RESEARCH MEMORANDUM

EVALUATION OF THREE INJECTORS IN A 2400-POUND-THRUST

ROCKET ENGINEUSING LIQUID OXYGEN AND
LIQUID AMMONIA

B y Robert C. Hendricks, Robert C. Elders, andJack C. Humphrey

Lewis Flight Propulsion Laboratory
Cleveland, Ohio

NACA RM E58B25

NATIONAL

ADvISm

C O M M I T r n FOR

moNAm1cs

USING LIQUID OXYGEN AND LIQUID AEpIONIA*

By Robert C. Hendricks,Robert

C. Ehlers, andJack

C. Humphrey

SUMMARY
The performance of t h r e e i n j e c t o r types was evaluated in a 2400pound-thrustrocket test chaniber. Each injector represents one of eighteen such u n i t s forning the i n j e c t o r f o r a 50,000-pound-thrust rocket
engine. %e i n j e c t o r s were designed t o compare the r e l a t i v e e f f e c t s of
f u e l and oxidant atomization.
*

Characteristic velocity and s p e c i f i c impulse w e r e obtained over a

range of oxidant-to-fuel ratios at a nominal chaniber pressure of 600
pounds per square
inch
absolute.
lhjectors
atomizing
the fuel (RMI-1
and RMI-2) gavecomparable results. No s t a b i l i z e d data w e r e obtained
w i t h the i n j e c t o r atomizing t h e oxidant o n l y (RMI-3) because of combustion instability.
me unsteady-state data i n d i c a t e d t h a t the performance o f this RMI-3 fnjector was lower than RMI-1 and RMI-2. For the
thrust-chaniber configuration used, conibustion i n s t a b i l i t y was not encountered w i t h the RMI-2 injector, but incipient
conibustion i n s t a b i l i t y
was encountered w i t h the RMI-1 injector.

At the request of the Air Force, t h e NACA Lewis laboratory invest i g a t e d the performance of three types of injector spuds, or elements,
designed by Reaction l&tors, Inc.
f o r use w i t h t h e oxygen-aumonia propellant c o d i n a t i o n . These spuds, i n groups of 18, are designed f o r a p p l i c a t i o n i n a backup version of the injector f o r the m99-RPII-1 engine.
The three injectors investigated w e r e d e s i g n e d to represent the r e l a t l v e
contribution of f u e l and oxidant atomization t o engineperformance.
The
evaluation was based on the performance of iqjectors featuring fuel atomization,oxidantatomization,
and both fuel and oxidant atomization.
Characteristic velocity and s p e c i f i c impulse were determined f o r each of
the i n j e c t o r types f'rom tests of a s i n g l e spud mounted in a 24W-poundthrust rocket c h d e r .
r

NACA RM E58B25

Apparatus

The three injectors evaluated axe showg i n f i g u r e 1.

Injectors.
Injector R M I - 1 consisted of 22 paZrseach of like-on-like fuel
and oxidantholes with surface impingement at 900. Injector RMI-2 consisted
of 22 p a i r s of like-on-like fuel holes with &face
impingement a t 900
and 22 showerhead oxidant holes. Injector
RMI-3 consisted of 22 p a i r s
with surface .impingement
a t 900 and 22
of like-on-like oxidant holes
.
."
.
.
showerhead fuel holes.
The i n j e c t o r s w e r e mad& of d c k e l " A . " The oxidant injection holes
were f e d from a reservoir on the upstream side of t h e Injection face.
The fuel injection holes were fed by p&S6ageS, croS8-driU.ed through t h e
i n j e c t o r , from a f u e l manifold around the circumference of the injector.
The i n j e c t o r was designed t o d e l i v e r 6.1 pound-mass per second oxidant
and 4.9 pound-mass per second f i e 1 a t an i n j e c t o r pressure drop of 189
pounds per squareinch.Fuelcoolantholes
were provided i n t h e b a s i c
design; however, they were eliminated during the course of the experiments t o prevent face burning. -

Injector holders: - The injector holders ( f i g s . 2 and 3) were made

from s t a i n l e s s s t e e l and w e r e designed t o provide a fuel manifold, an
oxidantreservoir, and f l a n g e s f o r mounting t o t h e t h r u s t s t a n d .
The
original design provided f o r t h e removal of the injector from t h e holder;
it u t i l i z e d a small bottoming l i p for positioning, and w
t o O-rings f o r
sealing t h e f i e 1 from t h e chamberand t h e atmosphere. The indector was
brazed t o a tube which was i n turn screwed into position in the holder.
Howe-ker,. t h i s holder- ( A ) burned on the f l a t face surface (fig. 2 ( a ) ) .
In an attempt to eliminate this burning, a modification was made as shown
i n figure 2(b); but t h i s modifiedholder (B) also burned. A f a r t h e r modif i c a t i o n t o t h e holder w a s made after a severe fuel leakage i n t o t h e
chaniber caused immediate burnout; t h i s modification (holder C ) i s shown
in figure 2(c).
Although holder C d i d notleak, t h e edge o f t h e insert
burned. The next step w a s t o eliminate the f u e l coolanthole8 izound
the periphery and mount t h e i n j e c t o r f l u s h i n the chamber. Two o f these
holders (type D) were used for W-1 and RMI-3 injectors, and a redesign
was used w i t h M - 2 . This modification(holder D) i s shown i n figure 3.
Description of chamber and nozzle.

1
The thrust chamber was a 5-2

inch-outside-diameter mild steel pipe, bored to 8 4.072-inch inside diameter. -The chambers us& with holder mdifications A, B, and C w e r e
7-9 inches i n length. The chamber usedwithholder
D was 8 inches i n
16
length. The chamber-pr'essure taps w e r e Installed near the- injector and

NACA RM ES8B25
9

thenozzle ( f Q . 3). !@e nozzles were-made of copper,chromium-plated

(0.003 t o 0.005 in. thick), and of mild steel, ceramic-coated (0.010 t o
0.020 in. t h i c k ) . The nozzle had a throat
area
of 3.259 square
inches
and a 4-to-1 contraction ratio.
The t h r o a t extended 0.125 inch downstream;
the divergent section w a s e1imrinFLt;ed. Eight b o l t s were used t o compress
the metal O-ring s e a l s and hold t h e assembly together (fig. 3).
Instrumentation
The engine was munted on a flexure-plate thrust staad equipped wfth
astrain-gageforce-measuring
load c e l l . Chamber pressure was measured.
b o t h w i t h strain-gage transducers and w i t h a recording Burdon-tube instrument. To obtain stable recording conditions during t h e s h o r t durat i o n runs, s h o r t water-cooled pressure leads t o t h e pickup were used t o
decrease the t i m e lag. Oxidantand f u e l flow w e r e each measured with w
to
instruments, a turbine-type flowmeter and a Venturi m e t e r equipped w i t h a
differential pressure transducer. mese
signals w e r e recorded by an oscillograph.Accuracies
of the load c e l l , pressure transducers, flowmeters, and recording oscillograph were rated a t 21 percent o r b e t t e r .
Temperatures of t h e l i q u i d oxygen were measured by copper-constantan
thermocouples with cold-junction thermocouples i n a bath of liquid nitrogen- The copper-constatan thermocouples were made f r o m calibrated wire
w i t h a l i n e drop of 3 microvolts o r less. me animonia temperatures were
recorded by iron-constantan thermocouples with cold-junction thermocouples
located i n a bath of melting i c e . Temperatures were recorded on the
aforementioned recording oscillograph and on s t r i p - c h a t r e c o r d i n g p o t e n tiometers. An accelerometer, installea on the engine to determine
whether
or not screaming occurred, responded t o radial accelerations caused by
pressure oscillations within the chaniber. m e signal from t h e accelerometer was received by an oscilloscope and recorded on film.
Calibrations. - The pressure transducers w e r e calibrated before
each s e r i e s of runs w i t h helium gas and standard gages w i t h accuracies
rated a t i1/4 percent. The thrust-measuring l o a d c e l l was also c a l i brated before each run by using a standard load . c e l l w i t h a rated accuracy
of L l / 4 percent. The constants of c a l i b r a t i o n v a r i e d between t h e s e r i e s
1
1
of runs about 2 2 percent
~
for the pressure transducers and 2% percent
f o r the thrust-measuring load c e l l .
lChe thermocouples w e r e c a l i b r a t e d p r i o r to t h e series of runs, and
check points w e r e made intermittently t o assure that t h e c a l i b r a t i o n
constant did not vary.
Ekrors i n measuring performance. - Although t h e instrumentation was
r a t e d a t t l percent o r b e t t e r , t h e r e were s e v e r a l e r r o r s t h a t reduced the

NACA RM E58B25

accuracyofperformance
measurements. D i f f i c u l t i e s that were experienced
i n measuring oxygen temperature increased-the error in determining oxygen
mass flow. The t h r u s t l o a d c e l l was used a t halfcapacity.
Errors w e r e
observed i n combustion-chaniber pressure, probably from tkpmal e f f e c t s on
t h e diaphragm ofthestrain-gagepressuretransducers.
The m
a
x
m
iu
m error
i n measuring s p e c i f i c impulse and c h a r a c t e r i s t i c v e l o c i t y was estimated
t o b e 25 percent.
Operational Procedure
A flow diagram of t h e t e s t f a c i l i t y i s shown i n figure 4 . The
engine was started w i t h gaseous oxygen and gaseous propane f e d a t r e l a t i v e l y low pressures (50 and 30 lb/sq in. gage, respectively) through the
i n j e c t o r . These propellants were ignited by a torch located outside the
nozzle. After i g n i t i o n of the gasesoccurred i n t h e chmiber, the main
propellant valves were opened t o obtain approximately 20 percent of f u l l
flow. After 1.5 seconds of this low propellant flow, thevalves were
opened t o full flow. Full propellant flow was maintained for periods o f
t i m e ranging from 2 t o 3 seconds, with one r u n of 7 seconds recorded f o r
t h e RMI-2 inJector.

The propellant tanks were pressurized to 975 pounds per square inch
gage f o r t h e Fuel and 1000 pounds per squase inch gage for the oxidant.
The propellant flows were regulated by positioning the propellant valves.
RESULTS AND DISCUSSIOH

Thirteen rum were made w i t h RMI-1, seventeen runs were made w i t h

RMI-2, and three runs were made w i t h RMI-3. The performance data obtained are shown in t a b l e I and me p l o t t e d i n figures 5, 6, and 7 .
The experimental performance c h a r a c t e r i s t i c s of i n j e c t o r s RMI-1 and
RMI-2 were comparable within the accuracy of the measurements. Steadystate performance f i g w e s a r e u n m a i l a b l e for the RMI-3 i n j e c t o r because
of combustion o s c i l l a t i o n s that limited the run duration to less than
1 second. The Unsteady-statedataindicated
that the performance of t h i s
injector was lower than R
M
I
l and RMI-E.
In general, the experimental
performance data were scattered as a r e s u l t of errors previously mentioned.
me curves obtained by linear regression (figs. 5 and 6 ) indicated t h e
characteristic-velocity standard error
of estimate to be 2186 f e e t p e r
The specificsecond for R
M
I
1 and 2133 f e e t p e r second f o r RMI-2.
impulse standard error of estimate w88 27 seconds f o r RMI-1 and 2 2 seconds
f o r RMI-2.
The curves indicated t h e characteristic velocity t o be 87 percent of
t h e o r e t i c a l f o r RMI-1 and 89 percent of t h e o r e t i c a l f o r RMI-2 ai an

NACA RM E58B25

"

oxidant-fuel weight r a t i o of 1.25.

The s p e c i f i c impulse was 82 percent
o f t h e o r e t i c a l f o r RMI-1 and 84 percent of theoretical for RMI-2 a t an
oxidant-fuel weight r a t i o of 1.25.

Becauseof
f o r RMI-3.

insufficient data, the regression

line waa not computed

I n the engine configuration u

sed, the e f f e c t of oxygen atomization
on performance appeared t o be small. The importance of m o n i a atomizat i o n was not determined. Similar studieswith ammonia-oxygen (ref'. l),
hydrocarbon-oxygen (ref. 2 ) , and'hydrocarbon-oqygen-fluorine (ref. 3) have
been made a t a 200-pound-thrust l e v e l . These studies have shown engine
Performance to be dependent on the atomization o f the least v o l a t i l e propellant, which i n t h e p r e s e n t i n v e s t i g a t i o n would be the ammonia. The
previous studies also showed that the performance obtained by atomizing
the ammoniawas lower than that obtained by atomizing t h e hydrocarbon.
This indicated that, with t h e ananonla-oxygen combination, a greater degree
of ammonia atomization i s necessary t o achieve comparable performance.

"

O p e r a t i o n a l d i f f i c u l t i e s w e r e experienced i n s t a r t i n g the i n j e c t o r s .
InJector RMI-1ms the most reliable on starts. Appasently the w e l l atomized propellants from the like-on-like holes had l e s e tendency t o
quench the ignitionsource.Frequentburnouts
were experienced. In
general, the trouble
occurred
on the uncooled injector-holder
face.
Essentially, four separate designs of injector holders were u t i l i z e d in a n effort to alleviate the face burning,
as previously described.
Screaming was suspected on several burnouts.
Accelerometer readings
w e r e made on t h e three types of injectors. %e R
M
I
1 i n j e c t o r showed
evidence of incipient screaming on transition from low flow t o high flow.
In a l l rn except one t h e large-amplitude vibrations darqened out after
highflows w e r e established. In the one run where no dampening occurred,
t h e cooledholder w a s burned. The RMI-2 i n j e c t o r gave no evidence of
screaming. An accelerometerrecordof
the FMI-3 i n j e c t o r w i t h holder D
( f i g . 3) indicatedlarge-amplitude pressure waves. Ilhe holder and inj e c t o r w e r e badly burned.

SUMMARY OF RESULTS
Ekperimental investigation of

the three RMI i n j e c t o r s showed:

1. The c h a r a c t e r i s t i c v e l o c i t y and t h e s p e c i f i c impulse of RMI-1 and

RMI-2 w e r e comparable.

NACA RM E58B25
'
I

2. For the colnbustor configuration used,combustion i n s t a b i l i t y

was not encountered w i t h t h e RMI-2 injector, but incipient conibustion
i n s t a b i l i t y wa8 encountered w i t h t h e FNI-1 injector.

..

3. RMI-3 gave no stabilized data because of conibustion i n s t a b i l i t y ;

however, indications were t h a t this injector had lover performance than
t h e other two Injectors.
4. Since adequate data were unavallahle for t h e RMI-3 indector, a
comparison of t h e effect of f u e l atomization on performance wa8 not obtained. Comparison of RMI-1 and XI-2 showed oxidantatomization t o
have a second-order effect on performance.

Lewis Flight Propulsion Laboratory

National Advisory Committee for Aeronautics
Cleveland, Ohio, March 7, 1958
REFERENCES

1. Priem, Richard J., and Clark, Bruce J. : Comparison of InJectors with

a EOO-Pound-Thrust Aumonia-Oxygen w i n e . NACA Rh4 E57H01, 1957.

a d Auble, C. M. : InjectionPrinciples from Combustion Studies i n a 200-Pound-IChrust Rocket Engine U s i n g Liquid Qxygeja
and Heptane. NACA RM F55C22, 1955.

2. Heidmann, M. F.,

A Study ofInjectionProcess
f o r E-Percent Fluorine
85-Percent Oxygen and Heptane i n a 200-Pound-Thru~t Rocket u t n e .
NACA RM E56Jl1, 1957.

3. Heidmann, M. F.:

... . ....
1

- ..

..

..

.. .

. .

.. .

.. ..

..

Ba.1

46657.1

81.8
8L.O
82.8
79.4
77.8
77.2
77.8

5lEl.O

79-8
80.6
B.3

m.7
386.0

31.1

235.8

P1.S

1n.s
m.7
186.4
m.7
B . 6

Sl61.8

=.e

8oBl.l
5118.8

1.0

5237.8
4346.8
am.3

89.5

4188.2

85.1

ei3
81.8

PIA

211.4

P1,2
a
7.s

a1.s

178.1

a7.S
n7.2

176
-

5oL5.9

a
1.s

174.6
181.8

.__

6296.8
m58 .4

a1.s

m.7
a4.4

la.8
116.0
118
lBB.8

a1.s

lEE.1
36a.4

177.8
112.8

. . . . . .. . . . .

.. .... .

a7.e

a 7 3
aB.8

a7.a
m.4
PI .6

..

..

. ..

.. .

. .-

.. . .

. ... .

.. .

L/BB

.. . . .

..

. .

. ..

. .

.. .

..

..

..

I-*

... .

. ... .

. .
. .. ..

.
L

FUEL -

.. .

P-A

.. . .

..
.

"

. ...

..

..

(b) Bolder E.

..
I

(c)

HDldeT c.

/CD-6002/

. .

EIACA RM E58B25

220

180

160

1.0

1.1

1.2
1.3
1.4
O x i m t - f i e l weight ratio, O/F

1.5

1.6

Figure 5. Performance of M-1 injector. Fuel and oxidant

like-on-like impingement.

E
W

cn

. . .

..
1

. .. . . .
I

220

200

160

..-l

6000

lbooo

1
Figure 6.

- Performance of RMI-2 injector.

Oxa
i ant sharerhead.

Fuel like-on-like impingement;

. . .. .

.. .. .. .

. . . . . . . . .

. .

Oxidant-fuel'weight ratio, O/F

- Performance of MI-3 injector. Q

x
d
lUke-on-like
a
u
t
'
impingement; fuel showerhead.

Figure 7.
f