Advanced Rocket Engines

Advanced Rocket Engines
Oskar J. Haidn
Institute of Space Propulsion, German Aerospace Center (DLR)
74239 Lampoldshausen
Germany
[email protected]
SUMMARY
Starting with some basics about space transportation systems such as the thrust equation and some
mission requirements, the paper explains the underlying physical and technical challenges every rocket
engine design engineer faces at the beginning of a project. A brief overview about the main subsystems of
a liquid rocket engine such as turbopumps and gas generators is followed by a more detailed description
of the thrust chamber assembly, the injector head, the ignition system and the combustion chamber liner
which includes the first part of the nozzle and finally the nozzle extension. The technological challenges of
these components are presented which result from the severe operating conditions and their current
design principles and production technologies.
Finally, a series of new concepts and techniques for propellant injection, ignition, combustion chamber
liners and nozzles are proposed. The major challenges of new materials and production technologies lay
in the still missing detailed knowledge about the behaviour of these materials under operating conditions,
material-related crack initiation and propagation laws and reliable life prediction tools.
INTRODUCTION
Rocket engines may either work with solid or liquid propellants or as a combination of both, hybrid
propulsion systems, such as the one for the SpaceShipOne project. Liquid rocket engines can be
subdivided into mono-propellant or bi-propellant systems. Mono-propellant engines operate either as a
simple cold gas system or apply a catalyst for an exothermal decomposition of the propellant, such as
hydrazine (N2H4) or laughing gas (N2O). Generally these type of engines are only in use for low thrust
satellite propulsion systems. Typical bi-propellant engines use either earth storable propellants, generally
combinations of nitric acid or its anhydride with derivates of hydrazine, i.e. asymmetric di-methylhydrazine (UDMH) or mono-methyl-hydrazine (MMH), mixtures of storable and cryogenic propellants,
liquid oxygen and kerosene or fully cryogenic, liquid oxygen and liquid hydrogen. Although rocket
engines show large differences depending on mission profile and staging of the launcher it is possible
useful to separate them in four major classes. Booster, main stage and upper stage engines and satellite
propulsion and attitude control systems [1,2].
Rocket engines are energy conversions systems with a heat release in the combustion chamber which
exceed by far typical values of nuclear power plants (~ 3-4 GW). While solid rocket engines may even
reach power levels of more than 30 GW, the most powerful liquid rocket engines have peak power values
of up to 20 GW but the majority works with values of less than 10 GW. Obviously such power levels are
only possible with high combustion chamber pressures and even higher propellant mass flow rates which
may exceed 1000 kg/s.
Haidn, O.J. (2008) Advanced Rocket Engines. In Advances on Propulsion Technology for High-Speed Aircraft (pp. 6-1 6-40). Educational
Notes RTO-EN-AVT-150, Paper 6. Neuilly-sur-Seine, France: RTO. Available from: http://www.rto.nato.int.
RTO-EN-AVT-150
6-1
1.
CHEMICAL PROPULSION SYSTEM BASICS
The basic concept of a rocket engine relays on the release of the internal energy of the propellant
molecules in the combustion chamber, the acceleration of the reaction product and finally the release of
the hot gases at the highest possible velocity in the convergent/divergent nozzle.
1.1
Thrust equation
The thrust equation which describes the basic relations of a chemical rocket was first given in 1903 by
the Russian Konstantin Tsiolkovskiy. Assuming a frictionless, 1-D flow of an ideal gas with negligible
entry velocity and the heat release in the combustion chamber at constant pressure with adiabatic walls,
see Fig. 1, the thrust equation can be written as:
pa
Tb, pb
uh
ua
p
Figure 1: Schematic of an ideal rocket engine
F = m u e + ( p e p )Ae
(1)
ue, the exhaust gas velocity at the exit area Ae,, pe, the exit and p the ambient pressure, respectively. The
first term in eqn. (1) is the momentum and the second, the pressure part of the thrust and replacing
c = u e + ( p e p )Ae / m or c = C F c * finally yields:
F = m c
or
F = m C F c *
(2)
CF is commonly known as the thrust coefficient, c, the effective exit velocity and c* the characteristic
velocity. It is useful to examine these coefficients in more detail since they give valuable hints about the
relevance of fluid properties and operational parameters. The characteristic velocity relates the combustion
chamber pressure to the burned propellants and thus represents the energy content of the propellant and
the combustion efficiency. The widely used term specific impulse is available with the gravitational
constant g0.
1
1
k +1
+1
2
2
k 1 RT
1
p
p
p
1
+
1
2
2
c
c
c* =
; CF =
1 e + e ; I sp,0 = ; (3)
2
M
g0
pc pc
1 + 1 pc
with , the ratio of cross section of nozzle exit and combustion chamber throat.
As equation (3) shows, the characteristic velocity increases with increasing combustion temperature
Tb , but decreases with growing molecular weight M and isentropic coefficient k of the exhaust gases. This
dependency on the isentropic coefficient holds also for the thrust coefficient but of much greater impact is
the ratio between exit pressure and chamber pressure and, in the second term the difference of exit and
ambient pressure. Although large exit velocities are quite favourable, the corresponding expansion in the
nozzle may depending of the height during ascent result in exit pressures smaller than the ambient
6-2
RTO-EN-AVT-150

pressure which finally yields a smaller thrust coefficient. Typical values of these characteristic properties
for a wide range of rocket engines are summarized in the following table 1 [3].
Table 1: Typical values of characteristic properties of rocket engines
Tc [K]
Pc [MPa]
M [kmol/kg]
c* [m/s]
[-]
K [-]
CF [-]
Isp[-]
2000-3900
1-26
2-30
900-2500
15-280
1,1-1,6
1,3-2,9
150-480
A rather interesting aspect about the impact of the combustion chamber pressure on the performance
of the engine is demonstrated in figure 2 which shows the vacuum specific impulse as a function of the
area ratio. While variations in the chamber pressure itself have only a minor influence, i.e. an increase by a
factor of six yields only a marginal (about 0.1 %) performance increase, an only minor increase of the
combustion efficiency (by 1% from 0.96 to 0.97) yields already an increase of the performance of about
1%.
4650
pc = 200 bar, Isp = 0,97
Vacuum Specific Impulse / m/s
4600
4550
4500
pc = 200 bar, Isp = 0,96
4450
4400
pc = 35 bar, Isp = 0,96
4350
4300
4250
50
100
150
200
250
300
350
400
N o z z le A re a R a tio A e /A t
Figure 2: Vacuum specific impulse as a function of the area ratio
with chamber pressure and combustion efficiency
1.2
Propellants
Following the previous chapter on characteristic properties allows a classification of typical

propellants for rocket engines. Key requirements are high combustion chamber temperatures, usually
achievable with a high energy content of the molecules, a small isentropic coefficient and the smallest
possible molecular weight of the exhaust gas mixture. Depending on the specific launcher system, rocket
engine cycle or mission requirements, values such as propellant density or temperature may become even
more important. Generally, the propellant combination H2/O2 is favourable for upper stage engines while
RTO-EN-AVT-150
6-3

booster engines either apply kerosene / oxygen or solid propellants and as such quite often a combination
of ammonia per-chlorate (AP), hydroxy-terminated poly-butadien (HTPB) and some ingredients which
have the function of either a binder or a moderator to control the heat regression rate and thus the overall
booster performance.
In the early days of rocketry hypergolic propellant combinations of nitrogen-tetroxide (N2O4) and
mixtures of N2H4 and UDMH or MMH have been used for almost any engine application although they
have a comparatively small specific impulse, are toxic and cause due to their aggressiveness problems
during storage and handling but nowadays they are very common only for upper stages and satellite
propulsion systems. The main reason for this wide-spread usage is based on their chemical reactivity; they
dont require an ignition system and are storable at room temperature. Nowadays booster engines using
kerosene / O2 are commonly used and the most impressing examples are the Energomash engines of the
RD-170 family which power the first stages of the Zenit (RD-171) and the ATLAS (RD-180) launchers.
Table 2 shows frequently used propellant combinations, their mixture ratios, specific impulse and
densities. For further information see i.e. [3, 4]
Table 2: Characteristic data of typical liquid propellants
Oxidizer
LO2
N2O4
rof
Isp
[-]
[s]
[kg/m3]
kerosene
2,77
358
820
LH2
4,83
455
700
LCH4
3,45
369
430
UDMH
1,95
342
791
MMH
2,37
341
880
UH25
2,15
340
850
Fuel
The influence of the propellant combination and the mixture ratio on the specific impulse is shown in
Figure 3. Three different classes of propellants can easily be detected. H2/O2 is by far (~30%) the most
efficient propellant pair followed by combinations of O2 and hydrocarbons and interestingly the specific
impulse decreases with increasing carbon content. Fuels which already contain oxygen atoms, alcohols,
follow by another 10% margin and have an almost similar specific impulse as storable propellant
combinations. For comparison reasons figure 3 includes as well propellant combinations of which use
hydrogen-peroxide H2O2 as oxidizer.
6-4
RTO-EN-AVT-150
4500
LOX-LH2
Ideal Specific Vacuum Impulse Ivac [Ns/kg]
4300
4100
3900
LOX-Kerosene
LOX-Propane
3700
3500
LOX-Ethanol
LOX-Methane
LOX-Methanol
3300
3100
90%-H2O2-Kerosene
90%-H2O2-Ethanol
2900
Ideal Specific Vacuum

Impulse
N2O4-MMH
2700
pc = 100 bar, A e/A t = 45, CET93

2500
0
Propellant Mixture Ratio O/F [-]
Figure 3: Ideal specific impulse of various propellant combinations
A possibility to combine the advantages of high propellant density at low altitude operation with high
specific impulse at high altitudes is a combination of three propellants. So-called tri-propellant
combinations of LOX, kerosene, LH2 or LOX, CH4, LH2 burn in the same engine LOX / kerosene at low
altitudes and switch to LOX / LH2 operation at high altitudes have been investigated in the past but have
never reached the level of flight hardware. The RD 701 engine (LOX / kerosene which has been
designed for the air-launch MAKS concept reached at least development level [1].
1.3
Engine feed systems
The easiest way to distinguish different engine types is to classify them according to their method of
propellant pressurization and transport. While only small and low pressure engines apply pressurized tanks
for propellant delivery, the majority uses turbo pumps in order to bring the propellants to the desired
pressure level.
1.3.1
Pressure-fed engines
Principally there are two types of pressure-fed systems, self-pressurization and pressurization by
foreign high pressure gas. Self-pressurization is typically applied by mono-propellant engines and is
achieved either by vaporization of the liquid propellant or thermal decomposition caused external heat
addition or catalytic decomposition. Pressure-fed bi-propellant engines typically apply high pressure (up to
30 MPa) helium bottles. In any case, the thrust level of pressure-fed engines is limited by the available
tank technology. An example of such a system is the AESTUS engine of ARIANE 5G. A flow schematic
of a pressure-fed engine is shown on the right hand side of figure 4 together with a gas generator cycle and
a staged combustion cycle.
RTO-EN-AVT-150
6-5
Pressure-fed system
Pump-fed systems
gas generator
Fuel
Fuel
OX
LOX
expander
Fuel
LOX
staged combustion
Fuel
LOX
Figure 4: Common Engine Cycles
1.3.2 Pump-fed engines

In contrast to pressure-fed systems, pump-fed engines use the principle of a turbo-charger to
pressurize the propellants. Parts of the propellants are fed into a separate combustion chamber called gas
generator or pre-burner which usually operates at a sufficiently high pressure level and with a mixture
ratio such that the exhaust gas temperature reaches values around 850 K. These gases are then fed to a
turbine which drives the propellants pumps. Depending on the pathway of the turbine exhaust gases one
distinguishes between different engines cycles. The cycle schematics in figure 3 show that the case
characterized by the direct release of the turbine exhaust gases into the ambient is called Gas Generator
Cycle. In the case that they are fed together with the remainder of the propellants into the main
combustion chamber, this cycle is called Staged Combustion Cycle. The differences between those two
options are first, much higher system pressures since the gas generator has to operate at a pressure
sufficiently high to compensate the pressure losses in the turbine, the cooling channels and the injector
head and second a better performance since all the propellants go through the main chamber and thus
produce both a higher mass flow rate and allow so for a larger nozzle expansion ratio and finally for an
equivalent increase in the specific impulse.
With the exception of Russian LOX/kerosene stage combustion engines which run oxygen-rich, all
other gas generator or staged combustion systems operate in a fuel-rich mode. As part of the Integrated
High Payoff Rocket Propulsion Technology (IHPRP) program [5], the US aims at the development of an
integrated power-head which focuses on the most critical sub-components of this cycle, the oxygen-rich
pre-burner and turbine as well as the injection system of the main chamber. China as well works towards
the development of staged combustion technology [6]. The attractiveness of a full flow cycle engine
results from a higher overall system performance and the elimination of critical inter-propellant seals [7].
Nevertheless, there has been only one rocket engine to date, the Russian RD-270, which had an ox-rich
and a fuel-rich pre-burner which made it full scale to the test facility but unfortunately never into flight
[1].
6-6
RTO-EN-AVT-150
1.4
Engine applications
Typically the propulsion systems and thus the rocket engines are tailored to the specific requirements
of the launching concept. This has consequences insofar as any improvement or modification of the engine
has significant impact on the entire launch system and, even more important, an improvement of the
engine, i.e. engine performance, may not necessarily improve the overall efficiency of the launcher. The
currently operational space transportation systems can be divided into two major types: Launchers with
large strap-on boosters with core and upper stages such as the European ARIANE, the Japanese HII, the
Russian Soyuz or the American Space Shuttle and launchers with a booster, an optional sustainer and an
upper stage such as the American ATLAS and Delta, the Russian PROTON, the Chinese Long March or
the international Sea Launch Vehicle [2].
Typical booster engines provide a high lift-off thrust (i.e. 3000 8000 kN) over a short burning time
(i.e. < 150 s). Table 3 shows characteristic values of booster engines. Engines with a thrust level smaller
than 3 MN are clustered together to provide the necessary lift-off thrust. For example, the Soyuz boost
system consists of an ensemble of four RD-107 engines, the YF-21 boost stage comes with 4 YF-20
engines and the ARIANE 4 L40 configuration was made up by 4 Viking 6 engines. Core engines provide
a comparable lower lift-off thrust (i.e. 1000 - 2000 kN) but with a higher specific impulse and with
operating times in the range of about 600 sec. The thrust levels of upper stage engines vary between 30
150 kN with operating times varying between 600 1100 sec. Typical representatives of booster, main
stage and upper stage engines are summarized in table 3, table 4 and table 5. Unsurprisingly, almost all of
the booster engines shown in table 3 are fueled with relatively high density propellants and only one, the
RS-68 operates with the unfavorable density combination LOX/LH2.
Table 3: Characteristic data of liquid booster engines [2, 8]
Engine
Cycle
RD-170
SC
LO2/Kerosene
7,65
310
25,1
150
RD-180
SC
LO2/Kerosene
3,82
311
25,5
150
RD-107
SC
LO2/Kerosene
0,81
257
5,9
119
F-1
GG
LO2/RP1
6,91
264
6,6
161
MA-5A
GG
LO2/RP1
1,84
263
4,4
263
RS-27
GG
LO2/RP1
0,91
263
4,8
265
RD-253
SC
N2O4/UDMH
1,47
285
15,2
130
YF-20
GG
N2O4/UDMH
0,76
259
7,4
170
Viking 6
GG
N2O4/UH25
0,68
249
5,9
142
RS-68
GG
LO2/LH2
2,89
360
9,7
249
RTO-EN-AVT-150
Propellant
combination
Thrust
[MN]
Chamber
spec. impulse
pressure
(sl) [s]
[MPa]
Rocket
Engine
Burn
time [s]
6-7

Table 4: Characteristic data of core and main stage engines [2, 8]
Rocket
Engine
Engine Propellant
Cycle combination
Thrust
(sl)
[MN]
Thrust
(vac.)
[MN]
Spec.
impulse
(sl) [s]
Spec.
impulse
(vac.) [s]
Chamber
pressure
[MPa]
Burn
Time
[s]
RD-108
SC
LO2/Kerosene
0.78
1.01
252
319
5.1
290
Viking 5C
GG
N2O4/UH25
0.68
0.75
249
278
5.9
142
YF-20B
GG
N2O4/UDMH
0.73
0.81
259
289
7.4
170
RS-68
GG
LO2/LH2
2.89
3.31
360
420
9.7
249
SSME
SC
LO2/ LH2
1.82
2.28
364
453
20.5
480
RD-0120
SC
LO2/ LH2
1.51
1.96
359
455
21.8
600
LE-7A
SC
LO2/ LH2
0.84
1.10
338
438
121
390
Vulcain 2
GG
LO2/ LH2
0.94
1.35
320
434
11.6
600
Table 5: Characteristic data of upper stage engines [2, 8]
Rocket
Engine
Engine
Cycle
Propellant
combination
Thrust
(vac.)
[ kN ]
Spec.
impulse
(vac.) [s]
Chamber
pressure
[MPa]
Burn
Time
[s]
11D58M
SC
LO2/Kerosene
79.5
353
7,6
680
RD-0210
SC
N2O4/UDMH
582
327
14.8
230
AESTUS
PF
N2O4/MMH
30
325
1,0
1100
J-2
GG
LO2/ LH2
890
426
4.4
YF-75
GG
LO2/ LH2
79
440
3.7
470
LE-5B
EC
LO2/ LH2
137
447
3.6
534
HM7-B
GG
LO2/ LH2
70
447
3.5
731
VINCI
EC
LO2/ LH2
180
465
6.1
RL-10B
EC
LO2/ LH2
110
462
4.3
700
Main stage engines operate as booster engines right from the start and generally have thrust levels
similar to clustered booster engines somewhere in the range between 800 kN to 2000 kN. The burn times
however are much longer and may reach 400 s to 600 s. An analysis of the data presented in table 4 in
more detail reveals some interesting features. Similar thrust levels and burning times of the SSME and the
RD-0120 are no longer surprising, knowing that engines were developed to serve a similar purpose:
provision of sufficient thrust for a manned vehicle in low earth orbit. The LOX/LH2 fired LE-7A and
Vulcain 2, show comparable thrust levels and operating times and are applied as core engines in a
launchers with two large thrust boosters each, the HII and the ARIANE 5, respectively. A comparison of
the data for the RD-108, Viking 5C, and YF-20B reveals similar parameters and again, these engines are
all applied as clusters of engines in the first stage of launcher without large boosters.
The key characteristic feature of an upper stage engine is the vacuum ignition capability. The broad
differences in propellant combination, engine cycle, and thrust level are explainable only by the different
6-8
RTO-EN-AVT-150

applications. The main mission of the J-2, the second and third stage engine of the Saturn V, was to enable
the spacecraft to leave the earth for the moon. The engine with the second-largest thrust level, the RD0210, has a considerably shorter burn time and brings the Block M of the Proton into GTO. All the other
engines have thrust values below 200 kN and operating times between roughly 500s and up to 1100s.
The thrust requirement for apogee engines and satellite thrusters for attitude control is much smaller
than for any other engine. While the thrust levels of the first may reach up to 600 N, the latter first may
have peak values in the order of less than 100 N. Generally, attitude control thrusters are monopropellant
engines with tens of N. The key characteristics of a satellite engine are extremely long (more than 10
years) and reliable functioning and the very high cyclic loads due to the pulse operation mode. Hence, a
satellite engines is optimized towards aspects of long-life valve operation such as seals and leakage and
corrosion.
2.
COMPONENTS OF LIQUID ROCKET ENGINES
The key components of a liquid propellant rocket engine are the devices for propellant delivery, the
turbines and pumps as well as the generators of the driving gases, i.e. gas generator or pre-burner, the
propellant injection system, the thrust chamber which combines the combustion chamber and a short part
of the diverging section of the nozzle which typically ends with a distribution manifold for the propellants
used as coolant for the thrust chamber liner, and, finally the thrust nozzle. Each of these devices has its
own design problems and criteria and the optimum of one component not necessarily leads to the optimum
of the entire system. The following section will touch briefly the key issues of each of the components but
will finally focus on the thrust chamber.
2.1
Gas generator and preburner
The main requirement for any gas generator or preburner is the delivery of a sufficient amount of
driver gas a designed pressure and temperature which guarantees a continuous propellant supply of the
thrust chamber. A quick glance at equation (4), yields that the necessary turbine power depends aside of
the desired the ratio of exit to entry pressure and the mass flow rate of the gases, solely on thermodynamic
properties, temperature, heat capacity and isentropic coefficient, and , of course the efficiency of the
device.
1
p2
P = m c pT1 1
p1
(4)
As already mentioned previously, the turbine exit pressure in gas generator cycle engine is
independent of the pressure in the thrust chamber. This is entirely different for staged combustion engines
where the exhaust gases are fed to the chamber. While the combustion pressure in the gas generator lay
frequently below the one in the thrust chamber, the pressures in the pre-burners are typically a factor of
two or three higher, see table 6 which summarizes characteristic data of sample gas generators and preburners. A high turbine power may be achieved by appropriate mass flow rates and pre-burner exhaust gas
temperatures. Due to the extremely high dynamic loads of the turbines, the entry temperature of the gas is
limited to about 850 K to ensure sufficient turbine life which leaves only gas generator mass flow rate and
pressure to adjust to the power needs. Extremely important is a homogenous temperature distribution at
the entrance to the turbine to avoid local overheat at the turbine blades through hot spots.
RTO-EN-AVT-150
6-9

Table 6: Characteristic data of sample gas generators and pre-burners [10]
Vulcain 2
SSME
LE-7
RD-0120
LOX/H2
LOX/H2
LOX/H2
LOX/H2
875
940 / 870
810
846
pGG [ MPa]
10.1
35 / 36
21.0
42.4
rof [ - ]
0.9
0.89 / 0.8
0.55
0.81
~55
54
m [ kg/s ]
9.7
80 / 30
53
78.6
887
P [ MW ]
5 / 14
56 / 21
4.5 / 19
62
44
93.5
pc [ MPa ]
11.6
20.6
12.7
21.8
6.6
25.7
Propellant
Combination
T
[K]
RF-1
RD-180
LOX/RP1 LOX/RP1
816
820
55.6
An analysis of all the gas generator engines shows that cryogenic engines such as SSME, LE-7, RD0120, Vulcain 2, operate independently of the cycle, the gas generator or pre-burner in a fuel-rich mode.
Taking into account that all liquid rocket engines operate fuel-rich and that the stoichiometric mixture is
for any propellant combination always larger than 2, reveals that the amount of oxidizer outnumbers that
of the fuel significantly. Hence, oxidizer-rich operation allows for a sufficient reduction of the turbine
entry temperature since the mass flow rates are available. Although there has been considerable effort in
the US to develop an oxidizer-rich cycle, only Russia seems to have mastered the problem of material
compatibility of pre-burner, piping system and turbine as well as the injector head with high temperature
(T > 800 K) oxygen-rich turbine gas since all Russian LOX/Kerosene engines operate oxygen-rich [9].
None of the American Lox/Kerosene engines, neither the old F1, nor the RS-27 and the MA-5A which are
all gas generator cycle engines are running oxygen-rich although a fuel-rich operation will have problems
with soot formation in the gas generator and subsequently soot deposition along the flow lines [1]. The
only explanation might be that for the relatively short operation times of less than 200 s and rather small
pressures in the gas generator soot deposition in the turbine might still be tolerable while for even shorter
times of 150 s or shorter, either soot formation and deposition in the turbine is increased drastically or the
key problem is soot deposition in the injection system and especially in the injectors.
In order to emphasize the difference between LOX/H2 and LOX/Kerosene as well as gas generator
cycle and staged combustion cycle engines, table 5 summarizes typical data such as operating temperature,
pressure and mixture ratio of the gas generator or pre-burner, the propellant mass flow rate, the power
requirement of the turbine and the pressure in the thrust chamber. For comparison reasons an American,
European, Japanese and Russian LOX/H2 engine are shown together with two, American and Russian
LOX/Kerosene engine. Typically all cryogenic engines have for each propellant a turbine and a pump.
Only the RD-180 has one single preburner which delivers the necessary driving gas to a one single stage
turbine which again provides for the thrust requirements of the two propellant pumps, a single stage
oxygen pump and a two stage fuel pump.
2.2
Turbopumps
The function of the turbopump system of a rocket engine is to receive the liquid propellants from the
vehicle tanks at low pressure and supply them to the combustion chamber at the required flow rate and
injection pressure. The energy to power the turbine itself is provided by the expansion of high pressure
gases, which are usually mixtures of the propellants being pumped. Radial pumps and axial turbines are of
common use. Figure 5 basic elements of a turbopump and their main function as well as the flow passages.
6 - 10
RTO-EN-AVT-150
cycle
Figure 5: Basic elements of a turbopump and their functional description [11]
2.2.1 Turbopump design considerations

Turbomachinery design is an iterative process which requires the interaction of many disciplines, such
as hydrodynamics, aerodynamics, mechanics, materials and processes, structural mechanics and analysis,
rotor dynamics, thermodynamics and last but not least instrumentation and testing. The set of
requirements, pump discharge pressure and flow rate, pump suction pressure, turbine drive cycle and
efficiency, fluid properties, throttling range, life, reliability, weight and size, and cost are often
contradictive and the final design is at best a compromise optimum for a specific application. The SSME
powerhead shown in Figure 6 demonstrates the compact design of a liquid rocket engine.
RTO-EN-AVT-150
6 - 11
Figure 6: SSME Powerhead with preburners, turbopumps and thrust chamber assembly [11]
2.2.2
Pumps
Pumps for engines with similar density fuel and oxidizer propellants such as RP-1/LOX and similar
discharge pressure requirements will typically be optimum at approximately the same speed. This permits
the fuel and oxidizer pumps to be placed on a common shaft and driven by a common turbine which is the
case for the F1, RS-27 or the RD-180. Maximum pump speed is generally limited by the suction
performance requirements to avoid cavitation to occur. Due to the large density difference between liquid
hydrogen and liquid oxygen and the pressure drop required for regenerative cooling cryogenic engines
usually have two separate pumps since the hydrogen pump has to operate at much higher speed to provide
the necessary high exit pressures.
Vehicle weight reductions achieved by thinner propellant tank walls result in lower pump inlet
pressures and as a consequence the pump speed has to be reduced. Lower fluid velocities would in turn
result in a larger inlet diameter and increased size and weight of the pump. Furthermore, typical high
performance staged combustion engines require very high pump discharge pressures. A typical way to
overcome these weight and size constrains of the power pack and still maintain sufficient margin for a safe
operation are additional low pressure boost pumps which are applied in the SSME or RD-180 engine.
The inducer diameter (inlet area) is selected to limit the fluid velocity so that the available inlet
pressure is considerably higher than the dynamic pressure of the fluid, typically 200% for LOX and 100%
for LH2. Once the inlet conditions are fixed, the shaft speed will be set such that cavitation at the blade
tips is avoided. Increasing blade thickness to react pressure and centrifugal forces sets another constraint
for tip speed since it decreases flow passage area and thus reduces the suction performance. Small flow
rate pumps are generally less efficient than large flow rare pumps because the clearance and surface finish
related losses cannot be scaled with size.
6 - 12
RTO-EN-AVT-150

2.2.3
Turbines
Optimum turbine efficiency requires a certain pitchline velocity which is a product of the shaft speed
and the turbine diameter. The minimum weight turbine has the highest speed and smallest diameter within
the structural and mechanical arrangement limitations.
There are two types of turbines, impulse and reaction turbines which in rocketry may have one or
more stages. Each stage consists of a stator part where the flow expands and accelerates and a rotor part
where the impulse on the blades sets the rotor in motion thus transforming kinetic in mechanic energy.
The impulse turbine is a constant pressure device since the area ratio between the blades is constant and
ideally the only the direction of the flow changes. This type is usually applied in cases characterized by a
high pressure drop and a small mass flow rate, i.e, gas generator cycle engine. Turbines with more stages
may be further distinguished in velocity compound, expansion only the first stator, or pressure compound,
expansion in every stator stage.
Reaction turbines distribute the expansion between stator and rotor and hence, the fluid passages
between the blades have a converging cross section. This type of turbine is usually applied in expander or
staged combustion cycle systems, typical reaction grade of 50% meaning half of the power of comes from
the expansion, where a minimum pressure loss between pump and combustion chamber is almost
mandatory. The efficiency of a turbine is a function of number of stages n and the velocity ratio u/c0,
which relates circumferential velocity at the mean diameter to the fluid velocity for an isentropic
expansion between turbine entry and exit pressure.
n
u
= f
c0
u
with
=
c0
2
i
(5)
c0
Figure 7 shows the turbine efficiency as a function of the velocity ratio and reveals that the type of turbine
too has an impact on the efficiency. Practical turbines have an about 5 and 10% smaller efficiency since
the relations given above do not account for losses due to tip clearance.
100
90
Turbine efficiency [%]
80
70
60
50
40
50% Reaction
Impulse (zero reaction)
Radial inflow turbine
30
20
10
0
0
0,2
0,4
0,6
0,8
Turbine velocity ratio u/c [-]
Figure 7: Single stage turbine efficiency [12]
RTO-EN-AVT-150
6 - 13

2.2.4
Materials for turbomachinery components
In the design of turbomachinery various materials, Aluminium alloys, stainless steels, high strength
steels, nickel base alloys, cobalt base alloys and titanium alloys are in use. Low pressure components are
typically cast from Aluminium to avoid welds. Nickel-base superalloys such as Inconel 718 are applied
used to cast pressure vessels where higher strength is required. The high strength-to-weight ratio of
titanium is utilized to obtain the high tip speeds required for LH2 impellers and inducers. The high
pressure high temperature hydrogen-rich steam may cause hydrogen embrittlement and thus requires the
protection of high-strength superalloys with appropriate copper or gold plating. Turbine blades are often
directionally solidified and thermally coated to cope with the extremely high heat fluxes. LO2 pumps
where contact with the inducer or impeller could result in ignition are coated i.e. with silver. Similar
materials and techniques are also used for potential contact with titanium impellers to preclude the
formation of titanium hydrides due to heat generation. The LOX turbopump of the Vulcain 2 engine
shown in figure 8 is a rather new example of an entire turbopump set, turbine plus pump, which
demonstrates again the compactness of the design.
Figure 8: LOX turbopump of the Vulcain 2 engine [12]
2.2.5
Bearings and seals
Aside pump and turbine any turbopump requires components such as bearings, seals, gears and the
suction and discharge ducts. There are two kind of bearings, rolling elements and fluid film ones and they
have to perform three primary functions, radial control of the rotor to prevent rubs and to maintain
clearance, axial control of the rotor to maintain rotor control during transient mechanic and thermal loads
and to react residual thrust loads and third, control of rotordynamics to provide for adequate radial
stiffness and damping. They are usually cooled by the fluid which quite often doesnt provide for
lubrication, operate at high speeds and are exposed to high transient radial and axial loads. While static
seal design in rocketry generally doesnt exceed standard procedures, dynamic seals which separate
stationary and rotating parts are critical. Dynamic seals just must not fail. There are different types of
seals, labyrinth, face contact or shaft contacting, floating ring or hydrodynamic face seals. Major design
problems are fluid compatibility, thermal gradients and dynamic loads.
6 - 14
RTO-EN-AVT-150
2.3
Thrust chamber assembly
In a liquid propellant rocket engine, the function of the thrust chamber assembly is to generate thrust
by efficiently converting the propellant chemical energy into hot gas kinetic energy. That conversion is
accomplished by the combustion of the propellants in the combustion chamber, followed by acceleration
of the hot gas through a convergent/divergent nozzle to achieve high gas velocities and thrust. The basic
elements of the thrust chamber assembly of a rocket engine consists of the following components which
are closely coupled in their functioning and therefore have to be designed in parallel: the cardan, welded to
the injector head which is which holds the propellant distribution manifolds and the ignition system and
the combustion chamber liner including the throat section and the first part of the nozzle to which the inlet
manifold for the coolant is welded. Looking at figure 9, a sketch of the Vulcain thrust chamber, simplifies
the explanation of these parts.
Figure 9: Cut through the Vulcain thrust chamber [12]
The cardan on top of the LOX dome which allows thrust vector control through gimballing, connects
the engine with the launcher. The injector head has various functions, it distributes the incoming
propellants according to the needs of the injectors, holds the ignition system and the propellant injectors
which decouple the combustion chamber from the fluid supply system. The liquid oxygen enters directly
RTO-EN-AVT-150
6 - 15

from the turbopump while the hydrogen has already passed the cooling channels and the outlet manifold.
At the lower end of the regenerative cooled liner the distribution manifold of the fuel is welded and which
acts also partially as a mount for the nozzle extension.
2.3.1
Ignition system
A successful ignition is the outcome of matching conditions of propellant mass flow rates and mixture
ratio and initiation energy in time and in space. However, the ignition of a rocket engine has also to be
smooth with negligible pressure peaks to reduce the risk of triggering combustion instabilities or
damaging the cooling channels. Combustion chamber pressure peaks may cause a partial blockage of the
propellant flow inside the chamber yielding extinction or damage the turbopumps. Hence, exact timing of
ignition initiation and propellants valve opening is essential. Furthermore, the local mixture ratio of the
propellants has to be sufficiently far inside the flammability limits, the energy provided by the ignition
system has be high enough to vaporize and preheat the gaseous propellants above the ignition temperature.
Once a chemical reaction has started, the local heat release has to be sufficiently high to overcome the
losses. With propellant temperatures below 100 K the process of vaporization and heating especially
during the start-up transient may exceed a couple of milliseconds and hence flame extinction may occur
locally. In combination with the high propellant mass flow rates even a few milliseconds may locally yield
a sufficient amount of premixed propellants which may have catastrophic consequences. Therefore a
coupled development of injection and ignition system is mandatory for a successful design. There are
quite a large number of ignition devices which either use pyrotechnical propellant charges, hypergolic lead
or a torch ignition system with gaseous propellants. An example of a solid propellant based system, the
pyrotechnical ignitor of the Vulcain engine is shown in Figure 10.
Figure 10: Pyrotechnic ignition system of the Vulcain engine
The methodology aims at a prediction of the propellant mass flow rates in order to establish a mixture
ratio map. Second, the hot exhaust gases from the ignition system are injected and mixed with the
propellants without a prediction of the chemical reaction. The main aim is to establish a temperature map
inside the combustion chamber. Based on the results of mixture ratio and temperature distributions a map
of local ignition probabilities is predicted and successful ignition is assumed when the area where ignition
probability exceed a predefined limit is sufficiently high.
2.3.2
Injector head
As part of the thrust chamber assembly, the essential function of the injector head assembly is to
uniformly inject liquid propellants into the combustion chamber at the proper oxidizer/fuel mixture ratio.
6 - 16
RTO-EN-AVT-150

The geometry of the propellant distribution manifolds should dampen secondary flows resulting from the
turbopump or piping systems. Injector heads in engines which work with storable or very cold cryogenic
propellants are frequently equipped with baffles or cavities which work as damping devices. Furthermore,
injection elements have a significant pressure loss, about 15-20% of the combustion pressure, work as
damping elements and at least partially decouple dynamically the propellant feed system from the
combustion chamber.
There are four different basic concepts of propellant injection: impinging injection, swirl injection,
parallel injection in form of a showerhead, and shear coax injection and combinations of these concepts,
i.e. a swirl coax injector which can be found in the HM-7B engines or a swirl and impinging as in the
AESTUS engine. Any injector has to introduce the propellants into the combustion chamber in a manner
to promote propellant mixing and droplet atomization by either impinging propellant streams, swirling,
shear mixing or other mechanical means of achieving maximum atomization. Within the combustion
chamber, ignition is achieved and vaporization of the atomized droplet is initiated by heat transfer from
the surrounding 3500 K hot gas. The size and velocity of the droplets change continuously during their
entrainment and acceleration in the hot gas combustion flow. As the combustion process progresses, the
vaporized propellants are mixed rapidly, heated and react, thereby increasing the gas phase flow rate
within the combustion chamber. Typically, a well-designed injector atomizes a large population of
propellant droplets to less than 100 microns in diameter. Therefore, with proper injector and combustion
chamber design, all droplet vaporization and combustion is completed before the combustion gas enters
the chamber's narrow throat area.
A major concern of any injector is the injector/wall interaction. In the vicinity of the face plate where
propellant mixing is poor oxidizer-rich gases mixed with cryogenic droplets may get in contact with the
combustion chamber walls. The result of this process, a combined physical and chemical attack, clearly
visible on the wall of the combustion chamber liner of the Vulcain engine shown in figure 11, is called
blanching. There are two methods to cope with this problem, first, injector trimming where the outmost
row of the injectors operates with a smaller mixture ratio to reduce gas temperature and avoid oxidizerrich gas hit the surface, and, second, injection of a coolant film which has a similar effect.
Figure 11: Footprint of Injector/Wall Interaction on the combustion

chamber liner of the Vulcain engine
RTO-EN-AVT-150
6 - 17

In cases where both propellants enter the injector in a liquid state the injector of choice would be
anyone of a kind of an impinging injector. An injector is called like-on-like if propellants of the same kind
impinge on one another and like-on-unlike if fuel and oxidizer jets impinge. The working principle of an
impinging injector is as follows: once the jets impinge on one another they form a liquid sheet which
either decomposes into smaller fragments and atomizes and, in case of hypergolic propellants, reacts
simultaneously (like-on-unlike) or interacts with another liquid sheet and the atomizes and reacts (like-onlike). The establishment of a lifted flame and a considerably large volume of partially premixed
propellants is a major drawback of any impinging injector. The typical high mass flow rates establish
within milliseconds a large enough amount of propellants which when they react suddenly may lead to
pressure peaks which in turn may trigger combustion instabilities.
Swirl injectors which introduce a tangential velocity component to both propellants are typical for
small thrusters are often some type of coax injector. They usually create due to the numerous small
recirculation zones in the vicinity of the face plate a very good mixing and combustion performance and
allow as well for a simple establishment of a cooling film. The major drawbacks are the comparatively
large pressure drop, the coupling between performance and thermal loads to the chamber walls and the
very stringent precision requirements for production. Nevertheless, there are large engine applications as
well such as the RD-170 engine family where this kind of elements are used among others to actively
control combustion stability by locally introducing via a this different kind of atomization a different
proplet size distribution of the liquid thus altering the propellant mixing and breaking any symmetry
established in the combustion chamber otherwise.
Shear coax injectors with or without swirl are the element of choice when one of the propellant is in
its liquid phase. The dominating physical phenomena in a shear injector are the destabilizing aerodynamic
forces which act on the liquid jet which may be characterized by dimensionless numbers such as velocity
ratio of the propellantsl, the appropriate Reynolds and Weber numbers, and, finally the momentum flux
ratio.
V =
wg
wl
dw
Re =
;
We = g
(w
wl )
(w )
J=
(w )
2
(6)
It is important to note that the liquid oxygen is injected below its critical temperature but above its critical
pressure, and, as a consequence, atomization in the near injector region is additionally influenced by
thermodynamic effects.
The basic functional principles of a shear coax injector can be explained using figure 12. The area
constriction at the entrance of the liquid oxygen guarantees a homogenous mass flow rate for every
element and is the main decoupling mechanism. The restriction and the expansion afterwards introduce
large scale turbulences in the liquid which support the atomization. Prior to the exit some of the gaseous
hydrogen is injected into the liquid which has two effects. First, it increases the pressure drop and thus
introduces engine throttling capabilities without jeopardizing the combustion performance and second it
improves atomization and mixing. The final exit cone has two further specifics, a tapering angle which
lays somewhere between 5 and 10 and a recess of a few millimetres. Both measures will again support
atomization and mixing. Typical exit velocities of liquid propellants range between 10 m/s 20 m/s. The
gaseous hydrogen enters the injector and is finally accelerated at the edge to velocities in the range of 250
m/s 350 m/s.
6 - 18
RTO-EN-AVT-150
LO2
GH2
GH2
Figure 12: Schematic of a shear coax injector element
Figure 13 offers a view inside the fuel dome of the Vulcain 2 engine. A dense package of more than
550 injection elements homogenizes the hydrogen supply. Obviously precision mounting of all these
elements is a time consuming and costly process. The exit of the ignition gas is visible in the centre.
Figure 13: Row of LOX injection elements inside the hydrogen distribution dome
Finally, a comparison of different injectors is given in table 6 which summarizes some characteristic
data of large liquid propellant rocket engines, F-1, RD-170, SSME, RD-0120, LE-7A, RS-68, and Vulcain
2.
This offers interesting comparison since the first two engines are LOX/ kerosene one gas generator and the
other staged combustion cycle, the following three staged combustion engines are working with LOX/
LH2 and the final two work both the same propellants but are both gas generator cycle engines.
RTO-EN-AVT-150
6 - 19
Table 7: Characteristic data of sample injectors [10]
Engine
F-1
RD-170
SSME
RD-0120
LE-7A
RS-68
Vulcain 2
6.9
7.6
1.8
1.52
0.86
2.9
0.94
Propellant
combination
LOX /
RP-1
LOX /
kerosene
LOX /
LH2
LOX /
LH2
LOX /
LH2
LOX /
LH2
LOX /
LH2
Injector type
Impinging
Swirl
coax
Shear
coax
Shear
coax
Shear
coax
Shear
coax
Shear
coax
Flow rate /
element [kg/s]
1.7
1.8
0.9
~1
0.85
~0.9
0.55
Thrust /
element [kN]
4.6
5.5
3.8
3.4
3.1
~ 3.2
1.6
Thrust (sl)
[MN]
2.3.3
Combustion chamber
As an integral and may be the most important component of the thrust chamber assembly, the
combustion chamber must be specifically designed to satisfy the operating requirements of engine. The
basic components of a combustion chamber are the coolant inlet and discharge manifolds, a cooled
internal liner consisting of tubes or channels and an external structural assembly which is capable to carry
all structural and flight loads such as thrust, pressure, and vibrations. The contour of the inner liner is
aerodynamically designed to provide maximum performance and thrust for the engine operating
conditions.
2.3.3.1
Combustion chamber cooling techniques
Combustion chambers are generally categorized by the cooling method or the configuration of the
coolant passages, where the coolant pressure inside may be as high as 30 MPa. The high combustion
temperatures which may exceed 3600 K and the high heat transfer rates, peak values may reach 100
MW/m2 encountered in a combustion chamber present a formidable challenge to the designer. To meet
this challenge, several chamber cooling techniques have been utilized successfully.
Regenerative cooling is the most widely used method of cooling a combustion chamber and is
accomplished by flowing high-velocity coolant over the back side of the chamber hot gas wall to
convectively cool the hot gas liner. The coolant with the heat input from cooling the liner is then
discharged into the injector and utilized as a propellant. Quite a large number of engines have a "tubular
wall" combustion chamber design, i.e. H-1, J-2, F-1 and RS-27 [1]. The primary advantage of the design is
its light weight and the large experience base that has accrued however this method exceed its application
limits at pressures in the range of 10 MPa. The best solution to date is the "channel wall" design, so named
because the hot gas wall cooling is accomplished by flowing coolant through rectangular channels, which
are machined or formed into a hot gas liner fabricated from a high-conductivity material, i.e. oxygen-free
copper or copper alloys such as Narloy Z (CuAgZr), see figure 14 which shows a cut through the
combustion chamber wall of the Vulcain engine.
6 - 20
RTO-EN-AVT-150
Figure 14: Cut through a combustion chamber wall with CuAgZr

liner and galvanic deposited Ni outer shell
Thrust chamber liner design is a complex process and necessitates a coupled solution of the hot gas
side heat transfer, the heat transport inside the walls and the coolant side heat transfer. Due to the large
thermal gradients plastic deformation occurs which has to be taken into account for an analysis of the
cyclic life of the liner. Cooling channel geometry design has to account for the thermal blockage caused
by the high heat fluxes and the comparatively drastic change in the fluid properties, heat capacity, density
and viscosity, which take place at cryogenic temperature and high pressures.
All the design bureaus and companies apply Nusselt-type correlations to describe the heat transfer
since a fully numerical solution of the coupled problems is far to complex. Major hurdles are the large
differences in the length and time scales, chemical non-equilibrium or finite rate chemistry, the continuous
transformation from sub- to supersonic velocities, and still missing models which describe satisfactorily
dominating phenomena such as fluid atomization, near critical fluid behaviour. Additionally, effects such
as dissociation due to high temperatures in the combustion chamber and recombination with appropriate
heat release in the cooler wall boundary layer and, finally catalytic effects at the liner surface are far too
complex and time consuming to allow for a coupled numerical solution. Usually the hot gas side heat
transfer is described in form of a Bartz-type correlation [13]:
Nu = 0,062 Re 0,8 Pr 0,3 (107 < Re 108, Pr~ 0,5)
(7)
There are various modifications around, almost every company works with their one correlation, which try
to account for local effects such as curvatures of liner and throat, area ratio, Mach number which all have
an influence on the heat transfer:
0,026 0, 2c p ( p ) g D
0,6 c ns t
0, 2
Dt Pr ns c * R
g =
0,68
1 T k 1 2 1
= wg 1+
Ma +
(
)
2
T
2
2
c
ns
RTO-EN-AVT-150
0,12
k 1 2
1+ 2 Ma
; Pr =
At
with
A
0,9
(8)
4k
and = 46 ,6 10 10 M 0 , 5T 0 , 6
9k 5
6 - 21

Nusselt-type correlations are as well applied for the prediction of the coolant side heat transfer. These
are usually verified nowadays by 3D CFD calculations which are still very demanding due to the operating
conditions. Especially the continuous but varying asymmetric heat input which results in drastic changes
of the local fluid properties (10-30% in density, heat capacity or viscosity are possible). Such variations
will have an influence on the pressure loss distribution along the cooling channel and thus also on the local
heat pickup. In order to minimize the pressure drop necessary for cooling the cross section of the cooling
channels is adjusted to the local heat loads on the hot gas side. Obviously only a coupled iterative
approach may yield the final solution of the appropriate liner and cooling channel design. A rather simple
example of such a correlation may read as:
T
Nu = K Re Pr w
Tb
a
2D
1 +
(9)
or a little bit more complex
Nu = K Re a Pr b
w
k

kw
cp
c
p
p 2D
1 +
L
p cr
(10)
Table 8 below summarizes the coefficients of the correlation (eqn. 10) for different combustion chamber
pressures, temperatures, cooling channel velocities, heat fluxes and propellant type.
Table 8: Coefficients for Nusselt correlations coolant side heat transfer
Fuel
Pc
Uch
q
Tb
Tw
K
a
b
c
d
K
MPa m/s MW/m2 K
420-810
0.00538 0.80 0.4 -0.125 0.242
C2H6 13.7 6-30
3-12.4 15-45 0.3-16.4 120-395
0.00545 0.90 0.4 -1.1 0.23
7-13.8 30-60 3.1-18.2 230-300 230-810 0.0280 0.80 0.4 0
0
0.00696 0.88 0.5 0
0
0.0220 0.80 0.4 0
0
CH4 27-34 55-238 2.6-139 146-275
27-34 55-238 2.6-139 146-275
0.0230 0.80 0.4 0
0
7-13.8 30-60 3.1-18.2 230-300 230-810 0.0230 0.80 0.4 0
0
0.0230 0.80 0.4 0
0
0.0230 0.80 0.4 0
0
RP-1 7-13.8 30-60 3.1-18.2 230-300 230-810 0.0440 0.76 0.4 0
0
7-13.8 30-60 3.1-18.2 230-300 230-810 0.0068 0.94 0.4 0
0
0.0056 0.90 0.4 0
0
0.193
0.27
0
0
0
0
0
0
0
0
0
0
0.395
0.53
0
0
0
0
0
0
0
0
0
0
-0.024
0
0
0
0
0
0
0
0
0
0
0
1
1
0
0
0
0
0
0
0
1
0
0
n
0
0
0
-1.0
-0.45
-0.57
0
-0.80
0
0
0
0
These coefficients should describe the influence of the entry conditions of the flow, cooling channel
geomery and any changes, curvature effects, catalytic surface reactions, influence of the strucutral heat
flow, thermodynamic (real gas, cryogenic conditions, vicinity to the critical point, varying fluid
properties), fluid mechanic (turbulence, stratification) or chemistry (pyrolysis). It worth mentioning that
all these coefficients are the result of experiments which surely are influenced by the experimental setup,
the sensors applied and the general operating conditions of the test [13, 14].
Film cooling provides protection from excessive heat by introducing a thin film of coolant or
propellant through orifices around the injector periphery or in the chamber wall near the injector or
chamber throat region. This method is typically used in high heat flux regions and in combination with
regenerative cooling. A sample correlation for film cooling prediction is given below:
6 - 22
RTO-EN-AVT-150
T aw T wg
T aw T co
= exp(
g
)
Fc pc C
(11)
Sample engines where film cooling is applied are the SSME, F-1, J-2, RS-27, Vulcain 2, and the RD171 and RD-180 with the latter two being the only ones where an additional cooling film is generated near
the throat [3].
Transpiration cooling, which is considered a special case of film cooling provides gaseous or liquid
coolant through the porous chamber liner wall at a rate sufficient to maintain the chamber hot gas wall to
the desired temperature. A typical correlation to determine the chamber wall is given in equation (12) [3]:
[ {
} ]
Taw Tco
0 ,1
= 1 + 1,18(Re b ) 1 Y Y Prm
Twg Tco
with
Y =e
G
37 c
Gg
( Re )
0 ,1
(12)
This technique was applied to cool the injector faces of the J-2 and the RS-44, as well the SSME.
However, the method has to date never been used to cool the chamber liner. The major reason might be
that transpiration cooling in combination with a metallic porous structure doesnt offer a big advantage
compared to conventional cooling since the maximum wall temperature will not change much and the
continuous adjustment of the coolant mass flow rate to the local thermal loads is difficult to achieve.
With ablative cooling, combustion gas-side wall material is successively sacrificed by melting,
vaporization and chemical changes to dissipate heat. As a result, relatively cool gases flow over the wall
surface, thus lowering the boundary-layer temperature and assisting the cooling process. Ablative cooling
may be applied either to the entire combustion chamber liner or to the throat section alone. Typically all
solid rocket boosters have ablative cooled nozzles. Sample chambers with partially ablative cooled throat
sections are the Viking or more recent the RS-68. Although successfully applied this cooling method has a
major drawback since it doesnt allow for modifications of the engine burn time.
A radiation cooled chamber transmits the heat from its outer surface, chamber or nozzle extension.
Radiation cooling is typically used for small thrust chambers with a high-temperature wall material
(refractory) and in low-heat flux regions, such as a nozzle extension. An example of an entirely radiation
cooled thrusters is the RS-21 Mariner/Viking Orbiter spacecraft, RL 10B or similar upper stage engines
have radiation cooled ceramic nozzles [1, 3].
2.3.3.2
Combustion chamber life
Cyclic fatigue of the liner material is one of the life limiting factors for a rocket engine. During the
start-up process the liner is first exposed to liquid hydrogen which causes the material to shrink and, since
the inner copper alloy liner cools down faster than the outer nickel shell, the fins of the cooling channels
are exposed to severe thermal stresses. At the end of the cool-down the entire chamber inner shrinks about
half the size of the cooling channel wall (Vulcain dimensions). After successful ignition, the temperature
and pressure in the combustion chamber increase very fast and the copper wall will expand accordingly.
Now, the nickel closeout is cooling by the liquid hydrogen and thus still at cryogenic temperatures. Hence,
the copper has to react to the thermal expansion alone which poses additional thermal stresses to the
material. With plastic deformation exceeding 3-4% a typical chamber has a life of less than 20 cycles.
Figure 15 shows the liner of the Vulcain engine with its typical failures at the cooling channels.
RTO-EN-AVT-150
6 - 23
Figure 15: Chamber liner with typical longitudinal failure
The methods available to predict cyclic life of a combustion chamber are highly based on experience
and experimental data. One reason is still missing detailed enough initial material properties and their
behaviour under operating conditions. Aside the above mentioned thermal stresses the liner material is
exposed to additional loads which will have an impact on material properties. First, mechanical loads such
as static pressure or pressure fluctuations or mechanical vibrations of various sources, second, material
weakening by hydrogen embrittlement, third, high temperature fatigue or creep, and last but not least a
chemical attack at the surface by OH or other radicals or a simple oxidation through exposure to oxygenrich gases. Anyone of these phenomena is difficult to access quantitatively in separate experiments and,
even more important there is no clear understanding about the interaction of these phenomena.
A typical approach for cyclic life prediction would be to determine the hot gas side and coolant side
heat transfer and then calculate the thermal field and the resulting stresses and strains in the material
iteratively. Then, based on the predicted strains coming from one cyclic load case the number of cycles to
failure is determined applying typical material failure relations derived from experiments, see figure 16
[15].
Normalized plastic strain
1
disk
flexion
compr.-tension
Manson-Coffin
0,1
0,01
0,1
1
10
100
1000
Cycles number N to rupture
10000
Figure 16: Cyclic material failure law
Quite often the thermal and mechanical analyses are performed in 2D and for a stationary thermal
field, however recent investigations have shown that 3D calculations and especially transient thermal
6 - 24
RTO-EN-AVT-150

analyses are important for reliable cyclic life predictions. The numerical prediction of total number of
cycles to failure is and will be for quite a while still out of reach for practical applications.
2.3.4
Nozzle
The nozzle uses the pressure generated in the combustion chamber to increase thrust by accelerating
the combustion gas to a high supersonic velocity. Commonly used nozzles are bell-shaped with a
parabolic contour which is either a truncated ideal or a thrust optimized one, see figure 17. While all
Russian rocket engines have nozzles with an ideal contour, American, European and Japanese engines fly
a thrust optimized nozzle design. The latter has a cross-sectional nozzle area which expands much faster
near the throat and thus results in a shorter nozzle at the same expansion ratio compared to an ideal nozzle.
This expansion however causes the formation of an internal shock and at certain stages during transient
start-up and shutdown a flow phenomenon develops which is described as restricted shock separation
since the flow separates but reattaches almost immediately again. Aside severe side loads, this
reattachment of a high speed hot jet poses high heat loads to the nozzles, too [16].
Figure 17: Typical bell-shaped nozzle
Nozzle extensions are seldom regenerative cooled more often the much simpler dump cooling is
applied. The most frequent method however is ablative or radiation cooling. Materials suited for these
applications are C/C or SiC structures, which infiltrated with poly-aromatic hydrocarbons and additionally
incorporated additives are decomposed in an endothermic reaction route.
Radiation cooling requires appropriate high temperature materials which often have a protective
coating to prevent oxidation. Only small thrusters for satellite applications have metallic nozzles build
from refractory metals such as Tungsten, Rhenium or other refractory metals. A design follows based on:
q = Twg
(13)
with , the emission coefficient, , the Stefan-Boltzmann radiation constant and Twg , the maximum hot
gas side wall temperature. It is worth mentioning that in case of radiation cooling, the engine itself, its
supply lines and especially sensors and cables have to be protected against high heat loads.
The key problem of main stage engines with long burn times is illustrated demonstrated in figure 18
which shows the change in specific impulse as a function of flight altitude for an engine with two different
expansion ratios 1 and 2. In order to avoid flow separation and induced heavy side loads during the first
flight phase engine 1 is designed with a short nozzle and thus has a relatively high impulse then. However,
RTO-EN-AVT-150
6 - 25
Spec. Impulse I sp [s]
this advantage wears out soon and in the later phases of the launch the impulse penalty may exceed 15%.
Engine 2 is designed for high impulse in the later flight phase and thus suffers from an impulse penalty
during the first flight phase. Anyhow, however the nozzle design may be sooner or later the impulse will
be less than possible.
Ideally adapted
nozzle
450
2
1
350
250
10
30
50
altitude [km]
Figure 18: Potential Isp penalty of main stage engines
2.3.5
Design Methodology
Current engines design techniques are based on a design methodology with load factors and safety
factors determined on more than four decades of experience, and are commonly gathered under the
heading of deterministic design. A deterministic structural analysis means that first, a load or condition is
solely based on a set of design points and load factors which account for some variability in the specific
load in order to be sure that the maximum load ever before will be accounted for, second, that values for
minimum material strength or allowable fatigue are taken that ensures variations in material properties are
taken into account, and, third, a safety factor is worked into the design which will cover all the remaining
unknowns in analysis, loads, fabrication or even human error. Finally, the product of the design load,
limiting factors, and safety factors must always be less than the strength or allowable fatigue to assure safe
operation. Usually, designs have to meet different conditions such as strength, fatigue, creep, deflection,
buckling or burst and therefore a typical design approach applies during the analysis maximum loads and
minimum strength conditions [17].
In typical design analyses, load factors are based on previous hardware failures that were initially,
analyzed by this same method - with factors that specified that none of the hardware was to fail. Taken all
component and sub-systems of an engine this approach is conservative one and usually yields part which
are over-designed. However, since the methodology aims at a no-failure design, there is no way to avoid a
partial over-design. Numerous design reviews, audits and quality checks during the entire process of
development are aimed at identifying analysis or manufacturing mistakes or integration, operation or load
conditions errors. However, extensive ground testing is still necessary since any new design may yield
load conditions and resulting failures that were unknown until the test [17, 18].
3.
ADVANCED CONCEPTS
Any new concept or design has to show the capability to improve reliability, performance or cost
effectiveness of the system or component and in recent years reliability and cost became dominating.
6 - 26
RTO-EN-AVT-150

Hence, current developments deal with low cost approaches such as the RS-68 or the more recent
European Vulcain X design studies. Within this chapter a series of new concepts and a novel design
methodology approach are discussed and their potential contribution to the above mentioned drivers will
be shown.
3.1
Advanced ignition devices
An ignition system is designed to meet the requirements of the mission, sea level, vacuum or multiple
ignitions and accounts for further boundary conditions such as propellant phase and temperature as well as
geometric and power constraints. Any improvement will have to account for these conditions and will
most likely be mission specific as well.
The simplest form of system improvement is a reduction of complexity by reducing the number of
parts. This may be combined with a reduction of component weight. An example could be to use the
original propellants instead of additional ones. For re-ignitable engines a simple pyrotechnic system isnt
sufficient and another form has to be used. Quite often gaseous pressurized propellants are applied which
are send to an ignition chamber at the appropriate mixture ratio and than injected into the combustion
chamber. Multiple ignitions require a sufficient amount of propellant and thus large enough tanks. The
entire system including valves ignition chamber a.s.o. may be omitted if another system with similar
reliability becomes available. Although not yet used in cars yet, the automotive industry currently
develops igniters for their internal combustion engines which apply semi-conductor lasers as we already
use in our MP3 or CD players [19]. A change from continuous wave to pulse mode operation allows for a
considerable increase in optical power. A laser-based ignition system offers the capability of distributed
ignition, the energy for ignition initiation can be delivered almost everywhere. This allows for an ignition
at places where unburned premixed propellants may accumulate, i.e. lifted flames, recirculation areas
within injectors. Hence, this method offers a potential for risk reduction towards initiation of combustion
instabilities. Nowadays, glass fibres are common for optical signal transmission. The combination of pulse
mode semi-conductor lasers optical fibres allows for a safer placement of the more sensitive parts of the
lasers and the above mentioned distributed ignition. This development seems worth considering since the
operating conditions in a car are demanding and the required reliability is extremely high.
There has been some effort at various places to develop this ignition technique. NASA has considered
laser-based ignition as the appropriate ignition method for the individual thrust cells of the XRS-2200, the
engine for the X-31 vehicle. Laser-based plasma generation and ignition of single injector cryogenic
propellants was investigated for example at DLR Lampoldshausen [20]. Figure 19 shows on its left side
on top the OH emission and below a Schlieren image with the dark spot just above the dark jet the resul
thigh temperature foot print of the generated plasma directly fter the laser shot. On its right side, the top
picture shows 800 microseconds later the OH emission of the developing flame front and below the
corresponding Schlieren image. It can be seen that the flame develops within the shear layer around the
oxygen jet before moving towards the injector face plate. At JAXA studies are underway which apply a
laser to heat a target in the chamber sufficiently fast so that the target glows and initiates ignition of the
propellant. The system might be improved if the surface of the target is activated to act as a catalyst or the
target is infiltrated with a liquid which when evaporated reacts hypergolic with one of the propellants.
RTO-EN-AVT-150
6 - 27
Figure 19: Laser-based ignition of a LOX/H2/ spray
3.2
Advanced injection techniques
High precision fabrication, qualification inspection and testing, high precision mounting and quality
assurance procedures for injector are time consuming and costly. There are two possible ways to alter the
procedures mentioned above, first, a reduction of the precision requirements through a simpler injector
design or, second, a reduction of the number of injection elements through an increase in the injector mass
flow rate.
3.2.1
High thrust/element injectors
An increase of the thrust per injection element requires improved atomization and mixing in order to
maintain similar combustion performance. In order to achieve a better atomization the area of interaction
between the gas and the liquid has to be increased. An improvement of the shear coax injector with the
liquid in the centre is the tri-coax injector which has a gas jet in the centre, an annual gap for the liquid and
another annular gap for the remaining part of the gas. The RD-170 and RD-180 already have the hot
oxygen-rich pre-burner gases in the centre and the kerosene in the annual gap around it. Examples of high
thrust elements, a cross coax and tri-coax injector, which are presently under investigation at Astrium are
shown in figure 20 [21]. A closer look at the thrust per element figures for different engines, see table 6,
reveals that engines such as the SSME reach a thrust/element of 3.8 kN out of a mass flow rate per
element of 0.9 kg/s compared to 1.6 kN out of 0.53 kg/s of the Vulcain 2 engine. Maintaining the
performance and increase the flow rate per element to SSME values would cut down the number of
elements by about 40%.
Figure 20: High thrust elements a) cross coax b) tri-coax
6 - 28
RTO-EN-AVT-150

3.2.2
Low precision injectors
An entirely different approach compared to the high precision requirement coax or tri-coax injectors is
an approach with a porous face plate through which all the gaseous propellants are fed into the combustion
chamber and large number of small LOX injectors. The advantage of this approach is that the dimensions
of each element are so small that deviations from the required dimensions are tolerable since each element
delivers something in the order of 0.1% of the total oxygen mass flow rate to the chamber and deviations
affect only negligible parts of the combustion chamber volume.
It is worth mentioning that the atomization process with this kind of injector is different compared to
the one with conventional shear coax injectors. Distribution of the gaseous fuel all over the face plate
considerably lowers the hydrogen velocity which in turn reduces the available aerodynamic forces
available for atomization which makes this approach less attractive. However, a closer look at the local
conditions reveals that these are not that far away from typical atomization conditions in an internal
combustion engine where the liquid is injected at high speeds into an almost stagnant air. Porous face
plates have already been considered for the HM7B and have been tested in the J2 and SSME engines and
have been flight proven with the RD-0120 although the face plate might not be consider porous since it
has more than 15000 holes of less than half a millimetre in diameter. However, all these engines still have
conventional shear coax injectors.
Recently, DLR performed a series of sub-scale test campaigns with LOX/GH2 and LOX/GCH4
applying a porous sintered metallic (copper as well as stainless steel) and a rigimesh face plate, see figure
21, with a large number of LOX posts made of cheep pipe material. The holes for the LOX posts were
simply drilled into the porous plate and the LOX posts itself were electron beam welded into the LOX
dome.
Figure 21: Sub-scale injector head with porous rigimesh face plate
The results were promising insofar as combustion efficiency determined by chamber pressure
measurements turned out to be almost negligible. Furthermore, heat transfer data indicates a more uniform
heat release without any heat transfer problems to the face plate itself. Currently, investigations are
underway with aim to replace both the metallic face plate and the metallic LOX post with ceramic
materials and have an entire ceramic injector head.
3.3
Advanced thrust chamber technologies
As already stated earlier, the two approaches to improve current technologies, one driven by cost
reduction constraints and the other driven by performance considerations. Improvements which increase
RTO-EN-AVT-150
6 - 29

engine reliability may be considered as cost reduction related. Any advanced thrust chamber technology
will be related to heat load management, i.e. protective coatings, improved cooling, new materials or
fabrication techniques with increased cyclic life behaviour, high temperature heat transfer, cooling and
related component life [22-24]. It is also worth mentioning that regenerative cooling reaches its
application limits when local heat fluxes exceed 80 MW/m2. Typical high pressure engines apply as
already mentioned a combination of regenerative and film cooling. Another method to reduce the
temperature gradient across the chamber wall is deposit a thin thermal barrier coating, usually Y2O3stabilized ZrO2 on the hot gas side surface. In any case, the implementation of new materials and concepts
requires the adaptation of existing design rules and tools, the determination of material properties and
behaviour laws, appropriate failure detection techniques and predictive tools for failure propagation, and,
first of all, reliable and reproducible material production itself.
Among the different methods which are proposed in order to meet the cooling and life requirements
three approaches are worthwhile to be considered in more detail. Two more conservative approaches
which maintain the concept of regenerative cooling, one which still keeps the standard combustion
chamber liner material CuAgZr but protects this liner with a thermal barrier coating employing extreme
thin ceramic top layers combined with appropriate metallic bond coats and another one which replaces
both the conventional liner material and fabrication method by forming the liner via vacuum plasma
spraying of CuCrNb powder and a more advanced approach which combines the advantages of a high
temperature, low density and porous carbon-fibre reinforced carbon (C/C) based combustion chamber
liner with the concept of transpiration cooling using liquid hydrogen without regenerative cooling [25-33].
While the first two techniques allow for a reduction of the thermal loads of the chamber liner and therefore
aim at an increase of chamber life, which may always be traded if necessary in reduced requirements for
the turbopump by a reduction of the necessary cooling pressure drop, the latter approach has almost three
goals in parallel, increase of cyclic life, considerable reduction of engine weight and improved production
cost.
An intermediate approach where conventional regenerative cooling is combined with nonconventional fabricated hybrid liner structure, fibre reinforced ceramic matrix composites such as C/C or
SiC/SiC, were coated applying different deposition techniques with thin metallic layers (copper, rhenium
coated with tantalum) are already under investigation [25].
As has been shown earlier, main stage engines with long burn times usually suffer from an almost
continuous mismatch of the nozzle expansion ratio to ambient pressure. Taking into account that none of
the various plug nozzle concepts which aim at a continuous adaptation of the expansion ratio to altitude
has made it from development into flight qualification [34, 35] the only remaining concept is that of a dual
bell, an extendable nozzle exit cone or an expendable nozzle insert [36].
3.3.1
C/C Thrust chamber liner with effusion (transpiration) cooling
In order to fully realize the potentials of CMC materials the final development goal would be an
integral thrust chamber with a CMC liner with an outer load bearing carbon-fibre reinforced polymer
(CFRP) and a CMC injection head. The main challenge of transpiration cooling is the continuous
adaptation of the local coolant supply to the varying heat fluxes. At each surface element of the liner static
and dynamic pressure of both coolant and hot gas have to be balanced to prevent entrainment of hot gases
into the porous C/C structure. A further key element will be the injector head including the injector
elements which have to be different in design since the heat pickup of the propellant in the cooling
channels will be missing. Especially the near injector region of the liner may need additional protection
against a potential chemical attack by oxygen molecules or OH radicals.
The principle of effusion cooling bases on the following two mechanisms: Passing through the porous
wall the coolant absorbs the heat conducted into the solid material of the wall. In this way, the porous wall
6 - 30
RTO-EN-AVT-150

may be considered as a counter flow heat exchanger. Heat conducted into the wall from the hot gas side is
transported backwards into the hot gas. The coolant having passed through the porous wall forms a fluid
film on the hot gas side wall surface and will act as a typical cooling film. Hence, a part of the convective
wall heat flux is absorbed by this film. The heated coolant at the film surface is transported downstream by
the momentum of the hot gas flow, partially entrained into the main flow and continuously replaced with
new coolant coming out of the wall.
The porous CMC is a carbon-fibre reinforced carbon (C/C) which has a high temperature resistance of
about 2500 K. The porosity can be adjusted to the application needs altering either the fabrication
procedure or the raw material. Another advantage of C/C is its almost negligible thermal expansion which
makes it resistive to thermal shocks or thermal cyclic fatigue. Typical high pressure rocket engines require
a sufficient high Reynolds number in the cooling channels to assure the necessary cooling efficiency and
to keep the wall temperatures below critical values. Pressure losses to assure sufficient cooling typically
may even exceed 10 MPa while for an effusion cooled liner less than 0.5 MPa will typically be sufficient
to keep the inner liner wall at tolerable temperatures. This possible pressure potential can be traded within
the system for a reduced turbopump discharge pressure, lower gas generator mass flow rates, an increased
engines performance or reliability or simply to enlarge the system margins.
DLR has been active for years to develop the fundamental and technological basis for this approach
[37]. Material and cooling technology have been verified in different test campaigns. Key objectives were
thermal cyclic fatigue tests at typical temperatures and pressure levels between 0.5 and 1 MPa, high
temperature failure tests which exceeded 2500 K and sub-scale tests at P8 with typical mixture ratios and
chamber pressures up to 10 MPa. One of the development goals was to overcome a potential disadvantage
of this cooling concept, the loss in specific impulse. The part of the propellant which is used as coolant
will enter the combustion chamber along the liner to the throat with continuously decreasing pressure and
subsequently will contribute less and less to the overall thrust.
Although the boundary layer of a typical conventional regenerative cooled liner has quite lower
temperatures compared to a effusion cooled C/C liner the resulting loss in specific impulse is still nonnegligible. A combined analysis of the problem however yielded a promising result. The specific impulse
may be omitted as long as the coolant flow rates can be kept below a limit of 0.7% of the total mass flow
rate. Figure 22 shows the surface temperature of the C/C as a function of the coolant mass flow rate and
reveals that the material allows even for flow as low as 0.5 % and still keeping the temperature threshold
of 2500 K.
Figure 22: Operating regime of a transpiration cooled C/C liner
RTO-EN-AVT-150
6 - 31

3.3.2
Thermal Barrier Coating (TBC)
There exist various methods to build up protective layers for thermal protection. Among them the
following techniques have been demonstrated successfully that they are suited for copper alloy substrates
typical for liquid rocket engine applications: electron beam physical vapour deposition (EB-PVD),
vacuum plasma spraying (VPS) and solution plasma spraying (SPS) [38-41]. A segmented sub-scale
model combustor with a EB-PVD zirconia TBC was tested successfully under representative operating
conditions at DLR [42].
The functional principle of a TBC, illustrated in figure 23, is to increase the hot gas side wall
temperature TW. This is achieved by applying a ceramic top layer, i.e. partially Yttrium stabilized Zirconia
(PYSZ), onto the copper wall. The PYSZ ceramic offers a thermal conductivity of about 1.5 W/(m K).
In order to prevent a coating overheat (> 1500 K) coating thicknesses of less than 50 m are required.
These high hot gas side wall temperatures compared to the 800 K maximum temperature of the uncoated
copper wall the convective wall heat flux is reduced remarkably.
Figure 23: Functional principle of a thermal barrier coating
There are several reasons for an application of a TBC to the liner of a combustion chamber, to provide
a protective layer against chemical attack of oxygen or OH radicals and hydrogen embrittlement and to
protect the copper liner from high temperatures and thermal shocks. With such a protective layer, the
application range of existing regenerative cooling thrust chambers can be enlarged towards higher
combustion chamber pressures or an increased thrust chamber life. In general a TBC acts as an insulator to
the inner liner, reducing the wall heat flux into the copper base material.
3.3.3
New materials and processes for conventional regenerative cooled chamber lines
Within the last years near net-shape formation of thrust chamber liners or nozzle throat sections has
been discussed intensively in the literature. Aside tungsten or rhenium which were proposed for satellite or
apogee application only recently high temperature [43, 44] high strength alloys such as CuCrNb are
proposed and developed which seem to be able to replace the standard CuAgZr material. This material
attracted increasing interest all over the world and in the meanwhile the technology and the material seem
to be ready to be applied for thrust chamber applications [45, 46]. However, as far as the open literature
shows only sub-scale level or component level testing of thermal spray formed hardware have been
performed yet [47, 48]. Results about test campaigns of high thrust engines are not available in open
literature. One of the reasons may be that the problems of identification, quantification and detailed
control of all the process parameters which may have an impact on key material properties such as
elasticity and strength, resistance to annealing or creep may not be solved yet. Furthermore, most of the
6 - 32
RTO-EN-AVT-150

engineering design tools which are used to predict the low cycle fatigue and failure of thrust chamber
liners have to be adjusted to this new material.
3.3.4
High area ratio cooling channels
The basic idea behind high aspect ratio cooling channels is to enhance the heat transfer and thus
influence the temperature gradients [49, 50]. A reduction in channel width will allow a larger number of
channels and longer fins and both effects will increase the heat transfer area. Furthermore, narrower
cooling channels allow for thinner liner walls, see figure 24. The limits for the liner walls thickness are
given by the finite non-homogeneities of the liner material and by the requirements of conventional
milling tools. To date thicknesses of less than 0.6 mm are hardly achievable.
Low aspect ratio design
High aspect ratio design
Figure 24: Sketch of a conventional and high aspect ratio cooling channel
DLR recently build and hot-fire tested a modular, sub-scale combustion chamber which had
incorporated in one segment with four different aspect ratios of cooling channels to determine the
effect of different aspect ratios on the heat transfer and especially on the temperature of the hot
gas side liner [51].
3.3.5
Advanced nozzles concepts
As already mentioned in chapter 2.3.4, any rocket engine with a burn time considerably higher than
that of a standard booster engine (~150 s) will suffer from an off design penalty of its nozzle due to the
changing ambient pressure conditions during ascent. The dual bell nozzle concept, see figure 25, aims at a
reduction of this performance losses. During sea level pressure conditions the flow separates at the contour
inflection of the nozzle and only the first part of it, the base nozzle experience fully attached flow
conditions. During accent of the launcher, the ambient pressure decreases and at a certain altitude the flow
jumps from the end of the first bell to the end of the nozzle extension, the second bell. The full flowing
dual bell nozzle now works under high altitude conditions. The period of transition is sensitive to outer
pressure field fluctuations and therefore of special interest. Figure 26 below demonstrates the advantage of
a dual bell in comparison with conventional bell nozzles with different expansion ratios. The light blue
curve shows the specific impulse of a booster engine with an expansion ratio of 32, while the green, black
and red curves show that of engines with expansion ratios of 45, Vulcain 2 (60), and 100, respectively. In
dark blue the specific impulse of a dual bell nozzle, designed with an expansion ratio of 45 for the first
bell and of 100 for the second bell are shown. Although the specific impulse drops considerably due to the
jump from first to the second bell this drawback is overcome later into the flight.
RTO-EN-AVT-150
6 - 33
Figure 25: Dual bell nozzle
440
Specific Impulse [s]
420
400
380
360
Vulcain 2
340
Area Ratio = 45
Area Ratio = 100
320
Area Ratio = 32
Specific Impulse Dual Bell
300
10
20
30
40
50
Altitude [km]
Figure 26: Specific impulse of nozzles with different expansion ratios
The experiments performed so far with cold gas facilities revealed that the jump and back-jump from
the first to the second bell and back are very fast, typically less than 10 ms. The separation from the first
bell is usually not entirely symmetric however, this asymmetry is limited to angles of less than 10 which
effectively limits the resulting side loads, too. Additionally, there is a hysteresis occurs effect insofar as
jump and back-jump dont occur at the same nozzle pressure ratio. While the jump with increasing
pressure ratio at a certain value, the back-jump at decreasing pressure ratios is delayed to a considerably
lower value [52]. This stabilizing effect increases the sensitivity of the nozzle against ambient pressure
fluctuations and thus enlarges the design margins of the nozzle and the nozzle operating conditions.
However, the establishment of safe jump conditions during ascent require an active control of the nozzle
pressure ratio by throttling of the engine.
Within the last 10 years the European FSCD consortium which features CNES, DLR, EADS, ESA,
ONERA, SNECMA, Volvo Aero Corp. investigated various nozzle contours in order to identify and
quantify the physical sources of flow separation and side-loads and to establish reliable criteria and models
to be implemented in engineering design tools [53, 54].
6 - 34
RTO-EN-AVT-150

3.3.6
Advanced design methodology
A clear alternative to the deterministic approach briefly described in chapter 2.3.5, is a probabilistic
design approach which considers the uncertainties of material properties, operating conditions and loads in
a more structured manner. In the probabilistic approach, design variables are neither seen as single values
nor are they weighed to an upper or lower bounds. They are instead represented by the actual distribution
or variation of the parameters. A distribution is a histogram of discrete values of the parameters or a
mathematical model that represents a smooth description of the variation. When the area under the
histogram or curve is normalized to a value of one, the function is called the probability density function.
The parameters are termed random variables. The distribution functions, then, are used to determine the
probability of occurrence of a given value of the random variable. A typical structural model for both
static and dynamic analyses bases on nominal hardware geometry. Resulting displacements, loads and
stresses are then compared to the material data, i.e. cyclic fatigue or ultimate strength. The material data
are taken as lower bound values of the available material test data. The ratios between the data and the
predicted values have to be within the specified safety margin. The predicted distributions show the
probability for a failure and not a value or safety. The sensitivity of each variable to failure which is part
of the result and this information helps the designer to get a better feeling for the impact of a variable. It is
essential to become knowledgeable of inherent risk of failure and assess it, identify the mayor causes and
minimize them within the design constraints.
A typical design process can be divided into conceptual design, preliminary design, detailed design
and development testing. The conceptual design process is performed in the classical deterministic way
and includes defining operating characteristics and configurations and the use of simplistic design
guidelines for definition of configuration. Initial sizing is based on deterministic analysis for primary loads
and the design features incorporate the essence of the load-carrying features of the hardware. Many details
are ignored at this stage and will be handled later and whenever appropriate, simple approximations are
utilized. The preliminary design typically entails a deterministic design analysis, analyses of failure modes
and a list of critical items. In a probabilistic design methodology, it is at this point that the initial decisions
are made as to which variables are crucial; an initial quantification of uncertainty is also determined. In
addition, the design must be critically reviewed to define possible failure modes and how they relate to the
uncertainty of variables. Typically the detailed design will still make use of deterministic analysis with
maximum loads, minimum properties and associated deterministic models and failure techniques. Each
element of a design requires a reliability estimate. Non-critical items where reliability is easily obtained, a
simple reliability estimator will be utilized. Critical items however require a detailed probabilistic analysis
which considers component load distributions, geometric tolerances and variations, material property
unknown or uncertainties, failure model characterization such as ultimate load, buckling or fatigue, and
allowances for human error, model error, fabrication and assembly. Using this detailed evaluation, loads,
responses and damage assessment will be quantified as a component reliability that has considered the
sensitivities of the hardware and uncertainties.
Generally, all probabilistic methods are approximate and the most common of the approximate
methods is the Monte Carlo simulation or a variation of it. Repeated numerical experiments are performed
that represent the total spectrum of the population of the problem. The method has an accuracy problem
when small probabilities are involved. The mean value is adequately represented with a reasonable
number of simulations. Sufficiently accurate solutions are with increasing numbers of simulations (less
then 10,000) and the method delivers perfect answers when 20,000 to 100,000 simulations can be done in
a reasonable amount of computer time. Obviously, the method has problems if each takes too time
consuming which quite often happens for large finite-element analysis.
Finally, important results from probabilistic analysis are the sensitivity responses which allow for a
quantitative ranking of different design variables relative to the deviation in the response of the structure.
The probabilistic approach identifies areas where tolerances might be loosened with negligible effect on
RTO-EN-AVT-150
6 - 35

reliability as well as areas where additional data is required to improve reliability. Although probabilistic
analysis requires more computational effort, the method provides much more information than
deterministic design and it is better suited to include uncertainties such as scatter in material data,
variations in thermal, mechanical, chemical static or dynamic loads tolerance in production precision.
Furthermore, it is a tool which allows for risk identification.
4.
CONCLUSIONS
Based on a description of the basic components and features of rocket engines areas have been
identified and the physical and technological limitations given in order to pave the way for a discussion of
possible improvements. The topics of gas generator or pre-burner as well as were not included in the areas
where potential advancements were proposed although some of the techniques proposed for thrust
chambers, ignition systems or injector should be applicable as well for other combustion devices.
The concept of an advanced laser-based ignition system was discussed and the current status given.
Additionally, an injection system design which aims at a much cheaper design with negligible impact on
performance with simple LOX posts and a porous face plate was proposed and sample results were given.
Various possibilities to improve thrust chamber cooling techniques involving new materials such as
CuCrNb, ceramic matrix composites or effusion cooling were explained, the different aspects shown.
Additionally, the advantages and the basic features of an advanced nozzle concept for a main stage engine,
the dual bell nozzle was described and the current statues of technology explained. Finally, a novel design
methodology, the probabilistic design approach was briefly described.
REFERENCES
[1] Sutton, G.P., History of Liquid Propellant Rocket Engines, ISBN 1-56347-649-5, American Institute
of Aeronautics and Astronautics, 2006
[2] Isakovic, S.J., Hopkins Jr., J.P., Hopkins, J.B., International Reference Guide to Space Launch
Systems, 3rd Ed. AIAA, 1999
[3] Huzel, D.K., Hwang, D.H., Modern Engineering for Design of Liquid Rocket Engines, ISBN 156347-013-6, American Institute of Aeronautics and Astronautics, 1992
[4] Sutton, J.P., Rocket Propulsion Elements, 3rd Edition, Wiley and Sons, New York, London, 1963
[5] NASA Facts, Next Generation Propulsion Technology: Integrated Powerhead Demonstrator,
Technology, FS-2005-01-05-MSFC, Pub 8-40355, 2005
[6] Cai, C., Jin, P., Yang, L., Du, Z., Xu, K., Experimental and Numerical Investigation of Gas-Gas
Injectors for Full Flow Stage Combustion Cycle Engine, AIAA-2005-3745, 2005.
[7] Davis, J.A., Campell, R.L., Advantages of a full-flow staged combustion cycle engine system, AIAA1997-3318, 1997
[8] M. Wade, Encyclopedia Astronautica, www.astronautix.com
[9] Farhangi, S., Hunt, K., Tuegel, L., Matthews, D., Fisher, S., Oxidizer-Rich Pre-burner for Advanced
Rocket Engine Application, AIAA 94-3260, 1994
[10] D. Haeseler, private communication
6 - 36
RTO-EN-AVT-150

[11] Cannon, J.L., Turbomachinery for Liquid Rocket Engines, Liquid Propulsion Systems - Evolution
and Advancements, American Institute of Aeronautics and Astronautics Professional Development Short
Course, 2003
[12] Albat, R., Langel, G., Haidn, O.J., Antriebssysteme, in Ley, Wittmann, Hallmann (Eds.)
Handbuch der Raumfahrttechnik, Hanser Verlag, to be published 2007
[13] Fisher, S.C., Popp, M., Quentmeyer, R.J., Thrust Chamber Cooling and Heat Transfer, in M.
Habiballah, M. Popp, V. Yang (Eds., LIQUID ROCKET COMBUSTION DEVICES Aspects of
Modeling, Analysis and Design, proceedings of he 2nd Int. Symposium on Liquid Rocket Propulsion,
Chatillon, 1995
[14] Liang, K., Yang, B., Zhang, Z., Investigation of heat transfer and coking characteristics of
hydrocarbons fuels, Journal of Propulsion and Power, Vol. 14, No. 5, 1998
[15] Riccius, J.R., Haidn, O.J., Zametaev, E.B., Influence of Time Dependent Effects on the Estimated Life
Time of Liquid Rocket Engines, Int. Symposium on Space Propulsion, Shanghai, Sept. 25-28, 2004, China
[16] Frey, M., Rydn, R., Alziary de Roquefort, Th., Hagemann, G., James, Ph., Kachler, Th., Reijasse,
Ph., Schwane, R., Stark, R.: European Cooperation on Flow Separation Control, 4th International
Conference on Launcher Technology, Liege, 2002.
[17] Hammond, W., Space Transportation: A System Approach to Analysis and Design, ISBN 1-56347032-2, American Institute of Aeronautics and Astronautics, 1999
[18] Newell, J.F., Rajagopal, K.R., Probabilistic Methodology A Design Tool for the Future,
www.engineeringatboeing.com/articles/probabalistic.htm, 1999
[19] Gerlinger, B., et al., Direkteinspritzender Ottomotor und Laserzndung, 5. Dresdner
Motorenkolloquium, 2003
[20] Gurliat O., Schmidt V., Haidn O.J., Oschwald M., Ignition of Cryogenic H2/LOX Sprays, Aerospace
Science and Technology, Vol. 7, Issue 7, pp. 517-531, 2003
[21] Haeseler, D., Mding, C., Rubinski, V., Khrissanfov, S., Kosmacheva, High Flow Rate Injection
Elements, 4th Int. Conference on Launcher Technology, Liege, 2002
[22] Popp, M., Schmidt, G., Rocket engine combustion chamber design concepts for enhanced life, AIAA1996-3303, 1996.
[23] Kindermann, R., Beyer, S., Sebald, T., Hollmann, C., Denkena, B., Friemuth, T., Kaufeld, M., Kolb,
U., Advanced Production and Process Technologies for current and future Thrust Chambers of Liquid
Rocket Engines, 4th International Conference on Launcher Technology, Liege, 2002.
[24] Quentmeyer, R.J., Rocket Combustion Chamber Life-Enhancing Design Concepts, AIAA-1990-2116,
1990.
[25] Elam, S., Effinger, M., Holmes, R., Lee, J., Jaskowiak, M., Lightweight Chambers Thrust Cell
Applications, AIAA 2000-3131, 2000.
[26] Shelley, J.S., LeClaire, R., Nichols, J., Metal-Matrix Composites for Liquid Rocket Engines, JOM,
Vol. 53 No. 4, pp. 18-21, 2001.
RTO-EN-AVT-150
6 - 37

[27] Dembinski, L., Grosdidier, T., Coddet, C., On the microstructure of vacuum plasma sprayed Cu-3%
Ag alloys, ITSC 2001 - Singapour - dite par C.C. Berndt - K.A. Khor et E.F. Lugscheider - ASM-TSS Materials park - OH-USA, p.633-637.
[28] Leonhardt, T., Hamister, M., Carln, J.C., Biaglow, J., Reed,B., Near-Net Shape Powder Metallurgy
Thruster, NASA TM-2001-210373, 2001
[29] May, L., Burkhardt, W.M., Transpiration Cooled Throat for hydrocarbon Rocket Engines, NASA
KEE6-FR, 1991
[30] Haeseler, D., Maeding, C., Rubinskiy, V., Gorokhov, V., Khrisanfov, S., Experimental Investigation
of Transpiration Cooled Hydrogen-Oxygen Subscale Combustion Chambers, AIAA 98-3364, 1998
[31] Meinert, J., Huhn, J., Serbest, E., Haidn, O.J.: Turbulent Boundary Layers with Foreign Gas
Transpiration, Journal of Spacecraft and Rockets, 2001, volume 38, number 2, pp. 191-198.
[32] Keener, D., Lenertz, J., Bowerson, R., Bowman, J., Transpiration Cooling Effects on Nozzle Heat
Transfer and Performance, Journal of Spacecraft and Rockets, Vol. 32, No. 6, 1995, pp. 981 985.
[33] Serbest, E., Haidn, O.J., Hald, H., Korger, G., Winkelmann, P., Fritscher, K., Advanced Technologies
and Materials for Future Liquid Rocket Engines, 12th European Aerospace Conference, Paris, 1999
[34] Hagemann, G., Immich, H., Nguyen TV., Dumnov, G.E., Advanced Rocket Nozzles, Journal of
Propulsion and Power, Vol. 14, No.5, pp. 620-634, 1998
[35] Korte, J.J., Salas, A.O., Dun, H.J., Alexandrov, N.M., Follet, W.W., Orient, G.E., Hadid, A.H.,
Multidisciplinary Approach to Aerospike Nozzle Design, NASA TM 110326, 1997.
[36] Frey, M., Hagemann, G., Critical Assessment of Dual-Bell Nozzles, Journal of Propulsion and Power,
Vol. 15, No. 1, pp. 137-143, 1999.
[37] Haidn O.J., Greuel, Herbertz, A., Ortelt, M., Hald, H., Application of Fiber Reinforced C/C Ceramic
Structures in Liquid Rocket Engines, SPACE CHALLENGE IN XXI CENTURY, Assovskiy I., Haidn OJ,
(Eds), ISBN 5-94588-036-1, Moscow 2005, pp. 46-72.
[38] Sebald, T., Beyer, S., Gawlitza, P., Mai, H., Pompe, W., Advanced Thermal Barrier Coatings for
High Heat Fluxes in Thrust Chambers of Liquid Rocket Engines, 4th International Conference on Launcher
Technology Space Launcher Propulsion, 2002, Lige.
[39] Gell, M., Xie, L., Ma, X., Jordan, E.H., Padture, N.P., Highly durable thermal barrier coatings made
by the solution precursor plasma spray process, Surface & Coatings Technology, 177-178 (2004) pp. 97102.
[40] Azzopardi, A., Mvrel, R., Saint-Ramond, B., Olson, E., Stiller, K., Influence of aging on structure
and thermal conductivity of Y-PSZ and Y-FSZ EB-PVD coatings, Surface and Coatings Technology 177178 (2004), pp. 131-139.
[41] Faraoun, H., Aourag, H., Grosdidier, T., Klein, D., Coddet, C., Development of modified embedded
atom potentials for the Cu-Ag system, Superlattices and Microstructures, Vol. 30, (5), pp. 261-271, 2001.
[42] Schulz, U., Fritscher, K., Peters, M., Greuel, D., Haidn O.J., Fabrication of TBC- armoured rocket
combustion chambers by EB-PVD methods and TLP, Science and Technology of Advanced Materials, 6,
2005, pp. 101 110.
6 - 38
RTO-EN-AVT-150

[43] Leonhardt, T., Hamister, M., Cerln, J.C., Biaglow, J., Reed, B., Near-Net Shape Powder Metallurgy
Rhenium Thruster, NASA/TM-2001-210373, 2001
[44] Flinn, E., Tough Coating for a Smoother Rocket Cone, Aerospace America, August 2004, pp. 24-25.
[45] Thomas-Ogbuji, L., Humphrey, D.L., Setlock, J.A., Oxidation-Reduction Resistance of Advanced
Copper Alloys, NASA/CR-2003-212549, 2003.
[46] Hickmann, R., McKechnie, T., Holmes, R., Elam, S., Material Properties of Net Shape, Vacuum
Plasma Sprayed GRCop-84, AIAA 2003-4612, 2003
[47] Ellis, D.L., GRCop-84: A High-Temperature Copper Alloy for High-Heat-Flux Applications,
NASA/TM-2005-213566.
[48] Verdy, C., Coddet, C., Thomas, J.-L., Cornu, D., Choulat, M., Highly Thickness Components for
Liquid Rocket Engine Produced by Low Pressure Plasma Spraying, International Thermal Spray
Conference, Seattle, 2006.
[49] Carlile, J.A, Quentmeyer, R.J., An Experimental Investigation of High-Aspect-Ratio Cooling
Passages, AIAA 92-3154, 1992
[50] Wadle, M.F., Comparison of High Aspect Ratio Cooling Channel Designs for a Rocket Combustion
Chamber with Development of an Optimum Design, NASA/TM-1998-206313, 1998
[51] Woschnak, A., Suslov, D., Oschwald, M., Experimental and Numerical Investigations of Thermal
Stratification Effects, AIAA 2003-4615, 2003.
[52] Stark, R., Haidn, O.J., Bhm, C., Zimmermann, H., Cold Flow Testing of Dual-Bell Nozzles in
Altitude Simulation Chambers, Proc. of EUCASS European Conference for Aerospace Sciences, Moscow,
2005.
[53] Frey, M., Rydn, R., Alziary de Roquefort, Th., Hagemann, G., James, Ph., Kachler, Th., Reijasse,
Ph., Schwane, R., Stark, R.: European Cooperation on Flow Separation Control, 4th International
Conference on Launcher Technology, Liege, 2002.
[54] Winterfeldt, L. Laumert, B., Tano, R., James, P, Geneau, F., Blasi, R., Hageman, G., Redesign of the
Vulcain 2 Nozzle Extension, AIAA-2005-4536, 2005
RTO-EN-AVT-150
6 - 39
6 - 40
RTO-EN-AVT-150

Advanced Rocket Engines

Uploaded by

Copyright:

Available Formats

Advanced Rocket Engines

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Advanced Rocket Engines

Uploaded by

Copyright:

Available Formats

Advanced Rocket Engines