of 0 with increasing E, is Lndicativeof the fact that the power applied to the diffuser jetprovides a greater increase of total ejector thrustthan would occur if the same percentage of power wereapplied to the free jet at fixed geometry.

Increases of 0 of as much as 30% are indicatedon Figures 5-9 for diFfuser jet power of approximately40% of the total power. A4indicated above this impliesa superiority in achievement of thrust by the applicationof power, for the C/JFD Ejector compared to a free jetat fixed nozzle exit area.

(2) Thrust augmentation at fixed mass flow (0.)

The quantity 0 is the ratio of total ejectorthrust to the thrust of a reference free jet having thesame mass flow and power requirement as those of theejector's jets.

Since the reference free jet utilized in evalu-ating 0 has a nozzle exit area whose relationship tothe nozzle exit areas of the ejector's jets changes withE (See Appendix I), it cannot be utilized to evaluatethe merits of changes in the diffuser jet power E, inthe manner suggested for 0p.

It is very useful however in comparison, at anygiven value of E, of the ejector's thrust with the thrustof a reference jet having the same mass flow and powerrequirement as the ejector. It is important however,in making this comparison, to establish the properrelationship between the stagnation pressure of thereference free jet (P om) and the stagnation pressuresof the ejector's jets "(Po0 and P.2) as indicated inAppendix I.

Variation of 0 with E is very small as indicatedby both theory and experiment on Figures 5-9. Asdiscussed above, this fact has very little physicalsignificance since the reference jet has a nozzle exitarea arm which varies with E.

As indicated on Figure 9, the C/JFD Ejector canprovide thrust augmentation of approximately 1.5,compared to a free jet having the same mass flow andpower requirement, with a very small area ratio solid

23

diffuser (1= 1.3) and a ratio of minimum ejector ejector

cross section to driving jet cross section (A/(•o+ s))of 4.0.

(3) Optimization parameter (Ol

Tae quantityfl is a dimensionless parameter, relatedto the thrust/power ratio. It is made dimensionless bymultiplication of the thrust/power by the average primaryjet velocity V

This parameter does not involve a comparison witha reference jet, as do the various thrust augmentationparameters used for the description of performance ofconventional ejectors.

Ejectors with more than one driving jet havingdifferent stagnation properties cannot be easily comparedto a single free jet; however, using AL, it is possible toevaluate the ejectors performance over a range of variablesand to select the optimal conditions as those which producea maximum value of L.

Comparison of the C/JFD Ejector's performance withthat of a conventional, single driving jet, ejectorcan be easily accomplished since the Aof the conventionalejector is rolated to the conventional thrust augmentation ýas follows, since

0 =FT (10

Cor a conventional ejector in which F is the thrust ofa reference jet having the same mass Wow and power asthe ejector, and

T12 ) (1-E) (11)ET/ 1 l F/l l n

for a C/JFD Ejector.

Therefore, for a conventional ejector in whichE =O, and F1 = FRm

- 2 0 (Conventional ejector) (12)

Tt is also interesting to note that for a free jet

V1- = 2.0 (free jet) (13)

24

I

In view of the simplicity of the parameterSl for

purposes of

a) Optimization

b) Comparison with other ejector systems

c) Design applications

it appears to be a universally applicable parameter whichcan be easily evaluated for any geometrical or thermo-dynamic configuration, and is independent of the choiceof reference jet properties.

On the basis offLh, it can be observed on Figure 9,that the C/JFD Ejector has achieved a thrust of almosttwice that achievable by a free jet with the same power,despite the fact that in that particular configuration ithas a small diffuser area ratio C • = 1.3) and a ratio ofejector minimum cross section/total nozzle exit area, of only 4.0.

2. COMPARISON OF C/JFD EJECTOR WITH CONVERrIONAL EJECTORS

A comparison of the performance of the C/JFD Ejectorwith that of several conventional, high performance ejectorson the basis offltis presented in Table II.

TABLE II

COMPARISON OF EJECTOR CHARACTERISTICS

Inlet Diffuser LengthEjector Area Area Ratio

Ratio Ratio (2)

((3C/JFD 4.0 1.3 3.0 3.94(3)Fancher (Ref.3) 19.0 1.9 5.1 3.56

Drummond (Ref.4) 19.0 1.5 7.0 2.60

Scott (Ref.5) 21.5 2.0 7.0 3.12

(1) Minimum Channel area/total jet exit areas(2) Overall length/minimum channel width

(3) E 0.161, - 30 deg. (Run 343)

25

Using the theory of Ref. 2 which produces goodagreement with experiment, as illustrated on Figures 5-9,the value of X%- increases by approximately 12% for anincrease in the C/JFD solid diffuser area ratio (3 )from 1.3(as in the above table) to g = 1.9. -his wouldindicate that at q = 1.9, the C/JFD Ejector would produceSL = 4.41, at the same inlet area ratio (A/(a + s) = 4.0).

This represents a considerable advantage in performanceover the best solid diffuser ejector reported to date witha geometrical advantage requiring only approximately 25%of the volume of the solid diffuser ejector.

3. C/JFD EJECTOR AS A CONTROL DEVICE

The jet flap diffuser provides a mechanism whichcan be utilized for control of the magnitude and directionof the thrust ejector.

a. Thrust Vectoring

Although detailed measurements were not made duringthe experiments reported in this document, it was observed,that the entire efflux from the ejector could be easilycontrolled directionally, by simple opening or closingof the valve in either supply leg to the diffuser jets.In addition, thrust vectoring could also be achieved bydifferential operation of the small flaps used to controlthe diffuser jet angle /3.

b. Thrust Magnitude Control

The magnitude of the thrust vector achievable by ajet flap diffuser ejector is controllable by two methods.

(1) Power Incremftnt Response

The application of additional power to the diffuserjet causes large changes inf't. As illustrated on Figure 9,'0- increases with increasing values of E, up to E = 0.18.

Thus it is apparent that the application of additional powerin this region will produce tUhrust increases larger -chanthose achievable from a free jet or from a conventionalejector whose. Lor 0 is independent of power level.

(2) Flap Angle Response

Control of the magnitude of the thrust vector canbe achieved by symmetrical changes in th* jet flap angleFor example, comparison of Figures 5 and 7 illustrate anincrease in thrust of approximately 20% as a result of achange of 10 degrees in A , with fixed pqwer.

26

SECTION VII

CONCLUSIONS

The Coanda/Jet Flap Diffuser Ejector, occupyingapproximately twenty five percent of the volume of thesolid diffuser ejector reported in Reference 3, hasdemonstrated experimentally that it can deliver ten

-percent more thrust per horsepower than the solid diffuserejector, despite the fact that the C/JFD Ejector had smallerinlet and diffuser area ratios than those of the soliddiffuser device to which it was compared.

Extrapolation of the performance, to a configurationhaving comparable diffuser area ratios, but with the samedisadvantage to the C/JFD Ejector in terms of the inletarea ratio, indicated a twenty five percent advantage inthrust per horsepower for the C/,'?D Ejector over the soliddiffuser ejector.

The use of jet flap diffusion also demonstrated aunique capability for production of large control forcesand for vectoring of those control forces by means ofsymmetrical or raymitetrical movement of the jet flaps.

The C/JFD Ejector also demonstrated the capabilityto produce control forces in response to incrementalpower which was considerably in excess of those availablefrom conventional ejectors or from free jets.

I 27

APPENDIX I

DETERMINATION OF EJECTOR PERFORMANCE

28

Appendix I

Determination of Ejector Performance

1. Experimental Measurements

The determination of the values of the performanceparameters A , 0 " and 0 from the experimental data iscomplicated by de fact That the ejector is two dimensionaland each jet (primary and diffuser) is comprised of twoelements, one on each side of the plane of symmetry.

As a result of the lack of exact symmetry due tomanufacturing tolerances, it was necessary to adjust thethermodynamic properties in a manner which created slightlydifferent values of the stagnation pressure and mass flowbetween the right and left side of each set of nozzles toproduce a symmetrical efflux.

These differences created some complexity in deter-mination of the velocities V.. and V. for use in evalu-ation of each of the performlAce parieters. The techniqueused is described below in detail.

The measurements made during the experiment includedthe following, for each test set up. (See Fig. 4 )

a. Stagnation temperature (To)0

b. Atmospheric pressure and temperature (p, , Tco)

c. Orifice pressure drops (AP'L,APlR,AP2LAP 2 R)

d. Nozzle plenum pressures (PolL' PolR' Po2L' Po2R)

e. Static pressure at diffuser jet (Pch.)

f. Total thrust (FT)

29

2. NOZZLE PERFORMANCE

The mass flow through each nozzle was then cal-culated from its orifice pressure drop, using the cali-bration constants as described in Section IV, Eq. 2.

The average jet velocities V I V., V Lwere then determined for eaph nozilT, uH1g til meaS1edstagnation temperature and pressure and the appropriatenozzle efficiency as described in Section IV, Eq. 4.

Using these values of mass flow and velocity foreach of the four nozzles, the values of the power re-quired to drive each nozzle was determined as follows.

Since the primary nozzles exhausted into a virtuallyundisturbed region, the pressure at these nozzle exitswas taken to be P . Therefore

ElL = 1/2 in1L VIL2 (14)

=1/2 mR V 1 (15)

and E1 = ElL + EIR (16)

the thrust of each primary nozzle is;

FIL i 1L V1L (17)

FIR " lR VlR (18)

and F1 F IL +F1R (19)

The average primary jet velocity was then determinedfrom the relationship

V F2 (El1L + El1R)

m Ii (20)

and the primary jet mass flow is,

m 1 = rol + mIR (21)

30

The rower required to drive the diffuser jetswhich exhaust into a region having a pressure somewhatbelow ambient is,

E 2E2L - 1/2 m2L V2 Lo (22)2L/2Li2 2R00

E2R 1/2 V (23)

and E2 E2L +E 2 R (24)

where

(25)

and the thrust of each diffuser jet is

F2L 2L V2jL (26)

F2 R " 2R V2 JR (27)

and F2 * F2L + F2R (28)

The average diffuser jet velocity was then deter-mined from the relationship;

2 (E2L + E2 R)

2j m2L + 2R (29)

and the diffuser jet mass flow is

m2 m2L + m2R (30)

31

3. PERFORMANCE PARAMETERS

Using these average values for E1 , E , F1 , F andm• and m2, the performance parameters were delermhned asfi11ows:

A. Optimization parameter (0)

The value of.C.is determined from its definition(Eq. 6) using the total thrust which was measured by a

wake survey as described in Section IV, the primary jetexit velocity Vlj and the total power ET, where

ET mE1 + E 2 (31)

thus T j (32)

b. Fixed mass flow thrust augnentation(0.)

To evaluate this parameter it is necessary todetermine the thrust (FRm) of a free jet having

m• W M m1 + m 2 (33)

and ERm = E1 + E2 (34)

Since the reference jet is a free jet (expandingfrom its stagnation pressure P to the ambient pressure Pa,its effective velocity is

/ 2ERm (5VRm / I iRm:(5

Rm

and therefore its thrust is

and the augmentation parameter 0m isFm

OM= FRm %(37)

32

The value of 0m obtained in the avove manner pro-vides information which relates the ejector's thrust to thethrust of a free jet gas generator having a mass flowand total jet power equal to those of the ejector's jets.However, in the general case, in which P • P , thestagnation pressure and the jet area of Ae reflrencejet both vary in relation to the ejector jets as thediffuser jet power ratio E changes. These relationshipscan be expressed as follows:

- -A)4 N

1'o~1- 1 (38)

and

where m = m2 /(mI + in) in the above equations.

These relationships imply that while 0 is anexcellent indicator of thrust augmentation availablefrom a particular ejector geometry at any given operating condition(fixed E and m), it is not useful as an optimization parameterssince the reference jet is in effect a "rubber" engine, in boththe geometric and thermodynamic sense. If however, the ejectorsjets are all supplied from a common source, the parameter 0can then be used to relate the ejectors thrust to that of tPegas generator having a fixed stagnation pressure (independentof E and m) and is "rubber" only in the geometric sense.

In this case since

Pol Po2 ' PoRm (40)

the ejectors nozzles are designed with areas related to thenozzle areas of the reference jet as follows:

a Rm /a = (mRm/il) l Vlj/SRm VRm) (41)

and within the approximation that all densities are equalthis can be simplified to

a•/a = mm/I (42)

33

and since s/a is chosen on the basis of other ejector andvehicle design considerations, the distribution of themass flow from the gas generator between the primary andthe difftser jet is easily determined, since

arm = a + s (43)

for this case in which P 01 Po 2 = PoRm

c. Fixed nozzle area thrust augmentation (0 p)

Since in the general case in which P 1 Pthe relationship among the nozzle exit areas of tna'ejector and the reference jet is complex and varies withthe parameters E( = E /E + E ) and m( m /m + m ), athird parameter 0 , h~vi•g a iixed referenge let eRitarea (a ) and re~erence jet power (E ) equal to thatof all Esector jets, was used to descibe the relativeperformance of the ejector. In this case the referencejet is related to the ejector jets as follows

rRP = a (44)

ERP = E1 + E2 (45)

From these relationships, the thrust of the ejectorF , can be expressed in terms of the thrust of the referenceJit as

FT FT/F1.P FR (1 + E2/E 1 )2/3= (l-E) 213 (46)

where 0 FT/F1 (47)

The parameter Oprelates the ejector thrust to thethrust of a fixed nozzle gas generator, and thereforeis more convenient than 0m for design purposes. Thjreference jet utilized to define this parameter (0 ) musthave a single stagnation pressure different from Peither of the ejectors jets, as in the case of 0 , or aswould be :equired for any performance parameter Wsinga single reference jet for comparative purposes. In thiscase, the stagnation pressure of the reference jet is rela-ted to the stagnation pressure of the ejectors primaryjet as follows.

P°- - (l-E) 2 / 3 (P . /pQ~P) (48)

34

Thus the relationship among the stagnation pressureof the reference jet and the ejector jets is dependentupon the power supplied to the diffuser jet, as in thecase of the reference jet used to define 0 , but thisreference jet is "rubber' only in the ther;odynamic sensesince its geometry is fixed by the relationship of Eq. 1-32.

35

APPENDIX II

TEST DATA

36

Jv OL4 OL A C; i 00 00U C 4 r: 0*0 10,

- 0 00 00 oaO oLA LO oL

~. 0 ow v rl Im vW coi HO 47 m

* U. 100 ýd\A A O, LA o00 oa LA oV) C coo 01 itrýtn ww N u- N t-

C ý I S__1._ o__ &_t

tm N v - m %D (A %

t.'. tn in tn 0 n Hn (

C- 0 0 0 0

N N N 4 N N N

37

LOD In o A LAL LAO m NoI a#1 *C1 owL OLNA:4f-?-I Ci HM '.Jr cq~O ' N _ w -i(' H O' -4q .Dr

- nt no 0t 00 0 m 0 001

in 0 om co m LAOn L c n tno uN 'I OCO LA LO JU I

-4 9. r 4 0 ;C .9 9 0;~ ; ; C; 0; ; a; C

_ _a E-j-4T

z N 1n H LAIn 1 W co N (

0 4 A OOAIrZ A ¼~~A 4 luýh

r

LAL LA 00 OLA LA OLA LA LA LAn

N~9 C14 N* CA**4N N

1~ N N O O 0

0. ;L4 0 n ko c'~ 0 L~ 0 0 0 0 L

C4 ON09- .- .q .- r- N N 0

(14 N

38 L

I;f- ~ ~ c -~0'U0OA inU LAO

%a r- q ;I % nt

* ~ t 00OA A 00 00 0U 18)C.. ov vm Nqo w' 0 N (N in~i %L'o

N9 I9 9 . .1 19 0I *L4AL4 L4t 9i Cý 9 OO (N 04N II He

LAL C4 I NC;'H 4 O a,~HU

( N c ? ,4 m NC i N oC0 Ne N N m m m m

C: C: ___4__ - ±n -o i. -D f- -m

0 10 0i 0 0

4 AbN N N4

r.4 P- ( H 0 jN a

-44 0

alt oH I i wI 1 1 0 ) C ýN4 t v~ I I flO 00 r 010 tO4 01 Coi 184) (4 10 4(

j-N C14 t-4 N N N N ff)

tn0 M I I 00 100 00p80012011 00 nu ot ,o ntCmC40 - o m mN Hr04c rN j0 e w m V

C40r I I N I 4c ý u nt Zc ýc

0001

r4C~~~~C C. Cr rw e4 -. ý ~ ON mv ~NN NOa

41

EU0 010 0 n0 0 0 0F. N N C% v0nI0

IT Iq n L

E-04 0 H r0 10 N

o1 oo 1 01 0o 01 0

0 N 4 N N N Nj m

40

a n oi Ln o t ooi n' in t inoq II qL L n L0.~~i in c4 wor qm ( nIm n rini to oco *n Ln mw :4 %Za:c 4 I I cvLn (n ao '

-N 1ý1

in tniLn initn tn ti n Ln in io incolI c oL.C 0% tln 07N -i 0 (n 0V m 0 c;% -14r cA %01

00 co cnc 00 'to ai ('n on~ qo,- 46 * a

C4 010 IC; 010 r0c NM m-wi CNOl -11col

0 0 0 ~e., C; U; 0;1C C4 0;

0 Q

(N 4 N NN Cq C 4 %

0_ i- IN CI m~ m010 m

___j 0H 0 - H~j 10Fi .

M v~ v vniwt

41

00 00 om Mt 0100o2* C4 LAL %""' 0;I 0;C4 (9 D i coo

-.. t-l uio-oIo n o~-e1t4% o

o LAnf uiLn 0000o LAO OLn Io m o v 0 m C'LA oc'0 c q

C; ~ ~~ CCrcoI N 0,

E- LA LAE- CS C4o0o~ o

co co co 0) c Cm In t n n L

_ _ _ý C (. Im C

_____ 0IvtnE4 0 r4 N NI

42

N HN H rI NO N A 04 M C-4A

-I-' cl I i M 0mrin

i7_.__ _. . . 1 / C;in_ ini ' - ,I ýC

m____4ao ' __ ~ cI~I)N L 4.4 ar inC to (nN or- at__ __ a_ _ w. q- ., o(

Nqm N N Ni 4 inM NN NN N N NN

0~~m 0~ 0

~ ''OAM N mO CO a%

E-4 0 0 01N

1<1 in in m1 0 10

01 N

<1 W v* 9- w m v 0- m 0 %0 9 0 m qv. -i %D-4 r% r'n ý-C C4 !r-.;a4. en 'o

C4 m 4m C - qS- - -

00 0 m 0 0 Ln tn nloo,0 q !oo tn tn o04r 0m m 0n tAo 'cocq I to rc-co -v %

oo ao~ co to io c n t in !tn~ a nu o~ vco

00 QUm 00 otn omlu~f 00 00 c oui1

>~rC % .~v. I%

iO m

N N

th t n i n i in IL, i

IL, 0 U coj a% v men

tn %t' i r-

44

00 O 00 o po

I> tn LAnf LAO I

C r IN v 4 0ji1

- NN

ocn t

C1 1___ C

0 t

0 0o

1 14t-Ijq

I4

~~v 0 0 N C

in 0( tOU MC i I I. (D1 coI Ch ~ r. c

4 c4 c4I* f*

tI an o L

0: r- co 0I~C4 eq N N

-- c o C4 MC4

P0 L4O IL

V..

N N

0~ *q'CI0~Ij7;m

___ M__I

46N coI

__ ~ HI .I a

I7% ini nac in ni%D in4 r4 N o N A

- ~ -H H-H- HN NMCN" ML

00~ no Malmo 00M M O'

4 C4 4 C4 j M 4( C 4 A 4 r 4A O A c4

u" i t cc innloo ino o cc a n 100v %D NC' Cr 0 D t

0C

5.. (0 c OH 1 0 0 o oH min~ MOt. amu

C% 0 0 9 0 0 0 *0 410 **

4 .U u0 in a in inI CO CO

0 00 CO CO CD 0 0

C% 0; C: 03 C; 0 0

- N C1 N N N C' jNI N

In M IMI nI

OD 0) 0 H () v 0

M t II -lY)M

47

kLA LAO LA 00 0A 0 OLA( C

wti N OD coc P% (r) in V

C4 14H jHC4 N') t

O.I mC H o w' e-r~ vin (v vA o N %J00

I IIII 00 LAL i A 0 0 Ot 0 0

m I0tn H- OH co '~om oo 0o' a -I w 9n mm mm wo . H N v m 9 % mm

Ho 00t~ co o OH C'4 C0* LA

C% N ZN N N

E-4 0 0 0 oH 0 o

co1) co oo 0) 0 - qlen'i ' ~ ~

8 ~ 9 9 9 9 47

om 00 tAO 00 ALO-W o, eoj q oAt co0 0%

C% 0

OL 00 m A LAW &f)

t< -- L cq LA~ 4A ~c; 4 c;

rs IA UA 00 00 0

(N t-r~ C% cn' coo

9n In L nin

E- 0 i i

__mN__NN

Hjz)OLA LAW) OLA LAO )LI

9 LA r-L (NkO Le4 o- r

E'448

APPENDIX III

PERFORMANCE DATA

49

L.

IIThe thrusts, mass flows and performance parameters,

derived by the methods described in Appendix I, using thedata presented in Appendix II, are presented in tabularform. The symbols and units used in this appendix areas follows.

Fpri - Primary jet thrust (lbs.)

F - Jet flap thrust (lbs.)j fd

m2 - Primary jet mass flow (slugs/sec.)

m2 - Diffuser jet mass flow (slugs/sec.)

B - Diffuser jet power ratio (E2 /(E 1 + E2 ))

R - Right side of ejector

L - Left side of ejector

50

I

H 0o0n N N

3C%4 CCo r-4Ci 0

H H H H H4

In Hi 0 Ir. co

co co In a% I.D

H- H H- H- H H

tn %D Nl 0 H'N 0 m3

C14

r- HH 00)C r- %0%0 C'4 C14 0 a a00 00 00 00 00 00 0 0

HH H) 00 H HH 00 00a

r- 00 H~O H- cHH CrOi 00 00 0O0 00 00 00 0 0 00 00

o) H) Olt In No 'o

c4 4 44 9ý 44C;

N1 N4 N N N N N

t- r4 0 .D M t-. Mfn WC' 0 (nco %D' 91 q 0 In 't. N O ( OtI-

4.4 o 4 a0 * a * a 4 0 4 .

%D C'l' HH H % v m C%~ HH0 1rz4

1.4 L4 0ý 0 S4 *ý ý L44

'D r-co a%0 H N0 0 0 0 H H H-N N N1 N N N (1

0N

%0 04 H

co (A (AN

0- HHA H

N coN

H H H H4 H H

cp % N 1 1 N Co r- - l * N

H0 N H H- H- H- H- - -II c

N~~ 0NN

*- C)a- 0 0 c

94 ~ o NC'4 1MM HH HO -I 04

41 CA tn 0o

0 ko in a%00

0j 0. LA CINr 4 N m NV N N1 N N

*4 A C1 v %D 0 00 co 0% O 0% m N

r. NH vM M Nm mN r

o h 10 ( c r ON vC HO0 fnc* * 0

NN N N N N

52

in '0U

v h in to to %0 0 co

N H- MY N- co N Nto Co aT.a 0 N CN eq N N m~ mY

o c 0 co LAC) m C)C

H V* 0o 0o N 0

Q H- HN - N

m C% N~ m O C N 0 n ntH 0~ H H rq N-

00 00 00 0000 00 00C! . SS S*

N

HO O H HO HO 001 a% m

00 0 00 0 00 00 0

.41 H- N 0 V '0 0 .0

4.) N% COq N* N N1

No N Ný N l c N No %DN rI

Nr4 '0'0 C14 NA L4AU

m4* co in o)' o3' l~ L oNr- co

N Nq NN

N 4 NN

53

40 ifl H

m N 0 N N Ho A

0~ ~ ~ m N r 1

0 o N% 0% 0

U') 00 0 N%

o 001 00 NC

'2'a 0 H 0D 0- 00' 1-0

oo 00 00 0 00 0 0 C

N%00 coN 0 00 400 NCO 100

HOq H H Ha NH HO HH- eH H H HH f4 $A#- HHV- H H

00 00 00 0 0 Q\0 00 00

0-4 9 0 0 0 0 0 0 0

40 400~ N U) Lnw. C % V~) m" Nl in

0 0

H04 NN~0

m 00 0 00 w2" 40o 00 %r %

00 OCq' v IH o co C 00 w7' 24

z 0H m'2 NOON NO

N NN N N

.94

('4f in 04IAIA'

c4 4 c4 c4 0; cc

'0 cIA IA r- C

('4 N' N' ('4

H H H4 H4 H

o Vc co -no f- -c-- -ci04 %Dc 0m0 l

'3 00 W IV H

o! H H- H- N' H -

OH- rir HH HH H H H H H-iH HHr

00( 00 00 00 00% 00

I- r- HO H HO HO OO r- 0 ~ HO

00 00 00 00 00 00 00H. c4 N

It% ('- m' H- cor nwi q - - t

4'4

H#.% .Ct (4C-.c 4 4uý u

N~~~ ~ ~ ~ cy - r 0 to c 0 4 - a-.In%Dc t*- %D% Cto3 %D OH C-.Ifl

C14 ('N N '3 '

55

N I HLn H. LANc Cl I C LA

C') H a% l

N N ~

H1 HD H4

N N CA Hn H

tn

v 0 0n

0V4 C 0L Hl o n% 0aH H r4 0H 00 0 0 0a

00 00 00 0 0 0 0 0 0 00

N z *Y 0 0 0 - Q 0 (

rý0 iH Ho 01 00 HOrI H - HO 00

00Go0 00 0 0 00a 00

V 1 1 Co IDH I AN

41 IT 0N CN N N HrNN4

N C)Ow NH,a IHi C)~ 0 00

fnt LA11t-0 %.0%0 %D0 C% %DN tn mA~

N NNo

56

00L Nco t-N 0 -

oN co 00 61)H0r

to It LN, %H U)

mN mN NN0 C14

Hv 14 Ci (N .4

%P nt mi VD co

-o r- 000

H- N

NO Oii oo 0 0HH HoI

000

-i0 0C HN LA rID N Cor-

(N C1 4N C14N

- -a%_ v_ C%_ .1 __Ir%D t v0 4 %wH f4(

W nIINr 4c.. n7

ii nco 0a k oa

LA L

en H t41

C'0 LA N C% HnN N 0 NjN

-n H- HV H H H H

LAr(4 .4 0 '

N wC14 LnA 0 N HY N n%r r r H H H H ar 0 0 0

co 00 0 0 C 0 40 c

HH4 HH C> O% 00 00m

H' 0- 4 I-I 0 0q 0- r- 0

00 00 00C 00 00 00 0

0 C: 0 C:'

N0 m Ho c to0

04 4 41 0 0 cn ( M

"n HO ON v c44 4 t4 r4NC,4 t H

U! LCa o in wH j 'N r o LAS4 %0 t%0 L4% n %0%0 .0 0n C0 0 0

LA4 '0 'ý 0 *'0 *1

(' (A1 N(m'

58

co 0

C14 0 co

C4 4n in

51 N~ m 0n C%4 04 N

N N

*- r 4 r 4 4- rir4 H 04 r4 I

N 0 0' 0 cc 0

r4 4 4- ) r4 a

0O C'0C Ln

DN NNN

H Noj H - 0L 0 I I0

4~L 4 44 4ý u; u 44

00~ 00 0 00 0 0 0014 LA v n 0n A t ý 4 ;

W- co CO 0 Nm'0N N C1 N(1

595

0n Ný ON f

H cN N Nn

H co

N N- C) N 0

N qt C14 v COI N Co nt04i #-I CO e4COr- C-) H .0

oo c C a 0 a C. 10 0

H - H H H H H qC H

0 0 0 0 00

0 000 0m0 00 m0 00

N

HO HO HO HO HO 0C N mko 01

UN 0% CO H oc c % (%

N? N N HN H

60

N Hn 0 ot

N M LA L

L~L

%D CN 0

Hn H N

N 0 N 0

0N H% 0 N %0 0

00 H0 0 H 0 0 00

0'. 1ý H'ý U! A C

00 00t 00 00 0 00 00

(1 I 44 C14 I 44 C1

2 __n __ _I _ _t l m (%0ON ' -0 OCq N Oi HOI 0O n

v-~ HO HO HO aH HO HH % k) nN

00 00I 0D0 00 00 00 A0

H 0 co Ch H

_ _ m_

_ _ _

ON HOH~j ~ He N

'~CO '~O 61 H ~ ~ Oi e{7

N H Oh 0ot

Ln N N coM

%D C% %; co

'3 co to 0

N~C CC0 0'

r-ro Li1- H N0 N- VH 0 0

N0 N0 '0 0 0 o0 0 0

O Hr Cm N 0 H% 0 0 G

OH ( Fe-4 H 0H 00 f-H 040 00400 00 00 00 00C 00 00

00 0 00 00 0 0 00

H1 . * o ** *

lu #3 0 '

C44 c4f N N N

No %DN MH LAA '3 i C% MN

ON ~ ~ ~ L HOcoHLAA cH N* * * * * * . * * .m

NN ~ J LA Hfn

61.

0! Cr) 1 0

0n C#) 0 m m

0 (nc

H HI H

Hv N N Hn H%

2:- I 2:i r- q .i 1 2:1 0 7 01HH 0H H0 0 00 00 00

Ho HH HOri0 - HO Hi HH4 H00 00 00 00 00 0 0 00

4JC %co H- 0

0 0 00 (I m %0.q

rN C1 (7N N N

NDC oc inO r- 00 i r -CAin0

V!r 1Aý CCH HHý HH7-ý ý1 71

C t I 'cco co m i C%)' C% N N r O

IV~r

62

0 c

HH

IT C%

tn 2VN

00

('4 `V *. .o

.. .

U, 0 o

C41

0. So 5

('IC' ('63

REFERENCES

i. Alperin, Morton; Harris, Gordon; and Smith, Charles"An Experimental Investigation of New Concepts inEjector Thrust Augmentation for V/STOL" TechnicalReport AFFDL-TR-69-52, June 1969 (Flight DynamicsResearch Corporation)

2. Alperin, Morton; Marlotte, Gary "A Jet Flap DiffuserEjector" Technical Report AFFDL-TR-71-66, June 1971 (FlightDynamics Research Corporation)

3. Fancher, Richard B. Capt. "A Compact Thrust AugmentationEjector Experiment" Technical Report ARL 70-0137, August1970 (Energetics Research Laboratory, Aerospace ResearchLaboratory, AMSC)

4. Drummond, A.N.: and Gould, D.G. "Experimental ThrustAugmentation of a Variable Geometry, Two-DimensionalCentral Nozzle Ejector" Aeronautical Report LR-328, January1962 (National Research Council of Canada)

5. Scott, W.J. "Experimental Thrust Augmentation of aVariable Geometry, Two-Dimensional, Coanda T..ll JetEjector" Aeronautical Report LR-34, January .964 (NationalResearch Council of Canada)

6. Morel, Jean-Pierre "Theoretical Solutions for the JetFlap Diffuser" Thesis, California Institute of Technology,May 1969.

7. Morel, Jean-Pierre and Lissaman, Peter B.S. "The JetFlap Diffuser: A New Thrust Augmenting Device" A/AA PaperNo. 69-777 July 1969, California Institute of Technology.

64