Propulsion and Power Research 2018;7(1):1–18

HOSTED BY http://ppr.buaa.edu.cn/

Propulsion and Power Research

www.sciencedirect.com

ORIGINAL ARTICLE

Review on film cooling of liquid rocket engines

S.R. Shinen, S. Shri Nidhi

Department of Aerospace Engineering, Indian Institute of Space Science and Technology, Thiruvananthapuram 695547, India

Received 13 October 2015; accepted 21 March 2016

Available online 13 March 2018

KEYWORDS Abstract Film cooling in combination with regenerative cooling is presently considered as
an efficient method to guarantee safe operation of liquid rocket engines having higher heat flux
Heat transfer;
Liquid rocket thrust densities for long duration. This paper aims to bring all the research carried out in the field of
chamber; liquid rocket engine film cooling since 1950. The analytical and numerical procedure followed,
Film cooling; experimental facilities and measurements made and major inferences drawn are reviewed in
Cooling effectiveness detail, and compared where ever possible. Review has been made through a discussion of the
analyses methodologies and the factors that influence film cooling performance. An effort has
also been made to determine the status of the research, pointing out critical gaps, which are still
to be explained and addressed by future generations.
& 2018 National Laboratory for Aeronautics and Astronautics. Production and hosting by Elsevier B.V.
This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).

1. Introduction extract the thrust. Extremely high heat flux levels and
temperature gradients are present not only in the immediate
Liquid rocket engines developed for space missions vicinity of the injector head, but also in the nozzle throat
encompass a wide spectrum of performance and structural region. It is seen that the maximum heat flux occurs in the
requirements. Thrust levels may vary from a few Newtons close proximity to nozzle throat, and an effective cooling of
to many thousands of Newtons, with burning time from the throat area is crucial for enhanced reliability and
fraction of a second to hours. In all these engines, the reusability. Regenerative cooling is the standard cooling
energy released by the propellants must be contained inside system for almost all modern main stage, booster, and upper
the thrust chamber and accelerated through the nozzle to stage engines [1]. Different cooling techniques such as film
cooling, transpiration cooling, ablative cooling, radiation
n
Corresponding author. Tel.: (91) 4712568427. Fax: +91-471 2568406. cooling, heat sink cooling and dump cooling have been
E-mail address: [email protected] (S.R. Shine). developed in the past to reduce regenerative cooling load
Peer review under responsibility of National Laboratory for Aeronautics and propellant requirements. Film cooling can be employed
and Astronautics, China. either at the combustion chamber or at the nozzle of a

https://doi.org/10.1016/j.jppr.2018.01.004
2212-540X & 2018 National Laboratory for Aeronautics and Astronautics. Production and hosting by Elsevier B.V. This is an open access article under the
CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
2 S.R. Shine, S. Shri Nidhi

Nomenclature X non dimensional distance from the coolant injection

point, x/D
c specific heat (unit: J/(kg·K)) α thermal diffusivity (unit: kg/(m·s))
CM turbulent mixing coefficient β tangential angle
D film cooling test section inside diameter (unit: m) δ factor distinguishing the initial mixing and developed
Da Damkohler number region of the jet
h convective heat transfer coefficient (unit: W/(m2·K)) χ entrainment factor for plane, unaccelerated flow with
H* heat release potential, (Tad-Tg,t)/Tg,t continuous slot injection
I engine thrust (unit: N) θ shape factor for the mixing layer profile
Isp specific impulse (unit: 1/s) γ azimuthal angle
L width of adiabatic wall (unit: m) ρ density (unit: kg/m3)
M blowing ratio, ρcvc/ρgvg ε adiabatic film cooling effectiveness, ðT ad −T g Þ=
m_ mass flow rate (unit: kg/s) ðT c −T g Þ
P pressure (unit: bar) η film cooling effectiveness, ðT−T g Þ=ðT c −T g Þ
Pr Prandtl number ω injection parameter
q heat flux (unit: W/m2)
Qs scaled heat flux ratio, (qhot-qcold)/(qmax-qcold) Subscripts
R chamber radius (unit: m)
S coolant slot height (unit: m) ad adiabatic
T local temperature (unit: K) c coolant
Tu turbulence intensity (unit: %) cc combustion chamber
Ν average velocity (unit: m/s) cold non reactive heat flux
v velocity (unit: m·s-1) g main stream flow
V vaporization rate of the liquid surface (unit: lbm·in-1·s-1) hot reactive heat flux
VR velocity ratio, vc/vg max maximum heat flux
We Weber number e entrainment
x axial distance measured from the coolant injection t total
point (unit: m)

rocket engine. Liquid film cooling with fuel or oxidizer as combustion chamber is shown in Figure 2. It can be
the coolant can be employed in the combustion chambers of observed that the coolant film produces a thermal insulation
gas generator/expander/staged combustion cycle engines. In effect and reduces the chamber wall temperature. Coolant
case of gas generator cycle, the turbine exhaust gas can be film may be generated by injecting liquid fuel or oxidizer
used as a gaseous film coolant in the combustion chamber through wall slots or holes in the combustion chamber, or
or nozzle sections. It is found that all these methods lead to through the propellant injector. The cooling effect will
reduced wall temperatures. persist up to the throat region in the case of a shorter
The mechanism by which film cooling maintains a lower combustion chamber. In a fully film-cooled design, injec-
combustor wall temperature is considerably different from tion points are located at incremental distances along the
that of convective cooling. Film cooling is accomplished by wall length. In liquid film cooling, the vaporized film
interposing a layer of coolant fluid between the surface to coolant does not diffuse rapidly into the main gas stream
be protected and the hot gas stream. The fluid is introduced but persists as a protective layer of vapor adjacent to the
directly into the combustion chamber through slots or holes wall for an appreciable distance downstream from the
and is directed along the walls (Figure 1). A typical terminus of the liquid film. The film coolant also forms a
temperature distribution from the hot combustion gases protective film which restricts the transport of the combus-
to the exterior of the chamber wall in a film cooled tion products to the wall, thus reducing the rate of oxidation
of the walls.
Engines like SSME, F-1, J-2, RS-27, Vulcain 2, RD-171
and RD-180 use film cooling technique for combustion
chamber cooling. Several open-cycle rocket engines have
turbine exhaust gas (TEG) delivered to the nozzle for film
cooling, including the F1 engine [2] and J2 engine [3] of the
United States, the upgraded LE5 engine [4] of Japan and
Vulcain 2 of the EADS Astrium. In Vulcain-2 engine, the
combustion chamber wall film cooling has been employed
from the injector face plate down to the nozzle throat
Figure 1 Schematic of the physical system. Section [5]. Engines like Vulcain 2 uses this cooling
Review on film cooling of liquid rocket engines 3

cooling investigations over the years have concentrated

mainly on gaseous film cooling applied to gas turbine
airfoils followed by work on film cooling of rocket
combustion chambers. Survey of current literature on airfoil
film cooling shows extensive studies for coolant holes with
various geometries and at various stream-wise injections.
Film cooling is applied to nearly all of the external surfaces
associated with the airfoils that are exposed to the hot
combustion gases such as the leading edges, main bodies,
blade tips, and end walls. In gas turbines, high pressure
bleed air from the compressor is exhausted from the internal
convective passages of the turbine blades, through holes
drilled into the blade surfaces. Though the processes are
similar, the key challenges involved in rocket engine film
cooling are different from that of airfoil film cooling. The
main differences between the two processes includes:
Figure 2 Typical temperature distribution of combustion chamber
(i) extreme heat flux, temperature and pressure conditions
across wall. present in modern rocket engines, (ii) the presence of highly
accelerated flows, (iii) the initial turbulent state of the
method at both locations. The overall dump cooled nozzle coolant and core streams, (iv) the use of gaseous and liquid
extension used on Vulcain-1 has been replaced by a coolants, (v) two phase flow conditions present with liquid
completely new nozzle design that provides turbine gas film cooling, (vi) surface curvature effects, (vii) the
film cooling injection at the nozzle extension [5]. This has requirement of flow uniformity and wall adherence, (viii)
produced a cooling effect on the nozzle structure. Similarly density gradients and compressibility effects (Coolant to
turbine exhaust gas (TEG) is injected in J-2X engine as a core stream density ratios are in the order of two or higher
coolant and pure hydrogen gas is used to cool LE-7A [6]. are typical in many engines), (ix) the effect of reactive
Film cooling of rocket engines will play a major role in coolant, (x) radiation effects, and (xi) unsteady flows
the development of reusable and booster rocket engines as (Unsteadiness in the core flow of a rocket engine can arise
this technique can increase the chamber life by reducing the from various sources including turbulent flow in the feed
thermal stresses. Efficient use of the coolant films is critical lines, fluttering of pump wheel blades, vibrations of control
because of the associated specific impulse reduction [7]. As valves, and unsteady motions in the combustion chamber
film cooling is such an important cooling technology for the and gas generator). Moreover, the film cooled length is an
durability of liquid rocket engines, this paper presents a important parameter in the design of rocket combustion
summary of past research in the area of film cooling. This chambers. It is also to be noted that in rocket engines, film
review looks at the features of gaseous and liquid film cooling is always applied in combination with other cooling
cooling processes, the methodologies and results in theore- methods (usually regenerative cooling).
tical, computational and experimental approaches and the Proprietary restrictions in space fairing countries and
major factors that influence film cooling performance. companies have resulted in very limited availability of
Every effort has been made to cover all the advancements detailed data about film cooling of liquid rocket engines in
published in this field. the open literature. Most of the studies available in the open
literature are at least 30 years old [11,12]. However, there
has been renewed interest in the subject due to the
2. Studies on film cooling development of new engines. The following discussions
are organized to provide an outline of two major divisions
Origin of film cooling studies extend way back into associated with rocket thrust chamber film cooling: gaseous
history of fluid mechanics starting from the days of and liquid film cooling. Research conducted in these areas
Reynolds [8] and Lamb [9] of the late 19th century. highlighting the main contributions is provided.
Reynolds [8] studied the behavior of vortex rings, a topic
closely related to the modeling of film cooling jets. The
application of a fluid film for the protection of surfaces in 3. Gaseous film cooling
aerospace related field is often attributed to Wieghardt [10].
He applied this technique to de-icing of aircraft wings by Gaseous film cooling is considered to have potential use
blowing warm air over them. Research on film cooling for in nuclear rockets and high-energy liquid chemical rockets.
turbine components and rocket combustion chambers In general, there are limited number of film cooling
started in 1950s. Since then, there have been numerous investigations dealing with gaseous-film cooling, applicable
experimental studies and several models have been devel- directly to a rocket engine. Among the earliest studies,
oped for the prediction of film cooling effectiveness. Film Lucas and Golladay [13] investigated the applicability of
4 S.R. Shine, S. Shri Nidhi

gaseous-film cooling to a rocket combustion chamber. The The above results established the feasibility of gaseous
effects of cooling a cylindrical portion of the combustion film cooling for thermal load reduction in conventional and
chamber with both tangential and inward-angled slot cool- advanced rocket nozzles. However, the effect of film
ant injection and cooling the nozzle with tangential injec- cooling on nozzle behavior in terms of flow separation
tion were studied. In each configuration, nitrogen was used and associated side loads can create limitations for this
as the coolant and in addition, limited data was also technique as evidenced by the reported breakdown of the
obtained with the cooled nozzle using propane as the nozzle wall during the early sea-level testing of the LE-7A
coolant. It was concluded that correlations for adiabatic engine [20]. Influence of film cooling on separation
wall film cooling effectiveness can be employed to predict characteristics during start up and shut down transients
the performance associated with non-reactive, non-conden- can be more substantial depending upon the phasing
sable gases. The requirement of the reactive coolant was between main flow and the secondary coolant flow
higher than the non-reactive coolant with equivalent trans- dynamics. Transient flow in a subscale LE-7A nozzle with
port properties. Multi-slot gaseous film cooling experiments film cooling was numerically investigated by Takahashi
for standard nozzle using hydrogen, methane and carbon et al. [21]. The nozzle had a joint section with backward
monoxide as coolants were carried out by Totten [14]. It facing step for film cooling between its upper section and
was concluded that multi-slot cooling is a most promising extension sections. Rapid movement of the separation point
means of cooling with high energy propellants. No throat was noticed when it passed through the joint section.
erosion was noticed with methane and hydrogen whereas Occurrence of large scale asymmetry of the flow and large
carbon monoxide induced a slight amount. side road was observed when the separation point passes
Back and Cuffel [15] had conducted an experimental through the joint section. The influence of film cooling on
investigation to delineate the structure of the flow field and nozzle flow separation has been experimentally and numeri-
temperature distributions in a shock wave/turbulent bound- cally studied by Reijasse and Boccaletto [22]. The study
ary-layer interaction with film cooling. The study revealed revealed a dynamic phenomenon of the separation shock
the influence of wall cooling on the flow field, wave front region during rapid film injection. Wang and Guidos
structure, and size of the flow separation region. The size [6] have conducted transient numerical side load analysis of
of the separation region became smaller compared to the a film cooled nozzle extension. Simulations were carried
case without cooling, and the separation and reflected shock out for engine start up and shut down transients. Simula-
waves merged together near the edge of the velocity tions showed generation of peak side load due to Mach disk
boundary layer, extending in to the free stream as one flow and the subsequent jump of the separation line.
wave. The ratio of heat transfer coefficient and wall shear Gaseous film cooling is found to be one of the most
stress with and without cooling was found to be 2.3 and promising candidates to achieve lower heat load to the
1.6–1.7. Stoll and Straub [16] used a parabolic finite nozzle wall. For gas-generator cycle engines, the turbine
difference boundary layer code with the k-ε model to exhaust gas (TEG) is readily available as film coolant. The
investigate film cooling and heat transfer in nozzles with characteristics during a steady operation are predictable, but
rounded, smooth throats. No influence of the throat on the film cooling at transient conditions is not well understood.
cooling film stability was noticed. It was also pointed out The side loads generated during the transient operation is to
that a coolant to mainstream mass flow ratio of 2.1% be studied in detail, as side loads are one of the decisive
yielded a significant decrease in heat transfer, both in factors for the use of film cooling during the transient
subsonic and supersonic regions. Arrington et al. [17] operation of rocket nozzles.
compared film cooling of a standard conical nozzle with a
bell nozzle. Gaseous hydrogen and gaseous oxygen were
used as propellants. Film cooling was successfully 3.1. Effect of various parameters
employed in both nozzle configurations. The effect of
gaseous film injection in dual bell nozzle was numerically It is seen that many factors affect film cooling perfor-
studied by Martelli et al. [18]. The injection was made in mance. The various parameters studied so far are listed in
the first bell through an axis-symmetric slot located in the Table 1. Each of these factors is not necessarily independent
divergent section, and it was found that the expansion fan of the others, and so a combination of any of these can
originating from inflection helps the film to protect the wall potentially change the film cooling performance. Therefore,
better. It was also observed that expansion generated by the a large number of operating conditions may be possible and
inflection point lowered the wall recovery temperature and hence there is inherent difficulty in predicting film cooling
reduced the mixing. This allowed the film to protect the performance.
wall for a longer distance. Matesanz et al. [19] have carried Marek and Tacina [23] employed tangential slots to inject
out numerical simulation of slot film cooling in convergent- coolant air inside a rectangular test section to study the
divergent nozzles using direct simulation of turbulence and effect of free-stream turbulence levels on film cooling
large eddy simulation (LES) algorithms. Predicted adiabatic effectiveness. The film cooling effectiveness decreased as
wall temperatures and film cooling effectiveness were much as 50% when the free-stream turbulence intensity was
within 20% of the corresponding experimental data. increased from 7% to 35%. Gau et al. [24] conducted
Review on film cooling of liquid rocket engines 5

Table 1 Factors affecting film-cooling performance.

Coolant/mainstream flow conditions Geometric parameters Other factors

Turbulence in main stream, Tu [23] Injector geometry [33,39,41] External shock wave [34,35]
Coolant Mach No [31] Injector orientation [43] Swirl in the mainstream [24]
Mainstream Mach No [29,30,44] Heat capacity, cc [35]
Blowing ratio, M [37,42]

experiments in a film cooled circular pipe with an abrupt efficient in supersonic flows than subsonic flows. This is
expansion of 2.4:1. These experiments demonstrated that mainly attributed to the following: supersonic flow in
the swirl in the mainstream had a significant effect on the cooling channels will provide thin boundary layers, smaller
film-cooling performance. turbulent length scales and reduced mixing. The spreading
Many studies have discussed film cooling under com- rate of the mixing layer can be controlled via velocity,
pressible conditions. Volchkov et al. [25] theoretically density ratios and convective Mach number. This may
analysed the effect of velocity and temperature compressi- further reduce mixing. Dellimore discussed the disagree-
bility on film cooling effectiveness. He concluded that ments in the literature over the compressibility effects in
compressibility effects are of the first order and can be film cooling problems. He showed that compressibility
neglected in practical combustion chambers. Repukhov [26] influences film cooling performance when the flow Mach
accounted the effects of compressibility on the turbulent numbers are high and the total temperature ratio is low. As
boundary layer equations and obtained similar results. the flow Mach number increases, the growth rate of the
However, Hansmann et al. [27] and Pedersen et al. [28] shear layer decreases and moves shear layer impingement
suggested that compressibility can significantly influence point farther downstream. This leads to improvement in
film cooling effectiveness. Their results showed higher film cooling performance. He also noticed that Volchkov
effectiveness at higher density and velocity ratios (coolant et al. [25] and Repukhov [26] used data corresponding to
to mainstream). Pedersen et al. [28] further showed that this low coolant Mach numbers and hence predicted negligible
effect is more predominant at higher blowing ratios. The effect of compressibility on film cooling performance.
role of mainstream flow velocity in a film cooling duct was The effect of the external shock wave on the film cooling
studied by O’Connor and Haji Sheikh [29] and Kuo et al. was investigated in the Mach 2.35 wind tunnel by Kanda
[30]. Kuo et al.'s experimental observations reported film et al. [34]. At higher pressure ratios, decrease in film
cooling to be effective in subsonic flow, but not in super- cooling effectiveness was noted which was attributed to the
sonic flow. They observed bending of the injected flow decrease in the local Mach number. An increase in heat
stream towards the upstream direction against the incoming transfer coefficients and adiabatic wall temperatures were
mainstream in the case of supersonic flow. This created a noted in the interacting region. Numerical investigations by
stagnation region and temperature rise in the flow field Peng and Jiang [35] confirmed the above findings. Results
leading to negligible film cooling effect on downstream by Peng and Jiang [35] showed reduction in adiabatic
wall regions where the supersonic flow was dominant. effectiveness resulting from impinging oblique shock wave.
However, there exists a vast amount of experimental data In addition, the oblique shock wave enhanced the mixing of
on supersonic film cooling which cover a wide range of the cooling stream with the free stream, especially for
Mach numbers and show that film cooling is more efficient lighter-gas coolants. Helium was more easily affected by
in supersonic flows than subsonic flow. Experimental the shock wave as it diffused faster than methane or
investigation of Juhany et al. [31] showed improvement nitrogen.
in film cooling effectiveness with the increase of coolant An investigation of film cooling in the laminar supersonic
Mach number. Comparison between Helium and air experi- regime was conducted by Yang et al. [36] and reported
ments indicated that the effectiveness increases with heat increased cooling with an increasing mass flow of the
capacity of the gas. Keener et al. [32] investigations in a coolant, with the slot height being unimportant. Investiga-
two-dimensional (2-D) Mach 2 nozzle showed that the tions on film cooling on a flat plate with a slot injection in
boundary-layer thickness has a major effect on the wall heat laminar supersonic flows were done by Heufer and Olivier
fluxes. High heat fluxes were also detected during the flow [37]. Experimental and numerical results showed no influ-
establishment, starting with the nozzle shock from the ence of the blowing parameters on the cooling effects. The
shock tunnel, due to the absence of the boundary layer. cooling effect was found to be influenced by the core gas
Supersonic film injection experiments for two types of flow conditions and was reported that the film cooling
injection geometries and pressure ratios were conducted by technique is highly effective under laminar flow conditions.
Aupoix et al. [33]. The coolant film was generated by Experimental data by Hombsch and Olivier. [38] also
vaporizing liquid nitrogen into air in a mixing chamber. The confirmed that film cooling is much more efficient in
experiments showed that film cooling is much more laminar flow than in turbulent flow. It was also shown that
6 S.R. Shine, S. Shri Nidhi

tangential injection supersedes angled slot injection. Experi- temperature predictions deviated significantly for the
ments by Arnold et al. [39] showed significant variations of unsteady case. Cruz and Marshall, Dellimore, Betti and
wall temperatures due to injector design and more pro- Martelli [44,46,47,51] etc. have used Spalart-Allmaras (SA)
nounced circumferential variations in wall temperature at turbulence model along with RANS based solver. It is seen
higher combustion chamber pressures. Experiments were that SA model with a constant turbulent Prandtl number
conducted in a subscale rocket combustion chamber with predicted mean temperature and velocity profiles very close
tangentially injected film of hydrogen. Tangential slot to experimental values. Dellimore [44] later used a blended
injection was investigated for various film-cooling para- k-oc , k-ω model and captured effectiveness curve moderately
meters in the same experimental set up [40]. Experimental well. Voegele et al. [48] used LOCI-CHEM RANS solver
investigations by Shine et al. [41] showed that injector to model a film cooled wall along with turbulence models
configuration had a major role in film cooling performance such as two equation shear stress transport (SST) model and
especially at higher blowing ratio. Shine et al. [42] had one equation SA model. The SA model predicted the initial
presented gaseous film cooling performance of straight decay region fairly well, SST model approached experi-
cylindrical cooling holes inside a cylindrical test section. mental values for the wall wake, but failed to predict the
Experimental and computational results showed that film film decay rate in the far field. Na et al. [49] used finite
cooling performance change with blowing ratio due to volume code with second- order upwind differencing and
differences in the magnitude of vortices created downstream eddy-diffusivity turbulence model to study film cooling. It
of the coolant injection and the jet exit momentum. Study was observed that for a flat plate configuration, SST
showed that an optimum blowing ratio exists for a given turbulence model provided better predictions, whereas the
geometric configuration. The flow field associated with realizable k-ε model was suited for a semi-cylindrical
cylindrical coolant holes inclined in tangential and azi- leading edge with a flat after body configuration. The
muthal direction, employed inside a circular pipe had been importance of radiative heat transfer in film cooled nozzle
examined by Shine et al. [43] The results indicated that was computationally analyzed by Badinand and Fransson
through the use of optimal coolant injector configurations, [50]. Finite volume method with the spectral line-based
maximum cooling performance could be achieved. Signifi- weighted sum of gray gases had been implemented along
cant increase in the heat transfer coefficients were noticed at with Navier-Stokes flow solver for calculating radiative heat
the downstream of injection for all the injector cases. transfer. It was observed that for the shocked nozzle, Mach
Review shows that there are strong effects of free stream disk emission is as intensive as the combustion chamber
turbulence, injector orientation and shock waves on the emission and the radiative heat flux is also intensive.
performance of film cooling. Effect of parameters such as Raijasse and Boccaletto [22] experimentally and numeri-
unsteady mainstream flow, shape of the coolant holes/slots, cally studied film cooling influence on the nozzle flow
and length of the coolant holes/slots etc. needs further separation phenomenon. Numerical study was carried out
investigation. A slot introduces coolant uniformly around with the RANS code named CPS developed by ONERA,
the circumference with less mixing with the overflowing Bertin Technologies and CNES. Numerical results pre-
mainstream than occurs with discrete coolant jets originat- dicted the experimental features and revealed separation
ing from a row of holes. Consequently, the slot provides an shock foot motion during film coolant injection. A
ideal performance and may be used as a basis of compar- 3DRANS solver, modified to simulate multi-component
ison. It can be noticed that some studies are carried out with mixtures of thermally perfect gases was developed by Betti
slot film cooling while others are performed with discrete et al. [52] for the analysis of liquid rocket engines. Spalart-
coolant holes. Allmaras one equation model was modified to take into
account compressibility effects had been used in the model.
Betti et al. [52] further analyzed thrust chamber film cooling
3.2. Recent numerical efforts efficiency using a multi-component RANS solver and
reproduced the test case results pertaining to oxygen/
Researchers have used Direct Numerical Simulation methane film-cooled thrust chamber. Matteo et al. [53]
(DNS), Large Eddy Simulation (LES) and Reynolds Aver- used EcosimPro, an object oriented tool and developed a
aged Navier-Stokes (RANS) to explore film cooling flows. quasi 2-D integral formulation to study a film wall jet in
The selection of method usually depends on the complexity combustion chambers. A DLR H2/O2 combustion chamber
of the problem and the fidelity of the results. Initial studies test results have been used to validate this model.
featured RANS based computations along with k-ε turbu- Maqbool et al. [54] conducted experiments with the
lence model. Stoll and Staub [16] used a finite difference objective of validating the film cooling performance predic-
boundary layer code to predict wall heat flux in a conver- tion of LOCI-CHEM, NASA's hot flow design tool. Schlie-
ging-diverging nozzle. Jansson [45] used both k-ε model ren images and heat flux measurements showed qualitative
and an algebraic stress model to simulate steady and consistency with the numerical results on most part of the
unsteady simulations. Mean velocity and temperature pro- film cooled wall with the exception on the upper non film
files were predicted fairly well for the steady state, while the cooled side. These models are capable of supporting the
Review on film cooling of liquid rocket engines 7

preliminary design of liquid rocket engine systems, analysing mixed with the mainstream in the boundary layer and
the performance of film cooling and predicting the loss in therefore exhibits a uniform temperature. The discontinuity
performance. of temperature appears only at the outer edge of the
The current computational efforts are mainly based on boundary layer. The growth of the boundary layer is similar
RANS with eddy-diffusivity or Reynolds stress models. It is to usual laws and do not have any influence of injection.
noted that LES and DNS approaches uses turbulent scales The equation for film cooling effectiveness is proposed in
of the flow directly and may be suited to film cooling flows terms of coolant Reynolds number (Rec), ratio of coolant
in contrast to RANS approach where turbulence is fully velocity to mainstream velocity (VR) and the non dimen-
modeled. Matesanz et al. [19] conducted LES and DNS sional distance from the coolant injection point (X). The
studies using a finite element Navier Stokes CFD code to following equations are proposed by various researchers.
study slot film cooling in converging-diverging nozzle. The Stollery and Ehwany [63] proposed
predictions were very good, but limited to fewer compar-
−0:8 0:2
ε ¼ 3:09 X=VR Rec ð1Þ
isons. Tyagi and Acharya, and Muldoon and Acharya
[55,56] also used LES and DNS approaches and were Kutateladze and Leont’ev [64]
successful in predicting important kinematic and thermal n
−0:8 0:25 o−0:8
properties. However, these studies were applicable to ε ¼ 3:1 4:16 þ X=VR Rec− ð2Þ
turbine film cooling cases.
Librizzi and Cresci [62]
n
−0:8 0:2 o−1
3.3. Side load analysis ε ¼ 3:0 3 þ X=VR Rec− ð3Þ

Transient nozzle side loads have been noticed for many Hartnett et al. [65]

−0:8 0:2
rocket engines during their initial development and has the ε ¼ 3:39 X=VR Rec ð4Þ
potential to cause structural damages [57–59]. When TEG
is used to cool nozzle extension part, this cause additional Tribus and Klein [66]

−0:8 0:2
side load physics due to interaction of the film coolant flow ε ¼ 4:62 X=VR Rec ð5Þ
with the Mach disk flow. Extensive experimentation has
been carried out by Boccaletto et al. [60] to characterize the It can be noted that the major differences in these
influence of coolant jet on flow separation. It was observed equations are mainly in the coefficients. This is mainly due
that positioning of the main jet separation line is clearly to the differences in the assumptions about temperature
influenced by the film cooling flow. In both steady and profile, i.e., the amount of mixing of the coolant with the
transient conditions, forward -backward movement of the mainstream. The more the mixing, the lower the coefficient.
shock system was noticed under various coolant flow rates Spalding [67] proposed a correlation based on the assump-
and the amplitude of the shock excursion was smaller in tion that the flow near the wall behaved more like a jet than a
transient conditions [22]. Wang and Guidos [6] has boundary layer. It was noticed that this assumption is valid
conducted transient 3-D numerical simulations to study when the coolant injection velocity is appreciably larger than
the side-load physics of a rocket engine consisting of a film that of the mainstream. The proposed correlation is
(  )0:5
cooled nozzle extension. During start up transients, a slight  1 

imbalance in the circumferential wall pressure between the ε ¼ 3:4 X 1−  ð6Þ
film coolant exit and the Mach disk flow separation line  VR 
caused the Mach disk flow to become asymmetric and Spalding [67] compared the theories and experimental
resulted in side loads. Wang solved transport equations data from various sources and proposed a correlation for
using a time-marching sub-iteration scheme to study the effectiveness that would fit all experimental data. Spalding
side load physics of a film cooled nozzle. Turbulence was had shown that it could be used over a broad range of
modeled based on an extended k-ε model [61] with wall geometrical slot arrangements and velocity ratio (vc/vg).
function approach. The characteristics and magnitude of the Their result for uniform properties is as follows.
side load calculations compared reasonably well with those For Xo7 : ε ¼ 1
of a LE-7A test [59]. It can be concluded that the side load For X Z 7 : ε ¼ 7=X
observed during the transient operation of film cooled Where,
rocket engines is an important issue and needs to be   0:5
investigated further.
0:8 −0:2  1 
X ¼ 0:91 X=VR Rec þ 1:41 X 1− 
VR 
3.4. Correlations for adiabatic effectiveness However, it was limited to tangential injection of the
coolant parallel to an adiabatic wall. Subsequently, Papell
Earlier theories on the prediction of film cooling effec- [68] presented further semi-empirical corrections to the
tiveness are based on the assumptions proposed by Librizzi equation to correct data obtained with coolant injection
and Cresci [62]. They assume that the coolant fluid is fully through angled slots and normal holes. Hatch and Papell
8 S.R. Shine, S. Shri Nidhi

[69] developed film cooling correlations for relatively high

gas stream temperatures. In contrast to above models, Hatch
and Papell assumed that the coolant film existed as a discrete
layer and no mixing. Empirical modifications are then made
to take care of this mixing phenomenon. Hatch and Papell's
equation for gaseous film cooling with angled slot is
  1=8
h g Lx Svg
ln ε ¼ ðmc Þc −0:04 ac f ðVRÞ þ ln cos ð0:8βef f Þ


f ðVRÞ ¼ 1 þ 0:4 tan −1 VR 1
−1 1
for VR Z1
ð7Þ
¼ 1:5  VR  ðVR−1Þ for VR o1
1

sin β
βef f ¼ tan −1 cos βþM1

Sellers [70] had shown that this equation could be used

for multiple-slot correlation as well as single-slot. Sellers
used K ¼ 0 and indicated that for rocket engine film cooling
Figure 3 Comparison of adiabatic effectiveness prediction between
experiments, a value of K ¼ 0.04 might be too large. He different studies.
argued that multiple-slot case could be analyzed by
considering the concept of progressively decreasing adia- effectiveness is given as
batic wall temperature. NASA SP 8124 [71] is based on the     
0:8
fact that the effects of turbulence between cooling film and cg x 1
ε ¼ 1:9Pr =1 þ 0:329
2=3
Reg −0:2 ω ð11Þ
mainstream cause an entrainment and is applicable to cc s M
gaseous film cooling.
The non-dimensional parameter ω depends on the angle
m_ e
ε′ ¼ ε  ð8Þ of injection between the hot-gas fluid and the film coolant.
m_ c A summary of different analytical and semi-empirical
m_ e m_ t −m_ e correlations for determining film cooling efficiency avail-
¼ f ðΨ r ; ς; x; rÞ ð9Þ able in open literature and their application in a GOX/
m_ c m_ c
kerosene combustion chamber is also described by Kirch-
Where Ψr and ς are the reference entrainment fraction and berger et al. [75].
mixing length height respectively. A correlation to predict The film cooling effectiveness values predicted by
slot film-cooling efficiency of a wall jet, under conditions of Arnold et al. [74], Simon [72] and Spalding [67] are
variable turbulence intensity, flow and temperature was compared with the experimental values of Shine et al.
developed by Simon [72]. Simon proposed the following [42] in Figure 3. Simon's model [72] predicts considerably
expressions as a result of analytical and experimental higher values compared to other models. This is due to the
considerations: increased dispersion of the coolant when injected through
1 discrete holes. Predictions from other models are very close
ε¼
x ð10Þ in the near injection regimes.
1 þ CM′ MS
The gaseous film cooling studies mentioned earlier and
Where CM′ ¼ f ðCM; δ; x; M; SÞ. their main focus of study are summarized in Table 2. It
Simon [72] considered that at the coolant exit, a semi- shows that many aspects of gaseous film cooling process
contained turbulent jet is formed near the wall. This zone is are now getting revealed. The feasibility of gaseous film
present till the inner edge of the growing main jet reaches cooling has been established and is found to be an effective
the wall. This is followed by the fully developed region cooling method both experimentally and numerically espe-
where complete mixing between the coolant and the main- cially for the nozzle cooling. It is also found that there are
stream happens. Dellimore et al. [73] extended Simon's strong influences of blowing ratios, coolant Mach number,
model to include the effect of variation in the mainstream external shock wave, free stream turbulence, and injector
pressure gradient. It was found that the response of the orientations on film cooling performance. Correlations are
coolant film to the pressure gradient depends on the velocity available to predict film cooling performance which helps in
ratio VR. A modified film-cooling model for application in a the design of cooling system of rocket nozzle. It is observed
combined convective and film-cooled combustion chamber that introduction of a coolant film has resulted in an
with an accelerated hot gas was developed by Arnold et al. increase in heat transfer coefficients downstream of injec-
[74]. The model was used to predict film cooling effective- tion. This leads to poor net heat flux reduction. There are no
ness at different combustion-chamber pressures and film apparent limitations on cooling capability, time, or chamber
blowing rates at sub-, trans-, and supersonic conditions. He pressure with film cooling. However, if one of the
used tangential injection with an angle of 15° between the propellants (usually fuel) or an inert fluid is used as a
hot gas and coolant. The prediction of film cooling coolant at the nozzle throat, there is a performance penalty
Review on film cooling of liquid rocket engines 9

Table 2 Gaseous film cooling studies applicable to rocket thrust acid oxidizer at 21.8 bar chamber pressure. Film coolants
chambers. such as water, aniline-alcohol fuel, 60-octane gasoline,
methyl alcohol, anhydrous ammonia, and JP-3 jet fuel with
Main focus of study a maximum flow rate of 7% of the engine propellant flow
Lucas and Golladay [13] Applicability of gaseous-film cooling
was used in the experiments. A reduction of heat flux to the
Back and Cuffel [15] Influence of wall cooling on the engine walls from 10% to 97% was noted. Film cooling
mainstream flow field investigations in a liquid ammonia-liquid oxygen rocket
Stoll and Straub [16] Assessment of heat transfer to nozzle wall engine was conducted by Morrell [77]. Various coolants
with film cooling studied were water, ethyl alcohol and liquid ammonia.
Arrington et al. [17] Performance in conical and bell nozzles
Martelli et al. [18] Applicability in duel bell nozzles
These experiments helped in estimating the film coolant
Matesanz et al. [19] Applicability of LES algorithms requirements of cylindrical rocket chambers having similar
Marek and Tacina [23] Effect of free-stream turbulence propellant injectors. Knuth [80] determined sufficient con-
Gau et al. [24] Effect of swirl in the mainstream ditions for the stability of thin liquid film flowing under the
O Connor and Effect of mainstream flow velocity influence of high velocity turbulent gas streams. The
Sheikh [29]
Kuo et al. [30] Effect of mainstream flow velocity
applicability of film cooling to rocket engines using earth-
Juhany et al. [31] Effect of coolant Mach number storable, space-storable and cryogenic propellant combina-
Aupoix et al. [33] Supersonic film injection tions were investigated by Stechman et al. [82]. In all
Kanda and Jiang [34] Effect of the external shock wave experiments, fuel was used as film coolant. The studies
Peng et al. [34] Effect of the oblique shock wave showed that N2O4/MMH, ClF5/MMH and other similar
Heufer and Olivier [37] Film cooling in laminar supersonic flows
Arnold et al. [39] Effect of injector design
propellant combinations are readily adaptable to small film-
Shine et al. [41–43] Effect of injector configuration cooled spacecraft engines. Totten [14] made a study on the
Spalding [67] Correlation for effectiveness behavior of a liquid film in a typical nozzle throat. It was
Hatch and Papell [69] Correlation for high gas stream observed that injecting the liquid in a straight converging
temperatures section and at the throat provided the best liquid film
Papell [68] Correlation for angled slots and normal
holes
cooling performance. Cook et al. [84] had noticed that
Sellers [70] Correlation for multiple-slot carbon deposition was the limiting factor of hydrocarbon
Simon [72] Correlation for variable turbulence fuels. Volkmann et al. [85] studied the effects of film
intensity, flow and temperature cooling on reducing the heat flux experienced at the throat
Dellimore et al. [73] Correlation for an accelerated hot gas of a rocket. Tests were conducted in a subscale LOX/RP-1
flow
Arnold et al. [74] Effect of Tangential slot injection
high pressure (138 bar) combustor. Peak heat flux reduction
Hombsch and Olivier [38] Different injection angles, coolant mass of 70% was observed with film cooling in his studies.
flow rates and free stream conditions Kirchberger et al. [86] conducted film cooling experiments
on a sub-scale heat-sink test article running on GOX and
kerosene. His results showed that kerosene was much more
(specific impulse loss). Film cooling with low molecular effective film coolant than nitrogen.
weight gases will produce less specific impulse loss and The above studies established the feasibility of using
may be more attractive. Further improvement into these liquid film cooling technique in rocket engines operating
kinds of investigations will have to focus on measurement with different propellant-oxidizer combinations. It may be
accuracy and numerical effort through LES/DNS modeling. noted that a wide variety of propellant combinations are
readily adaptable to film cooling. Engine parameters and the
coolants used in these studies are described in Table 4.
4. Liquid film cooling
Liquid film cooling process is different from that of 4.2. Injector types
gaseous film cooling because of the presence of phase
change during the cooling process which vastly increases Film cooling performance is affected by a number of
the cooling capacity. However, literature on liquid film geometric parameters. There are a few studies which deal with
cooling is quite limited compared to that of gaseous film the effect of angle of injection on film cooling performance.
cooling. Table 3 summarizes the major experimental, In most studies, coolant holes are tangentially angled to the
analytical and numerical studies conducted for liquid film hot gas flow. Coolant holes that are inclined in tangential and
cooling applicable to rocket combustion chambers. azimuthal direction toward the chamber wall are generally
referred as compound angle holes. However other factors
4.1. Feasibility studies such as shape of the hole, spacing between the holes, and
length of the hole etc. have not been discussed extensively as
One of the first experimental studies in film cooling is in the case of gas turbine film cooling [93]. Figure 4 shows
done at JPL by Boden [76]. The experimental engine was straight (holes parallel to the axial direction of the core gas
operated at 4448 N thrust with aniline alcohol fuel and nitric flow), tangential and compound angle arrangements.
10 S.R. Shine, S. Shri Nidhi

Table 3 Major liquid film cooling studies applicable to rocket thrust chambers.

Nature of study Focus of study, Injector types

Boden [76] Experimental Feasibility of films, Radial injectors

Morrell [77] Experimental Feasibility of different coolants, Vertical and Tangential slots
Kinney et al. [78] Experimental Performance study, Visualization of films, Porous and jet type
Abramson [79] Experimental Cooling of nozzles, Tangential slots
Knuth [80] Experimental Stability of liquid films, Radial injector
Warner and Emmons [81] Experimental Feasibility of H2 as a coolant, Dual slot radial injector
Stechman et al. [82] Experimental Propellants as coolant
Kesselring et al. [83] Experimental Development of analytical model, Tangential injector
Cook and Quentmeyer [84] Experimental Hydrocarbons as coolant
Volkman et al. [85] Experimental Cooling of nozzle throat
Arrington et al. [17] Experimental Cooling of bell nozzle
Kirchberger et al. [86] Experimental Kerosene as coolant
Crooco [87] Analytical Analysis of evaporation of liquid film
Knuth [80] Analytical Method for calculating evaporation rate of liquid film
Emmons [88] Analytical Determination of the heat transfer coefficient
Gater et al. [89] Analytical Analytical model including film instability and transpiration effects
Stechman et al. [82] Analytical Introduced ‘flow instability efficiency correction factor’
Grisson [90] Analytical Incorporated transpiration, radiative heat transfer and free-stream turbulence
Yu et al. [11] Analytical Swirling of the liquid film
Shembharkar and Pai [91] Numerical Couette flow model
Wang and Luong [92] Numerical Regeneratively cooled engine
Zhang et al. [12] Numerical Coolant loss is approximated by diffusion of vapor

Table 4 Comparison of various experiments.

Engine parameters Coolants used

Boden [76] I¼ 4448 N, water, aniline-alcohol,

gasoline,
Pcc ¼21.8 bar
Pcc ¼21.8 bar ethyl alcohol, anhydrous
ammonia, JP-3

Figure 4 Schematics of different cooling hole shapes is the

Morrell [77] I¼ 4448 N, water, ethyl alcohol, tangential angle and is the azimuthal angle, (from Ref. [43]).
Pcc ¼15.2–18.6 bar ammonia
Knuth [80] Test section with water effectiveness of the coolant. Morrell [77] had successfully
combustion products,
conducted liquid film cooling experiments with a vertical
Pcc E1 bar
rectangular slot injector and tangential type with slots cut at
45° to the axis. Tangential slot injection failed to improve
Stechman I¼ 44-4448 N, MMH, H2, B2H6
the effectiveness compared to the vertical slot injector.
et al. [82] Pcc ¼0.7–34.5 bar CH4
Kinney [78] used porous and jet type injectors with holes
cut at an angle of 25° to the axis. No significant difference
was noticed in the results with the two different coolant
Boden [76] used nine injector configurations out of injectors. Abramson [79] used annular slots inclined at 30°
which seven were oriented in the radial direction and two to the centre line of the nozzle in his internal film cooling
were drilled at an angle of 70° off the radial direction. The experiments of the exhaust nozzle of a liquid ammonia-
coolant was injected through multiple drilled holes dis- liquid oxygen rocket engine. Film cooling of the entire
tributed around the inner wall at one or more axial locations nozzle was achieved with film coolants such as water and
in the engine. An inner ring downstream of coolant injector anhydrous ammonia. Knuth [80] conducted experiments
was used to force the coolant axially before entering the with radial injector holes and determined sufficient condi-
engine. The data obtained was reviewed and correlated by tions for the stability of thin liquid film flowing under the
Welsh [94]. It was observed that injection from a single influence of high velocity turbulent gas streams. Gator et al.
axial location in the combustion chamber was the least [89] conducted experiments with a flat film and measured
efficient method and injection of coolant in a swirling the amount of liquid that remained attached to a wall with a
pattern had a negligible effect on cooling performance, knife-edge capture slot. Warner and Emmons [81] injected
although the addition of a deflector plate increased the coolant radially through circumferential slots inside the
Review on film cooling of liquid rocket engines 11

combustion chamber of a gaseous hydrogen-air rocket high temperatures prevailing in the combustion chamber. It
engine. He found that dual slot injection effectively reduced also does not account for the disturbances at the surface of
the quantity of coolant required to film-cool a given length the liquid film and the free-stream turbulence effects.
of surface compared to that required when a single slot is Kinney [78], Graham [97], Sellers [98] and Emmons [88]
employed. Kesselring et al. [83] performed tests with equated the convective energy transfer on the surface of the
tangential coolant injectors in a nickel calorimetric cham- liquid film from the hot gas stream to the energy utilized for
ber. The authors assumed that the liquid film immediately the phase change of the liquid coolant. It was assumed that
evaporated, based upon calculations of the normally the radiant energy transfer was negligible. Furthermore, in
expected heat flux without transpiration. Experimental and the experimental investigations conducted by Graham,
numerical investigation of tangential and compound angle Kinney and Sellers and in a portion of the experimental
cylindrical coolant holes were carried out by Shine et al. investigation conducted by Emmons, the wetted surface
[41]. It was noted that adding a compound angle produced was essentially adiabatic. Some of the unknown quantities
no improvement in effectiveness compared to tangential needed for completing the solution were approximated by
injection. Longer liquid film length for tangential injector means of empirical formulae, and the remaining were
was seen compared to the compound angle injector. determined experimentally. Sellers determined two
It can be concluded that tangential coolant injection is a unknowns and Graham and Emmons each determined one
good choice for liquid film cooling process wherein higher unknown from the experimental data. The disadvantage was
effectiveness and film uniformity is of concern. Although that the final analytical expression could not be applied with
several experimental investigations are available in the area confidence to situations that differ significantly from that
of liquid film cooling, they do not specifically deal with for which the unknowns were determined. Subsequently,
liquid film injector orientations. Therefore more investiga- heat and mass transfer analyses for the wall region wetted
tions need to be initiated to determine the relative perfor- by the liquid film was carried out by Gator et al. [89]. A
mance of different film cooling configurations. correlation procedure was suggested, which require experi-
mental data points, such as the measurement of the structure
of the boundary layer region above the liquid film. Film
4.3. Analytical and numerical models instability at the liquid-gas interface was considered as
important as transpiration effects, but the same was
The analytical and numerical models available in litera- considered only above a critical value of liquid flow rate
ture for the liquid film cooling are examined in this section. based on observations by Taylor et al. [99]. No assessment
One of the early works reported is that of Knuth [80] of the validity of the correlation procedures was carried out
wherein the analysis was based on an extension of the since accurate experimental data were not available.
Reynolds analogy to heat, mass and momentum transfer in Analytical methods suggested by Stechman et al. [82]
the turbulent core of a two component fully developed considered only the convective heat transfer from the main
turbulent pipe flow with unidirectional radial diffusion. He core gas and loss through the chamber walls. He used a
had considered only the unidirectional radial diffusion and modified Bartz [100] equation to calculate turbulent heat
neglected all other heat transfer mechanisms including the transfer coefficient from the combustion gases to the film
evaporation of liquid film due to convective and radiative coolant. The convective heat transfer coefficient from the
heat transfer. The method involves determination of the liquid film coolant to the chamber walls was calculated
evaporation rate and the surface temperature of the liquid assuming turbulent liquid flow on a flat plate and using a
wall film. Crocco [87] utilized the same interfacial energy ‘flow instability efficiency correction factor’. He assumed
balance as Knuth, and hypothesized that liquid film coolant (i) the coolant film exhibits no mixing or chemical reaction
evaporates and diffuses from the boundary towards the hot with the main core gas, (ii) the coolant film temperature
combustion products. The vapor is then confined in a profile does not change rapidly as the coolant moves down-
laminar sub layer that behaves as a thermal barrier. stream, (iii) the gradients across the coolant film are small,
Emmons's [88] analysis considered incompressible flow of and (iv) no heat transfer occurs in the chamber walls. The
the hot gas stream, flowing over a stable liquid film having predictions obtained from the model had errors ranging
a uniform temperature equal to the boiling point of the from -20% to 13% depending on the thrust chamber
liquid. The velocity profile within the sub layer was based configuration. Kesselring et al. [83] performed tests in a
upon a diffusivity variation relationship, originally sug- nickel calorimeter chamber using a propellant combination
gested by Rannie [95] and modified by Turcotte [96]. of OF2/B2H6 (oxygen difluoride/diborane). They had devel-
Turcotte's analysis of the sub layer considered the effect of oped an integral method to determine the film temperature
vapor injection upon turbulence. Using the Reynolds and film heat transfer coefficient. The model was based on
analogy, Emmons obtained an expression relating the heat an assumed cubic temperature profile through the wall.
transfer coefficient between the liquid film surface and the Liquid film cooling model by Huzel and Huang [101]
hot gas stream. None of the above analyses take into gave semi-empirical correlations for calculating the liquid
account the heat transfer by radiation which is significant at film cooled length (FCL). This is defined as the distance
12 S.R. Shine, S. Shri Nidhi

beyond which the flow ceases to exist. The entrainment of assumption of arbitrary values for pressure gradient and
the liquid film was accounted for the first time and is given local liquid film thickness. Liquid entrainment and free-
by stream turbulence effects were not considered. It is unclear
  whether the effects of transpiring vapour were properly
1 Am_c
FCL ¼ ln 1 þ ð12Þ included as a boundary condition in their model. Zhang
A V
et al. [103] have conducted numerical study of film and
where A ¼ A(Xe) and is the liquid entrainment parameter. Xe regenerative cooling in a thrust chamber at high pressure. It
is the function of ratio between the core flow temperature was observed that the liquid film reaches supercritical
and the interface temperature. After calculating the film regime at the introduction into the thrust chamber. How-
cooled length, NASA model gives an expression for the ever, it formed a thin layer with low temperature and
calculation of the film cooling effectiveness and is protected the wall. A one dimensional analytical model of
liquid film cooling in rocket combustion chambers operat-
η ¼ f ðθ; m_ c ; m_ e ; χÞ ð13Þ
ing at subcritical conditions is proposed by Shine et al.
Trotti [102] compared the predictions from the above [104]. Simplifying assumptions such as steady one-dimen-
model to that of experimental results from a high pressure sional flow, adiabatic film-wall interface, non-reactive cool-
combustion chamber which used gaseous oxygen as oxidi- ant and constant core gas temperature were made to develop
zer and Kerosene JetA-1 as fuel. He observed that this this model. The model incorporated mass transfer via
model generally over predicted in most cases. It was also entrainment by adapting suitable correlations from literature
noted that Stechman's [82] film cooling model was not pertaining to annular flow conditions. The results showed
better than NASA model as it was developed for rockets that the effects of radiation and coolant entrainment are
characterized by low combustion chamber pressure. significant. It was confirmed that the liquid film length
A numerical model, assuming a turbulent boundary layer decreases with increase in gas Reynolds number, coolant
flow for the hot gas stream and a Couette flow model for the inlet temperature and free-stream turbulence. The effect of
liquid coolant film was proposed by Shembharkarand and combustion chamber pressure was also investigated and
Pai [91]. The model predicted an exponential drop in found to be insignificant at higher pressures.
evaporation rate and did not account coolant transpiration The models described above are focused on the heat and
effects. Liquid film length predictions were significantly mass transfer at the interface. Majority of the models do not
higher than experimental results. An attempt was made by account for the interfacial instability and the annular
Grisson [90] to develop a general analysis of liquid film entrainment of liquid film. Flat plate correlations were
cooling. In Grisson's comprehensive model, transpiration generally used neglecting the effects of cylindrical combus-
effects, radiative heat transfer and free-stream turbulence tion chamber. Enough attention has not been given to the
were included. Flat plate correlations were used with a radiation heat transfer from the high temperature gases.
modified leading edge distance for convective heat transfer Recent research in annular two phase flow indicates the
calculations. The radiation calculations were based on presence of interfacial instability phenomenon affecting
Hottel's chart which over-predicted the radiative heat annular entrainment. The transpiration of vapour from the
transfer at high temperatures. The entrainment effects were liquid film decreases the normally expected convective heat
not considered in the analysis and the model was valid only flux that makes the radiation significant. Consequently the
at low coolant flow rates. Wang and Luong [92] developed results obtained through existing models differ significantly
a computational methodology to predict the hot-gas-side in the prediction of the liquid film length in practical
and coolant-side heat transfer in film cooling assisted, combustion chambers. The cooling mechanism of liquid
regeneratively cooled liquid rocket engine combustors. It film at high pressure would be different from that at low
was found that film cooling produced the low heat flux near pressure. In the supercritical regime, the flow is similar to
the injector face plate. Yu et al. [11] had performed a single-phase flow and all the thermal energy transferred
literature review on film cooling models along with the from the hot gases is devoted to heating up the film. Many
assumptions employed in the analyses. The paper described rocket engines are operating at supercritical conditions and
a model experiment and the important processes in film therefore, research on film cooling under supercritical
cooling were identified through an order of magnitude conditions would be beneficial.
analyses. Pertinent benefits of swirling of the liquid film
to reduce entrainment were also discussed in this paper.
Zhang et al. [12] numerically solved the governing equa- 4.4. Liquid film entrainment studies
tions for the liquid film and the gas stream coupled through
the interfacial matching conditions. The gas-liquid interface Limited studies are available in literature which char-
was assumed at the state of thermodynamic equilibrium. acterizes the mechanism of entrainment and film stability in
The radial component of velocity at the interface was liquid film cooling flows. Kinney et al. [78] had made
calculated based on the diffusion of coolant vapour from visual observations of liquid film flows on the inner surface
the interface to the core gas flow. The method involved the of the tubes containing flowing air. Water, water-detergent
Review on film cooling of liquid rocket engines 13

solutions, and aqueous ethylene glycol solutions were used In liquid film cooling flows, the liquid film flow is shear-
as film coolants with air stream momentum flux varying driven and the liquid film-gas core aerodynamic interaction
from 40,000 to 200,000 Pa. Kinney et al. [78] observed causes the liquid disintegration process. Therefore, a core
disturbances at the liquid-gas interface causing loss of flow Weber number, Wec representing the ratio of the
coolant, when the coolant flow rate was above a certain disrupting aerodynamic force to the surface tension retain-
value. This value was found to be varying with liquid ing force is usually used to characterize the flow conditions.
viscosity and surface tension, and did not show any change The weber number, Weg is defined as:
with air stream Reynolds number. Knuth [80] from his ρg V g 2 D
liquid-film stability experiments confirmed that longer Weg ¼ ð14Þ
σc
wavelength disturbances appeared only after some critical
flow rates of coolant. Small disturbances with wavelength The Weg values calculated for the film cooling experi-
of the order of ten times the film thickness were observed at ments are shown in Table 5. It is observed that liquid film
all flow rates. However, liquid droplets were entrained by cooling experiments are characterized by high aerodynamic
the gas stream from the crests of long wavelength dis- forces which results in higher Weg values.
turbances. Contrary to the observations by Kinney and
Knuth, Gater et al. [89] observed that the disturbances at the 4.5. Reduction in specific impulse
liquid-gas interface were dependent only on the momentum
flux of the gas stream. He proposed that the quantity of A film of liquid or gas flow through rocket nozzle throat
liquid entrained was a function of momentum flux of the with a temperature different from core gases can result in
gas and surface tension of the liquid film. However, it may performance losses [7]. Analysis by Coulbert [108] revealed
be noted that Gator's experiments were at lower gas mass that a typical performance loss due to film cooling is
flux conditions compared to earlier experiments. Coy et al. proportional to the quantity of coolant flow. The following
[105] conducted entrainment studies using slot injectors studies have reported reduction in specific impulse with film
with a Mach number of the test section of about 0.6 and gas cooling. Boden's [76] experiments showed reduction in
momentum fluxes varying from 30,000 Pa to 99,000 Pa. He specific impulse for all the film coolants. Except for low
concluded that there exists a critical flow rate of the coolant coolant flows, all film coolants caused lesser impulse
film beyond which any additional liquid injected would reduction in comparison to an inert coolant. Abramson [79]
become entrained into the gas phase. Miller and Coy [106] had reported that specific impulse reduction of the film
had measured the thickness of liquid film driven by high cooled engine was approximately 78% of the uncooled
momentum core gas flow. It was observed that at higher gas engine. The use of ammonia as a coolant resulted in a
momentum fluxes, the entrainment depended more on slightly higher specific impulse than was obtained with water
viscosity than surface tension of the liquid film. However, as coolant. Morrell's [77] experiments showed a 4% reduc-
a thorough understanding of the coolant entrainment tion in specific impulse (Isp) for water coolant flow of 5% of
mechanism is incomplete and uncertain at present and a total flow, 4% reduction in Isp for alcohol coolant flow of
detailed study in this aspect is essential to unravel the 15%, no reduction in Isp for ammonia coolant up to 11%, and
physical phenomenon behind the process. This can be due only 2% reduction in Isp for an ammonia coolant flow of
to the limitations of current instrumentation methods and 15%. Stechman et al. [82] also noted performance loss with
the complexity of the entrainment process. The liquid-gas film cooling. He assumed no mixing of the coolant and
interface characteristics of liquid film cooling were inves- mainstream and predicted the performance of the film cooled
tigated by Shine et al. [107]. It was observed that the engine. The performance obtained from his experiments was
disturbance waves started appearing at the liquid-gas inter- within the accuracy of the values predicted. Arrington et al.
face at coolant flows above a critical value. The liquid-gas [17] experiments with a standard conical nozzle and a bell
interface remained undisturbed for a fairly short distance of nozzle showed higher performance of the nozzles with lower
the order of diameter of the tube at low coolant flow rates. fuel film cooling. Film coolant injection will create a
This undisturbed distance increased with the increase in boundary layer film at a temperature different from the core
momentum flux ratio (I). It was also observed that the gas flow. The net thrust developed by the engine will be less
disturbance wave frequency mainly depended on the core than that would result if the fluid had been thoroughly mixed.
gas velocity, whereas the film thickness was a strong This gas stratification is the primary reason for the loss of
function of the momentum flux ratio. specific impulse associated with film cooling. An additional
specific impulse loss may be incurred due to the operation at
Table 5 Weg values for film cooling experiments. propellant mixture ratios other than optimum in order to
insure sufficient propellant as film coolant.
References Weg Film cooling with hydrogen, helium and ammonia is
Morrell [77] 95,000
advantageous because low molecular weight will contribute
Kinney et al. [78] 5000–65000 to a high ratio of temperature to molecular weight without
Knuth [80] 80,000–2,00,000 coolant combustion, and, hence, no appreciable performance
degradation will occur. With film cooling, there are no heat
14 S.R. Shine, S. Shri Nidhi

flux limitations as observed with regenerative cooling or time that exothermic coolants reduce the film cooling effective-
constraints with ablative cooling. Therefore the complete ness. Abramson's [79] experiments on film cooling of
cooling requirement of a rocket combustion chamber can be rocket nozzles showed that more coolant flow was required
met from the film cooling process if one is willing to pay the with a reactive coolant to cool the entire nozzle compared to
penalty resulting from temperature stratification and the an inert fluid. Welsh [94] noted that reactive film coolants
associated specific impulse loss. Other attractive features of caused less specific impulse reduction than would an inert
this process are: (i) the applicability to a wide range of engine coolant. Similar observations were obtained by Morrell [77]
size, (ii) no limit to the duration of operation for film cooling, while performing film cooling experiments inside a com-
and (iii) no storage problems involved when the fuel is used bustion chamber of liquid ammonia-liquid oxygen rocket
as the coolant. However, transient operation of the film engine. He conducted film cooling experiments with water,
cooling involves key challenges. If the film is established ethyl alcohol and liquid ammonia as film coolants. Stech-
before ignition, a fuel rich mixture in the combustion man et al. [82] observed that combustion chamber has to be
chamber will result. Starting the engine before the flow of operated fuel rich with high percentage of film cooling to
coolant will result in a deposit of condensable combustion reduce the specific impulse reduction. Takita [116] numeri-
products on the wall, which will plug the coolant injector cally studied cooling efficiency of hydrogen for a cylind-
passages [94]. The stability characteristics associated with rical body in supersonic air flow compared to cooling with
each coolant will be different and further research needs to be nonreactive gases. He observed that cooling efficiency
conducted to establish the guidelines during the starting and change is considerable according to whether a coolant is
pulsating operation of the engine. combustible or not. Kirk [117] studied surface heat flux
augmentation due to near-wall reactions over a film cooled
plate. He observed that the impact of near wall, secondary
4.6. Use of endothermic fuels
reactions on a film cooled surface is mainly a function of
5 non-dimensional groupings: Damkohler number (Da),
Coolant fuels such as liquid hydrogen, liquid ammonia
heat release potential (H*), scaled heat flux (Qs), and mass
and liquid methane contribute cooling effects through the
and momentum blowing ratios. The near wall reactions will
absorption of sensible and latent heat. A second category of
augment the surface heat flux which can be modelled as a
fuels provide endothermic reactions and offer an additional
convective heat transfer coefficient times the difference in
heat sink for cooling. The endothermic reactions will
driving (film) and the wall temperature. The driving
provide gaseous, lighter components with high heating
temperature may approach the order of adiabatic flame
values, short ignition delay time and rapid burning rates.
temperature at high Damkohler number, however the
A large number of such reactions have been reported by
changes in convective heat transfer coefficients are of lower
Lander and Nixon [109]. Sobel and Spadaccini [110] had
magnitude. Jang et al. [118] investigated film cooling of
demonstrated the use of JP-7, JP-8 and JP-10 as endother-
combustion chamber and nozzle experimentally and analy-
mic fuel for hypersonic scramjet cooling. The use of
tically for a hydrogen peroxide/kerosene bipropellant thrus-
endothermic fuels for rocket film cooling application is
ter. He predicted the film cooled length by using the model
another area of potential research. Many liquid rocket
proposed by Grisson [90] with a modification to account for
engines use RP-1 as propellant which has narrow density
the effect of the reactive coolant. The thermal decomposi-
and volatility range, lower sulphur, olefin and aromatic
tion of the hydrogen peroxide coolant was incorporated
content than aviation fuels [111]. However, RP-1 film
through an experimentally determined empirical constant.
would crack into small molecules forcing solid carbon
At higher coolant flow rates, higher rate of the thermal
and might be harmful to combustion chamber walls [112]. It
decomposition along with reduction in film cooled length
is reported that adding H2O into hydrocarbon fuel can
was observed. These studies show that film cooling with a
alleviate its coking behaviour especially at high temperature
reactive fluid is significantly important and more knowledge
[113,114]. RP-1 mixed with 5%–10% hydrogen or water is
of the process is vital to develop models which provide
proposed by Yang and Sun [115] for liquid rocket applica-
adequate levels of predictability of the film cooled length.
tions. Numerical simulations using RANS and k-ε model
was conducted and proposed that this mixture had the
potential to improve film cooling performance. An EDC
4.8. Oxidizer as the coolant fluid
model with 10 components and 17 step reaction is used for
simulating the combustion. It is also reported that adding
Few studies have been conducted to investigate the
5%–10% hydrogen to RP-1 coolant film reduced the
possibility of the use of liquid oxygen for the cooling of
specific impulse loss.
thrust chambers. Dederra and Kirner [119] had conducted
tests to study the feasibility of using liquid oxygen as a
4.7. Effect of thermal decomposition of coolant regenerative coolant. Price [120] observed that thrust
chamber cracks observed with cyclic testing with LOX as
Crocco [87] was one of the first researchers investigated the coolant have similar characteristics to those with liquid
film cooling of a reactive coolant and showed theoretically hydrogen as the coolant. He noted that in the event of small
Review on film cooling of liquid rocket engines 15

leak of coolant oxygen into the combustion chamber, the numerical investigations available pertaining to liquid film
following might happen. (i) oxidation of carbon film at the cooling. The current analytical models are very complex
chamber wall leading to higher chamber wall temperature. and contain many empirical correlations. Studies showed
This could cause oxidation of the metal and a catastrophic that film coolant entrainment results in significant loss of
failure. (ii) Liquid oxygen entering through cracks could the coolant and a reduction in liquid film cooled length. It is
film cool the carbon layer with no oxidation of carbon film clear that knowledge of this process is crucial, and
or metal wall. Therefore further experimentations were systematic visualization tests needs to be carried out with
suggested by Price [120] to confirm the two possibilities. sophisticated diagnostic techniques that are nowadays
The use of liquid oxygen as a film coolant may be an available.
interesting scenario and needs to be explored in future.

References
5. Conclusions
[1] D.K. Huzel, D.H. Hwang, Modern engineering for design
Presently film cooling has become an important technol- of liquid rocket engines, Progress in Astronautics and
ogy to tolerate high heat flux and high combustion Aeronautics, 147, AIAA, Washington, DC, 35, 1992.
temperature. This review has examined the origins of [2] D. Warren, C. Langer, History in the making the mighty
F-1 rocket engine, in: Proceedings of the 25th AIAA/
gaseous and liquid film cooling of rocket combustion
ASME/SAE/ASEE Joint Propulsion Conference, Monterey,
chambers and summarized the research carried out so far, CA, Paper No. AIAA 89-2387, 1989.
including experimental, analytical and numerical methods. [3] J.O. Vilja, G.L. Briley, T.H. Murphy, J-2S rocket engine,
The existing literature supports the idea of employing 29th AIAA/SAE/ASME/ASEE Joint Propulsion Confer-
gaseous film cooling into the throat and divergent portions ence and Exhibit, Monterey, California, Paper No. AIAA
of the nozzle of a rocket engine. Correlations are now 93-2129, 1993.
available to predict the film cooling effectiveness under [4] S. Kim, Numerical study of high area ratio H2/O2 rocket
varying injection conditions. The effect of variables such as nozzles, 24th AIAA/SAE/ASME/ASEE, Joint Propulsion
blowing ratios, injection Mach number, injector configura- Conference, Sacramento, CA, Paper No. AIAA 91-2434,
tions, accelerated hot gas, compressibility, pressure gradient 1991.
etc. have been quantified. However in some cases, increase [5] R. Arnold, D. Suslov, O.J. Haidn, Influence parameters on
film cooling effectiveness in a high pressure subscale
in heat transfer coefficient is noted which corresponds to
combustion chamber, 47th AIAA Aerospace Sciences
lesser net heat flux reduction. Current level in understand-
Meeting Including The New Horizons Forum and Aero-
ing the nozzle side loads during transient operation is space Exposition, Orlando, Florida, Paper No. AIAA 2009–
limited. The mechanisms that contribute to nozzle side 2453, 2009.
loads due to film cooling during engine transition opera- [6] T.S. Wang, M.J. Guidos, Transient three-dimensional side-
tions is to be identified and a more predictive ability has to load analysis of a film-cooled nozzle, J. Propuls. Power 25
be developed. Many studies have used subscale models to (2009) 1272–1280.
isolate the effects of different variables in their studies. [7] J.G. Campbell, D.R. Batha, M.D. Carey, A.R. Nagy, R.C.
Consequently many of the results have to be revaluated Stechman, Thrust chamber cooling techniques for space-
considering the actual operating environment of rocket craft engines, Marquardt Corp. Rept. 5981, Vol. II of
engines. Contract NAS7-103, 1963.
[8] O. Reynolds, On the resistance encountered by vortex rings
Liquid film cooling is usually employed in the combus-
and the relation between the vortex rings and the stream-
tion chamber walls with fuel as coolant. The task of
lines of a disc, Nature 14 (1876) 477–479.
developing a complete understanding of liquid film cooling [9] H. Lamb, Hydrodynamics, Cambridge University Press,
process starting from first principles is still far to be 1932.
reached, however, considerable progress have been made. [10] K. Wieghardt, On the blowing of warm air for de-icing
Models are available to predict the liquid film length with devices, FB.1900, Reports and Translations, No. 315, 1944.
reasonable accuracy for engines working under subcritical [11] Y.C. Yu, R.Z. Schuff, W.E. Anderson, Liquid film cooling
conditions. Many present engines are operating at super- using swirl in rocket combustors, 40th AIAA/ASME/ SAE/
critical conditions and therefore further research is to be ASEE Joint Propulsion Conference and Exhibit, Fort
carried out in this regard. The shape of the coolant hole and Lauderdale, Florida, Paper No. AIAA-2004–3360, 2004.
the angle of injection on film cooling performance is [12] H.W. Zhang, W.Q. Tao, Y.L. He, W. Zhang, Numerical
study of liquid film cooling in a rocket combustion
another area that needs further understanding. Tangential
chamber, Int. J. Heat. Mass Transf. 49 (2006) 349–358.
injectors are used in most cases and the effect of other
[13] J.G. Lucas, R.L. Golladay, An experimental investigation of
geometric factors has not been properly characterized. The gaseous film cooling of a rocket motor, NASA TN 0-1988,
current understanding of the liquid film entrainment effects 1963.
and physics is relatively sparse; however, the acceptance of [14] J.K. Totten, Ramjet structural elements, and non-regenera-
liquid film cooling is based on the available experimental tively cooled combustion chambers and nozzles, Vol. 9,
data. It is also observed that there are only very few Report No. 25, 116, Marquardt Corporation, 1964.
16 S.R. Shine, S. Shri Nidhi

[15] L.H. Back, R.F. Cuffel, Shock wave/turbulent boundary- [33] B. Aupoix, A. Mignosi, S. Viala, Experimental and
layer interactions with and without surface cooling, AIAA numerical study of supersonic film cooling, AIAA J. 36
J. 14 (1976) 526–532. (1998) 915–923.
[16] J. Stoll, J. Straub, Film cooling and heat transfer in nozzles, [34] T. Kanda, F. Ono, M. Takahashi, T. Saito, Y. Wakamatsu,
J. Turbomach. 110 (1988) 57–64. Experimental studies of supersonic film cooling with shock
[17] L.A. Arrington, B.D. Reed, A. Rivera, A performance wave interaction, AIAA J. 34 (1996) 265–271.
comparison of two small rocket nozzles, 32nd AIAA/ [35] W. Peng, P. Jiang, Influence of shock waves on supersonic
ASME/SAE/ASEE Joint Propulsion Conference and Exhi- film cooling, J. Spacecr. Rocket 46 (2009) 67–73.
bit, Lake Buena Vista, FL, AIAA Paper No. 96-2582, 1996. [36] X. Yang, K.J. Badcock, B.E. Richards, A numerical study
[18] E. Martelli, F. Nasuti, M. Onofri, Numerical analysis of of hypersonic laminar film cooling, Aeronaut. J. 107 (2003)
film cooling in advanced rocket nozzles, AIAA J. 47 (2009) 479–486.
2558–2566. [37] K.A. Heufer, H. Olivier, Experimental and numerical study
[19] A. Matesanz, A. Velazquez, M. Rodgriguez, Numerical of cooling gas injection in laminar supersonic flow, AIAA
simulation of slot film cooling in convergent-divergent J. 46 (2008) 2741–2751.
nozzles, 29th AIAA/ASME/SAE/ASEE Joint Propulsion [38] M. Hombsch, H. Olivier, Film cooling in laminar and
Conference and Exhibit, Paper No. AIAA 93-1977, 1993. turbulent supersonic flows, J. Spacecr. Rocket 50 (2013)
[20] Y. Watanabe, N. Sakazume, M. Tsuboi, LE-7A engine 742–753.
nozzle problems during the transient operations, 38th [39] R. Arnold, D. Suslov, O.J. Haidn, Circumferential film
AIAA/ASME/SAE/ASEE Joint Propulsion Conference cooling effectiveness in a LOX/H2 subscale combustion
and Exhibit, Paper No. AIAA 2002–3841, 2002. chamber, J. Propuls. Power 25 (2009) 760–770.
[21] M. Takahashi, T. Tomita, H. Sakamoto, H. Tamura, Y. [40] R. Arnold, D. Suslov, O.J. Haidn, Film cooling in a high-
Watanabe, M. Tsuboi, Influence of a gap for film cooling pressure subscale combustion chamber, J. Propuls. Power
on transient flow characteristics of rocket engine nozzle, in: 26 (2010) 428–438.
Proceedings of 38th AIAA/ASME/SAE/ASEE Joint Pro- [41] S.R. Shine, S.S. Kumar, B.N. Suresh, Influence of coolant
pulsion Conference and Exhibit, Indianapolis, IN, AIAA injector configuration on film cooling effectiveness for
Paper No. 2002–4147, 2002. gaseous and liquid film coolants, Heat. Mass Transf. 48
[22] P. Reijasse, L. Boccaletto, Nozzle flow separation with film (2012) 849–861.
cooling, in: Proceedings of 38th AIAA Fluid Dynamic [42] S.R. Shine, S.S. Kumar, B.N. Suresh, Internal wall-jet film
Conference and Exhibit, Seattle, Washington, Paper No. cooling with straight cylindrical holes, J. Thermophys.
AIAA 2008–4150, 2008. Heat. Transf. 26 (2012) 439–449.
[23] C.J. Marek, R.R. Tacina, Effects of free-stream turbulence [43] S.R. Shine, S.S. Kumar, B.N. Suresh, Internal wall-jet film
on film cooling, NASA TN D-7985, 1975. cooling with compound angle cylindrical holes, Energy
[24] C. Gau, K.A. Yihm, S.S. Chang, Swirling flow effect on Convers. Manag. 68 (2013) 54–62.
film cooling performance downstream of a sudden expan- [44] K. Dellimore, Modeling and Simulation of Mixing Layer
sion, J. Thermophys. Heat. Transf. 5 (1991) 89–95. Flows for Rocket Engine Film Cooling (Ph.D. thesis),
[25] E. Volchkov, S. Kutateladze, A. Leontev, Effect of University of Maryland at College Park, 2010.
compressibility and nonisothermicity on the efficiency of [45] L.S. Jansson, L. Davidson, E. Olsson, Calculation of steady
film cooling in a turbulent boundary layer, J. Appl. Mech. and unsteady flows in a film-cooling arrangement using a
Technol. Phys. 7 (1966) 93–94. two-layer algebraic stress model, Numer. Heat. Transf., Part
[26] V. Repukhov, Effects of compressibility and nonisothermal A 25 (1994) 237–258.
conditions on the performance of film cooling, J. Eng. [46] C. Cruz, Experimental and Numerical Characterization of
Phys. Thermophys. 19 (1970) 1401–1408. Turbulent Slot Film Cooling (Ph.D. thesis), University of
[27] T. Hansmann, H. Wilhelmi, D. Bohn, An experimental Maryland at College Park, 2008.
investigation of the film-cooling process at high tempera- [47] C. Cruz, A. Marshall, Surface and gas measurements along
tures and velocities, AIAA and DGLR, International Aero- a film cooled wall, J. Thermophys. Heat. Transf. 21 (2007)
space Planes and Hypersonics Technologies Conference, 181–189.
5th, Munich, Germany, 1993. [48] A.P. Voegele, A. Trouv, C. Cadou, A. Marshall, Rans
[28] D. Pedersen, E. Eckert, R. Goldstein, Film cooling with modeling of 2d adiabatic slot film cooling, 46th AIAA/
large density differences between the mainstream and the ASME /SAE/ASEE Joint Propulsion Conference and exhi-
secondary fluid measured by the heat-mass transfer ana- bit, AIAA Paper No. 2010–6735, 2010.
logy, J. Heat. Transf. 99 (1977) 620–627. [49] S. Na, B. Zhu, M. Bryden, T.I.P. Shih, CFD analysis of
[29] J.P. O’Connor, A. Haji-Sheikh, Numerical study of film film cooling, in: Proceedings of the 44th AIAA Aerospace
cooling in supersonic flow, AIAA J. 30 (1992) 2426–2433. Sciences Meeting, Reno, NV, 2006.
[30] S.L. Kuo, N.L. Zhang, W.J. Yang, Role of mainstream flow [50] T. Badinand, T.H. Fransson, Radiative heat transfer in film-
velocity in film cooling in a duct, J. Thermophys. Heat. cooled liquid rocket engine nozzles, J. Thermophys. Heat.
Transf. 10 (1996) 382–384. Transf. 17 (2003) 29–34.
[31] K.A. Juhany, M.L. Hunt, J.M. Sivo, Influence of injectant [51] B. Betti, E. Martelli, Heat flux evaluation in oxygen/
mach number and temperature on supersonic film cooling, methane thrust chambers by RANS approach, in: Proceed-
AIAA, J. Thermophys. Heat. Transf. 8 (1994) 59–67. ings of the 46th AIAA/ASME/SAE/ASEE Joint Propulsion
[32] D. Keener, J. Lenertz, R. Bowersox, J. Bowman, Tran- Conference, AIAA Paper 2010–6721, 2010.
spiration cooling effects on nozzle heat transfer and [52] B. Betti, E. Martelli, F. Nasuti, M. Onofri, Numerical study
performance, J. Spacecr. Rocket. 32 (1995) 981–985. of film cooling in oxygen/methane thrust chambers, in:
Review on film cooling of liquid rocket engines 17

Proceedings of the 4th European Conference for Aerospace [70] J.P. Sellers, Gaseous film cooling with multiple injection
Sciences, 2011. stations, AIAA J. 1 (1963) 2154–2156.
[53] D. Matteo, Francesco, M. Venanzi, M.D. Rosa, M. Onofri, [71] R.L. Ewen, H.M. Evensen, Liquid rocket engine self-
Modelling and simulation of film cooling in liquid rocket cooled combustion chambers, NASA SP 8124 (1977).
engine propulsion systems, in: Proceedings of the 48th [72] F. Simon, Jet model for slot film cooling with effect of free-
AIAA/ASME/SAE/ASEE Joint Propulsion Conference and stream and coolant turbulence, NASA TP 2655 (1986).
Exhibit, AIAA Paper 2012–3908, 2012. [73] K.H. Dellimore, C. Cruz, A.W. Marshall, C.P. Cadou,
[54] D. Maqbool, K. Dellimore, M. Collett, A. Voegele, J. Ruf, Influence of a stream-wise pressure gradient on film-cool-
A. Trouve, C. Cadou, Measurements of film cooling ing effectiveness, J. Thermophys. Heat. Transf. 23 (2009)
performance in supersonic environments, in: Proceedings 120–128.
of the 52nd AIAA Aerospace Sciences Meeting - AIAA [74] R. Arnold, D. Suslov, O.J. Haidn, Film cooling of
Science and Technology Forum and Exposition, AIAA accelerated flow in a subscale combustion chamber, J.
Paper No. 2014–0701, 2014. Propuls. Power 25 (2009) 443–451.
[55] M. Tyagi, S. Acharya, Large eddy simulation of film [75] C. Kirchberger, G. Schlieben, O.J. Haidn, Assessment of
cooling flow from an inclined cylindrical jet, ASME J. analytical models for film cooling in a hydrocarbon/gox rocket
Turbomach. 125 (2003) 734–742. combustion chamber, in: Proceedings of 48th AIAA/ASME/
[56] F. Muldoon, S. Acharya, Analysis of k-ε budgets for film SAE/ASEE Joint Propulsion Conference & Exhibit, 2012.
cooling using direct numerical simulation, AIAA J. 44 [76] R.H. Boden, Heat transfer in rocket motors and application
(2006) 3010–3021. of film and sweat cooling, ASME Trans. 73 (1951) 385–390.
[57] H.A. Cikanek, Characteristics of space shuttle main engine [77] G. Morrell, Investigation of internal film cooling of a 1000-
failures, in: Proceedings of 23rd AIAA Paper Joint Propul- pound thrust liquid ammonia liquid oxygen rocket, Tech-
sion Conference, AIAA 87-1939, 1987. nical Report, NACA RME51E04, 1951.
[58] L.H. Nave, G.A. Coffey, Sea level side loads in high-area- [78] G.R. Kinney, A. Abramson, J.L. Sloop, Internal film cooling
ratio rocket engines, in: Proceedings of 9th Propulsion experiments with 2 and 4 in. smooth surface tubes and gas
Conference, AIAA Paper 73-1284, 1974. temperatures to 2000- F in 2 and 4 in. diameter horizontal
[59] Y. Watanabe, N. Sakazume, M. Tsuboi, LE-7A engine tubes, Technical Report, NACA Report 1087, 1952.
nozzle problems during the transient operations, in: Proceed- [79] A.E. Abramson, Investigation of internal film cooling of
ings of the 38th AIAA/ASME/SAE/ASEE Joint Propulsion
exhaust nozzle of a 1000-pound thrust liquid ammonia-
Conference & Exhibit AIAA Paper 2002–2384, 2002.
liquid oxygen rocket, Technical Report, NACA
[60] L. Boccaletto, P. Reijasse, J.P. Dussauge, Dynamic beha-
RME52C26, 1952.
viour of a separation oblique shock during film cooling
[80] E.L. Knuth, The mechanism of film cooling, (Ph.D. thesis),
transient injection, in: Proceedings of the 37th AIAA Fluid
California Institute of Technology, 1954.
Dynamics Conference and Exhibit, AIAA Paper 2007-
[81] C.F. Warner, D.L. Emmons, Effects of selected gas stream
4003, 2007.
parameters and coolant properties on liquid film cooling,
[61] Y.S. Chen, S.W. Kim, Computation of turbulent flows
Trans. ASME J. Heat. Transf. 1 (1964) 271–278.
using an extended k-ε turbulence closure model, NASA
[82] R.C. Stechman, J. Oberstone, J.C. Howell, Design criteria
CR-179204, 1987.
for film cooling for small liquid-propellant rocket engines,
[62] J. Librizzi, R.J. Cresci, Transpiration cooling of a turbulent
J. Spacecr. Rocket. 6 (1969) 97–102.
boundary layer in an axis-symmetrical nozzle, AIAA J. 2
[83] R.C. Kesselring, R.M. Knight, B.L. McFarland, R.N.
(1964) 617–624.
[63] J. Stollery, A.A.M.E. Ehwany, A note on the use of a Gurnitz, Boundary cooled rocket engines for space storable
boundary-layer model for correlating film-cooling data, proplellants, Technical Report, Rocketdyne Report No.
Intern. J. Heat. Mass Transf. 8 (1965) 55–66. R-8766, 1972.
[64] S.S. Kutateladze, A.I. Leontev, The heat curtain in the [84] R.T. Cook, R.J. Quentmeyer, Advanced cooling techniques
turbulent boundary layer of a gas, Therm. Phys. High. for high pressure hydrocarbon fueled rocket engines, in:
Temp. 1 (1963) 250–258. Proceedings of the 16th AIAA/SAE/ASME Joint Propul-
[65] J.P. Hartnett, R.C. Birkebak, E.R.G. Eckert, Velocity sion Conference, Hartford, Connecticut, Paper No. AIAA
distributions, temperature distributions for air injected 80-1266, 1980.
through a tangential slot into a turbulent boundary layer, [85] J.C. Volkmann, L.M. Tuegel, J.M. McLeod, Gas side heat
J. Heat. Transf. 83 (1961) 293–306. flux and film coolant investigation for advanced LOX/
[66] M. Tribus, J. Klein, in: A. Symposium (Ed.), Forced Hydrocarbon thrust chambers, in: Proceedings of the 26th
Convection from Non-isothermal Surfaces, University of AIAA/SAE/ ASME/ASEE Joint Propulsion Conference,
Michigan Press, Ann Arbor, Mich, 1952. Orlando (Fl.), Paper No. AIAA 90-2184, 1990.
[67] D.B. Spalding, Prediction of adiabatic wall temperatures in [86] C. Kirchberger, G. Schlieben, A. Hupfer, H. Kau, P.
film cooling systems, AIAA J. 3 (1965) 965–967. Martin, S. Soller, Investigation on film cooling in a
[68] S.S. Papell, Effect of gaseous film cooling of coolant Kerosene/GOX combustion chamber, in: Proceedings of
injection through angled slots and normal holes, Report the 45th AIAA/ASME/SAE/ASEE Joint Propulsion Con-
No. NASA TN D-299, 1960. ference and Exhibit, Denver, Colorado, Paper No. AIAA
[69] J.E. Hatch, S.S. Papell, Use of a theoretical flow model to 2009–5406, 2009.
correlate data for film cooling or heating an adiabatic wall [87] L. Crocco, An approximate theory of porous, sweat, or film
by tangential injection of gases at different fluid properties, cooling with reactive fluids, J. Am. Rocket Soc. 22 (1952)
Report No. NASA TN D-130, 1959. 331–338.
18 S.R. Shine, S. Shri Nidhi

[88] D.L. Emmons, Effects of selected gas stream parameters [104] S.R. Shine, S.S. Kumar, B.N. Suresh, A new generalised
and coolant physical properties on film cooling a rocket model for liquid film cooling in rocket combustion cham-
motor, Technical Report, Report No. TM-62-65, Purdue bers, Int. J. Heat. Mass Transf. 55 (2012) 5065–5075.
University, 1962. [105] E.B. Coy S.A. Schumaker M.A. Lightfoot. Film cooling of
[89] R.A. Gater, M.R. Lecuyer, C.F. Warner, Liquid-film cool- liquid hydrocarbon engines for operationally-responsive
ing, Its physical nature and theoretical analysis, Technical space access, AIR FORCE RESEARCH LAB EDWARDS
Report, No. JPC-413, Jet Propulsion Center, Purdue Uni- AFB CA PROPULSION DIRECTORATE, No. AFRL-RZ-
versity, Lafayette, 1965. ED-TP-2010-087, 2010.
[90] W.M. Grisson, Liquid film cooling in rocket engines, [106] R.P. Miller, E.B. Coy, Studies in optimizing the film flow
Technical Report AEDC-TR-91-1, DTIC, 1991. rate for liquid film cooling, in: Proceedings of the 47th
[91] T.R. Shembharkar, B.R. Pai, Prediction of film cooling AIAA Joint Propulsion Conference, San Diego, Paper No.
with a liquid coolant, Int. J. Heat. Mass Transf. 29 (1986) AIAA 2011–5779, 2011.
899–908. [107] S.R. Shine, S.S. Kumar, B.N. Suresh, Numerical study of
[92] T.S. Wang, V. Luong, Hot-gas-side and coolant-side heat wave disturbance in liquid cooling film, Propuls. Power
transfer in liquid rocket engine combustors, J. Thermophys. Res. 2 (2013) 107–118.
Heat. Transf. 8 (1994) 524–530. [108] C.D. Coulbert, Selecting cooling techniques for liquid
[93] D.G. Bogard, K.A. Thole, Gas turbine film cooling, J. rockets for spacecraft, J. Spacecr. Rocket 1 (1964) 129–139.
Propuls. Power 22 (2006) 249–270. [109] H. Lander, A.C. Nixon, Endothermic fuels for hypersonic
[94] J.W.E. Welsh, Review of results of an early rocket engine vehicles, J. Aircr. 8 (1971) 200–207.
film-cooling investigation at the Jet Propulsion Laboratory, [110] D.R. Sobel, L.J. Spadaccini, Hydrocarbon fuel cooling
Technical Report TM-32-58, Jet Propusion Center, Cali- technologies for advanced propulsion, J. Eng. Gas. Turbines
fornia Institute of Technology, 1961. Power 119 (1997) 344–351.
[95] W.D. Rannie, Heat transfer in turbulent shear flow, J. [111] T.M. Lovestead, T.J. Bruno, A comparison of the hyperso-
Aeronaut. Sci. 23 (1956) 485–489. nic vehicle fuel JP-7 to the rocket propellants RP-1 and RP-
[96] D.L. Turcotte, A sublayer theory for fluid injection into the 2 with the advanced distillation curve method, Energy Fuels
incompressible turbulent boundary layer, J. Aeronaut. Sci. 23 (2009) 3637–3644.
27 (1960) 675–678. [112] C.M. Lee, K. Kundu, W. Acosta, Jet – a reaction mechan-
[97] A.K. Graham, in: Film Cooling of Rocket Motors, Part I: ism study for combustion application, in: Proceedings of
An Experimental and Theoretical Investigation of Film 27th AIAA/ASME/SAE/ASEE Joint Propulsion Confer-
Cooling of Rocket Motors, (Ph.D. thesis), Jet Propulsion ence, Paper 91–2355, 1991.
Center, Purdue University, Lafayette, 1958. [113] D.L. Linne, M.L. Meyer, T. Edwards, Evaluation of heat
[98] J.P. Sellers, Experimental and Theoretical Study of the transfer and thermal stability of supercritical JP-7 fuel,
Application of Film-cooling to a Cylindrical Rocket Thrust NASA Reports-Vol.107485, 1997.
Chamber, (Ph.D. thesis), Jet Propulsion Center, Purdue [114] T.S. Wang, Thermophysics characterization of kerosene com-
University, 1958. bustion, J. Thermophys. Heat. Transf. 15 (2001) 140–147.
[99] N.H. Taylor, G.F. Hewitt, P. Lacey, The motion and frequency [115] W. Yang, B. Sun, Numerical simulation of liquid film in a
of large disturbance waves in annular two-phase flow of air- liquid oxygen/rocket propellant 1 liquid rocket, J. Thermo-
water mixtures, Chem. Eng. Sci. 18 (1963) 537–552. phys. Heat. Transf. 26 (2012) 328–336.
[100] D.R. Bartz, A simple equation for rapid estimation of rocket [116] K. Takita, Film cooling effect of hydrogen on cylinder in
nozzle convective heat transfer coefficients, Jet. Propuls. - supersonic airflow, J. Spacecr. Rocket. 34 (1997) 285–289.
ARS J. 27 (1957) 49–51. [117] D.R. Kirk, Near-wall reaction effects on film-cooled surface
[101] D.K. Huzel, D.H. Huang, Design of liquid propellant rocket heat transfer, Diss. Mass. Inst. Technol. (2003).
engines, NASA SP 125 (1971). [118] D. Jang, Y. Kwak, S. Kwon, Design and validation of a
[102] M. Trotti, Modelling of liquid film cooling in a GOX liquid film-cooled hydrogen peroxide/kerosene bipropellant
kerosene rocket combustion chamber, Graduate thesis, thruster, J. Propuls. Power 31 (2015) 761–765.
POLITECNICO DI MILANO, 2012. [119] H. Dederra, E. Kirner, High pressure rocket engine liquid
[103] H.W. Zhang, Y.L. He, W.Q. Tao, Numerical study of film oxygen technology, in: Proceedings of 37th Congress of the
and regenerative cooling in a thrust chamber at high International Astronautical Federation, IAF Paper-76-174, 1976.
pressure, Numer. Heat. Transf., Part A: Appl. 52 (2007) [120] H.G. Price, Cooling of high-pressure rocket thrust chambers
991–1007. with liquid oxygen, J. Spacecr. Rocket 18 (1981) 338–343.