Pergamon Int. J. Mech. Sci. Vol. 38, Nos 8 9, pp.

951 965, 1996

Copyright (~, 1996ElsevierScienceLtd
Pnnted in Great Britain. All rightsreserved
0020 7403/96 $15.00+ 0.00

0020-7403(95)00089-5

RESPONSE OF LATERAL IMPULSE ON LIQUID HELIUM

SLOSHING WITH BAFFLE EFFECT IN MICROGRAVITY

R. J. H U N G and Y. T. LONG
Department of Mechanical and Aerospace Engineering, The University of Alabama in Huntsville,
Huntsville, AL 35899, U.S.A.

(Received 28 March 1995; and in revised Jbrm 5 September 1995)

Abstract Sloshing dynamics within a partially filled rotating dewar of superfluid helium II are
investigated in response to lateral impulse. The study, including how the rotating bubble of
superfluid helium II reacts to the impulse in microgravity, how amplitudes of bubble mass center
fluctuate with growth and decay of disturbances, how slosh reaction forces amplify and decay in
response to lateral impulse, and how sloshing dynamics differ with and without a baffle, are
investigated. The numerical computation of sloshing dynamics is based on the non-inertial frame
dewar bound coordinate. Results of the studies are illustrated. Copyright ~: 1996 Elsevier Science Ltd.

Keywords: microgravity, liquid helium, slosh dynamics, dewar with baffle, lateral impulse.

1. I N T R O D U C T I O N

In order to carry out scientific experiments, some experimental spacecraft use cryogenic
cooling for observation instrumentation and telescope, superconducting sensors for gyro
read-out and maintain very low temperatures near absolute zero for mechanical stability.
The approaches to both cooling and control involve the use of helium II. In this study, the
response of cryogenic systems to sloshing dynamics associated with impulsive oscillations is
investigated. Some spacecraft adopt the cooling and boil-off from the cryogenic liquid
helium dewar as a cryogen and propellant to maintain the cooling of instruments, attitude
control and drag-free operation of the spacecraft. Potential fluid management problems
may arise due to asymmetric distribution of liquid helium and vapor or to perturbations in
the free surface. Baffle-boards may be needed for the dewar to reduce the degree of
asymmetry and damps disturbances in the liquid-vapor distribution.
Cryogenic liquid helium II, at a temperature of 1.8°K, is used as the coolant
and propellant. With its superfluid behavior, there is a very small temperature gradient
in the liquid helium. In the negligibly small temperature gradient along the surface which
drives Marangoni convection, the equilibrium shape of the free surface is governed
by a balance of capillary, centrifugal and gravitational forces. Determination of
liquid-vapor interface profiles based on computational experiments can uncover details of
the flow which cannot be easily visualized or measured experimentally in a microgravity
environment.
The instability of the liquid-vapor interface can be induced by the presence of longitudi-
nal and lateral accelerations. Thus, slosh waves are excited, producing high and low
frequency oscillations in the liquid propellant. The sources of the residual accelerations
include effects of the Earth's gravity gradient and g-jitter. A recent study [1] suggests that
the high frequency accelerations may be unimportant in comparison with the residual
motions caused by low frequency accelerations.
The time-dependent behavior of partially-filled rotating fluids in reduced gravity envi-
ronments was simulated by numerically solving the Navier-Stokes equations subject to the
initial and boundary conditions [2-5]. At the interface between the liquid and the gaseous
fluids, both the kinematic surface boundary condition and the interface stress conditions for
components tangential and normal to the interface were applied [2-5]. The initial condi-
tions were adopted from the steady-state formulations developed by Hung et al. [4]. Some
of the steady-state formulations of interface shapes were compared with the available
experiments carried out by Leslie 1-6] in a free-falling aircraft (KC-135). The experiments
951
952 R.J. Hung and Y. T. Long

carried out by Mason et al. [7] showed that the classical fluid mechanics theory is
applicable to cryogenic liquid helium in large containers.
At temperatures close to absolute zero, quantum effects begin to be of importance in the
properties of fluids. At a temperature of 2.17 °K, liquid helium has a 2-point (a second-order
phase transition); at temperatures below this point, liquid helium (helium II) has a number
of remarkable properties, the most important of which is superfluidity. This is the property
of being able to flow without viscosity in narrow capillaries or gaps. At temperatures other
than zero, helium II behaves as if it were a mixture of two different liquids. One of these is
a superfluid and moves with zero viscosity along a solid surface. The other is a normal
viscous fluid. The two motions occur without any transfer of momentum from one to
another for velocities below a critical velocity [8-15]. For the components of normal and
superfluid velocities above a critical velocity, the two fluids are coupled through snarling in
a complex tangle [8-15].
The key parameter of critical velocity to distinguish one-fluid from two-fluid models is
a function of fluid temperature and container size. In other words, in considering the
dynamical behavior of helium II in a large rotating cylinder, a mixture of the superfluid and
the normal fluid without separation of the two fluids for fluid velocities greater than the
critical velocity is accounted for in the model computation. Density concentration of
superfluid is a function of temperature, which is also true for the surface tension and viscous
coefficient for helium II [8-15]. In this study, the theory of viscous Newtonian fluids is
employed with transport coefficients being a function of temperature.
In this paper, the response of superfluid helium II to lateral impulse is studied. In order to
reduce the degree of asymmetry in the liquid-vapor distribution and damp its associated
disturbances, a number of baffle boards [16] are installed inside the dewar. The sloshing
dynamics modulated liquid-vapor interface disturbances and sloshing reaction forces and
torques exerted on the dewar with and without the baffle [16] in response to lateral impulse
are also investigated.

2. NON-INERTIAL FRAME MATHEMATICAL FORMULATION

OF FUNDAMENTAL EQUATIONS
An experiment made by Andronikashvili [10-12] for rotating helium II shows that it is
necessary to exceed a critical velocity for the interaction between the normal and superfluid
components to establish entire bucket in rotation [8-15]. For the rotating dewar with outer
diameter of 1.56 m and inner diameter of 0.276 m, the critical velocities to assume the
interaction between the normal and superfluid components are 6.4x10 -7 and
3.6 × 10 -6 m/s, respectively [8-15]. With rotating speed of 0.1 rpm, the linear velocities
along the outer and inner walls of rotating dewar are 8.17 × 10 -3 and 1.45 x 10 -3 m/s,
respectively, which are at least several hundred times greater than that of the corresponding
critical velocities to assure the interaction between the normal and superfluid components
of helium II. Based on this illustration, the problems under consideration have the special
features to warrant an adoption of viscous Newtonian fluids formulation in this study.
These special features are as follows: (a) Fluid velocities are at least several hundred times
greater than that of the corresponding critical velocities; (b) These high fluid velocities can
produce great enough vortex lines to snarl in a complex tangle to assure an interaction
between the normal and superfluid components [8-15]; (c) These vortex lines snarling with
a complex tangle between the normal and superfluid components warrant the adoption of
Newtonian fluid model [8-15].
Consider a closed circular dewar partially filled with helium II while the ullage is filled
with helium vapor. The whole fluid system is spinning in the axial direction z of cylindrical
coordinates (r, 0, z), with corresponding velocity components (u, v, w). The governing equa-
tions for non-inertial frame spacecraft bound coordinates spinning along its z-axis has been
given in our recent studies [17-20]. In other words, dynamical forces, such as gravity
gradient [21-23], g-jitter [1-5] and angular accelerations, and centrifugal, Coriolis, surface
tension, viscous forces, etc., are given explicitly in the mathematical formulations [17-20].
In the computation of slosh reaction forces and moments, acting on the container wall of
 Response of lateral impulse on liquid helium sloshing 953

(A) (A) Three Dimetmional Profile of Baf'fle Boards

(B) Grid Points in Radial Axial Direction With

Baffle Boards

(C) Grid Points in R*di*l - Cireumferencial

Direction With Baffle Boards

z y(0 = ~r/2)
(B}~7::',::~m',;
~ ;',U,"ql~',~',~
[ li~,~iill
.............

. ~ - ~ J~ H-H44+H-H-HI
' )""I;"["')"I (c) ~ 1

o H;,)4~U~_t::, ,"m",",",", r
-70 -35 0 35 70 -70 35 0 35 70
RADIUS (era) RADIUS (era)

Fig. 1. (A) Three-dimensional configuration of a container with baffles, (B) Distribution of grid
points in the radial-axial plane, and (C) In the radial-circumferential plane.

the spacecraft, one must consider those forces and moments in the inertial frame rather than
in the non-inertial frame 1-17-20].
For the purpose of solving sloshing dynamic problems of fluid systems in orbital
spacecraft under a microgravity environment, one must solve the governing equations
I-17-20] accompanied by a set of initial and boundary conditions. A detailed illustration of
these initial and boundary conditions concerning the sloshing dynamics of fluid systems in
microgravity was precisely given in Hung and Pan 1,17, 18, 20]. The computational algo-
rithm applicable to cryogenic fluid management under microgravity is also given in Hung et
al. 1-24]. In this study, a dewar with an outer radius of 78 cm, and an inner radius of 13.8 cm,
top and bottom radius of 110 cm and a height of 162 cm has been used in the numerical
simulation. The dewar tank is 60% filled with cryogenic liquid helium and the ullage is filled
with helium vapor. The temperature of cryogenic helium is 1.8 °K. In this study the
following data were used: liquid helium density =0.146g/cm 3, helium vapor den-
sity = 0.00147 g/cm 3, fluid pressure = 1.66 × 1 0 4 dyn/cm 2, surface tension coefficient at the
interface between liquid helium and helium vapor = 0.353 dyn/cm, liquid helium viscosity
coeflicient=9.61xl0-5cm2/s; and contact a n g l e = 0 +. The initial profiles of the
liquid-vapor interface for the rotating dewar are determined from computations based on
algorithms developed for the steady-state formulation of microgravity fluid management
E25].
A staggered grid for the velocity components is used in this computer program. MAC
(marker-and-cell) method [26, 27] of studying fluid flows along a free surface is adopted.
VOF (volume of fluid) method [28] is used to solve finite difference equations numerically.
Approximate flow velocity is calculated from the explicit approximation of momentum
equations based on the results from the previous time step. Computation of pressure and
velocity at the new time step are, thus, obtained from iteratively solving pressure equation
through conjugate residual technique [29-31 ]. A detailed description of the computational
algorithm applicable to microgravity fluid management is illustrated in our recent works
[17-20]. As for the volume conservation of liquid, a deviation of less than 1% error
of volume is always guaranteed before a move to the next time step. Figures I(A-C)
show three-dimensional configuration of baffle installation, the distribution of grid points
for the dewar tank with baffle in the radial-axial plane and radial-circumferential plane,
954 R J. Hung and Y. T. Long

respectively, in cylindrical coordinates. The baffles are installed with three parallel plates
perpendicular to the rotating axis and four radial plates aligned with the rotating axis. The
inner radius, outer radius and thickness of the three parallel baffles chosen in this computa-
tion are 60, 77.48 and 0.1016 cm, respectively. These three baffles are installed at locations
z l , zz, z3 at 38.74, 80.94, and 84.4 cm, respectively. Four radial baffles are located at 90 ° each
apart with the same inner and outer diameters and thickness as that of the three parallel
baffles.

3. MICROGRAVITY SLOSHING DYNAMICS IN RESPONSE TO LATERAL IMPULSE

An example is given to illustrate sloshing dynamics in response to lateral impulse during
the time period of guidance and/or attitude controls. Dewar is spinning with a rotating rate
of 0.1 rpm during its normal operation. In this study, with the following form of lateral
impulse is assumed:
= [ax, ay, a~] = [1,0,0] 10 -2 go in cartesian coordinates
= [at, ao, a,] = [cos 0, - sin 0, 0] 10- 2 go in cylindrical coordinates for 0 < t ~ 10- z s

and
4=[0.0.0] for t > 1 0 - / s
where go( = 9.81 m/s z) is the Earth gravitational acceleration.
3.1. Bubble oscillations in a rotatin 9 dewar without baffles
The time evolution of the liquid-vapor interface without a baffle is first investigated.
Figure 2(A) shows the time sequence evolution of three-dimensional liquid-vapor interface
oscillations. The figure shows bubble oscillations at times t = 0.00, 155, 425, 725, 2320 and
4000 s. It shows that the bubble was deformed in response to the impulse and gradually
recovered back to the equilibrium shape at the end of the 4000 s simulation.
Figure 3(A) shows the time sequence of the liquid-vapor interface profiles in the vertical
r - z plane at 0 = 0 ° and 180 °. It indicates that the horizontal cross-section of the doughnut-
shape bubble was first pushed to the left (liquid to the right), then to the right (liquid to the
left), and vice versa.
Figure 4(A) shows the time evolution of the sloshing dynamics governed liquid-vapor
interface at z = 47.2 cm. It shows that a horizontal cross-section of the doughnut-shape
bubble was again pushed to the left (liquid to the right), then to the right (liquid to the left),
and vice versa. As impulse is pointing toward the eastern (positive-x) direction, the liquid is
pushed toward the eastern direction while the bubble is pushed toward the western
(negative-x) direction in the r-O plane. Back and forth oscillations are then induced due to
the surface tension effects. The oscillations are finally decaying out due to the viscosity and
surface tension effects. This scenario reflects the evolution of the decaying bubble oscilla-
tions in response to lateral impulse.
In order to illustrate the bubbles were experiencing finite deformation, Figs 5 - 7 show the
time evolution between time from 225 to 427 s for three-dimensional, vertical r - z plane flow
profile at 0 = 0 ° and 180 °, and horizontal r-O plane flow profile at z = 122 cm, respectively.
These three figures show the following results for a rotating dewar without baffles: (a)
A transition from left-side bubble swelling, equilibrium and right-side bubble swelling in
Fig. 5(A); (b) A flow transition from horizontal jet flow encircling around the central part of
bubble to the formation of circulating flow around the top-left, top-right, bottom-left and
bottom-right around the bubble in Fig. 6(A); and (c) A transition of flow patterns from
circulating flow with centers along 0 = 90 ° and 270 ° to circulating flow with centers along
0 = 0 ° and 180° in Fig. 7(A).
Figure 8 shows the time evolution of growth and decay of the fluid mass center
fluctuations with broken lines for a rotating dewar without baffles in response to the lateral
impulse. Figures 8(A-C) show the time evolution of the locations of the fluid mass center
fluctuations in xc, Yc, and fluctuating locations in the radial axis of the cylindrical coordi-
nates Rc [ = (x~z + yc2)1/2], respectively, for the container without baffle. It is due to the
 Response of lateral impulse on liquid helium sloshing 955

(A) INTERFACE OSClr .I.ATIONS WITHOUT BAFFLE

(a) t = 0.0 s (b) t = 155 s (e) ~ = 425 s

I:r.1
rO

(d)/- t = 72z5s (f) t = 4000 s

(B) 12WI'ERFACE OSCTI~I.ATIONS WITH BAFFL,E

i ~I; = • -- (e) t=425s

(d) t = 725 s (e) t = 2320 s (f) t = 4000 s

~ 2! ?-<

Fig. 2. Time sequence evolution of three-dimensional liquid vapor interface oscillations for a rotat-
ing dewar in response to lateral impulse for a container without and with baffles.

effect of Coriolis forces which carry over the lateral fluctuations from x- to y-axes even
though the impulse was activating on the x-direction only in the rotating dewar along the
z-axis. This means that the fluid mass center fluctuations are primarily in the x- and y-axes
and not in the z-axis. The locations of m a x i m u m absolute values of fluid mass center
fluctuations for the container without baffles in the radial direction are Max(Ixcl,
[Ycl) = (2.50, 1.96)cm; while the fluctuating values of the m a x i m u m fluid mass center
956 R . J . Hung and Y. T. Long

(A) INTERFACE OSCTLLATIONS W I T H O U T BAFFLE

(a) t= 0.0 s (b) t = 155 s (c) t = 425 s
i

..%u •

N
ul
~o ~ii i i~i

, , ,I

o (a) t=725 s (e) t= 2320 s (t') t= 4000 s

oo

"!
= , , , . . . . . . . . . . I " r -

~ (B) INTERFACEOSC'~.T.ATIONSWITHBAFFLE
~.) (a) t = 0.0 s (b) t = 155 s (e) t = 425 $

i i " i '

(d) t = 725 s (e) t = 2320 s (f) t = 4000 s

~ ..... ~.- ' :;"

i
: !

-TD -35 0 35 "¢D -70 -35 0 35 "TO -,';0 -35 ~ 35 70

RADIUS [crn) RADIUS (cm) RADIUS (cm)

Fig. 3. Time sequence evolution of liquid vapor interface oscillations for a rotating dewar in the
r-z plane at 0 = 0 ° a n d 180 °, in response to lateral impulse for a container without and with baffles.

disturbances in radial coordinates are Max(lAx~l, [ A y c l ) = ( 3 . 7 1 , 3 . 4 1 ) c m . It shows

[xol > iYo[ and Axe > Ay c for fluctuating fluid mass center disturbances in a rotating dewar
in response to the impulse along the x-axis. In Fig. 8, it shows that fluid mass center
fluctuations reach its peak value at time t = 82 s which is almost 82 s after the ending of
 Response of lateral impulse on liquid helium sloshing 957

(A) INTER~ACE OSCKJ.ATIONS W I T H O U T BAFFLE

(a) t = 0.0 s (b) t = 155 s (c) t = 425 s

\\
oi(O //
o

(d) t = 725 s (e) t = 2320 s (f) t = 4000 s

II

mO
~ ~ , , t , , i , • t ,"r ''~ r . . . . . . .

(B) INTERFACE OSCq'LTATIONS WITH BAFFLE

(d) t = 725 s (c) t = 2320 s (f) Z = 4000 s

L . - , , , . . . .

-~ - 3 5 0 35 70 -70 -35 0 35 70 -70 -35 0 0.5 "K)

RADIUS (.cm) RADIUS(crn) RADIUS(cm)

Fig• 4. Time sequence evolution of liquid-vapor interface oscillations for a dewar in the x - y (r O)
plane at height z = 47.2 cm, in response to lateral impulse for a container without and with baffles.

impulse activating on the fluid systems. It also indicates an exponentially decaying tendency
of the fluid mass center fluctuations for the c o n t a i n e r without baffle•
A c o m p a r i s o n of Figs 2 a n d 8 illustrates some peculiar behavior of cryogenic helium
fluids with t e m p e r a t u r e below ,i-point (2.17°K) in which helium d e m o n s t r a t e s a n u m b e r of
r e m a r k a b l e properties of superfluidity such as extremely low viscous a n d surface t e n s i o n
958 R.J. Hung and Y. T. Long

Three Dimensional Bubble Finite Deformation

(a) t=225 s (b) t=311 s (c) t=427 s

(a) t=225s (b) t=311 s (c) t=427 s

=C

Fig. 5. Time evolution of three-dimensional liquid-vapor interface oscillations for time between 225
and 427 s for a rotating dewar in response to lateral impulse without and with baffles.

Bubble Finite Deformation in r-z Plane at =0 and 180o

8
.9.o

o . . . . '

Ill

o , i , , i~ ? % ~ ,' ~ ' - i , , i . . . . . . . . . . .
-70 -3,5 0 35 70 -70 -35 0 35 70 70 -35 0 35 70

RADIUS (cm) RADIUS (cm) RADIUS (crn)

Fig. 6. Time evolution of liquid vapor interface oscillations for time between 225 and 427 s for
a rotating dewar in the r - z plane at 0 = 0 ° and 180" in response to lateral impulse without and with
baffles.

coefficients in r e s p o n s e to l a t e r a l i m p u l s e in a m i c r o g r a v i t y e n v i r o n m e n t . It c a n be c o n -
c l u d e d as follows: (a) I m p u l s e s t a r t e d to a c t u a t e o n the fluid s y s t e m at t i m e t = 0 w i t h p e a k
v a l u e a n d e n d at t = 1 0 - 2 s. H o w e v e r , fluid m a s s c e n t e r f l u c t u a t i o n s r e a c t e d g r a d u a l l y in
r e s p o n s e to the i m p u l s e w i t h m a x i m u m a m p l i t u d e of fluid m a s s c e n t e r f l u c t u a t i o n s in x-axis
 Response of lateral impulse on liquid helium sloshing 959

Bubble Finite Delormation in X-Y (r-o) Plane at z=122 cm

(a) I=225 s [bl t=311 s (c) t=427 s

(a) t=225 s fb) t=311 s (c) t=427 s

j .'_-'-'-:z:.'.',,

/70' ' i , , i , , i , , ~ ,
-'70 -35 0 35 70 -35 0 35 70 -70 -35 0 35 70
RADIUS (ore) RADIUS (ore) RADIUS (cm)

Fig. 7. Time evolution of liquid vapor interface oscillations for time between 225 and 427 s for
a rotating dewar in the x y (r-O) plane at height z = 122 cm in response to lateral impulse without
and with baffles.

at time t = 80 s and in y-axis at t = 90 s for action transferring from the x- to the y-axes due
to the effect of Coriolis forces. (b) Impulsive force was applied to the fluid element in the
positive x-direction which pushed the liquid in the positive x-direction and the bubble in the
negative x-direction. The displacement of the fluid mass center was sinusoidally shifting
continuously even long after the ending of the impulse. This displacement of the fluid mass
center fluctuations reached the maximum value at 82 s after the ending of the impulse. Then
the fluid mass center started to oscillate back and forth due to the effect of surface tension
forces, in addition to the reflection and bouncing of flows from the dewar side-wall. (c) The
oscillations of the fluid mass center fluctuations continued for a long period of time even
after the ending of impulse due to extremely low viscosity of helium II fluids. (d) It shows
exponentially decaying tendency of fluid mass center fluctuations in dotted lines for the
container without baffles.

3.2. Bubble oscillations in a rotating dewar with baffles

In this study, the time evolution of the l i q u i d - v a p o r interface oscillations for the
container with baffles due to the sloshing dynamics driven by the same impulse as that
actuating on the dewar without baffles has been investigated. Figure 2(B) shows the time
sequence evolution of the three-dimensional l i q u i d - v a p o r interface oscillations activated
by the impulse on the fluid systems with baffles. It shows that the bubble was deformed in
response to the impulse and was gradually recovered back to the equilibrium shape at the
end of the 4000 s simulation. In this case, the bubble was oscillating in the middle of the
dewar without contacting the edge of the baffles. A comparison of Figs 2(A and B) shows
that less asymmetric behavior of oscillations are characterized for bubble oscillations for
a dewar with baffles than that for a dewar without baffles.
Figure 3(B) shows the time sequence of the sloshing dynamics governed l i q u i d - v a p o r
interface profiles in the rotating dewar in response to impulse along the positive x-direction
in the vertical r - z plane at 0 = 0 ° and 180 ° for a dewar with baffles. It shows the following
conclusions: (a) the horizontal cross-section of the doughnut-shape bubble behaves in the
similar manner of leftward and rightward movement of oscillations for the containers with
960 R . J . H u n g a n d Y. T. L o n g

Time Evolution of Fluid Mass Center Variation

( With and Without Baffles )
(A)
3.0-

1 C.9
X
Z
0
Z -1.5" /Wii"hout Baffles
o
-3.0
c~ 0
(B)
g~ 3.0-

Baffles
1.5 ./...With
CL ..................
/'. - 7,,-?t"-22~:. ......... :-:........ ;.
0.0

,A -1.5

-3.0
-/Wit'hout Baffles

(c)
3.0] .'~i"-. r - Without Baffles
,,%u z°ll k-",<'r7..
"-_
C9
C~
o . o l , ","V, i'~,'- ": " ""v'""ZL:2":::::::"~:=" .....
o iooo 2000 3000 4ooo
Time (see)
Fig. 8. Time sequence evolution of fluid mass center fluctuations in response to lateral impulse for
a container with and without baffles. Fluid mass center fluctuations in (A) x-axis, (B) y-axis, and (C)
absolute values of radial direction.

and without baffles: (b) the container with baffles demonstrates a smaller amplitude of
fluctuations than that for the container without baffles; and (c) it shows a circumvention of
flow around the baffles for flow with baffles, and is not shown for flow without baffles.
Figure 4(B) shows the time evolution of the sloshing dynamics governed liquid-vapor
interface oscillations for the container with baffles in response to impulse along the positive
x-direction in the horizontal (r - 0) plane for height at z = 47.2 cm. It shows that a horizon-
tal cross-section of the doughnut-shape bubble also behaves, with an eccentric-type of
rotation, with leftward and rightward movement of oscillations for the container with
baffles, as that for the container without baffles, at smaller amplitudes of oscillations.
Figures 5(B), 6(B) and 7(B) illustrate the time evolution between time from 225 to 427 s for
three-dimensional, vertical r-z plane flow profile at 0 = 0 ° and 180 °, and horizontal r-O
plane flow profile at z = 122 cm, respectively, for the rotating dewar with baffles to show
that bubbles were experiencing finite deformation. The following results are concluded from
these figures: (a) A transition from left-side bubble swelling, equilibrium, and right-side
bubble swelling in Fig. 5(B); (b) A flow transition from horizontal jet flow encircling around
the central part of bubble to the formation of circulating flow around the top-left, top-right,
bottom-left and bottom-right around the bubble in Fig. 6(B); (c) A pattern of how flow
circumventing around the baffles in Figs 6(B) and 7(B); (d) A transition of flow patterns from
 Response of lateral impulse on liquid helium sloshing 961

circulating flow with centers along 0 = 45 ° and 225 ° to circulating flow with centers along
0 = 135 ° and 315 ° for the purpose of flow to circumventing effectively around the baffles in
Fig. 7(B).
Figure 8 shows the time evolution of the growth and decay of the bubble mass center
fluctuations with solid lines for the rotating dewar with baffles in response to the impulse.
Figures 8(A-C) show the time evolution of the locations of the fluid mass center fluctu-
ations in x~, Ye, and fluctuating locations in radial axis of cylindrical coordinates
R e [ = (x 2 + yZ)l/2], respectively, for the container with baffles. Similar to the container
without baffles, the effect of Coriolis forces which carry over the fluctuations from x- to
y-axes also applies to this case. This means that the fluid mass center fluctuations are
primarily in the x- and y-axes and not in the z-axis. The locations of maximum absolute
values of the fluid mass center fluctuations in the radial direction are Max(lxel,
lyel) = (2.24, 1.67)cm; while the fluctuating values of the maximum fluid mass center
disturbances in the radial coordinates are Max (Axe, Aye) = (2.64, 2.52) cm. Similarly, it also
shows Ix~l > [Yel and Axe > Aye for fluctuating fluid mass center disturbances in the
rotating dewar with baffles in response to the impulse along the x-axis.
A comparison of Figs 2 and 8 between the container with and without baffles also
illustrates some peculiar behavior of cryogenic helium fluids with temperature below
2-point (2.17°K) in which helium demonstrates a number of remarkable properties of
superfluidity such as extremely low viscous and surface tension coefficients in response to
impulse in microgravity environment. It can be concluded as follows: (a) Impulse started to
actuate on the fluid system at time t --- 0 with peak value and end at t = 10- 2 s for both the
container with and without baffles. However, the fluid mass center fluctuations reacted
gradually in response to the impulse with a maximum amplitude of fluid mass center
fluctuations in the x-axis about the same time as that between the container with and
without baffles at time t = 80 s, there is a 15 s delay of maximum amplitude fluid mass
center fluctuations in the y-axis for the container with baffles than that for the container
without baffles. (b) Fluid mass center locations were oscillating sinusoidally for both the
container with and without baffles long after the ending of the impulse. This displacement of
fluid mass center in the radial direction for both cases reached the maximum value at 82 s
after the ending of impulse. Then the fluid mass center started to oscillate back and forth
with a longer wave period of oscillations for the container with baffles than that for
container without baffles due to the greater effects of surface tension and viscous forces
between liquid-solid, vapor-solid and liquid-vapor interface for the container with baffÊes
than that without baffles, in addition to the reflection and bouncing of flows from the wall.
(c) The oscillations of fluid mass center fluctuations continued for a long period of time after
the ending of impulse due to extremely low viscosity of helium II fluid which eventually
decays the mass center fluctuations. A comparison of the container with and without baffles
shows that the container with baffles can reduce 13% of the maximum value of fluid mass
center dislocation and decrease 25% of the maximum amplitude of fluid mass center
fluctuations than that for the container without baffles.

4. CHARACTERISTICS OF SLOSH WAVES ACTIVATED SLOSH REACTION FORCE

In this study, characteristics of the slosh wave activated fluid reaction force and torque
fluctuations exerted on the dewar in response to lateral impulse are investigated. The
mathematical formulation of the fluctuations of slosh reaction forces and torques exerted
on the dewar are illustrated earlier [18-24]. With reference to Section 3 for sloshing
dynamics driven by impulse together with mathematical formulations [3, 18, 24], one can
calculate the slosh reaction force and its associated moments acting on the dewar.
Figure 9 shows the computed time variation of the fluctuations of slosh reaction force,
with broken lines for the container without baffles and solid lines for the container with
baffles, exerted on the dewar in response to the impulse. Figures 9(A-C) illustrate the time
evolution of the fluctuations of slosh reaction forces exerted on the container in x- and
y-components, and the total resultant force activating on the container, respectively. In
addition to the slosh reaction forces exerted on the container in response to impulse, the
962 R.J. Hung and Y. T. Long

Variations of Slosh Reaction Forces Acting on

The Dewar with and without Baffles
(A)
z°°'° l :',-. / WithoutBaffles
I~ "\"/'¢".......x:.....
r-,
iO0.O
i "'"""
00 - -

z -I00.0

-200.0
: v
L
Bar s
5~
Z
© O3) 200,0 -
!.,?...,~.,.~.
~ Without Baffles
100.0-

O.O-

,.A -lO0.O"
'J ~- With Baffles
<
-200.0
©
(c)
_/-withoutB fles
r-

~oo.o ~ ~-.--.~.-:~.g:..

0 1000 2000 3000 4000

Time (s)
Fig. 9. Time sequence of the fluctuations of slosh reaction forces exerted on the dewar in response to
lateral impulse for a container without and with baffles. Slosh reaction force fluctuations (A) along
x-axis, (B) along y-axis, and (C) absolute values of total slosh reaction force exerted on the
container.

exponentially decaying tendency of these slosh reaction forces are also plotted. As the
impulse was applied laterally along the x-axis in container bound coordinates, the slosh
reaction forces actuating on the container were in the radial direction along the x- and
y-axes, and there was no force in the z-axis.
There are some peculiar data differences shown between the results of the plots illustrated
in Fig. 9 and the data file. In the plotted figure, shown in Fig. 9, the maximum absolute
values of slosh reaction forces exerted on the container without baffles are Max (IFx], IF~ h
IF, I). = (190, 165, 195) dyn, and those of the slosh reaction forces exerted on the container
with baffles are Max(IFxh levi, IF, I)~ = (130, 120, 142) dyn. On the other hand, in the data
file, the maximum absolute values of slosh reaction forces exerted on the container without
baffle are Max(IFxl, IFyl, tF, I), = (1328, 0.41, 1329)103 dyn, and those of the slosh reaction
forces exerted on the container with baffles are Max(IFxl, IFrl, IFrl)b = (1411, 0.23, 1412)
103 dyn. These data can draw the following results: (a) For time t < 100 ms, slosh reaction
forces in response to lateral impulse introduce a large amplitude force greater than 106 dyn
exerted on the dewar at the moment when impulse was applied. (b) This result is nothing but
an indication of Newton's Third Law of Motion for action and reaction in which instan-
taneous reaction force of 106 dyn was immediately created by lateral impulse with acceler-
ation of 10-2 go, time period of impulse of 10-z s, and fluid mass of 215.7 kg [ 12]. (c)These
large amplitude slosh reaction forces damp out quickly at time t > 1 s.
 Response of lateral impulse on liquid helium sloshing 963

To understand these instant excitation and dissipation of slosh reaction forces in the
initial stage, the time-average of the impulse force in response from the cryogenic fluid to
the container are computed. The time-average impulsive force for impulse from time t = 0
to tk is

I d,k i - fi
~Fidt

*t~
(dyn).

t o dt

During the initial time period of 1 s, the major contributions are the component along the
x-direction and the contributions from the components in the rest of the directions are
negligibly small. Some selected data of the time-average slosh reaction impulsive forces for
the first 100 ms exerted on the container without baffles are 1226, 655, 429, 334, 263, 220,
191, and 122 (103 dyn) for the time period from t = 0 to 10, 20, 30, 40, 50, 60, 70, and 100 ms
respectively, while those of the time-average slosh reaction impulsive forces exerted on the
container with baffles are 1284, 687, 477, 372, 299, 249, 214, and 136 (103 dyn), respectively,
for the similar time period. After t > 1 s, a drastically great damping effect causes the
time average slosh reaction impulsive force finally reducing to 39 dyn for the container
with baffles in comparison with that of 76 dyn for the container without baffles for the time
period from t = 0 to 4000 s. Obviously this information is not properly illustrated in the
first second of the plot because these data coincide with the vertical axis of force in
Fig. 9.
A comparison of Figs 8 and 9 can draws the following conclusions: (a) There is a great
difference between fluid mass center fluctuations and slosh reaction force for the time period
in response to the impulse. Fluid mass center fluctuations of cryogenic helium are gradually
built and reach its peak value at t = 82 s after the ending of impulse while the instantaneous
and immediate response is reacted from the cryogenic helium to the container through the
production of slosh reaction forces. (b) A large slosh reaction force with a value of 106 dyn is
immediately produced in response to the impulse and acting on the container with baffles,
while a similar order of magnitude value of slosh reaction force is also immediately induced
in response to the impulse and acting on the container without baffles. After time t > 1 s, the
slosh reaction force for the container with baffles becomes smaller than that for the
container without baffles because of the viscous damping effect activating between the
fluid-solid interface for the dewar with baffles. (c) A similar result was concluded for the
time-average slosh reaction impulsive force acting on the container. A higher decay rate of
slosh reaction impulsive force is concluded for the container with baffles than that for the
container without baffles due to viscous damping on baffles at time t > 1 s. (d) The major
slosh reaction forces acting on the container are in the x - y plane and have no effect in the
z-axis because the lateral impulse is activating on the x-axis and transferring to the y-axis
through the effect of Coriolis forces in the rotating container. (e) The time evolution for the
amplitude fluctuations of slosh reaction forces activating on the container with baffles is
smaller than that for the container without baffles and slowly decays exponentially due to
viscous damping for time t > 1 s to the end of t = 4000 s. It shows 28% reduction of
maximum slosh reaction force fluctuations acting on the container with baffles than that on
the container without baffles for time t > 1 s. (f) Wave period characteristics for slosh
reaction force fluctuations activating on the container between dewar with and without
baffles are quite different. The container with baffles prolongs the wave period and makes
the amplitude of the fluctuations smaller than that of the container without baffles for time
t>ls.
The time-dependent numerical simulations of sloshing dynamics modulated fluid mo-
ments and moment arms in response to the impulse have been carried out. For the
container without baffles, it was found that both slosh reaction moment and moment arm
vanish during the entire time period of slosh reaction force oscillations. This implies that the
sloshing dynamics modulated total resultant reaction forces are constantly pointing
through the proof mass center of the container, and there was no induced torque produced
for slosh reaction in response to the impulse activating on the fluid system for the container
964 R.J. Hung and Y. T. Long

without baffles. The situation is slightly different for the container with baffles. It is due to
the fact that the existence of baffles makes the sloshing dynamics modulated total resultant
reaction forces slightly shift away from the proof mass center of the container. The absolute
maximum value of slosh reaction moment is Max(lMxl, [M,I, [ M z [ ) = ( 1 0 -3, 32,
2) 103 dyn-cm. The zero value of Mx is due to the fact that impulse is acting along the
x-axis.

5. DISCUSSION AND CONCLUSION

Sloshing dynamics of fluid system disturbances in response to lateral impulse due to the
activation of guidance and/or attitude controls of spacecraft operation [3, 32] in micrograv-
ity have been investigated. In this study, the effect of surface tension on a partially-filled
rotating dewar (liquid helium and helium vapor) activated by the time-dependent impulse in
the lateral x-direction with and without baffles have been carried out. Study is based on the
computation of three-dimensional and non-inertial frame Navier-Stokes equations subject
to the initial and boundary conditions [17-20-1. The initial condition for the liquid-vapor
interface profiles were adopted from the steady-state formulation of rotating dewar de-
veloped by Hung and Leslie [25].
In this study, the time evolution of the sloshing dynamics governed liquid-vapor
interface oscillations in response to lateral impulse have been investigated. The simulation
of fluid mass center fluctuations shows that the amplitude of bubble mass center reached the
maximum values at 82 s after the ending of the impulse. In the meanwhile, a great value of
slosh reaction force was reacted immediately right at the moment of the peak value of the
impulse imposed on the fluid systems. At the very beginning, the magnitude of slosh
reaction force activating on the container with and without baffles is about the same value.
After time t > 1 s, slosh reaction force activating on the container with baffles becomes
smaller than that for the container without baffles because of the viscous damping between
the fluid-solid interface for baffles. The maximum slosh reaction force fluctuations activat-
ing on the container without baffles are greater than that for the container with baffles. In
fact, it shows that dewar with baffles can effectively reduce 25% and 28% each of the
maximum value of fluid mass center fluctuations, and maximum slosh reaction force
fluctuations, respectively, in comparison with that of dewar without baffles.
Because of the peculiar physical properties of cryogenic liquid helium [8-15], extremely
low values of surface tension and viscosity coefficient make oscillations of bubble and
sloshing dynamics modulated fluctuating reaction forces actuating on the container in
response to the impulse continue for a long period of time and gradually decay exponenti-
ally. This paper also discusses how a container with baffles can help to damp out these
fluctuations. In this research, it has illustrated how the amplitude of cryogenic helium
sloshing dynamics disturbances grow and decay in response to the lateral impulse in
microgravity. A full understanding of the time history of growth and decay of disturbances
is important for the design of guidance and attitude controls of spacecraft utilizing
cryogenic liquid systems [3, 8-15, 32].

Acknowledgements--The authors appreciate the support received from the National Aeronautics and Space
Administration through the NASA Grant NAGS-938and NASA contract NAS8-39609/DeliveryOrder No. 103.
They would like to express their gratitude to Richard A. Potter of NASA/Marshall Space Flight Center for the
stimulating discussions during the course for the present study.

REFERENCES
1. Y. Kamotani, A. Prasad and S. Oastrach, Thermal convection in an enclosure due to vibrations aboard
a spacecraft. AIAA J. 19, 511 (1981).
2. R. J. Hung and K. L. Shyu, Liquid hydrogen suction dip and slosh wave excitation during draining under
normal and reduced gravity environment. Trans. Japan Soc. Aeronaut. Space Sci. 36, 217 (1994).
3. R.J. Hung, C. C. Lee and F. W. Leslie,Spacecraftdynamical distribution of fluid stresses activated by gravity
jitters induced slosh waves. J. Guidance, Control Dynamics 15, 817 (1992).
4. R. J. Hung, C. C. Lee and F. W. Leslie, Similarityrules in gravityjitter-related spacecraft liquid propellant
slosh wave excitation.J. Fluids Struct. 6, 493 (1992).
 Response of lateral impulse on liquid helium sloshing 965

5. R.J. Hung, Y. T. Long and H. L. Pan, Sloshing dynamics induced angular momentum fluctuations driven by
jitter accelerations associated with slew motion in microgravity. Trans. Japan Soc. Aeronaut. Space Sci. 37, 217
(1994).
6. F. W. Leslie, Measurements of rotating bubble shapes in a low gravity environment. J. Fluid Mech. 161, 269
(1985).
7. P. Mason, D. Collins, D. Petrac, F. Edeskuty, A. Schuch and K. Williamson, The behavior of superfluid helium
in zero gravity, Proc. 7th Int. Cryooenic Engineerino Conf., Surrey, England. Science and Technology Press
(1978).
8. S. W. Van Sciver, Helium Cryogenics. Plenium Press, New York (1986).
9. R. J. Donnelly, Quantized Vertices in Helium I1. Cambridge University Press, Cambridge (1991).
10. J. Wilks and D. S. Betts, An Introduction to Liquid Helium, 2nd edition. Clarendon Press, Oxford (1987).
11. R. J. Hung, H. L. Pan and Y. T. Long, Peculiar behavior of helium II disturbances due to sloshing dynamics
driven by jitter accelerations associated with slew motion in microgravity. Cryogenics 34, 641 (1994).
12. R. J. Hung and Y. T. Long, Effects of baffle on slosh reaction forces in rotating liquid helium subjected to
a lateral impulse in microgravity. Cryooenics 35, 589 (1995).
13. J. Wilks, The Properties of Liquid and Solid Helium. Clarendon Press, Oxford (1967).
14. F. E. Hoare, L. C. Jackson and N. Kurt, Experimental Cryogenics: Liquid Helium 1I. Butterworths, London
(1961).
15. R. J. Donnely, Superfluid turbulence. Scientific American 100, November (1988).
16. R.J. Hung and C. C. Lee, Effect of baffles on imbalance bubble configurations due to sloshing dynamics driven
by gravity gradient acceleration in microgravity. Canadian Aeronaut. Space J. 40, 185 (1994).
17. R.J. Hung and H. L. Pan, Differences in gravity gradient and gravity jitter excited slosh waves in microgravity.
Trans. Japan Soc. Aeronaut. Space Sci. 36, 153 (1993).
18. R.J. Hung and H. L. Pan, Gravity gradient or gravity jitter induced viscous stress and moment fluctuations in
microgravity. Fluid Dynamics Res. 14, 29 (1994).
19. R.J. Hung, H. L. Pan and F. W. Leslie, Fluid system angular momentum and moment fluctuations driven by
gravity gradient or gravity jitter in microgravity. Z. Fluwiss. Weltraumforsch. 18, 195 (1994).
20. R. J. Hung and H. L. Pan, Mathematical model of bubble sloshing dynamics for cryogenic liquid helium in
orbital spacecraft dewar container. Appl. Math. Modellin9 19, 483 (1995).
21. V. S. Advuduyesky (editor), Scientific Foundations of Space Manufacturin9. MIR, Moscow, USSR (1984).
22. R. L. Forward, Flattening space-time near the Earth. Phys. Rev. A26, 735 (1982).
23. C. W. Misner, K. S. Thorne and J. A. Wheeler, Gravitation. W. H. Freeman, San Francisco, CA (1973).
24. R. J. Hung, C. C. Lee and F. W. Leslie, Effect of baffle on the spacecraft fluid propellant viscous stress and
moment fluctuations. Trans. Japan Soc. Aeronaut. Space Sci. 35, 187 (1993).
25. R. J. Hung and F. W. Leslie, Bubble shapes in a liquid-filled rotating container low gravity. J. Spacecraft
Rockets 25, 70 (1988).
26. F. H. Harlow and F. E. Welch, Numerical calculation of time-dependent viscous incompressible flow of fluid
with free surface. Phys. Fluids 8, 2182 (1965).
27. C. W. Hirt and B. D. Nichols, Volume of fluid (VOF) method for the dynamics of free boundaries. J. Comput.
Phys. 39, 201 (1981).
28. M. G. Salvadori and M. L. Baron, Numerical Methods in Engineerin9. Prentice-Hall, Englewood Cliffs, NJ
(1961).
29. L. A. Hageman and D. M. Young, Applied lterative Methods. Academic Press, New York (1981).
30. D. Young, Iterative methods for solving partial difference equations of elliptical type. Trans. Amer. Math. Soc.
76, 92 (1954).
31. S. V. Patanker, Numerical Heat Transfer and Fluid Flow. Hemisphere-McGraw-Hill, New York, NY (1980).
32. R. J. Hung and H. L. Pan, Fluid force activated spacecraft dynamics driven by gravity gradient and jitter
accelerations. J. Guidance, Control Dynamics 18, 1190 (1995).