Journal of Mechanical Science and Technology 24 (10) (2010) 1997~2006

www.springerlink.com/content/1738-494x
DOI 10.1007/s12206-010-0618-7


Numerical and experimental investigation of vortex breaker effectiveness on the
improvement in launch vehicle ballistic parameters
Mahdi N. Mahyari
1,*
, Hasan Karimi
2
, Hasan Naseh
2
and Mehran Mirshams
2

1
Faculty of Mechanical Engineering, K.N Toosi University of Technology, Tehran, Iran
2
Faculty of Aerospace Engineering, K.N Toosi University of Technology, Tehran, Iran

(Manuscript Received July 1, 2009; Revised April 22, 2010; Accepted May 9, 2010)
----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Abstract

The focus of the present study is to investigate the effectiveness of installing vortex breakers at the outlet of launch vehicle tanks on
postponing vortex formation and decreasing the critical height of propellants while discharging. Analytical results in the absence of a
vortex breaker show that the effects of the Weber and Reynolds numbers in the flow field can be ignored for values greater than 720 and
1.1 10
5
, respectively; and critical height can be considered as a function of Froude number under aforementioned conditions. The ana-
lytical criteria are verified by two-dimensional, axis symmetrical, transient and two-phase numerical model. Eventually, some experi-
ments are conducted to examine the effectiveness of the applied vortex breakers in reduction of the critical height of propellant. Experi-
mental results show that a 30% decrease can be achieved in critical height by using a particular type of vortex breaker. Additionally, the
carried out simulations for an existing two-stage launch vehicle indicate a 13% increase in orbital altitude, which in turn proves the con-
siderable improvement in launch vehicle mass/energetic capabilities.

Keywords: Critical height; Launch vehicle; Propellant; Vortex breaker; Weber and Reynolds numbers
----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------



1. Introduction
Generally, vortex flow formation at the outlet of a tank is a
significant and undesirable phenomenon. Vortex formation in
flux fields with either free surface or near a sink (in several
systems such as water supply systems or liquid propellant
tanks of launch vehicles) is of great importance. As the tank
draining nears the end, some free surface disturbances may
occur as shown in Fig. 1(a). The height of the liquid in this
condition is called the dimple height. As discharge continues,
the dimple extends rapidly toward the outlet and forms a
swirling air funnel. This leads to the penetration of gas in the
outflowing liquid as depicted in Fig. 1(b). The height of the
liquid when the tip of the funnel reaches to the outlet is called
the critical height.
Gas presence in the expulsion liquid causes a sudden in-
crease in the turbo-pump rotational velocity and combustion
instability in the launch vehicle engine. Since feed pump op-
eration is not recommended in such conditions, the engine
must be stopped before any gas ingestion occurs. A common
approach to resolve this problem is to consider the overcharg-
ing of the propellant tanks followed by an adequate discharge
until the propellant height reaches the critical value. However,
the overcharged propellant will remain in the tank at the end
of the mission. It is obvious that the excess propellant lowers
the ballistic characteristics of a launch vehicle such as orbital
altitude and velocity. Applying a vortex breaker at the bottom
of a tank reduces the volume of the remaining propellant. Vor-
tex breakers postpone the time of vortex formation by dimin-
ishing the swirling flow and changing the pressure and veloc-
ity fields near the outlet.
Some analytical and experimental studies have been per-
formed on vortical flows in hydraulic engineering, in the past.
In some cases the free surface is located too close to the outlet,
as is done in a power plants intake [1, 2]. A simple device
was suggested for preventing vortex formation in a small scale
cylindrical container by L. Gowda et al. [3, 4]. The effects of
the initial tangential velocity and the intake eccentricity on the
critical height were examined by Piva et al. [5]. Also, an un-
steady three-dimensional numerical model was created by the
European Space Agency (ESA) in order to simulate the flow
field near the vortex breaker in the oxidizer tank of the
ARIANE 5 launcher [6] (see Fig. 2). On the other hand, to
date no research has been carried out to examine the effect of
applying a vortex breaker on the launch vehicle mass-
energetic capabilities. Therefore, the main aim of this paper is

This paper was recommended for publication in revised form by Associate Editor
Do Hyung Lee
*
Corresponding author. Tel.: +989126487870, Fax: +982177791045
E-mail address: [email protected]
KSME & Springer 2010
1998 M. N. Mahyari et al. / Journal of Mechanical Science and Technology 24 (10) (2010) 1997~2006


to examine the effect of using vortex breakers on the im-
provement of ballistic parameters of a launch vehicle.
The attempts by the launch vehicle community for attaining
an appropriate design by using the existing technologies and
instruments have led to successful results, for instance see [7-
9]. In this research, it is suggested that the optimization of the
design parameters is incorporated in the conceptual design
process by which the reduction of propellant and structure
mass can be achieved. This, in turn, leads to the improvement
in the ballistic characteristics. To this end, the amount of re-
duction in critical height is calculated experimentally, which
leads to a decrease in fuel and oxidizer mass. Then, the
achieved mass reduction is utilized in the performance simula-
tion algorithm and thereby the enhancement of the capabilities
of launch vehicle is illustrated.
Launch Vehicle Conceptual Design (LVCD) code was ap-
plied to illustrate the effect of vortex breakers on improvement
in launch vehicle performance [10]. In the LVCD software,
the results of various researches are employed to develop a
conceptual design algorithm for launch vehicle design [9, 11,
12]. Consequently, a launch vehicle with optimized major
design parameters would be the outcome of this developed
algorithm. Finally, the improvements in orbital altitude and
velocity of a two-stage launch vehicle are evaluated by this
code.

2. Dimensional analysis
For the purposes of the present study, let the critical height
be defined as the submergence of the outlet at which incipient
air entrainment is possible. Considering the critical height as
the dependent variable, the following functional relationship
may be written as

( )
1
, , , , , ,
cr
h f d D g = (1)

By choosing , d and as the repeating variables and
doing dimensional analysis of the variables of Eq. (1), the
following is obtained.

2
2
, , ,
cr
h D d d
f
d d gd



=


(2)

Since the parameter / D d is large enough in the tests, the
effect of the boundary of the tank on the vortex formation is
negligible and can be dropped from Eq. (2), i.e.

( )
3
Re, ,
cr
h
f Fr We
d
= (3)

It is necessary to find ranges for the dimensionless parame-
ters in Eq. (3) that /
cr
h d is not a strong function of these
parameters. This goal can be obtained analytically to reduce
the number of test iterations [13-17].

(a)


(b)

Fig. 1. Geometry of a tank near the outlet. (a) Formation of a dimple
on the surface. (b) The dimple extends and forms an air funnel.


Fig. 2. The geometry of liquid oxygen tank bottom of ARIANE 5 with
internal anti-vortex device [6].

M. N. Mahyari et al. / Journal of Mechanical Science and Technology 24 (10) (2010) 1997~2006 1999


3. Analytical consideration
The considered flow situation has been depicted in Fig. 1.
Since in the vicinity of the vertical axis the flow is assumed
steady, axis-symmetric and laminar, the equations of motion
for an incompressible fluid are as follows:

( )
1
0
z
r
v
r v
r r z

+ =

(4)
2
2 2
2 2 2
1 1
r r
r z
r r r r
v v v
v v
r z r
P v v v v
r r r r r z

+ =



= + + +




(5)
2 2
2 2 2
1
r
r z
v v v v
v v
r z r
v v v v
r r r r z

+ + =



= + +




(6)
2 2
2 2
1 1
z z
r z
z z z
v v
v v
r z
P v v v
g
z r r r z

+ =



= + + + +




(7)

The model is obtained by superimposing a radial flow on a
free vortex flow given by

( )
2
v f r
r

=

(8)

is the value of circulation in free vortex flow where the

radius is far enough from the axis. Near the pipe entrance
where r r <
o
and
cr
z h , the radial velocity can be ap-
proximately assumed as ar [1]. By using this assumption,
from the Eq. (4) the axial velocity is obtained as 2a z . Since
the axial velocity is equal to at
cr
z h = , then a can be
calculated by / 2
cr
h . Substituting Eq. (8) into Eq. (6) yields:

0
r
f d f
v
r dr r

+ =


(9)

Note that
r
v ar = , 1 f for r , and v

has no
singularity at 0 r = . Then, by substituting the solution of the
Eq. (9) into Eq. (8), a tangential velocity distribution is ob-
tained as follows.

2
1
1 exp
2 2
a
v r
r

=


(10)

The maximum tangential velocity (
,c
v

) occurs at the distance

c
r . This distance is measured from the axis and is given by

2
2.5
c
r
a

= (11)

Therefore, the maximum value of tangential velocity is given
by

2 2
,
0.00516
c
a
v

= (12)

Eq. (10) can be normalized by Eqs. (11) and (12) as follows:

2
,
1.4 1 exp 1.25
c
c c
v r r
v r r

=




(13)

The critical height is obtained from the pressure distribution
and by substituting Eq. (10) into Eqs. (5) and (7). Then inte-
grating on the resulting equation leads to

2
2 2
0
1
2
r
v
P P dr a r
r

= +

o
(14)

Where P
o
, the pressure at the tip of vortex due to surface
tension, is equal to 2 /
c
r . For distances beyond
c
r , P
tends to hydraulic pressure. The third term on the right side of
Eq. (14) is small compared to the second term for r r <
o
. In-
tegrating from axis to r r =
o
and by using Eqs. (13) and (14),
the following equation is obtained.

2
2
,
2
3.4 1- 0.5
2
c c
cr
c
v r
h
r g r g

= +




o
(15)

Where
c
r is given by Eq. (11). The radius ratio in Eq. (15)
can be replaced by
( )
2
5 /
cr
h r
o
. This value is small enough
so that the value of bracket in Eq. (15) can be assumed to be
unity. Considering Eqs. (11) and (12), Eq. (15) is simplified as
follows.

2
2
0.9 0.0043
cr
cr
h
h
g g

= + (16)

In terms of dimensionless parameters, Eq. (16) can be rewrit-
ten as the following form.

2 1/ 2
2 1/ 2 1
2 2
0.89 Re
0.0026 Re
cr cr
h h
Fr We
d d
Fr N

= +


+
(17)

A criterion for neglecting effects of surface tension may be
established by requiring that ratio between the surface tension
and circulation terms in Eq. (15) to be less than 10%. Thus

2
,
12
c c
v r

> (18)

2000 M. N. Mahyari et al. / Journal of Mechanical Science and Technology 24 (10) (2010) 1997~2006


Substituting Eqs. (11) and (12) into Eq. (18) leads to the fol-
lowing equation.

1/ 2 3/ 2 1/ 4
720 Re We Fr N

> (19)

The values of
1/ 2
Fr ,
3/ 2
N

and
1/ 4
Re for conditions of a
typical model are of the order of 1, 0.1 and 10, respectively.
Therefore, the criterion for neglecting the surface tension will
be 720 We > . Moreover, a criterion for neglecting the effect
of viscosity can be established by requiring that turbulent vis-
cosity be greater than 5 . This condition has less than a 10%
effect on the critical height [18]. It follows from Eq. (16) that
without the surface tension term, this criterion is met when the
following condition is satisfied.

2
3
3
2
Re 1.1 10
cr
h gd
d


>


(20)

The value of
3
( / )( / )
cr
h d gd

is of the order of 10, and
the criterion for neglecting viscous effects is
5
Re 1.1 10 > .
Under the conditions where Re > 1.1 10
5
and We > 720, Eq.
(17) can be simplified as follows:

1/ 3
1.55
cr
h
Fr
d
= (21)

Eq. (21) is the main goal of this section. This relation implies
the critical height of propellants can just be assumed as a func-
tion of Froude number provided that Re > 1.1 10
5
and
We > 720. In other words, neglecting the viscous and surface
tension effects of the flow, the number of the required tests
can be decreased significantly.

4. Numerical investigation
Generally, the vortex formation at the propellant tanks out-
let is a 3-D, unsteady, turbulent and two-phase phenomenon.
Thus, a complete numerical solution is time-consuming and
expensive. Instead, a 2-D, axis-symmetric, transient and two-
phase numerical model is established so as to simulate the
flow field. Both Reynolds and Weber numbers are greater
than the criteria mentioned in section (3). According to Eq.
(21), the Froude number is the most important parameter
which affects the critical height when Re > 1.1 10
5
and
We > 720. Under these conditions, the numerical model
should be carried out for various Froude numbers to show the
variation of the critical height.
The domain of the solution, a structured mesh, and the
boundary conditions are shown in Fig. 3. Air enters the inlet
boundary with constant pressure and water is expulsed with
constant mass flow rate at the outlet. Solution continues while
the volume fraction of water at the outlet is equal to 1. Passing
the air bubble through the outlet reduces the water volume
fraction. Therefore, the boundary condition at the outlet will
not be satisfied and solution will be stopped. The boundary
condition for the axis of symmetry can be expressed as
( )
0
/
r
r
=
=0, Where represents any arbitrary physical
quantity like velocity components or pressure.
The tracking of the interface between the phases is accom-
plished by the solution of a continuity equation for the volume
fraction of different phases. For the ith phase, this equation
has the following form [19]:

2
1
. 0 , 1
i
i i
i
V
t

+ = =

r
(22)

A single set of momentum equations is solved throughout
the domain, and the resulting velocity field is shared among
the phases. The momentum equations depend on the volume
fractions of all phases through the properties and [19].

( )
( )
( )
.
.
T
V
VV
t
P V V g F

+ =

= + + + +

r
r r
r r r
r
(23)

In general, for an n-phase system, the volume-fraction-
averaged (like density or viscosity) takes the following
form [19]:

1
n
i i
i

=
=

(24)

By using a 2-D axis-symmetric model, the critical height of
the fluid can be estimated as a function of Froude number
under the conditions mentioned for Eq. (21). But investigation
of the effects of the vortex breakers requires a fully 3-D nu-
merical model. However, it is possible to show the effect of a
circular plate located at the top of the intake on the flow field
using a two dimensional axis-symmetric model. Fig. 4 shows
a part of the unstructured mesh generated for this object.
The k model was employed to simulate the turbulence
in numerical code. A segregated solver was used to solve the
flow field equations, and the transient SIMPLE algorithm was
applied to couple the pressure and velocity fields. Also, con-
vection terms of governing equations were made discrete us-
ing the first-order method. In this numerical model, character-
Fig. 3. The geometry of tank with 30040 structured elements. Bound-
ary conditions can be supposed constant pressure at inlet and constant
mass flow rate at outlet.

M. N. Mahyari et al. / Journal of Mechanical Science and Technology 24 (10) (2010) 1997~2006 2001


istic time is a function of the free surface velocity and the
representative length of the free surface elements. The Cou-
rant number is a dimensionless parameter which compares the
time step in a calculation to the characteristic time of transit of
a fluid element across a control volume:

cell
characteristic
free surface
x
t CFL t CFL
v

= = (25)

Where CFL is Courant number,
cell
x is representative length
of free surface elements and
free surface
v is free surface velocity.
Generally, the Courant number is chosen as 0.25 in 2-D prob-
lems [20]. Therefore, the order of the time step can be calcu-
lated by dividing the order of representative length of interface
elements by the order of the free surface velocity as the fol-
lowing:

( )
( )
3
4
( )
( )
( )
1 10
0.25 2.5 10 sec.
1 / sec
cell
free surface
O x
O t CFL
O v
m
m

= =

= =
(26)

According to Eq. (26), the time step can be chosen equal to
2.5 10
-4
second in the solution. However, to enhance the
accuracy, the time step was considered to be 1.0 10
-4
second.
Fig. 5 shows contours of velocity magnitude for both the
gas and the liquid phases. The contours of the tangential ve-
locity are illustrated in Fig. 6. It can be seen that the tangential
velocity of a particle increases while it is moving towards the
outlet. Fig. 7 shows the stages of air funnel formation at the
interface of phases and the penetration of air towards the out-
let. Figs. 8-10 demonstrate contours of velocity magnitude,
tangential velocity and phases, respectively, but for the case
where a circular plate has been located on top of the intake
(see Fig. 4). Comparing Fig. 6 with Fig. 9 shows that tangen-
tial velocity decreases effectively by means of a flat plate.
Also, it can be seen that an air funnel does not appear at the
axis of the tank where maximum axial velocity occurs (Fig.
10). These two phenomena reduce the critical height by post-
poning vortex formation and diminishing swirling flow.

Fig. 5. Contours of velocity magnitude.



Fig. 6. Contours of tangential velocity.


(a) (b)
(c) (d)

Fig. 7. The stages of vortex formation and penetration of air towards
the intake.

Fig. 4. The geometry of tank with 23095 unstructured elements con-
sidering a circular plate near the outlet.

2002 M. N. Mahyari et al. / Journal of Mechanical Science and Technology 24 (10) (2010) 1997~2006



Fig. 8. Effect of a circular plate on velocity magnitude.



Fig. 9. Effect of a circular plate on tangential velocity.


(a) (b)


(c) (d)

Fig. 10. Effect of a flat plate on interface disturbance and reduction in
critical height during the time.

To verify whether the effect of the viscosity on the critical
height is negligible for Reynolds number greater than 1.1 10
5
,
the variation of the critical height was compared when Rey-
nolds number varies from 4.5 10
5
to 9.0 10
5
at a constant
Froude Number ( 3.882 Fr = ). As represented in Table 1,
numerical results show just 5.6% difference in critical height
when Reynolds number becomes twice.
Also, mesh independency of the numerical solution was ex-
amined by comparing the results of a coarse mesh with a finer
mesh. The differences in Table 2 show that the mesh used in
the model is fine enough.
Fig. 11 shows that the numerical results are in good agree-
ment with the analytical result (Eq. (13)) to estimate the varia-
tion of tangential velocity. Since the effect of the wall shear
stress was ignored for deriving Eq. (13), some deviation can
be seen from the numerical results near the wall, especially for
r/r
c
> 13.

5. Experiments
Some experiments are conducted to measure the dimple
height in propellant tanks by applying different vortex break-
ers. This is because the values of Reynolds and Weber num-
bers for both the model and the prototype conditions are
greater than the mentioned criteria (see Table 3). Thus, the
experiments are dedicated to investigating the effects of
Froude number on the dimple height for various outlets. In
other words, the Froude number is a parameter which can
relate the ratio of the inertia force to gravity force in flight
conditions. Under flight conditions, the Froude number must
be defined as a function of the axial acceleration (
x
a ), the
pitch angle ( ) the angle between launch vehicle axis and
local horizon, and local gravity ( g ) which is described be-
low.
Table 1. Independency of Reynolds number for critical height at Fr =
3.882.

Reynolds Number
Outlet Geometry
Re = 4.510
5
Re = 9.010
5

Difference
Without plate
(Fig. 4)
1.83 1.95 5.6%
With plate
(Fig. 5)
1.04 1.07 2.8%

Table 2. Independency of mesh size for critical height at Fr = 3.882.

Number of elements in mesh
Outlet Geometry
2010
3
3010
3

Difference
Without plate
(Fig. 4)
1.84 1.83 0.72%
With plate
(Fig. 5)
1.07 1.04 2.6%


Fig. 11. The comparison of the analytical and numerical results for the
tangential velocity variation.

M. N. Mahyari et al. / Journal of Mechanical Science and Technology 24 (10) (2010) 1997~2006 2003


( ) sin
x
Fr
a g d

=
+
(27)

Eq. (27) can be simplified to / gd when both
x
a and
are equal to zero. The tests are carried out using a trans-
parent cylindrical tank made of Plexiglas with a diameter of
1.25 m and a height of 2.0 m with a drain hole of internal di-
ameter 112 mm centrally located at the bottom of tank along
the vertical axis (see Fig. 12). The working liquid in the ex-
periments is water at room temperature. The pump discharges
the water in the propellant tank to the reservoir tank while
valves (4) and (5) are open and valves (6) and (7) are closed.
The dimple height is measured for various flow rates and dif-
ferent vortex breakers. By changing the status of valves (4, 5
closed and 6, 7 open), the propellant tank can be recharged.
The effects of four below devices are examined as vortex
breakers, experimentally.
1. A circular flat plate with the diameter of 3d .
2. A circular flat plate with the diameter of 3d with four ra-
dial blades (Fig. 13(a)).
3. A circular flat plate with the diameter of 3d with eight
radial blades (Fig. 13(b)).
4. A circular flat plate with a porous wall (Fig. 13(c)).

The vortex breakers are installed at a height of 0.25d from
the bottom of the tank. Increasing the installation height of the
vortex breakers reduces its effectiveness. A vortex breaker
with a diameter of 3d installed at a height of 0.25d provides a
good balance between pressure drop and effectiveness [21].
Plate with diameter greater than 3d or installation height less
than 0.25d leads to higher pressure drop and cavitation prob-
ability.
The variations of the dimensionless dimple height versus
Froude number for the four cases are demonstrated in Fig. 14.
The results show 11% decrease in the dimple height via instal-
lation of a flat plate with a diameter of 3d at the vicinity of
intake. Also, by adding four radial blades to the flat plate,
18.5% decrease in the dimple height is noticed. Results show
that a circular flat plate with a porous wall (Fig. 13(c)) is more
effective and 30% reduction can be obtained in the dimple
height and propellant weight (includes fuel and oxidizer
weight for two stages of launch vehicle) will be reduced by
280 kg.

6. LVCD software definition
Launch Vehicle Conceptual Design (LVCD) software was
Table 3. The characteristic parameters for the under study two-stage
launch vehicle.

2
nd
stage launch vehicle 1
st
stage of launch vehicle
Oxidizer
(N
2
O
4
)
Fuel
(UDMH
**
)
Oxidizer
(AK27
*
)
Fuel
(Kerosene)
unit Parameter
1458 795 1603 805
3
/ kg m
4.2 5.6 19 8.24 / . kg ms
4
10
25.1 24.1 - 54 / N m
3
10
7.6 5.3 109 31.1 / kg s m&
4.86 6.16 6.93 3.92 / m s
65 55
2
/ m s
x
a
0 25 deg
3.13 3.97 2.69 1.52 - Fr
6.2 3.2 6.6 4.3 -
-5
Re 10
5.1 4.6 - 5.2 -
-4
10 We
* AK27: 73% HNO
3
+ 27% H
2
O
** UDMH: Unsymmetrical Di-Methyl Hydrazine


Fig. 12. Experimental setup: (1) propellant tank, (2) centrifugal pump,
(3) magnetic flow meter, (4, 5, 6 &7) butterfly valves, (8) reservoir
tank.

(a) (b)

(c)

Fig. 13. Three types of vortex breaker.

2004 M. N. Mahyari et al. / Journal of Mechanical Science and Technology 24 (10) (2010) 1997~2006


employed to illustrate the improvement in performance of a
launch vehicle through the used vortex breaker. The main
objectives of this software are a reduction of the cost and time
of conceptual design phase. To this end, in this code, the
weight of structure, the calculations of propellant and mass
distribution of each stage, to launch maximum payload mass
to the orbit, and the pitch program trajectory, to achieve the
maximum final velocity, will be optimized [11].
Generally, in LVCD software, the calculations are per-
formed based on optimization, but herein just a brief review of
the most fundamental equations is presented to demonstrate
and clarify how the variables change.
Let us assume that launch vehicle mass breakdown for ith
stage is as follows (see Fig. 15):

0, , , , ,
; 1, 2,.., N
i st i P i gua i Pay i
M m m m m i = + + + = (28)

Where N indicates the number of launch vehicle stages and is
equal to 2 in this study. In addition, the characteristic velocity
achieved at the end of ith stage is [22]:

0, 1 ,
, , ,
0,
. .Ln ;
i st i
i sp i F i F i
i
M m
V g I
M

+
+
= = (29)

Where
, F i
denotes the ratio of the final mass to the initial
mass which is decreased by employing a vortex breaker. Also,
, sp i
I is called the specific impulse of the launch vehicle engine
and is obtained through dividing the launch vehicle engine
thrust to the exhaust mass flow rate.
7. Improvement in launch vehicle parameters
The ultimate goal of this effort is to simulate and improve
the performance of a two-stage launch vehicle. The results of
current simulation carried out by LVCD code for conditions,
represented in Table 3, are in Figs. 16 and 17. The simulation
results show that use of anti-vortex devices at the outlet of
liquid-propellant tanks leads to enhancement in engine burn-
ing time, thrust-to-weight ratio, axial acceleration (especially
in upper stages), and velocity and orbital altitude. Figs. 16 and
17 show, respectively, the improvement in orbital altitude and
axial velocity achieved by means of a vortex breaker. Fig. 17
shows a 13% increase in the orbital altitude by using a specific
type of vortex breaker (Fig. 13(c)) for the present two-stage
launch vehicle.

8. Conclusions
Analytical investigation of this study indicates that viscous
and surface effects of free-surface vortices can be neglected
when the intake Reynolds and Weber numbers are greater
than 1.1 10
5
and 720, respectively. Comparing the results of
numerical simulation for Re = 4.5 10
5
with 9.0 10
5
reveals
a maximum of 5.6% difference in critical height estimation at
a constant Froude number, which in turn verifies the analytical
criteria. So, in the experiments, the dimple height of the fluid
was considered as a function of Froude number and outlet
type. It is possible to achieve a 30% decrease by using a flat

Fig. 14. The effects of different vortex breakers on the dimple height.


Fig. 15. Stage and block mass breakdown concept.


Fig. 16. Improvement in orbital altitude.


Fig. 17. Improvement in launch vehicle velocity.

M. N. Mahyari et al. / Journal of Mechanical Science and Technology 24 (10) (2010) 1997~2006 2005


plate with porous wall. In this case, the mass of the two-stage
launch vehicle may be reduced by 280 kg and the result of
LVCD simulation shows a 13% increase in the orbital altitude.
This tacitly means that for a specific desired mission, with the
specified orbital altitude, by decreasing the remaining mass, it
is possible to carry a heavier payload mass to the same altitude.

Acknowledgment
The authors would like to express their deep gratitude to Dr.
Kh. Alipour, Dr. M. Mohseni and A. Kalabkhani for their
support and helpful comments.

Nomenclature------------------------------------------------------------------------
a : Factor of proportionality
x
a : Axial acceleration of launch vehicle at the end of
each stage
CFL : Courant number
d : Inner diameter of intake pipe
D : Diameter of tank
f : Function
F
r
: Body force vector
Fr : Froude number
g : Local gravitational acceleration
cr
h : Critical height
, sp i
I : Specific impulse of ith stage
m& : Propellant mass flow rate
0,i
M : Total mass of ith stage
, st i
m : Dry structure mass of ith stage
, P i
m : Propellant mass of ith stage
, gua i
m : Guarantee (remaining) mass of ith propellant
, Pay i
m : Payload mass of ith stage
N : Number of launch vehicle stages
N

: Circulation number
P : Pressure
P
o
: Pressure at the tip of the vortex
, , r z : Radial, tangential and axial coordinates
c
r : Radial distance of maximum tangential velocity
r
o
: Intake pipe radius
Re : Reynolds number
t : Time
: Mean velocity at the tank outlet
free surface
v : Free surface velocity
, ,
r z
v v v

: Radial, tangential and axial component of velocity

,c
v

: Maximum tangential velocity

V
r
: Velocity vector
i
V : Launch vehicle velocity at the end of ith stage
We : Weber number
i
: Volume fraction of ith phase in each cell
cell
x : Representative length of free surface elements
: Volume-fraction averaged of any physical quantity
i
: Volume fraction of ith phase

: Circulation at the far field

: Kinematic viscosity
: Pitch angle of launch vehicle at the end of each
stage
: Density
: Surface tension of liquid

References
[1] J. E. Hite and W. C. Walter, Velocity of Air Core Vortices at
Hydraulic Intakes, Journal of Hydraulic Engineering, 120
(3) (1994) 284-297.
[2] M. Maleewong, J. Asavanant and R. Grimshaw, Free Sur-
face Flow under Gravity and Surface Tension Due to an Ap-
plied Pressure Distribution: I Bond Number Greater Than
One-Third, Journal of Computational Fluid Dynamic, 19 (4)
(2005) 237-252.
[3] B. H. Lakshamana Gowda, Draining of Liquid from Tanks
of Square or Rectangular Cross Sections, Journal of Space-
craft, 33 (2) (1995) 311-312.
[4] S. Mizuki, B. H. Lakshamana Gowda and T. Uchibaba,
Visualization Studies using PIV in a Cylindrical Tank with
and without Vortex Suppressor, Journal of Visualization, 6
(4) (2003) 337-342.
[5] M. Piva, M. Iglesias, P. Bissio and A. Calvo, Experiments
on Vortex Funnel Formation During Drainage, Physica A
329 (2003) 1-6.
[6] J. Lacapere, Focus on the CFD for the ARIANE 5 Launcher:
Space Engineering Activities at Cryospace and Air Liquid,
Fluent News, (2005). Available online at www.fluent.com/
about/news/newsletters.
[7] M. J. Lewis and T. Rice, Design of a University Launch
Vehicle System, The 10th Applied Aerodynamics Confer-
ence, Palo Alto, CA, AIAA 1992-2663.
[8] S. Chakroborty, J. R. Wertz and R. Conger, The Scorpius
Expendable Launch Vehicle Family and Status of the Sprite
Small Launch Vehicle, The 1st Responsive Space Confer-
ence, Redondo Beach, CA, AIAA 2003-9005.
[9] L. Virgil, J. Hutchinson and J. R. Olds , Estimation of
Launch Vehicle Propellant Tank Structural Weight Using
Simplified Beam Approximation, The 40th AIAA/ASME/
SAE/ASEE Joint Propulsion Conference and Exhibit, Fort
Lauderdale, Florida, USA, AIAA 2004-366.
[10] M. Mirshams, H. Karimi and H. Naseh, Launch Vehicle
Conceptual Design (LVCD) Software based on multi-
parameter optimization, The 7th International. Aerospace
Engineering Conference, Tehran, Iran, Paper No. 281 (2008).
[11] A. Tewari, Atmospheric and Spaceflight Dynamics, Model-
ing and Simulation with Matlab and Simulink, Birkhauser,
Boston, USA (2007).
[12] M. Turner, Rocket and Spacecraft Propulsion, Principles,
Practice and New Developments, Springer, Chichester U.K.
(2009).
[13] N. Yildirim and F. Kocabas, Prediction of Critical Submer-
gence for an Intake Pipe, Journal of Hydraulic Research, 40
(4) (2002) 507-517.
[14] G. Echavez and E. McCann, An Experimental Study on the
2006 M. N. Mahyari et al. / Journal of Mechanical Science and Technology 24 (10) (2010) 1997~2006


Surface Vertical Vortex, Experiments in Fluids, 33 (3)
(2002) 414-421.
[15] B. H. Lakshamana Gowda, P. J. Joshy and S. Swarnamani,
Device to Suppress Vortexing during Draining from Cylin-
drical Tanks, Journal of Spacecraft, 33 (4) (1995) 598-600.
[16] G. Haller, on Objective Definition of a Vortex, Journal of
Fluid Mechanic, 525 (2005) 1-26.
[17] T. Alrutz and M. Rutte, Investigation of Vortex Breakdown
over a Pitching Delta Wing Applying the DLR TAU-Code
with Full Automatic Grid Adaptation, The 35th AIAA Fluid
Dynamics Conference and Exhibit, Toronto, Ontario, Can-
ada, AIAA 2005-5162.
[18] A. J. Odgaard, Free Surface Air Core Vortex, Journal of
Hydraulic Engineering, 112 (7) (1996) 610-620.
[19] C. Crowe, M. Sommerfield and Y. Tsuji, Multiphase Flows
with Droplets and particles, (1998).
[20] H. K. Versteeg and W. Malalasekera, An Introduction to
Computational Fluid Dynamics, The Finite Volume Method,
1st edition, Prentice Hall (1995).
[21] N. M. Beliaev, Calculation of Hydraulic Pneumatic System
in a Rocket, Mashinostroenieh, Moscow, Russia (1983).
[22] W. E. Wiessel, Spaceflight Dynamics, McGraw-Hill, USA
(1989).



M. Mahyari received his B.Sc. and
M.Sc. in Mechanical Engineering from
the K. N. Toosi University of Technol-
ogy of Iran in 2006 and 2008, respec-
tively. His research fields include ex-
perimental and computational fluid
mechanics and heat transfer, modeling
and simulation.
H. Karimi received his Ph.D. in Solid and Liquid Propellant
Engine Design from the Moscow state
Technical University in 1999. He joined
the aerospace engineering department of
K. N. Toosi University of Technology
in 1999 where he is presently an associ-
ate professor and head of the propulsion
group. He is also an organizing member
of the Iranian Aerospace Society. His
core research interests are gas dynamics, gas turbine, model-
ing, simulation and control of liquid and solid propellant en-
gines.

H. Naseh received his M.Sc. in Aero-
space Engineering from the K. N. Toosi
University of Technology, Iran in 2007.
His M.Sc. research involved conceptual
design of aerospace vehicles. His
research focuses on the space system
engineering, optimal control, orbital
mechanics and space propulsion
engineering. He is a Ph.D. student at K. N. Toosi University,
now.

M. Mirshams received his Ph.D. in
Aerospace Engineering from the Mos-
cow State Technical University in 1999.
He has been a faculty member at K. N.
Toosi University of Technology, Iran
since 1999. His Ph.D. research involved
space system engineering. His core re-
search interests are launch vehicle and
satellite system engineering, orbital mechanics, flight dynam-
ics and simulation engineering.