Chapter IV

Design Of Thrust Chall1bers

And Other COll1bustion Devices
While the proud designers of the various sub-
systems of a rocket engine each consider their
product as "the heart of the engine, .. the thrust
chamber assembly undeniably embodies the es-
sence of rocket propulsion: the acceleration and
ejection of matter, the reaction of which imparts
the propulsive force to the vehicle. The design-
er's goal is essentially to accomplish this with
a device of maximum performance, stability and
durability, and of minimum size, weight, and
cost.
The design of thrust chambers is one of the
more complex subjects in the field of liquid pro-
pellant rocket engineering. This is primarily due
to the fact that the basic processes, especially
the combustion within the thrust chamber, are
comparatively difficult to define and to study
analytically. Thus, during most engine develop-
ment programs, a major effort must be expended
toward the design and development of the thrust
chamber. A rational approach to this effort is
attempted here.
4.1 THE BASIC THRUST CHAMBER
ELEMENTS
The thermodynamic processes governing the
generation of thrust within a thrust chamber have
been treated in chapter 1. The primary function
of the thrust chamber is to convert the energy of
propellants into thrust. In a liquid bipropellant
rocket engine, this process is characterized by
the following basic functional steps:
1. The liquid propellants, at their proper mix-
ture ratio, are injected into the combustion cham-
ber through orifices in an injector, as jets at
velocities ranging from 20 to 150 feet per second.
These jets either impinge to form a mixed drop-
let spray, or run straight into the chamber hot
gas as a series of droplets. Part of the combus-
tion reaction may already take place in the liquid
state.
2. The droplets are subsequently vaporized
by heat transfer from the surrounding gas. The
size and velocity of the droplets change contin-
uously during their entrainment in the combustion
gas flow.
3. The vaporized propellants are mixed rap-
idly, further heated and promptly reacted at their
stoichiometric mixture ratio where ever they are
formed, thus effecting a continuous increase of
the gaseous mass flow rate within the combus-
tion chamber. This gas reaction is further aided
by the high-speed diffusion of active molecules
or atoms. The combustion is essentially com-
plete upstream of the chamber throat, when all
liquid droplets have been vaporized. Under cer-
tain conditions, shock and detonation waves may
be generated by local disturbances in the com-
bustion front, possibly caused by instability of
mixing process and propellant now prior to reac-
tion. These effects may trigger sustained pres-
sure oscillations at certain frequencies within
the thrust chamber, resulting in destructive com-
bustion instability. A major portion of the de-
sign and development effort, therefore, is directed
toward achievement of stable combustion.
4. As the gaseous products of the combustion
process pass toward and through the throat, they
are accelerated to sonic, and then to supersonic,
velocities within the diverging nozzle section,
and are finally ejected to the rear.
The basic elements of a thrust chamber re-
quired for its function, include a combustion
chamber section, an expansion nozzle section,
an injector, an ignition device (for nonhypergolic
propellant combinations), propellant inlets and
distributing manifolds, and interconnecting sur-
faces for,component and thrust mounts. The con-
struction of the various thrust chamber elements
depends largely on their specific operational
function. However, low weight and Simplicity,
which make manufacturing easier, are two impor-
tant factors to be considered at all times.
81
-
. .
82 DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
Figures 4- 1 and 4-2 a typical liquid
bipropellant rocket en ine thrust chamber assem-
bly. The illustrat&d thrust c amber assembly i s
composed of four major subassemblies or basic
el ements; namel y , t hrust chamber body, injector,
liquid oxygen dome, and i gniter.
The thrust chamber body subassembly j s of a
venturi shape, consistm[,: of a cyli ndrical section
in which th combustion occurs; a section nar-
rowing toward a throat; and a bell-shaped, ex-
panding nozzl e section throug/ which the com-
bustion gases are expelled (fig. 4-1) . The body
wall of this chamber is constructed of nickel
tubes running longitudinally, j oined by ver
brazing, and retained by external tension bands.
The tubes, of 0.012-inch waH thickness, are of
rectangular cross section of varying area, to
conform to t he thrust chamber shape. This con-
struction permits simple thrust chamber cooling
during operation, by flow:i ng fuel through the
tubes which form the chamber wall . The
under pressure, enters t e th;ust chamber bod,
at the fuel manifold inlet and is di stributed to
alternate thrust-chaml:> r t ubes. It then 110ws
down toward the thrust chamber nozzle exit
where the fuel return manifold reverses the flow
Fi ure 4-1. -7'hru t chambl?r a sembly.
i nto the return tubes. The fuel then flows
through an inje tor fuel creen i nto the radial
injector passages, and finally through th fuel
inj ector orifices ' nto the thrust chamber combus-
tion zone. The fuel manifolds of t hi s chamber
are made of 4130 steel or 347 stainless steel.
Other structural members, such as tension bands,
stiffening rin s, and outrig ers wer all made of
4130 steel. The oxid' zer (liquid oxygen) enter
the LOX dome under pressure t hrough a screened
central port and is di stributed within t he dome
directly to the liquid oxygen passages and
orifi ces (fig. 4-2).
The following are the operating characteri -
tics and principal dimensions of a hypothetical
t hrust chamber similar to the one shown in fi gure
4- 1:
1. . ....... . .......... . .... LOXf RP- l
2. Of F mi ;uure ratio ................... 2.30
3. Characteri ti c velocit , c*, ftlsec . ... 5400
4. Thmst coefficient, C( (ea . .. .. 1.489
5. Specific impulse (lshc (sea level ),
sec .......... . ............ . ...... 249
6. Total propell ant now rate. Ibl ec ..... 402
7. Thmst ( ea l evel), lb ... .. . .. ..... .. 100000
8. Chamber pre s ure (i njector end), p ia . 520
9. Chamber pre ure (nozzle s tagnation),
psia ..... .. .. . ... .. . . .......... . . 480
___ ,un MANIFOLD I LET
"<DE
.......... 'ROPEllANT NJClO PLATE
u UIO C
lOX
fUEl MANif OlD
Figure 4-2. --Thrust ch2mber injector .
... - ~
....... t .... ~ . ' ~ .
DESIGN OF THRUST CHAMBERS AND OTHER COMBUSTION DEVICES
83
10. Average gas specific heat ratio (y) . ... 1.233
11. Combustion chamber crosssection
area. in
2
...................... 244
(at injector)
12. Throat area. in
2
.................... 140
13. Nozzle exit area. in2. ............... 1120
14. Combustion chamber volume. in
3
...... 5320
(above the throat)
15. Combustion chamber length. in ....... 28.5
(injector to throat)
16. Characteristic chamber length. L*. in .38
17. Overall thrust chamber length. in ..... 73
18. Design contraction area ratio. fC ..... 1.60: 1
19. Design expansion area ratio. f ........ 8: 1
The thrust chamber injector (fig. 4-2) is a
round plate. honeycombed with circular and radial
inner passages. leading to drilled orifices. It is
constructed of 4130 steel with nickel-plated sur-
faces. and held in position at the fuel manifold
below the liquid oxygen dome with high-strength
bolts. The seals between injector and thrust
chamber body are of the O-ring type, made of
rub0er selected for compatibility with the fuel
(RP-l). A threaded hole is provided in the cen-
ter of the injector face to permit pyrotechnic
thrust chamber igniter installation. The injector
has 20 circular concentric copper rings which
contain the injection orifices and are fed from
the main propellant systems. Fuel and oxidizer
are kept separate by an elaborate distribution
system. feeding alternate rings. Fuel flows
through the outermost ring. through each alter-
nate inner ring. and through a central fuel disk
which is separately fed from an igniter fuel val ve
through an ignition fuel inlet port. Liquid oxy-
gen emerges from the remaining rings. The in-
jection orifices are so angled that the propel-
lants impinge in the thrust-chamber combustion
zone in a like-on-like pattern (liquid oxygen on
liquid oxygen and fuel on fuel). The primary
orifices are arranged in pairs. with a 0.416-inch
distance between centerlines and a 40 included
impingement angle. for both propellants. In other
designs. impingement angles or orifice separa-
tions are made different for the two propellants.
so as to effect their impingement in different
planes (multiplanar impingement as opposed to
uni planar) .
The liquid oxygen dome is a single-piece.
2014-T6 aluminum-alloy die forging. It provides
the inlet for the liquid oxygen. It also serves as
the thrust-chamber-t'o-vehicle attachment inter-
face. The flanges of the liquid oxygen dome and
the injector are sealed by a spiral-wound gasket
made of 304 stainless-steel strips with asbestos
fillers. This type of gasket is designed specifi-
cally for cryogenic and elevated-temperature
applications.
The electrically fired pyrotechnic igniter is
secured centrally to the injector surface by
means of a threaded joint. It is designed for one
start only and must be replaced after each firing.
It receives an electrical firing signal from wires
connected through the nozzle exit.
4.2 THRUST CHAMBER PERFORMANCE
PARAMETERS
The significance of the parameters which
express or influence the efficiency of thrust
chamber operation has been discussed in section
l.3 of chapter 1. Before discussing the details
of actual thrust chamber design. the following
summarizes these parameters and illustrates
their use for design calculations by applying
them to the engine systems of the Alpha vehicle
which was discussed in chapter III.
Specific Impulse. Is (sec)
From equations 1-31 and 1-31c:
The speCific impulse figure indicates the
overall quality of the thrust-chamber design. As
was learned earlier. it shows how much thrust is
generated for what "running propellant expendi-
tures. ~
Characteristic Velocity, c* (rt/sec)
From equation 1-32a:
(4-1)
Assuming that the propellant and mixture ratio
selection has been made. it can be expected that
the gas properties (y. R) will fall into a known
band. From there on. c* almost entirely depends
on the temperature of the gases. Obviously. this
temperature has a theoretical maximum for a
selected propellant combination. How close
84 DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
to this maximum the chamber will operate de-
pends on influences discussed for mixture ratio
in chapter II. Figures 4-3, 4-4, 4-5, and 4-6
illustrate this. It is seen that c* peaks at com-
bustion temperatures somewhat lower than maxi-
mum. Other conSiderations, such as bulk densi-
ties, which affect vehicle tank sizes, may cause
further adjustments of the mixture ratio for opti-
mum overall vehicle performance. Within these
boundaries, the quality of the combustion proc-
ess greatly depends on the design efficiency of
a thrust chamber assembly, in particular the
injector.
Thrust Coefficient, C f (dimensionless)
From equation 1-33a:
(4-2)
Let us assume that the performance of energy
generation through the combustion process, the
effects of which were just summarized for c*,
has been determined. Then, with a given gas
property (y), the performance of the remaining
thrust generating functions of the thrust chamber,
essentially those of the divergent nozzle. will
depend on the nozzle geometry (mainly ( which
determines the pressure ratio (Pe/(Pc)ns) and the
ambient pressure (Pa).
Performance Calculation
In actual design practice, the calculation of
the thrust chamber performance is based on theo-
retical propellant combustion data and the appli-
cation of certain correction factors as explained
in chapter 1. The theoretical propellant combus-
tion data are derived from thermochemical compu-
tations which equate the heat of reaction of the
propellant combination and the rise in enthalpy
of the combustion gases. Typical prol'ellant
combustion data at frozen composition are pre-
sented in figures 4-3 through 4-6. For given
propellant combinations and chamber-nozzle
stagnation pressures (Pc)n s, the values for the
combustion gas temperature (T c)ns, molecular
weight )IT, and specific heat ratio yare plotted
against the O/F mixture ratio rw. Performance
correction factors are determined by the theoreti-
cal assumptions and from earlier test data, as
well as from the chosen design configurations.
Typical performance calculation methods have
i
1
5
I

f-i I
I
I I
;

:
I
I
: I "-l
'--
,
-I'Y
I
I
."
I i
1/ I/i\ ..
V
I I '--tre)o,!
I
,
1 I I
I
: I
I
I
I
idt -)
\.
I
o I

t'.... i
; , i
I
" !

:
y/i ,-f-
r. j !
I
I
>=.. I 20
: :
I
I
J
I
I
I
I
!
I
5900
5800 u
i.'I
5 700'
t:
5600
g

u


5300
200 i 5
5
7

100 u
50 ..
35 ...
Figure 4-3.-Theoretical O
2
IRP-l combustion data
(frozen composition), (pcJns=lOOO psia.
'i
w

;z;
"'::>-'



:0 00
I'i

16
I.
12
10
t:


:

noo
u

, 0
.'
:> u

raw.. 5000

-Zw
-.-"'Q 5 4ax>f--+---+----iI'-

I- !(XX) f--t---+-i---+ ---r---f--+--+----'--., 30
8

I 28
0
;::
I 26



5 12_

r
u
;;:
1.22
iJ
"3
::>
10 II>

L2"O
,:.
..
..
Figure 4-4.-Theoretical 02/H2 combustion data
(frozen composition). (Pc)ns = 800 pSia.
DESIGN OF THRUST CHAMBERS AND OTHER COMBUSTION DEVICES
85
3
I
/
t'(
I .......
9
I V
7
,1/
V
I


V
V
/
2000
o I3S
i
[,reo
k+-
k'
"lm
'\
"!'....
t'-..

V

db
l...-
-
-
I--
8300
u
8200 i:j
8100 t
,:

...
,.
7900 u

7800 ii

u
7700 ;
7600 0.
".
7:100
_\ /Y
'<
l.t.-
1---
I--
I--
+--3
o
5 133
'" u 131
g 129
>' 127
-
I
.>
......
--

-
I
2.S 3.S 4.S 65 15 85 95
MIXTURE IItATIO. ' W.' Wf
20
I o
..
..
Figure 4-5. -Theoretical F 21H 2 combustion data
(frozen composition), (Pc)ns = 100 psia.
2

C>
f...-
r-
8
V
I
L..,...--
20
41--c:--.
.....
D--
VI
0
iY t-....
/
V

480
/\
i\.
/
)L(Tc:1ns
1\
/
V
J
6

1\

I 2
V
Irdb
5
'\
/ -
!--
p..;
I--
I--
4
........
...........
.......
r--
t--
3
t
,
80
7S
i
i
70

"
8S
..
..
0
i 12

...
l:
i
12
>'
12
0_6 01 08 09 10 1.1 1.2 1.3 14 1.5
MIXTURE RATIO. ' Wol Wf
Figure 4-6.-Theoretical N
2
0
4
1N
2
H
4
combustion
data (frozen composition), (Pc)ns=100 pSia.
been demonstrated earlier by sample calculation
(1-3). The following sample calculations illus-
trate the more specific approaches.
Sample Calculation (4-1)
Determine the design values of c*, Ct, and
(Is)tc for the engine thrust chambers of the
stages of the hypothetical Alpha vehicle, with
the following assumed design parameters:
(a) 750K A-I Stage Engine:
Propellants, L0
2
/RP-1; thrust chamber
O/F mixture ratio, 2.35; (Pc)ns, 1000
psia; propellant combustion data, figure
4-3; nozzle expansion area ratio, (= 14.
(b) 150K A-2 Stage Engine:
Propellants, L0
2
/LH
2
; thrust chamber
O/F mixture ratio, 5.22; (Pc)ns, 800
psia; propellant combustion data, figure
4-4; nozzle expansion area ratiO, (= 40.
Solutions
(a) A-I Stage Engine:
From figure 4-3 for L0
2
/RP-1 at (P
c
)ns=1000
psia and a mixture ratio of 2.35, the following
values are derived for the chamber product gases:
(T c)ns = 6000
0
F or 6460
0
R,
m = 22.5 Ib/mol, y= 1.222
Substitute into equation (1-32a):
Theoretical c* V
32
.
2
x 1.222 x 6460 x 1544/22.5
0.7215
= 5810 ftlsec
This value for c* can also be derived from fig-
ure 4-3.
For a good combustion chamber and injector
design, the c* correction factor for L0
2
/RP-1
and frozen composition will be about 0.975.
DeSign c*=5810x 0.975= 5660 ftlsec
For y=1.222, (=14, a theoretical vacuum C[
value of 1.768 can be derived from figure 1-11:
Th
. al C (Pa
eoretIC { at sea level = (Ct)vac - (P c)ns
=1.768 14x14.7 1.562
1000
86
DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
Sea level C [ can also be calculated using equa-
tion (1-33a), with the aid of equation (1-20),
With effective nozzle contour design, an over-
all C[ correction factor of 0.98 for L0
2
/RP-1
frozen composition can be used.
Design sea level C[= 1.562 x 0.98 = 1.531
From equation (1-31c):
D
11(1) 5660
x
1.531_
270
eSlgn sea eve s tc 32.2 - sec
(b) A-2 Stage Engine:
From figure 4-4 for LOz/LH
z
at (Pchs = 800
psia and a O/F mixture ratio of 5.22, the follow-
ing values are derived for the chamber product
gases:
(T c)n s = 5580 F or 6040 R,
m = 12 lb/mole, y= 1.213
Substitute into equation (1-32a):
Th t
1 * - \132.2 x 1.213 x 6040 x 1544/12
eore lCa c - 0.717
= 7670 ft/sec
Based on experimental data, a c* correction
factor can be assumed for the L0
2
/LH
2
frozen-
composition data of about 0.975.
Design c* = 7670 x 0.975 = 7480 ft/sec
For y=1.213, E=40, a theoretical vacuum C[
value of 1.876 can be derived from figure 1-11.
C f can also be calculated using equations (1-33a)
and (1-20). With effective nozzle contour design,
an overall C f correction factor value of 1. 01 can
be used for LOz!LH
z
frozen-composition data:
Design vacuum C[= 1.876 x 1.01 = 1.895
From equation (l-31c):
. 7480x1.895
Deslgn vacuum Ushc 32.2 440 sec
The reader should perform his own calcula-
tions for the A-3 and the A-4 engines, with the
aid of tables 3-4 and 3-5, and figures 4-5 and 4-6.
4.3 THRUST CHAMBER CONFiGURATION
LAYOUT
After major thrust chamber operating param-
eters such as type of propellants, thrust level.
chamber pressure, C[, c*, and Is have been es-
tablished from engine system requirements and
performance calculations, one of the fundamental
dimensions of the thrust chamber, the throat
area At, can be readily derived (eq. (1-33)). The
throat area At or throat diameter D
t
usually is
the starting point of a thrust chamber configura-
tion layout. The combustion chamber and nozzle
section are commonly designed as an integral
thrust chamber body. For light weight and ease
of manufacture, thrust chambers will have the
general shape of a pressure vessel with wall
surfaces of rotation and smooth contours.
Combustion Chamber Volume
The combustion chamber serves as an enve-
lope to retain the propellants for a sufficient
period (stay time) to assure complete mixing and
combustion before entering the nozzle. The vol-
ume of the combustion chamber thus has a defi-
nite effect on combustion efficiency. The theo-
retical required chamber volume is a function of
the mass flow rate of the propellants and their
average density, and of the stay time needed for
efficient combustion. The relationship can be
expressed by the following equation:
where:
V c = chamber volume, ft 3
IV tc = propellant mass flow rate, lb/ sec
V = average specific volume, ft 3 lIb
ts = propellant stay time, sec
(4-3)
A useful parameter, the characteristic length L*
(commonly pronounced "L-star"), can be used to
specify the propellant stay time in the combus-
tion chamber. L* is defined as the ratio of
chamber volume to nozzle throat area, and can be
expressed by the following equation:
(4-4)
DESIGN OF THRUST CHAMBERS AND OTHER COMBUSTION DEVICES 87
Since the value of At is in nearly direct propor-
tion to the product of W tc and V, L* is essen-
tially a function of ts. The effect of L* on c* in
an experimental combustion chamber is shown in
figure 4-7. The c* value increases with L* to an
asymptotic maximum. Increasing L* beyond a
certain point tends to decrease overall engine
system performance because of the following:
(1) Larger L* results in higher thrust chamber
volume and weight.
(2) Larger L* creates more surface area in
need of cooling.
(3) Larger L* increases friction losses at the
chamber walls.
In actual design practice, optimization analyses
will determine the minimum possible combustion
chamber L* consistent with efficient combustion.
Under a given set of operating conditions,
such as type of propellants, mixture ratio, cham-
ber pressure, injector design, and chamber geom-
etry, the value of the minimum required L* can
only be evaluated by actual firings of experi-
mental thrust chambers. L* values of 15 to 120
inches for corresponding propellant stay-time
values of 0.002-0.040 second have been used in
various thrust chamber designs. Typical L*
values for different propellants are given in
table 4-1. With At and minimum required L*
established, the required combustion chamber
volume V c can be calculated by equation (4-4).
I 50-50 N
Z
H4 - UDMH
OIF MIXTURE RATIO 1. 90
(Pc Ins' 275 PSIA. c' 2.00
L
-
5800
u
w
IJ)
5700


5600
a:-
w*
5500
tJu
-
5400
J:C
ug
5300
...J
.".. >--

/
V
/
I
W
>
!5200
10 20 40 50
CHAMBER CHARACTERISTIC LENGTH ( L*) IN.
Figure 4-7.-Effect of L* on c* value of experi-
mental thrust chamber.
TABLE 4-1.-Recommended Combustion Chamber
Characteristic Length (L*) for Various Propel-
lant Combinations
Propellant combination
Chlorine tnnuoridelhydrazine-base fuel.
Liquid fluorine/hydrazine ........ .
Liquid fluorine/liquid hydrogen (GH
2
injection) ....
Liquid nuorine/liquid hydrogen (LH
2
injection) .
Hydrogen peroxide /RP-l (including
catalyst bed) .....
Nitric acid/hydrazine-base fuel.
Nitrogen tetroxideihydrazine-base fuel.
Liquid oxygen/ammonia.
LiqUId oxygen/liquid hydrogen (GH
2
injection) .
Liquid oxygen/liquid hydrogen (LH 2
inj ection) ..
Liquid .
Combustion Chamber Shape
Combustion
chamber
characteristic
length (L*). in.
30-35
24-28
22-26
25-30
60-70
30-35
30-35
30-40
22-28
30-40
-to-50
As can be seen from equation (4-3), the stay
time ts is independent of the combustion chamber
geometry. Theoretic:111y. for a given required
volume, the chamber can be of any shape. In
actual design, however, the choice of the com-
bustion chamber configuration is limited. In a
long chamber with a small cross section, higher
nonisentropic gas flow pressure losses will re-
sult as explained in chapter 1. This approach
also dictates a longer thrust chamber space en-
velope and imposes certain space limitation on
the injector design to accommodate the neces-
sary number of injector holes. With a short
chamber of large cross section, the propellant
atomization or vaporization zone occupies a
relatively large portion of the chamber volume,
while the mixing and combustion zone becomes
too short for efficient combustion. Other factors,
such as heat transfer, combustion stability,
weight, and ease of manufacturing, are to be
considered in determining the final combustion
chamber configuration.
Three geometrical shapes which have been
used 10 combustion chamber design are shown in
figure 4-8. While the spherical and the near-
spherical chambers were used in earlier European
88
DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
designs, the cylindrical chamber has been used
most frequently in the United States.
The spherical or nearly-spherical chamber, as
compared to the cylindrical one of the same vol-
ume, offers the advantage of less cooling surface
and weight. A sphere has the smallest surface-
to-volume ratio. For equal strength of material
and chamber pressure, the structural walls of the
spherical chamber are about half the thickness of
the walls of a cylindrical chamber. However, the
spherical chamber is more difficult to manufac-
ture and has poorer performance under most cir-
cumstances. For these practical reasons, the
design details of the cylindrical combustion
chamber will be treated in this book. Several
SPHERICAL
COMBUSTION
CHAMBER
THROAT
I
NEAR SPHERICAL
COMBUSTION
CHAMBER
THROAT
I
CYLINDRICAL
COMBUSTION
CHAMBER
NOZZLE
THRUST CHAMBER
AXIS
NOZZLE
THRUST CHAMBER
AXIS
NOZZLE
THRUST CHAMBER
AXIS
Figure 4-8. -Frequently used geometrical shapes
[or combustion chambers.
novel thrust chamber designs will also be dis-
cussed.
In the design layout of the cylindrical com-
bustion chamber of a given At and L*, the value
of the contraction area ratio, (fe = (Ae! At)) can
be optimized through careful studies of the fol-
lowing factors:
(1) Combustion performance in conjunction
.... 'ith the injector design
(2) Chamber gas flow pressure drop
(3) Chamber wall cooling requirements
(4) Combustion stability
(5) Weight
(6) Space envelope
(7) Ease of manufacturing
For pressurized-gas propellant feed, low-
thrust engine systems contraction area ratio
values of 2 to 5 have been used. For most turbo-
pump propellant feed, high thrust and high cham-
ber pressure engine systems lower ratio values
of 1.3 to 2.5 are employed. The reader is also
referred to section 1.2 chapter I, "The Gas-flow
Processes in the Combustion Chamber and the
Nozzle. "
The basic elements of a cylindrical combus-
tion chamber are identified in figure 4-9. In
design practice, it has been arbitrarily defined
that the combustion chamber volume includes the
space between injector face I-I and the nozzle
throat plane II-II. The approximate value of the
combustion chamber volume can be expressed by
the following equation
INJECTOR
FACE
1
I
THROAT
CHAMBER :0:
OIA. Dc I
AREA Ac '
I
CHAMBER CONTRACTION
AREA RATIO
NOZZLE
THROAT
OIA 01
AREA AI
Figure 4-9.-Elements of basic cylindrical com-
bustion chamber.

DESIGN OF THRUST CHAMBERS AND OTHER COMBUSTION DEVICES 89
The total surface area of the combustion
chamber walls excluding injector face can be
approximated by the following expression:
Total area = 2LcY TTfcAt + csc e(c - l)Ae (4-6)
Nozzle Expansion Area Ratio
It was learned earlier that with all other
parameters fixed, in particular chamber pressure,
there is only one optimum nozzle expansion area
ratio for a given altitude or, more specifically,
ambient pressure. Except for those systems
which start in vacuum, ambient pressure will
have to be considered. This is especially true
for boosters which start at or near sea-level
conditions.
It is the ultimate purpose of a rocket engine
to lift vehicles to altitudes. Inherently, then,
ambient pressure will not be a constant (except
for high-altitude starts, as mentioned). It is,
therefore, extremely important for the designer to
know the trajectory of the vehicle to be propelled
or, more specifically, its altitude-versus-time
characteristics. With this information, the de-
signer is in a position to make a first, optimizing
selection of a nozzle expansion area ratio, for
best results throughout the entire trajectory. As
shown earlier, area ratio will be truly optimum
for only one speCific altitude. The optimization
for ambient pressure then is essentially an
averaging process.
Other considerations usually cause the de-
signer to deviate from the "paper optimum" for
the nozzle expansion area ratio. Some of the
most common are: weight, size, ease of manu-
facturing, handling, and cooling (heat transfer)
considerations.
Nozzle Shape
Most rocket nozzles are of the converging-
diverging De Laval type. Since the flow velocity
of the gases in the converging section of rocket
nozzle is relatively low, any smooth and well-
rounded convergent nozzle section will have
very low energy losses. By contrast, the contour
of the diverging nozzle section is very important
to performance, because of the very high flow
velocities involved.
The selection of an optimum nozzle shape for
a given expansion area ratio is generally influ-
enced by the following design considerations
and goals:
(1) Uniform parallel axial gas flow at the
nozzle exit for maximum momentum
vector
(2) Minimum separation and turbulence losses
within the nozzle
(3) Shortest possible nozzle length for mini-
mum space envelope, weight, wall fric-
tion losses, and cooling requirements
(4) Ease of manufacturing
In actual design practice, any abrupt change or
discontinuity in the nozzle wall contour should
be avoided to eliminate the possibility of shock
waves or turbulence losses. Theoretically, the
nozzle throat is simply the unique plane of mini-
mum cross-section area. In practice, a well-
rounded throat section is employed. Only at the
nozzle exit plane is a sharp edge used because
a rounded one would permit overexpansion and
flow separation.
1. Conical Nozzle
In early rocket engine applications, the coni-
cal nozzle, which had proved satisfactory in
most respects, was used almost exclusively.
The advantages of a conical nozzle are ease of
manufacturing and flexibility of converting an
existing design to higher or lower expansion area
ratios without major redesign of the nozzle con-
tour.
The configuration of a typical conical nozzle
is shown in figure 4-10. The nozzle throat sec-
tion has the contour of a circular arc with a
radius R ranging from 0.5 to 1.5 times the throat
radius Re. The half angle of the nozzle conver-
gent cone section can range from 20 to 45. The

I
4<---=-_-=-_--c. ----Tl-...-
i
THROAT
Figure 4-10.-Conical nozzle contour.
90 DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
divergent cone half angle a varies from approxi-
mately 12 to 18. The length of the conical
nozzle section can be expressed by the equation
R
t
(Vc-1) + R(sec a -1)
Ln
(4-7)
tan a
The conical nozzle with a 15 divergent half
angle has become almost a standard. as it is a
good compromise on the basis of weight. length,
and performance.
Since in a conical nozzle certain performance
losses occur as a result of the non axial compo-
nent of the exhaust gas velocity, a correction
factor A is applied for the calculation of the exit
gas momentum. This factor or thrust efficiency
is the ratio between the exit gas momentum of
the conical nozzle and that of an ideal nozzle
with uniformly parallel axial gas flow. The
value of A can be expressed by the following
equation:
A=i(ltcosa) (4-8)
where a = half angle of the conical nozzle.
For an ideal nozzle, A would be unity. For a
conical nozzle with a = 15 and A = 0.983, the exit
gas momentum or the exit velocity will be 98.3
percent of the ideal nozzle exit velocity calcu-
lated by equation (1-18). The value of the vac-
uum thrust coefficient of a nozzle is in direct
proportion to the thrust generated by the nozzle,
or to the nozzle exit gas velocity. Therefore,
the theoretical vacuum thrust coefficient (neglect-
ing friction and other flow losses) of a conical
nozzle with 15 half angle will be 98.3 percent
of the ideal nozzle thrust coefficient calculated
by equation (l-33a).
2. Bell Nozzle
For increased performance and shorter length.
bell-shaped nozzles have been developed. This
nozzle design employs a fast expansion or radial
flow section in the initial divergent region,
which then leads over to a uniform, axially
directed flow at the nozzle exit. The wall con-
tour is changed gradually enough so that oblique
shocks will not form.
Figure 4-11 shows the contour of a bell noz-
zle. A circular arc of selected radius R
J
is
[XIT
1-1.----- L, ______ P...,LANE
I
i THROAT
I
I
END '
""'NT 1
,
.-T ~ .. jfR,
fir-RIGHT I
R ' ~ CHARACT[NSTIC
I' 1 \ kERIIIEL (TN_O) p. Ii' INK) ,
L ; .' NOZZLE
-10-------------- - - - - ~ - ~
Figure 4-11. -Bell nozzle contour.
chosen for the nozzle contour MT upstream of the
throat. Contour TNE is the diverging portion of
the nozzle. The initial expansion occurs along
contour TN; contour NE turns the flow over to a
direction nearer to axial. For design conven-
ience, the contour TN is also a circular arc. with
a smaller radius R
2

For those familiar with compressible flow
theories. it is noted that, using transonic flow
analyses, a constant-Mach-number line TO can
be defined at the throat. Given the flow condi-
tion along TO and the solid boundary TN, a
kernel flow field TNKO can be generated by the
method of characteristics developed in gas dy-
namics. The kernel of the rocket nozzle contour
is defined as that portion of the supersonic flow
field determined entirely by throat conditions.
The last right characteristic line NK of kernel
TNKO, and thus the loeation of the point N along
contour TN, is to be determined by specific
design criteria.
The location of the end point E along contour
NE is defined by the given nozzle expansion
area ratio and nozzle length (distance between
throat and exit plane). Then the right character-
istic line NK can be determined by satisfying the
following conditions concurrently:
(1) A control surface PE can be generated
between the point E and a selected point
P along the line NK
(2) Mass now across PE equals the mass
now across NP
(3) Maximum thrust by the nozzle is attained.
By selecting points p', p", etc . along line
NK, a series of control surfaces P'E'. P" E",
etc .. can be generated to define points E I, E",
etc .. along the contour NE. Calculations for the
nozzle contour can be effectively performed by a
computer.
--
DESIGN OF THRUST CHAMBERS AND OTHER COMBUSTION DEVICES 91
.:r
...J
<I
10
9
9
0
9
8


'1
7 .':3<)
Ii
.,zo
.-10
/
/
6
/
9

/
60
(_BELL NOZZJS-
I I

P"
'J.lOINT

..,

V _/

V
L
V
r/
V

NOZZLE
/'
V
V
70 80 90
FRACTIONAL NOZZLE LENGTH BASED ON A 15 HALF ANGLE
CONICAL NOZZLE WITH ANY AREA RATIO.
Figure 4-12. - Thrust efficiency versus bell nozzle
length. (Shown [or comparison: effect o[ short-
ening conical nozzle, increasing half angle.)
Commonly, an equivalent 15 half-angle coni-
cal nozzle is used as a standard to specify bell
nozzles. For instance, the length of an 80-
percent bell nozzle (distance between throat and
exit plane) is 80 percent, or 0.8 of that of a 15
half-angle conical nozzle having the same throat
area, radius below the throat, and area expansion
ratio.
Figure 4-12 shows the thrust efficiency ,\
versus fractional nozzle length Lf for conical
and bell nozzles.
As may be seen, bell nozzle lengths beyond
approximately 80 percent do not significantly
contribute to performance, especially when con-
Sidering weight penalties.
3. Parabolic Approximation of Bell Nozzles
One convenient way to design a near-optimum-
thrust bell nozzle contour is through the use of
the parabolic approximation procedures as sug-
gested by G.V.R. Rao. The design configuration
of a parabolic approximation bell nozzle is
shown in figure 4-13. The nozzle contour imme-
diately upstream of the throat T is a circular arc
with a radius of 1.5 R
t
. The divergent section
nozzle contour is made up of a circular entrance

I
I
I
I
A
T I
R, 1 '-CIRCLE , I
. , I I
w----------------r---!...-- fllZU:
i L. I '<lS
THROAT
Figure 4-13. -Parabolic approximation o[ bell
nozzle contour.
section with a radius of 0.382 R
t
from the throat
T to the point N and a parabola from there to the
exit E.
For the design of a specific nozzle. the fol-
lowing data are required:
(1) Throat diameter, Dr, inches
(2) Axial length of the nozzle from throat to
exit plane, L
n
, inches (or the desired
fractional length Lf based on a 15
conical nozzle)
(3) Expansion area ratio c
(4) Initial wall angle of the parabola. B
n
,
degrees
(5) Nozzle exit wall angle, Be, degrees
The wall angles, en and Be are shown in figure
4-14 as a function of the expansion area ratio c.
-
I--
V
I--
r-
/' :..-
/':: t:::=
::::: f=-

V
l'--
""'-

""'-
....:::::

i
I
,
o
'0 20 30 40
EXPANSION AREA RATIO
Lt'"
Lt 70'/.
Lt 80'4
Lp 90',.
Lt:aIOO".
L
L
L
L
Lf' 60'/,
til 70'/,
,. eo'/.
f' 90'/,
,Z 10C".
Figure 4-14.-13
n
and 13
e
as function of expansion
area ratio c.
92 DESIGN OF LIQUID PROPELLANT ROCKFT ENGINES
Optimum nuzzle contours can be approximated
quite accurately by selecting the proper inputs.
Although no allowance is made for different pro-
pellant combInations . Gxperience has shown that
the effect of spocific heat ratio y upon the con-
tour is small . A computer program can be readily
set up to perform the calculation.
4. Annular ozzles
Based n the momentum theorem, for ideal ex-
pansion the thrust generated by a th:ust chamber
depends only upon the mass fl"lw conditions
(velocity and direction) t t. he nozzle exit. In
some nozzle desiO'ns, uch as annular nozzles,
the gas flow at tae throat is not necessarily
parallel to the axis , but the exit flow is similar
to that of a conical or bell nozzle and thus pro-
duce the same thrust results.
There are two basic types of annular nozzles:
the radial in-flow type (spike nozzle) and the
radial out-flow type (expansion-deflection or
E-D; reverse-flow or R-F; and horizontal-flow or
H-F nozzles). They are shown in figure 4-15,
together with conventional conical and bell noz-
zles. For compari"on of tl e effect of
type on size, all nozzles ShOWI are scaled to
the same thrust level , nozzle expansion area
ra io, and theoretical nozzl e efficiency. These
nozzles show potential of adapting their geom-
etry 0 space vehicle application, hecause short-
ened I ozzles r duce interstage struc ure weight
and will permit an increase in payload through
increased performance for a given length.
The nozzle expansion area ratio ( for an
annular nozzle is defined by equation (4-9):
Proj cted area of the
contoured nozzle wall Ae - Ap
( = Throat area = At (4-9)
where the projected area of the contoured nozzle
wall equals nozzle exit plane area Ae , less he
centerbody projected area Ap. ,Allother conve-
ient design parameter for annular is the
annuler diar.icter ratio, Dpl D
t
, where Dt i 3 t he
throat diameter of an equiv' lent
and D the centerbody diameter . The
Dpl D
t
13 an index of the annular nt)zzle design
AREA RATIO = 36: I
C
F
EFFICi ENCY;: 98.3% (ALT)
BELL
INJECTOR
COMBUSTIO'
CHAMBER
NOZZLE LENGTH = 100% NOZZLe LENGTH = 74.2'Yo
OVERAlL LENGTH = 100% OVERALL LENGT'"i = 78.",.
, . ,
INJECTOR
COMBUSTION

THROAT
SPIKE E- D '
Dp/D, = 1. 3 Dp/DT = 1.3
INJECTOR
NOZZLE LEN;;TH = 41.4% NOZZLE LENGTH = 41.4%
OVERALL LENGTH = 5 1'Y. OVERAlL LENGTH = 51 'Yo
OVEAALL L> AMETER = 100% OVERALL DIAMETER :: 100 % OVERALL DIAMETER = 105% OVERALL D! METER = IC2 . 5 %
Dp .'0T= 5
NOZZLE LENGTH:: 24.9
OVERALL LENuTH = 21 %
OVERALL DIAMETER = 130 %
THROAT
COMBUSTION
CHAMBER
INJECTOR
F
I
Dp THROAT
H-F , r
Dp/D , = 10
NOZZLE LENGTH = 14.5 %
OV:R " ENGTH = 12 %
OVERALL DIAMETER = 194 %
FiglJre 4-15. -Comparison of nozzle shapes.
DESIGN OF THRUST CHAMBERS AND OTHER COMBUSTION DEVICES 93
geometry as compared to a conventional nozzle.
The contour-calculating methods for annular
nozzles are similar to those for bell nozzles.
In a conical or bell nozzle, the gases may ex-
pand to pressures well below the ambient (sea-
level or low-altitude operation) before flow sepa-
ration from the nozzle wall occurs. As explained
in chapter I, for nozzles with large area ratios,
this overexpansion results in thrust losses at
low altitudes. Annular nozzles, because of their
special characteristics, are not subject to these
losses. As shown in figure 4-16 for an E-D noz-
zle (and equally applicable to other annular noz-
zles), the back pressure Pb at the back face of
the centerbody plays an important role in regu-
lating the nozzle flow. The value of Pb is a
function of the ambient pressure and generally is
lower than ambient. Downstream of the throat,
the expansion of the gases around the centerbody
shoulder C will continue unaffected until this
base pressure is reached. After the initial gas
expansion through the constant-Mach line CD,
the downstream flow of the gases is controlled
by the following two boundary conditions:
(1) The nozzle wall contour DE which turns
the gases to near-axial flow.
(2) The base pressure Pb which influences
the free stream surface of the inner jet
boundary.
Because of the curved-wall contour, the gases
are deflected, which leads to some compression
and local increases in wall pressure. A typical
nozzle wall pressure distribution for low-altitude
operation is shown in figure 4-16. This com-
INJECTOR
WALL
PRESSURE
NOZZLE WALL
E
,
1
1
STREAM 1
,
SURFACE, NOZZLE
------+-----
CENTER: AXIS
BODY I
I
I
I
I
I
I
Figure 4-16.-E-D nozzle at low altitude operation.
INJECTOR
WALL
PRESSURE
/'1
/" 1
",." 1
" I
/,/," FREE STREAM SURFACE
I
SHOCK :
1 NOZZLE
--- --1- - AxiS
1
I
I
I
,
,
I
I
,
I
,
Figure 4-17. -E-D nozzle at high altitude
operation.
pressive turning at the nozzle wall, which is
also typical for the spike nozzle, is responsible
for improved nozzle performance at low altitude.
Because of the self-adjusting nature of the inner
jet boundary, there is no flow separation from
the nozzle wall, as is the case for a conven-
tional nozzle.
At high-altitude operation the base pressure
Pb becomes so low that the nozzle flow con-
verges behind the center body , as shown in figure
4-17. Since the flow at the closure point must
be axial, a shock wave may occur depending on
the flow conditions. However, the expansion of
the gases may continue unaffected up to the end
of the nozzle. The nozzle wall pressure distri-
bution under this condition is also shown in
figure 4-17.
An improved spike nozzle concept is the
aerodynamic spike nozzle. I This nozzle concept
is a truncated annular spike nozzle (radial in-
flow type), which utilizes a small amount of sec-
ondary flow introduced into the nozzle base
region.
Performance of the aerodynamic spike nozzle
is a function of various nozzle geometric param-
eters, the amount of secondary flow, the manner
in which this secondary flow is introduced, and
the relative energy between the primary and sec-
ondary streams. To describe the flow field and
interrelated effect of truncating the spike nozzle,
I Source: AlA A Paper No. 66-828. "Liquid Rocket
Engines: Their Status and Their Future" By S. F.
Iacobellis.
94 DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
the base pressure and the base pressure increase
achieved through the secondary flow addition
requires a lengthy, detailed discussion; only the
basic operation can be presented here.
The primary flow (high-pressure gases) which
produces the major portion of the engine thrust
is exhausted from an annular-type combustion
chamber and expands against the metal surface
of the center trurnated-spike nozzle (fig. 4-18).
The characteric of the primary flow field up-
stream of the shown as region 1 in figure
4-18, are determined by the annular throat geom-
etry, the nozzle wall contour, and the ambient
pressure. The annular primary flow continues to
expand beyond the nozzle surface and encloses
a subsonic, recirculating flow field in the base
region (region 2). The pressure acting on the
nozzle base contributes additional thrust to the
nozzle.
When a small amount of secondary flow is
introduced into the base (added to the recirculat-
ing flow), the base pressure is increased further.
As the secondary flow is increased. the overall
nozzle efficiency (considering the additional
flow) increases because of this increase in base
pressure. There is a limit to this gain in effi-
ciency, and an optimum secondary flow exists
for each configuration.
The outer surface of the annular primary flow
is a free-jet boundary, which is influenced by
ambient pressure. This ambient pressure influ-
ence on the primary nozzle flow endows this type
of nozzle with altitude compensation. In oper-
ation at high-pressure ratios (i.e .. altitude con-
ditions). the outer free-jet boundary of the pri-
mary flow expands outward, governed by the
Prandtl-Meyer turning angle at the throat. At
low-pressure ratios (i.e .. sea level operation).
the relatively higher ambient pressure com-
tHN[R fItEf-JEl
iClJND4ltT
Figure 4-18. -Aerodynamic spike flow field illus-
trated under altitude conditions.
I-
Z
w
Ci
u.
u.
w
o
u
30

IDEAL NOZZLE (NO LOSSES)
.... - ....
HIGH-AREA-RATIO
AEROSPIKE NOZZLE
/
I

.....
.... //
,
/
v
, HIGH-AREA-RATIO
&- BELL NOZZLE
I
/::A-LEVEL / (VACUUM)
It---;-
I OPERATING RANGE
SEA LEVEL (VACUUMI
f.------- OPERATING RANGE
PRESSURE RATIO (pipe)
Figure 4-19. -Nozzle performance comparison.
presses the outer free-jet boundary of the primary
flow field. This compression increases the static
pressure on the nozzle wall and partially offsets
the negative effect of the higher ambient pres-
sure on the back side of the nozzle. The base
pressure also is increased with the higher am-
bient. because the compressed primary flow field,
which influences the base pressure, has higher
static pressures. This combination of flow field
effects provides the altitude compensation in-
herent to the aerodynamic spike nozzle.
Figure 4-19 presents the performance com-
parison of various nozzle designs. The nozzle
thrust coefficient C[ for an ideal nozzle (i.e., a
variable-area-ratio nozzle having the optimum
expansion for each chamber pressure to ambient
pressure ratio, PelPa) is shown together with
those of the high-area-ratio aerodynamic spike
and bell nozzle. As is evident, the C[ curve of
the aerodynamic spike follows the ideal nozzle
performance (altitude-compensation), rather than
dropping off rapidly like the bell design at low
Pe/Pa (i.e., sea level) operating points. All
nozzles have a higher C f at a high Pel Pa (i. e.,
vacuum).
The development of the annular-nozzle con-
cept may influence the design of rocket vehicles,
DESIGN OF THRUST CHAMBERS AND OTHER COMBUSTION DEVICES 95
especially in the areas of boattail structure and
mission staging optimization. The advantages
and disadvantages of annular nozzles are stun-
marized as follows:
Advantages
(1) Shortened nozzle length for the same per-
formance. or increased performance (higher
expansion area ratios) for a given length.
(2) Improved performance at sea level or low
altitudes. (Annular nozzles with high
expansion area ratios can be used for a
single-stage sea level to vacuum vehicle
mission.)
(3) The relatively stagnant region in the cen-
ter of the nozzle can possibly be used
for installation of gas generators. turbo-
pumps. tanks. auxiliary equipment. and
turbine gas discharges.
(4) A segmented combustion chamber design
approach can be used. easing develop-
ment effort (individual segments can be
built and tested during the early phases)
and improving combustion stability.
Disadvantages
(1) Relatively high cooling requirements.
because of higher heat fluxes and greater
surface areas to be cooled.
(2) Heavier structural construction in some
applications.
(3) Manufacturing difficulties.
Sample Calculation (4-2)
Layout the thrust chamber internal configura-
tion (cylindrical combustion chamber with bell
nozzle) for the engines on the Alpha vehicle with
the data derived from sample calculation (4-1)
and the following required chamber thrusts Ftc:
(a) A-1 stage engine: F
tc
,==747000 lb at sea
level
(b) A-2 stage engine: Ftc
2
= 149500 lb at
altitude
(c) A-3 stage engine: F tCl = 16000 lb at
altitude
(d) A-4 stage engine: Ftc. = 7500 lb at
altitude
The detailed calculations and their results
are presented in the following for the first-stage
engine only. For the other stages. the calcula-
tion results are summarized in figures 4-21 to
4-23. The reader is urged to conduct his own
calculations using the first stage as a guide. and
to compare his results with those shown.
Solution
A-I Stage Engine:
From sample calculation (4-1):
Design sea level C[= 1.531;
(P c)ns = 1000 psia; (= 14
Substitute into equation (1-33):
Throat area: A
- 747000 _ . 2
t -1.531 x 1000 - 487 m
Throat diameter: D
t
= l ~ x 487 = 24.9 in
R
_
24
.
9
-1245
t- 2 - . m
Exit diameter: De =vl4x 24.9=93.4 in
R
- 93.4 - 46 7 .
e- 2 - . m
Use a combustion chamber L* of 45 in for
L0
2
/RP-1 application. Substitute into equation
(4-4):
Chamber volume: V c = 487 x 45 = 21915 cu in
Use a nozzle convergent half angle of 20. a
contraction area ratio (c = 1.6. and a circular arc
of radius R = 1.5 R
t
or 18.68 in. for nozzle con-
tour upstream of the throat.
Chamber diameter: D
c
=vT.6x 24.9=31.5 in
R
_31.5_
1575
,
c--2-- . m
Use equation (4-7) to calculate the chamber
convergent cone length
Convergent cone length
_12.45 (v1.6 -1) + 18.68 (sec 20- 1)
- tan 20
4.515 12 4 .
=0.364 = . In
96
1
Dc' 31.50"
THRUST
DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
EXIT
INJECTOR
FACE
THROAT
....... --------E. 102.4" ------I-IE
N 12.99"
E
o
'46.7"
---'1--t----t- -------,--- --- --- --+--'--
CHAMBER AXIS
~ 0.,93.4"
16.6S"R
1
I
I
" L c.' ,,,.,.
Figure 4-20. -A-1 stage engine thrust chamber internal configuration layout:
f=14, 80% bell, L*=45", fc=1.6.
INJECTOR THROAT
FACE t-I""----
Et ::83.6"------1
I
I
Eo =35.5"
,
THRUST __ Dc:: I ~ _ " -+--,.,--
CHAMBER AXIS
I
11-3"- +-I--------Ln:: 83.6"-------1
Figure 4-21. -A-2 stage engine thrust chamber, internal configuration layout:
(=40,75% bell, L*=26", fc=1.6.
DESIGN OF THRUST CHAMBERS AND OTHER COMBUSTION DEVICES
INJECTOR THROAT
FACE
Eo: 31.25"
.-_-_--f-_+___ . _ . -+--1._
THRUST
I "
Oc: 14. 9 8
CHAMBER AX IS
,2.02R No' 5.64"

IS"-.I"------- L
n
'S8.1"
Figure 4-22. -A-3 stage engine thrust chamber, internal configuration layout:
(=35,70% bell, L*=28", (c=2.
INJECTOR
FACE
CHAMBER AXIS - -
THROAT
t---E t 46.1"


Dc"O.IS"
EXIT
IB" -+.1- .. -L"' .
Figure 4-23. -A-4 stage engine thrust chamber, internal configuration layout:
(=35,70% bell, L*=32", (c=2.
O. :S2.5"
D.: 42.3"
97
98
DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
Using the frustrum cone volume equation and
neglecting the slight rounding of the throat, the
approximate convergent cone volume is obtained:
Volume =j x12.4 815.75)2+ (12.45f + 15. 75x12.45 ]
= 7760 cu in
Required volume for cylindrical chamber section
= 21915 - 7760 = 14155 cu in
Required length for cylindrical chamber section
=14155/1.6A
t
=18.17 inches
Distance from injector face to throat
=18.17+12.40=30.57, say 31 inches
Design an "80-percent bell" nozzle configura-
tion using the parabolic approximation procedure.
The nozzle contour downstream of the throat will
be a circular arc of radius 0.382 R
t
, or 4.75
inches. By definition, the nozzle length Ln will
be 80 percent of the length for an equivalent 15
half-angle conical nozzle. Substitute into equa-
tion (4-7)
L =0.8x[1245(\'14-1)+4.75(SeC 15
0
-1)]
n tan 150
= 0.8 x 128 = 1024 inches
The parabolic contour wall angles en and ee can
be derived from figure 4-14, for (= 14 and L[=0.8;
en = 274 and ee = 9.8. The location of Nand E
along the nozzle contour, with respect to throat
and nozzle axis, can be calculated
Nc =0.382 R
t
sin en = 2.19 inches
N a = Rc + 0.382 Rt(1- cos en) = 12.99 inches
E
t
:= Ln = 102.4 inches
Ea = Re = 46.7 inches
With the aid of the established coordinates for
points Nand E, and the angles en and ee, a
parabola can be fitted to complete the contour.
The general layout of the A-1 stage engine thrust
chamber is shown in figure 4-20. With the aid of
a computer program, more accurate calculations
of the divergent nozzle contour can be made by
the method of characteristics.
Since the calculations for the thrust chamber
configuration are based on the calculated design
C [ value which has to be verified by later actual
testing, a slight change of chamber pressure is
usually allowed to compensate for C [ deviations
in order to meet the required thrust value.
4.4 THRUST CHAMBER COOLING
Techniques and Their Selection
Because of the high combustion temperatures
(4000 to 6000 F) and the high heat transfer
rates from the hot gases to the chamber wall (0.5
to 50 Btulin
2
-sec), thrust chamber cooling be-
comes a major design consideration. For short-
duration operation (up to a few seconds), un-
cooled chamber walls can be used. In this case,
the heat can be absorbed by the sufficiently
heavy chamber wall material which acts as a
heat sink, before the wall temperature rises to
the failure level. For most longer duration appli-
cations, a steady-state chamber cooling system
has to be employed. One or a combination of the
following chamber cooling techniques is often
used:
1. Regenerative cooling.-Regenerative cool-
ing is the most widely applied method and uti-
lizes one or possibly both of the propellants, fed
through passages in the thrust chamber wall for
cooling, before they are injected into the com-
bustion chamber. (See par. 4.1 and fig. 4-1.)
2. Dump cooling.-With this principle, a small
percentage of the propellant, such as the hydro-
gen in a L0
2
/LH
2
engine, is fed through pas-
sages in the thrust chamber wall for cooling and
subsequently dumped overboard through openings
at the rear end of the nozzle skirt. Because of
inherent problems, this method has only limited
application.
3. Film cooling.-Here, exposed chamber wall
surfaces are protected from excessive heat with
a thin film of coolant or propellant which is in-
troduced through manifolded orifices in the cham-
ber wall near the injector. and usually in several
more planes toward the throat. The method has
been widely used, particularly for high heat
fluxes, either alone or in combination with
regenerative cooling.
4. Transpiration cooling. - Transpiration cool-
ing is accomplished by introducing a coolant
DESIGN OF THRUST CHAMBERS AND OTHER COMBUSTION DEVICES 99
(either gaseous or liquid propellants) through
porous chamber walls at a rate sufficient to
maintain the desired combustion gas side cham-
ber wall temperature. This method is essentially
a special type of film cooling and has been
widely used.
5. Ablative cooling.-In this process a sacri-
fice of combustion-chamber gas-side wall mate-
rial is made by melting and subsequently vapor-
izing it to dissipate heat. As a result. relatively
cool gases flow over the wall surface. thus
creating a cooler boundary layer. assisting the
cooling process. Ablative cooling has been used
in numerous designs. initially mainly for solid
propellant systems. but later equally success-
fully for low pc. pressure-fed liquid systems.
6. Radiation cooling.-With this method, heat
is radiated away from the surface of the outer
thrust chamber wall. It has been successfully
applied to low heat flux regions, such as nozzle
extensions.
The selection of the best cooling method for
a given thrust chamber depends on many design
considerations. There are no simple-and-fast
rules. However. the following are the main fac-
tors which influence the selected design
approaches:
1. Propellants.-The properties of the com-
bustion products, such as temperature, specific
heat. specific weight, viscosity. etc .. have a
direct bearing on the heat transfer rate and in
turn affect the chamber cooling requirements and
methods. The cooling properties of the propel-
lants and their relative flow rate decide whether
they are suitable or sufficient for regenerative or
film cooling. Therefore, in evaluating a chamber
cooling system. the propellants involved will be
one of the primary design considerations.
2. Chamber pressure.-High chamber pressure
is linked with higher combustion gas mass flow
rates per unit area of chamber cross section and
thus raises the heat transfer rate. Combined
regenerative and film-cooling methods are usually
employed for the stringent requirement of higher
chamber pressure applications.
3. Propellant feed system.-The type of pro-
pellant feed used in an engine system deter-
mines the pressure budget for the system. In a
turbopump-fed engine system, more pressure drop
is usually available for chamber cooling. The
availability of this pressure drop permits the use
of regenerative cooling which requires propellant
pressure sufficient to force the coolant through
the cooling passage before entering the injector.
A pressurized-gas-fed engine system usually has
more stringent pressure limitations and operates
on relatively low chamber pressures. This sug-
gests the application of film, ablative. or radia-
tion cooling.
4. Thrust chamber configuration.-The geo-
metric shape of the chamber affects local
combustion gas mass flow rates and wall surface
areas to be cooled. This influences the choice
of cooling method. It can also limit the design
arrangements for regeneratively cooled tubular
wall thrust chambers.
5. Thrust chamber construction material. - The
properties of the thrust chamber materials will
affect the cooling system design profoundly.
Strength at elevated temperature, combined with
heat conductivity properties of a metal. will
1etermine suitability for regenerative cooling
systems. For film-cooled chambers higher allow-
able material working temperatures are desired
to reduce heat transfer rates and thus film cool-
ant flow rates. The application of radiation-
cooling to a chamber depends largely on the
availability of high temperature (3000 F and up)
refractory alloys. The success of ablative cool-
ing depends entirely on the availability of suit-
able composite plastic materials.
In practice. the design of thrust chamber
cooling systems is a major link in the complete
engine system design. It cannot be treated inde-
pendently. without due consideration of other
engine system aspects. For instance. optimiza-
tion of the chamber pressure value for a high-
performance engine system is largely limited by
the capacity and efficiency of the chamber cool-
ing system. In turn. chamber pressure will affect
other design parameters such as nozzle expan-
sion area ratio, propellant feed pressure, and
weight. Because of the complex interrelation
between these factors, the complete analySis of
chamber cooling systems is a specialized field
and requires thorough knowledge of heat transfer,
fluid mechanics, thermodynamics, and thermal
stresses. The engine system designer, therefore,
will enlist the services of heat transfer spe-
cialists.
100
DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
Gas-Side Heat Transfer
One of the primary steps in the design of a
thrust chamber cooling system is the analysis of
the heat transfer from the combustion gases to
the chamber walls (gas-side heat transfer). Be-
cause of the very high surface velocity of the
gases along the chamber walls, the heat transfer
occurs mainly through forced convection; i.e.,
through the transfer of heat energy resulting from
the relative motion of different parts of a fluid.
Before the gases can transfer heat to the wall,
the heat energy must pass through a layer of
stagnant gas along the wall, called the boundary
layer.
The basic correlation for this complicated
convective heat transfer can be expressed by the
following equation:
( 4-10)
where
q = Heat flux or heat transferred across the
stagnant gas film per unit surface area
per unit time, Btu/in 2-sec
hg = Gas-side heat transfer coefficient, Btu/
in
2
-sec-deg F
Taw = Adiabatic wall temperature of the gas,
deg R= (Tc)ns xturbulent boundary layer
recovery factor (ranging from 0.90 to
0.98)
T w g = Hot-gas-side local chamber-wall tempera-
ture, deg R
The determination of the gas-side heat trans-
fer coefficient hg is a rather complex problem.
The convection phenomenon as it occurs in
rocket thrust chambers eludes complete under-
standing. Attempts to compare analytical results
with experimental heat-transfer data obtained on
rocket thrust chambers have often shown disa-
greement. The differences are largely attributed
to the initial assumptions for analytical calcula-
tions. For example, there is good evidence that
oxidizing and reducing atmospheres covering a
wide range of temperature exist locally in the
combustion product gases within the thrust cham-
ber, because of the imperfect mixing of the pro-
pellants at the injector face. This results in
deviations from calculations based on the as-
sumption of homogeneous product gases.
However, it has been established by experi-
ment that the heat-transfer coefficient is pre-
dominantly influenced by the mass vt::!ocity or
the mass flow rate per unit area of the gas,
subject to the exponent 0.8. In comparison, all
other factors are relatively minor. A rough
approximation of hg can thus be expressed by
the following equation:
( 4-11)
where
p' = Free stream value of local gas density,
lb/cu in
V = Free stream value of local gas velocity,
in/sec
Thus, under normal circumstances, the heat-
transfer coefficient varies with the chamber pres-
sure to the 0.8 power and throughout a given
chamber inversely with the local chamber diam-
eter to an exponent of 1.8.
Based on experience with turbulent boundary
layers, some relatively simple correlations for
the calculation of the gas-side heat-transfer
coefficient have been developed. A much-used
form is that credited to Colburn
N u = C Re 0.8 P r 0.34
where
Nu = Nusselt number = hg D/k
C = Dimensionless constant
Re = Reynolds number = p'VD/1l
V = Free stream velocity, in/sec
P r = Prandtl number = /IC p/ k
D = Hydraulic diameter, in
( 4-12)
k =Gas thermal conductivity, Btu/sec-in2-
deg F/in
Il = Viscosity, lblin sec
C p = Specific heat at constant pressure, Btu/lb-
deg F
or as Bartz has shown
h =[0.026 (pc)nsg)0.8 (Dt)O.!]
g D 0.2 P 0.6 c* R
t r ns
(
A )0.9
At a (4-13)
where
R = Radius of curvature of nozzle contour at
throat, in
a = Correction factor for property variations
across the boundary layer
DESIGN OF THRUST CHAMBERS AND OTHER COMBUSTION DEVICES 101
A = Area under consideration along chamber axis
The value of a can be evaluated in terms of noz-
zle stagnation temperature, local gas-side cham-
ber wall temperature, and local Mach number.
values of a for various Twg/(Tc)ns and y, as
computed by Bartz, are shown in figure 4-24.
If P
r
and Jl data are not available for particu-
lar combustion gas mixtures, the following equa-
tions can be used for approximate results:
4y
P
r
= 9y- 5
Jl = (46.6 x 10-
1
) m OST06
where T=temperature of gas mixture, oR
(4-15)
(4-16)
Equations (4-13), (4-14), (4-15), and (4-16)
can be used to calculate the approximate hg val-
ues along the thrust chamber walls. However,
the calculated values can be expected to be
lower than the actual ones if the following con-
ditions exist:
1.2
CT I. 0 l::I=t:"::::::.

o.sH-+--
.we9LJ:
JO
==:='!"zo 30 40

CONTIIACTJON I!XPANSJON
Figure 4-24.-Values of correction factor a for
property variation across boundary layer.
(1) A substantial fraction of the combustion
gases are strong radiators.
(2) There is substantial dissociation, with
subsequent recombination near the wall.
(3) There are strong high-frequency flow
instabilities.
The calculated values may be higher than the
actual ones, because of the following:
(1) The combustion reactions may not be
completed in the chamber.
(2) The combustion gases may deposit solids
on the chamber walls, which act as
insulators.
In certain propellant combinations, the com-
bustion products contain small amounts of solid
particles. These solids tend to deposit on the
chamber walls, and form a rather effective insu-
lating layer. A quantitative evaluation of the
insulation effectiveness of this layer, necessary
for correct heat transfer calculations, has been
accomplished only experimentally.
In the case of the L0
2
/RP-1 combination,
carbon solids are deposited on the chamber
walls. After a firing, the carbon gives the in-
terior of the thrust chamber the appearance of
being freshly painted black. The outer surface
of the carbon appears sooty and can easily be
removed by light rubbing. Underneath the exte-
rior soot layer is a harder, graphitelike layer
which can also be removed, but is more tena-
cious. This carbon deposit significantly in-
creases the gas-side thermal resistance. The
temperature of the carbon deposit at the hot
gas-side interface approaches the gas tempera-
ture as the carbon thickness increases.
The values of the thermal resistance of the
carbon deposit based on actual experimental
testing results of a thrust chamber burning
L0
2
/RP-1 are shown in figure 4-25.
For the heat transfer calculation of the gas-
side heat transfer with solid deposit on chamber
walls, the following equations can be used
( 4-17)
where hgc = overall gas-side thermal conductance,
Btulin 2- sec-deg F
1
hgc = 1
-+Rd
hg
(4-18)
102 DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
2400

OONTF;,ACllON __ (1ft - EXPANSION
AREA RATIO
Figure 4-25. - Thermal resistance of carbon de-
posit on chamber walls L0
2
IRP-l, mixture
ratio = 2.35, (Pc)ns=lOOO psia.
where Rd thermal resistance caused by the solid
deposit, in
2
-sec-deg F/Btu
When there is no solid deposit, Rd = 0 and
hgc = hg, and equation (4-10) is used for heat
transfer calculations.
Sample Calculation (4-3)
Determine the approximate design gas-side
overall thermal conductance hgc in the combus-
tion chamber, at the throat, and at the exit noz-
zle point of f = 5, for the regenerati vely cooled
thrust chambers on the A-l and A-2 stage
engines.
Solution
CiJ A-l Stage Engine
First, let us consider equation (4-13). The
combustion reactions are assumed to be homoge-
neous and complete. From figure 4-3 the follow-
ing values are derived for the chamber product
gases, for L0
2
/RP-1 at (Pc)ns = 1000 psia and a
mixture ratio of 2.35:
(Tc)ns = 6000 For 6460 R,
m =22.5 Ib/mol, y= 1.222
The design (Tc)ns
= Theoretical (T c)ns x (c* correction factor)2
= 6460 x (0.975)2 = 6140 R
(See eq. 1-32a and 1-41).
From sample calculation (4-1):
Design c* = 5660 ft/sec
From sample calculation (4-2):
D
t
=24.9 in
Mean radius of the throat contour
= 18.68+4.75 11.71 in
2
9 (1544
R 1.22_ x 2
9
5
Cp = l)J = (1.222 -1) :778 = 0.485 Btu/lb-deg F
From equation (4-15):
4 x 1.222
P
r
= (9 x 1.222) _ 5 = 0.816
From equation (4-16):
J1 = (46.6 x 10-
1
) x (22.5).5 x (6140)6
= 46.6 X 10-
10
x 4.76 x 188
= 4.18 x 10-
6
lb/in-sec
From equation (4-13):
h =[0.026 x(C4.18xlO-6)02XO.485)
g 24.92 0.8166
x (1000 x 32.2)
0
8 x (At\ 0.9
5660 11.71 J ,A/ a
(
A )0.9
=0.01366xO.046x4.02x1.078x At a
(
A )009
=0.0027x i a
Since the carbon deposit temperature ap-
proaches the gas temperature, a (T wg/(Tc)ns)
value of 0.8 is used to determine the a values
from figure 4-24
At the combustion chamber:
DESIGN OF THRUST CHAMBERS AND OTHER COMBUSTION DEVICES 103
(
At )0.9 (1) 0.9
A = 1.6 =0.655, a= 1.05
hg = 0.0027 x 0.655 x 1.05
= 0.00185 Btulin 2- sec-deg F
At the throat:
(
A )0.9
At = 1, a= 1
hg=0.0027xlxl=0.0027 Btulin
2
-sec-deg F
At the exit nozzle point of
(
At)0.9 (1)0.9
(=5, if = 5" =0.235, a=0.8
h
g
= 0.0027 x 0.235 x 0.8
= 0.000507 Btulin 2 -sec-deg F
The experimental data of figure 4-25 can be
used to determine the values of thermal resist-
ance Rd, for the carbon deposit. The thermal
resistances are
in 2 -sec-deg F in
2
-sec-deg F
1670 Btu ,1125 Btu '
in 2- sec-deg F
and 1645 B
tu
for points at the combustion chamber, the throat,
and the exit nozzle area ratio of (= 5.
Substitute into the equation (4-18); at the
combustion chamber
hgc 1 1 =0.00045Btu/in2-sec-degF
0.00185 + 1670
at the throat
hgc = 1 1 =0.00067 Btulin
2
-sec-deg F
0.0027 + 1125
At the exit nozzle of f= 5.
hgc 1 1 =0.000276 Btulin
2
-sec-deg F
0.000507 +1645
A-2 Stage Engine
Again, the combustion reactions are assumed
to be homegeneous and complete. From figure
4-4, the following values are deri ved for the
chamber product gases, for L0
2
/LH
2
at (Pc)ns
= 800 psia and a mixture ratio of 5.22:
(T c)ns = 5580 F or 6040 R,
)IT = 12 lb/mol, y= 1.213
The design (Tc)ns=
Theoretical (T c)ns x (c* correction factor)2
= 6040 x (0.975)2 = 5740 R
From sample calculation (4-1):
Design c* = 7480 ft/sec
From figure 4-21:
D
t
=11.2 in
Mean radius of the throat contour =
8.4 + 2.14 - - 27 .
2 -t!. In
9 1544
yR
Cp = (y-1)J= (1.213 -1) x 778 =0.943 Btu/lb-deg F
From equation (4-15):
4 x 1.213
P
r
= (9 x 1.213) - 5 = 0.820
From equation (4-16):
JL = (46.6 x 10-
1
)(12)5 (5740)6
= 46.6 X 10-
10
x 3.47 x 180
=2.92x 10-
6
Iblin-sec
From equation (4-13):
h = [0.026 x (2.92 x 10-
6
)2 x 0.943)
g 11.22 0.8206
x (800 x 32.2 )0.8 x (11.2\ 0IJ(At)0.9
\' 7480 5.27) A a

--
104 DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
fA \09
= 0.01605 x 0.0828 x 2.69 x 1.078 x \TJ a
I.
A
)09
= 0.00385 x 'T a
Since there is no solid deposit on the chamber
walls, an average gas-side wall temperature of
1500R is assumed, and a (Twg/(Tc)ns) value of
(1500/5740) or 0.26 is used to determine the a
values from figure 4-24.
At the combustion chamber:
(
A
t
)0.9 L 1 ) 0.9
A = \f6 = 0.655, a= 1.38
hgc = hg = 0.00385 x 0.655 x 1.38
= 0.00348 Btulin 2- sec-deg F
At the throat:
fA )09
\-1 =l,a=1.35
hgc = h
g
= 0.00385 x 1 x 1.35
= 0.00520 Btulin
2
-sec-deg F
At the exit nozzle point of (= 5:
f At )0.9 f 1 )
0
9
\A =\"5 =0.235, a=1.16
hgc = hg = 0.00385 x 0.235 x 1.16
=0.00105 Btulin
2
-sec-deg F
Regenerative Cooling
The heat transfer in a regeneratively cooled
chamber can be described as the heat flow be-
tween two moving fluids, through a multilayer
partition. Figure 4-26 shows this process sche-
The general steady-state correlation
of heat transfer from the combustion gases
through the layers, which include the metal
chamber walls, to the coolant can be expressed
by the following equations:
-1fCHAMBER
I I INNER WALL
GAS SIDE
BOUNDARY LAyER1 r. f.- COOLANT SIDE
IJJ
a:
::>
.....
e:{
a:
w
Q.
::!!
w
.....
Taw-----......I'
COM BUSTfON
GAS
BOUNDARY LAYER
COOLANT
Twe
'.J:t------- Teo
RADIAL DIST ANCE FROM
CENTER OF CHAMBER
Figure 4-26.-Heat transfer schematic for regen-
erative cooling,
1
H = 1 t 1 (4-22)
-+-+-
hge k he
where
q = Heat flux, Btulin 2-sec
hge = Overall gas-side thermal conductance,
Btulin
2
-sec-deg F (see eq. 4-18; with-
out deposits, hge=h
g
)
he = Coolant side heat-transfer coefficient,
Btulin 2- sec-deg F
k = Thermal conductivity of chamber wall,
Btu lin 2- sec-deg F lin
= Chamber wall thickness, in
Taw = Adiabatic wall temperature of the gas,
deg R
T wg = Gas-side wall temperature, deg R
T we = Coolant side wall temperature, deg R
T CD = Coolant bulk temperature, deg R
H = Overall heat-transfer coefficient, Btulin 2_
sec-deg F
The bulk temperature Teo of the coolant in-
creases from the point of entry until it leaves the
cooling passages, as a function of the heat
absorbed, and of the coolant flow rate. Proper
balance of these parameters, to maintain the
chamber walls at temperatures below those at
DESIGN OF THRUST CHAMBERS AND OTHER COMBUSTION DEVICES 105
which failure might occur because of melting or
stress, is one of the major criteria for the design
of regeneratively cooled thrust chambers. For
metals commonly used in thrust-chamber walls,
such as stainless steel, nickel, and Inconel, the
limiting hot-gas-side wall temperature is around
1500-1800 F. The resultant differences be-
tween combustion gas temperature and wall tem-
perature range from 2500 to 6000
0
F.
Assume a station in the thrust chamber with
gas temperature Taw and coolant bulk tempera-
ture T co. Referring to equation 4-21. it is seen
that the heat flux q, which must be the same
through all layers, is a function of the tempera-
tures, and of overall heat transfer coefficient H.
The value of H is composed of the individual
coefficients for the boundary layers and the
chamber metal wall (eq. 4-22). The smaller H,
the smaller is q. However, it is one of the major
design goals to keep coefficient hge low, but
heat transfer coefficient he and conductivity Uk
high, in relation to h
ge
. Since the temperature
differentials are inversely proportional to the
heat-transfer coefficients of the heat flow paths,
the temperature drop will then be steepest be-
tween hot gas and inner chamber wall. The
effect is analogous to voltage drops along resis-
tors in electrical circuits.
It is noted that the heat absorbed by the pro-
pellant used for regenerative cooling raises
temperature of the propellant, and thus the energy
level before it is injected into the combustion
chamber. However, this effect on overall engine
performance is slight, the gain usually being
less than 1 percent. On the other hand, regener-
ative cooling with attendant pressure losses
requiring additional turbopump power or higher
gas pressurization levels imposes a performance
penalty.
Coolant Side Heat Transfer
The coolant side heat-transfer coefficient he
is influenced by many factors. At the high heat
fluxes and temperatures encountered in thrust
chamber operation, the propellants used for cool-
ing may become corrosive, may decompose or
deposit impurities upon the heated surface,
thereby reducing cooling effectiveness. It is
impossible to calculate the he values under
these conditions without experimental data.
The characteristics of coolant side heat
transfer depend largely on the coolant pressure
and coolant side wall temperature. In figure
4-27, the heat flux is plotted versus wall temper-
ature for a constant coolant pressure, bulk tem-
perature, and flow velocity. Curve A indicates
the behavior of heat transfer at coolant pres-
sures below critical. Line segment Al -A2 repre-
sents the heat transfer without boiling when the
wall temperature is below the saturation temper-
ature of the coolant corresponding to the fluid
pressure. As the wall temperature at A2 exceeds
the saturation temperature by a certain margin
(50
0
to 100 F), bubbles will form within the
coolant layer close to the wall. The bubbles
grow continuously out into the colder liquid
stream until condensation at the vapor to liquid
surface begins to exceed the rate of vaporization
at the base of the vapor bubble, whereupon the
bubbles start to collapse. This process, which
occurs at high frequencies, is described as
"nucleate boiling." It substantially increases
the heat-transfer coefficient, resulting in little
increase in wall temperature for a wide range of
heat fluxes. The heat transfer with nucleate
boiling is represented by line A
2
-A
3
At A
3
,
further increase in the heat flux abruptly leads
to such a dense bubble population that the bub-
bles combine into a vapor film with an attendant
large decrease in heat-transfer coefficient. The
region of heat transfer with film boiling is repre-
sented by line A
3
-A
4
The resulting increase in
-u
f7
W
x
(J)
::>1\1
-I
Z
II.. ,
~
::>
r-
ILl (II
:J:
7
6
A3 -. - -
A4
r-- - - r--
5
~
CURVE A
(P
eo
= 1/2 Pe.meAL)
4
,""
..- B3
3 ,
B2
2
V
A
r--
r-f
"-
CURVE B
A I ~
'I"
(P
co
> PC.ITICAL)
I 1 J
o BI2 4 6 8 10 12 14 16 18 20
COOLANT SIDE WALL TEMP.
TWC (OF) X 10-
2
Figure 4-27. -Heat flux versus coolant side wall
temperature of typical propellant in various
heat transfer regions.
106 DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
wall temperature is so high that failure of the
wall material often occurs. The heat flux at A3
is defined as the upper limit of nucleate boiling
of the coolant quI. which therefore should be
used as the design limit for a regenerative cool-
ing system.
Curve B indicates the heat transfer behavior
of a coolant above critical pressure. Since no
boiling can occur. the wall temperature continu-
ally increases with increasing heat flux. Line
B 1- B 2 represents the heat-transfer region. when
the wall temperature is below the coolant critical
temperature. The heat-transfer coefficient re-
mains essentially constant. As the wall temper-
ature reaches the critical temperature B2 and
higher. a gradual transition to a stable super-
critical vapor-film boundary layer begins. which
results in somewhat lower heat-transfer coeffi-
cients. Line B2-B3 represents the heat transfer
in this region. Wall failure temperatures are
usually reached at lower temperatures when the
coolant is above the critical pressure than when
it is below it. Where possible. a coolant oper-
ating pressure between 0.3 to 0.7 of critical
pressure should be used to take advantage of the
high heat-transfer coefficients available with
nucleate boiling. However. in most systems.
particularly those which are fed from a turbo-
pump. the cooling jacket pressure. which is
equal to or larger than the sum of chamber pres-
sure and injection pressure. is supercritical.
For the nonboiling subcritical temperature re-
gions of both, subcritical and supercritical cool-
ant pressures (A
1
-A
2
and B]-B2 in fig. 4-27), the
relationship between wall temperature and heat
flux, which depends on the heat transfer coeffi-
cient he. can be predicted with sufficient accu-
racy for design purposes with the help of the
Sieder-Tate equation (eq. 4-23) for turbulent heat
transfer to liquids flowing in channels:
JI 0]4
Nu=CIReoSPr04 8w) ( 4-23)
where
c] = a constant (different values for various
coolants)
Nu = Nusselt number=hedlk
Re = Reynolds number=pVeodlJI
P r = Prandtl number = JIC pi k
JI = coolant viscosity at bulk temperature
JIw = coolant viscosity at coolant sidewall
temperature
d = coolant passage hydraulic diameter. in
k = coolant thermal conductivity. Btu/sec-in 2_
deg F/in
p = coolant density, lblin 3
Veo = coolant velocity, in/sec
C p = coolant specific heat at constant pres-
sure, Btu/lb-deg F
The heat flux at the upper limit of nucleate
boiling qui can be estimated from
where
C
2
qnonboiling Peo
G
( 4-24)
= constant. its value depending on
coolant used
qnonboiling = heat flux without nucleate boiling,
Btu/in
2
-sec
Peo = coolant pressure, psia
G = coolant maximum flow rate per
unit area. lblin 2-sec
When the heat is transferred through a vapor-
film boundary layer (coolant at supercritical
pressure and temperature, region B
2
-B
3
in fig.
4-27). the coolant-side heat-transfer coefficient
he can be estimated from
_ 0.029 C p/1 02 (G o.s (Teo) 0.55
he - -)
P//3 d
02
Twe
( 4-25)
where
C
p
= coolant specific heat at constant pres-
sure. Btu/lb-deg F
/1 = coolant viscosity, lblin-sec
P r = Prandtl number
G = coolant weight flow rate per unit area,
Ib/in
2
-sec
d = coolant passage hydraulic diameter, in
Teo = coolant bulk temperature. deg R
Twe = coolant side wall temperature. deg R
The bulk temperature of most coolants should
be kept below the critical temperature. since the
vapor-film heat-transfer coefficient would be too
low to cool the wall effectively. The cooling
capacity of the liquid-state regenerative coolant
system can be estimated by
(4-26)
DESIGN OF THRUST CHAMBERS :\ND OrHER COMBUSTION DEVICES 107
Figure 4-2B.-Coaxial ::;hell t.'lrust chamber cuta-
way. Note overheated and burnt.-throllgh spot
nn clianJber
where
Q c = coolant capacity, E: u/ sec
we = cool a. t mas now rate, lb/ sec
C p = C00lant i1eat at 00nstant pres-
sure. Btu/ lb-deg F
Tee = c )olant critical temperature, deg R
'T ci = coolant inlet temperature, deg R
The allowed value of the total charr;ber wall gas
heat-transfel rate Q should be kept below Qc by
a afe filar in (Q Qc)
However there is no buch limitation for h -
drogen when u ed as a coolant. Hydro en has
excellent heat-tran fer characteristics with a
rea onabl ltigh heat-tran fer co ffic .. e t even in
the supercritical pr s ure and temperature region.
Usually liquid hydrogen enters the cham r 0001-
nt pa sa under superc:- ritical pressur and
reacltcls up r ritical t mp rature a short dis-
tance fro'll the i nl et .
001 Dt pas ... 0 area at various points
alon the chamber walls are designed to waintain
the proper oolant velocity dictated by the heat-
tran fer 0 fficients determined by the heat-
transfer calculatIOn. Ther3 are several basic
design approaches for regenerative-cooled thrust
chambers. Axial-flow cooling jackets, made up
of tubes , are us"'d in the design of large thrust
chambers (3000 pounds of thrust and up); coa:.ial
shells separated by helical ribs or wires are
typical of the smaller thrust chamber designs.
Figure 4-1 shows a lal'ee regenerative cooled
tubular wall thrusc chamber . Figure 4-28 repre-
sents a typif'al co!txial shell desir,D for a smaller
thrust chamber.
In thi design, the coolant p ssage is defined
as the rectan 'ul ar area between inner and outer
sh('ll and two a ,iacent ribs, wh' h are wrapped
helically around the inner shell or liner .
Tubular Wall Thrust Cbamoer Desig]1.
In the design of tubular wall thrust ch mbers,
the number of coolant tubes required :s a
tion of the chamber geometry, the coolant weight
flow rate per unit tube :....ea, t he maximum allow-
able tube wall stress, and fabrication eonsidera-
tions, The critical cooling region i near the
throat where the heat flux is 'lighest. I is this
regicn, ,hen, which determines the Gumhel' of
tubf;s required for a given coc1ant flow rate. For
easier fabrication and lower tuoo cross
sections of cir ' ular shape re vreferred. How-
ever, other shapes are often u;3t !d to meet certain
fl0'w-area requjrements . The Ell ss analysis of
the tubes is based upon three primary (;ons'dera-
tions: the hoop stress caused by coolant pres-
sure, the thermal tress caused by temperatur9
gradient across the tube section and the wall ,
and the bendmg stre s caus d by Jistortion in-
d ced by the pressure differe tial between adja-
cent tulJes (if any) or by other effects such as
discontinuities. The tube desi n tress is based
on the combi ned strt3SS from til e above three con-
siderations. It has found that the maximum
combined stress will 0 'cur at the throat region.
As ohown in figure 4-29, the com-
bined tdngential stresses of the clrcular-tube
will be at ection A-A and can be expressed by
where
S
_ (Pco - Pg) r
t -
t
(4-27)
St = comt"ned angential tensile ' res , Ib/ in 2
108
DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
q = heat nux, Btulin
2
-sec
r =tube radius. in
=tube wall thickness, in
Peo = coolant pressure, lblin 2
Pg =combustion gas pressure, lb/in
2
E = modulus of elasticity of tube wall mate-
rial, lb/in 2
a = thermal expansion coefficient of tube wall
material. inlin-deg F
k = thermal conductivity of tube wall material,
Btulin 2- sec-deg F lin
v = Poisson's ratio of tube wall material
M A = bending moment caused by discontinuity,
in-lblin (no effect of pressure differen-
tial between adjacent tubes for circular
tube design)
Since the combustion-gas-side portion of the tube
(zone I) has a much higher mean temperature than
that of the back side tube portion and chamber
outer shell (zone II), the thermal expansion of
zone I will be restrained by zone II. Because of
the considerably greater mass of zone II, thermal
inelastic buckling is induced under certain con-
ditions, in zone I, in the longitudinal direction.
The longitudinal thermal stress can be estimated
by
(4-28)
where
SI = longitudinal thermal stress, Iblin
2
t\ T = mean temperature difference between zone
I and zone II, deg F
S I should be kept at a level not higher than
0.9 Sc, below.
The critical stress for longitudinal inelastic
buckling on zone I can be estimated by
S - 4Er
E
c
t
c - (VEt ..,,; E c ) 2 \/3 (1 - V 2) r
( 4-29)
where
Sc = critical stress for longitudinal inelastiC
buckling in zone I, lblin
2
Er = tangential modulus of elasticity at wall
temperature, Iblin 2
E c = t.angential modulus of elasticity from com-
pression stress-strain curve, at wall
temperature, lb/in 2
An elongated cross section tube design is
shown in figure 4-30. Equations (4-27), (4-28),
and (4-29) can be also applied to calculate the
stresses for this design. Here again the maxi-
mum combined stress is at section A. The bend-
ing moment at section A, M
A
, should take into
consideration the pressure differential (if any)
between adjacent tubes.
It
THRUST CHAMBER
EQU/V. CHAMBER
INTERNAL RAD.
/
Figure 4-29. -Circular tube wall of regeneratively
cooled thrust chamber.
CHAMBER
JACKET
t
THRUST CHAMBER
EQUIVALENT CHAMBER
INTERNAL RADIUS
Y
Figure 4-30.-Elongated tube wall of regenera-
tively cooled thrust chamber.
DESIGN OF THRUST CHAMBERS AND OTHER COMBUSTION DEVICES 109
(4-30)
where
MA = combined bending moment at section A,
in-Iblin
MA = bending moment due to discontinuity
KA = dimensionless design constant based on
test results (range 0.3-0.5)
= length of nat portion on tube wall, in
t'.\ p = pressure differential between adjacent
tubes,lblin
2
Substituting equation (4-30) into equation (4-27),
the maximum stress of the elongated tube can be
calculated.
The working loads induced in a regenerative
tubular wall chamber by the chamber pressure are
designed to be absorbed by a chamber jacket or
tension bands wrapped around the tube bundle.
Coaxial Shell Thrust Chamber Design
In a coaxial-sheIl-type thrust chamber as
shown in figure 4-28, the outer shell is sub-
jected only to the hoop stress induced by the
coolant pressure. The inner shell, however, is
subjected to the combination of compressive
stress caused by the pressure differential be-
tween the coolant and combustion gases, and of
the thermal stress caused by the temperature
gradient across the wall. The maximum stress
occurs at the inner-wall surface of the inner
shell and can be calculated from the following
equation.
(Pea - Pg) R Eaqt
Sc = t + 2 (1 - v) k
( 4-31)
where
Sc = combined maximum compressive stress,
lb/in 2
q =heat nux, Btulin
2
-sec
R = radius of the inner shell, in
= thickness of the inner shell, in
Pea = coolant pressure, lblin 2
Pg = combustion gas pressure, Ib/in
2
E = modulus of elasticity of inner shell mate-
rial, lb/in 2
a =thermal expansion coefficient of inner
shell material, inlin-deg F
k = thermal conductivity of inner shell mate-
rial, Btulin 2- sec-deg F lin
v = Poisson's ratio of inner shell material
Pressure Drop in Cooling Passages
It is desirable to design a cooling passage
with minimum pressure drop. Abrupt change of
now direction and sudden expansion or contrac-
tion of now areas should be avoided. The inner
surface of cooling passages should be smooth
and clean. The pressure drop in a cooling pas-
sage can be treated as that in a hydraulic con-
duit and be calculated accordingly.
where
v: 2
t'.\ = [ ! : : . ~
p d 2g
( 4-32)
t'.\p = coolant pressure drop through the portion
of cooling passage under consideration,
lb/in 2
L = length of that portion, in
d = equi valent average diameter of that por-
tion, in
p = average density of the coolant, lblin 3
Vea = average coolant now velocity, in/sec
g =mass conversion factor, equal to gravita-
tional constant, 32.2x 12 in/sec
2
[ = friction loss coefficient, which is a func-
tion of the Reynolds number, and of
cooling passage conditions such as
surface smoothness, geometric shape,
etc. This can be determined only ex-
perimentally (also see fig. 7-20 and
table 7-3)
Sample Calculation (4-4)
Determine the cooling passage and the tube
design at the throat for the thrust chambers of
the A-1 and A-2 stage engi-nes.
Solution
(!) A-l Stage Engine
The fuel, RP-l, is used as the coolant. Since
the cooling requirement is most stringent at the
throat, the tube design for this station will serve
as the starting point for the entire chamber. For
high strength, Inconel X is chosen as the tube
WI. a-_
110
DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
material. Based on experimental test results
which showed good solid carbon deposits, design
values not exceeding 1000
0
For 1460
0
R may be
permitted for gas-side tube-wall temperature.
Specifically for the throat region, a T wg value of
1188
0
R is taken. Using the results of sample
calculation 4-3, the value for the adiabatic wall
temperature Taw can be calculated by multiply-
ing (T ch s by the estimated stagnation recovery
factor of 0.923, (. T
aw
=6140 x O.923=5667R.
From the same caleulation, the overall gas-side
thermal conductance at the throat region is
hgc =0.00067 Btulin
2
-sec-deg F. Substitute into
equation (4-10), to obtain the heat nux at the
throat:
q = (5667 - 11 88) x 0.00067 = 3.00 Btulin 2-sec
From supplier's specifications, the following
average data are obtained for Inconel X at 1000
0
-
1200
0
R: Coefficient for thermal expansion,
a = 8 x 10-
6
in/in-deg F; modulus of elasticity,
E=28x10
6
psi; thermal conductivity, k=3.19
x 10-
4
Btulin
2
-deg F/in; Poisson's ratio,
v=0.35.
A circular tube configuration with an internal
diameter d and a wall thickness t of 0.020 inch
is used. The assumption for thickness is to be
verified by heat transfer and stress calculations.
From equation (4-19), the coolant side wall tem-
perature then is;
T =T,
we wg k 3.19x10-4
A double pass design is used; i.e., the coolant
passes down through alternating tubes and up
through adjacent tubes.
For an "up" tube, assume a coolant bulk tem-
perature Teo = 600
0
R at the throat (the more
severe case, since the coolant has passed the
throat region before, on the way down). This
temperature is well below the critical one and
can be expected to remain so for the remainder
of the passage. Total temperature increase for a
typical thrust chamber design is of the order of
100
0
F between cooling jacket inlet and outlet.
The heat-transfer coefficient required to permit
the calculated heat flux for the temperature dif-
ferential assumed can now be calculated from
equation (4-20):
he q = 3.00
Twc-Teo 1000-600
= 0.0075 Btulin
2
-sec-deg F
The relationship between required he and correct
tube diameter is established by equation (4-23),
and experimental data for RP-1 eC
I
=0.0214):
or, substitutin& corresponding terms:
The following additional relationships exist:
N b f b N
rr[Dr+0.8(d+0.04)]
um er 0 tu es = ---:-:---::'-:--:-----'-'--
(d + 0.04)
_(0.8d+24.93) (b)
- (d+0.04)
From sample calculation (4-2), Dr = 24.9
inches. The factor 0.8 considers the fact that
the tube centers are located on a circle, rather
than a straight line.
For our double-pass design, the coolant
velocity then is
W[
P 827x8 2106
Veo= N (lTd2) = r.Nd2p = Nd 2p (c)
"2-4-
From table 3-2: w[=827lb/sec; p=local
value of fluid density.
RP-1 at 600
0
R has the following properties:
/1 = 4.16 x 10-
5
lblin-sec
k = 1. 78 x 10- 6 Btulin 2 -sec-deg F lin
Cp = 0.5 Btu/lb-deg F
For RP-l at 10000R, I1w=0.416x10-5lblin-sec.
Now substitute known values and equation (c)
into equation (a)
DESIGN OF THRUST CHAMBERS AND OTHER COMBUSTION DEVICES 1lI
~
2106 dO.
8
P
d--
Nd
2
0.0075d 0.0214x P
1.78 x 10-
6
4.16xlO-
s
X(0.5 x 4.16 x 10-
5
).4 X( 4.16 X 10-
5
)014
1.78 x 10-
6
0.416 x 10-
5
4220 d = 115000 ( ~ d ) 0.8
as an interpolation between fuel pump outlet
pressure and injector manifold pressure.
Combustion gas pressure at the throat
L
(
2)Y-I _ .
Pg = Pt = (Pc)ns y + 1 = 1000 x 0.b62 = 562 PSI a
(y = 1.222 from sample calculation (4-2); use
table 1-2).
N = 62.4 d-
22S
(d) Thus maximum tensile stress at the inner tube
wall face can be determined using equation
Substitute equation (d) into equation (b) (4-27):
624u- 225 rr(O.8d+24.93)
. (d+0.04)
d=0.85 in
Substitute (d) into equation (b)
N= 94.5
Since for two-pass design a whole, even tube
number is needed, the design value of N = 94 is
used. Substituting this into equation (b)
d = 0.855 in
For the determination of a new tube design, re-
peated calculations, with varied assumptions,
will be required. An experienced designer will
require fewer approaches, particularly if test
results of prior, comparable designs are avail-
able. However, even for complicated conditions,
great amounts of data can be generated in a rela-
tively short time if an electronic computer is
available.
From table 3-2:
- 50.45 - 00292 lb/' 3
p- 1728 - . In
Substitute into equation (c)
2106
Veo
94 x (0.855)2 x 0.0292
= 1051 in/sec or 87.6 ft/sec
At the throat Peo = 1500 psia is established,
S _ (1500 - 562) x 0.427
t- 0.020
28 x 10
6
x 8 x 10-
6
X 3.0 x 0.02 6 MA
+ + -
2 x (1- 0.35) x 3.19 x 10-
4
t
2
= 20000+ 32 500 + 15000 MA
= 52 500 + 15 000 M A
Based on suppliers' recommendation, F
ty
= 82 000 was used for Inconel X at 1000
0
R.
. 82000-53700
MaxImum allowable MA = 15000
= 1.88 in-Ib/in
From experience it can be assumed that the
bending moment due to discontinuity in this case
will be less than 1.88 inlIb/in. Thus the as-
sumption of 0.020 in thickness for the tube wall
is sufficient. Summarizing the tube configuration
at the throat:
d=0.855 in, t=0.020 in, N=94
A-2 Stage Engine
The fuel, hydrogen, is used as the coolant.
Again, Inconel X is chosen as the tube material.
To avoid the "hot shortness ~ or low-ductility
properties of Inconel X in the range 1200
0
-
1400
0
F, the mean temperature of the tube wall
must be kept under 1000
0
F (or 1460
0
R). The
value for adiabatic wall temperature Taw of the
gas can be calculated using an assumed stagna-
tion recovery factor of 0.92.


112 DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
Taw = (Tc}ns x 0.92 = 5740 x 0.92= 5270 R
From sample calculation (4-3) the overall
gas-side conductance at the throat region hgc
= 0.00520 Btulin
2
-sec-deg F. Substitute into
equation (4-19), to obtain the heat nux at the
throat:
q = (5270-1600) x 0.00520 = 3670 x 0.00520
= 19.10 Btulin
2
-sec
A value of 1600 R will be used for the gas-
side wall temperature T wg at the throat region.
From supplier's speCifications the following data
is obtained for Inconel X at 1600 R or 1140 F:
a = 8.2 x 10-
6
in/in-deg F; E = 24 X 10
6
psi;
k=3.86x 10-
4
Btulin
2
-sec-deg Flin
v=0.35.
Use a circular tube configuration with an in-
ternal diameter d and a wall thickness t of 0.008
inch.
From equation (4-19) the coolant side wall
temperature
T = 1600 _19.10 x O.OOS 1600 - 396 = 1204 R
we 3.S6 x 10-
4
A mean value will be used for the wall tem-
perature
Assume a coolant bulk temperature, T co
= 135 R at the throat; then, from equation (4-20),
Substitute into equation (4-25),
0.0179 p,.. -- ~
0.029 C 112 (GO.B) (T )0.55
p
r
2l3 (dO.
2
) T we
(a)
From figure 4-21. D
t
= 11.2 in.
The following relationships exist:
,,[D
t
+ 0.8 (d x 0.016)]
Number of tubes N= (d+0.016)
_ ,,(O.S d + 11.213) (b)
- (d +0.016)
A H ~ - p a s s design is used (Le., the coolant
enters the fuel manifold at the f = S nozzle plane,
flows to f = 30 and back. then passes through the
throat and combustion chamber zone before it
enters the injector). From table 3-3:
wt=54.5lb/sec
Coolant weight now rate per unit area
(c)
For hydrogen at 135 R:
P
r
=0.82; C
p
=3.5 Btu/lb-deg F;
11 = 0.367 X 10-
6
lblin-sec
Substitute these values and equation (c) into
equation (a):
0.0179 = 0.029 x 3.5 x (0.367 x 10-
6
).2
(0.S2)2/3
or
[(
69.3)0.8J
x ~ x (135 )0.55
d
O
.
2
1204
N = 3.91 d-
225
Substitute equation (d) into equation (b)
391a2.25 ,,(0.8d+ll.213)
. (d+0.016)
d=0.185 in
Substitute into equation (b)
N =" (O.S x 0.17 + 11.213) = 178
(0.17+0.016)
DESIGN OF THRUST CHAMBERS AND OTHER COMBUSTION DEVICES

. j
Maximum ensile is now check d at the
inner wall surface u in equation 4-27):
At the throat, estimated P co = 1200 psia;
Pg =Pt =(Pc)os (2/ (y+ 1))y / (Y- l = 00 xO.554 =443
psi a (y =1.213 from sample calculation (4-1))
S _ (1200- 443) 0.0925
t - 0.008
= 68750 + 93 900 MA
0.008
MA
+6---
(0.008 1.
Based on suppliers' recommendations ,
F ty = 81000 psi for Inconel X at 1200
0
R
81000 - 68750
Maximum allowable MA 93900
= .131 in-lblin
F om experience, i can be as 'umed thr,t.L
bending moment due to discontinuity will be l p.8 .
than 0.131 in-lblin. Thus the s l ection 0 C.O
inch tube hickness is valid. Summarizin
tllb0 configuration at the throa :
d =O.185 inch, t = 0.008 inch, N=178
As a general esign vid, figures 4-31 and 4-32
present construction detail for a typical re ener-
atively cooled thrust chamber. The chamber
showlJ is very similar to the one pr sented in
figures 4-1 and 4-2. Figure 4-31 shows clearly
how the tube shape changes along the longitudi-
nal axis . In typical manufacturing proces ,
tubes of uniform circular cross-section area ar
first cut to length, then swaged. The latter
operation can best be accomplished by i nternal
hydraulic pressure in a die. The tubes are filled
with wa:-, and are then bent (preshaped), in a
special fixture, to tile thrust-chamber contour.
Figure 4-31.-Typical regeneratively cooled tube wall thru t chamber.
Figure 4-32. -Detaii ,of ' njectoc manifoldi ng and return manifold of typical cegener Hvely cooled tube
wall chamber.
.. .. .. .
~ ... ~ . . ...... .
114
DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
The tube is then placed in a die of varying
cross-sectional area. Hydraulic pressure applied
to the inside of the tube forces it to aline with
the die and to assume final shape. In prepara-
tion for assembly, a trimming process usually
follows.
In preparation for assembly into a chamber,
the tubes are arranged on a brazing fixture (core).
For proper brazing great care is required to as-
sure even distribution of the gaps between tubes.
Earlier chamber models were then hand brazed, a
process requiring many weeks and considerable
skill. More recently, furnace brazing has been
successfully applied, drastically cutting chamber-
assembly time.
Dump Cooling
The dump cooling technique may be particU-
larly effective for applications in hydrogen-
fueled, low-pressure systems (Pc < 100 psia), or
for nozzle extensions of high-pressure hydrogen
systems. A small amount of the total hydrogen
flow is diverted from the main fuel feed line, fed
through cooling passages and ejected. The heat
transfer mechanism is similar to that of regener-
ative cooling. The coolant, however, becomes
superheated as it flows toward the nozzle exit,
where it is expanded overboard at reasonably
high temperatures and velocities, thus contribut-
ing some thrust. Application of dump cooling is
often limited, however, by various technical dif-
ficulties, such as discharge nozzle design at
low coolant flow rates.
The type of coolant path for a dump-cooled
thrust chamber is selected to assure maximum
overall engine system performance. Two possi-
ble paths are shown in figure 4-33:
(1) Axial flow: a one-pass longitudinal pas-
sage, using double-wall (fig. 4-33A) or
tube-wall designs (fig. 4-33C). Both are
open ended, with provision for expansion
of the dumped superheated hydrogen gas
at the nozzle exit.
(2) Circumferential flow: a double-wall de-
sign having a spiral flow path for the
coolant and provision for expansion of
the dumped superheated hydrogen gas in
the axial direction (fig. 4-33B). A
spiral-wound tube design may also prove
advantageous (fig. 4-33D).
A. LONGITUDINAL PASSAGE
DOUBLE - WALL CHAMBER
C. LONGITUDINAL TUBEwALL
CHAMBER
B. SPIRAL PASSAGE
DOUBLE-WALL CHAMBER
D. SPIRAL TUBEWALL
CHAMBER
Figure 4-33.-TypicaJ dump-cooled chamber fab-
rication methods.
The various constructions differ considerably
in complexity and fabrication cost. Selection
will depend on an optimum tradeoff between reli-
ability, performance, cost, and weight. The
longitudinal passage designs are often employed
for larger coolant flow rates, as related to the
physical size of the thrust chamber. The spiral
passage designs are used for smaller coolant
flow rates to alleviate the difficulties in main-
taining proper flow velocities and dimensional
clearances in the coolant passage.
Because the hydrogen coolant gas can be dis-
charged at relatively high temperatures (1000 R
and up), overall engine system performance will
not be affected appreciably by the dump-coolant
flow.
.. : ... ..
..
DESIGN OF THRUST CHAMBERS AND OTHER COMBUSTION DEVICES 115
Film Cooling
Porous wall materials, or slots and holes pro-
vided in thrust chamber walls, can serve to in-
troduce a coolant, a process commonly referred
to as film cooling. Because of interaction be-
tween coolant film and combustion gases, as a
result of heat and mass transfer, the effective
thickness of the coolant film decreases in the
direction of flow. In most case s, therefore,
additional coolant must be injected at one or
more downstream chamber stations. Figure 4-34
shows a model of the film-cooling process. The
coolant is introduced through rows of holes. The
fluid introduced through row" A" will cover the
wall surface between "A" and "B." Fluid from
row .. B" will cover the surface between .. B" and
"C," etc. In an optimum design a flow rate from
each row is provided which is just sufficient to
cover the area to be cooled.
Although heat protection exclusively by film
cooling has not been applied in the past for the
major operational rocket engines, it is signifi-
cant that in practice regenerative cooling is
nearly always supplemented by some form of film
cooling. In most instances, a fuel-rich gas
boundary layer is created by the injection of fuel
from the outermost circle of injector orifices,
toward the chamber wall.
h, IT .
HEAT TRANSFER ====r=:;;
j " ,COMSUSTION CHUSER
./ I .-'
Figure 4-34.-Film-cooling model.
An important advantage of film cooling is the
fact that it reduces heat transfer through the
wall. Consequently, thermal stresses become
less critical. This is an important considera-
tion, as thermal stresses may establish the
feasibility limits of conventional regenerative
cooling.
Liquid Film Cooling
It would appear, and has been verified experi-
mentally, that for simply reducing the heat trans-
fer to the wall, film cooling would be more effec-
tive with the coolant injected as a liquid rather
than a gas. When the coolant film is liquid, it
should behave essentially as an isothermal heat
sink, as it evaporates and diffuses into the free
stream. However, this process results in two-
phase flow, consisting of an annular liquid-
coolant film and a combustion gas core. This
effect introduces coolant losses which reduce
the theoretical cooling potential. Disturbances
in the form of capillary waves appear on the
surface of the liquid film adjacent to the com-
bustion gases and cause accelerated coolant
loss.
The theoretical equation by Zucrow and
Sellers can be used for design calculations of
liquid-film-cooled thrust chambers.
where
G
c
GC _ 1 H
G
g
- Tfc a (1 + bCpvc/Cpg)
(4-33)
= film coolant weight flow rate per unit
area of cooled chamber wall surface,
lb/in
2
-sec
G g = combustion gas weight flow rate per
unit area of chamber cross section
perpendicular to flow, lblin 2-sec
Tfc = film cooling efficiency
H = film coolant enthalpy
_ Cpvc(Taw-Twg)
- CpIc(T
wg
- Tco)+!iHvc' Btu/lb
CpIc = average specific heat at constant pres-
sure of the coolant in the liquid phase,
Btu/lb-deg F
C
pvc
= average specific heat at constant pres-
sure of the coolant in the vapor phase,
Btu/lb-deg F

.. ~ .... ~ .
116
DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
C pg = average specific heat at constant pres-
sure of the combustion gases, Btu/lb-
deg F
Taw = adiabatic wall temperature of the gas,
deg R
T wg = gas-side wall temperature and coolant
film temperature, deg R
Teo = coolant bulk temperature at manifold,
deg R
~ H ve = heat of vaporization of coolant, Btu/lb
a =2Vd/Vm[
b =(Vg/Vd)-l
[ = applicable friction coefficient for the
two-phase flow between combustion
gases and liquid film coolant
V d = axial stream velocity of combustion
gases at edge of boundary layer, ft/sec
V m = average axial stream velocity of com-
bustion gases, ft/sec
V g = axial stream velocity of combustion
gases at the center line of the thrust
chamber, ft/sec
In practice the theoretically determined film
coolant flow would be inadequate because of
losses, Therefore, the film-cooling efficiency "Ie
is introduced to correct for this. Liquid-film-
cooling efficiency values range from about 30 to
70 percent. They are determined experimentally
in actual hot firings of a specific design or test
model.
Hydrocarbon fuels have been found to be very
effecti ve liquid film coolants. Their effecti ve-
ness is attributed to their action as both film and
deposition cooling agents. As was mentioned
earlier, these fuels deposit carbon on the wall,
which serves as an effective heat insulator.
Gaseous Film Cooling
With the increasing use of hydrogen, gaseous-
film cooling has become important. Even if hy-
drogen were injected as a liquid for film-cooling
purposes, the film between the combustion gases
and the chamber wall would be heated within a
very short distance to temperatures above the
critical, after which the film would behave as a
gas.
For design calculation of gaseous-film-cooled
thrust chambers, the theoretical equation of
Hatch and Papell can be used. This equation
can be written as follows:
where
Taw-Twg
Taw-Teo
( 4-34)
Taw = adiabatic wall temperature of the gas,
deg R
T wg = maximum allowable gas side wall tem-
perature, deg R
Teo = initial film-coolant temperature, deg R
e = base of natural logarithms, 2.718
h g = gas-side heat-transfer coeffiCient,
Btulin
2
-sec-deg F
G
c
= film-coolant weight flow rate per unit
area of cooled chamber wall surface,
Iblin 2-sec
C
pve
= average specific heat at constant pres-
sure of the gaseous film coolant,
Btu/lb-deg F
"Ie = film-cooling efficiency
The film-cooling efficiency "Ie corrects for the
amount of gaseous-film coolant lost into the com-
bustion gas stream without producing the desired
cooling effect. Values range from about 25 to 65
percent, depending upon coolant injection geom-
etry and on flow conditions.
The above equation assumes that a balance
exists between heat input and coolant tempera-
ture rise. The heat input is based upon the gas-
side heat-transfer coefficient hg and the differ-
ence between the adiabatic gas temperature at
the wall and the coolant film temperature. The
heat absorbed is proportional to the heat capac-
ity of the coolant film from initial to final tem-
perature values. Once equilibrium is reached, no
heat is transferred to the wall (adiabatic condi-
tion) and the chamber wall surface will have
achieved the film-coolant temperature corre-
sponding to the various axial locations. Accord-
ingly, the wall-surface temperature will range
axially from the value of initial coolant tempera-
ture to a maximum allowable design wall temper-
ature, at which point the next film-coolant injec-
tion station must be provided. It is the specific
aim of film-cooled thrust chamber design to
accomplish cooling with an optimum number of
coolant injection stations.
Figure 4-35 shows an experimental hydrogen
film-cooled thrust chamber. Cooling is provided
by four film-coolant injecting rings upstream, and
one downstream of the throat. Axial coolant
injection, in the direction of combustion gas
.. e..: 1 T ',j ..
_" "" ... oF.
DESIGN OF THRUST CHAMBERS AND OTHER COMBUSTION DEVICES
117
Figure 4-35.-Experimental hydrogen/oxygen, film-cooled thrust chamber.
flow. greatly improves film-cooling efficiency,
whereas normal injection results in the escape,
any benefit, of large portions of the
coolant into the combustion gas stream. In a
typical case, the film coolant flow was approxi-
mately 3 percent of the propellant.
Sample Calculation (4-5)
For the design of the hydrogen film-cooling
system for the thrust chamber of the A-3 stage
engine, the following data are given at the throat
section:
hg = 0.0011 Btu lin 2- sec-deg F
Taw =5240
0
R
Twg = 1900
0
R max
Teo =50
0
R
C
pve
= 3.6 Btu/lb-deg F average
Assuming a value of 0.5 for film-cooling
efficiency, determine the film-coolant weight
flow rate per unit area of cooled chamber surface
in the throat section.
Solution
Substitute the given data into equation (4-34):

0.0011
5240-1900 = - G
c
X36Xo.S
5240- 50 e
(
000061)
1.554 = e G
c
G - 0.00061 0.00061 = 0001392 lblin 2-sec
c-ln 1.554 0.439 .
To calculate the heat flux for a regenerative
cooling system, with added film cooling, a cor-
rected value (Taw) must be used in equation
(4-10) or (4-17). This corrected adiabatic wall
temperature can be determined experimentally
under the specific thrust chamber operation and
film-cooling conditions. As to the gas-side heat-
transfer coefficient, it was found that there is
practically no difference with and without film
cooling. Thus, the normal gas-side heat-transfer
coefficient hg can be used in equation (4-10).
Note that if a hydrocarbon fuel is used as the
film coolant, the effect of carbon deposition must
be taken into account (eq. 4-17).
Transpiration Cooling
Figure 4-36 shows the principle of transpira-
tion cooling. The coolant is introduced through
numerous drilled holes in the inner chamber wall.
In other designs, the wall may be made of porous
material. In both cases, the permeable chamber
inner liner is enclosed by an outer shell (similar
to fig. 4-28), forming a jacket from which the
coolant emerges. For adequate design, the total
coolant flow requirement and coolant weight flow
rate per unit area of cooled chamber wall (lb/in 2_
sec) must be determined and then implemented
by a practical method.
,
118
DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
Mhbatie ""a:..l ':'e=p., 'l'a!.' Best '!:ratlsre:r 1Dto \Jill
Cal! Side Fill: Coe!'!"ldetlt, he
Coolant EtW.k temp. f r co
Figure 4-36.-Transpiration cooling model.
Transpiration coolant flow requirements de-
termined from theoretical equations turn out to
be !?ignificantly lower than those for film cool-
ing. This is due to the more efficient coolant
distribution. The Rannie equation for transpira-
tion cooling can be used to calculate the theo-
retical coolant flow requirements:
where
Taw = adiabatic wall temperature of the gas,
deg R
T wg = gas-side wall temperature, deg R
T co = coolant bulk temperature, deg R (entering)
G c = transpiration coolant weight flow rate
per unit area of cooled chamber wall
surface, Ib
/
in
2
-sec
G g = combustion gas weight flow rate per unit
area of chamber cross section perpen-
dicular to flow, lblin 2 -sec
P rm = mean film Prandtl number
e = base of natural logarithms, 2.718
Reb = bulk combustion-gas Reynolds number
Test data from various transpiration-cooling
experiments have been in good agreement with
the Rannie equation. However, the equation pre-
dicts coolant flows slightly lower than those
required in experiments. It is recommended that
an efficiency factor of approximately 0.85 is
used for calculations.
The porous material u sed for the transpiration-
cooled chamber walls must be selected and di-
mensioned for correct hydraulic resistance to
render the required coolant flow rate per unit
surface area. It must also be able to withstand
the stresses caused by the pressure differential
between coolant and combustion gases, and
thermal stresses. These requirements impose
certain limitations on the selection of materials
and on construction methods. The mechanical
design of the coolant distribution system, there-
fore, is an important factor for successful appli-
cation of transpiration cooling.
Ablative Cooling
Ablatively cooled thrust chambers have many
advantages for upper-stage applications. They
are designed to meet accumulated duration re-
quirements varying from a few seconds to many
minutes. Most designs are limited to lower cham-
ber pressure applications, 300 psia or less.
When assisted by film cooling. or by throat in-
serts made from refractory materials, successful
firings have been made up to a chamber pressure
level of 1000 psia. In general, ablative chamber
construction is rugged, exterior wall tempera-
tures are held to a minimum and the cost is low.
Ablative cooling is accomplished by the
pyrolysis of resins contained in the chamber wall
material. The thrust chamber construction will
vary with mission requirements. As shown in
figure 4-37, chamber and nozzle are composed of
an ablative liner, a thin layer of insulation, and
a high-strength outer shell. The ablative liner
is fabricated from a phenolic-resin-impregnated
high-silica fabric which is wrapped in tape form
on a mandrel at optimum orientation. The thick-
ness is programed as a function of chamber sta-
tion to provide adequate strength, charthickness,
and minimum weight for a particular mission. A
wrap of oriented phenolic-impregnated asbestos
is placed on the outer (far) surface of the abla-
tive liner as an insulator.
The high-strength outer shell is composed of
layers of unidirectional glass cloth for longitudi-
nal strength, and of circumferential-wound glass
filaments for hoop strength. The glass wrap is
bonded with epoxy resin. Aluminum alloy and
DESIGN OF THRUST CHAMBERS AND OTHER COMBUSTION DEVICES 119
AIL4TIV! lINII!'RS
J
\
\
IJOIt 111.101. liON -
CClOUD t40ZILE !XT!tfiION
------L----
L INSULAnON AND
..,. ....
Figure 4-37. -Ablatively cooled thrust chamber.
stainless steel also are sometimes used for the
outer shell. The combined thermal resistance of
the ablative liner and of the insulation layer
protects the outer shell and keeps it at moderate
temperatures.
Figure 4-38 shows an ablative-cooled thrust
chamber fitted with a throat insert. Both 98 per-
cent tungsten/2-percent molybdenum alloy and
pyrolytic graphite have been successfully em-
ployed as insert materials. The tungsten-
molybdenum alloy has given the best results.
Although pyrolytic graphite has a much lower
density than tungsten alloy, and therefore has a
substantial weight advantage, it is vulnerable to
fracture from thermal shock, making design and
installation critical. The throat insert is in-
stalled with heavy graphite backups for better
structural results. Insert and backups are bonded
to the thrust chamber main ablative liner with
epoxy adhesives. These adhesives have per-
formed satisfactorily up to 500
0
F. Certain
ceramic materials, such as silicon carbide, also
have been used successfully as throat inserts in
space engine applications.
The design of an ablative thrust chamber for
a given mission depends on the accuracy of pre-
dicting the depth of char during exposure, and on
the soak-back temperature variation in the insu-
lation surrounding the charred portion of the
thrust chamber wall during and after the hot fir-
ing. Test data from hot firings with various
ablative thrust chambers indicate that the charring
process in the combustion chamber (including
throat), that is, the relation between mass pyro-
lyzed and heat absorbed, can be expressed by
the following equation:
I
,
TION ...1oICI 1'041 IIAOIArlOM _
IOHOING --, COOleD IIIOZZll!! UT8IWOM
\
\
!"UNCI! '0111 !
IMJ'I!CTQR \
/
LHMOHnlltf...cTl4
ounlll 5JoI!u..
Figure 4-38. -Ablatively cooled thrust chamber
with throat insert for high chamber pressure
applications.
a=c 2kt In (1+ RrRvCpi:aw-Td)\los
RrRvCpp \ I' J
where
[
(Pc)
ns
l
04
(4-36)
100J
a = char depth, in
c = correction factor based on experimental
data for the specific design at the
throat section, and o!l a nozzle stag-
nation pressure of 100 psia
Rr = weight fraction of resin content in the
ablative material
Rv = weight fraction of pyrolyzed resin
versus total resin content Rr
C p = heat capacity at constant press!Jre of
pyrolysis gases, Btu/lb-deg F
p = density of ablative material, lb/in 3
k =heat conductivity of char, Btu/sec-in
2
-
deg Flin
= thrust chamber firing duration, sec
Lp = latent heat of pyrolysis, Btu/lb
Taw = adiabatic wall temperature of the gas,
deg F
T d = decomposition temperature of resin,
deg F
(Pc)ns =nozzle stagnation chamber pressure,
psi a
Results predicted by equation (4-36) have
been compared with char depth data obtained
from firings of Refrasil-filled phenolic chambers.
They were found to agree very closely with the
experimental data. However, for areas down-
stream of the throat, char depths were found to
.----
120 DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
be somewhat greater than predicted. and when
using the equilibrium gas temperature. Tempera-
ture recovery in the boundary layer may be one
cause for the discrepancy. A modified equation
is used. therefore, to predict char depths in the
nozzle areas:
(4-37)
where
b = a constant depending upon the nature of the
ablative shield (to be determined experi-
mentally)
f = nozzle expansion area ratio at the investi-
gated section
e = base of natural logarithm s. 2.718
The char-rate analysis is characterized by
physical dimensions and the formation of a char
layer that progresses from the heated surface
toward the supporting wall. During the pyrolysis
of the resin. the formation of a hard carbona-
ceous surface of increasing thickness is vital
because it resists thermal and mechanical abla-
tion and chemical attack. At the charring inter-
face, which slowly travels away from the hot
chamber gases, a large amount of heat energy is
absorbed by pyrolysis; i.e., melting and vapori-
zation of the bonding material. As gaseous
pyrolysis products flow through and out of this
char layer, they control the heat flux to the walls
by their own endothermic decompositions, and by
migration into the boundary layer.
No gross dimensional change occurs due to
energy considerations throughout most of the
thrust chamber; however, mechanical erosion is
evidenced in some designs in the throat region,
due to the high prevailing shear stresses. At
chamber pressures below 150 psia, throat erosion
is generally reduced. Throat erosion rates vary
from 0.0005 to 0.00005 in/sec
The adaptation of ablative thrust chamber
technology to the special field of space engines
has been significantly advanced during recent
years. Approaches typical for this type of engine
will be discussed in chapter Xl.
Sample Calculation (4-6)
The following design data are given for the
ablatively cooled thrust chamber of the A-4 stage
engine:
C=1.05; Rr=0.3; Rv=0.41; Cp =O.38Btu/lb-
deg F; p=0.0611blin
3
; k=9.8x 10-
6
Btu/
in
2
-sec-deg Flin; L
p
=686 Btu/lb; Taw
=5060
o
R; Td=1460oR; b=0.0335
Determine the char thickness at the throat and
combustion chamber section, and in the nozzle
at station f = 5, after firing for the design dura-
tion of 410 seconds.
Solution
From table 3-5: (Pc)ns = 100 psia
Substitute this and given data into equation
(4-36). The char thickness at the throat and
combustion section results as:
r 2x9.8xl0-
6
x410
a = 1. 05 La. 3 x 0.41 x 0.38 x 0.061
x In (1 + 0.3 x 0.41 x 0 6 ~ ~ (5060-1460) )J0.5 x (1)04
=1.05x[2.82xln 1.245J.5=0.828 in
Char thickness at nozzle station f = 5, using
equation (4-37):
a = bt 0.5 e- 0C124U = 0.0335 x (410).5 x (2. 718)--Q0247X5
1
= 0.0335 x 20.248 x (2.7182)8.1
=0.599 in
Radiation Cooling
Cooling by radiation heat transfer is practical
only for thrust chamber nozzle extensions, where
pressure stresses are lowest. High metal-wall
temperatures are required to attain the heat
fluxes needed. Assuming negligible temperature
drop through the metal and coatings, if any, the
steady-state heat transfer for a radiation-cooled
nozzle, as shown schematically in figure 4-39,
can be expressed by the following correlation:
(4-38)
where
hgc = overall gaS-Side thermal conductance,
Btulin
2
-sec-deg R
DESIGN OF THRUST CHAMBERS AND OTHER COMBUSTION DEVICES 121
COMBUSTION
GASES
..
..
RADIATION
4
q : CT Twg
Figure 4-39.-Schematic of radiation cooling.
Taw = adiabatic wall temperature of the gas,
deg R
Twg = gas-side wall temperature = bulk wall tem-
perature, deg R
f = total emissivity of outer wall surface
a =Stefan-Boltzmann radiation heat transfer
constant, 0.3337 x 10-
14
Btu/in
2
-sec-
(deg R)4
The design approach to radiation cooling is
to determine a T wg value which will satisfy both
equation (4-38) and the structural capability of
the wall material used under operational condi-
tions.
Only alloys which possess short-time strength
in the temperature range of 2600 R to 3500 R
have been successfully applied to radiation cool-
ing. A molybdenum alloy containing 0.5 percent
titanium, and a 90 percent tantalum-10 percent
tungsten alloy appear to have sufficient short-
time strength for use at 3500 R. Because of the
low emissivity of molybdenum and also for re-
sistance against oxidation, a coating of MoSi 2 is
required on both sides of the metal. Titanium
alloys and other commercial alloys, such as
Haynes 25, have been operated successfully at
2600 R. If a temperature capability higher than
the working range of bare metals is required,
insulating coatings of ceramic materials on the
gas-side wall surface may be needed. Because
of their brittleness and coefficient of thermal
expansion relative to that of the alloys, experi-
enced judgment is advised before using these
coatings for a specific application.
Sample Calculation (4-7)
The follOwing design data are given for the
A-4 stage chamber nozzle extension at station of
area ratio = 8:
h
gc
=7.1 x lO-
5
Btulin
2
-sec-deg R; Taw=49000R
Assuming a total emissivity of 0.95 of outer
wall surface, determine the bulk temperature and
heat-radiated flux.
Solution
Substitute data into equation (4-37):
7.1 x 10-
5
(4900- Twg)
=0.95xO.3337x10-
14
x (Twg)4
(T wg)4 = 22.4 x 10
9
x (4900 - T wg)
T
wg
=2660 R
Heat flux = 7.1 x 10- 5 (4900 - 2660)
= 0.159 Btulin 2-sec
4.5 INJECTOR DESIGN
The function of an injector, which is located
in general, at the forward end of the combustion
chamber as shown in figures 4-1 and 4-2, is sim-
ilar to that of the carburetor of an internal com-
bustion engine. The injector introduces and
meters the propellant flow to the combustion
chamber, and atomizes and mixes the propellants
for satisfactory combustion.
Design Objectives
A great number of injectors have been devel-
oped and many details of successful injector
designs are now available. However, there still
are no hard-and-fast rules to assure a successful
design. In the past, most injectors were de-
signed by a trial-and-error approach, with the
help of previous test data. While good results
have eventually been obtained, it was usually at
the expense of large amounts of time and money.
A more rational approach toward the design of
injectors is through understanding and prediction
of the chemical and physical processes that are
encountered within the combustion chamber. and
using this information as a basis for initial in-
jector deSign. For a given propellant combina-
tion, the chemical reactions and the kinetics of
stream breakup, mixing, droplet formation, and
heat transfer should be studied and clearly un-
derstood, before the approach to the design of an
injector is established.
122
DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
There are numerous requirements to qualify a
given injector for operational use. The following
are the most important objectives for injector
design:
1. Combustion stability. -In combination with
a given combustion-chamber configuration and
for a given propellant combination, an injector
should give smooth combustion, during engine
start and stop transients as well as during
steady-state operation.
Depending upon the propellants and their
ignition characteristics, the arrival sequence of
oxidizer and fuel streams during start is of great
importance. Any accumulation of unburned pro-
pellants in the combustion chamber prior to igni-
tion must be prevented to avoid destructive
chamber-pressure surges. Similarly, during
engine shutoff, chamber overheating and burnout
may be prevented by maintaining a fuel-rich mix-
ture. Arrival sequences are best controlled by
propellant valve timing. Furthermore, minimum
feed-line and injector-manifold volumes between
propellant valves and injector face will mate-
rially improve propellant sequencing during start
and stop.
To prevent chamber-pressure fluctuations from
affecting the propellant flows and thus from in-
ducing combustion instability, sufficient pres-
sure drop through injector orifices must be main-
tained. Effective and even mixing of the
propellants will be achieved through the choice
of a suitable injector Impingement pattern. This
will help to minimize accumulation within the
combustion chamber of unburned propellants
which could cause local detonations and thus
trigger combustion instability. Under certain
conditions, combustion instabilities of the tan-
gential oscillation mode can be prevented by
isolating local detonations by partitioning the
injector face into several compartments, as
shown in figure 4-40.
2. Performance.-Combustion performance of
an injector is influenced by: propellant mass
distribution; local mixture ratios; degree of
mixing of injected propellants, in either the liq-
uid or the gaseous phase, or both; droplet atomi-
zation and vaporization; rate of heat input; and
chemical reaction rates. These are predominantly
a function of suitable manifolding and proper
selection of injector-hole patterns. The more
Figure 4-40. -Baffled injector.
DESIGN OF THRUST CHAMBERS AND OTHER COMBUSTION DEVICES 123
thorough the mixing and uniform the distribution
of the oxidizer relative to the fuel, produced by
the injector, the more rapidly will the combustion
products reach the equilibrium composition nec-
essary for optimum performance. Although turbu-
lence induced by the combustion probably con-
tributes a major portion of the energy required for
gas-phase mixing, thorough premixing 01 the
liquid propellants must be accomplished by the
injector if maximum performance is to be achieved.
Furthermore, reaction between certain speCific
propellant combinations such as hypergolic pro-
pellants cannot reliably be initiated and main-
tained without it, since the energy released by
liquid-phase reactions supplements the kinetic
energy available for the process of atomization
through combustion-gas evolution. In addition,
the heat release from liquid-phase reaction ac-
celerates the process of vaporization.
Experience has shown that for a given injec-
tion velocity, propellant-droplet size is reduced
with decreasing injector-orifice size. Smaller
droplet size, in turn, results in a higher overall
vaporization rate, as a function of increased
total droplet surface area. This is true whether
the heat of vaporization is supplied internally
via liquid phase reaction or externally by heat
transfer from the hot gaseous combustion prod-
ucts. Consequently, injector designs with the
largest practical number of injection elements
can be expected to be the most efficient ones in
a given combustion chamber volume.
3. Structural integrity.-An injector should be
able to withstand the maximum loads incurred
during all phases of engine operation. Sufficient
cooling must be provided to prevent the injector
face or any other portion from overheating.
4. Hydraulic qualities.-The holes or orifices
of the injector must be designed to effect pre-
determined pressure drops at specific flow rates,
and to atomize the propellants properly. A low
injector pressure drop is desirable from the
standpoint of overall engine-system performance.
However, minimum pressure drop is determined
from combustion-stability considerations.
5. Combustion chamber heat protection.-An
injector should be designed to avoid formation of
hot spots or streaks on the combustion chamber
wall. Complete mixing of the propellants will
prevent oxidizer-rich peak temperature zones
from forming, although this may not prevent
streaks of high mixture ratio (O/F) from occa-
sionally reaching the chamber wall. To offset
this, a special set of fuel holes is often provided
at the periphery of the injector, close to the
chamber wall. Excess fuel along the chamber
wall is thus provided which tends to lower the
OIF mixture ratio of any errant streak. It also
assists in cooling the chamber wall.
6. Special requirements. -Certain engine sys-
tems are required to operate at off-nominal con-
ditions, such as at lower thrust levels during
throttling, or other than nominal mixture ratios
as a result of propellant-utilization control. In
these cases, injectors must be capable of oper-
ating reliably under modified as well as rated
conditions.
Injector Configurations
A typical injector design construction and
propellant-distribution method is illustrated in
figure 4-2. Different distribution methods are
shown in figures 4-41 and 4-42. The injector in
figure 4-42 uses an integral faceplate. This plate
is secured to the main injector body by brazing it
at the periphery and at posts which are an inte-
gral part of the main body. A fuel compartment
is located immediately behind the faceplate, and
fed from an inlet passage. The oxidizer com-
partment is separated from the fuel by a parti-
tion. The fuel is injected through orifices
drilled in the faceplate, while the oxidizer is
injected through orifices drilled in the posts.
The injector construction for a typical liquid-
bipropellant gas generator is illustrated in figure
4-43. The copper injector body is secured to the
stainless-steel outer shell by brazing. The
oxidizer inlet forms an integral part of the injec-
tor body. Fuel is supplied through a manifold in
the outer shell. In this injector, 2 fuel streams
impinge on each oxidizer stream, producing a
total of 44 impingement pOints.
A variety of injector patterns have been de-
signed to satisfy the needs of various propellant
combinations. In most cases, for good mixing
the injected streams are made to impinge at a
predetermined pOint. The impingement point
should be as close to the injector face as heat-
transfer conditions permit. The arrangement in
which all impinging points are the same dis-
tance from the injector face is called uniplanar
4C;l1'i'.Uif
124 DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
.
:
.

i

. .
0" i
. ::. ..
... r
&i



g
i
i

i!
.
Ii
DESIGN OF THRUST CHAMBERS AND OTHER COMBUSTION DEVICES
125
GIMBAL BEARING
MOUNTING SURFACE
ASI
CAVITY
POSTS
INTEGRAL
FACE PLATE
FUEL INLET
PASSAGE
OXIDIZER INLET
Figure 4-42.-1ntegral face plate injector.
impingement. If two or more different impinging-
point distances are used (fig. 4-44), the arrange-
ment is called biplanar or multiplanar impingement.
Numerous tests have been conducted to deter-
mine impingement-angle and distance effects.
Large included angles will enhance stability,
but can result in some of the propellants splash-
ing back on the injector face, which can cause
burnouts. Close spacing of the impinging holes
in a pair has similar advantages and disadvan-
tages, as has increased spacing between pairs.
The satisfactory d e s i g ~ value for the included
angle is usually found to be between 20 and
45. The injector face can be further protected
against overheating by circulating the propel-
lants on the back side of the faceplate or by
introducing film coolant (propellant) on the
surface. Some of the impingement patterns used
are described below:
1. Showerhead (fig. 4-44a).-This pattern em-
ploys non impinging oxidizer and fuel streams
which emerge normal to the injector face. It
relies entirely on combustion chamber turbulence
for mixing. While being the simplest to fabri-
cate, the showerhead injector exhibits poor per-
formance in most applications, with the excep-
tion of certain cryogenic propellant combinations.
2. Doublet (fig. 4-44b).-In this design, oxi-
dizer and fuel jets are made to impinge in pairs.
Thus good liquid-phase mixing and atomization
is obtained. One of the disadvantages of this
doublet arrangement is that even if the injector
holes have been accurately drilled, the resultant
angle of momentum vector, or beta angle, {3, will
vary with mixture ratio, particularly if a large
impinging angle is used. This variation can
adversely affect combustion performance and
chamber-wall heat transfer. The doublet design
is frequently used in systems using liquid
oxygen.
3. Triplet (fig. 4-44c). - Two streams of one
propellant impinging symmetrically on one stream
of the other propellant will eliminate the change
of vector angle {3, as a result of mixture-ratio
variations. This arrangement also provides
intimate mixing. Application and propellant
combination will determine whether two oxidizer
jets will impinge on one fuel jet, or vice versa.
Injectors using this triplet pattern have given
high combustion performance. These injectors
have been widely used for various propellant
com binations.
4. Quintuplet (fig. 4-44d). -Four streams of
one propellant impinging on one stream of the
other propellant in a symmetrical quintuplet pat-
tern provide excellent mixing and performance.
This design has been applied for various propel-
lant combinations.
5. Self-impinging (fig. 4-45a).-This pattern,
also known as a like-on-like impingement, gener-
ally employs self-impinging pairs of fuel and
oxidizer. Mixing is accomplished in the combus-
tion chamber by volatilization of the propellants
and by turbulence. This design usually provides
good inherent combustion stability, at a moderate
performance level. Applications have been suc-
cessful for both cryogenic and storable hyper-
golic propellant combinations. A modification of
this design provides for secondary impingement
of the two propellants following self-impingement.
"
".
LM C(X)LIN{;
36 fUEL IIOL.:5
.0465" I>IA.
COPPFn INJI':CTQn
TOTAL CUARAIICE .004"
EACH SURFACE PLATED
WITH .0002" NICKEL
WITH .0008" GOLD
rURIIACE BRAZED
INJECTOR III MANIFOLD
DRILLl:D AT $> ANGLI:
44 HOLES
.oU' OIA.
FUEL II HOLES






_
UEL X
. .on" iliA. ZO AN!JI.II:
--l 1--. ZOO"
TYPICAL SPRAY PATTmN
LOX PURGE
. APPROX. ASSY. WT. e., LB!I.
HOLES .210" DlA.
4 .144" OJA
HOLES .1"''' DIA.


Figure 443. -Bipropellant gas generator injector.






0.0 00

.0
o 0
o 0
0
00
c;
0
0
o 0
0 ()
0
0 0 0
0
00 0
00 0
00 000 ()
0 0
0 0
o ()
0 000 0
0
0 0
0
0
0 0 0 0 0
0 0

0
0
0 0 0 0 0
0
0
0
0 000

'0 0
o 0
0 0 0
0
0
0 0 00 000 0
00
0
o 0 o 0
0 0
00 0 0
o 0 00 00 0 0
0 o
000 00
0





U"UI4."k 'IA'I"r,,'uN

-.-"1
". 4'-

rUEL MANIFOLDS








o
m
'"

z
o
'TI
r
D
S
o
-g
::0
o
-g
m
r

z
-i
::0
o
n
'"
m
-i
m
z
!
z
m
'"
DESIGN OF THRUST CHAMBERS AND OTHER COMBUSTION DEVICES III
O'(lOlln
IIIIAHIFo...O
",,,,,, ..
"&WOLD
(0) SHOWER HEAD
(c) TRIPLET
OXItMlDt , ..... ;-,
MA.NtFOl.O
'NJ(C'l'OIft


'A",

(O'OOUEILET
Figure 4-44. -Injector impinging patterns.
(c)"WG SLOT (d) SPl.ASH PlATE
Figure 4-45.-Injector impinging patterns.
6. Coaxial (fig. 4-45b).-This injector em-
ploys two concentric tubes for the two propel-
lants, which are injected coaxially.
7. Ring slot (fig. 4-45c).-The ring-slot inJec-
tor employs concentric pairs of annular slots
which eject the propellants as conical sheets.
The slots are so arranged that fuel and oxidizer
sheets impinge much in the same manner as in
the doublet-type injector.
8. Splash plates (fig. 4-45d).-This injector
is designed for good propellant mixing while the
propellants are still in the liquid state. For this
purpose, injected propellants are denected by
the splash plates. The plates will be kept cool
by the impinging liquid propellants which do not
ignite until they have left the plate.
9. Premix.-Figure 4-46 shows a typical
premix-type injector. Fuel and oxidizer are in-
jected radially into the premixing chamber. where
they are intimately mixed before entering the
combustion chamber by a gas jet introduced tan-
gentially at the chamber end. The length and
diameter of the premixing chamber with relation
to the mass now of propellants is critical, as is
the reaction time of the propellants.
10. Throttieable injector.-Certain require-
ments for space vehicle missions, such as orbit
corrections, rendezvous and docking maneuvers.
and lunar soft landings, demand engine systems
capable of thrust control. Figure 4-47a shows a
typical throttlable injector with variable injection-
slot areas. This is a very effective means of
controlling the propellant nows and injector
pressure drops at various engine thrust levels.
The addition of moving parts, however. causes
design complications. Another approach to a
thrcttlable injection system is the H aeration
method, as shown in figure 4-47b. I An inert gas
is introduced into the injector propellant mani-
fold for reduced thrust levels. Through variation
of the propellant/gas mixture ratio, the effective
density of the propellants can be varied over a
wide range to achieve any desired thrust level
without affecting combustion stability. The gas
can be supplied by the same source used to
pressurize the propellant tanks. This method
has increased the range of rocket-engine throt-
tling up to a ratio of 100 to 1.
Design Calculations
For the design of injectors, various parame-
ters. such as injector pressure drop, impingement
lAstronautics magazine. December 1962. pp. 36-37.
,
128
DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
and resultant vector angle relationship, and
structural loads, can be calculated with reason-
able accuracy.
Injection Velocity
The propellant (oxidizer or fuel) injection
velocity V (in/sec) can be calculated from the
basic relation:

Ap
(4-39)
where IV is the propellant weight flow rate, lb/
sec; A is the calculated injector orifice area,
in 2; and p is the propellant density, lb/in 3 .
GAS
MANIFOLO
GAS JET
TANGENTIAL
ORIFICE
FUEL
MANIFOLD
OXIDIZER
MANIFOLD

CHAM8ER
Figure 4-46.-Premixing type injector.
MOVABLE
PINTLE RING"
i
FUEL
MANIFOLD
(A) VARIABLE AREA INJECTOR
INJECTOR
FACE
Injection Pressure Drop
The injection pressure drop (lblin
2
) can
be calculated as
( 4-40)
where g is the gravitational constant, in/sec
2
;
Cd is a dimensionless discharge (velocity and
jet contraction) coefficient which is a function
of injector orifice configuration.
The value of this coefficient ranges from 0.5
to 0.92 and can be determined accurately by ex-
periment flow tests). Injector orifices
with well-rounded entrance and smooth bore give
high values of the discharge coefficient. For a
given injection velocity, a higher value of dis-
charge coefficient gives a lower injection pres-
sure drop. The rule-of-thumb design value for
injector pressure drop varies from 15 to 20 per-
cent of the chamber-nozzle stagnation pressure.
Resultant Angle of Impinging Streams
The angle between the thrust chamber axis
and the resultant momentum vector of a pair of
impinging streams is defined as the beta angle,
(3. By definition when the vector is directed
toward the chamber wall, the beta angle is posi-
tive, and when the vector is directed toward the
FUEL
THRUST
CHAMBER
THROTTLE.
VALVE ....
OXIDIZER I
THROTTLE
VALVE
INJECTOR
INERT GAS
CONTROL
VALVES
(B) AERATION THROTTLEABLE INJECTING SYSTEM
Figure 4-47. -Throttleable injecting methods.
DESIGN OF THRUST CHAMBERS AND OTHER COMBUSTION DEVICES 129
central axis of the thrust chamoer. the beta angle
is negative. With a hypergolic-type propellant
combination, a positive beta angle (2 to 5")
tends to increase the combustion performance by
causing recirculation and better mixing of the
liquid propellants along the chamber wall. How-
ever. in a cryogenic propellant combination.
where gaseous mixing is predominant. the com-
bustion performance will not be noticeably
affected by the beta angle. A negative beta
angle should be used in this case to avoid the
possibility of hot streaks on the chamber wall
caused by excessive heat transfer.
The beta angle may be readily calculated.
from the principle of conservation of momentum:
For the impinging streams shown in figure
4-48, al and a2 are the respective angles be-
tween the thrust chamber axis and the streams;
IV 1 and IV 2 are the weight flow rates; and Viand
V 2 are the injection velocities.
Injection Momentum Ratio
The injection momentum ratio can be defined
by the expression
~ THRUST
....... ----:--
CHAMBER
INJECTOR
FACE
CHAMBER
WALL
Figure 4-48. -Resultant angle of impinging
streams.
( 4-42)
where IVo and WI are weight flow rates, and V 0
and V I are injection velocities for oxidizer and
fuel. The injection momentum ratio is a useful
injector design parameter for the prediction of
combustion stability and performance of certain
propellant combinations. In the design of oxygen-
hydrogen injectors. the value of the momentum
ratio varies from 1.5 to 3.5 for liquid hydrogen
injection. and from 0.5 to 0.9 for gaseous hydro-
gen injection.
Structural Loads
The main loads to be considered in the struc-
tural design of injectors result from propellant
pressures behind the injector face, and in the
manifolds. During steady-state main-stage oper-
ation. the pressure load on the injector face is
equal to the injector pressure drop:
( 4-43)
The pressure load in the injector manifolds is
equal to the sum of the injector-end chamber
pressure and the injector pressure drop:
(4-44)
where PI is the pressure load on the injector
face. Pm is the pressure load in the manifold;
b.Pj is the injector pressure drop; and (Pc)i is
the injector-end chamber pressure.
During start transients. however. maximum
pressure loads on the injector may be substan-
tially higher than during steady state. When the
propellant valves are opened rapidly. propellants
rushing into the empty injector passages can
cause severe hydraulic ram. This pressure load
can be estimated empirically as
PI=P
m
=4P
p
(4-45)
where Pp is the propellant pressure at time of
valve opening.
Sam pIe Calculation (4-8)
Using data given in tables 3-2 and 3-3, deter-
mine the injector orifice sizes, injection veloci-
ties and momentum ratios for the A-1 and A-2
engines.
130
DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
Solution
A-l Engine
Thrust chamber propellant flow rates are 1941
lb/sec (oxidizer) and 827 lb/sec (fuel); propel-
lant densities are 71.38 lb/ft3 (oxidizer) and
50.45 lb/ft
3
(fuel); injector pressure drops are
200 psi for both propellants. Based on compo-
nent test results an injector orifice discharge
coefficient Cd of 0.75 is used for both sides.
Substituting it into equation (4-40), and convert-
ing feet to inches:
For the oxidizer side:
200 = 1 t 1 941 2
2 x 32.2 x 12 x 71.38 x \0.75 x Ao}
Total oxidizer injector orifice area: Ao = 32.4 in
2
For the fuel side:
200 = 1 ( 827 J 2
2 x 32.2 x12 x 50.45 x 0.75 x Ai}
Total fuel injector orifice area: A[= 16.4 in
2
An injector pattern of 700 pairs of self-
impinging streams is used for both oxidizer and
fuel. The following orifice areas and diameters
result:
- 32.4 - 0 0232' 2
a o -1400-' 10
d
o
=0.172 in
_16.4 -0 0117' 2
a[-1400-' 10
d[ =0.122 in
From equation (4-39), and using orifice areas
obtained above for available tlP and Cd. mean
injection velocities are determined:
For the oxidizer:
v 0 = A:; 0 = = 1452 in/sec or 121ft/sec
32.4 x 1728
For the fuel:
Vl= IV[ = 827 1728 in/secor 144 ft/sec
AIPI 16 4 x 50.45
. 1728
In support of the injector development engi-
neer, who may wish to compare with earlier test
data, the injection momentum ratio is determined,
using equation (4-42):
121x1941
Rm = 144 x 827 = 1.97
(b) A-2 Engine
From table 3-3, the propellant flow rates for
the thrust chamber are 285.2 lb/sec (oxidizer)
and 54.5 lb/ sec (fuel); the propellant denSities
are 71.38 lb/ft 3 (oxidizer) and 0.72lb/ft3 (fuel;
gaseous hydrogen at 180
0
R); the injector pres-
sure drops are 160 psi (oxidizer) and 60 psi
(fuel). With the coaxial injecting pattern (fig.
4-45b), experimental tests give a value of 0.62
for the oxidizer-side discharge coefficient and a
value of 0.9 for the fuel side. Substitute these
into equation (4-40).
For the oxidizer side:
160= 1 I 285.2 J 2
2 x 32.2 x 12 x 71.38 x \0.62 x Ao7
Total injector oxidizer orifice area: A
o
=6.43in
2
For the fuel side:
60 = 1 ( 54.5 )2
0.9xA[
Total injector fuel orifice area: A[= 13.79 in
2
Use a total of 300 coaxial elements for the
injector. The individual orifice areas and diam-
eters will be
do = 0.165
Use a tube inner wall thickness of 0.025 inch.
The diameters for the annular fuel orifice will be
d[, = do + 2 x 0.025= 0.215 in
DESIGN OF THRUST CHAMBERS AND OTHER COMBUSTION DEVICES 131
From equation (4-39), the injecting velocities
are
For the oxidizer:
VO=AWO 285.2 1076in/secor89.6ft/sec
oPO 6 41x 71.38
. 1728
For the fuel:
V t= Wt 54.5 9.500 in/sec or 790 ft/sec
Alpt 072
13.79 x 1728
From equation (4-42), the injection momentum
ratio
R
= 285.2 x 89.6 = 0 ~ 9 3
m 54.5 x 790 .0
Experimental Evaluation of Injector Designs
The design of an injector can be improved
through experimental testing. Three types of
tests are usually employed: hydrostatic pres-
sure, water flow, and hot firing. The hydrostatic
pressure tests are used to determine whether the
injector structure will withstand the required
pressure loads. The water flow tests are used
to evaluate the following design characteristics:
1. Effective injector pressure drop.-Tbe data
from the water-flow tests can be used to deter-
mine the orifice-discharge coefficient and to pre-
dict the injector pressure drop for the design
propellant, with corrections for density and
viscosity.
2. Injection pattern. -Injection pattern and
impingement can be observed, and faulty opera-
tion can be detected and corrected.
3. Atomization.-Water-flow tests at veloci-
ties corresponding to those employed in actual
service indicate the quality of atomization to be
expected with the actual propellants.
The true injector operational characteristics,
such as performance, combustion stability, and
heat-transfer characteristics for main-stage con-
ditions, as well as start-and-stop transients, can
only be fully evaluated by hot firing tests, in a
thrust chamber of representative design or a
"workhorse" eqUivalent.
It is often beneficial if during the hot-firing
tests of a given injector configuration, certain
operational parameters such as injector pressure
drops, thrust chamber shape, and L* can be
changed to determine the effects on performance
and stability. In such an experimental evalua-
tion program injector, orifice hole patterns can
be redrilled or holes plugged, until an optimum
configuration is obtained.
Heat-transfer characteristics are an important
factor when evaluating an injector design.
Temperature-measuring instruments embedded in
chamber walls and injector face are required to
measure heat-transfer rates and to detect local
hot spots. Instrumentation for measuring propel-
lant flows, chamber pressure, and combustion
vibration characteristics are Similarly important
for determining the true levels of injector per-
formance and stability.
4.6 GAS-GENERATING DEVICES
In liquid propellant rocket engine systems,
gases are required to power the propellant feed
systems and other subsystems. Bottled com-
pressed gases such as helium can be employed
for these purposes; however, the use of higher
temperature gases generated by suitable devices
gives much higher overall system performance.
Gases at temperatures ranging from 400
0
to
1000
0
F have been generated for pressurizing
propellant tanks. Gases in the range from 1200
0
to 1700
0
F are used to drive gas turbines for
pump-fed systems. Where ever possible, the
engine system primary propellants are used for
gas-generating purposes in the interests of over-
all system simpliCity. However, for certain
applications, such as for a pressurized gas-fed
system or for starting a turbopump-fed system,
high-pressure propellants other than those tapped
off the primary system are required to supply the
gas-generating systems.
Most operational engine systems use special
devices for gas generation. However, tapping
hot gases from the main chamber has shown
promise for certain applications. The following
is a list of design objectives for operational gas
generators:
(1) Ability to produce gases safely, with re-
quired properties (temperature, pressure,
132 DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
nonexplosive) in a compact unit, at the
required now rate.
(2) Ability to start and stop smoothly, with-
out abrupt temperature surges, pressure
oscillations, or overnow of unburned
propellants.
(3) Ability to operate over a wide range of
propellant now rates and (in the case of
bipropellants) mixture ratios, and to
respond closely to the control system.
(4) Ability to maintain safe shutdown without
complicated purging and draining systems.
(5) Ability to restart safely (restartable
engine systems only).
Additional design requirements depend on the
particular engine system involved.
Gas generators can be classified according to
the propellants employed:
(1) Solid propellant systems
(2) Liquid monopropellant systems
(3) Liquid bipropellant systems.
Solid Propellant Gas Generators
Solid propellant gas generators are applied to
liquid propellant engine systems for limited-
duration applications only, such as for turbine
spinners for engine start, or as pressurants for
short-duration pressure-fed systems. The tem-
perature of the gases generated by solid propel-
lants is generally in excess of 2000 F and is
not suitable for un cooled components over ex-
tended durations. Diluents can be used with a
loss of basic simplicity.
Figure 4-49 shows the typical design of an
operational solid-propellant gas generator used
to supply power to the turbine for engine start.
It is built in the form of a cartridge that bolts to
a nange at the liquid bipropellant gas generator
(shown in fig. 4-51). The particular cartridge
shown is a disposable type that cannot be re-
loaded and reused. Upon an engine start signal,
two initiators or igniters set off solid propellant
combustion.
Twenty milliseconds after start, the solid
grain produces a gas-now rate of approximately
4.68 pounds per second. It will maintain this
now rate for approximately 1.0 second. This
gas generator operates at 2500 F and at a cham-
ber pressure of 1000 psia. The product gas
renders an approximate characteristic velocity of
4000 ft/sec. A burst diaphragm located just
upstream of the gas generator orifice is used to
seal the unit during storage. It is ruptured by
the increased gas pressure at start. Body and
end cap are made of 4130 steel.
IGNITER
PELLETS
3S GRAMS
TAPE MIL-T-6ft4I
'" POLY BAG
CAP
, PLACES ON TABS
'OUTER GRAIN ..
CENTER GRAIN': ..
INNER GRA'"
. . ..... : .:::
.. '., '.
" .. ' ..
ALUM. CUP
11-0.120"
HOUS
Figure 4-49.-DisposabJe solid propellant gas generator (SPGG).
DESIGN OF THRUST CHAMBERS AND OTHER COMBUSTION DEVICES 133
Solid propellants burn uniformly at the ex-
posed surfaces at a rate which is primarily a
function of the temperature and the pressure of
the surrounding gases. The correlation for a
given propellant can be expressed as
(4-46)
where
R = propellant linear burning rate. inl sec
k 1 = constant representing the linear burning
rate of a given propellant. at a given
initial temperature and a chamber pres-
sure of 1000 psia
Pc = chamber pressure. psia
n = constant allowing for the sensitivity of the
propellant burning rate to changes in
pressure. at a given temperature.
The weight flow rate of a solid propellant gas
generator can be calculated by
(4-47)
where
IV g = weight flow rate through gas generator.
lb/sec
Ab = burning area. in 2
p P = propellant density. lb/in
2
It can be seen that a solid propellant gas gener-
ator of constant flow requires a propellant grain
design which assures consta,nt burning area.
The correlation between pc. A b. and gas gener-
ator orifice area can be expressed as
(4-48)
where
Ao=gas generator orifice area. in
2
k2 = constant characteristic for a given propel-
lant at a given temperature.
Liquid Monopropellant Gas Generators
Monopropellants such as hydrogen peroxide
(H
2
0
2
) and hydrazine (N
2
H
4
) have been used as
gas generants in many applications. They per-
mit a simple generator system and do not require
mixture-ratio adjustments. The advantage of
these systems is that they are relatively easy to
control and that the gases are generated at pre-
dictable temperatures. However. unless the
mono propellant is also employed as one of the
engine system main propellants. the monopropel-
lant gas generator system introduces a third
propellant often requiring special handling and
tankage.
Figure 4-50 shows the schematic of a typical
monopropellant gas generator using 90 percent
hydrogen peroxide. The catalytic screen pack or
bed consists of alternate layers of stainless-
steel mesh and silver-plated brass wire screens
secured by perforated end plates or grids. which
are applied with a preload of approximately 800-
1000 pounds per square inch of bed cross-
sectional area. The length of the catalytic bed
generally ranges from 2 to 3 inches. The allow-
able design throughput. that is. propellant flow
rate per unit cross-sectional area of catalytic
bed. is about 0.4 Ib/in
2
-sec. The propellant
pressure drop across the bed can be approxi-
mated by
where
t.Pb
( 4-49)
= pressure drop through the catalytic
bed. psi
= throughput. lb/in 2-sec
= gas generator chamber pressure at the
end of the catalytic bed. psia
= accumulated running time. sec
C 1. C 2 = design constants determined experi-
mentally for a given bed configuration.
II I \ ,\
Figure 4-50.-Schematic of monopropellant gas
generator.
134
DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
Sample Calculation (4-9)
A hydrogen peroxide monopropellant gas gen-
erator attached directly to the inlet flange of a
turbine has the following design data: Turbine
inlet pressure, 340 psia; total turbine nozzle
throat area, 0.776 in 2. Assume a c* value of
3080 ft/sec for 90 percent H
2
0
2
and catalytic
bed design constants C 1 = 7.2 x 10
4
and C
2
= 0.021
psi/sec. Determine the gas generator propellant
flow rate W g, catalytic bed area, and pressure
drop after 480 seconds of accumulated running
time.
Solution
Gas generator pressure = turbine inlet pressure
= 340 psia. Equivalent gas generator throat
area = total turbine nozzle throat area = 0.776 in 2.
Substitute into equation (1-32):
3080= 340 x 0.776 x 32.2
Wg
Propellant flow rate Wg=2.761b/sec
Use a design value for catalytic bed throughput,
G b, of 0.4 lblin 2-sec
Catalytic bed area = 7 ; . ~ 6 = 6.9 in 2
Substitute given data into equation (4-49):
7.2 x 10
4
x (0.4)195
Bed pressure drop t.Pb = 09
(340) .
+ 0.021 x 480
=63.5+10=73.5 psi
Liquid Bipropellant Gas Generators
This gas-generating system is used more
widely in liquid rocket engine systems than any
other, since it makes use of the primary rocket
engine propellants. Bipropellant gas generators
react in the same manner as the main thrust
chamber, except that the oxidizer-fuel-mixture
ratio will be adjusted to yield the desired gas
temperatures and chemical properties. The
gases generated may be either fuel rich or oxi-
dizer rich, depending on the propellant combina-
tion. Oxidizer-rich gases tend to accelerate
erosion of structural members, while fuel-rich
gases continue to burn with ambient air after
discharge, requiring special measures to prevent
damage.
Figure 4-51 and table 4-2 describe a typical
liquid bipropellant gas generator system. It is
designed to produce hot gases using primary
engine propellants (L0
2
/RP-1) for driving the
turbine of a pump-fed system. The control sys-
tem consists of two normally closed, linked
poppet valves that control the flow of propellants
to the gas generator injector. The valve assem-
bly includes an oxidizer strainer, oxidizer pop-
pet, fuel poppet. timing orifice, actuating piston,
and valve main body. The valve assembly is
actuated by gas pressure which forces the piston
on the fuel side down to open the fuel poppet. A
yoke integral with the piston actuates the oxi-
dizer poppet. The valve design, through a com-
bination of manifold volumes and LOX poppet
adjustment, effects a slight oxidizer lead to
prevent detonations, and a fuel-rich cutoff to
eliminate the possibility of turbine burning. The
propellants flow through the poppets to the in-
jector and into the gas generator combustor, are
mixed, and burn within the inner chamber and
combustor body. Ignition of the propellants is
accomplished by two pyrotechnic igniters. A
gas duct with two flanges opposite to each other
is located at the end of the combustor body.
These flanges connect with the solid propellant
gas generator turbine spinner and the turbine
inlet.
The basic design parameters for bipropellant
gas generators are similar to those for thrust
chambers. The total throat area of the turbine
nozzles may be considered to be the equivalent
throat area of the gas generator combustion
chamber. In calculating combustion chamber
characteristic length L*, the volume between
injector and turbine nozzle throats is used, with
a correction factor allowing for the specific
design configuration. Because of temperature
limitations of the turbine construction materials,
gas generators are rarely operated at gas temper-
atures higher than 1800" F. In most designs no
cooling is provided for the combustion-chamber
wall and other surfaces exposed to the hot
gases.
DESIGN OF THRUST CHAMBERS AND OTHER COMBUSTION DEVICES 135
Figure 4-51. -Liquid bipropellant gas generator.
The maximum available energy per pound of
gas generator propellants is obtained when the
products of combustion are expanded isentrop-
ically through a supersonic nozzle to ambient
pressure. This quantity of energy is termed the
available energy content t'iHt, and is expressed
by the correlations in equations (6-16), (6-17),
and (6-18).
136
DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
TABLE 4-2.-0perating Characteristics of a
Typical Liquid Bipropellant Gas Generator
System, as Shown in Fig. 4-51
OXidizer.
Fuel. ..
Total propellant now rate
O:F mIXture ratio
Oxidizer now rate ...
Fuel !low rate ..
GG chamber pressure (injector end)
GG chamber temperature.
Oxidizer-side pressure drop of line, valve,
and inJector. . . . . . . . . . .. , ..
OXidizer side pressure drop across orifice.
GG oxidizer supply line takeoff pressure
(total at mam OXidizer pump discharge) ..
Fuel side pressure drop of line, valve,
and inJector ..
Fuel sIde pressure drop across OrifICe
GG fuel supply line takeoff pressure (total
at mam fuel pump discharge ..
Thrust Chamber Gas Tapoff Systems
Liquid oxygen
RP-1
1734 lbisec
0.342
4.42 Ib/sec
12.92 Ib'sec
612.1 psia
1200' F
114 psi
121 psi
846 psia
216 psi
80 psi
907 psi
With these systems, combustion product gases
are bled from the main thrust chamber and ducted
to the turbine, where they are used as the work-
ing fluid, This arrangement eliminates the need
for a separate gas generator system and contrib-
utes significantly to the simplicity and its relia-
bility potential of the engine system, The tech-
nique has been successfully developed by North
American Aviation's Rocketdyne Division and
appears to be particularly promiSing for hydrogen
engines requiring throttling, Figure 4-52 shows
a schematic of a tapoff system,
In a separate gas generator system, gases are
produced and "tailormade" for turbine-power
purposes only, with the benefit of relatively high
liquid supply pressures, In a tapoff system, the
bulk of the extremely hot gases of the main
chamber would not be suitable as the turbine
drive fluid, in view of the limitation of the tur-
bine construction materials, It has been suc-
cessfully demonstrated, however, that by with-
drawing chamber gases from the boundary zones
only, and through proper shaping and location of
the bleedports, any desired turbine inlet temper-
ature (usually less than 1700 F) can be reliably
and repeatably produced, for a given thrust cham-
ber and injector design, The bleed gases thus
withdrawn have been found to possess thermo-
dynamiC properties comparable to the products of
a liquid bipropellant gas generator utilizing the
OXIDIZER
- 0=l
fUEL PUtf.P
! I
TURBIHE -.. I ,; J /
.
,:.S I
,-p /
t. L HOT G"S
OOCT \ /
U
Figure 4-52, -Schematic diagram of thrust chamber
gas tapo([ system.
same primary propeUan ts, However, a tapoff
turbine must be designed to operate at an inlet
pressure lower than the thrust chamber pressure,
Furthermore, a tapoff engine system will require
some sort of simple starting device, such as gas
spinner. With the aid of a hot gas regulating
valve placed at the turbine inlet, hydrogen tapoff
engine systems have been successfully throttled
to thrust ratios of 10:1 or better.
4.7 IGNITION DEVICES
The initiation of the release of the chemical
energy stored in liquid rocket propellants is
accomplished by a number of methods from which
the engine designer will select the most suitable
for a particular system:
(1) Igniters
(2) Hypergolic ignition
(3) Catalysts
All of these methods have been in use for
both thrust chambers and gas generators, Note
that in liquid propellant rocketry, gas generators
have been used not only for turbine power but for
propellant-tank pressurization as well. The
selection of the preferred ignition method de-
pends on the chosen type of propellants, whether
bipropellants or mono propellants are used. on
the size of the combustion chambers, the heat
release per unit time required in relation to the
amount of propellants entering, and on a number
of other considerations which will be discussed,
All ignition methods, particularly those for
bipropellant syst.ems, have one overriding re-
quirement in common: minimum ignition delay,
DES!GN OF THRUST OiAMBERS AND OTHER COMBUSTION DEVI CES 137
If the propellants entering the combustion cham-
ber are not promptly ignited. explosive mixtures
can form and detonate with damaging results.
The assurance of faultless . gnition is dependent
on the selection of the ignition method. the
quality of design. and on adequate heat release.
Igniters
These are defined as devices which release
heat and thus initiate ti e reaction of the main
propellants which subsequently sustains itself.
Igniters derive their power from an outside
source or from a limited amount of energy stored
as solid propellants within themselves. Follow-
ing ignition. igniters do not participate further in
the' combustion process. Some of the principal
igniter types are discussed below.
Pyrotechnic Igniters
These are literally slow-burning fire crackers.
somewhat modified for rocket engine application.
Burning time is in the range of a few seconds.
For thrust chamber use they can be mounted to
the injector face or inserted f.om below at the
end of a wooden or plastic stick (fig. 4-2). For
better heat distribution, multiple units firing ill
different directions. have been used. as a rule
radially outward from the center across the in-
jbctor face. In other deSigns they have been
mounted to pinwheels, achieving distribution
through rapid rotation.
Although pyrotechmc. tgniters are used, they
should be considered obsolescent. To achieve
adequate heat release (or modem large engines,
their size becomes impractIcal. Also, ejection
of their inert parts can cause damage to the
delicate thin-walled tubes 0 modem chamber
walls. Under cryogenic conditions, they have
exhibited a teudency to cause ignition delays,
complete duds, or explosive popping.
Pyrotechnic igniters for gas generators and
small thrust chambers have been mounted in
recesses as screw-in-type plugs. (See figs. 4-53
and 4-54.) The igniters e initiated by electri-
cally triggered oC whicb there are a
variety ot t ypes in use. The need to CODDect
wires tr, tbe pyrotech.nic igniters is aootber in-
CODveDisnce. cbeckout of tbe
iDtecrity aDd readiDau of pyroteCbDic ipitel'll is
\
. difficult. It is one of the inherent shortcomings
of solid propellants that they cannot be switched
on briefly for checkout and tben stopped again.
4-$3. -Radudly ouward tJr1q pyrot.eclWc
.,11M1 eellM 01 bJJ&oI --ua,.
138 DESIGiII OF LIQUID PROPELLANT ROCKET ENGINES
Figure 4-54. -Gas generator igniter with built-in
fusible link.
Assurance of their thereror.e. is by
statistical and sampling methods. For opera-
tional application. additional safety margins are
aecured by redundancy. However. this tends to
increase bulkiness further. Also. pyrotechnic
igniters are unsuitable for repeated starts.
EDgine manufacturers alave bee_tl procuring
pyrotecbnic \gniters from sources specializing in
this field. Typical eDIIIPles of selected paralD-
ecera U8 follows:
For a l&OOOG-powad tbruat, LOXlRP-lear,iDe:
Main chGmber: centrally mounted unit;
propellant weight, 33 gms; 4.5 sec
burning time; heat release, 45 Btu/sec;
perchlorate-type fuel. Electrically initi-
ated. (Later models of this engine used
hypergolic slugs.)
Gas generator: Pyrotechnic: propellant
weight, 2 gms; 8 sec buw:ning time; heat
release. 1. 6 Btu/oec; perchlorate-type
fuel.
Vernip.r engines (earlier models only):
Similar to gas generator units.
It is vital that of the specifications
fall within a stated band. "" or instance. it can
be specified that in a famPLy of samples no
igniter will fire at currents below 1 amp, and
tbai all must fire below 4 amp. The first condi-
tion concerns the avoidance of accidental firing
due to stray currents introduced by extraneous
RF signals or other sources; ,he latter is simply
a reliability requirement . Similarly, minimum
and maximum burning times. i. 6., tolerances of
the nominal times, will i;9 dictated by start-
sequence conditions.
In operational designs, the pyrotechnic i" Jnit-
ers rrequently do not ignite the main propellants
directly, but ignite a pilot fiamo fed by a small
portion of the main fuel. The pilot name then
ignites the main propellants.
HypergoUc Igniters
The term "hypergolic" was coined by the
German chemist Noeggerath about 1942 and is
composed of elements of the Greek words for
"high energy liquid. II '!be term now denotes a
bipropellant combination ignites spontane-
ously when the two meet . Such a
system was used as a..a ignition source for the
earlier German A-4 (.mgine (later called V-2
engine). utilizing bydrazine-hydrate (N
2
H ... H
2
0)
and 80 percent hydrogen peroxide (H
2
0
2
) . Thin
tubes supported by a wooden stick were im;e.ted
into the t.hrust ctJ8lllber from below. Upon an
"ign tion" signa'!. a ground-mounted supply unit.
including remotf!ly operated valves. ted the two
compooents to the injector elements where
tbey burned wf,tb a spontaneously igniting bot
name. Thill method may also have been used
,.porart1y aD otber system". However. its
cla ........ 'tbe tteqaeo, clngiDi of feed linea.
DESIGN OF THRUST CHAMBERS AND OTHER COMBUSTION DEVICES 139
and the need to eject a cor.diderable amount of
inert solid materi21 made it undesirable. Also,
adaptation to repeated starts would be complex
and would require vehicle mounting, thus adding
inert flight weight.
HypergoUc Slugs
A mora elegant way of using the hypergolic
effect for main-propE'llant ignition is through use
of a hypergolic slug. In this design a small
amount fluid is used which is hypergolic with
one of the main propellants but not with the
other. The fluid is stored in a cylindrical car-
tridge which has burst diaphragms at either end.
The cartridge, in tum, is loaded into a housing
which is part of a bypass line paralleling a high-
pressure main propellant feed line (fig. 4-55). If
a fluid is chosen which is hypergolic with the
oxidizer but neutral to the fuel, it is installed in
the fuel system, and vice versa. The former
type is the more common one. Here a fuel by-
pass line feeds an injection element in the cen-
ter of the injector, or a set of elements evenly
distributed over the injector face. When the
pumps start and outlet pressures rise, the oxi-
dizer valve is opened. As pressures rise further,
the burst diaphragms in the hypergolic-fluid con-
tainer rupture and the fluid meets w1.th the oxi-
dizer in the chamber, igniting spontaneously.
The fuel following the slug sustains the ignition
name. The main fuel vai ! then opened and
all parameters reach main-stage level. Since the
igniter elements carry fuel fed from the main
source following ignition, they continue to par-
tiCipate> in the combustion, undistingu' shed from
the remainder of the main injector. (See scbe
matic, fig. 2-13.)
. ".,. 4-i1. -H1Pf1r101 ,.,., CIJICl1qe ud ""'"
Typical of fluid hypergolic with oxygen is
t"ie>t.hyl alurrunum. For optimum behavior, i.e.,
fm ,oinimum ignition delays and avoidance of
undbsirable deposits in fuel and sensing lines,
optilnum mixtures of the two have been success-
fully established experime:ltally. In a typical
application, an amount of 6 cubic inches has
been found adequate, although 9 cubic inches
are actually used for maximum safety ma.rgin. By
comparison, a 1.5-million-pound thrust engine
uses 35 cubic inches. The hypergolic slug
method, first explored at the German Peenemtinde
installation, is well developed and bas found
wide application. Limiteci design and develop-
ment work has been done more recently toward
repeated-start units, using a device resembling
an automobile brake master cylinder witil its
replenishing features. However, tbe bypergolic
slug is truly a single-start device. Because of
relative bulkiness, .the bypergolic slug is not
recommended for small units sucb as gas gener-
ators , where pyrotechnic igniters or spark plugs
are preferred.
Spark Plugs
S(lark plugs and their accessories have been
developed to high levels of efficiency and relia-
bility for liquid rocket engine use. They are
eminently suitable for repeated starts. For
direct ignition, however, tbey are confined to
relatively small combustion (Sef> fig.
4-56.) In a typical 200000-pound tbrust el(lgine,
tbe gas-generator spark plugs fire at the ra.te of
50 sparks per second, releaSing approximately
1/10 joule per spark. This corresponds to 5
joules/sec or 0.005 Btu/sec per plug. The
etriciency of spark generation from tbe electrical
4-56. -Spart aue.bly. At
SCIft'-ia .,.,t plu,; at leit, cable coaectGr.
140 DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
storage device is approximately 20 percent. The
overall efficiency of the spark ignition system is
approximately 10 percent. Thus, 100 watts of
electrical power is required. If a 28- V source is
lIsed, the nominal current is 4 amp when two
redundant systems are used in parallel. A Typi-
cal voltage at the spark plug is 15000 V.
Augmented Spark Igniters
The limitation of direct spark ignition to
small units has led to the design and develop-
ment of augmented spark igniters (ASI). In this
design a spark plug, similar to the above, fires
into a small chamber about the size of a gas
generator. A small amount of the main propel-
lants is fed into this chamber where they ignite.
The hot flame generated in turn ignites the main
propellants. Figure 4-57 shows the principles of
an earlier augmented spark igniter design. Per-
manently located at the injector end of the thrust
chamber, this igniter directs its products of com-
bustion across the face of the main injector. It
is film cooled by the tangential injection of fuel.
The oxidizer is injected through two copper
tubes which impinge at the centerline of the
igniter, resulting in a hot-core type of combus-
tion. The igniter is made of 4130 steel and has
a convergent, throatless nozzle. This results in
a wide dispersion angle of gases emanating from
the nozzle. The igniter is capable of an unlim-
ited number of starts. since the spark plug is so
located that the combustion does not seriously
affect its life. The igniter has proven operable
over a wide range of mixture-ratio and pressure
conditions. It continues to operate throughout
Figure 4-57. -Augmented spark igniter.
main stage. which prevents any of the main
chamber gases from backing up into the igniter.
Approximately one-half percent of the main pro-
pellant flow rates has been found adequate for
the augmented spark igniters.
Special Designs
For the engine used in the Saturn S-I booster.
gas-generator ignition by the solid-propellant
turbine spinner has been successfully developed.
The turbine spinner will be discussed in a later
chapter.
Hypergolic Main Propellant Ignition
In preceding paragraphs it was learned that
hypergolic fluids are being used as ignition
sources for main propellant combinations which
by themselves are not hypergolic (i.e., "anergol"
propellants). Hypergolic liquid main propellants
have attracted attention since the early days of
modern rocketry. Their use permits a substantial
simplification of the engine system through elim-
ination of the entire ignition system. leaving
entry timing the only functional requirement.
This gain however. is not entirely without penal-
ties. The practical hypergolic propellant com-
binations have a somewhat limited specific im-
pulse. Furthermore. some are highly corrosive
and/or pose handling and storage problems which
the engine designer has to consider_
A number of hypergolic main propellant com-
binations have been in successful operational
use for many years. During World War II. several
guided-missile systems using hypergolic propel-
lants, were under development in Germany. Hydra-
zine hydrate (N
2
H
4
H
2
0) and high-percentage
hydrogen peroxide were usen in certain Messer-
schmitt rocket fighters and the antiaircraft mis-
sile Enzian. Amines and nitric acid were used
for the AA rocket Schmetterling. Optolines and
nitric acid with sulfuric acid additives (approxi-
mately 10 percent were applied in the Peenemiinde
developments of the antiaircraft rocket Wasserfall
(17000-pound thrust) and the small 1300-pound
Taifun. "Optoline" was a generic term for vari-
ous mixtures of aniline. hydrocarbons. and other
substances.
In the United States. several propulsion sys-
tems utilizing hypergolic propellants have been
developed.
DESIGN OF THRUn CHAMBERS AND OTHER COMBUSTION DEVICES 141
To make available to the high performing but
anergol propellant combinations the simplicity of
hypergolic behavior, the effectiveness of small
amounts of additives ("sweeteners") has been
successfully demonstrated.
Catalysts
In a general sense, catalysts are not igniters
but initiators and sustainers of reactions, which
themselves remain unchanged during these reac-
tions. In rocketry, catalysts have been used
predominantly to initiate and sustain the com-
position of monopropellants ("Monergols "), nota-
bly that of hydrogen peroxide. Several opera-
tional or near-operational systems existed during
World War II, such as the earlier Messerschmitt
Me-163 which used hydrogen peroxide with
potassium permanganate solution as the catalyst.
Probably the most widely used application of
this principle during that period was for the tur-
bine steam-generating system of the German A-4
(later called V-2) ballistic missile, which em-
ployed 80 percent hydrogen peroxide with either
potassium permanganate or sodium permanganate
as catalysts.
Because of the need for relati vely elaborate
timine, valving, and interlocking devices, the
use of liquid catalysts was soon found to be
cumbersome and undesirable. Application of
solid catalysts, therefore, as they were being
used for underwater torpedoes, led to design and
development work toward use of these systems
for rocket application. They never reached
maturity for the German World War II systems,
but were perfected after the war by the British
and to a limited degree by the United States.
The Redstone rocket steam plant, using solid
catalysts, has consistently and successfully
operated in many flights, among them the first
U. S. manned rocket flight by Commander Shepard.
Another successful development are the AR air-
plane superperformance rockets. Analogous to
certain British systems', the AR systems decom-
pose hydrogen peroxide fed through a solid cata-
lyst bed consisting of impregnated wire screens.
Since the specific impulse of decomposed hydro-
gen peroxide alone is low (below 200, depending
on concentration and design parameters), RP ftiel
is injected below the decomposition chamber.
Because of the sufficiently high temperature of
MAIN
FUEL
VALVE
OXIDIZER PUMP
(H,OZI
THRUST
CONTROL
VALVE
MAIN
OXIDIZER
VALVE
Figure 4-58. -Schematic of a Rocketdyne AR-l
superperformance rocket engine.
the decomposition gases (1400 F), the RP
ignites and burns spontaneously with the free
oxygen of the decomposed H
2
0
2
. (See fig. 4-58.)
In this manner, the solid catalyst indirectly
serves as an ignition system. While the specific
impulse with RP afterburner is still moderate
(approximately 245 seconds for the AR), these
systems offer great versatility, storability, and
extreme simplicity, including throttling to low
levels and restartability.
More recently, it has been successfully dem-
onstrated that catalytic operation offers a simple
alternative to augmented spark ignition for hydro-
gen systems.
In a process patented for Engelhard Indus-
tries, Newark, N. J., a gaseous mixture of oxygen
and hydrogen is fed through a catalyst bed of
palladium-impregnated alumina (Al
2
0 3) pellets,
by which the mixture is ignited. Installation of
this igniter is similar to an augmented spark
igniter.
Ignition Detection
The reader familiar with the news stories
about rocket launchings over the past years is
well aware of the consequences of rocket stages
failing to ignite: loss of mission. This is
,.
.... ....... .
.. ...... ... ... .
.. ..
r
142
DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
equally true for liquid- and for solid-propellant
systems. With the former, however. an additional
hazard exists in case of ignition failure: that of
accumulation of explosive propellant mixtures
which can be accidentally set off with cata-
strophic consequences. This consideration has
always been a concern with unmanned vehicles,
but has become even more important for manned
ones.
Recognition of these potential dangers has
prompted extensive investigation of means to
detect reliably absence or presence of ignition
in liquid propellant rocket engines. Only upon
an "ignition OK" Signal should the engine-start
sequence be permitted to proceed. This refers
mainly to the thrust chamber. For gas gener-
ators, redundancy appears to be adequate pro-
tection for most applications. Desirable detec-
tion systems must judge ignition both qualitatively
(absence or presence) and quantitatively (ade-
quate heat release). Not all methods are equally
good in both respects.
In some form or another, the engine designer
will have to provide means for ignition detection.
A survey follows of several which have found
operational application.
Visual Detection
For the German A-4 (V-2) and the early U. S.
Redstone missiles, visual observation by the
test conductor was used. Man in this case was
the interlocking device and would initiate the
next sequence step only if, in his judgment,
ignition was adequate. This simple procedure
was satisfactory because these early systems
employed a prestage, during which the main pro-
pellants were admitted under tank head only.
The resulting relatively low flow rates were then
increased by starting the turbo pump upon a
"prestage OK" signal.
With the increasing size of modern rocket
engines, visual observation became unreliable.
The type of installation of these systems in
static firing stands and on launch tables made
direct observation difficult. Furthermore, im-
proved igniters, developed to keep the ignition
flame concentrated where it should be, i.e., at
the injector face, resulted in little or no visible
fire emerging at the chamber exit. With the dis-
appearance of the prestage step, the visual
problem increased because of the large amount
of oxidizer present in full flow ignitions which
shrouds the ignition flame. Thus, means had to
be found to detect ignition by other means.
Optical Detection
Ground-mounted optical devices can be moved
up close to the chamber exit. A number of types
have been investigated, such as simple light or
infrared-sensitive cells. They were found, how-
ever, to be subject to the limitations mentioned
for human observers. It is possible to mount the
optical devices into the chamber wall facing
toward the inside near the injector face; how-
ever, the devices used thus become vehicle
mounted and require interfaces to ground-support
equipment. Also, "windows" in the chamber wall
represent undesirable surface discontinuities. It
is unlikely, therefore, that optical devices will
find wide application for ignition detection.
Pyrometers
Heat-sensitive pyrometers are closely related
to the optical devices and subject to the same
limitations.
Fusible Wire Links
For many applications these are simple and
reliable devices. A wire is strung across the
chamber exit which, when fused by the ignition
flame, interrupts a circuit and signals "ignition
OK." Through proper selection of wire gage,
material and distance from the chamber exit
and/or center, some quantitative judgment is
obtained. The wire can be ground mounted or
chamber mounted. It must be isolated and should
have spring loading, like the well-known electric
fuses, to assure positIve separation.
Wire links have a number of shortcomings.
The fused wire ends may touch other metal parts
and thus reconnect the circuit before the relay
drops out. SuitabJ e circuitry and mounting must
therefore be applied. If a pyrotechnic igniter is
used, the wire can be broken by inert particles,
or even by a dud igniter coming out of the cham-
ber, giving an incorrect "ignition OK" Signal.
This has been overcome by providing redundancy
using several wires in parallel, all of which
DESIGN OF THRUST CHAMBERS AND OTHER COMBUSTION DEVICES 143
must be broken before the sequence can proceed.
In another design the wire has been mounted as
a loop placed in a groove on a wooden or plastic
stick. It is thus supported against all reason-
ably expected mechanical damage and adequate
insulation is maintained after fusion.
Pressure-Sensing Devices
Because of the need to mount the fusible
wire s at the exit of the thrust cham ber, they are
subject to some of the limitations noted for vis-
ual and optical methods. It has been attempted,
therefore, to sense the pressure rise in the com-
bustion chamber resulting from the burning igniter
flame. However, since the pressure rise is small
(a few psi at best), reliable discrimination is
difficult. Furthermore, the sensing-pressure
switches must be able to withstand the much
higher pressures during subsequent main stage.
Pressure-sensing devices have potential for
multi start engines.
Resistance Wires
Another method designed to overcome the
shortcomings of fusible wire links is the appli-
cation of resistance wires. Constructed like a
glow plug and connected to a bridge circuit, the
resistance wire will Signal by a distinctly differ-
ent resistance in the presence or absence of
ignition. The art is to find that spot in the
thrust chamber or gas generator which experi-
ences a clear temperature rise as a function of
ignition, yet remains cool enough to prevent
fusion of the wire. Resistance-wire sensors are
ideally suited for repeatable start engines.
Indirect Methods
In conjunction with hypergolic slug ignition,
other approaches to ignition sensing have been
developed. In one design an electric contact
assures that a cartridge is actually installed.
Thi s doe s not assure, however, that the cartridge
is loaded or completely loaded, nor that the
downstream lines are not clogged or that the
diaphragms will burst. Weighing of the cartridge
and purging of all lines must be included in the
firing preparation.
In another arrangement a pressure switch
senses pressure buildup in the igniter injection
line upon rupture of the hypergolic fluid cartridge
burst diaphragms. The switch Signal then initi-
ates the next sequential step. A modification of
the system substitutes a pressure-actuated valve
for the switch with similar effects. This method
does not assure. however, that the cartridge is
properly filled with the right amount of the cor-
rect fluid.
Spark igniters use electric devices which
ascertain that the plug is sparking based on
conductivity effects due to ionization near the
electrodes.
The methods in the above list. which un-
doubtedly is not complete. are described as in-
direct because none of them directly and reliably
detects ignition; i.e .. the release of adequate
heat. This is a drawback and cannot entirely be
offset by weighing, certifying, and inspecting.
4.8 COMBUSTION INSTABILITY
"Combustion instability" is defined in terms
of amplitude of pressure fluctuations in the com-
bustion chamber. Chamber-pressure fluctuations
are always present during normal, stable opera-
tion of a rocket engine system. These fluctua-
tions are generally quite random. showing fre-
quency spectra which are essentially continuous
in nature, with few, if any, recognizable peaks.
However, in case of instability. large concentra-
tions of vibratory energy appear at one or more
frequencies in the spectrum. They can easily be
recognized against the normal random-noise
background.
It has been found experimentally that the
amplitude of the chamber pressure oscillations
which will cause detrimental physical or opera-
tional effects varies widely for different thrust
chambers and engine systems. Also. in a given
chamber or system the effects of various types
of instability can be quite different at the same
amplitude. Thus it is difficult to assign a quan-
titative value to the amplitude at which the com-
~ u s t i o n chamber should be considered as running
unstable.
It is an interesting observation that the first
large liquid-propellant rocket-propulsion system,
the German A-4 (V-2) rocket. never experienced
combustion instability in over 4000 launchings
and in several times as many static chamber and
144
DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
engine firings. The term "combustion instabil-
ity" was literally unknown. It has been sug-
gested that the Germans may have had instabil-
ity, but did not know it because of their poor
high-speed recorders. To this it can be said
that for those instabilities which have caused
today's real difficulties, the high-speed recorder
chart is merely a postmortem confirmation of the
cause of often very costly failures, which did
not require any special instrumentation but 'were
unmistakable even to the untrained visual ob-
server. It is much more likely that there is a
relationship between the low performance level
of t.he A-4 or the chamber geometry, or both, and
the inherent stability.
Experimentally it has been found that as long
as the ratio of the peak-to-peak amplitude of
pressure oscillation to average chamber pressure
is kept below 0.10, there is usually no physical
damage to the chamber. However, while a 10-
percent variation in pressure for any class of
instability may not appear to have detrimental
physical effects instantly or within a short
period. it would be unacceptable for longer range
rocket vehicle missions. One practical way to
detect combustion instability and to prevent it
from causing damage during engine operation is
by monitoring the vibratory acceleration of the
system. Accelerometers are mounted on the
system to monitor in all three coordinates. They
are connected to discriminator circuits which are
set to specific g-load limits. When these limits
are exceeded, a counting instrument begins to
record. As soon as an allowable cumulative
number of oscillations is exceeded, an engine
cutoff signal is automatically triggered to pre-
vent damage.
Types of Instability
For proper remedial action, it will often be
important to know whether the observed oscilla-
tions are of a longitudinal. radial, or tangential
mode, or a combination of these. These three
modes and their normal acoustic frequency are
indicated in figure 4-59. Furthermore. in certain
cases. it will be extremely valuable to know
whether the oscillations originated in the thrust
chamber, or in the feed system, or whether they
originate in an interaction of characteristics,
harmless when separate but destructive when
L.QN(2IT!,!QI NAL. TANI2ENTIAL .BAQlAl..
MQQE.
~
MQQE.
Ae 0.59Ac l22Ae
N=- N=-- N=--
2Le de de
Figure 4-59.-Three modes of instability. Le =
combustion chamber length (injector [ace to
throat); de = combustion chamber d i am e t e r;
N = normal acoustic frequency; Ae = velocity of
sound in chamber.
combined, of both feed system and chamber.
This will be further discussed in connection
with methods to improve stability.
The effects of the oscillations on an engine
system are very much dependent on frequency.
These effects may range from Simple shaking
(usually at the lower frequencies). possibly re-
sulting in an eventual mechanical failure after
sometimes prolonged exposure, to "acoustic"
vibrations (usually at the higher frequencies)
capable of destroying the entire system in a few
hundred milliseconds. But, how low is "low"
and how high is "high "? As pointed out earlier,
the general field of combustion stability is ex-
tremely complex and it would be far beyond the
scope of this book to attempt to present a gen-
eralized theory of the subject. Each system.
because of configuration and dimensions, be-
haves somewhat. differently and requires special
treatment. The thrust chamber designer must
have a basic understanding of the stability prob-
lem. and it is felt that this can best be conveyed
by describing it in terms of a typical system, for
which a substantial amount of experimental data
exists.
The frequency of the chamber pressure oscil-
lations in a given chamber is determined by the
geometry of the system as well as by complex
interactions between the fluid flow in the pro-
pellant lines, the physical and chemical process
DESIGN OF THRUST CHAMBERS AND OTHER COMBUSTION DEVICES 145
.. ' ..

i ..

___ J ___ J. ___ l. _____
. .. ..
THRVST CHAMB!:R "W(TURf 'U,rlO
Figure 4-60. -Approximate vibration characteris-
tics at 150000 lb thrust level.
of combustion, and the dynamics of the combus-
tion gases in the chamber. It has been found
that each of the frequency components in the
instability spectrum of a thrust chamber is pre-
dominantly innuenced by only one physical
process so that it is possible to group the ob-
served instabilities in broad general classes,
which are: high frequency or gas dynamics, low
frequency or hydrodynamics, and intermediate
frequency or combustion dynamics, listed in the
order of their relative importance. Figure 4-60
presents vibration amplitudes of various fre-
quencies versus mixture ratio of a typical L0
2
/
RP-1 engine at the 150000-pound thrust level.
The graph indicates the large difference in vibra-
tion amplitude between a stable and an unstable

/ligh-Frequency Instabilities
High-frequency instabilities at frequencies of
approximately 1000 cps and over are sometimes
referred to as "damaging acoustiC," or "scream-
ing" modes of instability. They are gas-dynamic
instabilities which are both sustained and initi-
ated by the combustion process and are believed
to be concentrated in the uppermost portion of
the combustion where they cause in-
creased heat-transfer rates to the injector suf-
ficient to melt and burn it through within a few
hundred milliseconds. They also frequently have
serious damaging effects upon other parts of the
rocket engine system. High-frequency instabili-
ties are further characterized by instantaneous
initiation (a few milliseconds from absence to
full amplitude), and by extreme difficulty in
eliminating them once they are initiated. They
do not occur at, nor can they be generally damped
to, low-amplitude levels. They are either pres-
ent at high amplitudes or not at all. It is be-
lieved that these oscillations are predominantly
of the radial and tangential types.
In many systems extremely unpredictable
high-frequency instability has occurred. It is
often, but by no means always, connected with
the buildup phase to main stage. Systems which
ran stably during numerous successive tests can
become unstable without warning or subsequent
clear indication of a cause. Figure 4-61 shows
the starting of a typical high-frequency instabil-
ity, indicated by the SUdden shift in the acceler-
ometer trace due to high-amplitude chamber pres-
sure oscillations. All types of rocket propulsion
systems, including solid systems, have been
plagued by high-frequency instability. A consid-
erable amount of research and engineering has
been devoted to the explanation and elimination
of this phenomenon.
Unless a run in which instability is encoun-
tered is terminated within fractions of a second,
serious damage to the engine hardware almost
always occurs. It is assumed that the rapid gas
pulsations directly interface with propellant
injection, their mixing, and with the combustion
process, upsetting the condition in the boundary
zones, in particular at the injector face, to such
an extent that the heat transfer to the metal parts
increases at a high rate. Within seconds, or
even fractions thereof, the injector can burn
through, permitting propellant mixing behind the
injector face. This, in turn, leads to explosions
which often completely destroy the system.
It has been observed that the degree and
speed of damage is somewhat related to the
level of energy release occurring in the combus-
tion chamber. This may explain why "bursts of
Figure 4-61. -High frequency combustion
ity shown on oscillograph for engine accelerometer.
146
DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
instability" during shutdown cause damage less
frequently than they do during buildup and main
stage.
Low-Frequency Instabilities
This type of instability, at frequencies below
approximately 180 cps, is a hydrodynamic oscil-
lation, characterized by a cause-and-effect-type
coupling between combustion process and pro-
pellant feed system flow processes. The phe-
nomenon is sometimes referred to as "chugging."
Its secondary effects can be serious indeed.
"Chugging" may trigger destructive high-frequency
instability. Also, prolonged chugging can lead
to loosening of bolts and other vital connections
and to ruptures in general. Low-frequency in-
stability is self-sustaining but may damp out.
As a rule, it is predictable from analytical and
from test result studies. In figure 4-62 the high-
speed-pressure instrumentation measurements
indicating chugging clearly show that the oscil-
lations of propellant feed system pressures are
at the same frequency as the "rough combustion
cutoff accelerometer" reading.
Chugging occurs most frequently during
buildup and shutdown of an engine system, or
when operating at off-rated operating levels,
such as at incorrect mixture-ratio values. Like-
wise, too high, and particularly, too Iowa thrust
level can lead to chugging. This is especially
important for systems requiring throttling to a
lower-than-rated thrust level during flight. If
sustained, chugging will cause measurable per-
formance losses, which are attributable to widely
fluctuating mixture ratios.
The chugging phenomenon is frequently asso-
ciated with the quality and promptness of igni-
tion of the entering propellants. This can be
descnbed as "flame holding characteristics,"
"combustion timelags," "flame propagation
velocity," or other terms, which ultimately are
all traceable to the excessive accumulation of
unburned fuel, with subsequent detonation or
cyclic higher-than-rated combustion. The re-
sulting excessive chamber-pressure spikes effect
a reduction, or even reversal, of the propellant
flows. This will cause rapid collapse of the
chamber pressure, allowing propellants to rush
in again, thus repeating the cycle. It is readily
evident that the physical dimensions of the com-
I I

I .,' I I
,."" .. i ..... ... .w-.. " ... , .
I I' I

,I I
'''J[CTICIrI
",[SSulll[
Ii' I I
furl Utl[T
M.U;lfOLO 'R[SSUI![
I
Figure 4-62.-High-speed pressure measurements
of low-frequency instability.
bustion chamber and of the propellant ducts, and
the magnitude of the propellant flow rates and
their ratio to one another (in a bipropellant sys-
tem), are critical to the phenomenon of low-
frequency system oscillations.
Intermediate-Frequency Instabilities
This instability, with frequencies ranging
from 200 to 1000 cps, is sometimes referred to
as combustion dynamics or "buzzing." It is
characterized by a spring-and-mass-type coupling
between combustion process and propellant feed
system flow processes. It is often present in
only a portion of the feed system, or is confined
to the combustion chamber, or, in a bipropellant
system, to one of the two propellant systems
only. It appears to be initiated by the combus-
tion process and to be sustained by acoustic
resonance of a critical portion of the system.
Some researchers have shown that in a system
having a pump, the pump may be the prime source
of these oscillations.
This type of instability has not appeared to
be a problem in the development of large engines.
While it can occur occasionally in large engines.
it appears to be much more prevalent in small-
scale low-thrust systems. Systems oscillations
of the buzzing type are undesirable because of
their adverse effects on engine reliability and
DESIGN OF THRUST CHAMBERS AND OTHER COMBUSTION DEVICES 147
' 1 J ~ 1 . . IKJI!C;l'IOl't
IOttUSVIII!
'UI!:I0 IIIIIL[T
IIIAli!Il"o\'O ,"fUUfII!
'uf:L .. U.... OI.lTl..I1'
'fIIUSUlit.
itOveM co ... usno"
(U'O.,
ACCILUOMITIIII
'ulL "'*' I_If
'1II[S5111t1
Figure 4-63. -High-speed pressure measurements
of intermediate frequency instability.
performance. When exposed to prolonged buzz-
ing. critical parts of the engine may fail be-
cause of material fatigue. and thus cause sec-
ondary major failures. In addition. measurements
have shown that a performance (Is) loss of up to
approximately 7 percent can be incurred. which
is largely attributed to widely nuctuating mixture
ratios.
Figure 4-63 is the record of high-speed-pressure
measurements of a typical test afnicted by buzz-
ing. The oscillation is attenuated in the fuel
pump outlet pressure, and nonexistent in the
pump inlet pressure. This indicates that the
buzzing is limited to only a part of the feed
system.
Field of Stability
For a given engine system, combustion stabil-
ity limits can be defined experimentally in terms
of certain operational parameters such as cham-
ber pressure, injection tlP. and mixture ratio.
Figure 4-64 presents the stability field for a
typical L0
2
/RP-1, 150000-pound nominal thrust
engine system. The stability field. together
with heat transfer limits. can serve as a guide to
the design of a safely operating engine system.
DeSign Approaches Toward Control of Combustion
Instability
The problem of controlling combustion insta-
CHAN8R ,.ft[SSlJR(
Figure 4-64. -Field of stability and safe operating
region of a typical engine system.
bility in liquid propellant rocket engine systems
can be attacked in several ways. including sys-
tems design, choice of propellant combination.
and operating conditions, and through the use of
special control devices. Two basic approaches
toward eliminating combustion instability are
employed. The first is to eliminate triggering
processes and/or to limit the drivin!? energy per
cycle to a value below the damping per cycle
inherent to the system. This approach is typi-
fied by design investigations in which injector
configurations are varied to give different atomi-
zation and propellant distribution characteristics
with varying resistance to initiation of instabil-
ity, or in which propellant additives are used
which modify the physical properties controlling
spray formation. chemical kinetics. etc. Experi-
mental studies. both with full-scale engines and
with research model thrust chambers. have shown
that injector modification yields relative stabil-
ity ratings. differing by a factor from 5 to 6 from
the most stable to the least stable configuration.
The second approach is to introduce addi-
tional damping in the system through the use of
mechanical or other devices so that any opera-
tional disturbance. or oscillation triggered by
the disturbance, are sufficiently and rapidly
damped out. Experience over the past several
years has shown that the destructive transverse
acoustic modes of instability can be most effec-
tively combated through the use of this second
approach. Mechanical devices, such as barnes
or a divergent wall gap, have been found to

1,,' .. .. - ~ ,. .......
- l.ie .... :........ .
148
DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
introduce sufficient damping into the system so
that it will recover from an instability triggered
by an explosive charge as large as can be used
without damaging the thrust chamber in some
other manner. This ability of a system to re-
cover from a triggered instability has been des-
ignated as "dynamic stability." A prerequisite
for any propulsion system to operate reliably is
that it should exhibit dynamic stability with
respect to all modes of instability. As a mini-
mum requirement it should be "dynamically
stable" at least with respect to the destructive
transverse (radial and tangential) acoustic
modes.
The successful application of the above
methods has been based primarily on criteria
established empirically in research model thrust
chambers, together Vlith testing in actual engine
systems. However, the understanding of the
fundamental physical principles of the damping
processes is still limited.
Prevention or Triggering Processes
The most desirable design method of con-
trolling instability is the prevention of those
physical or chemical processes which trigger
and/or sustain the resonant modes of the com-
bustion chamber or engine system. While a great
number of studies, in which different design
parameters were varied systematically, have
been made by various investigators, the results
have failed to yield truly generalized design
criteria. This can be traced to the fact that
basic processes which trigger and sustain the
various types of instability have not been iso-
lated. Thus, while a parameter which controls
one type of instability may have been estab-
lished on an engineering basis, this same design
criterion may be enhancing another type of insta-
bility. The folloVling is a general discussion of
the prevention of triggering instability in various
com!)onent and subsystem designs:
1. Propellant feed system design.-Past ex-
perience has indicated that certain combustion
instabilities, such as buzzing, are sustained
through an interaction between feed system and
combustion dynamics. It is believed that hy-
draulic resonances are a major factor in sustain-
ing this type of instability. The requirement is
to design a feed system whose hydraulic charac-
teristics will not trigger the interact!on with the
combustion process.
2. Combustion chamber design.-Analytical
studies and experimental results have indicated
that the geometrical configuration of the combus-
tion chamber will determine the type of frequency
of the acoustic modes of instability. Chambers
having large length-to-diameter ratios appear to
be quite prone to large-amplitude longitudinal
instability. On the other hand, chambers having
small length-to-diameter ratios appear to be
sensitive to the transverse modes. Also, small-
diameter chambers are much more stable than
large-diameter chambers. The requirement is to
design a chamber geometry which will have least
tendency to trigger instability, in conjunction
with other considerations.
3. Injector design.-The injector deSign ap-
pears to be a most critical factor in triggering
instability. In turn, it offers great potential for
controlling instability-triggering processes
through variation of parameters. Of the most
common types of injectors, the self-impinging
injector (fig. 4-45a) has been chosen by many
investigators as the best compromise between
performance and stability. There is some indi-
cation that longitudinal instability may be en-
hanced if the propellant travel time from the
injector face to the point of impingement is close
to the half-period (or an odd multiple of the half-
period) of the longitudinal mode oscillations.
There are strong indications that increasing the
injection tlP to too great a value may cause the
thrust chamber to operate unstably in the trans-
verse acoustic modes. The effect of injection
tlP on the longitudinal acoustic modes and on
hydrodynamic instabilities appears to be just the
opposite, with stability improving as tlP in-
creases. The propellant distribution across the
injector face has a significant relation to trig-
gering transverse modes of instability.
4. Propellant combination and mixture ratio.-
Control tests with various propellants have
shown that there are certain oxidizers or fuels
which can be triggered into instability more
easily than others, in general or at certain mix-
ture ratio conditions. In L0
2
/LH
2
systems, a
lower hydrogen injection temperature tends to
trigger instability.
5. Engine system operating characteristics.-
In some cases, combustion instability can be
DESIGN OF THRUST CHAMBERS AND OTHER COMBUSTION DEVICES 149
avoided by modification of engine-system oper-
ating characteristics. For instance, if the cham-
ber pressure buildup period during engine start
is excessive, "chugging" may have time to de-
velop. This low-frequency instability, in turn,
can trigger the destructive acoustic modes of
instability by contaminating one propellant sys-
tem with the other propellant by driving it back
through the injector ports. The triggering chug-
ging can be eliminated through a fast chamber
pressure buildup.
Application or Damping Devices
In further support of design for stability,
special damping devices can be provided for
"dynamic stability" in the combustion chamber
and engine system. The following are several
devices which have been applied with good
results:
L Feed system hydraulic capacitor.-Self-
sustained combustion and feed system instabil-
ties of the buzzing type can be eliminated by
introducing hydraulic capacitors in the feed sys-
tem. A hydraulic capacitor is any device which
will increase the effective compressibility at a
given point in the liquid system. Figure 4-65
shows the schematic of an experimental hydraulic
capacitor. It is an isolation-type capacitor with
large capacitance and zero resistance between it
and the system. Its function is to isolate the
transmission of pressure disturbances through
the system above a given frequency. Another
type is called absorption capacitor, which has
small capacitance and high resistance between
it and the system. Its function is to absorb the
oscillatory energy of the system by damping or
attenuating the resonant frequencies of the
system.
2. Combustion-chamber baffles.-The use of
combustion-chamber baffles has been found to be
Figure 4-65.-Schematicof experimental isolation
type hydraulic capacitor.
the most effective method of suppressing trans-
verse acoustic modes of combustion instability.
This has been demonstrated in both full-scale
thrust chamber and small-scale models, operating
with various propellant combinations. These
baffles are usually designed to be secured to the
injector face as shown in figure 4-40. Adequate
cooling means should be provided to keep the
baffles from burnout. The depth or height of the
baffles is a function of the distance of the
combustion-flame front from the injector face.
Experimental evaluations should be conducted to
support the design and development effort.
3. Chamber divergent wall gap.-It was found
experimentally that leaving blank an annular
portion of the propellant injection area adjacent
to the combustion chamber wall, as shown in
figure 4-66, improved the capability of the com-
bustion chamber to recover from triggered insta-
bilities. It was further determined that by filling
this "wall gap" with a contoured filler block,
"dynamic stability" could be drastically improved
in most cases. The exact shape of the contour
is critical. and experimental evaluations are
required to determine the most effective design
configuration.
4. Acoustic chamber Iiners.-The feasibility
of using resonant and nonresonant acoustic sup-
pressing liners on the combination chamber walls
has been demonstrated in turbojet engine appli-
cations. In this case the principle of suppres-
sion is similar to the use of patches of acoustic
tile to reduce the sound level in a room, whereby
the energy absorbed from the mode will reduce
its amplitude. Figure 4-67 shows a typical
arrangement. The combined area of the sup-
pressor openings must be of the order of from 3
INJECTOR FACE
------ - ----- -
CONTOURED
WALL GAP
FILLER BLOCK
Figure 4-66. -Combustion chamber divergent wall
gap.
150 DESIGN OF LIQUID PROPELLANT ROCKET ENGINES
INJE'CTOR FACE
-:--:0---
. ;"."., .. ::."
ACOUSTIC SUPPRESSION LINER
Figure 4-67.-Combustion chamber acoustic liner.
to 4 percent of the total chamber wall surface
area. Furthermore, the suppressor thickness
should be maximum in the area of maximum pres-
sure variation; i.e., near the injector face.
Rating Stability
It is desireable to establish the combustion
stability level of a particular engine system
without an excessive number of tests. This can
be accomplished by perturbing a normally stable
system by suitable means until instability is
initiated. The relative stability of various sys-
tems is then judged as a function of the magni-
tude of perturbation needed to reduce the insta-
bility. The larger the perturbation, the more
stable the system. The perturbation can be
introduced in either the propellant feed system
or in the combustion chamber. Several effective
methods exist for inducing different types of
instability.
Feed System Perturbations
Perturbation in the propellant feed system to
induce disturbances in the chamber of the low-
frequency hydrodynamic type, can be introduced
by-
(1) Explosive charges in the fuel feed system
(2) Single-stroke positive displacement
pistons
(3) Oscillating pistons
(C!' DIRECTED EXPLOSIVE
PULSE
r EXPLOSIVE
/--'"
ChARCE
\
I\, / j
"---./
(b) ND1<4 DIRECTED
eXPLOSIVE PULSE
r CONTROL
I VALVE
/---p-
, '> GAS
SUPPLY
It) DIRECTED NON
GA.) FLOW
Figure 4-68. -Combustion chamber perturbation
methods.
Combustion Chamber Disturbances
The introduction of disturbance in the cham-
ber proper offers a simpler method of inducing
instability. Transverse acoustic modes have
been initiated most successfully by the following
methods:
1. Directed explosive pulses.-The directed
explosive pulse method of inducing instability,
as shown in figure 4-68a, uses a high explosive
charge mounted in an external fixture which is
attached to the combustion chamber in such a
way that the gas pulse resulting from the detona-
tion enters the chamber with any desired orien-
tation.
2. Nondirected explosive pulses.-The closest
simulation of localized random detonations which
can occur in a chamber during normal operation
due to accumulation of unburned propellants is
achieved by the nondirectional explosive pulse
method as shown in figure 4-68b. An explosive
is placed into a thin-walled Micarta shell which
is designed to be mounted at any desired posi-
tion in the chamber.
3. Directed nonexplosive gas flows.-In this
method, as shown in figure 4-68c, a flow of gas
from a regulated high-pressure source is con-
trolled by a fast-acting valve. This valve is
placed as near to the chamber as possible. This
method additionally permits a better definition of
the parameters associated with the disturbance.