Sie sind auf Seite 1von 21

42nd AIAA/ASME/SAE/ASEE Joint Propulsion Conference & Exhibit AIAA 2006-4534

9 - 12 July 2006, Sacramento, California

Integrated Modeling and Analysis for a LOX/Methane


Expander Cycle Engine: Focusing on Regenerative Cooling
Jacket Design

Reuben Schuff*, Matthew Maier*, Oleg Sindiy*, Christopher Ulrich*, and Susanne Fugger*
Purdue University, West Lafayette, IN, 47907

A fully integrated engine analysis methodology was created and employed to study a 25,000
pound-force thrust class expander cycle liquid oxygen/methane engine for Crew Exploration
Vehicle applications. All major components for the engine were modeled, including the
combustion thermal chemistry, the turbopump assembly, the regenerative cooling jacket and the
injector. Significant attention was assigned to the regenerative cooling analysis with respect to
varying channel geometry dimensions. Chamber pressures ranging from 500 to 1000 psia, were
parametrically studied for varying aspect ratios of 4 to 8, channel widths of 0.020 to 0.080 inches,
and 100 to 200 number of channels. Two of the three geometric parameters were held fixed,
while trends in the third were studied at constant chamber pressures. The channel cross-section
was held constant with respect to axial position in the engine. From the geometry cases studied,
the initial baseline channel geometry was determined and chamber pressures were studied in
detail from 500 to 1000 psia. A chamber pressure of 850 was determined as most favorable with
respect to overall engine performance and design metrics. Using the initial baseline geometry, a
constant cross-section and stepped channel design were compared. The performance for the
stepped channel configuration is notably superior to the constant cross-section channel case with
respect to pressure loss, and chamber mass metrics. Detailed results are presented for all engine
components for the most favorable design.

Nomenclature
A = area ρ = density
AR = aspect ratio η = efficiency
D = diameter µ = viscosity
H = height Subscripts
L = length am = arithmetic mean between freestream and wall values
P = pressure c = methane coolant
Pr = Prandtl number cham = chamber
Re = Reynolds number chan = cooling channel
T = temperature cu = copper
dx = axial step in chamber fin = fin
f = friction factor g = combustion gas side
convective heat transfer
h = h = hydraulic
coefficient
k = thermal conductivity lig = ligament
p = perimeter o = stagnation
t = thickness r = stagnation recovery
w = width rib = fin rib
m = mass flow sw = skirt wall
q ′′ = heat flux wall = wall of the combustion chamber fin
Q = heat transfer rate wc = wall coolant side
ε = emissivity wet = wetted surface
σ = Stefan-Boltzman constant wg = wall combustion gas side

*
Graduate Student, School of Aeronautics and Astronautics, 315 N. Grant St., AIAA Member.

Copyright © 2006 by the American Institute of Aeronautics and Astronautics, Inc. All rights reserved.
I. Introduction

I n the United States, methane fueled in-space propulsion is being explored as part of the national initiative to
return people to the moon and to expand human exploration to Mars. This study was initiated in anticipation of
the requirements for the Crew Exploration Vehicle service module main engine. A methane fueled expander cycle
engine capable of delivering vacuum thrust in the range of 25,000 lbf was assumed. This paper provides an
overview of the advantages of using a methane fueled expander cycle engine, an overview of the analysis performed
to parametrically study the performance characteristics of a liquid oxygen/methane (LOX/CH4) expander cycle
engine, with particular focus assigned to selection of channel geometry in the regenerative cooling jacket, and
selection of a chamber pressure with respect to the overall performance of the engine.

The results presented in this study are from a fully integrated MATLAB simulation for an expander cycle engine
design which includes routines for engine performance, turbo-machinery, regenerative cooling jacket and injector
analyses. The performance of each routine was validated with respect to existing data in the literature. In cases
where specific design parameters and subsystem data were unavailable for comparison, the routines employed
historical design guidelines. Furthermore, significant attention was devoted to parametrically studying the impact of
the cooling channel geometry on the regenerative cooling channel jacket performance. Also, a range of chamber
pressures was investigated for the baseline geometry and a discussion on the implication of chamber pressure with
respect to design parameters and the overall engine performance is presented.
Turbine
Fuel Pump
II. Expander Cycle Engine Oxidizer Pump
LCH4 LOX
Over four decades of continuous reliable operations of
Pratt & Whitney’s RL-10 LOX/hydrogen, upper stage
expander cycle engines make these engines ideal for the
Valve
renewed need for access from low-earth-orbit to the
International Space Station and plans for future Lunar and
Injectors
Mars exploration. The expander cycle engines are known CH4 gas

Chamber
for their longevity and capability for multiple restarts,
throttling, high potential for maximum specific impulse in Regenerative
Cooling
a minimum dimensional envelope, and prolonged record
of safety.1 The expander cycle engine, shown in Figure 1,
is the simplest of the pump-fed engines because it does not
require components such as a gas generator or preburner,
has higher performance potential than engines driven by
gas generator cycles, and lower turbopump pressure
requirements than staged combustion engines.2 Nozzle

III. Methane Fueled Engines Figure 1. LOX/CH4 Expander Cycle Engine Flow
Consideration of methane as a fuel for space propulsion is a relatively new trend in the aerospace industry. This
trend arises from the many advantages that a methane-fueled propulsion system can provide over modern systems
and over other proposed hydrocarbon systems. Methane is of particular interest in the United States because it can
be theoretically produced on Mars for manned and unmanned space flight applications.3

Methane, as a rocket fuel, has many performance and property characteristics that make it desirable for applications
in aerospace propulsion. Out of common hydrocarbons, methane has the highest vacuum specific impulse of about
370 seconds; in comparison, kerosene, the closest chemical hydrocarbon has a vacuum specific impulse of about
355 seconds. Methane is a soft cryogenic that is not corrosive and has very low toxicity; thus it is easier to store,
requires less insulation and fewer handling concerns than comparable hydrogen fuel systems.1 Liquid methane is
about six times denser than liquid hydrogen; thus, methane tanks weigh much less and/or require less storage
volume than comparable hydrogen tanks. Furthermore, methane has exceptional heat capacity properties that
provide “superior cooling properties” for expander cycle engine applications; methane fuel has lower pressure drops
in regenerative cooling channels as compared to kerosene fuel.4 Several studies show that methane coking and soot
decomposition are of minimal concern as compared to kerosene and other complex hydrocarbon fuels; this allows
for reusability, multiple-restarts and longer burn times.4 Additionally, methane is a natural gas that is relatively easy
to extract on Earth and is about 5 to10 times cheaper to acquire and store than liquid hydrogen.

2
Although no known LOX/CH4 engines have been flown for aerospace applications, space agencies in the United
States, Russia, Europe and Japan have been considering methane fueled propulsion systems for various applications;
these include: first stage, second stage, booster, upper stage and in-space main propulsion system.

Single-shaft (direct drive) turbopump, methane fueled expander cycle engines are not only feasible, but are often the
optimal solutions for hydrocarbon engines under investigation for future use in aerospace applications.4,5 In the
United States, Pratt & Whitney has performed two recent studies on modifying its existing RL10B-2 LOX/hydrogen
expander cycle engine to run on LOX/CH4 propellant.1,6 These two studies were used to validate the power head
simulation.

IV. Analysis Methodology


In order to investigate the effects of cooling channel design for an expander cycle, an integrated engine system
analysis must be performed. The top level approach used for the simulation analysis is summarized in Figure 2.

CF curve fit f(Pc,ε) Chamber


CEA
Pc, F, ε & Nozzle O/F
F Performance
At = Geometry
PcCFηCFηc*
Comb Gas
Props f(Z)
AR, N, b Massflows
Isp
Update
Channel ηfp ηop, ηt
Geometry Update
∆Pfp

Power Head
no

Pumps Update Pinj,f = Pinj,o Pc,calc = Pc


∆Pfo
Guess: no
∆Pfp, yes yes
P,T ∆Pfp
∆Pfo ∆Pfo
ηfp, ηop,
ηt PT no
Regen Injector
Coolin Turbine ηfp,calc = Analysis
/ Pump η Results
Fluid Sizing
yes
Props

Figure 2. Engine Simulation Flowchart

The first step in this simulation requires the selection of a design case: chamber pressure, thrust, and expansion ratio.
For all design studies performed, the desired thrust and the expansion ratio of the nozzle were kept constant. The
throat area is calculated from a curve-fit for the perfectly expanded thrust coefficient. The chamber and nozzle
geometry are defined by the calculated throat area and specified values for expansion ratio, contraction ratio,
characteristic length, and design parameters for curvature and turning angles given by Ref. 7. Using the chamber and
nozzle geometry, the variable gas properties are obtained for the combustion products.

Subsequently, guesses for turbine and turbopump efficiencies must be made with a guesses for the required pressure
rise across the fuel and oxidizer pumps. Once the geometry is obtained, the cooling channel geometry is generated
for the desired case. There are three nested convergence looped within the main routine. The first convergence loop
within the power head routine updates the LOX pump pressure rise to yield equal injection pressure for both the CH4
and LOX. During the convergence, for matching LOX/CH4 chamber pressure, the regenerative cooling analysis is
performed to specify the turbine inlet conditions, and to determine if the current cooling geometry allows for engine
operation. The next convergence loop updates the fuel pump pressure to obtain the specified chamber pressure. The
injector for the engine is also analyzed to determine the pressure drop, injector geometry, and the efficiency for the
characteristic velocity. Once the power head has converged on the desired chamber pressure, the pumps and turbine
are sized, and efficiencies are calculated. The final convergence loop updates the pump and turbine efficiencies.
The fuel pump efficiency is the criteria for convergence; the LOX pump and turbine efficiency variations are
significantly less than for the fuel pump. Fuel and oxidizer injection temperatures are also updated at this point in
order for a more accurate calculation of the combustion gas properties.

3
A. Performance Analysis
Given desired propellants and performance conditions, such as chamber pressure and O/F (Oxidizer to Fuel) ratio,
chamber and nozzle geometry is then required to continue the analysis of the engine. The next important element
for the design of the baseline geometry is the throat area, which is calculated using the equation for thrust.8
The thrust coefficient is approximated by a 3D curvefit of the perfectly expanded thrust coefficient, which is a
strong function of expansion ratio and a weak function of chamber pressure. A characteristic velocity efficiency, ηc*,
of 0.97 and a thrust coefficient efficiency, ηcf, of 0.95, were included to calculated throat area. The combustion gas
thermal chemistry was analyzed marching through the chamber and nozzle using the NASA Chemical Equilibrium
Analysis program (CEA).9 This allows variable gas transport properties to be available for the heat transfer analysis.
B. Thermal Analysis Approach
In an expander cycle engine one of the controlling parameters is the amount of energy absorbed by the coolant in the
regenerative jacket. This energy transfer is directly related to the size and performance of the rocket engine which is
the main goal of our integrated engine analysis; a great deal of concern was therefore placed in the results of the
thermal analysis. The analysis scheme marches through the cooled nozzle and chamber, beginning at the aft most
location where the coolant enters the jacket and continues until the coolant exits the channel near the injector face.
At each discrete axial location, dx, a control volume analysis is performed.
1. Balanced Heat Rate Approach
The basic analysis approach employed was to assume a steady state quasi one-dimensional energy balance across
the regenerative cooling jacket. To further simplify the analysis additional assumptions were made: all energy
transferred across the coolant wall is absorbed by the coolant, at each axial step the heat transfer coefficient along
the coolant wall is constant, and axial temperature gradients at each step in the coolant channel are ignored. The
general equation describing the energy balance is given in Eq. (1):

k
Q = hg (Tr − Twg ) Ag = cu (Twg − Twc ) Ag = hc (Twc − Tc ) Alig + q " fin A fin (1)
tw
In Eq. (1), Alig is the ligament or non-finned convecting surface area in the x-z plane and Afin is the convecting
surface area of the fin or rib in the x-y plane, as shown in Figure 3.
The combustion gas side heat transfer coefficient was found from
a form of Bartz10 correlation in Eq. (2):
0.8 0.2
k  0.4  ρ am   µ am 
hg =  g 0.8
 0.026 Re Pr     (2)
  ρ   µ0 
 Dg 
The combustion gas properties for Eq. (2) were updated at each
step along the combustion chamber and nozzle using the CEA
code.9 The coolant side heat transfer coefficient was calculated
using a Seider-Tate11 correlation presented in Eq. (3):
0.14
k  0.8 1  µ 
hc =  c  0.027 Rec Prc 
3 c
 (3)
 Dh   µ wc 
As shown in Figure 3 the rib surface provides additional wetted
coolant surface area providing a fin effect for the cooling channel.
To calculate the additional heat transfer from the finned surface
area, a fin efficiency was calculated as outlined in Ref. 12
assuming an adiabatic tip as shown in Eq. (4).
η fin = tanh ( mH c ) ( mH c ) (4)
Where m, an intermediate parameter, is defined by Eq. (5):
m = 2hc ( wrib + dx) (k wall wrib dx) (5)
Figure 3. 2-D Single Coolant Channel
From the fin efficiency, a corresponding heat flux is calculated for Cross-Section Schematic
the fin as shown in Eq. (6).

4
q" fin = hc (Twc − Tc )η fin (6)

Nozzle Skirt Thermal Analysis: Balanced Heat Flux


The nozzle skirt analysis is very similar to that of the energy balance presented in the previous section. From Eq. (7)
we can see the difference from the cooling channel energy balance; the cold wall side is irradiated to an assumed
surrounding temperature of 0 oR. Additionally the hot combustion gas side surface area is assumed to be
approximately equal to the cold side surface area, resulting in a flux balance.

k ws
q′′ = hg (Tr − Twg ) = (Twg − Twc ) = ε wsσ sbTwc4 (7)
t ws

Where in Eq. (7) σsb is the Stefan-Boltzman constant and εws is the emissivity of the niobium nozzle skirt material.
2. Pressure Loss across Cooling Channel
The pressure loss across the coolant channel was calculated with Eqs. (8) and (9), with the mass flow per channel.
The friction factor is calculated per recommendations given in Ref. 18. The entrance and exit losses to and from the
cooling channel, from the respective manifolds, was accounted for by an additional pressure reduction scaled with
dynamic pressure. However, upon examining the results of the analysis, the dynamic pressures are extremely low
and the resulting entrance and exit losses are nearly negligible in comparison to the pressure loss across the channel

∆P = ( fLcham pwet Achan ) q (8)

q = m 2 2 ρ chan Achan
2
(9)

3. Thermal Analysis Validation


Two different sets of data were available for validation purposes, a computation study presented in Ref. 13 and
experimental data from the RL-10.14 The study in Ref. 13 provided data for a LOX/CH4 system using high aspect
ratio cooling channels. However, there were several differences in the design, specifically design thrusts were much
lower, on the order of 1000 lbf and consequently the Reynolds number regime differed from our design case. The
RL-10 data provided a comparison using a similar thrust and Reynolds numbers, however, the RL-10 is a
LOX/hydrogen system using tubular cooling channels. A detailed comparison to both sets of data allowed us to
examine the fidelity of our thermal model. Using Bartz correlation to predict the gas side heat transfer coefficient,
Eq. (2), and the Seider-Tate correlation to predict the coolant side heat transfer coefficient, Eq. (3) was determined
to be suitable within engineering accuracy. The friction factor used in the pressure loss calculation, Eq. (8), was
modified by a factor of four based on the comparison to both sets of data.
C. Turbo-machinery Analysis
In the turbopump assembly power balance, the power extracted by the turbine from the supercritical methane fuel
must fulfill the combined power needs of the fuel and oxidizer pumps; this allows the engine to operate at the
desired chamber pressure. To perform a turbopump assembly power balance analysis, propellant storage properties,
effects of the feedlines, engine performance, regenerative cooling, and injector analyses must be integrated into this
routine. A design analysis reference15 was used as a primary source for procedures needed to perform the turbo-
machinery analysis.

Typical recommended tank storage pressure and temperatures for LOX are 43 psi and 175°R and for methane are 25
psi and 200°R; these values can also be assumed as the input parameters to the single-stage pumps in the turbo-
pump assembly.6 Typical pressure losses in the feedlines are in the range of 5 to 15%. In this study, pressure line
loss values from Ref. 1 were used for comparison purposes; line losses of 10% were assumed between the LOX
pump exit and injector entry, 2.5% between the methane pump and regenerative jacket entry and 2% between the
turbine exit and methane injector entry.1

5
D. Injector Analysis
In order to complete a comprehensive engine analysis, the effects of the injector on the flow of the LOX and
methane must be examined. Based on the available references, we developed a preliminary iterative injector design
methodology that is capable of converging over a range of injector inlet conditions. The injector analysis consists of
calculations for injector pressure drops for fuel and oxidizer; as well as combustion efficiency, fuel to oxidizer
injection velocity ratio and the primary geometric design for the injector elements. The geometric design of the
injector analysis was driven by the structural considerations.

The pressure drop calculations across the fuel and oxidizer injectors were based on procedures described in literature.
7,16,17
The procedures for expansion and viscous loss calculations for manifold and dome were obtained from Ref. 18.
To evaluate the injector geometry, an empirical correlation obtained from Ref. 19 was used to calculate the
characteristic velocity efficiency. Additionally, a stress analysis for the injector was integrated, based on Ref. 20.
An injector mass estimation was obtained based on simple geometrical shapes and material properties.
1. Injector Geometry Design
For a propellant combination of liquid oxygen and gaseous methane, coaxial concentric tubes are recommended as
they provide reasonable wall compatibility, low pressure drops and excellent atomization, which has been
demonstrated by their extensive use on the J-2, RL-10, and M-1.16 Further more, an aluminum alloy, such as Al-
5083, is recommended for most injector components as it has been extensively and successfully used with most
propellant combinations and provides a good corrosion-resistance as well as good low-temperature ductility, which
is required for cryogenic propellants.16 Copper is recommended for the material of the injector face plate because its
melting point and allowable stress are capable of sustaining the chamber temperatures and thermal stresses expected
for this engine.

6.
5
4

7 . 62
The injector analysis provides
geometric values for LOX post
diameter, LOX post thickness,
methane post diameters, methane
manifold height, dome thickness,
dome plate thickness, and face plate
thickness, while the geometries of
igniter, flange, and inlet of LOX and
6
1
.
7
3
1
R
6
.
3
5

methane are kept constant. The


injector and injector post schematics,
respectively, are shown in Figure 4
and Figure 5; a description of the
geometric parameters is given in
Table 6 where the final design point
Figure 5. Injector results are given.
Figure 4. Injector Schematic
Post Schematic

2. Injector Optimization Analysis


To optimize the injector geometry for a desired pressure drop and high injector performance, every station of the
injector has to be evaluated for both LOX and methane characteristics and performance. Since the injector thermal
analysis to determine the actual temperatures at every station was outside the scope of this study, constant densities
and dynamic viscosities for both LOX and methane, obtained from the injector inlet conditions, were assumed. A
flowchart of the injector analysis is provided in Figure 6.

6
The injector analysis logic converges to match two different
criteria: it forces the pressure drop across LOX and methane Pressure Drop Calculation
injector elements to be in the range of 10% to 12% of the
inlet pressure for LOX and 14% to 18% for methane by Analysis of pressure drop
altering the injector post diameters, and it additionally forces
the injector geometry to provide a sufficient combustion
efficiency value as this is a revealing parameter to evaluate Increase
Update
Analysis of dynamic
pressure in manifold
number of
the injector. elements Geometry

Stress analysis
One main influence on combustion efficiency is thrust per
element, T/E, corresponding to the varying number of
injector elements. Typically, increasing the number of No
Do all analyses match
Yes
their criteria?
elements results in a higher efficiency; however, T/E also
affects combustion stability; a higher T/E yields higher Efficiency = desired
No
stability. At the same time it provides less combustion efficiency?
efficiency as the number of elements decreases, while a
lower value of T/E provides less stability but higher
Yes
efficiency.

Ref. 21 suggests that reasonable values for this type of Results Mass Estimation

design are 100 lbf to 250 lbf thrust per element because for
Figure 6. Injector Analysis Flowchart
values less than 100 lbf T/E combustion instabilities may
occur, while the combustion efficiency decreases rapidly for values greater than 250 lbf T/E. This design guideline
was used to determine the minimum and maximum number of injector elements.

As shown in Figure 6, the optimization scheme also encompasses a check for the dynamic pressure within the
injector manifold. To avoid high viscous losses in the manifold, a dynamic pressure of less than 1% of the static
pressure is desired. Moreover, a stress analysis is performed based on procedures described in Ref. 20 and
correlations provided by Ref. 22 and 23. The stresses and deflections were not allowed to exceed limits induced by
material properties and maximum allowable deflection values. For both injector plates, a conservative safety factor
of 5 was used to account for the fact that the present calculations for stress analysis do not consider the effects of the
holes in the plates.

The total mass estimation of the entire injector system is based on the volume of simple geometric shapes, i.e. a
frustum represents the thrust transmitting cone, a ball cap represents the dome, tubes represents the injector posts,
and disks represent all plates. The injector mass analysis also encompasses the flanges for the dome, the dome plate
and the face plate.

V. Design Study Methodology


The methodology of the design study centered on determining a cooling channel baseline geometry by studying a
limited number of channel geometry parameters. Then, chamber pressures effects, given the baseline cooling
channel geometry, were studied to select a design chamber pressure. Finally, the design was improved by adding a
stepped cooling channel configuration.
A. Geometric Parametric Study
The cooling channel geometry is a principal parameter controlling the regenerative cooling thermal analysis.
Varying the channel geometry has profound implications on the heat transfer and pressure loss analyses.
Additionally, the channel geometry impacts the chamber and nozzle mass. While there are only two parameters
which define the channel at any axial location along the engine, namely width and height, allowing these to vary
with respect to axial location in the engine introduces a large number of independent parameters. A parametric
approach is taken to understand the basic trends of varying the channel geometry parameters.

In order to study the effects varying the channel geometry has on the regenerative cooling thermal analysis, a
baseline case was required. Implementing a channel geometry which is invariant with respect to axial location in the
chamber and cooled nozzle drastically reduces the number of independent parameters controlling the geometry. This

7
condition on axial variation facilitated studying the effects of cross sectional shape and allowed selection of an
initial baseline geometry.

Figure 7 shows a cross-section of a coolant channel at an arbitrary axial location in the chamber. The land or rib
thickness, as denoted in the figure, is a function of the available chamber (or nozzle) circumference and therefore
mandates knowledge of the chamber (or nozzle) radius at the axial location where the channel is being defined. To
define the chamber and nozzle radial coordinate as a function of axial coordinate the approximations given in Ref. 7,
24, and 25 are employed. A characteristic length L* of 35 inches is selected based on available data in Ref. 7. For

Figure 7. Detail of Cooling Channel Cross-Section Referenced from the Center of a Radial Cross-
Section of the Chamber
chamber radii much greater than the channel height, the land or rib may be approximated as rectangular. θ is defined
as 360 degrees over the number of channels. The channel width and land thickness are related by Eq. (10), where
chamber wall thickness is a specific constant and chamber radius R is dictated by the aforementioned
approximations used to determine the chamber and nozzle geometry. Finally the channel height is known from a
specified aspect ratio AR, where AR is the ratio of channel height to width.

wchan = R + t wall tan θ − wrib (10)

Table 1: Channel Geometry Used in Parametric Study to In an effort to detail how varying the
Determine Baseline Cooling Channel Geometry, Pc = 750 psia channel geometry parameters affect the
Aspect Ratio Channel Width [inches] Number of Channels thermal analysis of the regenerative cooling
4,5,6,7,8 0.050 150 jacket, the aspect ratio, channel width, and
number of channels were parametrically
0.020, 0.035, 0.0425, studied at a constant chamber pressure.
6 0.050, 0.0575, 0.065, 150 Table 1 lists the scenarios investigated; from
0.0725, 0.080 this study, we are able to arrive at a baseline
100, 112, 125, 137, cooling channel geometry. All cases used a
6 0.050 nominal chamber pressure of 750 psia
150, 162, 175, 187, 200

B. Chamber Pressure Study and Stepped Geometry


Upon determining a baseline channel geometry from the aforementioned parametric study, the chamber pressure
was varied from 500 to 1000 psia. The design point chamber pressure is determined primarily from the total mass of
the engine and the performance of the turbopump assembly. The chamber pressure is limited by the turbine inlet
temperature and the maximum gas wall temperature. The turbine inlet temperature did not prove a limiting factor,
however the maximum gas wall temperature is an important constraint.

After a design chamber pressure was established, improvements to the engine mass and pressure loss in the cooling
channel are achieved by implementing a stepped channel. A channel step occurs both forward and aft of the throat
region; these axial locations are determined based on the region of elevated heat transfer rate near the throat.

8
C. Fabrication Constrains of Channel Geometry
The fabrication constraints Table 2. Fabrication Criteria Available in Literature
available in the literature were Ref 26 (1998) Ref 13 (1982) Ref 27 (1976)
used as design guidelines. The Channel width (wcham), inch ≥0.02 ≥0.03 ≥0.04
most recent data is from Ref. 26, Channel height (Hc), inch ≤0.20 NA NA
therefore it is considered most Land thickness (wrib), inch ≥0.02 ≥0.03 NA
applicable. Data on the channel Aspect ratio (Hc/wchan) ≤8 ≤4 NA
wall thickness is taken from Ref. Wall thickness (twall), inch NA ≥0.025 ≥0.028
13 and Ref. 27 . Table 2
summarizes the fabrication constraints.
D. Channel Structural Analysis
A preliminary analysis was performed to define an envelope of cooling channel geometries. The fabrication
constraints set the lower limits for the majority of the key parameters, i.e. channel width, channel height, and land
thickness. The creep and low cycle failure mechanisms were studied by reviewing historical data given by Ref. 28.
For the operating times and design cycles, these modes of failure were not significant.

The pressure stresses and thermal stresses were analyzed individually. Ref. 29 states that the maximum stress in the
cooling channel is due to the bending stress at the channel midpoint. The thermal stress is approximated by
modifying equations from Ref. 7 by replacing the heat flux term with the thermal conductivity α and the temperature
difference across the channel as given by Ref. 29. The result of this approximation yields a design limit of 0.100
inches for the channel width.

VI. Results & Discussion


A. Engine Performance

Table 3. Baseline LOX/Methane Engine Parameters In order to find the effects of varying cooling
Required Thrust 25000 [lbf] channel geometry the baseline chamber and nozzle
O/F 3.5 presented in Table 3 was established. This baseline
Contraction Ratio 3.5 nozzle was necessary to establish the combustion gas
Expansion Ratio 180 properties, which are crucial for a proper
Cooling Channel Expansion 15 thermodynamic analysis, and to evaluate the effects
Characteristic Length 35 [in] of the changing cooling channel geometry on the
Contraction Angle 20o requirements for the power head.
o
Percent of 15 Cone 85%
The characteristic length, contraction angle, and
percent of 15o cone were determined historically. In order to generate the baseline nozzle geometry for a 25000 lbf
LOX/CH4 engine the oxidizer to fuel ratio, expansion ratio, and cooling channel expansion ratio must be chosen.

Figure 8 shows that there is an optimal value for O/F ratio in terms of specific impulse. This value is 3.3 but a value
of 3.5, as seen in Table 3, was used for all scenarios in this study. This was suggested by previous studies.1

An expansion ratio for the cooling channels was chosen at the point where radiation cooling was not sufficient to
keep the nozzle wall temperature below the melting point. This nozzle skirt was analyzed as niobium, melting point
of 4400 oR. Examining Figure 9, an expansion of 15 was chosen to stay within 90% of this value.

9
Figure 8. Vacuum Specific Impulse vs. Oxidizer to Fuel Figure 9. Gas Side Wall Temperature Vs. Expansion
Ratio Ratio (Pc = 850)

Studying how Isp varies with expansion ratio, shown in Figure 10, there is not an obvious choice for an expansion
ratio. In trying to achieve a high performance for an engine design a high Isp is desired, i.e. higher expansion ratio.
However, since as expansion ratio increase, the engine diameter increases, this was the limiting factor in choosing
expansion ratio. Given typical limitations on exit diameter for upper stage engines an expansion ratio of 180 was
chosen from Figure 11 for the baseline geometry.
8

6
Exit Diameter [ft]

Pc
Pc==1000
1000[psi]
[psi]
2
Pc
Pc==700
700[psi]
[psi]
Pc
Pc==400
400[psi]
[psi]
1
0 20 40 60 80 100 120 140 160 180 200
Expansion Ratio
Figure 10. Vacuum Specific Impulse vs. Expansion Figure 11. Engine Diameter vs. Expansion Ratio
Ratio
B. Channel Geometry Variations on Regenerative Jacket
As previously demonstrated in the fundamental equations involved in the thermal analysis, increasing aspect ratio
has two competing effects on the heat transfer. The first effect is that increasing aspect ratio, given and fixed channel
width, serves to increase the hydraulic diameter and wetted perimeter of the channel. This affects the coolant side
heat transfer coefficient, in a decreasing manner, two fold because the hydraulic diameter appears in the
denominator of the heat transfer coefficient and the wetted perimeter appears in the denominator of the Reynolds
number as shown in Eq. (11). The decreasing heat transfer coefficient consequently decreases the heat transfer rate.
Second, increasing aspect ratio increases the area of the fin, and consequently increases the heat transfer rate.
Increasing channel width has an analogous effect to increasing aspect ratio, the primary difference being that the
channel width acts on the area of the ligament, not on the area of the fin.

Rec = 4m µc pwet (11)

10
The heat transfer coefficient is the dominating effect of the heat transfer rate because two of its constituent terms,
hydraulic diameter and Reynolds number, are affected in the same manner with changing aspect ratio. The area of
the fin and ligament are present singularly in the numerator of the heat rate balance; this results in the heat transfer
effects being dominated by the hydraulic diameter and wetted perimeter effects on the heat transfer coefficient. The
fin area effect is additional retarded by the fin efficiency term, which decreases with increasing aspect ratio.

An increasing heat transfer rate is characterized by decreasing the maximum gas wall temperature and increasing the
temperature rise in the channel. Figure 12 and Figure 13 show the resulting effects on the heat transfer rate for the
maximum gas wall temperature and the temperature rise in the channel, respectively. As expected, the dominance of
the hydraulic diameter and wetted perimeter causes the heat transfer rate to decrease with increasing aspect ratio
(given a constant width). Consequently, the maximum gas wall temperature increases and the temperature rise
across the channel decreases. Similarly, increasing channel width has the same effect; the maximum gas wall
temperature increases and the temperature rise across the channel decreases.

At first examination these treads may appear contrary to the traditional wisdom that increasing the fin size increases
the heat transfer capabilities of the system. However, this traditional wisdom holds valid; but the competing effects
between the heat transfer coefficient and area terms must be isolated to observe this. It is possible to plot a curve on
Figure 12 and Figure 13 which holds the coolant heat transfer coefficient constant. This is accomplished by fixing
the product of wetted perimeter, raised to the 0.8, and hydraulic diameter. That is to say, holding the relationship in
Eq. (12) constant holds the heat transfer coefficient constant. A curve satisfying the condition of Eq. (12) is added to
Figure 12 and Figure 13.
(1 Pwet )0.8 Dh = const (12)
p p ( [p ] )
3200 820
Wth = 0.0425 [in]
Wth = 0.0575 [in]
3000 Wth = 0.0725 [in] 800
Constant h c

780
2800
Increasing Width
760
2600
[o R]

∆ T c [ R]
max

740
Twg

2400

720
Increasing Width
2200
700
Wth = 0.0425 [in]
Wth = 0.0575 [in]
2000 Wth = 0.0725 [in]
680
Constant hc

1800 660
4 4.5 5 5.5 6 6.5 7 7.5 8 4 4.5 5 5.5 6 6.5 7 7.5 8
Channel Aspect Ratio Channel Aspect Ratio

Figure 12. Maximum Gas Wall Temperature vs. Figure 13. Temperature Rise Across Cooling Channel
Aspect Ratio vs. Aspect Ratio

Following the curve for constant heat transfer coefficient reveals that increasing aspect ratio does in fact increase the
heat transfer effects. That is, as aspect ratio increases, noting that we are following the curve of constant coolant side
heat transfer coefficient, the heat transfer rate increases, causing the maximum gas wall temperature to decrease and
the temperature rise across the channel to increase. This supports our conception that the larger fin provides more
heat transfer. Here we have decoupled the competing effects driving the heat transfer coefficient and area terms.

In comparing the two points where the curve of constant heat transfer coefficient crosses the curves of constant
width, we are forced to deal with yet another coupling of effects, the area of the ligament and the area of the fin. The
ligament and fin area are not independent as we travel along the curve of constant heat transfer coefficient. As we
jump from one curve of constant width to another, with a lesser width, the ligament area decreases, while increasing
aspect ratio serves to increase the fin area. The fact that increasing aspect ratio results in decreasing the maximum
gas wall temperature and increasing the temperature rise in the channel shows that the fin area effect dominates.
This also follows our intuition, being that the fin has a much larger area than the ligament.

11
1400 The effects on pressure loss across the cooling channel
depend on two terms, the hydraulic diameter and cross-
1200
Wth = 0.0425 [in]
Wth = 0.0500 [in]
sectional area of the channel. Both terms appear in the
Increasing Width
Wth = 0.0575 [in]
Wth = 0.0650 [in]
denominator of the pressure loss equation; however the
1000 Wth = 0.0725 [in]
Wth = 0.0800 [in]
cross-sectional area term is raised to the second power.
Constant h c This quadratic dependence results in the shapes of the
800 curves shown in Figure 14. This clearly shows that
∆ P c [psi]

higher cross-sectional areas are desired, and


600 consequently higher aspect ratios and channel widths.

400 The number of channels affects three terms in the heat


transfer rate equation, the coolant heat transfer
200
coefficient, the area of the ligament and the fin
efficiency. For a given channel geometry, chamber and
0
4 4.5 5 5.5 6 6.5 7 7.5 8 nozzle, as the number of channels increases, the mass
Channel Aspect Ratio
flow per channel decreases, which decreases the
Figure 14. Pressure Loss Across the Cooling Reynolds number and consequently directly decreases
Channel vs. Aspect Ratio the heat transfer coefficient. Similarly, the ligament
area decreases with increasing number of channels. The effects of the decreasing heat transfer coefficient and
ligament area both serve to decrease the heat transfer rate. The land width, which is a term in the fin efficiency
calculation, decreases with increasing number of channels. However, a global statement can not be made for how
the fin efficiency varies with respect to the number of channels. The fin efficiency is calculated as a function of axial
location in the chamber and cooled nozzle, and is a complex function of land width, coolant heat transfer coefficient
and channel height. The complex variations in the fin efficiency are responsible for the treads seen in Figure 15 and
Figure 16. We see that there is an optimal point with respect to maximum gas wall temperature near 150 channels.
773
2370 Pc = 750 [psi]
Pc = 750 [psi] 772

2360
771

2350
770
Twg max [oR]

769
∆ Tc [ R]
o

2340
768

2330 767

766
2320
765

2310 764
100 110 120 130 140 150 160 170 180 190 200 100 110 120 130 140 150 160 170 180 190 200
Number of Channels Number of Channels

Figure 15. Maximum Gas Wall Temperature vs. Figure 16. Temperature Rise Across Cooling Channel
Number of Cooling Channels vs. Number of Cooling Channels
600
Pc = 750 [psi]
550 As presented in Figure 17, the pressure loss varies
500 quadratically with the number of channels. This is a direct
450
consequence of the mass flow term in the pressure loss
calculation.
∆ Pc [psi]

400

350

300

250

200

150
100 110 120 130 140 150 160 170 180 190 200
Number of Channels

Figure 17. Pressure Loss Across the Cooling


Channel vs. Number of Channels

12
C. Baseline Channel Geometry
The parametric study on the cooling channel geometry parameters allowed the selection of an initial baseline
channel geometry. As previously noted, this channel geometry is invariant with engine axial location. A guiding
constraint on selecting the channel geometry is that the maximum hot gas wall temperature must be limited below
the melting point of Oxygen Free High Conductivity copper, which is 2440 oR. The best design choice for the
channel geometry maximizes the temperature rise across the channel, while minimizing the pressure loss incurred.
The maximum gas wall temperature and the temperature rise across the channel both show that small widths are
advantageous, however the converse is true of the pressure loss. We are able to achieve the highest thermal
performance while maintaining a reasonable pressure drop with an aspect ratio of 8, and a channel width of 0.0425
inches. The baseline design uses 150 cooling channels. This is based on the optimal point with respect to maximum
gas wall temperature as previous discussed. The maximum gas wall temperature is a limiting factor in the maximum
operable chamber pressure. Therefore, minimizing the maximum gas wall temperature is considered primarily
important. The temperature rise across the channel and the pressure drop for the 150 channel point are acceptable for
the design.

Figure 18 shows the geometry for the baseline case selected. The left hand side of the figure shows the cross-
sectional view of the channels; note that the figure only shows the machined channel, the top face would be closed
out by the cooling jacket, typically electrodeposited nickel as in the case of the SSME.27 The right hand side of the
figure shows the chamber and cooled portion of the nozzle. Respective axial locations corresponding to the channel
cross-sections shown on the left are given, however, this not significant here because the channel shape is invariant
with axial location. This channel geometry facilitates the study on chamber pressure variations, and is useful in
describing the geometry effects on the thermal parameter, but is not used as the final design point case.

Figure 18. Baseline Channel Geometry


D. Chamber Pressure Effects on Regenerative Cooling Jacket
2500 Overall, the thermal parameters, Twg and ∆Tc, saw an
increase in magnitude with an increasing chamber pressure.
2400
This is a direct result of the increase in physical thermal
conditions in the combustion chamber which in effect
increase the combustion gas side heat transfer and therefore
Twg max [oR]

2300
the temperatures seen. In Figure 19, the trend of increasing
combustion gas side wall temperature with chamber
2200 pressure is evident.

2100
The selected baseline design point chamber pressure allows
a 4% margin to the melting temperature of OFHC. This
aggressive design allowed an overall minimum engine
2000
500 600 700 800 900 1000 weight as described in following sections. It is recognized
Chamber Pressure that this temperature is well above the typical structural
Figure 19. Combustion Gas Side Wall limits for OFHC, however other cooling methods, not
Temperature vs. Chamber Pressure included in the model such as fuel film cooling, can be used
to mitigate this high wall temperature.

13
E. Chamber Effects on Turbopump Assembly
The direct drive single-shaft, as oppose to a geared or dual shaft, turbopump assembly was chosen for the
turbopump assembly operations because historically it requires the least amount of hardware design and control
complexity, is the cheapest to produce and 0.8
provides the safety record needed for manned in-
space propulsion applications. Direct drive 0.75
single-shaft turbopumps are designed to rotate at
a fixed rotational speed for the fuel and oxidizer 0.7

Fuel Pump Efficiency


pump; this allows for minimum hardware
0.65
complexity and cost but also induces the fuel
pump to operate at lower efficiencies as 0.6
compared to a geared or dual-shaft systems.
Specifically, in these operations, the fuel pump 0.55
is forced to rotate at the same speed as the
0.5
oxidizer pump because of cavitation Direct Drive Shaft
Geared Shaft
considerations, which in turn causes the fuel 0.45
pump to operate at lower stage specific speeds 500 600 700 800 900 1000

and thus lower efficiencies. Figure 20 shows Chamber Pressure


that 15 to 30 % penalty in fuel pump efficiency Figure 20. Fuel Pump Efficiency vs. Chamber Pressure for
occurs for direct drive operations for the Direct Drive & Geared Operations
chamber pressures explored.

For the cooling jacket channel geometry and chamber pressure scenarios explored in this study, Figure 21
demonstrates that the LOX pump for a single stage scenario remains relatively close to the peak of pump efficiency
versus stage specific speed curve. On the other hand, the methane pump, which is already forced to operate at
lowered efficiency for the single-shaft direct-drive operations, is further forced to operate at lower efficiencies for
increased chamber pressures; this confirms the notion that increased chamber pressure requirements further strain
the fuel pump performance to provide the higher
0.9
pressure rise. Figure 22 shows the empirical
efficiency curve used in the estimation of the 0.85

fuel and oxidizer pump calculations. 0.8


Turbine
0.75 Ox Pump
Furthermore, the turbine efficiency is based on a Fuel Pump
0.7
parabolic curve with respect to the Parson Factor
Efficiency

0.65
(isentropic spouting velocity of the turbine over
the maximum turbine pitch velocity).5 For the 0.6

baseline scenario explored, the turbine efficiency 0.55


is actually able to increase its efficiency as 0.5
chamber pressures increase; this behavior is
0.45
produced because the chamber pressures
explored place the turbine efficiency on the 0.4
500 600 700 800 900 1000
ascending portion of the efficiency vs. Parson Chamber Pressure
Factor curve (as per Figure 23). In order to Figure 21. Turbopump Efficiencies vs. Chamber Pressure
maximize the total turbine assembly for Baseline Scenario
performance, Figure 21 shows that a chamber
pressure near the intersection of the turbine efficiency and the pump efficiency curves, about 850 psia, is desired.

14
0.7

0.6

0.5

Turbine Efficiency
0.4

0.3

0.2

0.1

0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1

Isentropic Spouting Velocity / Max Turbine Pitch Velocity

Figure 22. Pump Efficiency vs. Stage Specific Figure 23. Single Stage Turbine Efficiency vs.
Speed (Ref. 15) Parson Factor (Ref. 5)
F. Chamber Pressure Effects on Injector
The thrust per element ratio was calculated for
the range of the chamber pressures studied and is
1000
comparable to historical data. Figure 24 shows
SSME
the range where our engine is placed in
comparison to the historical data of other Apollo Lunar Descent J-2
Thrust per Element [lbf]

3 2
S M URF
engines for thrust per element verses chamber Design Range
pressure.16,30 Aestus 20kN Type
100
(4496 lbf)
The analysis performed for the injector RL 10A-3
optimization with respect to varying chamber
pressures shows several important trends for the
geometric parameters explored. As observed Aestus 3kN type
from Figure 25, the mass of the injector (674 lbf)

decreases with higher chamber pressures, 10


because higher pressures reduce the chamber 100 1000
Chamber Pressure [psi]
diameter. Additionally, the number of injector
elements required also decreases with increasing Figure 24. Engine Placement in Comparison to Historical
chamber pressure as shown in Figure 26. This Data16,30
behavior can be explained with the iteration on combustion efficiency, which is calculated with the empirical
correlation from Ref. 19: the combustion efficiency increases with a higher equivalent chamber length; this length
increases with the square root of the chamber pressure and reciprocally proportional to the chamber diameter, which
leads to fewer elements being needed to meet the required combustion efficiency.
50 190

48
180
46
Number of Elements

170
44
Mass [lbm]

160
42

40 150

38
140
36
130
34
120
32

30 110
500 550 600 650 700 750 800 850 900 950 1000 500 550 600 650 700 750 800 850 900 950 1000

Chamber Pressure [psi] Chamber Pressure [psi]


Figure 25. Injector Mass vs. Chamber Pressure Figure 26. Number of Injector Elements vs.
Chamber Pressure

15
Figure 27 shows that, except for a slight increase for the dome, the thicknesses of the injector components stay
reasonably constant with varying chamber pressure requirements; this is explained by the fact that the static pressure
in the dome raises proportional to the chamber pressure while the pressure differences loading the other injector
elements stay at a nearly constant level. This yields to the conclusion that the geometry for the first guess might need
to be changed to lower values to decrease the mass of the injector. But, nevertheless, the stress analysis of the
injector dome plate and of the injector face plate is only a first order approximation and should be either revised or
affirmed by a more detailed structural analysis, in the future.

For increasing chamber pressure and at the same time decreasing mass flow, we observed from Figure 28 that the
methane element areas decrease while the LOX areas increase. This may be explained by the fact that a decreasing
mass flow yields lower velocities on the same area; thus, the area has to be increased to obtain sufficient results, and
as the number of elements goes down with higher chamber pressure, the area has to be increased to get a reasonable
pressure drop.

This is only true for LOX as it is assumed to act as an incompressible liquid and therefore is assumed to have a
nearly constant density and dynamic viscosity. For methane, the increasing chamber pressure causes higher
densities; thus, the area has to be decreased to obtain the required pressure drop for methane.
0.7 0.02
Injector Geometric Parameter [in]

tfaceplate LOX
CH4
0.6 hm an 0.018
tdom eplate

0.5 tdom e 0.016


tLOX
Area [in2]

0.4 0.014

0.3 0.012

0.2 0.01

0.1 0.008

0 0.006
500 550 600 650 700 750 800 850 900 950 1000 500 550 600 650 700 750 800 850 900 950 1000
Chamber Pressure [psi] Chamber Pressure [psi]
Figure 27. Injector Geometry vs. Chamber Pressure Figure 28. Injector Post Geometry vs. Chamber
Pressure
The trend in the methane post area shown in Figure 28 may be explained by Figure 29. Due to the complexity of the
convergence scheme the inlet pressure for methane strongly increases for chamber pressures above 800 psia in
contrast to the LOX inlet pressure. This yields a higher pressure drop for methane and thus smaller injector areas,
while the LOX injector post areas decrease as they have to be readjusted.
90 11.5

11
80

10.5
Pi,CH4 - Pi,LOX [psi]

70
10
v CH4/v LOX

60
9.5

50 9

8.5
40
8

30
7.5

20 7
500 550 600 650 700 750 800 850 900 950 1000 500 550 600 650 700 750 800 850 900 950 1000

Chamber Pressure [psi] Chamber Pressure [psi]


Figure 29. Injector Inlet Pressure Difference Figure 30. Injection Velocity Ratio vs. Chamber
Pressure

16
Figure 30 shows that the velocity ratio decreases with higher chamber pressure; this is explained by the fact that, as
shown in Figure 31, at a constant mixture ratio the density ratio of LOX to methane decreases faster than the area
ratio. For a better view, Figure 32 shows the area ratio itself.

100 0.8
Density and Area Ratios LOX/CH4

Density Ratio
90 Area Ratio 0.75
combined
80
0.7
70
0.65

ALOX/ACH4
60

50 0.6

40 0.55

30
0.5
20
0.45
10

0 0.4
500 550 600 650 700 750 800 850 900 950 1000 500 550 600 650 700 750 800 850 900 950 1000

Chamber Pressure [psi] Chamber Pressure [psi]


Figure 31. Comparison of Area Ratio and Density Figure 32. Injector Post Area Ratio
Ratio vs. Chamber Pressure
G. Chamber Pressure Effects on Systems Masses
Figure 33 presents the variations on mass of each engine component as well as the total engine mass. We show that
the mass of the injector, chamber, regenerative cooling, and nozzle skirt decrease as chamber pressure increases.
This is expected because as chamber pressure increases, the throat, i.e. the engine, shrinks for a given desired thrust
and expansion ratio. The total turbopump assembly mass is proportional to the power required from the turbine. As
the desired chamber pressure increases, the power needed from the turbine increases to provide the higher pressure
rises needed across the pumps and consequently, the overall turbopump assembly mass increases. For the fixed
regenerative cooling geometry described, Figure 33 shows that increasing the desired chamber pressure increases the
turbopump assembly mass. A relative minimum for the total engine mass appears, however the increase in
performance from increasing chamber pressure results in a much larger mass savings for the total vehicle. This
results in the desired chamber pressure to be as high as the cooling cycle and power head allow it to be.

450

400

350

300 Total
Power Head
Mass [lbm]

250 Injector
Chamber+Regen
200 Skirt

150

100

50

0
500 550 600 650 700 750 800 850 900 950 1000
Chamber Pressure

Figure 33. Engine Mass vs. Chamber Pressure

17
H. Implementation of Stepped Channel and Impact on Design
The channel geometry transitions from the baseline geometry, both forward and aft of the throat region, to a larger
width, lower aspect ratio geometry. The locations for these transitions were determined by the region of elevated
heat flux near the throat. The overriding goal was to reduce the pressure loss through the channel without impacted
the thermal performance. Figure 34 shows the cross sections for the channel forward, aft and at the throat. The step

Figure 34. Stepped-Channel Geometry


transitions occur at the given axial location shown on the right hand side of the figure. Table 4 summarizes the step
channel design. A comparison of the hot gas wall side temperature between the baseline design and the stepped
channel is given in Figure 35 and the key parameters are compared in Table 5. We determined with such a large
decease in ∆Pc the resulting decrease in ∆Tc tradeoff was acceptable.
Table 4. Summary of Step Channel Design
2500
Throat Fwd & Aft of Constant x-section
Region Throat Region Stepped
AR 8 1.33 2000
Width [in] 0.0425 0.150
Number of Channels 150 150
1500
Contraction/Expansion NA 3.5/2.8
Twg [ oR]

Ratio for Step


1000
Table 5. Summary of Stepped Channel
Improvements
500
Constant Cross-Section Stepped
Pin [psia] 2299 1838
Pout [psia] 1762 1680 0
-15 -10 -5 0 5 10 15
∆Pc [psi] 537 158 Axial Length [in]
∆Tc [oR] 766 737 Figure 35. Gas Wall Temperature vs. Streamwise
Engine Mass [lbm] 327 274 Axial Position

It is well documented that near the throat the gas side heat transfer coefficient increases significantly and as a result
the wall temperature also increases. To adequately deal with the high heat transfer rate, the geometry remained the
same as defined for the baseline case in this region. In the reduced combustion gas heat transfer regions upstream
and downstream of the throat we increased the channel cross-sectional area to decrease the pressure drop and
maintain a combustion gas side wall temperature less than at the throat, as shown in Figure 35. Consequently,
increasing the channel cross-sectional area Achan, decreased the coolant heat transfer coefficient hc, therefore the
stepped-channel case presented a tradeoff in performance parameters, temperature rise (∆Tc) and pressure drop (∆Pc)
of the coolant along the entire length of the channels. The ∆Pc and ∆Tc decreased 340% and 4%, respectively.

18
VII. Summary of Final Design Point Results

The injector optimization code estimated that 124 elements are needed in the injector; Table 6 provides the values
for the design point injector geometry. The overall design of the turbopump assembly for stepped cooling channel
geometry and 850 psia chamber pressure is shown in Table 7. The selected design point has the following
parameters as seen in Table 8 and a full schematic of the engine is presented in Figure 36.

Table 6. Injector Design Point Geometry Table 7. Turbopump Assembly


Geometric Rotational
Description [in] 32726
Parameter speed [rpm]
DiLOX LOX Post Inner Diameter 0.104 Mass [lbm] 116
DoLOX LOX Post Outer Diameter 0.164 CH4 LOX
DCH4 CH4 Annulus Diameter 0.209 Turbine
Pump Pump
tLOX Post Thickness 0.03 Efficiency [%] 59 85 53
R Recess Length 0.2 NPSH [ft] 60 197 ---
Digniter Igniter Diameter 0.3 Impeller Tip
Manifold Diameter = Dome 770 352 ---
Dman = Ddome = Dc 8.346 Speed [ft/s]
Diameter = Chamber Diameter Power [hp] 501 240 741
DinletLOX LOX Inlet Diameter 0.75 Inlet Tip
DinletCH4 CH4 Inlet Diameter 1.25 1.59 1.75 ---
Diameter [in]
tdomeplate Dome Plate Thickness 0.4 Exit Tip
tfaceplate Face Plate Thickness 0.6 5.39 2.46 ---
Diameter [in]
hman Manifold Height 0.5 Mean Pitch
Dome Thickness = Cone --- --- 9.19
tdome = tcone 0.259 Diameter [in]
Thickness
r1 Ball-Cap Radius 5.01
tflange Flange Thickness 1.0
hflange Flange Height 0.2

Table 8. Engine Parameters


Vacuum Thrust [lbf] 25,900
Chamber Pressure [psia] 850
Mixture Ratio 3.5
Engine Flow Rate [lbm/sec] 68.05
LOX flow [lbm/sec] 52.93
Fuel flow [lbm/sec] 15.12
Delivered Vacuum Isp [sec] 380.6
Throat Area [in2] 15.63
Throat Diameter [in] 2.23
Contraction Ratio 3.5
Expansion Ratio of Cooled Nozzle 15
Expansion Ratio of Entire Nozzle 180
Turbine Bypass [%] 10
Characteristic Exhaust Velocity Efficiency ηc* 0.97
Thrust Coefficient Efficiency ηcf 0.95

19
η = 53%
P = 25 η = 59% RPM = 32726 η = 85% P = 43
T = 200 HP = 501 HP = 240 T = 175
m = 15.12 m = 52.93
LCH4 LOX
P = 1670 P = 1015.8
T = 942.2 T = 854.4
m = 13.61 m = 13.61
P = 1056
P = 1885 T = 175.2
m = 52.93

Line ∆ P
T = 212.5

Line ∆ P
m = 15.12
P = 1670
T = 942.2 P = 995.5
m = 1.51 T = 854.4
m = 15.12

Line ∆ P P = 1670 P = 951


T = 947.2 T = 175.2
m = 15.12 m = 52.93
LEGEND
P = 1838 PUMP WITH INDUCER
Pc = 850 psia T = 210.0
Pc== 15.63
At 850 [psia]
in22 m = 15.12
TURBINE

At = 15.63 100.6 in VARIABLE POSITION VALVE


mdot = 68.1[inlbm/sec
]
m = 68.1 [lbm/s]
Ispd = 380.6 sec LCH4
Isp = 369.2 [sec]
F = v25900 lbf LOX
F = 25,000 [lbf] CH4 gas
Expansion = 180:1
ε = 180:1
PRESSURE P = [PSIA]
TEMPERATURE T = [DEG R]
MASS FLOW m = [(LBM/S]
59.9 in

Figure 36. Expander Cycle Engine Design for 25000 lbf of Thrust

VIII. Conclusion
This study takes advantage of all the pertinent data available in the open literature to derive a fully integrated engine
model including the combustion thermal chemistry, the turbopump assembly, the regenerative cooling jacket and the
injector. The turbopump analysis shows that there is a substantial penalty for choosing a single shaft direct drive
turbopump assembly as compared to a geared arrangement. However, cost and reliability shift the design choice to
the single shaft design. The variations in pump and turbine efficiency are shown with respect to chamber pressure
and a preferred region exists at the chosen design point. The injector analysis provides an optimized design, which it
is believed will provide stable combustion, and high combustion efficiency.

High aspect ratio rectangular cooling channels were studied in detail. The results of the parametric study on cooling
channel geometry show that as the channel geometry changes, the coolant heat transfer coefficient dominates the
heat transfer rate as compared to the area terms. In general a small flow area tends to increase heat transfer; however
the pressure loss across the coolant channel restricts the extent that the area can be reduced. Higher aspect ratio
cooling channels are shown to be advantageous in balancing the pressure loss requirements with the heat transfer
demands. The maximum allowable chamber pressure is limited by the survivable gas wall temperature. A stepped
channel configuration is shown to be superior to a channel which is invariant along the entire chamber and cooled
nozzle, with respect to pressure loss and over all engine mass, while maintaining the thermal performance of the
regenerative cooling jacket. Further optimization of the channel is possible based on the trends and tradeoffs
demonstrated in the project. Additionally, the model developed could facilitate studies on different propellant
combination, geared turbopump assemblies, injector designs, and expansion ratios.

Acknowledgments
This study was conducted as part of AAE 590C – Propulsion System Design Course in the Department of
Aeronautical and Astronautical Engineering at Purdue University. The authors wish to thank Professor Stephen D.
Heister for his wisdom, guidance and support.

20
References
1
Crocker, A.M., Peery, S.D., “System Sensitivity Studies of a LOX/Methane Expander Cycle Rocket Engine,” AIAA-1998-
3674, 34th Joint Propulsion Conference & Exhibit, Cleveland, OH, 1998.
2
Knab, O., Frohlich A., Wennerberg D., Haslinger W., “Advanced Cooling Circuit Layout for the Vinci Expander Cycle
Thrust Chamber,” AIAA 2002-4005, 38th Joint Propulsion Conference & Exhibit, Indianapolis, IN, 2002. .
3
Zubrin, R.M., “Mars and Lunar Direct: Maximizing the Leverage of In-Situ Propellant Production to Develop a Coherent
Architecture for the Space Exploration Initiative,” AIAA 92-1669, AIAA Space Programs and Technologies Conference,
Huntsville, AL 1992.
4
Burkhardt, H., Sippel M., Herbertz A., Klevanski J., “Kerosene vs. Methane: A Propellant Tradeoff for Reusable Liquid
Booster Stages,” Journal of Spacecraft and Rockets, Vol. 41, No. 5, 2005. DLR, German Aerospace Center, Space Propulsion
Institute, Germany.
5
Pempie, P., Frohlich, T., Vernin, H., “LOX/Methane and LOX/Kerosene High Thrust Engine Trade-Off,” AIAA 2001-3542,
37th Joint Propulsion Conference & Exhibit, Salt Lake City, UT, 2001.
6
Brown, C.D., “Conceptual Investigations for a Methane-Fueled Expander Rocket Engine,” AIAA 2004-4210 40th Joint
Propulsion Conference & Exhibit, Fort Lauderdale, FL, 2004..
7
Huzel, D.K., Huang, D.H., “Design of Liquid Propellant Engines,” NASA SP-125, 1967.
8
Sutton G.P., Biblarz O., Rocket Propulsion Elements, 7th ed., John Wiley & Sons, Inc. New York, 2000.
9
Gordon, S., McBride, B.J., “Computer Program for Calculation of Complex Chemical Equilibrium Compositions and
Applications,” NASA Reference Publication 1311, 1994.
10
Heister, S.D., “AAE 539: Advanced Rocket Propulsion Notes,” Purdue University, Spring Semester 2005.
11
Sieder, E.N.,Tate G.E., “Heat transfer and Pressure Drop of Liquids in Tubes,” Industrial and Engineering Chemistry, V. 28,
n 12, Dec. 1936, pp. 1429-1435.
12
Incopera, F.P., DeWitt, D.P., “Fundamentals of Heat and Mass Transfer,” 5th Ed., John Wiley & Sons, Inc., New York, 2002.
13
Schoenman, I., “Low-Thrust Isp Sensitivity Study,” NASA CR-165621, 1982.
14
Design Study of RL10 Derivative Rinal Report, Vol II, Engine Design Characteristics Appendices, FR-6011,
circa 1973.
15
Humble, R.W., Henry G.N., Larson, W.J., Space Propulsion Analysis and Design, McGraw-Hill Companies, Inc., New
York, 1995.
16
Gill, G.S., Nurick, W.H., “Liquid Rocket Engine Injectors,” NASA SP-8089, 1976.
17
McHale, R.M., Nurick, W.H., “Noncircular Orifice Holes and Advanced Fabrication Techniques for Liquid Rocket Injectors
(Phases I, II, III, and IV)”, NASA CR-134315, 1974.
18
Miller, D.S., Internal Flow Systems, Vol. 5”, BHRA Fluid Engineering, 1978.
19
Tamura, H., Ono, F., Kuma kawa A., Yatsuyanagi, N., “LOX/Methane Staged Combustion Rocket Combustor
Investigation,” AIAA-87-1856, 23rd Joint Propulsion Conference & Exhibit, San Diego, CA, 1987
20
Seely, Fred B., Smith, James O.: Advanced Mechanics of Materials, 2nd ed., John Wiley & Sons Inc., 1967.
21
Salmi, R.J., Wanhainen, J.P., Hannum, N. P., “Effect of Thrust per Element on Combustion Stability Characteristics of
Hydrogen-Oxygen Rocket Engines,” NASA TN D-4851, 1968.
22
College of Engineering, University of Michigan, “Cylindrical Pressure Vessels,” ME 211 Supplemental Course Material,
URL: http://www.engin.umich.edu/students/support/ mepo/ELRC/me211/flash2/cylindricalRev1.swf [cited Dec 2005].
23
College of Engineering, University of Michigan, “Spherical Pressure Vessels,” ME 211 Supplemental Course Material URL:
http://www.engin.umich.edu/students/support/mepo/ELRC/me211/flash2/spherical_02.swf [cited Dec 2005].
24
Rao, G.V.R, "Approximation of Optimum Thrust Nozzle Contour," Jet Propulsion, Vol. 30, No. 6, June 1960, pp. 561.
25
Rao, G.V.R., “Exhaust Nozzle Contour for Optimum Thrust,” Jet Propulsion, Vol 28, No. 6, June 1958 pp. 377-382.
26
NASA/TM-1998-206313, “Comparison of High Aspect Ratio Cooling Channel Designs for a Rocket Combustion Chamber
with Development of an Optimized Design,” M.F. Wadel. Lewis Research Center.
27
Hyde, J.C., Gill, G.S., “Liquid Rocket Engine Nozzles,” NASA SP-8120, 1976.
28
Esposito, J.J., Zabora, R.F., “Thrust Chamber Life Prediction, Vol I. Mechanical and Physical Properties of High
Performance Rocket Nozzle Materials,” NASA CR-134806, 1975.
29
Cook, R.T., Coffey, G.A., “Space Shuttle Orbiter Engine Main combustion Chamber Cooling and Life,” AIAA-73-1310, 9th
Propulsion Conference, Lake Buena Vista, FL, 1973.
30
Obermaier, G., Taubenberger, G., Fehyl, D., “Coaxial Injector Development for Storable Propellant Upper Stage
Turbopump Engines,” AIAA-1997-3095, 33rd Joint Propulsion Conference & Exhibit, Los Angeles, CA, 1997.

21

Das könnte Ihnen auch gefallen