m. TX CH1?ICAIINOTES

NAT 1017AL ADVISORY COMM IT!?EE!-- .

,: ------

: - .“,:

Npw. 399

.—

4+-.d.

. . .

YOR ARRONAUT ICS

SOME CHARACTERISTICS OF FUEL SPRAYS A!?

LOW- INJECTION PRESSUR%S .. ‘-;.;->

3y A. M. Rothrock and C. D. Waldron.

J

10 bu returrie~ lothe Illes of the Langley “Memorial Aeronautical a

Ld3mtwYs .. .

Washington’Ifovenber, 1931

--

. .

. -.....

..

‘uf’.-.=.> .,..“

“b;...

_, ,,.._._,, .

- ““:-.,::::”2”“;;;;;:;-:..... ............,.~_:.:,.- —:NATIONAL ADVISORY COMMITTEE FOR.AE~O~AUTICS

TECHNICAL NOTE NO. 399

SOME CHAFUCTERISTICS OF INJEL SPRAYS AT

LOW-INJECTION PRESSURIS

,. By A; M. Rot~ro@c and C, D, Waldron

Summary

This report presents the. results.of tests conductedat the Langley Memorial Aeronautical Laboratory, LangleyField, Ta., to determine s~me of.the characteristics oftke fuel sprays. obtained from an 0.008-inch and a 0.020-inch open nozzle when injection pressures from 100 to 500poun$s per square inch were used. Fuel 0:1 and gasolinewere injected into air at densities of atmospheric and0;325 pound per cubic foot.

It was found that the penetration rate at these lowpressures was about the same as the rate obtained withhi~her pressures. Spray cone-angles were small and indi-vidual oil drops,were visible in all the sprays. Gasolineand fuel oil sprays had similar characteristics.

Introduction

Interest has-recently been shown in spark-ignitionengines using gasoline or Biesel oil and having the fuelinjected into the inlet manifold or into the engine cyl-inder during the suction and compression stroke when theair density is low. .Since in this type of engine injec-tion pressures of several thousand pounds per square inchare not required. to give t~e necessary fuel spray penetra-tion with round-hole orifices, a low-pressure injectionsystem could be used, provided that the low-injectionpressure gave sufficient atomization and dispersion ofthe fuel spray. Consequently, it seemed desirable to in- ~vestigate the characteristics of the sprays produced withlow-injection pressures. TO do t,his, several series of ..-”

high-speed motion picturesof such sprays were obtainedat the Langley Memorial Aeronautical Laboratory, Langley

.. . . . ...

Field, Va. This report presen$s the results of tho in-vestigation.

.,. .. .. ,. ... ,:, . . . :.-., .,,,.., ,., ,..

Apparatus”’and Methodsh,..,.:,.-..., ..--,

. .

The apparat,us.:wsecl was the N.A.C..4. spray photography

eqnipment which is descri%ed in reference 1. A diagram-matic sketch of the injection system is shown in Figure 1.The ball check valve mounted in the open nozzle holder isshown in I?igure 2. This system gave one injection of fueliato a chamber with glass windows while 25 electric con-deaser”s were successively discharged at tbe rate of 2,000per second across. a spark gap in a reflector. The lightfrom these discharges was focused l)y the reflector ontotho fuel spray.. In the camera a .fila moving at tha rat+ ___of 2,000 inches per second recorded—a picture of the sprayeach time it was illuminated hy a spark discharge.

The steel injection tube was 50 inches long and Badan inside and ah outside diameter of one-eighth and”one’-fourth inch, respectively.

,,The values of spray-tip penetration were obtained by

dr.i?.witiga smooth curve through,.the tips of the spray fmag.eson each film and ,then measuring the he”ight of this curveat various time positions. The start of the spray was con-sidered as the point at which this curve intersected ahorizontal line drawn throu:gh the images of the dischargeorifice.

?2he “images of the sprays were examined for distin-:;uishable fuel. drops. Nothing more than this could beleari~ed about, atomization in the sprays from the examina-tioil of.the phdt~graphs, for, according to Lee (reference2), th~ external appearance of a “spray gives little in-dication of the actual’ atomizatio,n ‘or distribution in thespray. Several pictures-were taken tinder the same con-ditions for each test. Thq variation in penetration wasilot large, l)eing atiout the same as that found by Beard-slby (reference 3) when he did not. control the initial‘.pies~ure in”’the injectio~ tu~e. The same trouble. wasexperienced with.,a~r, ir. the system as discussed in refer-ence: L&J.:The, kif?icvlty was elim$nated~ as before, bycarefully f.i.lling.thq line with fuel” %efor? each injec-tion. . ,.. ..

. .

N. A. C.A. Technical Note” Ndi 399 3

Tes”ts were ma&e with””a 0.020 -iach ,?rifice with alength-diameter ratio of 2 and an’ 0.008 di”arnet8rorificewith a length-diameter ratio”of’1/2,first with no checkvalve in the line and then with the ball check”valve,which had an opening pressure of 145 to 1S0 pougds per :.square inch. The injection pressures were varied from “‘;100 to 500 pounds per square inch. Bdth”atmospheric “(0.0765 pound per cubic foot) and 0.325 pound ~er cubic ‘:foot” (50 pounds per square inch pressure) chamber air .“,densities were used. The air in the chamber was at roomtemperature.

The fuels tested were fuel oil and gasoline. Thefuel oil had a specific gravity of 0.83.and a viscosityof 0.0221 poise at 100° F. The gasoline had a specificgravity of 0.75 and a viscosity of 0.00’79 poise at 100° F.

Test Results and Discussion

.

Figure 3 shows three series of photographs obtainedwith the gasoline. At an injection pressure of 100 poundshper square inch the individual “fuel drops were visible andthe spray cone-angle was indefinite. When the injectionpressure was increased to 500 pounds per square inch, thenumber and size of the visible drops decreased and thecone-angle was 12°. The spray appears to be better dis-tributed and atomized. When the, spray-chamber air densitywas increased from atmospheric to 0.325inch, the cone-angle was i.ncreas.ed to 25

~ound per squareand the distri-

h.ztion and atomization were apparently still further ”im-proved. The. appearance of the syrays with fuel oil wassimilar to those obtained with gasoline.

,.

,

●

Enlargements of photographs of the gasoline spraysare shown in Figure 4. At the injection pressure of 100pounds per. square inch the spray consisted of a centralcore of individual drops “surrounded by an envelope ofdrops of different sizes. The large i.niiistinct drops areprobably caused by smaller drops out of focus. At theinjection pressure of 300 pounds per sq=.are inch thereare a few individual drops visible on the edge of thespray, but most Of the spray appears to he well atomized.The cone-angle with the “gasoline was 3° to 4° greater thanwith the fuel oil.

4 N. A. C.A. Technical Note No. 399.,.

i?igures 5 and 6 show the effect of injection pressureand of the check valve on the penetration of the spraysfrom a 0.020-inch orifice. The penetration was in each ~case greater without the check valve (0 potind per square .inch Valve-”opening pressure) than with the check valve. .In each,f.igurti the curve for 100 pounds per square inchinjection pressure with the check valve’ in the line iscon?ide:ra%ly” less tb.an than the penetration under the other ‘conditions . This decrease in penetration is caused. lythe opening pressure, of the check valve being greater thanthe injection ~ressure maintained in the high-pressurereservoir. (See also reference 5.) For these conditions,under which curves A were obtained, the spray consistedof a stream of drops. With the exee”ption of the 100 “poundsper square inch injection pressure, the penetration ofthe fuel oil was slightly greater’”than that of the gas-oline, (See also reference 6.) The values of penetrationcompare favorably witi~ those obtained at much higher in-jection pressures. (Reference 7. )

.Tigures 7 and 8 show the effect of injection pres-

sure on the penetration of fuel” oil and gasoline spraysfrom an 0.,008-’inch orifice and a 0.020-inch orifice. Asbe”fore, the penetration increased with the injection pres- .sure and decreased when the oheck valve was placed in theline, Comparing the figures, it is seen that the penetra-tion of.the gasoline was in some cases greater than thatfor the ftiel oil when the check valve was not in the line.(See ‘reference 6,) This was probably caused-by the dif-ference in the f’1’owconditions through the orifice withthe’two fuels. with the 0.020-inch orifice the ReynoldslTu.mberof the f-low conditions through the orifice was

,

.

.—

—..— .

—

..

—

.

.

—

greatei than 2,000 for ,both fuels. “T?ith the 0.008-inchorif-i’ceand fuel oil, the Reynolds Num%er’was greater than ~ ‘-2,000 Orily for the-pressures greater than 100 pounds per -square inch, but with the gasoline as the fuel the Reyn-olds Number was greater than 2,000 for all the injectionpressures tested. Con~equently, with the gasoline the t*

flow through the orifice was well. within the “turbulent.—

ran~e for all the test conditions; mhiletith fuel oil theflow was turbulent with the O’.b2O-inch orifice at all thepressures tested, but with the 0.003-inch orifice theflow changed from semiturbulent to ttirbulent. As the flowconditions with the fuel oil and the 0,008-inch orificediffered,

-Ff ‘Ue’ -.~a “direct comparison from the s

densities can not be made between Fi@res 7 a

.

N. A. O,’A.

When the chamberpheric to 0.325 pound

Technical Note No,..3.99 5

air density was increased from atmos -per cubic foot (fig. 9), the lighter

fuel pe~etrated the greater distance, although the penetra-tion with ”eithei;fuel was less than that oltained at atmos-ph.etic air” density. The greateb penetration was yrobablycaused, as before, by the change in the flow conditions.

Coticlusions . ‘, ; .’

The following conclusions are drawn from the testresults presented:

.“.:.1. Injection pressures %elow 500 pounds per square

inch give penetration rates comparable viith the rates ob-tained with higher injection pressures.

2. “Atomization is.very.poor with 100. pounds per .square inch injection pbeeeufea, but appears to be fairwith 400 and 500 pounds per square inch injection pressures.

3. Atomization and penetratio-n rates change littlewith orifice size.

4. The penetration rate decreases as the density ofthe air into which the ftiel is injected.is increased.

,..

5. Gasoline and fuel oil give about the same penetra-tion rates and external spray appearance. .

.Langley .Memorial Aeronautical Laboratory,

National Advisory Committee for Aeronautics,Langley Field, Vs., November 4, 1931.

.

.

4

6

.-

1.

2,

3.

4,

5.

6.

,.,

7.

X.A. C..Ai Technical .Not”e So,. 399;

..” References U

.,..

Beardsley, Edward: G, : The I?.A. C.A. Photographic Ap- .2paratus for Studying Fuel Sprays from Oil EngineInjection Valves and Test Results from Several

—.—

Researches , T.R. ~Oo 274, N.A.C.A. , 1927.

Lee, Ilana W.: The Xffect of Nozzle Design and Operat--.

ing Conditions on the Atomization and Distributionof Fuel Sprays for High-Speed Compression-IgnitionEngines. Report to be published.

Beardsley, E. G.: Some Factors AffectinR the Reprodu,c- ..—

i3ility of Penetration and the Cut-Off of Oil Spraysfor Fuel Injection Engines. !2,R. No.258, N.A,C.A, ,1927.

Rothrock, A. U., and Lee, D, W,: Some Characteristics .,

of Fuel Sprays:from Open I?ozzle$. T.N. ~Os 356,E;A;C.A. , 193Q.; . “

.. ~.—-

“Rothroclz, A, M.: Pressure Fluctuations in a Common-Rail Fuel Injection System.

.T.R. NOo 363, N,A.C,A. ,

1930.. .

JQachim,W. F., and Beardsley, E, G,: The Effects of.Fuel and Cylinder Gas Densities on the Character-

.-—

istics, of Fuel Sprays for Oil Engines. T,R. ~~Oo

281, N.A.C,A. , 1927.

Gelalles, A, G.: Effect of Orifice Length-DiameterRatio on Penetration and Cone-Angle. Report to bepublished. ,,

.,

.

.

* r.

.1 *

Initial+mtmlro

~tTOl &17e

ErYuzhcmer-.

z Iiigbpreaetm

freservoir

Flg.1 Fuel upray iqeotiom wataa

—

, .

.

Q.;:./~//;.....;-,/,/..:<...;,:.,:ormn nozzles

‘(+ ‘

$0 0.008 inch, + = .+.;./..’/,./ ...,..1/

.<.,..,.,,.,,4 a

‘/’,,/,~,-’. ‘,

/..;,; :.,”.

.’/’,

‘;.’7”‘;, 643

~<;..;

“’//,’ .’,,

h,

cdaU3

Ii’ig.2 OpBn-nozz10holder with ball-chock valve

1,

. , . r

.013 .012 .066 .& .& .~WsIe, SeCon6

100 lb.~sq.in.injeotionpressure. Atmosphere chember alr density.EO &e& ~ve.~

s

.0b9 .036 ● ti7 .006 “ .065 .CKi4 .063 .C02 .Obl oTime , Seco?ld.

500 lb. /sq. tn. Injection premme. 0.3251b./cu.ft. ohember air density. w

Ho Cheokwlm. ~

Fig. Z (%mdlne sprays obtainedfrom the O.OX)-Inch orific~ m

.

.

●

✎

✎

✎

N.A.C.A. %ohnioal Note No. 399 Fig. 4

. ..100 lh./sq. in. injection pressue 300 111./sq. in. injection pressu,re

NO oheok valve. Atmosphere chamber air density.

Fig. 4 Enlargement of gasoline sPaws from O.0~%.nch orifice

,

.

N.A.C.A. Technical Note No. 399 Fig. 5

5

m

1

0

H /B—..

/ /

/r

E

/

/

// I’hj.pre$s.V< .ve op

s )./sq.:.3.

/ A, lWB, 300

} / c.

// jy

f

/’/ ‘

D, 100I

E, 200 I

F. 300

H, 500 () /

4 / ‘/

/ ‘

1 .001 . m2 .003 .004 .005

Fig. 5 Effect of injectionAtmospheric chamber

Time, secondF ..

pressure on penetration.O.02&inch orifice.pressure. Fuel oily

. . . . .

5

r!flI ,~/Bl

4

0

I // I

1?/

// , 4/ //

I

//”‘E / ‘

/‘/

,<

f , ,

I I //

I // / ./

tI

//

I

1 / In.ectioh pres mue, Valve own. 2!rew r e

/’lb. jt q.in. Ib.jfl(in.

A. 1()0 16q

/ ‘ B, 3()0 160

c, 5 16()

D, o/

E, 3()0 o

F, 5()0 o

.001 .002 .003 .004 .005 .006

!Mme, secondfi

Fig. 6 Effect of injection pressure on pmmtration,O.020-in& orifice. ~huoEphOric

Lhamber pre6sure. GasolineX

‘z

:*

c1:+

-,.. . . . , .

.

51I I t / I

I I I I

z r 1 I. I I //l / -IT-I--” L-’ I I I

8—J –., . 1 1:e, I’ialve Iopen. kreiiik+

Ari. f q.zn./ “

A,

B,14

c.1414! i

D,

~

30E,

I o5 0

0 ,001 .002 ,003 .004 .005‘l!im~,socond~

.006

Fig, 7 “Effect of i~-ection pressue on penetration. O.M@inch orifice.

Atmospheric chamber pres~ue. Fuel oilY

IJ.A.C.A..

Technical Note No. 399

A,B,

c,a,E,l?,

Iajection pres~ue,111./sq.in.

100”

.s00

100

500

Fig. 8

dValve open. resmxelb./sq.in.

.

160160160

000

5

//

/ /

4 /

1’// ~ /

/ /A

i /’/

E /IQ07 , }~ ) / ——

/ (u / /;.s /s

/ / /I$ /

(I)2 /g /, ~Q1 /

/

// ;(/

1/

‘) “J_ —. —.. — .

/

t

0 .001 .002 .003 . @04 .005!l%ne,second[

,,

Fig..8 Effect of injection pressure on penetration.0.00~inchorifice. Atmospheric chamber pressure. Gasoline

Y

N.A.C.A. Technical Note No. 399

0

Fig.

A,B,

Injection pressure, 500 lb./sq.in.

160 1%./sq.in. valve openingo 11 II II 11

—— — GasolineFuel oil

pressure!1

.001

9 Effect of0.020~hch

.002 .0Q3 .004Time, second~

injection pressure on penetration.orifice. 0.325 lb./cu.ft. air density

Fig. 9