Beruflich Dokumente
Kultur Dokumente
cooling of
liquid rocket
engine thrust
chambers
Marco Pizzarelli
ASI UNCLASSIFIED - pag 1 Agenzia Spaziale Italiana
Regenerative cooling of LRE thrust chambers
Outline:
Thrust chamber
environment (basics)
𝑻𝟎𝒈
𝑰𝒔𝒑 ~
𝕸
where
• 𝑇./ : temperature of the combustion
products
• 𝔐: molecular mass of the combustion
products
From (2)
ASI UNCLASSIFIED - pag 5 Agenzia Spaziale Italiana
Thrust chamber environment (basics)
The strength of most materials declines rapidly at high temperatures. For rocket engine
applications, the temperature where a material loses 60 to 75% of its room temperature
strength is often selected as the maximum allowable wall temperature. This temperature is
well below the material melting point
Thrust coefficient 𝑪𝒇 versus nozzle area ratio (note: 𝑰𝒔𝒑 ~𝑪𝒇 ) (2)
ASI UNCLASSIFIED - pag 7 Agenzia Spaziale Italiana
Thrust chamber environment (basics)
Combustion chamber temperature and pressures are “high”
Necessity to:
• cool the wall to a temperature considerably below its maximum allowable temperature
(i.e., much below the combustion temperature) or to stop operation before the wall
becomes too hot
• use high strength materials (the thickness, therefore the mass, of the thrust chamber
wall depends strongly on the stresses it can support)
Rocket engine cooling is necessary for strength considerations. In fact, enough heat has to
be absorbed to keep the walls at a sufficiently low temperature, so that the wall material is
strong enough to withstand the stresses imposed by the fluid pressure, thermal gradients,
and other loads. Consequently, the design of a thrust chamber is mainly a thermo-
mechanical problem that requires a proper characterization of the heat transfer and the
mechanical loads
ASI UNCLASSIFIED - pag 8 Agenzia Spaziale Italiana
Thrust chamber environment (basics)
Regenerative cooling:
the walls of large liquid-propellant rocket engines (that are always bi-propellant rockets)
usually consist of an array of suitably shaped tubes machined or brazed together to form the
thrust chamber. The fuel or the oxidizer is used as a coolant flowing in such tubes before it is
injected in the combustion chamber. This method is called regenerative cooling because of
the similarity to steam regenerators
RL-10 cutaway
ASI UNCLASSIFIED - pag 9 Agenzia Spaziale Italiana
Thrust chamber environment (basics)
Notes on regenerative cooling:
• This cooling technique mainly applies in mid-to-high
thrust levels because the heat flux to the chamber walls
increases as the hot-gas mass flow rate (and thus the
thrust) increases
• In regenerative cooling the heat absorbed by the coolant is not wasted; it augments the
initial energy content of the propellant prior to injection, increasing the exhaust velocity
slightly (0,1 to 1,5%) and most noticeably in small thrust chambers, where the wall-
surface-to-chamber volume ratio is relatively large
• Generally the fuel is used as coolant because of the tube wall oxidization concern if
oxidizer is used as coolant
ASI UNCLASSIFIED - pag 10 Agenzia Spaziale Italiana
Thrust chamber environment (basics)
From a thermal point of view, regenerative cooling consists of the steady flow of heat
from a hot gas through a solid wall to a cool fluid. In fact, the heat from the combustion
gases conducts through the walls –mainly in the radial direction- and is convected away by
the fluid flowing in the cooling channels
Such problem can be reduced to a one-dimensional cooling jacket model (ignoring the heat
transfer through the walls that separate the coolant passages)
From (1)
𝒬̇ or 𝑞> : wall heat transfer rate per unit area (or wall heat flux) [W/m2]
Velocity boundary layer (δ): large velocity gradients
Thermal boundary layer (δT ): large temperature gradients
𝜇𝑐E δ
𝑃𝑟 = =𝑓 (experimental evidence)
𝑘 δT
In most situations: δ~δT (hence Pr~1)
ASI UNCLASSIFIED - pag 13 Agenzia Spaziale Italiana
Convective heat transfer (fundaments)
The governing equations for steady, two-dimensional, incompressible flow with constant
properties and under the Prandtl’s hypothesis (thin boundary layer):
𝜕𝑢 𝜕𝑣
+ =0
𝜕𝑥 𝜕𝑦
𝜕𝑢 𝜕𝑢 1 𝑑𝑝 𝜕c𝑢
𝑢 +𝑣 =− +𝜐 c
𝜕𝑥 𝜕𝑦 𝜌 𝑑𝑥 𝜕𝑦
c
𝜕ℎ 𝜕ℎ 𝑑𝑝 𝜕c𝑇 𝜕𝑢
𝜌𝑢 + 𝜌𝑣 =𝑢 +𝑘 c+𝜇
𝜕𝑥 𝜕𝑦 𝑑𝑥 𝜕𝑦 𝜕𝑦
where:
𝑥 : direction parallel to the wall (velocity component 𝑢)
𝑦 : direction normal to the wall (velocity component 𝑣)
𝜌, 𝑝, 𝑇 : density, pressure and temperature
ℎ : specific enthalpy [J/K]
𝜇 : dynamic viscosity [Pa s]
𝜐 = 𝜇/𝜌 : kinematic viscosity [m2/s]
𝑘 : thermal conductivity [W/(m K)]
ASI UNCLASSIFIED - pag 14 Agenzia Spaziale Italiana
Convective heat transfer (fundaments)
Simplifications of the energy equation:
c
𝜕ℎ 𝜕ℎ 𝑑𝑝 𝜕c𝑇 𝜕𝑢
𝜌𝑢 + 𝜌𝑣 =𝑢 +𝑘 c+𝜇
𝜕𝑥 𝜕𝑦 𝑑𝑥 𝜕𝑦 𝜕𝑦
st c
• the viscous dissipation 𝜇 can be generally neglected (because of the low fluid velocity)
su
• in case of calorically perfect gas (i.e., ℎ = 𝑐E 𝑇) and no axial pressure gradient (e.g., flat plate)
it becomes:
𝜕𝑇 𝜕𝑇 𝜕c𝑇
𝑢 +𝑣 =𝛼 c
𝜕𝑥 𝜕𝑦 𝜕𝑦
where 𝛼 = 𝑘/𝜌𝑐E is the thermal diffusivity [m /s] 2
w
• in case of an incompressible fluid (i.e., 𝑑ℎ = 𝑐 d𝑇 + d𝑝) it becomes:
x
c
𝜕𝑇 𝜕𝑇 𝜕 𝑇
𝑢 +𝑣 =𝛼 c
𝜕𝑥 𝜕𝑦 𝜕𝑦
ASI UNCLASSIFIED - pag 15 Agenzia Spaziale Italiana
Convective heat transfer (fundaments)
In case of no axial pressure gradients (i.e., flat plate) the momentum and energy equations
becomes:
𝜕𝑢 𝜕𝑢 𝜕c𝑢
𝑢 +𝑣 =𝜐 c
𝜕𝑥 𝜕𝑦 𝜕𝑦
𝜕𝑇 𝜕𝑇 𝜕c𝑇
𝑢 +𝑣 =𝛼 c
𝜕𝑥 𝜕𝑦 𝜕𝑦
({~ |{} )
• 𝑢y = 𝑇 y → 𝑇(𝑦) = 𝑇> + 𝑢(𝑦)
z
st
• wall shear stress: 𝜏> = 𝜇
su >
s{ {} |{~ st
• wall heat flux: 𝑞> = −𝑘 = −𝑘
su > z su >
‚ƒ}
Combining the above relations: 𝑞> = 𝑇> − 𝑇… .
„z
In other terms:
𝒒𝒘 = 𝒉𝒄 𝑻𝒘 − 𝑻…
𝜕𝑢 𝑈
𝜏> = 𝜇 ≈𝜇
𝜕𝑦 >
δ
𝜕𝑇 𝑇> − 𝑇…
𝑞> = −𝑘 ≈𝑘
𝜕𝑦 >
δT
Hence:
𝑘𝜏> 𝑘 𝑘
ℎ• = ≈ ≈ (δ ≈ δT if 𝑃𝑟 ≈ 1)
𝜇𝑈 δ δT
Note that:
• the convective heat transfer coefficient does not depend on the wall temperature
• where the boundary layer thickness is minimum, the heat transfer coefficient is maximum
• Nusselt number
ℎ• 𝑥
𝑁𝑢‘ =
𝑘
• Reynolds number
𝜌𝑈𝑥
𝑅𝑒‘ =
𝜇
‚ƒ}
Rearranging the relation ℎ• = we obtain the Reynold’s analogy (flat plate):
„z
1
𝑁𝑢‘ = 𝐶• 𝑅𝑒‘
2
In case of 𝑃𝑟 ≠ 1 and turbulent flow (Colburn’s analogy):
1
𝑁𝑢‘ = 𝐶• 𝑅𝑒‘ 𝑃𝑟w/•
2
ASI UNCLASSIFIED - pag 19 Agenzia Spaziale Italiana
Convective heat transfer (fundaments)
In case of pipe (i.e., axial pressure gradients) the Colburn’s analogy for turbulent flow is still valid:
𝟏
𝑵𝒖 = 𝑪𝒇 𝑹𝒆 𝑷𝒓𝟏/𝟑
𝟐
where:
• Reynolds number
𝐺𝐷 𝑚̇𝐷
𝑅𝑒 = =
𝜇 𝜇𝐴
(𝐷: pipe diameter; G: mass flux [kg/(s m2)]; 𝑚̇: mass flow rate [kg/s])
• Nusselt number
ℎ• 𝐷
𝑁𝑢 =
𝑘
• Mass flux:
1
𝐺 = £ 𝜌 𝒖 ¤ 𝒏 d𝐴
𝐴
¦
ASI UNCLASSIFIED - pag 20 Agenzia Spaziale Italiana
Convective heat transfer (fundaments)
For a smooth pipe with turbulent flow having 3 ¤ 10¨ < 𝑅𝑒 < 10ª :
0,046
𝐶• =
𝑅𝑒 .,c
This expression, often referred to as “Dittus Boelter equation”, is useful in a fairly wide range
of pressure gradients
Surface roughness can have a large effect on the friction and heat transfer (that is, increasing of
both 𝐶• and 𝑁𝑢)
Heat transfer
characterization:
hot-gas side
𝒒𝒈 = 𝒉𝒈 𝑻𝒘𝒂 − 𝑻𝒘𝒉
where:
• 𝑞/ is the convective heat flux
• ℎ/ is the heat transfer coefficient
• 𝑇>´ − 𝑇>µ is the driving potential
• 𝑇>´ is the adiabatic wall temperature
• 𝑇>µ is the hot gas side wall temperature
Note that:
• one dimensional isentropic flow modeling is assumed for the hot-gas flow
• the hot-gas flow in a thrust chamber can be still assimilated to a boundary layer problem
and thus 𝒉𝒈 is only a weak function of the wall temperature
• in case of high velocity, the driving potential is not based on the free stream
temperature but on the adiabatic wall temperature
ASI UNCLASSIFIED - pag 24 Agenzia Spaziale Italiana
Heat transfer characterization: hot-gas side
The adiabatic wall temperature is the temperature that would be attained at wall in case of
adiabatic condition: 𝑞/ = 0. Its deviation from the stagnation temperature of the free stream is
evaluated by the recovery factor 𝑟:
𝑻𝒘𝒂 − 𝑻𝒈
𝒓=
𝑻𝟎𝒈 − 𝑻𝒈
where
• the adiabatic wall temperature is essentially equal to the free stream stagnation
temperature (i.e., the combustion temperature)
• in case of adiabatic condition the flow at wall almost reaches the combustion temperature
• if the wall has a flow obstruction or a wall protrusion, then the kinetic gas energy is locally
converted back into thermal energy essentially equal to the stagnation temperature and
pressure of the combustion chamber. Since this would lead to local overheating and failure
of the wall, rocket engine inner walls have to be smooth
ASI UNCLASSIFIED - pag 26 Agenzia Spaziale Italiana
Heat transfer characterization: hot-gas side
Many (although sometimes more complicated) correlations for the Nusselt number similar to
the Dittus Boelter equation (𝑁𝑢 = 0,023 𝑅𝑒 ., 𝑃𝑟 .,•• ) are presented in the literature
Almost all the proposed correlations are based on the assumption that 𝑁𝑢~ 𝑅𝑒 ., and can be
applied to determine (at best) within about ±𝟐𝟎% the convective heat transfer
The fluid properties can be evaluated at local bulk temperature or at a given “film
temperature” (e.g., the average value between the wall and the free stream temperature)
.,Â
ℎ/ 𝐷 𝑚̇𝐷 𝜇𝑐E .,¨ 𝜌´Å 𝜇´Å .,c
= 0,026
𝑘. 𝜇. 𝐴 𝑘 . 𝜌′ 𝜇.
where:
• the subscript “0” refers to properties evaluated at the combustion temperature
• the subscript “am” refers to properties evaluated at the average temperature between the
wall and the free stream temperature
• 𝜌′ is the free-stream value of the local gas density
𝑵𝒖~ 𝑹𝒆𝟎,𝟖
where
𝑚̇
𝑅𝑒~𝐺𝐷~
𝐷
(𝑚̇ = 𝐺𝐴 is the hot gas mass flow rate, which is
constant through the thrust chamber in case of
steady state conditions)
𝒎̇𝟎,𝟖
𝒉𝒈 ~ 𝟏,𝟖
𝑫 throat section
For a given mass flow rate, the maximum heat
transfer coefficient occurs at the throat of the
thrust chamber (experimentally, the maximum
ℎ/ is observed slightly upstream because of
variable flow properties)
𝒑𝟎𝒈 𝑨𝒕
𝒎̇ = 𝜞
𝑹𝑻𝟎𝒈
where
• 𝑝./ is the stagnation pressure at the throat
(it is assumed that the flow is isentropic and thus the stagnation pressure is constant all along
the thrust chamber → 𝑝./ is the chamber pressure)
• 𝑇./ is the stagnation temperature at the throat
(it is assumed that the flow is adiabatic and thus the stagnation temperature is constant all
along the thrust chamber → 𝑇./ is the chamber temperature)
ÐÑÒ
ÎÏw |Ó ÐÔÒ
• 𝐴Ì is the throat area, 𝑅 is the hot gas constant, and Γ = γ is a fluid constant
c
•¾
weakly depending on γ = . For γ = 1.2 → Γ ≃ 0,65
•¿
Thus, the hot gas side heat transfer coefficient can be expressed as:
., .,×
𝑝./ 𝐴Ì
ℎ/ ~ .,c
ASI UNCLASSIFIED - pag 29 𝐷Ì 𝐴 Agenzia Spaziale Italiana
Heat transfer characterization: hot-gas side
The expression:
𝒑𝟎,𝟖
𝟎𝒈 𝑨𝒕
𝟎,𝟗
𝒉𝒈 ~ 𝟎,𝟐
𝑫𝒕 𝑨
highlights that:
• the cooling requirement increases rather rapidly with increasing chamber pressure
• the maximum heat transfer coefficient occurs at the throat of the thrust chamber
• scale laws can be used to characterize the hot gas side heat transfer
The heat transfer rate to the thrust chamber walls can be augmented by the hot gas energy
radiation
Gases do not radiate over a continuous spectrum of wavelengths but rather over discrete
“bands” (i.e., far from the “blackbody” emission spectrum). Moreover, gases absorb
radiation as they radiate it. The most important radiating gases from common propellants
are CO2, H2O, CO, etc. (in general, polyatomic gases radiate more strongly than diatomic
gases)
The radiant power from hot gas to a “unit area” wall can be modeled as:
𝒒𝒓 = 𝝐𝒈 𝝈 𝑻𝟒𝒈
where
• 𝜖/ is the gas emissivity (non-dimensional < 1); it is a complex function of the gas
properties and 𝜖/ = 1 in the case of “blackbody” emission
• 𝜎 is the Stefan-Boltzmann constant (𝜎 = 5,6697 ¤ 10|Â W/mc K ¨ )
• 𝑇/ is the gas temperature
The wall is usually assumed to absorb all the incident radiation (like a “blackbody”) while
reradiating negligible energy because 𝑇>/ ≪ 𝑇/
Hence, 𝑞à = 𝜖/ 𝜎 𝑇/¨ approximately represents the net radiant heat flux to the wall
In general, radiative heat flux increases with increasing temperature, pressure and chamber
throat diameter
If the gases contain solid or liquid particles, these may appreciably contribute to the
emitted radiant energy. In this case, the particles radiate over a continuous spectrum of the
wavelengths (i.e., luminous flames in contrast to the non-sooty flames that radiate very little
in the visible region of the spectrum because the emitting bands of CO2, H2O, CO are all in
the infrared region)
Generally, radiative heat transfer in rocket thrust chamber is generally modest (especially
in the absence of solid particles), from 5% to 35% of the total heat transfer
Heat transfer
characterization:
coolant side
𝒒𝑳 = 𝒉𝑳 𝑻𝒘𝒄 − 𝑻𝑳
where:
Note: in case of pipe flow (i.e., considering the coolant flowing in a number of tubes that
constitute the thrust chamber) 𝑇â represents the average coolant temperature (bulk
temperature) instead the free-stream temperature. In this case, the fluid properties (𝜇, 𝑘, and
𝑐E ) are evaluated at such average temperature. Finally, if the pipe cross section is not
circular, the hydraulic diameter is used in the formulas
The bulk temperature of the coolant increases from the point of entry until it leaves the cooling
passages, as a function of the heat absorbed and of the coolant flowrate, as imposed by the
energy balance equation:
Note that in the energy equation the assumption of low velocity has been made and thus the
stagnation and the static conditions are equal
None of the common propellants can absorb more than a few percentage of the heat of
combustion (without vaporizing or decomposing and thus becoming unsuitable as coolants).
The heat of combustion can be of the order of 100 MJ per kg of fuel while ∆𝐻â can be of the
order of 1 MJ/kg. This large difference is due to the fact that the combustion is a much more
energetic phenomenon than convective heat transfer
ASI UNCLASSIFIED - pag 35 Agenzia Spaziale Italiana
Heat transfer characterization: coolant side
Notes on coolant pressure drop
Provided that the coolant remains chemically stable, the cooling system should be designed so
that the fluid absorbs all possible heat transferred from the hot-gas. Of course, the coolant
pressure drop must be properly regulated because higher pressure drop allows a higher
coolant velocity in the cooling channel (and thus a better cooling), but requires a heavier feed
system, which increases the engine mass and thus also the total inert vehicle mass
In order to limit the coolant pressure drop, abrupt change of flow direction and sudden
expansion or contraction of flow areas should be avoided. Moreover, the inner surface of
cooling passages should be smooth and clean
The pressure drop ∆𝑝 in a channel of length L, diameter 𝐷 and without concentrated pressure
loss can be evaluated as follows:
𝐺 c 𝑓𝐿
∆𝑝 = ¤
2𝜌 𝐷
where
• 𝑓: friction factor coefficient (it is mainly a function of 𝑅𝑒 and relative surface roughness)
• 𝐺: coolant mass flux
• 𝜌: coolant density
ASI UNCLASSIFIED - pag 36 Agenzia Spaziale Italiana
Heat transfer characterization: coolant side
Typical rocket coolants operate in different thermodynamic regimes that can have a
remarkable influence on the cooling performances
Some examples and their typical state in the
cooling circuit with respect to the critical
point 𝑝• and 𝑇• :
• Water (liquid)
𝑝• = 221 bar and 𝑇• = 647 K
• Hydrazine (liquid or sup. pres.)
𝑝• = 147 bar and 𝑇• = 652 K
• Nitrogen tetroxide (liquid or sup. pres.)
𝑝• = 101 bar and 𝑇• = 431 K
• Methane (sup. pres.)
𝑝• = 46 bar and 𝑇• = 190 K
• Kerosene (sup. pres.)
𝑝• = 20 bar and 𝑇• = 678 K
Note: in most systems, particularly those
• Hydrogen (sup. pres.)
fed from a turbopump, the coolant
𝑝• = 13 bar and 𝑇• = 33 K
pressure is supercritical
ASI UNCLASSIFIED - pag 37 Agenzia Spaziale Italiana
Heat transfer characterization: coolant side
• Forced convection
(when 𝑝 < 𝑝• and 𝑇>• < 𝑇ëì or when 𝑝 > 𝑝• )
single phase heat transfer
• Nucleate boiling 𝑝
(when 𝑝 < 𝑝• and 𝑇>• − 𝑇ëì ≲ 50K)
small vapor bubbles causes local turbulence
increase and ℎâ greatly increases
• Film boiling
(when 𝑝 < 𝑝• and 𝑇>• − 𝑇ëì ≳ 50K)
From (2)
a gaseous film insulates the wall, often
causing its melting 𝑇>• − 𝑇â
ñò
• Nitrogen tetroxide: 𝑞Å´‘ = 6 ÷ 11
ÅÓ
(𝑝• = 101 bar and 𝑇• = 431 K) 𝑝
ñò
• Kerosene: 𝑞Å´‘ = 2 ÷ 6 Ó
Å
(𝑝• = 20 bar and 𝑇• = 678 K)
𝑇>• − 𝑇â
ASI UNCLASSIFIED - pag 39 Agenzia Spaziale Italiana
Heat transfer characterization: coolant side
Many semiempirical relations have been found to determine the coolant heat transfer in round
tubes and for liquids far from boiling and supercritical fluids sufficiently far the from critical
point. These correlation are typically of the “Dittus Boelter”-type and have a typical uncertainty
of 20÷30%:
𝑵𝒖 = 𝑨 𝑹𝒆𝟎,𝟖 𝑷𝒓𝒂
Modification terms are typically added to take
into account for:
• entrance (increases ℎâ )
• roughness
(can greatly
increases ℎâ )
(NASA) (NASA)
Heat transfer
characterization:
wall conduction
where
Steady-state heat
transfer
𝑞/ + 𝑞à = 𝑞> = 𝑞â
where
‚}
• 𝑞> = 𝑇>µ − 𝑇>• : conductive heat flux through the wall
∆𝐿
𝒒𝒓
𝒒𝒘 = 𝑯 𝑻𝒘𝒂 − 𝑻𝑳 +
𝒉𝒈
𝟏 𝟏 ∆𝑳 𝟏
= + +
𝑯 𝒉𝒈 𝒌𝒘 𝒉𝑳
𝑻𝑳 + 𝜺𝑻𝒘𝒂 𝒒𝒓
𝑻𝒘𝒉 = +
𝟏+𝜺 𝒉𝒈
where:
∆𝑳 𝟏
𝜺 = 𝒉𝒈 +
𝒌𝒘 𝒉𝑳
ASI UNCLASSIFIED - pag 46 Agenzia Spaziale Italiana
Steady-state heat transfer
𝑞/ + 𝑞à 𝐴/ = 𝑞> 𝐴> = 𝑞â 𝐴â
where:
• 𝐴/ : gas side area
• 𝐴> : effective wall area
• 𝐴â : effective coolant side area
The formulas of the conductive heat flux 𝑞> and the hot gas side wall temperature 𝑇>µ does not
change, provided that the coolant heat transfer coefficient and the wall thermal conductivity are
replaced by the “equivalent” values:
𝐴â
ℎây = ℎâ
𝐴/
y
𝐴>
𝑘> = 𝑘>
𝐴/
Note that, since 𝐴â and 𝐴> are larger than 𝐴/ , the equivalent values 𝒉y𝑳 and 𝒌y𝒘 are larger than
the original one 𝒉𝑳 and 𝒌𝒘
Including this geometric effect implies a larger conductive heat flux 𝑞> and a lower hot gas side
wall temperature 𝑇>µ
The nozzle throat region has usually the highest heat-transfer intensity and is therefore the most
difficult to cool. For this reason the cooling channels are often designed so that the coolant
velocity is highest in this region by restricting the coolant passage cross section
𝒑𝟎,𝟖
𝟎𝒈
𝒒𝒘𝒕 ~
𝑫𝟎,𝟐
𝒕
Finally, because the thrust is F = 𝐶ÿ 𝑝ì/ 𝐴Ì , comparing engines with equal thrust coefficient 𝐶ÿ gives
F~𝑝ì/ 𝐷Ìc and thus:
𝒑𝟎,𝟗
𝟎𝒈
𝒒𝒘𝒕 ~
𝑭𝟎,𝟏
The peak heat flux (at the throat region) increases almost linearly with 𝑝./ because higher 𝑝./ is
linked with higher combustion-gas mass flow rate per unit area of chamber cross section and
therefore with higher heat transfer coefficient:
𝐹 𝐹 𝒎̇𝒈
𝐢𝐟 𝐅 = 𝒄𝒐𝒔𝒕 and 𝒑𝒐𝒈 ↑ ⇛⇛⇛ 𝐴Ì = ↓; 𝑚̇/ = = 𝑐𝑜𝑠𝑡; 𝑮𝒕 = ↑ ; ℎ/Ì ↑
𝐶ÿ 𝑝ì/ 𝐶ÿ 𝐶 ∗ 𝑨𝒕
Moreover, increasing the engine thrust, the heat flux slightly reduces. This is due to the fact that
larger engines have larger throat area and thus slightly smaller heat transfer coefficient:
𝐹 𝐹 𝒎̇𝒈 𝒑𝒐𝒈 𝑚̇/., 𝐺Ì.,Â
𝐢𝐟 𝒑𝒐𝒈 = 𝒄𝒐𝒔𝒕 and 𝑭 ↑ ⇛⇛⇛ 𝐴Ì = ↑; 𝑚̇/ = ↑; 𝑮𝒕 = = ∗ = 𝒄𝒐𝒔𝒕; ℎ/Ì ~ w, = .,w ↓
𝐶ÿ 𝑝ì/ 𝐶ÿ 𝐶 ∗ 𝑨𝒕 𝑪 𝐷Ì 𝐴Ì
ASI UNCLASSIFIED - pag 50 Agenzia Spaziale Italiana
Steady-state heat transfer
𝒑𝟎,𝟗
𝟎𝒈
The relation 𝒒𝒘𝒕 ~ implies that:
𝑭𝟎,𝟏
• the maximum allowable heat flux limits the chamber pressure and thus the engine
performances
• cooling is relatively easier in large-thrust engines
Different propellants
have different heat
transfer rates:
p.,×
.4
q2® = C .,w
F
where the parameter
C depends on the
adopted propellants.
𝒒𝒘 𝑻𝒘𝒉
𝒉𝒈 + +
𝒉𝑳 + -
𝒌𝒘 /∆𝑳 + -
In fact:
• an increase of ℎ/ , or ℎâ , or 𝑘> /∆𝐿 leads to an increase of the overall heat transfer coefficient 𝐻
w w ∆𝐿 w
where = + + and thus of 𝑞> = 𝐻 𝑇>´ − 𝑇â
5 µ¸ ‚} µ6
∆𝐿 w {6 Ï8{}9
• an increase of ℎ/ implies an increase of 𝜀 = ℎ/ + and thus of 𝑇>µ = while an
‚} µ6 wÏ8
increase of ℎâ or 𝑘> /∆𝐿 implies a decrease of 𝜀 (and thus of 𝑇>/ )
It is one of the major design goals to keep coefficient 𝒉𝒈 “low” and the coefficient 𝒉𝑳 and
∆𝑳 /𝒌𝒘 “high” in order to reduce the hot gas side wall temperature 𝑻𝒘𝒈
Thermo-mechanical
characterization
• The walls of the thrust chambers are subjected to radial and axial loads from the chamber
and coolant pressure, flight accelerations, vibration, and thermal stresses. They also have to
withstand a momentary ignition pressure surge or shock, often due to excessive propellant
accumulation in the chamber (this surge can exceed the nominal chamber pressure)
• The thermal stresses induced by the temperature difference across the wall are often the
most severe stresses and a change in heat transfer or wall temperature distribution will
affect the stresses in the wall. In particular, the most severe thermal stresses can occur during
the start, when the hot gases cause thermal shock to the wall
Issues to be taken into consideration about the selection of the thrust chamber materials:
• The strength of the chamber wall against the thermal gradients and the high pressure in the
cooling system (with respect to the hot-gas pressure)
• The chemical resistance of the material to the low-velocity coolant on one side and to the
high-velocity hot-gas on the other
• The method of fabrication (metal forming and welding or machining, casting, etc.)
ASI UNCLASSIFIED - pag 54 Agenzia Spaziale Italiana
Thermo-mechanical characterization
Estimation of the wall stress (static loads)
Ignoring the walls that separate the coolant passages (i.e., assuming a one-dimensional cooling
jacket model), the wall material undergoes a combination of a constant compressive stress 𝑆E ,
caused by the pressure differential between the coolant and combustion gases, 𝑝•ì − 𝑝/ , and the
thermal stress 𝑆Ì caused by the temperature gradient across the wall. The thermal stress induce
compression on the hot-gas side and expansion on the coolant side. The maximum (compressive)
stress occurs at the hot-gas side and can be evaluated as:
Note that, according to the cooling jacket method, wall temperature difference in wall material is:
𝑞> ∆𝐿
ASI UNCLASSIFIED - pag 55 ∆𝑇 = Agenzia Spaziale Italiana
𝑘>
Thermo-mechanical characterization
At a given thrust level regenerative cooling is feasible only if the combustion pressure is
below a limiting value 𝒑𝒍𝒊𝒎 𝟎𝒈 and for a given wall thickness ∆𝑳
𝒍𝒊𝒎
. These parameters are
established by structural requirements to accommodate pressure and thermal stresses as well as
by fabrication feasibility limits
This limit arises because:
• the pressure differential stress 𝑆E is
proportional to:
ECD |E·¸ à E·¸
𝑆E = ~
Ɖ Ɖ
(note that 𝑝•ì − 𝑝./ ~𝑝./ )
In other words:
• the thicker the wall the more it supports the pressure load. The request for thick wall is
more pronounced with increasing chamber pressure
• the thinner the wall the more it supports the thermal load. The request for thin wall is
more pronounced with increasing chamber pressure
ASI UNCLASSIFIED - pag 56 Agenzia Spaziale Italiana
Thermo-mechanical characterization
The limiting values of chamber pressure 𝒑𝒍𝒊𝒎𝟎𝒈 and wall thickness ∆𝑳
𝒍𝒊𝒎
are dependent on
cooling system configuration and material properties
(NASA)
With respect to stainless steels, copper alloys generally permit higher chamber pressure with
wall thickness between 0,5 and 1mm (which is still feasible by means of milling technique
because copper is a highly ductile material). For this reason, copper alloys are an excellent
choice but for propellant combinations with corrosive or aggressive oxidizers (nitric acid or
nitrogen tetroxide) stainless steel is often used as the inner wall material, because copper would
chemically react. Maximum allowed temperature is about 850K for high-performance copper
alloys (such as CuAgZr) and 900K for stainless steel. Thermal conductivity of copper alloys is
300-380 W/m K while of stainless steel is about 20 W/m K
ASI UNCLASSIFIED - pag 57 Agenzia Spaziale Italiana
Thermo-mechanical characterization
Note that the copper alloys led to “high” value of 𝑘> /∆𝐿, which is beneficial for the heat
transfer. In fact, considering typical values of the thermal conductivity and wall thickness of
copper alloys structures, 𝑘> /∆𝐿~10H ÷ 10ª W/m2K whereas the hot gas and coolant side heat
transfer coefficients are ℎ/ , ℎâ ~10• ÷ 10H W/m2K. This leads to:
• higher wall heat flux 𝑞>
• lower hot gas side wall temperature 𝑻𝒘𝒉
• lower temperature difference in the wall 𝑇>µ − 𝑇>•
• lower thermal stress 𝑺𝒕
In fact:
‚ ‚
• increasing } under the hypothesis } ≫ ℎ/ , ℎâ
∆𝐿 ∆𝐿
then:
w w ∆𝐿 w w w
• = + + ≈ + ↓
5 µ¸ ‚} µ6 µ¸ µ6
• 𝑞> = 𝐻 𝑇>´ − 𝑇â ↑
∆𝐿 w µ¸
• 𝜀 = ℎ/ + ≈ ↓
‚} µ6 µ6
{6 Ï8{}9
• 𝑇>µ = ↓
wÏ8
∆𝐿 ∆𝐿
• 𝑇>µ − 𝑇>• = 𝑞> = 𝐻 𝑇>´ − 𝑇â ↓
‚} ‚}
EFù} ∆â EF({}L |{}C )
• 𝑆Ì = = ↓
c w|G ‚} c w|G
ASI UNCLASSIFIED - pag 58 Agenzia Spaziale Italiana
Regenerative cooling of LRE thrust chambers
Different thrust
chamber designs
Tubular wall without outer shell (NASA) Tubular wall with outer shell (2)
This type of cooling system is made of singularly formed tapered tubes (to reduce the tube
area in the throat region) that are brazed together. Tube cross section can be circular or not.
This construction technique generally imposes the use of stainless steel or nickel alloys –e.g.
Inconel- (low thermal conductivity materials) with relative small thickness (even 0,2 mm can
be realized because of the relative strength of these materials). Even if external bands of high-
strength steel are generally added to contain the pressure loads, this design is limited to
relatively low heat flux (i.e., 𝒑𝟎𝒈 < 𝟏𝟎𝟎 bar). The primary advantage of this design is its light
weight (because of the use of high strength materials) and the large gained experience in US
(e.g., F-1, RL-10, J-2, RS-27)
ASI UNCLASSIFIED - pag 60 Agenzia Spaziale Italiana
Different thrust chamber designs
Tubular-wall design used on relatively low chamber pressure thrust chambers or nozzle extensions
H-1: 𝑝./ ~50 bar F-1: 𝑝./ ~70 bar RL-10: 𝑝./ ~20 ÷ 40 bar
(stainless steel 347; ∆𝐿Å=M ~0,25mm) (Inconel X-750; ∆𝐿Å=M ~0,46mm) (stainless steel 347; ∆𝐿Å=M ~0,25mm)
J-2: 𝑝./ ~50 bar LE-5A: 𝑝./ ~40 bar SSME nozzle: 𝑝./ ~200 bar
(stainless steel 347; ∆𝐿Å=M ~0,25mm) (nickel alloy) (stainless steel A-286)
This type of cooling system is realized by machining (typically milling) rectangular grooves of
variable width and depth into the surface of a relatively thick contoured high-conductivity
chamber and nozzle wall liner; an outer shell of nickel-alloy (high strength material) is added to
enclose the coolant passages
The advantage over the tubular-wall design is the possibility to use high-conductivity material
for wall construction, such as copper alloys, and relatively small thickness (below 1mm).
Consequently, this design can be used to extremely-high-heat-flux. This design, that is more
recent in US than tubular-wall, is adopted in all relatively high pressure thrust chambers, that
is, having 𝒑𝟎𝒈 > 𝟏𝟎𝟎 bar (e.g., SSME-MCC, RS-68, Vulcain)
ASI UNCLASSIFIED - pag 62 Agenzia Spaziale Italiana
Different thrust chamber designs
Channel-wall design generally used on relatively high chamber pressure thrust chambers
Merlin 1C: 𝑝./ ~100 bar Aestus (note the liner made of steel instead of copper alloy): 𝑝./ ~11 bar
When the Space Shuttle Main Engine (SSME) began development, it was apparent that the
higher combustion chamber temperature and pressure required much stronger construction, as
well as an extremely high heat transfer capability. This dictated a channel-wall construction
along with a copper alloy hot wall for high thermal conductivity
ASI UNCLASSIFIED - pag 64 Agenzia Spaziale Italiana
Different thrust chamber designs
Historical note on US thrust chambers
The advancement of the US thrust chamber design has permitted to improve the maximum
allowable heat flux (at the throat) and thus the chamber pressure
channel wall design
up to 160 MW/m2
(𝒑𝟎𝒈 = 𝟐𝟎𝟎 bar)
disruptive
tubular wall design improvement up to 35 MW/m2
(𝒑𝟎𝒈 = 𝟓𝟎 bar)
Rocket engines time evolution of the maximum throat heat flux (c)
The Main Combustion Chamber of the Space Shuttle Main Engine is regeneratively cooled with
hydrogen and has 390 coolant slots (430 in later versions) in the copper alloy (NARloy-Z, a
copper-silver-zirconium alloy with significantly greater strength than pure copper but with only slightly
lower thermal conductivity) liner; the slots are closed out with a thin layer of electrodeposited
copper (as a hydrogen barrier) and then electrodeposited nickel (for strength)
Sketch of the RD-107 engine: channel-wall design for the thrust RD-107 engine clearly showing
chamber and sandwich-wall design for the nozzle extension (NASA) the copper alloy inner liner
ASI UNCLASSIFIED - pag 67 Agenzia Spaziale Italiana
Different thrust chamber designs
Moon race rocket engines: F-1 (first stage of Saturn V, that brought 12 men on the moon)
Main characteristics:
• propellant: LOX /kerosene (RP-1)
• cycle: gas-generator (open cycle)
• chamber pressure: 70 bar
• thrust (sea level): 677 ton
(still the highest for a flown engine)
Bifurcation joint (from 178
• throttle range: N.A. to 356 Inconel X-750 tubes)
• Isp (seal level): 263 s
• exit to throat ratio: 16
• dry weight : 8,4 ton
• thrust-to-weight ratio: 94
• height: 5,8 m
• diameter: 3,7 m
≈
thrust
With respect to the tubular-wall design of F-1 (not developed anymore), the thermo-mechanical
superiority of the channel/sandwich-wall design (and other peculiarities) of NK-33 permits to
have about the same thrust of one F-1 using 4 x NK-33 and:
• +110% chamber pressure
• +34 s (+13%) Isp
• -43% weight
• -25% volume
Clearly, going to the moon is not just a matter of propulsion!
ASI UNCLASSIFIED - pag 70 Agenzia Spaziale Italiana
Regenerative cooling of LRE thrust chambers
Overview of advanced
concepts:
• Chemical stability of coolants
• High aspect ratio cooling channels
• Hot-gas side wall surface roughness
• Hot-gas side wall oxidation
• Different heat transfer mechanism in the face plate region
• Hot gas side carbon deposition
• Combustion instabilities
• Thrust chamber life
• Influence of the thrust chamber dimension on the cooling requirements
• Methods of increasing the heat transfer parameters
• Additional methods to reduce the wall heat flux
• Heat transfer in nozzle extension
• Heat transfer in gas-generators/preburners
Hydrogen is always supercritical in the channels and is an excellent coolant, has a high specific
heat (low temperature gain in the cooling system), and is chemically stable (it leaves no
residues)
ASI UNCLASSIFIED - pag 72 Agenzia Spaziale Italiana
Overview of advanced concepts
High aspect ratio cooling channels
Channel wall design permits the Cut of the Vulcain
construction of high aspect ratio thrust chamber wall
cooling channels (height to base made of copper alloy
liner and galvanic
ratio up to 10) deposited Nickel outer
shell (5)
This geometry leads to a larger
number of channels and longer fins
and both effects will increase the
cooling heat transfer area.
(NASA)
The limits for the liner walls
thickness are mainly given by the
requirements of conventional
milling tools. To date thicknesses
of less than 0.5 mm are hardly Effect on cooling channel aspect ratio on wall temperature (ASTRIUM)
achievable
ASI UNCLASSIFIED - pag 73 Agenzia Spaziale Italiana
Overview of advanced concepts
Hot-gas side wall surface roughness
Surface roughness can have a large effect on the heat
transfer coefficient and thus on the wall heat flux
(which can be increased by a factor up to 2) and wall
temperature. In addition, it is generally measured that
hot-gas side surface roughness increases with chamber Typical SSME-MCC heating evolution with test
run-time. Major surface roughness on the hot-gas side duration –up to 10% increase in 300 sec.- (NASA)
may have dramatic consequences. In fact, if the nozzle
inner wall has a flow obstruction or a wall protrusion,
then the kinetic gas energy is locally converted back
into thermal energy essentially equal to the stagnation
temperature and pressure in the combustion chamber.
Since this would lead to local overheating and failure
of the wall, nozzle inner walls have to be smooth.
Hot-gas side wall oxidation Typical SSME-MCC hot spots (NASA)
Since the rates of chemical oxidizing reactions between the hot gas and the wall material can
increase dramatically with wall temperature, cooling also helps to reduce the oxidation of the
wall material and the rate at which walls would be eaten away. The oxidation problem can be
minimized not only by limiting the wall temperature, but also by burning the liquid propellants
at a mixture ratio where the percentage of aggressive gases in the hot gas (such as oxygen) is
very small, and by coating certain wall materials with an oxidation-resistant coating
ASI UNCLASSIFIED - pag 74 Agenzia Spaziale Italiana
Overview of advanced concepts
Different heat transfer mechanism in the face plate region
Comparisons of analytical results with experimental heat transfer data obtained on rocket thrust
chambers have often shown disagreement. Major deviations generally affect the face plate region
because each injector configuration produces different combustion characteristics. This results in
deviations from calculations based on the assumption of homogeneous product gases (as in the
Bartz’s equations)
In addition, assumption of purely radial heat flow has been made. In reality, circumferential
variations are present, especially in the injector region. Peaks of heat flux result in longitudinal
discolorations (streaks) of the inner surface of the chamber after a firing
Evidence of multi-
dimensional
environment (and
also wall
oxidation) in a
combustion
chamber (NASA)
(NASA)
The carbon deposit thermal resistance when 𝑝./ ~70 bar can (NASA)
be as high as 7000 cm2K/kW while the material thermal
resistance ∆𝐿 /𝑘> can be as low as (in case of copper alloy
structure) 10 cm2K/kW. The equivalent material thermal
resistance is practically the one of the carbon deposit
(∆𝐿 /𝑘> |;ù =10+7000 ≈ 7000 cm2K/kW) and thus the hot-gas
side wall temperature 𝑇>/ greatly increases (although the
carbon deposit withstands higher temperatures than metal
alloys and thus may protect the metal liner)
Combustion instabilities
Vulcain thrust
chamber liner
with typical
longitudinal
failures (5)
(NASA)
“NARloy Z” (CuAgZr) elasticity modulus ( NASA) “NARloy Z” (CuAgZr) yield and ultimate strengths ( NASA)
• creep: plastic deformation under the influence of mechanical stresses (also below the yield
strength of the material); creep is more severe in materials that are subjected to heat for long
periods
• fatigue: the weakening and damaging caused by repeatedly applied loads. The nominal
maximum stress values that cause such damage may be much less than the strength of the
material (i.e., the ultimate tensile strength limit, or the yield strength limit)
No existing materials combine good creep and fatigue properties at elevated temperatures
and thus the strength of materials decreases with increasing temperatures
ASI UNCLASSIFIED - pag 79 Agenzia Spaziale Italiana
Overview of advanced concepts
Aside the discussed static pressure and thermal stresses, the thrust chamber material is exposed
to additional loads which have an impact on the engine life:
For a small rocket engine, the integrated heat flux over the entire inner surface is below 2% of
the heat of combustion. This percentage is smaller for larger rocket engines. In fact, for constant
chamber pressure 𝑝./ , increasing the thrust chamber dimension (i.e., increasing the throat
diameter 𝐷Ì and thus the thrust F = 𝐶ÿ 𝑝./ 𝐴Ì ) implies that:
• the heat transfer rate absorbed by the coolant 𝑄â (i.e., the integrated wall heat flux) increases
with 𝐷Ì.,Â
.,Â
𝑝./
𝑄â ~𝐿𝐷Ì 𝑞>Ì ~𝐿𝐷Ì .,c ~𝐷Ì.,Â
𝐷Ì
(note that the chamber length L is considered constant because the time required for complete
combustion does not appreciably change with chamber dimension)
𝑄â = 𝑚̇â ¤ ∆𝐻â ~ 𝑚
Ṙ / ∆𝑇â and 𝑄â ~𝐷Ì.,Â
Å̇6 ~Å̇¸
1
∆𝑇â ~
𝐷Ìw,c
That is, the larger the engine the smaller the temperature gain. This is important when the
maximum permissive coolant temperature has to be limited (for kerosene carbon deposition at
wall or for safety reasons with hydrazine). Moreover, this also suggests that expander cycle
engines (where the larger the coolant temperature gain the larger the chamber pressure) are
more efficient if the thrust is limited (small engines)
• the heat of combustion of one kg of hot-gas products at chamber pressure 𝑝./ = 197.5 bar
originating from liquid oxygen and hydrogen at the mixture ratio O/F = 6 is 𝐻Q = 10,6 MJ/kg
(obtained under chemical equilibrium condition). The rate of the heat of combustion is:
kg MJ
𝑄Q = 𝑚̇/ ¤ 𝐻Q = 501.6 ¤ 10,6 = 5317 MW ≃ 5,3 GW
s kg
(Italian average electric power consumption is ~ 40 𝐺𝑊)
• the wall heat transfer rate entering the coolant (hydrogen) is:
𝑘𝑔 𝑘𝐽 𝑘𝐽
𝑄â = 𝑚̇â ¤ ∆𝐻â = 13,15 ¤ W4079,9 − 673,5 [ = 44,8 𝑀𝑊 −−→ 𝑄â = 0,008𝑄Q
𝑠 𝑘𝑔 𝑘𝑔
;‘=Ì ;M̵´@Eu =M@;Ì ;M̵´@Eu
(EY•.ª ë´à (EY•Â× ë´à
ASI UNCLASSIFIED - pag 83 {YcªH Z) {YHw Z) Agenzia Spaziale Italiana
Overview of advanced concepts
Methods of increasing the heat transfer parameters
Aestus
𝑇>,Å´‘ ~
ASI UNCLASSIFIED - pag 86 1400 K Agenzia Spaziale Italiana
Overview of advanced concepts
Heat transfer in gas-generators/preburners
In a gas-generator or staged combustion rocket engine type the hot combustion gases that drive
the turbine (with rotating speed up to 50000 rpm) are burned in a separate combustion chamber
(referred to as gas generator or preburner, respectively). The higher the gas temperature at
turbine inlet the lower the required turbine flow (of great relevancy for gas generator engine)
State of the art turbine blade materials
(such as single crystals which have been
unidirectionally solidified) and special
alloys can allow turbine inlet temperatures
up to 1400-1600 K; however, reliability and
cost considerations have kept actual
turbine inlet temperatures at conservative
values, such as 900 to 950 K, using lower
cost steel alloy as the material. Such
temperatures, which are obtained with
mixture ratios far from stoichiometric
(usually fuel rich), are sufficiently low, so
that the combustion chamber and the hot
turbine hardware (blades, nozzles,
manifolds, or disks) still have sufficient Measurements of static stress to cause rupture after 100 hours
(typically the “stress to density ratio” is the relevant variable) (1)
strength without needing forced cooling
ASI UNCLASSIFIED - pag 87 Agenzia Spaziale Italiana
Regenerative cooling of LRE thrust chambers
Experimental
characterization of hot-
gas side heat transfer
(DLR)
This apparatus permits to have reliable estimations of the convective heat transfer coefficient
because it is a weak function of the wall temperature. That is, same hot-gas flows have same
convective heat transfer coefficients even if the wall temperature is different due to the use of
different cooling systems and fluids (e.g., water in the circular sections of the calorimetric
chamber and one of the propellants in the regenerative cooling system of the actual thrust
chamber)
ASI UNCLASSIFIED - pag 89 Agenzia Spaziale Italiana
Experimental characterization of hot-gas side
heat transfer
The procedure to characterize the hot-gas side heat transfer coefficient is:
• the average heat flux in the section 𝑞/ is measured from the increase in water
temperature from inlet to outlet
• the adiabatic wall temperature 𝑇>´ is estimated (e.g., using a one-dimensional model for
the hot-gas expansion and a recovery factor)
• the hot-gas side wall temperature 𝑇>µ can be measured using thermocouples in the
material (eventually more than one in the radial direction, also to evaluate the local 𝑞/ if
the material thermal conductivity is known)
Then, the average heat transfer coefficient in the section is computed as:
𝑞/
ℎ/ =
𝑇>´ − 𝑇>µ
• Scaling is defined as the ability to design new combustion devices with predictable
performance on the basis of test experience with specifically scaled hardware (mainly in
terms of dimensions, flow rate, and pressure). This approach is the only way to evaluate
novel and innovative designs by hot-fire test with considerably lower development cost
• Particularly relevant in this context is the dimensional analysis, where variables that may
influence the system are combined into non-dimensional groups. Scaling is obtained by
keeping these groups constant between the subscale and the full-scale devices
• Scaling of hot-gas side convective heat transfer (mainly the hot gas side heat transfer
coefficient) can be achieved by means of subscale calorimetric thrust chambers
• A beneficial effect of using subscale thrust chambers is that measurement accuracy of heat
flux deteriorates with increasing thrust chamber dimension (i.e., larger thrust, pressure,
vibrations, heat flux, temperature gradients, etc.)
A subscale water-cooled
calorimetric chamber was
used to measure the heat
flux at different sections.
The nominal thrust level
was 40 klbf (17.8 ton), that is
approximately 1/10 scale,
with respect to thrust, of the
full-scale SSME. The throat
diameter was 84 mm while
full-scale SSME diameter is
262 mm. Injection elements
are 600 and 61, respectively,
but the geometry of the
injection element was
identical
(NASA)
ASI UNCLASSIFIED - pag 92 Agenzia Spaziale Italiana
Experimental characterization of hot-gas side
heat transfer
The water-cooled calorimeter chamber had 58 independent coolant circuits, each with separate
temperature and pressure measurements
The subscale chamber had the same combustion chamber length (injector face to throat) of 356
mm, throat convergence ramp angle, throat contour radius of curvature, combustor contraction
ratio (combustion chamber to throat area) of 2,96, and nozzle expansion ratio (exit to throat
area) of 5 as the full-scale MCC
(NASA)
ASI UNCLASSIFIED - pag 93 Agenzia Spaziale Italiana
Experimental characterization of hot-gas side
heat transfer
The average heat flux in each section 𝑞/ is measured from the increase in water temperature
from inlet to outlet
Heat flux profile 𝑞/ for a test with p.4 = 108.7 bar and O/F = 6 (total heat load was 9079 kW) (3)
The (average) hot gas heat transfer coefficient in each of the 58 circumferential cooling circuit
was calculated as:
𝑞/
ℎ/ =
𝑇>´ − 𝑇>µ
where 𝑞/ and 𝑇>µ are experimentally measured while the 𝑇>´ is calculated using a recovery
factor and one-dimensional chemical equilibrium at the appropriate test chamber pressure and
mixture ratio
These data were used to predict the heat transfer coefficient in the same chamber at the
maximum SSME chamber pressure condition. For each axial location, ℎ/ was scaled from p./w =
108,7 bar (subscale condition) to p./c = 204,7 bar (full-scale conditions) by:
.,Â
p./c
ℎ/c = ℎ/w
p./w
• From the injector face to ~127 mm downstream, ℎ/ was the same as the scaled value from
𝑝./ = 108,7 bar (subscale condition) to 𝑝./ = 204,7 bar (full-scale condition) because the
injector elements were the same and heat transfer rates near the injector were then primarily
influenced by the distance from the injector
• Further downstream, where heat transfer rates are primarily convective (velocity driven), the
full-scale ℎ/ corresponded to axial locations where the hot gas Mach number 𝑀 was the same
as in the calorimetric chamber
., .,×
𝑝./ 𝐴Ì 𝐴Ì
ℎ/ ~ .,c where = 𝑓(𝑀)
𝐷Ì 𝐴 𝐴
In particular, the throat diameter effect is not taken into account (that is, the scaling due to
absolute dimensions of the chambers is ignored and thus ℎ/ could be over-predicted). In
particular, for the throat diameter:
• 𝐷Ì = 84,1 mm for the subscale chamber
• 𝐷Ì = 261,75 mm for the full-scale SSME chamber
ASI UNCLASSIFIED - pag 97 Agenzia Spaziale Italiana
Experimental characterization of hot-gas side
heat transfer
3) Scaling Techniques for Design, Development, and Test, C.E. Dexter et al., In
Progress in Astronautics and Aeronautics Series, Vol. 200, published by AIAA
4) Many NASA technical reports from the 60s, 70s, and 80s and devoted to
theoretical and experimental studies on rocket engine thrust chambers
Cover illustration by NASA: Space Shuttle Main Engine – Main Combustion Chamber
ASI UNCLASSIFIED - pag 99 Agenzia Spaziale Italiana
Bibliography
c) Space Propulsion Analysis and Design, R.W. Humble, G.N. Henry, and W.J.
Larson, published by McGraw-Hill
Notes:
• the figures that are not expressively cited are mainly taken from the web
• Russian rockets are barely represented in this presentation only because of the
poor data and available literature
ASI UNCLASSIFIED - pag 100 Agenzia Spaziale Italiana