48th AIAA Aerospace Sciences Meeting Including the New Horizons Forum and Aerospace Exposition AIAA 2010-1297

4 - 7 January 2010, Orlando, Florida

Simulating Self-Pressurization in Propellant Tanks

A P Winter1 and J G Marchetta2

University of Memphis, Memphis, TN 38152

Recent studies focusing on predicting the self pressurization play a significant role in the
design of cryogenic storage systems since tank pressure must be controlled for long duration
space missions. Incident solar radiation heats the fluid in the tank over a period of time,
vaporizing the liquid and as a result increasing the tank pressure. The objective of current
research is to develop a complete finite volume based Computational Fluid Dynamic (CFD)
model of tank pressurization in reduced gravity using an Energy of Fluid (EOF) approach,
and to use the simulation to aid in the design and prediction of propellant management
technologies aimed at controlling tank pressurization. The FLUENT pressure-based
computation model is significantly enhanced to include the EOF method, which will solve the
energy equation in terms of internal energy. The enhanced model will numerically predict
thermodynamic properties in each computational cell. The simulation results will provide
temperature and pressure histories for a given tank geometry and fill. The current study
technical objectives are to complete the model development and validate the model using
existing computational data for saturation temperature and evaporated liquid fraction.

Nomenclature
Aface = area of the computational cell face
Cv = specific heat at constant volume
ds = distance between neighboring cell center points
e = specific internal energy
∆e = latent heat, eg - el
E = internal energy
f = vapor phase fraction
g = vapor phase indicator
j = iteration level
k = thermal conductivity
l = liquid phase indicator
n = time level
nb = neighbor cell center point
P = pressure
ρ = density
sat = saturation point
t = time
T = temperature
p = cell center point
Q = heat
R = fluid gas constant
V = volume
W = work

I. Background

C ryogenic propellant tanks in space are exposed to incident solar radiation, which heats the liquid in tank over
time. As the temperature increases, the liquid vaporizes causing an increase in system pressure, i.e. self

1
Graduate Research Associate, Mechanical Engineering, AIAA Student Member.
2
Assistant Professor, Mechanical Engineering, AIAA Lifetime Member.

1
American Institute of Aeronautics and Astronautics

Copyright © 2010 by the American Institute of Aeronautics and Astronautics, Inc. All rights reserved.
pressurization. Cryogen vaporization causes mass loss of the valuable propellant. To decrease propellant loss during
long duration space missions, new, reliable management of cryogenic propellants in low gravity environments is
continuously being investigated. The self pressurization in cryogenic propellant tanks has been the focal point of
several past experimental and numerical studies as the need for on-orbit storage and transfer of propellant from one
tank to another tank has been identified. Mixing and active cooling are techniques which have been investigated as
means to control tank self-pressurization.
Multiple methods have been proposed for controlling the pressure in the tank. One approach is to insulate the
tank. The insulation would reduce the rate of self-pressurization in the tank, but over time, the pressure in the tank
will still rise. Another approach is to make stronger tanks, but such tanks have an increase in mass and could be cost
exorbitant. Tank venting is a different approach. However, the costly propellant may be lost to space during the
venting process. Due to the prospect of utilizing tank venting, an adaptation of the idea, has been proposed called a
Thermodynamic Vent System, or TVS1,2,3. This system would take a small amount of the bulk fluid in the tank and
pass it through a Joule-Thompson valve. After leaving the Joule-Thompson valve, the cold fluid would then pass
through a heat exchanger to continue its evaporation and absorb heat from a separate stream of fluid from the tank.
After the evaporation process for the first stream of liquid is complete, it would be vented overboard. This process
would result in a pressure and temperature decrease of the fluid. The TVS solution represented progress toward a
more comprehensive on-orbit experiment program that was eventually terminated by NASA for budgetary reasons.
In order for the venting approach to be practical, the liquid must also be positively positioned4 so that the liquid
propellant is opposite the vent minimizing venting the liquid portion of the tank contents to space.
A second approach to controlling tank pressurization is active mixing. An effective way to achieve active
mixing in the tank is to establish a jet of liquid along the tank’s centerline5,6. Some of the benefits of the jet-induced
active mixing include a reduction in thermal stratification, a catalyst in the self-pressurization process, faster on-
orbit transfer times, and the possible use of small heat exchangers, such as the TVS. The jet injects cooler propellant
and the subsequent mixing of the cooler propellant in the tank reduces the temperature of the bulk liquid. If the
magnitude of momentum of the jet is large enough, a geyser at the liquid-vapor interface can form. This jet increases
the area of the free surface helping to promote condensation which would further reduce the pressure. If the jet’s
momentum is high enough, the geyser will strike the opposing end of the tank. When the jet strikes the opposite end
of the tank, it can either form a separate pool or flow down the tank walls re-mixing with the bulk fluid. There will
be a cooling effect on the wall as the liquid flows down it. A negative aspect to active mixing is an addition of
excessive kinetic energy to the bulk fluid. The kinetic energy will eventually result in undesirable heat generation
through viscous dissipation. A significant stage in designing an appropriate jet is the optimization of the jet size and
intensity with respect to tank size and propellant volume. It is necessary to first classify and make predictions to the
behavior of the flow in order to optimize the system.
A small number of notable experiments have been conducted to investigate self-pressurization in cryogenic
propellant tanks. Aydelott1,2 performed the first small-scale drop-tower experiments using axial jets injected into
four different tanks partially filled with Ethanol. Poth et al.3 experimental investigations suggested a mixing jet
could be used to minimize thermal stratification and reduce tank self-pressurization. Boeing5,6,7,8 conducted Tank
Pressure Control Experiments (TPCE) which obtained pressure histories in a tank experiencing jet-induced
condensation, and recorded photographs of the liquid’s position. The first flight experiments demonstrated jet-
induced pressure collapse in a low gravity tank, but they did not record the velocity and temperature fields in the
tank. The second flight experiments recorded both temperature and pressure profiles. Chato9 conducted twenty-two
normal gravity experiments of hydrogen resupply utilizing an initial receiver tank chill down. The results of the
non-vented hydrogen fill of an insulated aluminum tank with varying degrees of chill down and inlet configurations
were reported. The results verified that successful no-vent fills were achievable for initial tank wall chill down.
Lin et al.10,11,12 reported experimental results for the self-pressurization and thermal stratification of a liquid
hydrogen tank subjected to low heat flux in 1g conditions. A further investigation was performed to account for jet
mixing to control tank pressure13.
Numerical models of tank self-pressurization have evolved over past years. Hochstein et al.14 enhanced the
SOLA-ECLIPSE Code with a complete lump node analysis to predict self-pressurization rates of liquid hydrogen
scale model tanks and compared the predictions to experimental data obtained from Aydelott15,16. The experiment
showed that the pressure rise was lower for reduced gravity than normal gravity conditions and that the location of
the heat source was important. The computational prediction was in good agreement with the experimental data or
the pressurization was over predicted. The computational model also suggested that sub-cooling significantly
affecting the pressurization rate. The research was further developed to include a multi-dimensional model of
pressurization and expulsion of slush hydrogen in tanks17,18. This new model was used to study the influence of
ullage boundary heat flux rates on the pressurization process. An investigation of both liquid hydrogen and

2
American Institute of Aeronautics and Astronautics
hydrogen slush was performed to depict the differences between the two. It was determined that the ullage
boundary heat flux rates drastically affected the pressure process.
A model was developed by Fite19 employing a finite difference approximation with a lumped parameter
approach which successfully predicted the transient response of two independent thermodynamic properties using
liquid hydrogen as the transfer in a non-vent fill of a storage tank. The results obtained conformed to the
experimental data obtained by Chato9, but exhibited unapparent oscillations in the pressure history.
A lumped thermodynamically based model of the ullage coupled to the transport equations in the liquid has been
the topic of a number of publications. Panzarella and Kassemi20 used the model to compare the results from
microgravity, zero-gravity and ground based studies for several different liquid-vapor configurations and fill levels.
The results showed that buoyancy and natural convection cannot be ignored when predicting the self-pressurization
of large cryogenic tanks. There were discrepancies reported between the numerical and thermodynamic model for
pressurization for each case. Barsi et al.21 investigated the effects of the heat power distribution and fill fraction of a
model fluid experiment in normal gravity had on pressurization. Discrepancies between experimental results and
the proposed thermodynamic model were documented and future work was proposed. In 2006, the work was further
continued by developing a model to examine test points in the ZBOT test matrix22. An axial liquid jet to mix the
liquid, and two strategies to remove energy from the system: a sub-cooled liquid jet and cold-finger cooling were
incorporated and the results were documented.
The abovementioned research performed by Barsi et al. was enhanced by employing an active vapor approach23.
With the knowledge obtained from previous research it was considered necessary to numerically compute
conservation equations in both the liquid and vapor region. By using an integral mass balance, the inter-phase mass
exchange was taken into account. The results obtained were compared to previous lump vapor models. Uniform,
vapor and liquid heating configurations were investigated. A conclusion was drawn that the active vapor approach
predicted the pressurization process with more reliability. Future work was proposed to compare the numerical
results obtained by the new model to experimental results.
Hochstein4,24 enhanced a computational model that simulated jet-induced mixing in a reduced gravity. He
compared the simulation results to the available experimental data presented by Aydelott1,2. The challenge has
always been to determine whether the simulation yields acceptable predictions as compared to the experiment data.
Aydelott performed each case in his experiment test matrix only once and he did not quantify uncertainty. The
simulation used a Jones and Launder25 k-ε model with the Pope round jet correction26 to model two dimensional
axisymmetric turbulence jet flows. The results of these simulations showed reasonable agreement with the bulk
mixing times, but did not show good agreement for Aydelott’s dimensionless geyser heights. Thornton et al27,28,29
improved the Aydelott’s original correlation by re-evaluating the dimensionless parameters describing the geyser’s
height. More recently, a fully three-dimensional simulation of cases within the Aydelott experiment test matrix was
undertaken using FLUENT by Marchetta et al30. They were able to predict all four of the flow regimes identified by
Aydelott.
Simulation of the tank self-pressurization process in low gravity requires accurate models of two-phase fluid
flow, heat transfer, capillary effects, and phase change. One major difference between the current study and previous
studies is that the Volume of Fluid (VOF) approach is not required since the phases, and subsequently the interface,
are defined by the internal energy in the computational cells. In previous modeling efforts, a lumped
thermodynamically based model of the ullage is coupled to the transport equations in the liquid and a cell by cell
mass balance along the interface is employed to account for the bulk evaporation process31. Lumps are treated as
homogeneous regions and Biot Numbers are assumed to be less than 0.1 in the low gravity environment.
Complicated geometries with differential heating sources are difficult to simulate using lump nodes. Field variables,
such as velocity, cannot be assessed using lump node models and convection coefficients have to be assumed.
These methods do not sufficiently account for the latent heat associated with the heat transfer with phase change
problems. This has proven to be a disadvantage when modeling the liquid-vapor interface.

II. Introduction to Computational Model

The Energy of Fluid (EOF) method developed by Anghaie et al.32,33 can be directly applied to all heat transfer
phase change problems including those that the value of latent heat is significantly larger than the sensible internal
energy in the computational cell. The EOF method is similar in concept to the well-established enthalpy
formulation used in finite volume simulations of bulk evaporation and condensation processes. When compared by
Anghaie, each method produced the same results and the computation times were similar; however, a major
disadvantage to the enthalpy formulation is a transient pressure term which adds more complexity to the problem.
The EOF method uses the complete form of Navier-Stokes equations which includes all relevant phase change

3
American Institute of Aeronautics and Astronautics
parameters such as the latent heat, surface tension, shear stress and gravity. The total internal energy is a
discontinuous function of temperature for a pure substance undergoing evaporation or condensation. This
discontinuity presents a challenge in maintaining solution stability when utilizing the transport equations in both the
liquid and vapor regions. Numerical difficulties normally associated with the steep gradients across the liquid-vapor
interfacial boundary are resolved with the EOF iterative scheme. This is achieved by that the discontinuity with
phase change is smeared over a small temperature range at the interface which consequently causes a sensitivity of
the vapor phase fraction to the change in temperature.
For complex geometries, the commercial software, FLUENT34, is used to model the fluid flow and heat transfer
problem presented herein. Along with its many features, FLUENT can model transient, three dimensional,
incompressible flows and is enhanced to for modeling EOF. The pressure based solver in FLUENT will be used to
solve for mass and momentum. The mass and momentum equations in FLUENT, will be utilized for future cases
where convection is considered. A user-defined function is established to solve the energy equation in terms of
internal energy. To solve the discretized energy equation, the energy equation is written in a linear form designed
for an IMSL C Numerical Math Library to be employed. The temperature solution is obtained by calling a routine
within the library which can handle sparse matrices in a linear form. For the validation case where convection is
neglected, e.g. bulk evaporation and condensation, only the weak form of the energy equation and property
relationships are utilized to solve for internal energy. FLUENT uses a control volume based technique and upwind
differencing to alter the governing equations to algebraic equations that can be solved numerically. Second order
upwind differencing is utilized for all FLUENT simulation predictions presented. FLUENT uses a co-located
scheme; whereby pressure, velocity, and temperature are all stored at cell centers. The Pressure-Implicit with
Splitting of Operators (PISO) pressure-velocity coupling scheme is used to obtain a semi-implicit pressure
correction equation. The pressure-correction equation is subsequently solved using the algebraic multigrid (AMG)
method. Temporal discretization is accomplished using implicit time integration, which is unconditionally stable
with respect to time step size. Solutions are subsequently iterated at each time level until the convergence criteria
are met.
The objective of current research is to complete the development of the finite volume based Computational Fluid
Dynamic (CFD) model of tank pressurization using an EOF approach. The enhanced model will numerically predict
thermodynamic properties in each computational cell. Transient properties are required to determine whether the
point of interest is in the liquid phase, saturated phase, or vapor phase. Once developed, code verification and
validation will be performed to assess the fidelity of the enhanced model. Test cases used by Anghaie et al are
utilized herein to verify and validate the current model. The simulation results will provide temperature, evaporated
liquid fraction and pressure histories for a given tank fill. Evaporation and condensation is determined by
comparing the internal energy in each computational cell to the internal energy at saturation.

III. Mathematical Formulation and Computational Discretization

The first law of thermodynamics for a closed system is expressed as

δ Q = dE + δ W (1)

where the internal energy term can be expressed in terms of enthalpy by

dE = dH − ( PdV + Vdp ) (2)

For the phase change process under investigation, the system is a rigid and enclosed container with constant volume
which results in no mechanical work. Therefore the first law reduces to

δ Q = dE = dH − VdP (3)

Equation 3 illustrates that the heat exchange is equal to the change in internal energy but not equal to the change
in enthalpy. As previously stated, using the enthalpy formulation of the energy equation introduces a transient
pressure term which makes the problem more complex; consequently it is desired to use a weak form of the energy
equation based on internal energy. Neglecting viscous heating, the energy equation for pure conduction in
microgravity which does not include the transient pressure term can be expressed as

4
American Institute of Aeronautics and Astronautics
 ∂e
ρ = ∇ ⋅ ( k ∇T ) (4)
∂t
where the released or absorbed latent heat associated with the phase change is equal to the change of internal energy.
Anghaie expresses the specific internal energy for the liquid, vapor, and mixed phases in the form of a summation of
the sensible heat and latent heat

(
e j = Cv ,l T j + f j −1
(T
sat
j
−T j )) + f j −1
∆e + Cv , g f j −1
(T j
− Tsatj ) (5)

where f is equal to 0 for the liquid phase at temperature T, 1 for the vapor phase at temperature T and a fractional
value for the mixed phase at temperature Tsat as determined from Equation 6.

e − el
f = (6)
eg − el

Substituting Equation 5 into the weak work of the energy equation gives

∂ ∂f
ρi
∂t
( Cv ,iT ) = ∇ ⋅ ( ki ∇T ) − ρi ∆e
∂t
(7)

where the subscript i = g when f = 1, and i = l when 0 ≤ f < 1. The discretized energy equation presented by
Pantankar35 for each computational cell yields

a pTpj = ∑ anbTnbj + b j + S (8)

where the subscript p represents the cell point of the control volume to be solved and nb represents a neighboring
cell point. The coefficients are computed as follows:

a p = ∑ anb + a 0p
(9)

knb Aface
anb =
ds (10)

ρ CvVcell
a 0p = (11)
dt

b = a 0pTp0 (12)

0
where Tp is the temperature value obtained from the previous time step and knb is the harmonic mean between the
thermal conductivity of the cell point and that of the neighboring cell point. The above relations also pertain to the
case when the cell point is adjacent to a neighboring boundary of known heat flux or temperature.
To incorporate the vapor phase fraction into the discretized energy equation, Anghaie defines the linear source
term as

5
American Institute of Aeronautics and Astronautics
 V
S = ρ∆e
dt
( f Pn -1 − f Pj -1 ) (13)

where the value of the latent heat is defined as a function of saturation temperature.

∆e = 275778.2 − 465.3Tsat − 0.20391Tsat 2 (14)

To implement the EOF method, the system is initialized with values for the pressure, temperature, saturation
temperature and vapor phase fraction. Since the source term is a function of the previous iteration and time values
of the vapor phase fraction, a value for temperature for each cell can be obtained. Once a temperature is calculated
the internal energy is updated. The vapor phase fraction is updated with the updated internal energy. Once the
iterative update converges on a solution for the vapor phase fraction by a specified residual value, the iterative loop
can end. The temperature solution is used to calculate the system pressure and saturation temperature for the next
time step. Anghaie approximates the saturation pressure by the equation of state for an ideal gas in which he uses
the bulk vapor density and temperature. For a cell whose vapor phase fraction is greater than zero the volume
weighted average for vapor pressure

no.of cells

∑ (ρ g RTgVcell f p )
i
Psat = Pg = i =1
no.of cells
(15)
∑ (V
i =1
cell fp )
i

Once the system pressure is calculated, the updated saturation temperature is obtained using the Clausius-Clapeyron
Equation relating the saturation temperature as a function of saturation pressure

D
Tsat = (16)
C − ln( Psat )

where the constants C and D for Refrigerant R-12 are 14.5936 and 2416.89, respectively.

IV. Model Validation Cases

The first test case currently being used to validate the model is pure bulk evaporation of refrigerant R-122. A
cylindrical container with a height of 0.04 m and radius of 0.01 m is filled 25% with saturated liquid and the rest
with saturated vapor both at 260 K. The top wall is heated and maintained at 500 K while the bottom wall is
maintained at 260 K. The boundary conditions of the side walls are insulated. Due to the significant difference
between the liquid density and vapor density, the density of the fluid is calculated based on the system pressure and
temperature. Additional material properties are held constant for the liquid and gas state individually. All material
properties at the interfacial boundary vary with the vapor phase fraction, i.e. quality.
Regarding the large difference between the liquid density and vapor density, a small change in the vapor phase
fraction can cause a significant change in the system pressure. Consequently any amount of increase in evaporation
raises the pressure very rapidly. To dampen the rate of change of the saturation pressure and ultimately the
saturation temperature, Anghaie utilized an under relaxation factor to modify the values by a weighted average of
the results from the previous and current time step. The optimal range of values for the relaxation factor reported
is 0.15-0.3 for the saturation temperature and 0.01-0.03 for the vapor pressure established by several numerical
experiments36. As the relaxation factor is increased the rate of change of the solution is increased. Thus a relaxation
factor of 0.15 will slow down the change of a solution between time steps more than a relaxation factor of 0.3. The
residual to regulate convergence of the vapor phase fraction is 1e-4. The saturation temperature and evaporated
liquid fraction is compared to the time history data reported by Anghaie2. At the maximum simulation time of
3000 s, Anghaie reports a saturation temperature of 271 K and an evaporated liquid fraction of 0.0135.
A 6976 cell mesh of the above container dimensions was used to test a range of relaxation values.
Figure 1 represents the saturation temperature as a function of flow time for four sets of relaxation values. While

6
American Institute of Aeronautics and Astronautics
Anghaie completed a simulation of 3000 s, the current simulation was terminated at an earlier time. For each set of
relaxation factors, the relative error was calculated between data points extrapolated from the time history curves
provided by Anghaie and the simulation results for both the saturation temperature and evaporated liquid fraction.
Significant relative errors were produced in the early time steps for the saturation temperature. Table 1 details the
relative error values. Although the error for a relaxation factor of 0.25 for the saturation temperature and 0.02 for
vapor pressure is the smallest for the results compared for saturation temperature, it has the greatest error for the
results compared for evaporated liquid fraction. Evaluating Anghaie’s data for both saturation temperature and
evaporated liquid fraction, a relaxation factor of 0.2 for the saturation temperature and 0.03 for vapor pressure is a
better fit for the present investigation and is used for the mesh convergence study. Taking into consideration the data
reported in the current simulations, it is pertinent to analyze the time histories for both the saturation temperature
and evaporated liquid fraction. As illustrated in Figure 2, the rate of vaporization for the favorable set is more
gradual and within the bounds reported by Anghaie than that of a relaxation factor of 0.25 for the saturation
temperature and 0.02 for vapor pressure. Given the time history of the evaporated liquid fraction, had the simulation
completed a flow time of 3000 s, the evaporated liquid fraction for a relaxation factor of 0.2 for the saturation
temperature and 0.03 for vapor pressure could in time report minimal error relative to the data reported by Anghaie.
While Anghaie did not document the time history for vapor pressure, Figure 3 illustrates the pressure history for
each set of relaxation factors for the current simulations.
Anghaie examined three different uniform mesh sizes for a cross section of the cylindrical tank with a time step
∆t = 1 s. The 61 x 11, 51 x 11 and 41 x 11 meshes produced identical results at ∆t = 1 s. To contrast Anghaie’s
computations, three mesh sizes were examined with a time step ∆t = 1 s. Based on the preceding simulations
discussed, the relaxation factor used for a mesh convergence study was 0.2 for saturation temperature and 0.03 for
the vapor pressure. Each meshed volume was three-dimensional with the above specified container dimensions. As
the mesh concentration increased, the rate of vaporization decreased. This was expected recognizing the change in
saturation temperature with respect to flow time. Figure 4 illustrates the time histories of the saturation temperature
for a 6976, 5800, and 3825 cell mesh. Despite the fact that mesh convergence was not established, the more
concentrated 6976 cell mesh improved the refinement such to better depict the latent heat of vaporization at the
interfacial boundary. As shown, the saturation temperature for each mesh investigated appears identical up to
approximately 150 s of flow time. As the flow time increases past 150 s, the saturation temperature begins to
increase earlier for the 6976 cell mesh. The liquid in the cells which has not evaporated, do so at a slower rate since
the saturation temperature has began to rise more rapidly.

V. Conclusion
A goal for the current research objective was to validate a complete finite volume based Computational Fluid
Dynamic (CFD) model of tank pressurization in reduced gravity using an Energy of Fluid (EOF) approach. Based
on the simulation results more numerical experiments should be conducted to optimize the relaxation factor used to
dampen the rate of change for the saturation temperature and vapor pressure. Since the saturation temperature is
directly calculated based on the vapor pressure, any corrections made to the solution in order to select an optimal
relaxation factor for the problem will affect both the saturation temperature and vapor pressure. A mesh
convergence study was conducted; however, mesh convergence was not demonstrated. A time convergence study
was not conducted. Additional computation experiments are required to converge on a most favorable mesh size
and time step. Future technical objectives are to finalize the model development, continue to verify and validate the
model using existing computational and ground experiment data, and utilize the enhanced simulation to model
active cooling using jet-induced geysers. The simulation will be used to calculate the effectiveness of using axial
jets for cooling and the predictions will be compared to available experimental results.

Table 1. Relative error between saturation temperature and evaporated liquid fraction
values reported by Anghaie and current simulation values for each set of under relaxation
factors.

Tsat 0.25 Tsat 0.3 Tsat 0.2 Tsat 0.2

Pressure 0.02 Pressure 0.02 Pressure 0.03 Pressure 0.03
Saturation
2.56 3.89 3.34 3.45
Temperature
Evaporated
2.63E-3 9.82E-5 2.04E-5 2.20E-5
Liquid Fraction

7
American Institute of Aeronautics and Astronautics
 265.0

264.5 T sat 0.25 P 0.02

T sat 0.3 P 0.02
264.0
Saturation Temperature, K T sat 0.2 P 0.03
263.5 T sat 0.25 P 0.03
263.0 Anghaie

262.5

262.0

261.5

261.0

260.5

260.0
0 25 50 75 100 125 150 175 200 225 250
Flow Time , s

Figure 1. Comparison between different sets of relaxation factors for

the evolution of saturation temperature in a 0.04 m height 0.01 m
radius cylindrical container.

0.014 T sat 0.25 P 0.02

T sat 0.3 P 0.02
0.012
Evaporated Liquid Fraction

T sat 0.2 P 0.03

0.010 T sat 0.25 P 0.03
Anghaie
0.008

0.006

0.004

0.002

0.000
0 25 50 75 100 125 150 175 200 225 250
Flow Time , s

Figure 2. Comparison between different sets of relaxation factors for

the evolution of evaporated liquid fraction in a 0.04 m height 0.01 m
radius cylindrical container.

8
American Institute of Aeronautics and Astronautics
 235
T sat 0.25 P 0.02
230 T sat 0.3 P 0.02
T sat 0.2 P 0.03
225
T sat 0.25 P 0.03
Pressure, kPa

220

215

210

205

200
0 25 50 75 100 125 150 175 200 225 250
Flow Time , s

Figure 3. Comparison between different sets of relaxation factors for

the evolution of vapor pressure in a 0.04 m height 0.01 m radius
cylindrical container.

268

267 6976 Cell Mesh

5800 Cell Mesh
266
3825 Cell Mesh
Saturation Temperature, K

265
Anghaie

264

263

262

261

260

259
0 25 50 75 100 125 150 175 200 225 250 275 300 325 350
Flow Time , s

Figure 4. Comparison between different mesh sizes for the evolution

of saturation temperature in a 0.04 m height 0.01 m radius
cylindrical container.

9
American Institute of Aeronautics and Astronautics
 References
1
Aydelott, J.C.: Axial Jet Mixing of Ethanol in Cylindrical Containers During Weightlessness. NASA TP-1487 (1979).
2
Aydelott, J.C.: Modeling of Space Vehicle Propellant Mixing. NASA TP-2107 (1983)
3
Poth, L.J., and Van Hook, J.R., “Control of the Thermodynamic State of Space-Stored Cryogens by Jet Mixing,” Journal of
Spacecraft, Vol. 9, No. 5, pp. 332-336, May 1972.
4
Hochstein, J.I., Gerhart, P.M., and Aydelott, J.C., “Computational Modeling of Jet Induced Mixing of Cryogenic Propellants
in Low-G,” AIAA-1984-1344, AIAA/SAE/ASME/ASEE 20th Joint Propulsion Conference, Cincinnati, OH, June 1984.
5
Messerole, J.S., “Mixing Induced Ullage Condensation and Fluid Destratification,” AIAA-87-2018,
AIAA/SAE/ASME/ASEE 23rd Joint Propulsion Conference, San Diego, CA, June-July 1987.
6
Boeing Aerospace, “Preliminary Experiment Requirements Document,” NASA Lewis Research Center IN-STEP Flight
Development Program, Contract NAS3-25363, 1989.
7
Albayyari, J.M., Gerner, F.M., and Bentz, M.D.: Axial Jet-Induced Mixing in Low Gravity: Results of the Reduced Gravity
Tank Pressure Control Experiment. The 1st Pacific Symposium on Flow Visualization and Image Processing, Honolulu, 1997.
8
Hasan, M.M, Lin, C.S., Knoll, R.H., and Bentz, M.D., “Tank Pressure Control Experiment: Thermal Phenomena in
Microgravity,” NASA TP-3564, March 1996.
9
Chato, D.J. “Ground Testing on the Nonvented Fill Method of Orbital Propellant Transfer: Results of Initial Test Series,”
AIAA-91-2326, AIAA/SAE/ASME/ASEE 27th Joint Propulsion Conference, Sacramento, CA, June 1991.
10
Hasan, M.M, Lin, C.S. and Van Dresar, N.T., “Self-pressurization of a flightweight liquid hydrogen storage tank subjected
to a low heat flux.” NASA TM 103804, 1991.
11
Van Dresar, N.T., Lin, C.S. and Hasan, M.M., “Self-pressurization of a flightweight liquid hydrogen tank: Effects of fill
level at low wall heat flux.” NASA TM 105411, 1992.
12
Lin, C.S., Van Dresar, N.T. and Hasan, M.M., “Pressure control analysis of cryogenic storage systems.” J. Propulsion and
Power, Vol. 20, No. 3, pp. 480–485, 2004.
13
Lin, C. S., Hasan, M. M., and Van Dresar, N. T., “Experimental Investigation of Jet-Induced Mixing of a Large Liquid
Hydrogen Storage Tank,” AIAA Paper 94-2079, July 1994.
14Hochstein, J.I., Ji, H-C., Aydelott, J.C., “Prediction of Self-Pressurization Rate of Cryogenic Propellant Tankage,” Journal
of Propulsion and Power, Vol. 6, No. 1, pp. 11-17, 1990.
15
Adelott, J.C., “Effect of Pressurization on Self-Pressurization of Spherical Liquid-Hydrogen Tankage,” NASA TN D-4286,
1967.
16
Adelott, J.C., “Normal Gravity Self-Pressurization of 9 inch (23cm) Diameter Spherical Liquid Hydrogen Tankage,” NASA
TN D-4171, 1967.
17
Sasmal, G.P., Hochstein, J.I. and Hardy, T.L., “Influence of Heat Transfer Rates on Pressurization of Liquid/Slush
Hydrogen Propellant Tanks,” AIAA-93-0278, The 31st AIAA Aerospace Sciences Meeting, Reno, NV, January 1993.
18
Sasmal, G.P., Hochstein, J.I., Wendl, M.C. and Hardy, T.L., “Computational Modeling of the Pressurization Process in a
NASP Vehicle Propellant Tank Simulation,” AIAA-91-2407, AIAA/SAE/ASME/ASEE 27th Joint Propulsion Conference,
Sacramento, CA, June 1991.
19
Fite, L.W. “Characteristics if Nonvented Propellant Transfer,” Ph.D. Dissertation, Memphis State University, Memphis,
TN, 1993.
20
Panzarella, C.H. and Kassemi, M., “Self-Pressurization of Spherical Cryogenic Tanks in Space,” Journal of Spacecraft and
Rockets, Vol 42, No. 2, pp. 299-308, 2005.
21
Barsi, S., Kassemi, M., Panzarella, C.H., Alexander, J.I.D., “A Tank Self-Pressurization Experiment Using a Model Fluid
in Normal Gravity,” AIAA-2005-1143, The 43rd AIAA Aerospace Sciences Meeting, Reno, NV, January 2005.
22
Barsi, S. and Kassemi, M., “A Numerical Study if Tank Pressure Control in Reduced Gravity,” AIAA-2006-0936, The 44th
AIAA Aerospace Sciences Meeting, Reno, NV, January 2006.
23
Barsi, S., Panzarella, C.H., and Kassemi, M., “An Active Vapor Approach to Modeling Pressurization in Cryogenic Tanks,”
AIAA-2007-5553. AIAA/SAE/ASME/ASEE 43rd Joint Propulsion Conference, Cincinnati, OH 2007.
24
Hochstein, J.I., “Computational Modeling of Jet-Induced Geyser Mixing in Cryogenic Propellant Tanks in Low-G,”
Doctoral Dissertation, University of Akron, 1984.
25
Jones, W.P., and Launder, B.E., “The Prediction of Laminarization with a Two-Equation Model of Turbulence,”
International Journal of Heat and Mass Transfer, Vol. 15, pp. 301-314, 1972.
26
Pope S.B., “An Explanation of the Turbulent Round Jet/Plane Jet Anomaly,” AIAA Journal, Vol. 16, No. 3, pp. 279-281,
March 1978.
27
Thornton, R.J., “A Computational and Dimensional Analysis of Microgravity Propellant Tank Geyser Prediction,”
Magisterial Thesis, The University of Memphis, Memphis, TN, 2000.
28
Thornton, R.J., and Hochstein, J.I., “Microgravity Geyser and Flow Field Prediction,” AIAA-2000-0858, The 38th AIAA
Aerospace Sciences Meeting, Reno, NV, January, 2000.
29
Thornton, R.J., and Hochstein, J.I., “Microgravity Geyser Analysis and Prediction,” AIAA-2001-1132, The 39th AIAA
Aerospace Sciences Meeting, Reno, NV, January, 2001.
30
Marchetta, J.G., and Benedetti, R.H., “Three Dimensional Modeling of Jet-Induced Geysers in Low Gravity,” AIAA-2007-
0955, The 45th AIAA Aerospace Sciences Meeting, Reno, NV, January 2007.

10
American Institute of Aeronautics and Astronautics
 31
Panzarella, C.H. and Kassemi, M., “On the Validity of Purely Thermodynamic Description of Two-Phase Cryogenic
Storage Tank,” Journal of Fluid Mechanics, Vol 484, pp. 136-148, 2003.
32
Ding, Z. and Anghaie, S. “Numerical Modeling of Bulk Evaporation and Condensation with Constant Volume,”
International Journal for Numerical Methods in Engineering, Vol 39, pp. 219-233 1996.
33
Anghaie, S. and Ding, Z., “Modeling of Bulk Evaporation and Condensation,” NASA CR 198392, 1996.
34
FLUENT Inc.: FLUENT 6.3 User’s Guide, Lebanon, NH, 2008.
35
Patankar, S.V., Numerical Heat Transfer and Fluid Flow, Hemisphere Publication, New York: 1980.
36
Z. Ding and S. Anghaie, “Numerical investigation of the two-phase equilibrium state in a cylindrical nuclear fuel cell under
zero gravity condition,” Proc. 6th AIAA/ASME Joint Thermophysics and Heat Transfer Conf., Colorado Springs, CO,
AIAA 94-1995, 1994.

11
American Institute of Aeronautics and Astronautics