JOURNAL OF SPACECRAFT AND ROCKETS

Engineering Notes
Rankine Vortex Formation in potential occurrence in various engineering applications such as
spacecraft systems, hydraulic structures, and metal casting. When a
Cylindrical Tanks with Curved Base: liquid column filled in a cylindrical tank or container given with an
An Experimental Investigation initial rotation is drained through a port located at the cylinder base,
initially a dip develops at the liquid surface. When the liquid level
reaches a certain height known as “critical height” hc , this dip
Sankar Harish,∗ S. K. Hariharan,∗ Ashiq Shajahan,∗ Adithya develops into an air core vortex that subsequently enters the drain
Ganapathiraju,† Mahadev Prabhu,† and R. Ajith Kumar‡ port [1]. The time instance at which the air core enters the drain port is
called “critical time” [2]. When the air core enters into the drain port,
Amrita Vishwa Vidyapeetham, Amritapuri 690 525, India it reduces the flow area and consequently slows down the fluid
https://doi.org/10.2514/1.A35199 discharge. This phenomenon can have serious undesirable effects
on the performance of liquid propellant rocket engines and also could
lead to underutilization of the propellant [1]. It is this specific
Nomenclature practical application to propulsion systems that forms the major
Downloaded by mahadev prabhu on April 4, 2022 | http://arc.aiaa.org | DOI: 10.2514/1.A35199

D = tank diameter, m motivation for this work.

d = drain port diameter, m Lubin and Springer [3] conducted draining of initially quiesent
g = acceleration due to gravity, m ⋅ s−2 immisible liquids from cylindrical tanks. They were able to come up
h = height of the dome, m with an empirical relation that connects critical height with fluid
Hc = non-dimensional critical height (hc ∕hi ) density ratio and fluid drain rate. A numerical study conducted by
hc = critical height/critical submergence, m Sohn et al.[2] found that critical time of air core vortexing is greatly
hi = initial height, m influenced by the size of the tank and drain port. They also found that
n = initial rotation, RPM initial rotation provided to the liquid column can strengthen the air
P = pressure at center of the drain port, Pa core vortex by reducing the critical time of vortex formation. An
P∞ = hydrostatic pressure at the bottom of the cylinder, Pa experimental study conducted by Ajith Kumar et al. [4] revealed that
R = radius of curvature, m critical height of vortex formation can be highly influenced by the
r = distance measured in radial direction, m location of the drain port at the tank base (measured from the tank
tc = critical time, s center) and diameter of the drain port. They also found that, for drain
Td = nondimensional time of emptying, td ∕to ports with zero eccentricity (concentric port), vortexing can be sup-
td = time required to completely empty the tank with rotation, s pressed when the ratio of the port diameter to the tank diameter is less
to = time required to completely empty the tank without than 0.3. Further studies by Ajith Kumr et al. [5] revealed that the ratio
rotation, s of port diameter to tank diameter required to mitigate air core vortex is
td0 = theoretical time required to completely empty the tank, s a function of drain port eccentricity. Experimental studies conducted
β = volume loss coefficient by Ramamurthi and Tharakan [6] on draining of liquids through
ε = loss coefficient circular, stepped, and bell mouth ports showed that stepped ports are
μ = dynamic coefficient of viscosity of water, N ⋅ s ⋅ m−2 very effective in suppressing the air core vortex when compared to
ρ = density of draining fluid, kg ⋅ m−3 other drain port shapes. Experimental investigations by Gowda [7] on
ρair = density of air, kg ⋅ m−3 draining of liquids through tanks with square and rectangular cross
σ = surface tension of water, N ⋅ m−1 sections revealed that these tanks are effective in suppressing the air
ω = angular velocity, s−1 core vortex compared to cylindrical tanks (circular cross section). He
tentatively explained that flow assymetry and sharp corners associ-
Subscripts ated with square and rectangular cross-sectional tanks are responsible
for the vortex suppresion. Particle image velocimetry (PIV) analysis
c = critical conducted by Sohn et al. [8] verified and confirmed the result of
d = with rotation Gowda [7]. They also found that strength of air core vortex is a strong
i = initial function of the corner radius of the square cross-sectional tank.
o = without rotation Mechanism of air core generation is well explained by Nazir and
Sohn [9,10]. Their study revealed that axial and swirl velocity can
significantly influence the phenomenon of air core vortexing. Exper-
imental and numerical investigations conducted by them revealed
that strength of air core vortex is a strong function of initial height [9]
I. Introduction
and temperature [10] of draining fluid. Prabhu et al. [11] allowed
V ORTEX air core or vortex funnel formation in liquid draining
has been a spotted topic of continued research interest due to its
liquid to drain through two drain ports located at equal eccentric
locations at the bottom of the draining cylinder. They found that air
core vortex is suppressed when the size of the drain ports are equal.
Received 5 July 2021; revision received 2 March 2022; accepted for An experimental study conducted by Prabhu et al. [12] on draining of
publication 3 March 2022; published online 31 March 2022. Copyright © liquids through polygonal ports revealed that ports with octagonal
2022 by the American Institute of Aeronautics and Astronautics, Inc. All cross section are very well suited for mitigating air core vortex.
rights reserved. All requests for copying and permission to reprint should be In all the reported studies, storage vessels (cylindrical tanks) with a
submitted to CCC at www.copyright.com; employ the eISSN 1533-6794 to flat base were considered to study the Rankine vortex formation.
initiate your request. See also AIAA Rights and Permissions www.aiaa.org/
randp.
Other than cylindrical tanks, aerospace industries also make use of
*Under Graduate Student, Department of Mechanical Engineering. spherical tanks for storing the fuel. As monolithic spherical tanks
†
Ph.D. Scholar, Department of Mechanical Engineering. made by titanium alloy are weight efficient, they were used for
‡ satellite propellant tanks [13]. In the last two decades, the commercial
Professor, Department of Mechanical Engineering; amritanjali.ajith@
gmail.com (Corresponding Author). satellite industry has received high demands for satellites with larger
Article in Advance / 1
 2 Article in Advance / ENGINEERING NOTES

Fig. 2 Schematic sketch of tank base with dimensions.

diameter varies in the range of 6–12 mm with an increment of 2 mm.

The side view of the adapter is shown in Fig. 3. It should be noted that
Fig. 1 Schematic representation of problem statement. adapters are carefully fabricated such that curvature of the top side of
the adapter is matching with the radius of curvature of the tank base
Downloaded by mahadev prabhu on April 4, 2022 | http://arc.aiaa.org | DOI: 10.2514/1.A35199

and they are watertight.

capacity and longer orbital life. A propellant tank with increased The experimental setup is shown in Fig. 4. The cylindrical tank
volumetric capacity becomes an essential need to meet these inserted with a suitable adapter was firmly placed on the rotating
requirements. Due to design constraints of launch vehicles, satel- stage of the experimental setup. To ensure proper tank rotation and
lites can grow in height rather than girth. Consequently, satellite safety, an acrylic base of the cylindrical tank was screwed to the
propellant tanks grow by adding cylindrical center sections or in- rotating stage. After inserting a stopper into the adapter (stopper will
creasing the length of the center sections rather than by increasing prevent drainage of water through drain port), water was filled inside
their diameters. As a result, a cylindrical central section is added to the tank to the required height, which will be here onward called the
increase the volumetric capacity of traditional monolithic spherical initial height hi . The height of the water column was measured from
fuel tanks [13]. The new design of propellant tanks [13] is able to the drain port. In the current study, the value of initial height was fixed
meet the current demand of higher volumetric capacity in the aero- at 300 mm (see Fig. 1). The tank placed on the rotating stage was
space industry.
provided with suitable values of initial rotation using a speed con-
It is reported that propellant tanks of spacecraft and rockets are
troller. In the current study, the value of initial rotation was varied
highly prone to a sloshing phenomenon due to various maneuvers
from 0 to 80 rpm with an increment of 20 rpm. The selected range of
performed by the spacecraft [14]. In the context of sloshing of liquids,
initial rotation (0–80 rpm) is appropriate for satellite propellant tanks
the curved surface of the spherical tanks could also help to reduce the
[16,17]. After providing 4 min of rotation, the tank was brought
impact pressure when compared to cylindrical tanks [15]. Hence,
usage of cylindrical tanks with a curved base is highly recommended to rest without any jerks. Once the tank came to rest, the stopper
for the propellant tanks of spacecraft and rockets. Furthermore, the attached to the adapter was pulled out to allow draining of liquid. A
shape of propellant tanks and the tank bottom in particular can vortex air core was formed inside the tank, and the free surface height
significantly influence the vortexing phenomenon [7,8,16]. This of the water column when the air core enters the drain port was
practical relevance motivated authors to study the effect of the measured using a vertical graduated scale attached to the cylinder. For
vortexing phenomenon in cylindrical tanks with a curved base. all draining experiments, the total time needed to empty the tank was
With this impetus in mind, an air core vortexing phenomenon that also measured using a stopwatch. For each value of initial rotation
occurs during liquid draining from a cylindrical tank with a curved and port diameter, the experiment was repeated 10 times and it was
base was experimentally investigated in the present study. It is found that the uncertainty in the measurements of critical height and
specifically highlighted that detailed studies on vortex air core for- time of emptying was 0.42 mm and 0.15 s, respectively.
mation in curved-bottom tanks are not reported in the literature, and Liquid drained from the tank was collected in a collecting tank that
hence it is thought to be a novel investigation in this direction. was placed on a discharge measuring system (DMS). This device
Figure 1 illustrates the configuration of the present study. The current works on the principle of piezoresistive effect. It can give the value of
study is intended to bring to light the influence of radius of curvature discharge at an interval of 0.5 s after the commencement of draining
of draining base R on the critical height of vortex formation and the of liquid. The current study has also attempted to bring to light the
time of emptying the tank. For this purpose, an experimental study instantaneous behavior of the vortexing phenomenon.
was carried out considering a broad range of R values.

II. Experimental Arrangement III. Results and Discussion

This paper investigates how the radius of curvature of the tank base Experimental results on the effect of radius of curvature of the tank
affects the vortexing phenomenon during liquid draining from a base R on the phenomenon of air core vortexing for various values of
cylindrical tank with diameter D 96 mm and height 450 mm. In the initial rotation are presented in this section. Strength of Rankine
present study, the authors have considered four tank bases with vortex (vortex air core) is quantified by the parameter critical height
various values of radius of curvature. Out of the four tank bases, hc and it is nondimensionalized with respect to the initial height hi of
two tanks have concave curvature, one base carries a convex curva- the liquid column (H c  hc ∕hi ). Time required to completely empty
ture, and the remaining one base is flat (R  ∞). Radius of curvature the tank with rotation td is nondimensionalized with respect to the
of the four bases are 94, 70, and −188 mm (negative sign refers to time required to completely empty the tank without rotation to .
convex base), and infinity. A schematic sketch of the tank base is Nondimensional time of emptying (T d  td ∕to ) and nondimen-
shown in Fig. 2. sional critical height Hc are evaluated for all values rotation n, and
For each tank base made of acrylic, there is a provision to change nondimensional radius of curvature D∕R. It should be noted that
the drain port diameter by inserting suitable adapters (see Fig. 2) into D∕R equal to zero refers to the flat base as it is having infinite radius
the central hole at the base. In the present study, the value of the port of curvature.
 Article in Advance / ENGINEERING NOTES 3

Fig. 3 Side view of the adapter. Fig. 5 Variation of Hc with n for various values of D∕R at d∕D
0.10.
Downloaded by mahadev prabhu on April 4, 2022 | http://arc.aiaa.org | DOI: 10.2514/1.A35199

Figure 5 shows the variation of nondimensional critical height with emptying. As the value of D∕R increases, nondimensional time of
initial rotation for various values of D∕R at d∕D equals 0.10. It can be emptying also increases. This means curved bases will give rise to
observed that for all values of D∕R, critical height increases with reduced discharge compared to flat bases. It should be noted that, as
initial rotation and this observation complies with the observation critical height of vortexing increases, strength of the vortex is also
made by Prabhu et al. [16]. D∕R  0 (flat base) gives rise to the increased and this will result in more blockage of the drain port. As a
lowest value of critical height, and the convex curved base begets the result, the discharge reduces and consequently time required to
highest critical height. For concave curved domes, critical height empty the tank increases. A large convex radius of curvature (D∕R 
increases with D∕R irrespective of the value of initial rotation and is −0.5106) can enhance nondimensional time of emptying, resulting
typically shown in Fig. 5. Figure 6 illustrates the video screen shots of in reduced values of liquid discharge, as shown in Fig. 7. For other
air core vortexing at the instance of critical height formation for values of d∕D, trends of T d are closely similar and hence they are not
d∕D  0.10 at 80 rpm. The result presented in Fig. 5 is reflected in presented here.
these video screen shots. For other values of d∕D, similar results are It was mentioned by Prabhu et al. [16] that the vortexing phe-
obtained and hence they are not presented here. nomenon ispinfluenced










by Reynolds number ( ρD2 ω∕μ), Froude
Figure 7 shows the typical variation of nondimensional time of number (ω D∕g), and Weber number (D3 ω2 ρ∕σ). The results
emptying with initial rotation for various values of D∕R at d∕D  shown in Figs. 5 and 7 correspond to the Reynolds number range
0.10. In the current study, variation of nondimensional time of empty- of 0–86,672, Froude number range of 0–0.828, and Weber number
ing with initial rotation is very similar to that of the critical height. That range of 0–849.
is, at any value of initial rotation, flat bottom gives the lowest value and It should be noted that the convex curved base gives rise to the
convex bottom gives the highest value for nondimensional time of highest vortexing. At the highest value of initial rotation tested

Fig. 4 Experimental setup details.

 4 Article in Advance / ENGINEERING NOTES

Hence, pressure at the drain port center P is found mathematically

from the measured values. A mathematical treatise has been carried
out to find the pressure at the drain port center.
Time required to empty the cylindrical tank with a curved base is
given by Eq. (1), which can be derived from continuity and Bernoulli’s
equations. Here, h is the height of the dome (curved base) measured
from the drain port center (see Fig. 2) and is a function of radius of
curvature of the dome R and diameter of the cylinder D. This func-
tional relationship can be derived from the geometry of the curved
base and is given in Eq. (2). The loss coefficient (ε  td0 ∕td ) is the
ratio of theoretical time of emptying td0 and actual time of emptying
td , where theoretical time of emptying is a hypothetical quantity
defined as the time required to empty the tank without any losses.
Losses incurred during draining are broadly classified as swirl losses
and losses occurring in the absence of rotation (LAR) [11,18]. Prabhu
et al. [11] has mentioned that, when liquid is drained without rotation,
losses are limited to frictional loss (friction between draining cylinder
and liquid) and entry loss. They have also assumed that these losses
do not change with rate of fluid rotation provided. Entry losses are the
head loss associated with draining fluid when fluid enters the drain
port from the large cylindrical section. When a rotating liquid column
Downloaded by mahadev prabhu on April 4, 2022 | http://arc.aiaa.org | DOI: 10.2514/1.A35199

is allowed to drain, swirl losses also occur along with LAR. That is, in
this case (case with fluid rotation) total head loss will consist of
Fig. 6 Video screen shots of vortexing at the instance of critical height frictional loss, entry loss, and swirl loss. It should be noted that the
formation for d∕D
0.10 and n
80 rpm. right-hand side of Eq. (1) (td ) is directly measured from the experi-
ment and all other parameters except loss coefficient ε are known
quantities, and hence value of loss coefficient can be obtained. It
should be also noted that the loss coefficient can attain a maximum
value of 1, which indicates no loss (td  td0 ), and it can attain a mini-
mum value of 0, indicating maximum head loss of the fluid. Figure 8
shows the variation of loss coefficient with initial rotation for various
values of D∕R at d∕D  0.10. As seen in Fig. 8, for all values of
D∕R, the value of loss coefficient is relatively high for the no rotation
case, which means that total loss (frictional loss and entry loss) is the
minimum and the current study also assumes that these losses are
invariant with fluid rotation [11]. It could also be inferred from Fig. 8
that swirl loss incurred during rotation is significantly higher than the
total loss in the no rotation case. Hence, in the present analysis LAR is
neglected.
The discharge of liquid Q from the drain tank through the drain
port just at the beginning of draining is given by Eq. (3). From Eq. (3),
it can be inferred that the area available at the drain port for liquid to
drain will be equal to loss coefficient ε times the area of the drain port.
This means the area blocked by the air core at the drain port Aair can be
determined from Eq. (4). The radius of air core vortex at the drain port
rair can be determined from Aair and is given by Eq. (5). Hai-Feng and
Hong-Xun [19] have mentioned that, in a Rankine vortex system, the
Fig. 7 Variation of Td with n for various values of D∕R at d∕D
0.10. radius corresponding to the maximum tangential velocity Rmax is the

(n  80 rpm), both the convex curved base and the concave curved
base (D∕R  1.3714) give approximately 6% more critical height
and 22% more time of emptying compared to the flat base tank. This
directly implies a reduction of the discharge flow rate by 22% due to
the introduction of concave and convex curvatures at D∕R  1.3714
and -0.5106, respectively. To specifically highlight, the absolute
value of D∕R for the convex curved base is much smaller (very mild
curvature) compared to that of the concave curved base. This means
that in a flat-bottom fuel tank, a slight convex curvature resulting
from possible precipitation of foreign particles at the tank base or
getting introduced due to manufacturing errors can intensify the
vortexing phenomenon in the tank. From the results, it could be
inferred that curved-bottom tanks are more prone to the vortexing
phenomenon, and hence these tanks should be equipped with proper
vortex arresting mechanisms. In contrast, a conical base suppresses
the vortexing phenomenon [16]. This indicates that the vortexing
phenomenon is extremely sensitive to the geometry of the tank base.
The current study also tries to find out the reason for such enhanced
vortexing in a curved base. It is reported that formation of a low static
pressure zone at the vicinity of the drain port is the major reason for Fig. 8 Variation of loss coefficient with initial rotation for various values
ingestion of air from the ambient to form an air core vortex [7,12,16]. of D∕R at d∕D
0.10.
 Article in Advance / ENGINEERING NOTES 5

vortex air core radius (Rmax  rair ). By applying Bernoulli’s equation 0.5D2  h2
in the free vortex region (r > Rmax ) and applying Euler’s equation in R (2)
2h
the forced vortex region (r < Rmax ) of the Rankine vortex, expression
for pressure as a function of radius Pr can be determined, and is d2 p










given by Eq. (6). As density of air ρair is much less than density of QεΠ 2ghi (3)
4
water ρ, the term 1∕2ρair ω2 r2  in Eq. (6) can be neglected. The
resulting equation is given by Eq. (7), from which one can infer that
d2
pressure inside the forced vortex is not a function of radial distance r. Aair  1 − ε  Π  (4)
This follows that the pressure at the center of the drain port P has the 4
same value as the pressure at r  Rmax as revealed in Eq. (7). It should r








be noted that P∞ is defined as the pressure at an infinite radial location Aair
rair  (5)
from the center of the cylinder. However, in this study maximum Π
radial location is constrained by the walls of the draining base. On the
walls, fluid is static due to no slip condition, and pressure over there 1 1
is approximately equal to hydrostatic pressure (P∞  ρghi ) [20]. Pr  P∞ − ρRmax ω2  ρair ω2 r2 (6)
2 2
Figure 9 is a typical plot between P and initial rotation n for various
values of D∕R at d∕D  0.10. For other values of d∕D, similar 1
trends are observed and hence they are not presented here. Figure 9 P  P∞ − ρRmax ω2 (7)
2
shows that for any value of initial rotation, flat bottom gives rise to
higher static pressure at the drain port, and for curved base (both Π  4−1  D2  β  hj−1 − hj 
convex and concave) the value of pressure is comparatively lower. Qj  (8)
0.5
Downloaded by mahadev prabhu on April 4, 2022 | http://arc.aiaa.org | DOI: 10.2514/1.A35199

It is highlighted here that the pressure difference at the drain port is

less than 1% when flat bottom changes to curved bottom. Prabhu
et al. [16] have obtained the static pressure distribution near the This study also attempts to bring to light the instantaneous behav-
vicinity of the drain port for the flat-bottom and conical-bottom tank ior of the vortexing phenomenon. Recession height as a function of
to explain vortex suppression taking place in conical-bottom tanks. time can be obtained from instantaneous discharge by using Eq. (8).
From their work, it could be observed that, as shape of the base In Eq. (8), β is called volume loss coefficient. It is introduced to
changes from flat to conical, change in pressure on the surface of the account for the loss in liquid volume due to air core formation inside
drain port is not significant (pressure change is in the order of 1%). the tank. The value of β is calculated from the video analysis of
This complies with the observation of the current study. It is also draining liquid from the tank at various values of initial rotation and
observed from the static pressure contour of Prabhu et al. [16] that at port diameter. It is found that before the formation of critical
other axial locations (few distances above the drain port), the change height, the value of β is approximately equal to 1, and after the critical
in pressure due to change in shape from flat to conical is greater than height formation, β will be less than 1. For Fig. 10, the value of β
70% and this is thought to have contributed to vortex suppression in was calculated to be 0.95 at d  10 mm and n  80 rpm. In Eq. (8),
the conical base. This leads to an inference that as the shape of the “j” is the time step and each time step has an interval size of 0.5 s.
cylinder changes (either from flat to conical or from flat to dome), Hence, Qj is the discharge received in the jth time step and its value is
change in pressure at the drain port is very little (order of 1%), which obtained from the DMS device. For the first time step, hj−1  h0 
cannot be expected to bring a notable change in the discharge. Rather, 300 mm (initial height of the liquid). Figure 10 shows the variation in
as aforementioned, it could be the change in static pressure at other recession height with time for initial rotation of 80 rpm at
axial locations that could influence the discharge. However, the D∕R  0 (flat tank base) and D∕R  −0.5106 (convex tank base)
explanation provided here could be considered as tentative. A more for d∕D  0.10. It could be qualitatively inferred that at any instance
in-depth investigation employing particle image velocimetry or a of time, recession velocity (slope of the graph shown in Fig. 10) is
numerical investigation would possibly provide more details into high for D∕R  0 compared to D∕R  −0.5106. This indicates the
the physics of the phenomenon. slower draining of liquid in the convex base tank due to more intense
  air core vortex formation resulting in more time of draining (Fig. 7).
2D2 p




p


 4 4  R  h2.5 2  h2.5 From Fig. 10, it can also be found that, for the convex base, at 51 s
td  p






h − h  p






−
εd2 2g εd2 2g
i
3 5 (end of draining) recession height is not zero but recession velocity is
(1)

Fig. 9 Variation of P with initial rotation for various D∕R at Fig. 10 Variation of recession height with time for d
10 mm and
d∕D
0.10. n
80 rpm.
 6 Article in Advance / ENGINEERING NOTES

zero. This means that complete draining of liquid is not possible in the 1993, pp. 786–788.
convex-shaped base. At the end of draining, some residual amount of https://doi.org/10.2514/3.26395
liquid gets collected at the cylinder base due to base curvature [7] Gowda, B. H. L., “Draining of Liquid from Tanks of Square or Rec-
(convex) and this hinders complete draining. This is another demerit tangular Cross Sections,” Journal of Spacecraft and Rockets, Vol. 33,
No. 2, 1996, pp. 311–312.
of a convex-shaped base. For other D∕R ratios, the graphical trends of [8] Sohn, C. H., Ju, M. G., and Gowda, B. H. L., “PIV Study of Vortexing
recession height will lie between D∕R  −0.5106 and D∕R  0. During Draining from Square Tanks,” Journal of Mechanical Science
and Technology, Vol. 24, No. 4, 2010, pp. 951–960.
https://doi.org/10.1007/s12206-010-0207-9
IV. Conclusions [9] Nazir, K., and Sohn, C. H., “Study of Aircore Phenomenon and Influ-
In the present study, effect of radius of curvature of the tank base on ence of Water Height During Liquid Draining,” Journal of Mechanical
Rankine vortex formation when a liquid column is rotated and Science and Technology, Vol. 31, No. 8, 2017, pp. 3831–3837.
drained out of a cylindrical tank was experimentally investigated. It https://doi.org/10.1007/s12206-017-0727-7
was found that by altering the value of nondimensional radius of [10] Nazir, K., and Sohn, C. H., “Effect of Water Temperature on Air-
Core Generation and Disappearance During Draining,” Journal of
curvature of the tank base D∕R from zero (flat-bottom tank), the air Mechanical Science and Technology, Vol. 32, No. 2, 2018, pp. 703–
core vortexing phenomenon intensifies (higher critical height), 708.
adversely affecting the time of draining (more time to drain the https://doi.org/10.1007/s12206-018-0118-0
liquid). Establishment of low values of static pressure at the center [11] Prabhu, M., Kumar, R. A., Gopikrishnan, T. H., Deshpande, P. J.,
of the drain port and its vicinity is found to be the major reason Anandhakrishnan, U., Kiran, A. S., and Govindu, R. P.,, “Rankine
for vortex enhancement in curved bases. Concave curved base tanks Vortex Formation During Draining: A New Twin Port Suppression
are commonly used as propellant tanks for spacecraft, rockets, and Strategy,” Journal of Applied Fluid Mechanics, Vol. 13, No. 1, 2020,
satellites due to their high volumetric capacity and less impact pre- pp. 147–160.
https://doi.org/10.29252/jafm.13.01.30202
Downloaded by mahadev prabhu on April 4, 2022 | http://arc.aiaa.org | DOI: 10.2514/1.A35199

ssure they experience during sloshing. Results indicate that air core
[12] Prabhu, M., Kumar, R. A., Nair, R. R., and Anith, T. A., “Liquid
vortexing gets amplified in cylindrical tanks with a concave curved Draining Through Polygonal Ports: An Investigation on Gas Entraining
base. Hence, this study suggests that cylindrical tanks with concave Vortices,” Journal of Spacecraft and Rockets, Vol. 56, No. 5, 2019,
curved bases should be equipped with suitable vortex suppression pp. 1651–1657.
mechanisms. Study results also reveal that convex curved tanks with https://doi.org/10.2514/1.A34458
high values of radius of curvature (very mild curvature) are also prone [13] Tam, W., Hersh, M., and Ballinger, I., “Hybrid Propellant Tanks for
to vortexing. In a flat- bottom fuel tank, development of slight convex Spacecraft and Launch Vehicles,” 39th AIAA/ASME/SAE/ASEE Joint
curvature resulting from precipitation of foreign particles or manu- Propulsion Conference and Exhibit, AIAA Paper 2003-4607, July
facturing errors of the tank base can aggravate the process of air 2003.
entrainment and the consequent vortexing phenomenon. Further- https://doi.org/10.2514/6.2003-4607
[14] Lenahen, B., Bernier, A., Gangadharan, S., Sudermann, J., and Marsell,
more, cylindrical tanks with a convex curved base will not be able B., “A Computational Investigation for Determining the Natural
to drain the liquid completely out of the tank. Hence, it is recom- Frequencies and Damping Effects of Diaphragm-Implemented Space-
mended that the use of a convex curved base must be avoided in craft Propellant Tanks,” Collection of Technical Papers—AIAA/ASME/
propellant tanks. ASCE/AHS/ASC Structures, Structural Dynamics and Materials
Conference, AIAA Paper 2012-1886, April 2012.
https://doi.org/10.2514/6.2012-1886
Acknowledgments [15] Xue, M. A., Chen, Y., Zheng, J., Qian, L., and Yuan, X., “Fluid
The authors would like to thank Harikrishnan M.R., Graduate Dynamics Analysis of Sloshing Pressure Distribution in Storage Vessels
of Different Shapes,” Ocean Engineering, Vol. 192, Nov. 2019, Paper
student of Mechanical Department, Amrita School of Engineering,
106582.
Amritapuri, for his sincere help and support. https://doi.org/10.1016/j.oceaneng.2019.106582
[16] Prabhu, M., Sreenath, K., Ajith Kumar, Y., Jayakumar, J. S., and Joshy,
P. J., “Rankine Vortex Suppression in Tanks with Conical Base: A
References Numerical Investigation,” Journal of Spacecraft and Rockets, Vol. 58,
[1] Agarwal, D., Basu, P., John Tharakan, T., and Salih, A., “Prediction of No. 2, 2021, pp. 326–333.
Gas-Core Vortices During Draining of Liquid Propellants from Tanks,” https://doi.org/10.2514/1.A34794
Aerospace Science and Technology, Vol. 32, No. 1, 2014, pp. 60–65. [17] Lee, D. O., Park, G.,, and , and Han, J. H., “Experimental Study on On-
https://doi.org/10.1016/j.ast.2013.12.001 Orbit and Launch Environment Vibration Isolation Performance of a
[2] Sohn, C. H., Son, J. H., and Park, I. S., “Numerical Analysis of Vortex Vibration Isolator Using Bellows and Viscous Fluid,” Aerospace Sci-
Core Phenomenon During Draining from Cylinder Tank for Various ence and Technology, Vol. 45, Sept. 2015, pp. 1–9.
Initial Swirling Speeds and Various Tank and Drain Port Sizes,” Journal https://doi.org/10.1016/j.ast.2015.04.012
of Hydrodynamics, Vol. 25, No. 2, 2013, pp. 183–195. [18] Prabhu, M., Kumar, R. A., Warrier, R., Mouli, B. S. N. V., and Kiran,
https://doi.org/10.1016/S1001-6058(13)60353-4 S., “Liquid Draining Through Multiple Ports: An Investigation on
[3] Lubin, B. T., and Springer, G. S., “The Formation of a Dip on the Surface Air Core Vortex Formation,” Fluid Dynamics, Vol. 56, No. 5, 2021,
of a Liquid Draining from a Tank,” Journal of Fluid Mechanics, Vol. 29, pp. 718–731.
No. 2, 1967, pp. 385–390. https://doi.org/10.1134/S0015462821050116
https://doi.org/10.1017/S0022112067000898 [19] Hai-Feng, L. I., and Hong-Xun, C., “Experimental and Numerical
[4] Ajith Kumar, R., Joykutty, J., Shaji, R. K., and Srikrishnan, A. R., Investigation of Free Surface Vortex,” Journal of Hydrodynamics,
“Vortex Suppression Through Drain Port Sizing,” Journal of Aerospace Vol. 20, No. 4, 2008, pp. 485–491.
Engineering, Vol. 29, No. 4, 2016, Paper 06016002. [20] Basu, P., Agarwal, D., Tharakan, T. J., and Salih, A., “Numerical Studies
https://doi.org/10.1061/(asce)as.1943-5525.0000609 on Air-Core Vortex Formation During Draining of Liquids from Tanks,”
[5] Ajith Kumar, R., Nair, R. R., Prabhu, M., and Srikrishnan, A. R., “Vortex International Journal of Fluid Mechanics Research, Vol. 40, No. 1,
Formation During Draining from Cylindrical Tanks: Effect of Drain Port 2013, pp. 27–41.
Eccentricity,” Journal of Aerospace Engineering, Vol. 30, No. 5, 2017, https://doi.org/10.1615/InterJFluidMechRes.v40.i1.30
Paper 06017001.
https://doi.org/10.1061/(ASCE)AS.1943-5525.0000731 M. S. Miller
[6] Ramamurthi, K., and Tharakan, T. J., “Shaped Discharge Ports for Associate Editor
Draining Liquids,” Journal of Spacecraft and Rockets, Vol. 30, No. 6,