M72- 143 02

NASA TECHNICALM E M O R A N D U M

NASA TM X- 67979

IX

<to<

ra, (p&f A (gjp

Y

PEAK AXIAL-VELOCITY DECAY WITH SINGLE-

AND MULTI-ELEMENT NOZZLES

by Uc He von Glahn, D0 E. Groesbeck, and R0 G. HuffLewis Research CenterCleveland, Ohio

TECHNICAL PAPER proposed for presentation atTenth Aerospace Sciences Meeting sponsored by theAmerican Institute of Aeronautics and AstronauticsSan Diego, California, January 17-19, 1972

https://ntrs.nasa.gov/search.jsp?R=19720006653 2018-02-13T23:22:00+00:00Z

PEAK AXIAL-VELOCITY DECAY WITH SINGLE- AND MULTI-ELEMENT NOZZLES

by 0. H. von Glahn, D. E. Groesbeck, and R. G. HuffNational Aeronautics and Space Administration

Lewis Research CenterCleveland, Ohio

Abstract

Jet peak-velocity decay data were obtainedfor a variety of circular and noncircular single-element and multi-element nozzles for applicationto externally-blown-flap STOL aircraft. Thesedata permit a rational approach, in terms of ele-ment type and element spacing, to nozzles designedto promote mixing of the jet exhaust with the sur-rounding air. Rapid mixing and the resultinglower axial jet velocity decreases the noisecaused by the interaction of jet impingement onthe flap assembly of EBF STOL aircraft. Empiricalrelationships are presented that permit the pre-diction of peak axial-velocity decay curves for awide spectrum of mixer-type nozzles. The data areuseful also in the design of ejector-type noisesuppressors and for the suppression of VTOL down-wash velocities caused by vertically orientedexhaust nozzles.

Introduction

The air transportation system of the 1980 'sis expected to include substantial numbers of STOLaircraft. Of the several lift augmentation con-cepts proposed for STOL aircraft, the presentstudy is concerned with externally blown flaps(fig. 1). Experimental studies have shown thatthe impingement of the engine exhaust jet on de-flected flaps can cause an unacceptable increasein the aircraft noise signature. vl~3) The in-crease in noise level is a 6-power function of theimpinging jet velocity on the flap surfaces^ ' andis also proportional to the surface area scrubbedby the jet. The jet-flap interaction noise can belowered by reducing the impinging velocity on the

^>2) This velocity reduction generally mustbe accomplished in a specified distance from thejet exhaust plane to the flap.

The jet velocity impinging on the flap can bereduced by: (1) a reduction of the jet velocityat the exhaust by utilizing a large bypass-typefan engine and (2) by use of a mixer- type nozzle,consisting of multi-elements rather than a singlelarge exhaust nozzle of equal total area. Theindividual small elements of a mixer nozzle pro-mote an initially rapid mixing with the surround-ing air resulting in a rapid axial velocity decay.

It is the purpose of this paper to summarizethe results of an experimental study, (4) conductedat the NASA Lewis Research Center, on the peakaxial -velocity decay obtained with circular andnoncircular single-element nozzles and severalmulti-element mixer nozzles. Empirical equationsare developed for estimating peak axial-velocitydecay curves for a wide range of nozzle configu-rations.

Apparatus

Test Stand

The test stand used in the present work is

shown in figure 2. Pressurized air at about 289 Kis supplied to a 15.25-cm diameter plenum by twindiametrically opposed supply lines. Flexiblecouplings in each of the twin supply lines isolatethe supply system from a force measuring system.In the present study no thrust measurements wereused. The test nozzles were attached to a flangeat the downstream end of the plenum.

Airflow through the overhead main supply linewas measured with a calibrated orifice. The nom-inal nozzle inlet total pressure was measured witha single probe near the plenum exit flange.

Free jet surveys were made with a traversingpitot-static probe at several downstream stations(up to about 50-cm) from the test nozzle exitplanes. Thereafter, a single pitot-static probelocated on the approximate centerline of the noz-zle was used at downstream distances up to 300 cm.

The measurements from the traversing probewere transmitted to an x-y-y' plotter whichyielded direct traces on graph paper of the totaland static pressure distribution radially acrossthe jet. All other pressure data were recordedfrom multitube water or mercury manometers.

Nozzles

Peak axial velocity degradation data were ob-tained with single-element nozzles that includedthe following cross sections: circular, trape-zoidal, triangular, and rectangular. The studiesincluded variations in nozzle aspect ratio andnozzle area.

Multi-element nozzles included multitubes (upto 19 tubes) and multi-lobe nozzles (up to12 lobes). The spacing between elements and cir-cumferential rings of elements (for multitube noz-zles) were studied to evaluate these geometry ef-fects on peak axial-velocity degradation.

Procedure

Initially, the traversing probe was posi-tioned 0.317-cm from the nozzle exit plane andradial pressure-traverses were made in the planeof the largest element dimension (i.e., diameterof a circular element, width of a rectangular ele-ment, etc.). Pressure measurements were obtainedat nominal nozzle pressure ratios of 1.15, 1.31,1.53, 1.87, and 2.3. This procedure was then re-peated at nominal distances from the nozzle exitplane of 13-, 25-, 3%-, and 51-cm. At the sametime a single pitot-static probe was moved manu-ally to various locations from 100- to 300-cmdownstream from the nozzle exit plane and pressuredata recorded.

Results and Discussion

General

According to the literature, (5,6) (-he peak

axial velocity decay, U/Uj, downstream of the jetcore varies as a function of X~̂ - for circularnozzles to X~l" for infinite or large aspectratio rectangular (slot) nozzles. (All symbolsare defined in the appendix.) For other singleelement geometries the decay appears to vary be-tween these exponents. The axial distance is non-dimensionalized by the effective diameter of theelement; i.e., X/De.

For multi-element nozzles, the initial peakaxial-velocity decay (fig. 3) is substantially thesame as that for an individual element. However,at some distance downstream of the nozzle exitplane, the individual jets coalesce sufficiently toform a large diameter coalescing core and a veryslow peak-velocity decay occurs. Once the coa-lesced core has fully formed, normal mixing againoccurs with an associated rapid velocity decay.The literature^>") nondimensionalizes the multi-element nozzle decay distance by use of an effec-tive diameter based on the total nozzle exhaustarea. However, from the wide range of configura-tion variables covered herein, it was determinedthat the effective diameter of a single-elementwas more useful for correlation purposes and prac-tical applications.

The peak velocity ratio, U/Uj, at a givenaxial station has been found to increase with in-creasing jet Mach number.('•' In general, corre-lation of the jet Mach number was achieved bydividing the axial distance parameter, X/De, by

Data for representative single-element(conical convergent nozzle) and multi-element(multitube nozzle) configurations are shown in fig-ure 4 using the /I + M.S factor to correlate vari-ations in Mj. The Mach number correlation factorfor the multitube nozzle applies to the entire de-cay curve including the coalescing core and coa-lesced core regions. In order to avoid confusingsubsequent figures with a large number of datapoints, hereinafter only the nominal 0.99 jet Machnumber data, unless specifically noted, will beshown.

The following sections will present decaydata for a variety of single-element nozzles andmulti-element configurations together with corre-lating equations. Typical radial profile of ve-locity ratio at the point of departure of the coa-lescing core from the single element curve alsowill be shown and the jet spreading angle will bediscussed briefly.

Single-Element Nozzles

Four basic single-element nozzle types werestudied in order to provide sufficient informationfor establishing correlation equations on which tobase multi-element mixer nozzle designs. Thesebasic nozzles consisted of circular, rectangular,trapezoidal, and triangular cross-sections. It wasestablished early in the program that the locationof a baseplate at the nozzle exhaust plane did notaffect the peak axial-velocity decay of single-element jets. Consequently, many of the single-element configurations were simple orifice-typenozzles rather than tubular-type nozzles.

Circular nozzles. The peak axial-velocitydecay data for a conical convergent nozzle (7.63-cmI.D.(7)) were used as a standard for comparison ofthe present tube (2.36-cm I.D., 10-cm long) and

circular orifice data (2.46-cm I.D.) obtained inthe present study. The velocity decay data forthese configurations are shown in figure 5 interms of the ratio of the local peak velocity tothe jet exhaust velocity, U/U^, as a function ofthe decay distance parameter X(CnDe/l + Mj)~l.It is apparent that the velocity decay data forall four nozzle types are identical when an appro-priate nozzle coefficient, Cn, is used. This co-efficient is a function of three variables:

(1) The ratio of the measured mass flow tothe calculated ideal mass flow.

(2) An entrance factor including a secondarydependency (<10-percent) on the number of elementsused. (The latter is believed unique to the sup-ply line and diffusers used upstream of the testnozzles in the present test rig.)

(3) A jet Mach number correction for the ori-fice configurations. The overall coefficients (Cnvalues) are shown on the succeeding figures.

The curve shown in figure 5 is calculatedfrom the following empirical equations:

1 +

-I/a

(1)

The exponent a is a complex function of nozzleexit geometry. Over the range of nozzle geome-tries studied herein, the following equation, ob-tained by crossplotting of experimental data, pro-vides good correlation of single-element velocitydecay data:

- I))'1a = 4(2 - (b/b))l + (D/D - I)' (2)

For a circular nozzle the a-exponent reduces to 4.

Rectangular nozzles. The peak axial-velocitydecay data for four rectangular-orifice nozzleswith aspect ratios (length/height) varying from1.5 to 12 are shown in1 figure 6 together with cal-culated velocity decay ' curves. (All pertinentnozzle dimensions are summarized in table I.) Ingeneral, the data show that with increasing aspectratio, the peak velocity initially decays morerapidly. As jet mixing with the surrounding airproceeds farther downstream from the nozzle exitplane, the velocity decay for all nozzles becomessubstantially the same.

Triangular and trapezoidal nozzles. Data forsingle-element triangular -orifice nozzles and thetrapezoidal nozzles O fell within the envelope ofthe rectangular nozzle data shown in figure 6 andare correlated well by equations (1) and (2).

Multi-Element Nozzles

The peak axial-velocity decay curves for sev-eral categories of multi-element nozzles , photo-graphs of which are shown in figure 7, were deter-mined. The first category consisted of coplanarmultitube nozzles including single and multiplerings of tubes, the second consisted of singlering coplanar multi-lobe nozzles while the finalcategory consisted of several specialty, non-

coplanar nozzles.

Empirical equations were developed to corre-late the peak axial-velocity decay of these multi-element nozzle types in terms of the significantflow regimes shown previously in figure £ and interms of pertinent nozzle dimension parameters.These equations are limited by the general nozzlegeometries tested; however, they are useful in pre-dicting the decay curves for many practical nozzleconfigurations.

Correlation

' The decay curve was divided into severalregions shown in figure 8. Equations were thendeveloped to predict the departure point of thecoalescing core from the single-element decay curve(point (T)) . Examination of the data showed thatthe velocity ratio in the coalescing core decayregion had a slope of -0.2 with respect to axialdistance (region denoted by (T) to Q2~)) . The non-dimensional displacement parameter, D^ , of thecoalesced-jet decay curve from the single-elementcurve was then determined. The value of U/Uj atpoint (T) was then correlated in terms of the coa-lescing core decay slope and the displacement dis-tance DX. Finally, the correlation equation forthe decay curve of the fully coalesced core wasestablished.

Empirical equations for the preceding pointsand regions are given in the following paragraphs.

The departure pointfollowing equation:

is calculated by the

where

1 +

-i

(4)

and

'(i)

-1

1 + 0.33 - (5)

The following table summarizes the necessary ratiosof (r/s) and f(w) for use in equation (5).

Nozzle

Center element with 1 or 2rings of multi-elements

r/s

rl/sl

f(w)

2 rings of multi-elements, r2/s2no center element

The displacement of the fully coalescedmulti-element core from the single element curve(Dx in fig. 8) is calculated from the followingequation:

1 +

-1/2

(6)

When the U /U values are 1, the terms c

in equation (6) reduces to

The velocitv decay in the coalescing core(region (T) to (2\) is given by the.following re-lationship :

©-©(7)

In the coalesced core region (fig. 8) the ve-locity ratio is given by:

1 +

-I/a

0+0(8)

The intersection of curves calculated from equa-tions (7) and (8) provides the location of point(T) in figure 8.

Coplanar Multitube Nozzles

Multi-element nozzles consisting of six cir-cular tubes (each 2.36 cm I.D. and 10.16 cm long)were studied to determine the effect of the cir-cumferential spacing between adjacent tubes on thepeak axial-velocity decay. (Pertinent nozzle di-mensions and configuration code designations aregiven in table I. Typical data obtained withthese nozzles are shown in figure 9 together withcalculated decay curves. Initially, the velocitydecay curve for all nozzles coincides with thatfor a single tube. As the circumferential spacingbetween the individual jets was decreased, the de-parture point of the coalescing-core curve fromthe single-tube curve occurs at increasinglylarger values of U/Uj (i.e., shorter axial dis-tances from the nozzle exit plane).

In order to accommodate larger total flows(within the requirements of achieving a required

velocity decay in a given axial distance), multi-ple rings of multi-elements are required. Nozzlesconsisting of two rings of multitubes each with6 tubes in the inner ring and 6 and 12 tubes, re-spectively, in the outer ring (all tubes had anI.D. of 2.36 cm) were studied to determine the in-terference effects of the outer ring of multitubeson the peak axial-velocity decay. As shown infigure 10 the use of multi-rings of multi-elementscaused the departure point of the coalescing corefrom the single element curve to occur at increas-ingly large values of U/U^ as the number of ele-ments in the second ring increased (decreasingcircumferential spacing ratio).

The addition of a center tube on the axis ofmultitube nozzles also caused a significant in-crease in the U/U^ value at which the coalescingcore departed from the single-element curve(fig. 10). The departure point from the singleelement curve, however, was determined by the cir-cumferential spacing ratio of the inner ring oftubes. The outer (second) ring of tubes did notinfluence the departure point (for nozzles with acenter tube) provided the circumferential spacingof these tubes was equal to or greater than thatfor the tube spacing of the inner ring.

Coplanar Lobe-Type Nozzles

Two lobe-type nozzles were studied. Thefirst type consisted of flat-ended trapezoidaltubes while the second type consisted of rounded-ended trapezoidal tubes (figs. 7(b) and (c), re-spectively) .

Three flat-ended trapezoidal nozzles with thesame total area were tested to determine theeffect of element spacing and element number onthe peak axial-velocity decay. Data obtained withthese nozzles are shown in figure ll(a) togetherwith the calculated velocity decay curves. It isevident from the data that, for these nozzles,doubling the number of elements while maintainingthe radial height and the circumferential spacingratio constant caused only a small increase in theU/Uj value at the departure point from thesingle-element curve. As would be expected, anincrease in the circumferential spacing ratio from1.0 to 3.0 (by increasing the element radialheight and reducing its width) for the six-elementnozzle resulted in a significantly lower U/Ujvalue at the departure point.

The peak axial-velocity decay for a round-ended trapezoidal 8-lobed nozzle is shown in fig-ure ll(b) together with calculated velocity-decaycurves. The data trends shown are similar tothose for the flat-ended trapezoidal nozzles.

In order to promote a greater decay of thepeak axial-velocity, alternate lobes of the 8-lobed nozzle were canted 10° outward from the noz-zle centerline. Canting the alternate lobescaused the velocity decay to be reduced by aAU/Uj of about 0.12 over that with the uncantedlobes. The analysis presented previously hereinhas not been extended to include the effects oflobe canting; consequently, the curve shown infigure 11(b) is estimated rather than calculated.

Nonplanar Multi-Element Nozzles

This section presents peak-axial velocity

decay data for several noncoplanar-type nozzlestypical of high-bypass fan-jet engines in whichthe bypass exhaust plane is some distance upstreamof the core jet exhaust plane. Such nozzles mayhave a difference in velocity between the twojets; consequently, the data presented herein wereobtained at nominal secondary to core velocityratios, US/UC, of 0.7 and 1.0. The velocity decaydata shown are presented for a nominal core-jetMach number of 0.99.

Multitube bypass-type nozzles. The firstnozzle tested consisted of 8 tubes (each 1.4-cmI.D.) for the core jet and 8 round-edge orifices(each 2.54-cm I.D. and 10.16 cm long) for the sec-ondary jet. The core and secondary jets werealigned radially. The peak axial-velocity decaydata for this configuration is shown in figure 12together with the calculated velocity decaycurves. The data analysis showed that, for thepresent configuration, the core jet determined thepeak axial-velocity decay. Therefore, the valueof De in the abscissa of figure 12 is that forthe core tube and the axial distance, X, is meas-ured from the exit plane of the core jet. A re-duction of the jet velocity ratio, US/UC, from 1.0to 0.7 caused a decrease in U/Uj at the point ofdeparture from the single-element curve. The noz-zle was tested with and without a conical centerplug between and around the core tubes. Use ofthe plug did not affect the departure point.

The velocity decay obtained with a 3-tubecore nozzle (2.36-cm I.D. tubes, no center plug)of about equal area to that for the original8-tube core nozzle, is also shown in figure 12 fora nominal US/UC value of 0.7. The use of thesmaller number of large-diameter tubes caused anincrease in the U/Uj value at the departurepoint compared with the original 8-tube core con-figuration.

Radial Profiles of Velocity

The radial profiles of jet velocities at thedeparture point of the coalescing core from thesingle-element curve for multi-element nozzles areneeded in order to calculate the jet-flap inter-action noise for externally-blown-flap STOL sys-tems. In the present study, these profiles wereobtained from the jet velocity contours which, inturn, were obtained from the radial profiles ofvelocities measured at specified axial stationsdownstream of the nozzle exhaust plane. Typicalprofiles are shown in figure 13 in terms of U*/Ujas a function of R/Rn. When the departure pointoccurs at a high value of U/Uj (fig. 13(b)) theflow characteristics of the individual elementsare easily identified in the profile. On theother hand, when the departure point occurs atlower values of U/Uj , the profiles are morenearly uniform up to a jet radius ratio approach-ing 0.8. For considerations of STOL jet-flap in-teraction noise, this suggests that a larger areaof the flap is scrubbed by the impinging jet froma mixer nozzle compared with that for a conven-tional nozzle. Since the jet-flap interactionnoise is proportional to AU!j,(l) the noise re-duction expected from the use of a lower jet im-pingement velocity at the flap is not fullyrealized due to the smaller but opposite effectof the increased scrubbed area on noise genera-tion.

Spreading of Multi-Element Jet Wakes

The jet spreading half-angles for severalmulti-element-nozzle types are summarized in fig-ure 14. The half-angles noted are averages forthe regions indicated in the accompanying sketchof a typical velocity decay curve.

The data showed that these half-angles werea function of jet mixing (velocity decay) as il-lustrated by the decay curve. Initially in thesingle-element core region (U/Uj - 1), denoted byregion A in figure 14, the spreading half-anglevaried from 6° to 8°. The larger angles were gen-erally associated with those nozzles having thelargest spacing between adjacent elements. As thejet velocity began to decay (region B) the jetsbegan to fill in Che center portion of the config-uration resulting in a spreading half-angle ofonly 1° to 4°. As coalescing between adjacentjets proceeds, including the departure point ofthe coalescing core from the single-element curve(region C), the jet spreading half-angle increasedto near 6°. Thereafter, region D, the angle againdecreased somewhat, with values in the range of3° to 5°.

The preceding jet-wake spreading half-anglesare useful in determining the vertical distancebelow a wing that a jet exhaust must be located inorder to avoid scrubbing the wing surface with ahigh velocity jet for externally-blown-flap STOLaircraft. Furthermore, these half-angle data arerequired in the design of ejector shrouds usingmulti-element mixer nozzles in order to determinewhen the flow fills the ejector.

Concluding Remarks

In the use of a mixer nozzle for reducing thejet-flap interaction noise from an externallyblown flap for STOL aircraft applications, notonly must the effect of the reduction of the im-pinging velocity on the flap be Considered, butalso the larger jet impingement area on the flap.This increased area is caused primarily by thelarger overall dimensions of the mixer-nozzle jetcompared with that for a conventional circular-nozzle jet. Thus, the full jet-flap interactionnoise benefits resulting from the velocity decayassociated with a mixer nozzle may be significantlyreduced by the larger jet impingement area.

For a given velocity decay requirement, theminimum number of elements for a multi-elementnozzle appears to be obtained when the designvalue of X(CnDe/l + Mj)~l for the nozzle is atthe departure point of the coalescing core fromthe single element curve. This criteria could re-sult in only a small number of elements for agiven nozzle application. Although this designcriteria should cause minimum internal flow lossesand external drag increases, it could result inlittle, if any, of the jet exhaust noise suppres-sion commonly associated with multi-element noz-zles. From the point-of-view of jet exhaust noisereduction, therefore, it may be desirable to usemore but smaller elements and accept some smalldrag increase, due to an increase in the overallnozzle size, for the aircraft cruise condition.

On the basis of the preceding brief remarks,there are obvious performance trade-offs and com-

b,h,L,!t,R,r,s,w

promises that can be exercised in the design ofmixer nozzles for specific applications. The em-pirical relationships for predicting peak velocitydecay curves for jets presented herein are an im-portant step in establishing rational design pro-cedures for mixer nozzles. Use of present tech-nology for predicting internal nozzle-flow lossesand aerodynamic penalties associated with thelarger mixer-nozzle surfaces and cross-sectionalprofile can provide the additional necessary in-formation to achieve optimum mixer nozzle configu-rations .

area of core nozzle, cm

2area of single element, cm

area of bypass nozzle, cm

nozzle dimensions (see Table I), cm

effective nozzle (or orifice) coeffi-cient

effective diameter of circular nozzlewith exit area equal to that of non-circular single element (De for acircular nozzle equals the nozzlediameter), cm

effective diameter of circular nozzlewith exit area equal to that of totalmulti-element nozzle area, cm

hydraulic diameter of nozzle element,cm

analytical displacement parameter

jet Mach number

overall nozzle radius, cm

ratio of effective spacing between ad-jacent jets (including nozzle wallthickness) at nozzle exit plane toeffective element width (see table I)

local peak axial-velocity of jet, m/sec

local jet velocity, m/sec

jet exhaust velocity, m/sec

ratio of bypass jet velocity to corejet velocity

axial distance downstream of effectivenozzle exit plane ̂

e,T

M.3

Rn

s/w,r/w

U /Us c

Subscripts:

nozzle

outer

center tube

1 first ring of multi-elements

2 second ring of multi-elements

REFERENCES

1. Dorsch, R. G., Krejsa, E. A. and Olsen, W. A.,"Blown Flap Research," TM X-67850, 1971,NASA, Cleveland, Ohio.

2. Kramer, J. J., Chestnutt, D., Krejsa, E. A.,Lucas, J. G., and Rice, E. J., "Noise Reduc-tion," Aircraft Propulsion. SP-259, 1971,NASA, Washington, D.C., pp. 169-209.

3. Goodykoontz, J. H., Olsen, W. A. and Dorsch,R. G., "Preliminary Tests of the Mixer NozzleConcept for Reducing Blown Flap Noise," TMX-67938, 1971, NASA, Cleveland, Ohio.

4. Groesbeck, D., Huff, R. and von Glahn, U.,"Peak Axial-Velocity Decay with Mixer-TypeExhaust Nozzles," TM X-67934, 1971, NASA,Cle ve land, Oh io.

5. Landis, F. and Shapiro, A. H., "The TurbulentMixing of Co-Axial Gas Jets," Proceedings ofHeat Transfer and Fluid Mechanics Institute,Stanford University Press, Stanford, Calif.,1951, pp. 133-146.

6. Weinstein, A. S., Osterle, J. F. and Forstall,W., "Momentum Diffusion From a Slot Jet Intoa Moving Secondary," Journal of Applied Mech-anics. Vol. 23, No. 3, Sept. 1956, pp. 437-443.

7. Higgins, C, C. and Wainwright, T. W., "DynamicPressure and Thrust Characteristics of ColdJets Discharging from Several Exhaust NozzlesDesigned for VTOL Downwash Suppression," TND-2263, 1964, NASA, Washington, D.C.

8. Higgins, C, C., Kelly, D. P. and Wainwright,T. W., "Exhaust Jet Wake and Thrust Charac-teristics of Several Nozzles Designed forVTOL Downwash Suppression," CR-373, 1966,NASA, Washington, D.C.

TABLE I NOMINAL NOZZLE DIMENSIONS

Circular Tube-Type Nozzles(Tube I.D., 2.36 era)

Nozzletype

0-6-0*

e,T'cm

2,cm

Vcm

sl/wl S2/W2 rl/wl r2/wl

3. 11

8.2010.0

8.58.5

10.3

J. J-O

5.085.08

3.185.045.04

10.10.

6.10.10.

22

3522

11.11.

7.11.11.

jj

44

5544

ii

11

.15

.15

.344

.15

.15

3.31.25

1.693.31.25

1.1.

3441515

1.151.15

.3441.151.15

0-6-60-6-121-6-01-6-61-6-61-6-12*Multitube nozzle designationscenter tube is used, secondtubes in the first ring andof tubes in the second ring

first number indicates ifnumber indicates number ofthird number indicates number

Rectangular Nozzles

Aspect ratio

12631.5

Height,h,cm

1.272.545.085.08

Length,L,cm

15.25

I

7.62

Trapezoidal Nozzles

(a) Flat-ended

Number of De,elements

1266

De,cm

2.183.143.19

De,T-cm

7.557.707.80

h,cm

3.153.155.23

cm

0.861.610.90

cm

1.563.302.16

R0'cm

3.073.003.06

cm

6.36.28.4

si/w:

1.151.013.06

(b) Round-ended

3.81 10.8 6.25 1.19 2.51 2.59 8.8 1.37

Noncoplanar Nozzles

Number ofelements

e,cm

Core Bypass Core Bypass

1.412.36

2.542.54

4.062.54

10.210.2

sl/wl

- ovOI

W

Figure 1. - Externally-blown-flap STOL airplane.

SECTION A-A

AIR FLOW

OUTLETORIFICE

STATIC TAPS-y ~~~—^FLEXIBLE JOINTS

ORIFICE PLATE Vh ^THERMOCOUPLE TYPE K

'AIR FLOWr NOZZLE MOUNTING1 FLANGE

SUPPORT CABLESi |

r TOTAL PRESSUREPROBE

AL AXIAL THRUST LOAD CELL

1.22MJL

FLOOR

Figure 2. - Schematic diagram of test rig.

SINGLE ELEMENTDECAY

l.Or

COA-^LESCED.

COREDECAY

10 100RATIO OF AXIAL DISTANCE TO EFFECTIVE

DIAMETER OF ELEMENT —+

Figure 3. - Schematic of multi-element nozzle peakaxial-velocity decay.

i.o.8.6

.4

O

3

Jet Machnumber

o .36a .78o 1.18

(a) CONVERGENT NOZZLE; DIAM, 7.62 CM;REFERENCE 7.

1 10AXIAL DISTANCE PARAMETER, X(De

(b)6-TUBE NOZZLE; TUBE DIAM, 2.36 CM; CIRCUM-FERENTIAL TUBE SPACING, 2. 7 CM.

Figure 4. - Typical peak axial-velocity decay obtained atvarious jet Mach numbers with single- and multi-element nozzles.

1.0.8

.6

oo

<£

.2

NOZZlf TYPE

CONICAL CON-VERGENT

TUBEROUND-EDGE

ORIFICESQUARE-EDGE

ORIFICE

DIAM, REF. C

7

CM

7.62

2.46

2.46

CALCULATED

.4 .6 .8 1

n

0.91

2.36 PRESENT .91.91

.81

.2

Figure 5. - Peak axial-velocity decay obtained withcircular nozzles. Nominal jetMach number,0.8.

E-6703

1.0

Aspect Orificeratio, entranceL/h

Nominalcn

0.67

40 50

Figure 6. - Peak axial-velocity decay obtained with rectan-gular orifice-type nozzles.

(a) MULTITUBE NOZZLE. (b) FLAT-ENDED TRAPEZOIDAL NOZZLE.

C-71-4014

(cl ROUND-ENDED TRAPEZOIDAL NOZZLE. ALTERNATE LOBESCANTED 10° OUTWARD FROM NOZZLE CENTERLINE.

Figure 7. - Typical multi-element nozzles.

(d) BYPASS-TYPE NOZZLE WITH 8 CORE TUBES AND 8 SECONDARY-FLOW ORIFICES.

1.0

o ̂o •=)

.1

DEPARTURE POINT OF COALESCING CORE FROMSINGLE-ELEMENT CURVE

START OF FULLY COALESCED CORE DECAYCOA-

SINGLE-ELEMENTDECAY

SINGLE ELEMENTDECAY CURVE

COALESCINGCORE

10

LESCEDCORE

DECAY

100

Figure 8. - Significant mixed-flow regimes formulti-element nozzles.

CIRCUMFERENTIALSPACING RATIO,

Sj/Wj

iff^5o"i__

o;

o3

1.0.8

.6

.4

o

o 0.344o .615

- o 1. 15* 1.83

1^— — a CALCULATED— """̂ v.

~ ° ° ^^S r̂\jOr ^^^^^^^*^^3 .̂

-o o -̂~ !̂>.- ° ° smx ?^§

ELEMENT^' \

ll L4 6 8 10 20 40 60 80 100

Figure 9. - Effect of element circumferential spacing onpeak axial-velocity decay for 6-tube nozzle.

_i.o1 -8

o -6

^ .4

ooSi .2

.1

MULTITUBE CONFIGURATION

- n - o-o "o o~ o

I i I i I6 8 10

X(CnDe

60 80100

Figure 10. - Effect of center tube and/or multi-ringsof tubes on peak axial-velocity decay of multitubenozzles. Nominal C> 0.828; S/w PJ/WJ =

LOBE CIRCUMFERENTIALNUMBER SPACING RATIO,

Sj/Wj

1.0

.8

.6

.4

SINGLE ELEMENT

Qi

OO

£ 1.0.8

.6

.4

(a) FLAT-ENDED TRAPEZOIDAL NOZZLES; NOMINALCn, 0. 77.

LOBE CANT ANGLE,DEC

o 0 ALL LOBESQ 10 ALTERNATE LOBES

CALCULATEDESTIMATED

l\ \ I i -I i 1 i4 6 8 10 20 40 60 80 100

(b)8-LOBED, ROUND-ENDED TRAPEZOIDAL NOZZLE; NOM-INAL Cn, 0.67.

Figure 11. - Peak axial-velocity decay obtained withtrapezoidal multi-element nozzles.

COO

iW

1.0.8>.6

oD

o

OO

NUMBEROF CORE

ELfMENTS

JET EXHAUSTVELOCITY RATIO,

us/uc

NOMINALCn

6 8 10X(CnDe

40 60 80100

Figure 12. - Typical peak-axial velocity decay fornon-coplanar multi-element nozzles. Secondarynozzle, 8-circular orifices.

NOZZlf SPACINGRATIO,sl'wl

NOZZlf SPACINGRATIO,

1.151.15.344

(a) SINGL£-RING MULTI-TUBE NOZZLES.

(b) DOUBLE-RING MULTI-TUBE NOZZLES.

NUMBEROF LOBES

SPACINGRATIO,sl/wl

.4 .8 0 .4JET VELOCITY RATIO, if/U;

(c) TRAPEZOIDALLOBED NOZZLES.

(d) NON-COPLANARNOZZLES (8-TUBECORE NOZZLE).

Figure 13. - Typical radial profiles of jet velocities atdeparture point of coalescing core from single-element decay curve.

1.0

JET WAKE REGIONSA—-I— BH—C

or^vO

W

oo

.1100

AVERAGE JET SPREADING HALF-ANGLE,DEC

NOZZLE TYPE

MULTITUBESSINGLE RINGDOUBLE RING

TRAPEZOIDAL LOBESNON-COPLANAR

us/uc, i.oUS/UC, 0.7

JET WAKE REGION

A

6±L7il8±1

—— ——

B

2-31

3-4

23

C

66il

6

45

D

55

4±1

34

Figure 14. - Relation of jet-wake spreading half-angleto velocity decay.

NASA-Lewis-Com'l