Вы находитесь на странице: 1из 11

Ocean Engineering 108 (2015) 552562

Contents lists available at ScienceDirect

Ocean Engineering
journal homepage: www.elsevier.com/locate/oceaneng

Fluid damping in rectangular tank tted with various internal


objects An experimental investigation
Santosh Kumar Nayak a, Kishore Chandra Biswal b,n
a
b

Government Polytechnic, Sambalpur, Odisha, India


Department of Civil Engineering, National Institute of Technology Rourkela, Odisha, India

art ic l e i nf o

a b s t r a c t

Article history:
Received 29 December 2013
Accepted 25 August 2015
Available online 15 September 2015

The potential of internal objects in changing the dynamic system characteristics of mobile liquid carrying
rectangular containers is experimentally investigated in the present study. This study involves identication of system characteristics such as natural frequency and damping. Three different congurations of
centrally installed internal objects; bottom-mounted vertical bafes, surface-piercing wall-mounted
vertical bafes, and bottom-mounted submerged-blocks have been tried out as potential passive slosh
damping devices. A series of painstaking experiments has been conducted in a rigid rectangular tank
model on a shake table under lateral harmonic excitation. A frequency response of various internal object
arrangements on free surface elevation has been studied. Having identied the system characteristics,
the sloshing responses of liquid (water) to harmonic sinusoidal loading for different bafe congurations
are investigated. The time variation of the free surface elevation to the bafe conguration and height
have been highlighted. Sine sweep and Logarithmic decay method have been resorted to in the
experimental discourse. The parametric study shows that the surface-piercing wall mounted bafes are
the most effective one in slosh damping among the three congurations.
& 2015 Elsevier Ltd. All rights reserved.

Keywords:
Sloshing frequency
Frequency response
Fluid damping
Internal objects
Bafes

1. Introduction
Low frequency uctuation of free surface liquid in a partially
lled liquid container is regarded as sloshing, which becomes
violent under favourable condition. Sloshing is an important
physical phenomenon in many elds of engineering interest in
which a partially lled liquid container is an important structural
component of the system. The problem of liquid sloshing in containers of various shapes has received considerable attention in
transportation engineering from the middle of the last century. To
begin with, the requirement for a precise assessment of the
sloshing induced hydrodynamic forces was felt in aerospace
applications. The vicious motion of the liquid fuel in the tanks of
aerospace vehicles was studied by Graham and Rodriguez (1952).
In offshore applications, the effect of free surface liquids on board
can create problems that may lead to loss of stability of the ocean
going vessels (Cleary, 1982; Bass et al., 1980).
Liquid carrying road tankers constitute an important problem
and have dragged the attention of many researchers involved in
the eld of slosh dynamics. Sloshing has often been a potential
source of danger for road tankers resulting in accidents due to
n

Corresponding author.
E-mail address: kcbiswal2003@yahoo.co.in (K.C. Biswal).

http://dx.doi.org/10.1016/j.oceaneng.2015.08.042
0029-8018/& 2015 Elsevier Ltd. All rights reserved.

poor manoeuvrability. Precise estimation of sloshing related


hydrodynamic loads is one of the major design issues of liquid
transporting containers. The sloshing and its associated forces
depend on a wide variety of variables such as shape and dimension of the container, liquid ll depth, amplitude and frequency of
external excitation, exibility of the container, etc. The slosh
generated on account of the possible critical combination of the
above parameters, should not exceed a certain prescribed maximum value; and if it so does, it needs to be suppressed by some
means. The inherent damping due to liquid viscosity is very useful
in small size containers. In relatively large containers, natural
damping on account of the viscous boundary layers has often been
found inadequate to counter violent sloshing and hence several
articial means have been tried to undermine this threat. The use
of passive anti-slosh devices such as bafes, oating cans, oating
lids and mats are the most sought after means for slosh suppression. Bafes and other such ow obstructing internal objects are
found to have given promising results in controlling and preventing the vehicle instability during manoeuvring which has
inspired many researchers to conduct analytical, numerical and
experimental methods of investigation to corroborate the reliability of these damping devices.
The damping of bare wall containers owes its origin to three
sources: (i) viscous dissipation at the free surface of liquid, (ii)
viscous dissipation at the side walls and bottom of the tank, and

S.K. Nayak, K.C. Biswal / Ocean Engineering 108 (2015) 552562

(iii) viscous dissipation in the interior of the tank. In case of a


bafed tank, an additional source due to relative motion between
the liquid and the bafe wall comes into the damping domain. It is
often not feasible analytically to accommodate damping contribution from so many sources. Hence, the effect of different types
of bafes is usually determined experimentally.
The conguration of slosh suppression devices such as shape,
size, stiffness, perforation, gaps, and location has been discussed
extensively in NASA Space Vehicle Design Criteria, NASA-SP-8031
(1969). Laws and Livesey (1978) studied the effects of the screen
on velocity distribution that effects a change in the ow direction
and a reduction in pressure. Evans and McIver (1987) presented an
approximate solution based on eigenfunction and Galerkin
expansions and studied the effect of vertical bafe on sloshing
frequencies in a rectangular water tank. Watson and Evans (1991)
further extended this technique to study the resonant frequencies
of uid in rectangular and circular containers having internal
bodies such as surface or bottom-mounted rectangular blocks and
submerged circular cylinders. Jeyakumaran and McIver (1995)
offered an approximate method for the calculation of the oscillation frequencies in a rectangular tank with a cylinder as an
internal structure. Porter and Evans (1995) presented a Galerkin
based approximate method to study sloshing problems in rectangular tanks with internal structures. Using linear water wave
theory, Choun and Yun (1996, 1999) studied the effects of the
dimension and location of the submerged block on the sloshing
characteristics of the liquid in a rectangular tank. Their ndings
show that the existence of internal block decreases the sloshing
frequencies. The wave elevations increase in the vicinity of the
block and a large hydrodynamic force can be exerted on the tank
wall and block when the block is closer to the wall. Armenio and
Rocca (1996) adopted the nite difference method (FDM) to solve
the two-dimensional Reynolds averaged NavierStokes (RANS)
equations in order to account for the viscosity and vorticity, both
of which play a dominant role in liquid tank with internal element.
They reported that the presence of a vertical bafe considerably
reduced the sloshing loads in the whole range of roll frequencies.
Tait et al. (2005) developed numerical ow models to simulate
tuned liquid dampers (TLD) with slat screens and assessed the
model efciency with experimental results. Gavrilyuk et al. (2006)
proposed fundamental solutions of the linearised problem on uid
sloshing in a vertical cylindrical container having a thin rigid-ring
horizontal bafe. The analytically oriented approach not only
provided good approximations of natural frequencies and modes
but also captured the singular asymptotic behaviour of the velocity
potential at the sharp bafe edge. Mitra and Sinhamahapatra
(2007) developed a pressure-based nite element method and
analysed the slosh dynamics of a partially lled rigid rectangular
container with rectangular shaped bottom-mounted submerged
components. Gavrilyuk et al. (2007) used a nonlinear asymptotic
modal method based on the Moiseev asymptotic ordering to study
the nonlinear resonant sloshing in the bafed cylindrical tank.
They studied the inuence of size and the vertical location of the
bafe on the effective frequency domains of the steady state
resonant waves. Askari and Daneshmand (2009) analytically
investigated the coupled vibration of a partially lled cylindrical
container with a thin-walled and open-ended cylindrical shell as
an internal body. Askari and Daneshmand (2010) and Askari et al.
(2011) investigated the effects of a rigid internal body on dynamic
characteristics of a cylindrical container partially lled with liquid.
Mitra et al. (2010) made a numerical assessment of the slosh
dynamics of a liquid lled container due to the presence of multiple submerged block and bafe over a partition wall. Hasheminejad and Aghabeigi (2012) studied the effect of vertical bafes on
two-dimensional liquid sloshing characteristics in a halffull nondeformable horizontal cylindrical container of elliptical cross

553

section. They employed the linear potential theory in conjunction


with the successive conformal transformation technique to arrive
at the solution. Wang et al. (2012) proposed an improved semianalytical scheme to study the frequencies and modes of liquid
sloshing in a rigid cylindrical container with multiple rigid annular
bafes.
Gedikli and Erguven (1999) developed a numerical model
based on linear boundary element method to study the effect of
rigid ring bafe in damping liquid oscillation in a rigid cylindrical
tank. Studied the slosh damping potential of different bafe congurations in heavy-duty trucks with circular and elliptical tanks
of various ll depths using the commercial CFD solver FLUENT. Cho
and Lee (2003a, 2003b) employed a coupled ALE nite element
method to examine numerically the damping effects of disc-type
elastic bafe on the dynamic characteristics of cylindrical fuelstorage tank boosting with uniform vertical acceleration. Cho and
Lee (2004) carried out a parametric investigation on the twodimensional nonlinear liquid sloshing in a bafed tank under
horizontal forced excitation. A time-incremental nonlinear nite
element method (FEM), based on the fully nonlinear potential ow
theory in the semi-Lagrangian numerical approach, was used for
numerical analysis. They showed that the liquid motion and
dynamic pressure variation above the bafe were more signicant
than those below the bafe. In addition, they suggested that the
quantities of interest in the liquid sloshing are strongly dependent
on the bafe design parameters. Cho et al. (2005) conducted a
frequency domain parametric investigation of sloshing damping
characteristics in 2-D bafed tank subjected to forced lateral
excitation. Finite element method based on the linearised potential ow theory was used as the numerical tool for the study. An
articial damping term was employed to the kinematic free surface condition to reect the imminent dissipation effect in resonant sloshing. Biswal et al. (2003) studied the natural frequencies
and modes of liquid in a liquid-lled rigid cylindrical container
with and without bafes by using the nite element method.
Using FEM, Biswal et al. (2006) studied 2D nonlinear sloshing
response of the liquid in non-deformable cylindrical and rectangular tanks with rigid bafes subjected to harmonic base excitation. Bafes close to the free surface of the uid were found more
effective in reducing the effect of sloshing. Belakroum et al. (2010)
studied the vibratory behaviour of three different congurations of
tanks equipped with bafes using a Galerkin Least Square (GLS)
nite element model based on Arbitrary LagrangianEulerian
(ALE) description of the NavierStokes equations.
Strandberg (1978) carried out an experimental investigation on
dynamic performance and stability of horizontal circular, elliptical,
and super-elliptical tank shapes of equal capacity. He also examined the overturning limit for halffull elliptical containers with
various bafe congurations and established the preference of
vertical bafe over a horizontal one as an anti-sloshing device in
the elliptical container. Warnitchai (1998) used both analytical and
experimental investigations to determine the damping ratios of
various ow dampening devices: tank with two circular section
poles, a tank with a at plate, and tank with a wire-mesh screen
with a specic wire diameter and solidity ratio. Experimentally
studied the pressure distribution on the walls of bafed and
unbafed cylindrical tank. Maleki and Ziyaeifar (2008) proposed a
theoretical damping model to investigate the damping effect of
bafes in circular cylindrical liquid storage tanks, and simultaneously carried out experiments for verication of the theoretical
models. Their study concerned two kinds of bafes; horizontal
ring and vertical blade bafes. They observed that the damping
ratio of the sloshing mode in the presence of these bafes
depended on the tank and bafe dimensions in addition to the
location of the bafe. The ring bafes were found to be more
effective in reducing the sloshing oscillations. They also showed

554

S.K. Nayak, K.C. Biswal / Ocean Engineering 108 (2015) 552562

that the damping ratio depended on the sloshing wave amplitude.


Panigrahy et al. (2009) used two different congurations of bafe
in which they rst considered a combination of two horizontal
bafes and two vertical bafes with four holes each, whereas the
second one was a ring bafe. Eswaran et al. (2009) studied the
effect of bafes on liquid sloshing in a partially lled cubic tank.
The use of the rst category of bafes is not convincing as they
have not conducted any parametric study on the use of such bafes. They used only a given set of vertical and horizontal bafe
with xed dimensions considering them as a bafe-group ignoring
the individual effect of the horizontal bafe. However, in contravention they concluded that the use of ring bafe is more effective
as compared to the conventional horizontal bafe. The study thus
made is not comprehensive enough. The reason for the provision
of holes in the vertical bafes used in the study is also not convincing. Having used a xed number of holes, the effect of holes is
not studied. Akyildiz et al. (2013) conducted a series of experiments in a cylindrical tank subjected to rolling motion and found
that ring bafe arrangements are very effective in reducing the
sloshing loads.
From the review of the literature made as above, the following
summary of observations is deduced.
1. The liquid viscosity in tanks without bafes does not have much
effect in reducing the sloshing amplitudes in tanks of practical
dimensions.
2. The effects of bafe conguration on dynamic system characteristics, namely the fundamental sloshing frequency and
damping of liquid in the tank are signicantly different from
those in the tank without a bafe.
3. The analytical means for damping estimation are not easy due
to the complications involved in the boundary conditions, and
the inclusion of bafes makes it even more difcult.
4. Experimental investigations are the most reliable ones for
estimating the damping characteristics of the liquid sloshing.
5. The comparison among different bafe congurations is seldom
reported.
6. Although, the computational methods provide powerful means
in the solution of mathematical models for analysing the real
world problems, yet developing the precise numerical models
considering bafe damping effects may be a challenging task.
Furthermore, most of the studies on damping on account of the
bafes involve cylindrical tanks with single or multiple ring bafes. In this respect, rectangular tanks with surface-mounted bafes and bottom-mounted submerged blocks have received very
little attention. The major objectives of the present investigation
are to examine the relative effectiveness of various congurations
of internal objects in slosh damping in order to enrich the test data
bank for the verication of the numerical models and also to
recommend the most effective arrangement of internal objects for
sloshing attenuation. A series of experiments are performed for
partially-lled rectangular tanks with three different congurations of internal objects, namely; bottom-mounted vertical bafes,
surface-piercing vertical bafes, and bottom-mounted submerged
blocks in tanks with two different aspect ratios (ll depths). The
parameters are systematically changed to conduct free vibration
analysis, determine the damping ratio, and measure the sloshing
wave height for specic aspect ratio and internal object parameters. During the experiments, the tank is subjected to lateral
harmonic excitation dened by the expression: x = x0 sin 2ft ,
where x0 is the amplitude and f is the frequency of excitation.
With the above objectives in view, the experimental domain is
xed up. The experimental programme includes:

 Frequency response study of sloshing in a rectangular tank t





ted with three different congurations of centrally placed


internal objects using the sine sweep test.
Free vibration study for identication of important dynamic
property of liquid-damping in term of the hydrodynamic
damping ratio.
Dynamic response analysis of the tankliquid system with
centrally placed internal objects subjected to sinusoidal horizontal excitation at resonant frequencies.
Comparative study on the relative effectiveness of various
internal objects on sloshing damping.

2. Experimental work
The prediction of hydrodynamic loads on uid-lled tankers
moving on the road is often an important requirement in the
design. The extent of sloshing and thus the magnitude of sloshing
loads are modied due to energy dissipation. Particularly, signicant energy dissipation may occur on the account of ow
separation effects as the uid oscillates past bafes or other
obstacles in the tank. When a partially lled container such as a
road tanker is moved, it rapidly dissipates its kinetic energy due to
the presence of the bafes, and as a result, the sloshing is attenuated. The objective of the experiments discussed in this paper is
to investigate the effects of three different congurations of
internal objects and their important parameters on energy dissipation and sloshing attenuation. In the following sections, the
experimental procedure is explained in detail. The summary of the
observations from an extensive investigation is presented.
2.1. Experimental test setup
Experimental tests were carried out using a harmonically
excited shake table in the Structural Engineering Laboratory of the
Department of Civil Engineering at National Institute of Technology, Rourkela, India.
The experimental test setup as illustrated in Fig. 1 consists of:
i. a Millennium make, 1 m  1 m shake table of 100 kg payload
and capable of giving the horizontal harmonic motion of
maximum amplitude capacity 20 mm;
ii. shake table control unit to vary the frequency of motion with
an increment of 0.01 Hz within a frequency range of 0.40
2.20 Hz;
iii. a model rectangular tank of 57 cm length, 31 cm width, and
45 cm height (the length and width of the tank are same as that
of the one taken by Liu and Lin (2009)) constructed of 11.7 mm
thick Perspex sheet to prevent deformation of the tank due to
the hydrodynamic impact of water and thus to ensure the
assumption of rigid tank;
iv. two vertical entry liquid level sensors of Honeywell make tted
at both ends of the tank and kept at a distance of 0.06 m from
the walls perpendicular to the direction of excitation, to measure the sloshing elevations;
v. a dynamic LVDT (Endevco Model-SX20MER300) tted to the
shake table for accurate measurement of the displacement
amplitude of the shake table;
vi. Multichannel data acquisition system PCI 6255 from National
Instruments;
vii. a PC loaded with LabView SignalExpress software from
National Instruments for data acquisition and post-processing;
viii. two sets of bafes for two different congurations of varying
dimensions, h/d (bottom-mounted and surface-piercing), made
of Perspex sheet, with all xtures for installing them with the
tank; and one set of submerged-blocks also made of Perspex

S.K. Nayak, K.C. Biswal / Ocean Engineering 108 (2015) 552562

555

Fig. 1. Experimental setup and instrumentation: (a) Details of the instruments used in the experiment, (b) photograph of tank without any internal objects, (c) photograph of
a tank with bottom-mounted bafe, (d) photograph of surface-piercing bafe in the tank.

Table 1
Parameters of tank and internal objects for two different lling depths.
Length of tank
Width of tank
Height of tank
Still water depth
Aspect ratio
Relative height of bottom-mounted bafe
Relative height of surface-piercing bafe
Width of submerged block
Relative height of submerged block

Fig. 2. Snapshots of internal objects and their xtures.

sheet (snapshots of bafes and submerged blocks are demonstrated in Fig. 2)


ix. accessories and ttings for xing the bafes in the tank.
Two liquid depths are considered, coincident with d/L 0.26
(15 cm water depth) and 0.52 (30 cm water depth). d is the height
of still water in the tank and h is the height of the internal object
used for damping potential. The parameters of the tank and
internal objects of three different types are shown in Table 1.
2.2. Fixing the bafes and submerged blocks
The bafes of two different congurations and their supporting
setup are demonstrated in Fig. 2. The bottom-mounted vertical
bafes are supported rmly to the base of the tank in the middle
since this position is expected to correspond to the maximum
relative horizontal velocity. The rmness of the bafe is ensured
by the compressive force given by the two vertical stainless steel
rods screwed to the two clamps xed at two ends of the bafe. The

L
B
H
d
d/L
h/d
h/d
w
h/d

57 cm
31 cm
45 cm
15 cm and 30 cm
0.26 and 0.52
0.2, 0.4, 0.5, 0.6 and 0.8
0.2, 0.4, 0.5, 0.6 and 0.8
L/3
0.2, 0.4, 0.5, 0.6 and 0.8

stainless steel rods are in turn clamped to two horizontal rods


provided at the top of the tank. Adequate compressive force is
applied by pushing the rods down while tightening them with the
horizontal rods pulled upward. For larger bafes an additional
provision of bolts, two each at both the ends of the bafe, is made.
By tightening these bolts, the friction between the ends of the
bolts and the wall of the tank provides the bafe with additional
strength required to resist the drag force of water on the bafe and
prevent its lateral movement. The surface-piercing wall-mounted
vertical bafe with its top ushed with the quiescent free surface
of the liquid is xed to the wall of the tank by four sets of bolts,
two each at both ends. Again by tightening these bolts, the friction
between the end of the bolts and the wall of the tank provided the
bafe with the required strength to resist the drag force on the
bafe and prevent it from moving. The bottom-mounted submerged blocks are xed to the bottom of the tank in the same
manner as is done in the case of bottom-mounted vertical bafes.
The clamp conguration is slightly different as can be marked
from the Fig. 3. Fig. 4 depicts the problem denition where d is the
depth of still liquid in the tank, h is the height of the bafe, and
max is the maximum height of the sloshing from the quiescent
free surface.
Experiments are carried out in tanks of two aspect ratios (d/L)
of 0.26 and 0.52 corresponding to ll depths of 15 cm and 30 cm

556

S.K. Nayak, K.C. Biswal / Ocean Engineering 108 (2015) 552562

Vertical rods

Baffle

Clamp

Baffle

Supporting
Bolts

Supporting Bolts

57 cm

57 cm

Clamps

Submerged block
57 cm

Fig. 3. Internal objects and associated supporting xtures. (a) Bottom-mounted bafe, (b) Surface-piercing bafe and (c) Bottom-mounted submerged block.

Fig. 4. Problem denition. (a)Bottom-mounted bafe, and (b) Surface-mounted bafe.

respectively, with two different bafe congurations (bottommounted and surface-piercing) and bottom-mounted submerged
block. The relative height of bafes (h/d) as well as that of the
submerged blocks is varied as 0.2, 0.4, 0.5, 0.6 and 0.8.

3. Experimental results and discussion


3.1. Validation of experimental result with numerical analysis
A numerical simulation of the tank sloshing is undertaken
using the nite element model developed by the Nayak and Biswal
(2013) to validate the experimental measurement. Fig. 5 depicts
the time history of free surface sloshing elevation near the left wall
of the tank (length 570 mm, width 310 mm and water
depth 15 mm) undergoing sinusoidal excitation with an amplitude of 5 mm and excitation frequency corresponding to the fundamental sloshing frequency. It is observed from Fig. 5 that the
numerical prediction compares well with the experimental data.

3.2. Frequency response study of sloshing by sine sweep test


For a certain liquidtank system, with xed system parameters,
excitation frequency has a signicant effect on the amplitude of
sloshing elevations and thus on the whole lot of associated
dynamic characteristics. In this regard, the frequency response
study of sloshing is carried out in tanks with two aspect ratios (ll
depths); d/L0.26 and 0.52. Three different congurations of
centrally placed internal objects are used as passive slosh damping
devices to assess their slosh damping potential. The relative bafe
heights (h/d) for each case of ll depth were varied as 0.2, 0.4, 0.5,
0.6, and 0.8. The effect of height of the submerged block (width,
w0.33L) was investigated only for the tank with low ll depth, d/
L0.26.
Lateral sine sweep tests were conducted in tanks without and
with various congurations of internal objects as mentioned earlier to identify the natural frequencies of the systems. Each of the
liquidtank systems were excited with sinusoidal harmonic loading dened by expression x = x0 sin 2 ft (where x0 is the amplitude of excitation and f is the excitation frequency in cycles/s) over

100
75
50
25
0
-25
-50
-75
-100

Experiment
FEM

10

12

14

Time (s)
Fig. 5. Comparison of free surface sloshing elevations near the left wall for tank
under harmonic excitation with displacement amplitude, x0 5 mm and excitation
frequency, f (cycles/s) corresponding to the fundamental sloshing frequency.

Displacement (mm)

10

-5

-10
0

10

15

20

Time (s)
Fig. 6. Harmonic excitation with a forcing frequency, f 0.97 cycles/s (fundamental
sloshing frequency of tank without any internal objects) and displacement amplitude x0 5 mm.

a wide range of excitation frequencies (0.40 to 2.20 Hz) which


included the natural sloshing frequencies of the system. A typical
harmonic force given to the shake table at an excitation frequency
of 0.97 Hz as recorded by the dynamic LVDT is depicted in Fig. 6.
The displacement amplitude of the excitation was kept constant at
5 mm. The shaking was allowed over a signicantly long period so
as to allow the sloshing reach steady state. The time domain
sloshing elevation data for each excitation frequency was recorded
with the help of vertical entry liquid level sensor, xed near the
two end walls perpendicular to the direction of excitation, via NI
6225 data acquisition system (DAQ) and LabView SignalExpress
software. The peak amplitude of sloshing was then recorded and
all the sloshing peaks were plotted against their respective excitation frequencies to obtain the frequency response curve of a
particular liquidtank system. The excitation frequencies corresponding to the sloshing peaks of frequency response plots gave
the respective natural frequencies of sloshing. One such frequency
response plot for a tank with ll depth d/L 0.26 and without any
internal element is demonstrated in Fig. 7. The rst two natural
sloshing frequencies obtained for the above tank are 0.97 Hz and
2.02 Hz respectively. The fundamental modes of sloshing frequencies for other cases of tanks are similarly obtained.
It can be seen that presence of internal objects reduced the
fundamental frequency of the liquid. Between two different bafe
congurations, for a certain bafe height the decrease in sloshing
frequency was more in case of surface-piercing bafe. Of the three
different congurations, a maximum reduction in fundamental
frequency was observed in the case of submerged block.
For a certain bafe conguration, the effects of bafe height on
the fundamental sloshing frequencies in tanks with different ll
depths are compared in Fig. 8. The fundamental frequency
decreases with an increase in the height of the bafe. For a given
relative bafe height, the sloshing frequency of liquid in the tank
with a high ll depth and bottom-mounted bafes is always more
than that for low ll depth. In case of surface-piercing bafe, the
sloshing frequencies in both the ll depth cases considered for the

Maximum relative elevation


( / f 1) x 100

Amplitude (mm)

S.K. Nayak, K.C. Biswal / Ocean Engineering 108 (2015) 552562

557

120
100
80
60
40
20
0

0.5

1.5

2.5

Nondimensional frequency (/ 1)
Fig. 7. Frequency response of free surface wave elevation at left wall of tank of
aspect ratio, d/L 0.26, x0 0.005 m. sloshing elevation at different excitation
frequencies; f1 sloshing elevation at excitation frequency equal to the fundamental sloshing frequency.

study come closer as the relative height of the bafe increases. As


can be seen from Fig. 8(b), for bafe height h/d 40.5, the fundamental natural frequency is not much affected.
Fig. 9 shows comparative results of the effects of bottommounted and surface-piercing bafes on the fundamental frequency of the bafed rectangular liquid tank. One can nd from
Fig. 9(b) that the reduction in the fundamental frequency due to
the surface-piercing bafe of the same height is signicantly more
than that due to the bottom-mounted bafe.
Fig. 10 illustrates the effect of the submerged block on the
fundamental frequency of liquid in the tank with ll depth 0.26.
The results are compared with those obtained for bottom-mounted bafe of equal height. It may be noted that submerged blocks
are bafes with comparatively larger width. One can see that a
submerged block of similar height causes a considerably greater
decrease in frequency than bottom-mounted vertical bafe.
3.3. Effect of bafe on sloshing response
The basic concept behind the passive damping techniques of
sloshing in tanks is a manipulation of tank geometry so as to break
the ow eld into several sub-elds which helps in the dissipation
of kinetic energy of the liquid. With an aim to judge the effectiveness of bottom-mounted bafes studied in this work for
sloshing damping, the frequency response of free surface sloshing
elevation near the left wall was studied. The frequency responses
of free surface elevation for unbafed and bafed tanks with two
different bafe heights are shown in Fig. 11.
The damping effect is reected by the reductions observed in
maximum relative elevations, especially close to the fundamental
frequency. Dramatic reduction in sloshing wave height was
noticed over a wide range (0.402.20 Hz) of lateral excitation
frequencies in the presence of bafe height, h/d 0.8. In particular,
under the resonant condition, more than 80% reduction in wave
height was noted. The percentages attenuations of maximum
elevations for three different congurations of internal objects are
similarly determined.
The percentage attenuations of maximum sloshing elevations
due to the presence of internal objects in tank with d/L 0.52
could not be presented because it was not possible to nd the
maximum sloshing elevation in the tank without any internal
objects as the water sloshed out of the tank due to lack of enough
freeboard.
3.4. Hydrodynamic slosh damping
To investigate the hydrodynamic slosh damping effect of bafes
on the rst mode sloshing, the method of logarithmic decay of
wave amplitude of free vibrating liquid is employed. In this
method, the tank was harmonically excited at a frequency close

558

S.K. Nayak, K.C. Biswal / Ocean Engineering 108 (2015) 552562

Fig. 8. Comparison of fundamental sloshing frequencies as a function of bafe height in tanks with ll depth, d/L 0.26 and 0.52: (a) bottom-mounted vertical bafes;
(b) surface-piercing wall mounted vertical bafes.

Fig. 9. Sloshing frequencies in the tank as a function of bafe height (h/d) with (a) bottom-mounted vertical bafe (b) surface-piercing bafe.

elevation (0, t) during the free vibration phase was recorded. The
rate of decay () of wave amplitude was then computed, from
which the damping ratio () was calculated as follows:

Fig. 10. Comparison of sloshing frequencies in the tank with d/L0.26 due to
bottom-mounted vertical bafes and submerged components.

Maximum relative elevation


( / f 0 ) x 100

120
100
h/d=0.0
h/d=0.2
h/d=0.8

80
60
40
20
0
0.2

0.4

0.6

0.8

1.2

1.4

1.6

Frequency (Hz)
Fig. 11. Frequency response of the free surface wave elevation at the left wall of
tank with d/L 0.26 due to bottom-mounted vertical bafes of different heights;
sloshing elevation at different excitation frequencies; f0 maximum sloshing
elevation in the tank without any internal component at excitation frequency equal
to its fundamental sloshing frequency.

enough to the rst mode frequency so as to induce rst mode


sloshing. The displacement amplitude of excitation was kept at
5 mm. Sufcient time was allowed for the sloshing to attend
steady state and then the shake table was swiftly switched off to
initiate free vibration. The time history of free surface wave

1
ln i
i + j
j

2
1 2

, is the damping ratio

where i and i + j are the sloshing amplitudes in i and ij cycles of


oscillations.
Experiments were conducted for each bafe with water height
equal to 15 and 30 cm, with aspect ratios (d/L) of 0.26 and 0.52
respectively. As explained in the previous paragraph, the time
history of sloshing elevations during free vibration phase was
recorded by vertical entry liquid level sensors, located near the
two walls perpendicular to the direction of excitation. Similar
experiments were also conducted for bottom-mounted submerged
blocks.
Time history of sloshing wave heights during the free vibration
phases for different cases of ll depths and internal object
arrangements are shown in Figs. 1216. According to the gures,
the systems that show lower settling time responses have higher
damping ratios and vice versa.
3.5. Effect of internal objects on uid damping
The effect of heights of various internal objects and ll depth on
the damping ratio is illustrated in Fig. 17 for two different ll
depths. It is observed that for bottom-mounted bafes of certain
relative height (h/d), the damping ratio decreases with an increase
in the liquid ll depth. The fact of the matter may be attributed to
the increase in the convective mass of liquid for higher ll depth.
Irrespective of the ll depth and type of bafe (bottom-mounted

S.K. Nayak, K.C. Biswal / Ocean Engineering 108 (2015) 552562

Fig. 12. Time decay of sloshing amplitude in free vibration phase in a tank of aspect
ratio d/L 0.26 with bottom-mounted vertical bafes of various heights.

559

Fig. 14. Time decay of sloshing amplitude in free vibration phase in a tank of aspect
ratio d/L 0.26 with bottom-mounted submerged blocks of various heights (width,
w0.33L).

Fig. 13. Time decay of sloshing amplitude in free vibration phase in a tank of aspect
ratio d/L 0.52 with bottom-mounted vertical bafes of various heights.

or surface-piercing), the damping ratio increases with an increase


in the relative height (h/d) of the bafe. In case of bottommounted bafe, as the top of the bafe approaches the quiescent
free surface, the damping ratio signicantly increases. This fact is
consistent with the ow dynamics as there is a higher concentration of convective mass near the free surface of the uid
block. The relative motion between the convective uid block and
the bafe helps in dissipating the sloshing energy and thus the
increase in the damping ratio.

Fig. 15. Time decay of sloshing amplitude in free vibration phase in a tank of aspect
ratio d/L 0.26 with surface-piercing vertical bafes of various heights.

In case of surface-piercing wall-mounted bafes, the variation


of damping ratio as a function of relative bafe height for two
different liquid ll depths shows a typical trend. Up to a certain
ratio of bafe height to liquid depth (h/d), the damping ratios
obtained for tank with higher ll depth was found to be more than
those for the low ll depth and thereafter a reverse trend was
marked. This observation could not be explained in the present
study and needs to be further investigated.

S.K. Nayak, K.C. Biswal / Ocean Engineering 108 (2015) 552562

Sloshing attenuation (%)

560

100
80
60
40

Bottom-mounted baffle
Surface-mounted baffle

20

Bottom-mounted submerged block

0
0

0.2

0.4

0.6

0.8

h/d

Fig. 18. Comparison of percentage attenuation of sloshing elevation as a function of


relative object height (h/d) in a tank with ll depth, d/L 0.26 for three different
bafe object conguration.

Fig. 19. Comparison of damping ratio as a function of relative bafe height (h/d) in
a tank with ll depth, d/L 0.26 for bottom-mounted vertical bafes and submerged blocks, w/L 0.33.
Fig. 16. Time decay of sloshing amplitude in free vibration phase in a tank of aspect
ratio d/L 0.52 with surface-piercing vertical bafes of various heights.

150

Bottom mounted
(d/L = 0.26)
Surface-piercing
(d/L = 0.26)
Bottom-mounted
(d/L = 0.52)
Surface-piercing
(d/L = 0.52)

0.08
0.06
0.04
0.02
0.00
0

0.2

0.4

0.6

0.8

h/d
Fig. 17. Comparison of damping ratios of various internal objects, as a function of
relative object height (h/d), in tanks of two different aspect ratios (d/L 0.26 and
d/L 0.26).

100

Amplitude (mm)

Damping ratio ()

0.10

h/d=0.0

h/d=0.2

h/d=0.5

h/d=0.8

50

-50

-100
0

10

15

20

25

30

Time (s)

Irrespective of the ll depths considered in the present study,


for a certain bafe height (h/d) the damping ratio is always more
due to surface-piercing bafe. However, in case of high ll depth,
the slosh damping effect of the surface-piercing bafe is signicantly more than that due to the bottom-mounted bafe of
similar height. Furthermore, the damping effect of both bafe
congurations becomes equal as the bafe heights approach the
height of liquid in the tank. It can be seen that for bafe height
equal to 80% of the height of water in the tank, the damping ratios
for both bafe congurations approach each other. It is quite
obvious, as in both the cases the tank is almost divided into two
separate tanks.
In addition, percentage attenuation of free surface sloshing
elevation, as a function of relative object height for three different
internal objects in the tank with ll depth 0.26 is depicted in
Fig. 18. One can see that the performance of surface-piercing bafe
as the passive slosh damper is better than bottom-mounted bafe.
The slosh attenuation performance of submerged block improves
signicantly compared to the bafes as the h/d ratio goes on
increasing. The same could not be presented for a tank with ll
depth 0.52 as the oscillating water splashed out of the tank due to
lack of sufcient freeboard.
As observed from Fig. 10, submerged blocks are found to reduce
the fundamental sloshing frequencies more as compared to the

Fig. 20. Temporal evolution of the free surface sloshing elevations near the left wall
of the tank with bottom-mounted bafes of varying heights (h/d) under harmonic
excitation with displacement amplitude, x0 5 mm and excitation frequencies,
f (cycles/s) corresponding to their respective fundamental sloshing frequencies.

bottom-mounted vertical bafes of similar heights. However, such


reductions do not translate into an increase in the damping ratio.
This fact can be observed from Fig. 19, where it is seen that
compared to the bottom-mounted vertical bafes, submerged
blocks of equal height with a comparatively more width
(w/L 0.33) have less damping potential.
3.6. Forced vibration response
The tank with different heights of bottom-mounted vertical
bafes was subjected to forced harmonic vibrations with excitation frequencies equal to the fundamental sloshing frequencies of
water of respective tankliquidbafe congurations. The excitation amplitude was kept constant at 5 mm for all the cases. The
bafe height relative to the water depth was varied as h/d 0.0,
0.2, 0.5, and 0.8. The temporal evolutions of free surface sloshing
elevation measured from the bottom of the tank for a tank with ll
depth, d/L 0.26 with bottom-mounted bafe as the internal
object is illustrated in Fig. 20. The plots of the temporal evolution

S.K. Nayak, K.C. Biswal / Ocean Engineering 108 (2015) 552562

150

h/d=0.0
h/d=0.5

Amplitude (mm)

100

h/d=0.2
h/d=0.8

50
0
-50
-100
0

10

15

20

25

30

Time (s)
Fig. 21. Temporal evolution of the free surface sloshing elevations near the left wall
of the tank with surface-piercing bafes of varying heights (h/d) under harmonic
excitation with displacement amplitude, x0 5 mm and excitation frequencies,
f (cycles/s) corresponding to their respective fundamental sloshing frequencies.

Bottom-mounted
Surface-piercing

Amplitude (mm)

60

561

of the internal objects regardless of the congurations. Hydrodynamic damping due to centrally placed internal objects is a
function of relative liquid depth, relative bafe height, and position of the bafe (bottom-mounted or surface piercing). The
damping coefcient increases with relative bafe height. With the
top of the bafe approaching closer to the quiescent free surface,
the damping ratio signicantly increases. This fact is consistent
with the ow dynamics as the convective mass concentrates more
near the free surface of the uid block. Submerged blocks were
found to reduce the fundamental sloshing frequencies more as
compared to the bottom-mounted vertical bafes of similar
heights. However, such reductions did not reect in an increase in
the damping ratio. Surface-piercing bafes give higher damping
than bottom-mounted bafes of similar heights. Overall, bafes
can be used effectively to damp liquid sloshing near resonance
conditions. Many such experimental investigations are necessary
to unravel the complexities involved in sloshing.

Acknowledgements

40
20

The nancial support received from Department of Science and


Technology, Government of India under research Grant (SR/S3/
MERC-0078/2010, 02.02.2011) for this research is greatly
acknowledged.

0
-20
-40
-60
0

10

15

20

Time (s)
Fig. 22. Comparison of temporal evolution of the free surface sloshing elevations
near the left wall of the tank due to bottom-mounted and surface piercing bafe of
equal height (h/d 0.2) in a tank with ll depth, d/L 0.26. Excitation frequencies
are equal to respective fundamental sloshing frequencies and excitation amplitude,
x0 5 mm.

of free surface sloshing elevation for different bafe heights clearly


underlined the damping effect of respective bafes. It was
observed that bafe of height equal to eighty percent of the height
of water in the tank dramatically reduced the sloshing.
A plot on the similar line as above, for surface-piercing vertical
bafes of various heights in a tank with same ll depth of 0.26 is
also depicted in Fig. 21. From a close comparison of Figs. 20 and 21,
one can notice that for a certain height of the bafe, the surfacepiercing vertical bafe is potentially more effective in slosh
damping than bottom-mounted vertical bafe. To make the above
observation further clear, a comparison of temporal evolutions of
sloshing for two congurations of the bafes of equal height,
h/d 0.2 was made on Fig. 22. Fig. 22 clearly shows that surfacepiercing bafe is more effective in slosh damping.

4. Conclusion
The main endeavour of the present study was to investigate the
hydrodynamic damping potential of three different congurations
of centrally installed internal objects perpendicular to the direction of lateral excitation in a rectangular tank partially lled with
water. The objects used are bottom mounted vertical bafes, surface-piercing vertical bafes, and bottom mounted submerged
blocks (w 0.33 L) in a tank with two different ll depths (ratio of
liquid height to tank width in the direction of excitation),
d/L 0.26 and 0.52. The relative height of internal objects (h/d) was
varied as h/d 0.2, 0.4, 0.5, 0.6, and 0.8.
The experimental study showed that the presence of centrally
installed internal objects decreases the fundamental frequency of
liquid and such decrease is monotonic with the increase in height

References
Akyildiz, H., Erdem Unal, N., Hafzullah, Aksoy, 2013. An experimental investigation
of the effects of the ring bafes on liquid sloshing in a rigid cylindrical tank.
Ocean Eng. 59, 190197.
Armenio, V., Rocca, M.L., 1996. On the analysis of sloshing of water in rectangular
containers: numerical study and experimental validation. Ocean Eng. 23 (8),
705739.
Askari, E., Daneshmand, F., 2009. Coupled vibration of a partially uid-lled
cylindrical container with a cylindrical internal body. J. Fluids Struct. 25,
389405.
Askari, E., Daneshmand, F., 2010. Free vibration of an elastic bottom plate of a
partially uid-lled cylindrical container with an internal body. Eur. J. Mech. A/
Solids 29, 6880.
Askari, E., Daneshmand, F., Amabili, M., 2011. Coupled vibrations of a partially uidlled cylindrical container with an internal body including the effect of free
surface waves. J. Fluids Struct. 27, 10491067.
Bass, R.L., Bowles, E.B., Cox, P.A., 1980. Liquid dynamic loads in LNG cargo tanks.
SNAME Trans. 88, 103126.
Belakroum, R., Kadja, M., Mai, T.H., Maalouf, C., 2010. An efcient passive technique
for reducing sloshing in rectangular tanks partially lled with liquid. Mech. Res.
Commun. 37, 341346.
Biswal, K.C., Bhattacharyya, S.K., Sinha, P.K., 2003. Free-vibration analysis of liquidlled tank with bafes. J. Sound Vibr. 259, 177192.
Biswal, K.C., Bhattacharyya, S.K., Sinha, P.K., 2006. Non-linear sloshing in partially
liquid lled containers with bafes. Int. J. Numer. Methods Eng. 68 (3), 317337.
Cho, J.R., Lee, H.W., 2003a. Dynamic analysis of bafed liquid-storage tanks by the
structural-acoustic nite element formulation. J. Sound Vibr. 258, 847866.
Cho, J.R., Lee, H.W., 2004. Numerical study on liquid sloshing in bafed tank by
nonlinear nite element method. Comp. Methods Appl. Mech. Eng. 193,
25812598.
Cho, J.R., Lee, S.Y., 2003b. Dynamic analysis of bafed fuel-storage tanks using the
ALE nite element method. Int. J. Numer. Meth. Fluids 41, 185208.
Cho, J.R., Lee, H.W., Ha, S.Y., 2005. Finite element analysis of resonant sloshing
response in 2-D bafed tank. J. Sound Vibr. 288, 829845.
Choun, Y.S., Yun, C.B., 1996. Sloshing characteristics in rectangular tanks with a
submerged block. Comput. Struc. 61 (3), 401413.
Choun, Y.S., Yun, C.B., 1999. Sloshing analysis of rectangular tanks with a submerged structure by using small-amplitude wave theory. Earthq. Eng. Struct.
Dyn. 28, 763783.
Cleary Jr., W.A., 1982. Subdivision, stability, liability. Mar. Technol. 19, 228244.
Eswaran, M., Saha, U.K., Maity, D., 2009. Effect of bafes on a partially lled cubic
tank: Numerical simulation and experimental validation. Comput. Struct. 87,
198205.
Evans, D.V., McIver, P., 1987. Resonant frequencies in a container with vertical
bafe. J. Fluid Mech. 175, 295307.
Gavrilyuk, I., Lukovsky, I., Trotsenko, Yu, Timokha, A., 2006. Sloshing in a vertical
circular cylindrical tank with an annular bafe. Part 1. Linear fundamental
solutions. J. Eng. Math. 54, 7188.

562

S.K. Nayak, K.C. Biswal / Ocean Engineering 108 (2015) 552562

Gavrilyuk, I., Lukovsky, I., Trotsenko, Yu, Timokha, A., 2007. Sloshing in a vertical
circular cylindrical tank with an annular bafe. Part 2. Nonlinear resonant
waves. J. Eng. Math. 57 (1), 5778.
Gedikli, A., Erguven, M.E., 1999. Seismic analysis of a liquid storage tank with a
bafe. J. Sound Vibr. 223, 141155.
Graham, E.W., Rodriguez, A.M., 1952. The characteristics of fuel motion which
affects airplane dynamics. J. Appl. Mech. 19, 381388.
Hasheminejad, S.M., Aghabeigi, M., 2012. Sloshing characteristics in half-full horizontal elliptical tanks with vertical bafes. Appl. Math. Model. 36, 5771.
Jeyakumaran, R., McIver, P., 1995. Approximations to sloshing frequencies for rectangular tanks with internal structures. J. Eng. Math. 29, 537556.
Laws, E.M., Livesey, J.L., 1978. Flow through screens. Annu. Rev. Fluid Mech. 10,
247266.
Liu, D., Lin, P., 2009. Three-dimensional liquid sloshing in a tank with bafes. Ocean
Eng. 36, 202212.
Maleki, Abbas, Ziyaeifar, Mansour, 2008. Sloshing damping in cylindrical liquid
storage tanks with bafes. J. Sound Vibr. 311, 372385.
Mitra, S., Sinhamahapatra, K.P., 2007. Slosh dynamics of liquid lled containers
with submerged component using pressure based nite element method. J.
Sound Vibr. 304, 361381.
Mitra, S., Khoo, B.C., Kumar, Arun, Sinhamahapatra, K.P., 2010. Inuence of various
internal components on slosh dynamics: a study using nite element method.
Int. J Fluid Mech. Res. 37, 359381.

NASA/SP-8031, 1969. Slosh Suppression, NASA Space Vehicle Design Criteria


(Structures): Slosh Suppression. NASA (Washington, DC), United States.
Nayak, S.K., Biswal, K.C., 2013. Quantication of seismic response of partially lled
rectangular liquid tank with submerged block. J. Earthq. Eng. 17, 10231062.
Panigrahy, P.K., Saha, U.K., Maity, D., 2009. Experimental studies on sloshing
behaviour due to horizontal movement of liquids in bafed tanks. Ocean Eng.
36, 213222.
Porter, R., Evans, D.V., 1995. Complementary approximations to wave scattering by
vertical barriers. J. Fluid Mech. 294, 155180.
Strandberg, L., 1978. Lateral Stability of Road Tankers, VTI Report No. 138A, vol. 1.
Tait, M.J., El Damatty, A.A., Isyumov, N., Siddique, M.R., 2005. Numerical ow
models to simulate tuned liquid dampers (TLD) with slat screens. J. Fluid Struct.
20, 10071023.
Wang, J.D., Lo, S.H., Zhou, D., 2012. Liquid sloshing in rigid cylindrical container
with multiple rigid annular bafes: free vibration. J. Fluids Struct. 34, 138156.
Warnitchai, P., 1998. Modelling of liquid sloshing in rectangular tank with ow
dampening devices. Eng. Struct. 20 (7), 593600.
Watson, E.B.B., Evans, D.V., 1991. Resonant frequencies of a uid container with
internal bodies. J. Eng. Math. 25, 115135.

Вам также может понравиться