Seismic Response of Liquid Storage Tanks Isolated by Sliding Bearings

Engineering Structures 24 (2002) 909921
www.elsevier.com/locate/engstruct
Seismic response of liquid storage tanks isolated by sliding
bearings
M.K. Shrimali, R.S. Jangid
*
Department of Civil Engineering, Indian Institute of Technology Bombay, Powai, Mumbai 400 076, India
Received 23 July 2001; received in revised form 12 November 2001; accepted 13 December 2001
Abstract
The response of liquid storage tanks isolated by the sliding systems is investigated under two horizontal components of real
earthquake ground motion. The continuous liquid mass is lumped as convective mass, impulsive mass and rigid mass. The corre-
sponding stiffness associated with these lumped masses is calculated depending upon the properties of the tank wall and liquid
mass. The governing equations of motion of the tank with a sliding system are derived and solved by Newmarks step-by-step
method with iterations. The frictional forces mobilized at the interface of the sliding system are assumed to be velocity dependent
and their interaction in two horizontal directions is duly considered. A parametric study is also conducted to study the effects of
important system parameters on the effectiveness of seismic isolation of the liquid storage tanks. The various parameters considered
are (i) the period of isolation (ii) the damping of isolation bearings and (iii) the coefcient of friction of sliding bearings. It has
been found that the bi-directional interaction of frictional forces has noticeable effects and if these effects are ignored then the
sliding base displacements will be underestimated which can be crucial from the design point of view. Further, the dependence of
the friction coefcient on relative velocity of the sliding bearings has no signicant effects on the peak response of the isolated
liquid storage tanks. 2002 Elsevier Science Ltd. All rights reserved.
Keywords: Liquid storage tank; Sliding isolation system; Bi-directional interaction; Seismic response; System parameters
1. Introduction
The integrity of a structure can be protected from the
effects of severe earthquakes either through the concept
of resistance or isolation. In designing a structure by
resistance, it is assumed that the earthquake forces are
directly transmitted to the structure and each member of
the structure is required to resist the maximum possible
forces that may be induced by earthquakes based on
various ductility criteria. In the category of earthquake
isolation, however, one is interested in reducing the peak
response of the structure through implementation of cer-
tain isolation devices between the base and foundation
of the structure which prevents the transmission of earth-
quake acceleration. The main concept in isolation is to
increase the fundamental time period of structural
vibration beyond the energy containing periods of earth-
quake ground motion. The other purpose of an isolation
* Corresponding author.
0141-0296/02/$ - see front matter 2002 Elsevier Science Ltd. All rights reserved.
PII: S0141- 0296( 02) 00009- 3
system is to provide an additional means of energy dissi-
pation, thereby reducing the transmitted acceleration into
the superstructure. This innovative design approach aims
mainly at the isolation of a structure from the supporting
ground, generally in the horizontal direction, in order
to reduce the transmission of the earthquake motion to
the structure.
A variety of isolation devices including elastomeric
bearings (with and without lead core), frictional/sliding
bearings and roller bearings have been developed and
used practically for aseismic design of buildings during
the last 20 yr [1,2]. A signicant amount of recent
research in the base isolation has focused on the use of
frictional elements to concentrate exibility of the struc-
tural system and to add damping to the isolated structure.
The most attractive feature of the frictional base isolation
system is its effectiveness for a wide range of frequency
input. The other advantage of a frictional type system is
that it ensures the maximum acceleration transmissibility
equal to the maximum limiting frictional force. The sim-
plest sliding system device is a pure-friction (P-F) sys-
910 M.K. Shrimali, R.S. Jangid / Engineering Structures 24 (2002) 909921
tem without any restoring force [3]. More advanced
devices involve P-F elements in combination with a
restoring force. The restoring force in the sliding system
reduces the base displacements and brings the system
back to its original position after an earthquake. Some
of the commonly proposed sliding systems with restor-
ing force include the resilient-friction base isolator (R-
FBI) system [4], the friction pendulum system (FPS) [5],
Electricite de France system (EDF) [6] and elliptical rol-
ling rods [7]. The sliding systems perform very well
under a variety of severe earthquake loading and are very
effective in reducing the large levels of the superstruc-
tures acceleration without inducing large bearing dis-
placements [4,8]. In addition, the sliding systems are
also less sensitive to the effects of torsional coupling in
asymmetric base-isolated buildings [9].
There have been several studies investigating the
effectiveness of seismic isolation for buildings but very
few studies are reported for the seismic isolation of
liquid storage tanks which has a vital and strategic use.
Kim and Lee [10] experimentally investigated seismic
performance of liquid storage tanks isolated by elasto-
meric bearings and found that the isolation system is
effective in reducing the dynamic response. Malhotra
[11,12], Chalhoub and Kelly [13] and Shrimali and Jan-
gid [14] studied the seismic response of isolated liquid
storage tanks and observed that isolation is quite effec-
tive in reducing the earthquake forces. It is to be noted
that in all the above studies elastomeric bearings were
used and there is a need to study the performance of
sliding systems for seismic isolation of liquid storage
tanks.
In this paper, the response of liquid storage tanks iso-
lated by the sliding systems under two horizontal compo-
nents of earthquake ground motions is investigated. The
specic objectives of the present study can be summar-
ized as (i) to present a method for earthquake analysis
of liquid storage tanks supported on sliding systems by
duly incorporating the effects of bi-directional interac-
tion and velocity dependence of frictional forces of the
isolation system, (ii) to investigate the effects of bi-
directional interaction of frictional forces under the
earthquakes (by comparing the response of the system
with and without interaction), (iii) to study the effects
of velocity dependence of the friction coefcient of the
sliding system on the peak response of the system and
(iv) to study the inuence of important parameters on
the effectiveness of sliding systems for liquid storage
tanks. The various important parameters considered are:
the period, damping, friction coefcient of the sliding
system and the aspect ratio of the tank.
2. Model of liquid storage tank and the sliding
system
Fig. 1 shows a model of a liquid storage tank sup-
ported on the sliding system. The sliding system is
Fig. 1. Model of a liquid storage tank mounted on a sliding system.
installed between the base and the foundation of the
tank. The tank is modelled by a lumped mass model
suggested by Housner [15] and Rosenblueth and New-
mark [16]. The contained continuous liquid mass is
lumped as convective, impulsive and rigid masses
referred as m
c
, m
i
and m
r
, respectively. The convective
and impulsive masses are connected to the tank wall by
corresponding equivalent springs having stiffness k
c
and
k
i
, respectively. The damping constant of the convective
and impulsive masses are c
c
and c
i
, respectively. The
system has 6 degrees of freedom under bi-directional
earthquake ground motion, 2 degrees of freedom at each
lumped mass in two horizontal x- and y-directions. These
degrees of freedom are denoted by (u
cx
, u
cy
), (u
ix
, u
iy
)
and (u
bx
, u
by
) which denote the absolute displacement of
convective, impulsive and rigid masses in the x- and y-
directions, respectively. The tank model is assumed to
have a deformable cylindrical shell. The parameters of
the tanks considered are liquid height H, radius, R and
average thickness of tank wall, t
h
. The effective masses
are dened in terms of the liquid mass, m from the para-
meters expressed [17] as
Y
c
1.013270.87578S 0.35708S
2
(1)
0.06692S
3
0.00439S
4
Y
i
0.15467 1.21716S0.62839S
2
(2)
0.14434S
3
0.0125S
4
Y
r
0.01599 0.86356S0.30941S
2
(3)
0.04083S
3
where S H/ R is the aspect ratio (i.e. ratio the liquid
height to radius of the tank) and Y
c
, Y
i
, and Y
r
are the
mass ratios dened as
Y
c
m
c
m
(4)
Y
i
m
i
m
(5)
Y
r
m
r
m
(6)
m pR
2
Hr
w
(7)
The natural frequencies of sloshing mass, w
c
and
impulsive mass, w
i
are given by the following
expressions
w
i
P
H
E
r
s
(8)
w
c
1.84
g
R
tanh

1.84
H
R
(9)
where E and r
s
are the modulus of elasticity and density
of tank wall, respectively; g is the acceleration due to
gravity; and P is a dimensionless parameter expressed as
P 0.037085 0.084302S0.05088S
2
(10)
0.012523S
3
0.0012S
4
The sliding system is considered as isotropic (i.e.
same coefcient of friction in two orthogonal directions
of the motion in the horizontal plane) and the restoring
force provided by the sliding systems is considered to
be linear (i.e. proportional to relative displacement). The
additional damping (other than friction) is assumed as
viscous damping. The frictional forces mobilized at slid-
ing system are assumed to be coupled in two horizontal
directions and the friction coefcient is assumed to be
dependent on the relative velocity. The limiting value of
the frictional force, Q
s
to which the sliding system can
be subjected in a particular direction is expressed as
Q
s
mMg (11)
where m is the friction coefcient of the sliding system;
and M (i.e. m
c
+m
i
+m
r
) is the effective mass of the tank
(the mass of tank wall is neglected since it is very small
in comparison to the effective liquid mass).
The coefcient of sliding friction, m at a resultant slid-
ing velocity z
b
x
2
b
y
2
b
, may be approximated from
[18] by the following equation
m m
max
(m) exp (a|z
b
|) (12)
where x
b
and y
b
are the velocities of the sliding system
relative to the ground in the x- and y-directions, respect-
ively; m
max
is the coefcient of friction at large velocity
of sliding (after leveling off); m is the difference
between the friction coefcient at large and zero velocity
of the system; and a is a constant which depends upon
bearing pressure and condition of interface and its value
is taken as 20 s/m.
3. Governing equations of motion
The equations of motion of an isolated liquid storage
tank subjected to earthquake ground motion are
expressed in the matrix form as
[m]{z} [c]{z} [k]{z} {Q} (13)
[m][r]{u
g
}
where {z} {x
c
, x
i
, x
b
, y
c
, y
i
, y
b
}
T
and {Q}
{0, 0, Q
x
, 0, 0, Q
y
}
T
are the relative displacement and
frictional force vector, respectively; x
c
u
cx
u
bx
and
y
c
u
cy
u
by
are the displacements of the convective
mass relative to bearing displacements in the x- and y-
directions, respectively; x
i
u
ix
u
bx
and y
i
u
iy
u
by
are the displacements of the impulsive mass relative to
bearing displacements in the x- and y-directions, respect-
ively; x
b
u
bx
u
gx
and y
b
u
by
u
gy
are the displace-
ments of the bearings relative to the ground in the x- and
y-directions, respectively; [m], [c] and [k] are the mass,
damping and stiffness matrices, respectively; [r] is the
inuence coefcient matrix; {u
g
} {u
gx
, u
gy
}
T
is the
earthquake ground acceleration vector; (u
gx
, u
gy
) and (Q
x
,
Q
y
) are the ground accelerations and the frictional forces
in the x- and y-directions of the system, respectively; and
T denotes the transpose.
4. Criteria for sliding and non-sliding phases
In a non-sliding phase (x
b
y
b
0 and x
b
y
b
0) the resultant of the frictional forces mobilized at the

sliding system interface is less than the limiting frictional
force (i.e. Q
2
x
Q
2
y
Q
s
). The system starts sliding
(x
b
y
b
0 and x
b
y
b
0) as soon as the resultant
of the frictional forces attains the limiting frictional
force. Thus, the sliding of the system takes place if
Q
2
x
Q
2
y
Q
2
s
(14)
Note that Eq. (14) indicates a circular interaction
between the frictional forces mobilized at the interface
of the sliding system as shown in Fig. 2(a). The system
remains in the non-sliding phase inside the interaction
curve. Further, the governing equations of motion in two
orthogonal directions of the structures supported on the
sliding type of isolators are coupled during the sliding
phases due to interaction between the frictional forces.
However, this interaction effect is ignored if the struc-
tural system is modelled as a 2-D system. In such cases
Fig. 2. Interaction and incremental frictional forces in two orthogonal
directions of the sliding system.
the corresponding curve which separates the sliding and
non-sliding phases is a square as shown in Fig. 2(a) by
dashed lines.
Since the frictional forces oppose the motion of the
system, the direction of the sliding of the system with
respect to the x-direction is expressed as
q tan
1
y
b
x
b
(15)
5. Solution of equations of motion
During the non-sliding phase (x
b
y
b
0 and x
b
y
b
0), the rigid mass sticks to the foundation and the
system behaves as two single degrees of freedom (i.e.
convective and impulsive mass) in two orthogonal hori-
zontal directions. These equations can be solved by an
exact or numerical integration technique until the fric-
tional forces mobilized at the sliding surface are less than
the limiting value. As soon as the frictional force attains
the limiting value the sliding phase of motion begins in
which the additional degrees of freedom of the rigid
mass are included in the response analysis. Since the
frictional forces mobilized in the sliding system are
coupled and non-linear functions of the displacement
and velocity of the system are in two orthogonal direc-
tions, as a result, the equations of motion are to be solved
in the incremental form during the sliding phase of
motion. Newmarks method has been chosen for the sol-
ution of governing differential equations, assuming lin-
ear variation of acceleration over the small time interval,
t. The incremental equations in terms of unknown
incremental displacements are expressed as
[k
eff
]{z} {P
eff
} {Q} (16)
where [k
eff
] is the effective stiffness matrix; {z}
{x
c
, x
i
, x
b
, y
c
, y
i
, y
b
}
T
is the incremental dis-
placement vector; {P
eff
} is the effective excitation vec-
tor; {Q} {0, 0, Q
x
, 0, 0, Q
y
}
T
is the incremental
frictional force vector; and Q
x
and Q
y
are the
incremental frictional forces in the x- and y-directions,
respectively.
In order to determine the incremental frictional forces,
consider Fig. 2(b). At time t the frictional forces are at
point A on the interaction curve and move to point B at
time t+t. Therefore, the incremental frictional forces
are expressed as:
Q
x
Q
t+t
s
cos (q
t+t
)Q
t
x
(17)
Q
y
Q
t+t
s
sin (q
t+t
)Q
t
y
(18)
where Q
t t
s
is the limiting frictional force at time t+t
which is obtained by Eq. (11) depending upon the coef-
cient of friction, m expressed by Eq. (12); and the
superscript in the above equations denotes the time.
Since the frictional forces are opposite to the motion
of the system, therefore, the angle q
t t
is expressed in
terms of the relative velocities of the system at time
t+t by
q
t+t
tan
1
y
t+t
b
x
t+t
b
(19)
Substituting for q
t t
from Eqs. (17) and (18), the
incremental frictional forces [3] are expressed as
Q
x
Q
t+t
s
x
t+t
b
(x
t+t
b
)
2
(y
t+t
b
)
2
Q
t
x
(20)
Q
y
Q
t+t
s
y
t+t
b
(x
t+t
b
)
2
(y
t+t
b
)
2
Q
t
y
(21)
In order to solve the incremental matrix Eq. (16), the
incremental frictional forces (Q
x
and Q
y
) should be
known at any time interval. The incremental frictional
forces involve the system velocities at time t+t by Eqs.
(20) and (21) which in turn depend on the incremental
displacements (x
b
and y
b
) at the current time step. As
a result, an iterative procedure is required to obtain the
required incremental solution. The steps of the procedure
considered are as follows
1. Assume Q
x
Q
y
0 for iteration, j 1 in Eqs.
(20) and (21) and solve Eq. (16) for x
b
and y
b
.
2. Calculate the incremental velocity x
b
and y
b
using
the x
b
and y
b
.
3. Calculate the velocities at time t+t using incremental
velocities (i.e. x
t t
b
x
t
b
x
b
and y
t t
b
y
t
b
y
b
) and compute the revised incremental frictional
forces Q
x
and Q
y
from Eqs. (20) and (21), respect-
ively.
4. Iterate further, until the following convergence cri-
teria are satised for both incremental frictional
forces i.e.
(Q
x
)
j+1
||(Q
x
)
j
|
|(Q
x
)
j
|
e (22)
|(Q
y
)
j+1
||(Q
y
)
j
|
|(Q
y
)
j
|
e (23)
where e is a small threshold parameter. The
superscript to the incremental forces denotes the iter-
ation number.
When the convergence criteria is satised, the velocity
of the sliding structure at time t+t is calculated using
incremental velocity. In order to avoid the unbalanced
forces, the acceleration of the system at time t+t is
evaluated directly from the equilibrium of the system in
Eq. (13). At the end of each time step the phase of the
motion of the system should be checked. The response
of the sliding structures is quite sensitive to the time
interval, t and initial conditions at the beginning of
sliding and non-sliding phases. The number of iterations
in each time step is taken as 10 to determine the
incremental frictional forces at the sliding support.
The base shear is a measure of the hydrodynamic
forces generated in the tank which is directly pro-
portional to earthquake forces exerted in the tank. There-
fore, the effectiveness of base isolation is measured in
terms of reduction of the base shear generated in the
tank during earthquake. The base shear is directly pro-
portional to the axial compressive forces induced in the
cylindrical tank wall which causes the buckling [11].
The base shear generated in the x- and y-directions of
the tank are expressed by
F
bx
m
c
u
cx
m
i
u
ix
m
r
u
bx
(24)
F
by
m
c
u
cy
m
i
u
iy
m
r
u
by
(25)
where F
bx
and F
by
are the base shear in the x- and y-
directions of the tank, respectively.
6. Numerical study
The seismic response of liquid storage slender and
broad tanks isolated by the sliding system is investi-
gated. Three types of commonly used sliding base iso-
lation systems i.e. the pure-friction (P-F) system, the
friction pendulum system (FPS) and the resilient-friction
base isolator (R-FBI) are considered for the present
study. The properties of the sliding system (such as stiff-
ness, damping and friction) are kept the same in both
the x- and y-directions of the system. As a result, the
sliding isolation system can be completely dened by
the three parameters namely the period of isolation, T
b
(i.e. T
b
2p/ w
b
; and w
b
k
b
/ M), the damping ratios,
x
b
(i.e. x
b
c
b
/ 2Mw
b
) and the coefcient of friction (i.e.
m
max
and m). However, other tank parameters such as
damping ratio of convective mass (x
c
) and the impulsive
mass (x
i
) are taken as 0.5 and 2%, respectively. The tank
wall considered is made of steel with a modulus of elas-
ticity of E 200 MPa and mass density, r
s
7900
kg/m
3
.
Three real earthquake ground motions are used to
study the response of isolated tanks. The details of these
motions are given in Table 1. The components S90W,
N90E and N90E of Kobe, Loma Prieta and Imperial Val-
ley earthquake ground motions, respectively are applied
in the x-direction of the tank. The other orthogonal
components are applied in the y-direction. The displace-
ment and acceleration spectra of the above earthquake
ground motions are shown in Fig. 3. The seismic
response of the isolated tanks is compared with the cor-
responding response of non-isolated tanks in order to
measure the effectiveness of the sliding systems. The
response quantities of interest in both the x- and y-direc-
tions of the tank are: base shear (F
bx
, F
by
), displacements
Table 1
Properties of earthquake ground motions
Earthquake x-direction y-direction
Component PGA (g) Component PGA (g)
Imperial Valley, 1940 (El-Centro) S90W 0.2144 S00E 0.3486
Loma Prieta, 1989 (Los Gatos Presentation Center) N90E 0.6079 N00E 0.5704
Kobe, 1995 (JMA) N90E 0.6297 N00E 0.8345
Fig. 3. Displacement and acceleration spectra of three earthquake ground motions (a) x-direction components and (b) y-direction components.
of convective mass (x
c
, y
c
), impulsive mass (x
i
, y
i
) and
displacements of sliding system (x
b
, y
b
). For comparative
and detailed parametric study two different types of
tanks, namely the broad and slender tanks are con-
sidered. The properties of these tanks are: (i) aspect ratio
(S) for slender and broad tanks is 1.85 and 0.6, respect-
ively; (ii) the height, H, of water lled in the slender
and broad tanks is 11.3 and 14.6 m, respectively; (iii)
the natural frequencies of convective mass and impulsive
mass for the broad and slender tank are 0.123, 3.944 Hz
and 0.273, 5.963 Hz and (iv) the ratio of tank wall thick-
ness to its radius (t
h
/R) is taken as 0.004 for both the
tanks. Note that the same value of t
h
/ R 0.004 is used
in deriving Eqs. (13) and (10). The base shear of the
tank is normalized by the effective weight of the tank,
W (i.e. W Mg).
The time variation of base shear and relative displace-
ments of the convective mass, impulsive mass and rigid
mass isolated by the FPS system for the slender tank
is shown in Figs. 4 and 5 for the x- and y-directions,
respectively under Imperial Valley, 1940 earthquake
ground motion. The isolation parameters considered are
T
b
2 s, m
max
0.05 and m 0. It is observed that
there is signicant reduction in the base shear and impul-
sive displacement of the tank implying that the sliding
system is quite effective in reducing the earthquake
response of the tanks. On the other hand, the sloshing
displacement of the tank remains the same for both iso-
lated and non-isolated conditions. This is due to the fact
that the period of sloshing mass is 3.66 s which is well
separated from the period of the isolation systems hence
isolation has no signicant effect on sloshing displace-
ment. The peak bearing displacements are 6.19 and 5.12
cm in the x- and y-directions of the tank, respectively
which are considerably less in magnitude to accommo-
date the sliding system.
The peak response of non-isolated and isolated tanks
under three earthquake ground motions are shown in
Tables 2, 3 and 4 for P-F, FPS and R-FBI systems,
respectively. It is observed that due to isolation, peak
Fig. 4. Time variation of response of slender tank in the x-direction
isolated by the FPS system under Imperial Valley, 1940 earthquake
(T
b
2 s, m
max
0.05 and m 0).
base shear and peak impulsive displacement are signi-
cantly reduced. Moreover, the peak impulsive displace-
ment is quite small in comparison to sloshing and base
displacement which is further reduced due to isolation
of the tank. The percentage reductions of peak base shear
with interaction effect in slender and broad tanks isolated
by P-F, FPS and R-FBI due to Imperial Valley are 79.84,
79.84, 88.39 and 70, 76.47, 84.41, respectively. Simi-
larly due to the Loma Prieta earthquake the percentage
reductions of peak base shear are 90.63, 58.33, 83.99
and 83.37, 45.81, 80.45, respectively. The percentage
reductions due to the Kobe earthquake are 88.79, 58.89,
90.6 and 87.21, 55.17, 89.18 respectively. The above
reduction in base shear indicates that the sliding systems
are relatively more effective for slender tanks in com-
parison to broad tanks. Further, there is no signicant
change in the sloshing displacement of both slender and
broad tanks due to Imperial Valley, 1940 earthquake but
the other earthquake ground motions have some inu-
ence on it. It is to be noted that there is signicant slosh-
ing displacement in the slender tank in the y-direction
for both isolated and non-isolated conditions under
Loma Prieta, 1989 earthquake motion. This is due to the
fact that at the sloshing period of slender tank (i.e. 3.66
s) this earthquake has a maximum displacement spectra
ordinate (see Fig. 3), as a result, the sloshing displace-
Fig. 5. Time variation of response of slender tank in the y-direction
isolated by the FPS system under Imperial Valley, 1940 earthquake
(T
b
2 s, m
max
0.05 and m 0).
ment is amplied. The maximum bearing displacement
is found to of the order of 50 cm for a slender tank
isolated by the R-FBI system under Loma Prieta, 1989
earthquake motion.
In Tables 24, the response of isolated tanks is also
compared with the corresponding response without con-
sidering the interaction of friction forces of the sliding
systems (i.e. earthquake response of tanks in two hori-
zontal directions obtained as a 2-D idealization). It is
observed that the base shear for both slender and broad
tanks is not signicantly inuenced by the interaction
of frictional forces for all sliding systems. Further, the
convective and impulsive displacements due to the inter-
action effect are slightly decreased in the slender tank
while in the broad tank it is increased. The bearing dis-
placement is found to be increased due to the interaction
of the friction forces of the sliding system. This is due
to the fact that when the interaction is taken into con-
sideration the system starts sliding at a relatively lower
value of the frictional forces mobilized in the sliding
system (refer to the sliding Eq. (14)), as a result, the
sliding displacements in the isolated system are
increased. Thus, if the interaction of the frictional forces
of the sliding system is ignored then the sliding displace-
ments will be underestimated which can be crucial from
the design point of view of the isolation system.
Table 2
Peak seismic response of liquid storage tanks isolated by P-F system (T
b
, x
b
0 and m
max
0.1 and m 0)
Earthquake/tank condition Response quantity
F
bx
/W x
c
(cm) x
i
(cm) x
b
(cm) F
by
/W y
c
(cm) y
i
(cm) y
b
(cm)
Slender tank
Imperial Valley, Non-isolated 0.319 41.54 0.39 - 0.491 23.71 0.74 -
1940
Isolated (with interaction) 0.100 39.00 0.22 3.80 0.100 22.73 0.22 4.02
Isolated (no interaction) 0.100 39.84 0.26 2.33 0.100 24.87 0.29 3.21
Loma Prieta, 1989 Non-isolated 0.771 36.14 0.96 - 1.056 142.8 1.53 -
Kobe, 1995 Non-isolated 0.594 52.05 0.66 - 0.883 37.85 0.91 -
Broad tank
1940
Table 3
Peak seismic response of liquid storage tanks isolated by FPS (T
b
2 s, x
b
0, m
max
0.05 and m 0)
F
bx
/W x
c
(cm) x
i
(cm) x
b
(cm) F
by
/W y
c
(cm) y
i
(cm) y
b
(cm)
Slender tank
1940
Broad tank
1940
Table 4
Peak seismic response of liquid storage tanks isolated by R-FBI system (T
b
4 s, x
b
0.1, m
max
0.04 and m 0)
F
bx
/W x
c
(cm) x
i
(cm) x
b
(cm) F
by
/W y
c
(cm) y
i
(cm) y
b
(cm)
Slender tank
1940
Broad tank
1940
Table 5
Effects of m on the peak seismic response of liquid storage tanks isolated by FPS system (T
b
2 s, x
b
0 and m
max
0.05)
Earthquake F
bx
/W x
c
(cm) x
i
(cm) x
b
(cm) F
by
/W y
c
(cm) y
i
(cm) y
b
(cm) m
m
max
(%)
Slender tank
Imperial Valley, 1940 0.0 0.103 47.60 0.11 6.20 0.098 29.41 0.11 5.12
10 0.103 47.90 0.11 6.16 0.096 29.58 0.11 5.06
20 0.102 48.12 0.09 6.14 0.095 29.74 0.10 5.01
30 0.102 48.33 0.08 6.12 0.095 29.87 0.10 5.04
Loma Prieta, 1989 0.0 0.159 52.67 0.15 14.20 0.440 207.90 0.40 40.46
10 0.159 52.65 0.15 14.20 0.440 207.89 0.40 40.44
20 0.159 52.65 0.15 14.20 0.440 207.88 0.40 40.42
30 0.159 52.64 0.15 14.19 0.439 207.87 0.40 40.40
Kobe, 1995 0.0 0.195 73.05 0.17 14.69 0.363 64.27 0.31 31.12
10 0.194 72.98 0.16 14.68 0.359 64.23 0.31 31.13
20 0.193 72.95 0.16 14.68 0.357 64.19 0.31 31.09
30 0.191 72.91 0.16 14.64 0.356 64.14 0.31 31.04
Broad tank
Imperial Valley, 1940 0.0 0.076 52.27 0.39 3.04 0.078 43.76 0.42 3.39
10 0.076 52.28 0.38 3.09 0.076 43.92 0.38 3.44
20 0.076 52.29 0.36 3.15 0.076 44.06 0.35 3.50
30 0.077 52.33 0.34 3.23 0.077 44.11 0.32 3.55
Loma Prieta, 1989 0.0 0.182 24.95 0.56 13.74 0.330 64.13 0.98 30.42
10 0.182 24.99 0.56 13.75 0.331 64.18 0.99 30.45
20 0.182 25.02 0.56 13.75 0.331 64.23 0.99 30.48
30 0.182 25.03 0.56 13.76 0.331 64.27 0.99 30.50
Kobe, 1995 0.0 0.177 25.59 0.61 15.07 0.347 42.10 1.12 30.97
10 0.178 25.58 0.62 15.09 0.347 42.12 1.12 30.99
20 0.178 25.55 0.62 15.11 0.346 42.14 1.12 31.01
30 0.178 25.51 0.63 15.12 0.346 42.13 1.12 31.02
6.1. Effects of velocity dependent friction coefcient
The friction coefcient of various sliding isolation
systems is typically dependent on the relative velocity
at the sliding interface and it will be interesting to study
these effects on the peak response of isolated liquid stor-
age tanks (refer to Eq. (12) for the expression of velocity
dependent friction coefcient). In Table 5, the peak seis-
mic response of slender and broad tanks isolated with
the FPS system is shown for different values of m (i.e.
0, 0.1m
max
, 0.2m
max
and 0.3m
max
). Note that m 0
denotes that the friction coefcient of the sliding iso-
lation system is independent of the velocity at the sliding
interface (i.e. Coulomb-friction idealization). It is
observed from Table 5 that the dependence of the fric-
tion coefcient on the relative sliding velocity has no
noticeable effects on the peak response of liquid storage
tanks isolated by sliding systems for all the earthquake
ground motions. A similar trend in the results was also
observed for the tanks isolated by the P-F and R-FBI
systems. Thus, the effects of dependence of friction coef-
cient on the sliding velocity may be ignored for
determining the peak response of isolated tanks. These
effects are similar to that observed by Fan and Ahmadi
[8] for buildings isolated by the sliding systems.
Fig. 6. Effects of isolation period on the resultant seismic response
of liquid storage tanks isolated by the FPS system (m
max
0.05 and
m 0).
6.2. Effects of isolation period
So far the effectiveness of a sliding system for isolat-
ing the liquid storage tanks is investigated for the xed
parameters of the sliding systems. However, it will be
interesting to study the inuence of isolator parameters
(such as period, damping and friction coefcient) on the
beahviour of isolated tanks. The variation of resultant
base shear, F
bz
(i.e. F
2
bx
F
2
by
), sloshing displacement,
z
c
and bearing displacement, z
b
for both slender and
broad tanks is shown against the period of isolation, T
b
in Figs. 6 and 7 for the FPS and R-FBI systems, respect-
ively. The gures indicate that the base shear decreases
with the increase of exibility of isolation systems. This
is due to the fact that with an increase of isolation period
the system becomes more exible and, as a result, trans-
mits less earthquake acceleration into the tanks leading
to a reduction in the base shear. However, the effect of
isolation period is not found to be signicant on the
sloshing displacement in both tanks. The base displace-
ment increases with the increase of isolation period for
Imperial Valley, 1940 and Loma Prieta, 1989 earth-
quakes whereas it decreases for Kobe, 1995 earthquake
ground motion. This trend is similar to the displacement
spectra of these ground motions as shown in Fig. 3.
Fig. 7. Effects of isolation period on the resultant seismic response
of liquid storage tanks isolated by the R-FBI system (x
b
0.1,
m
max
0.04 and m 0).
Thus, the effectiveness of sliding systems for tanks
increases with the increased exibility.
6.3. Effects of coefcient of friction
The effects of coefcient of friction, m
max
on the
resultant response of isolated tanks are shown in Figs.
8 and 9 for the FPS and R-FBI systems, respectively. It
is observed that the peak resultant base shear decreases
initially and after that it increases with the increase of
friction coefcient. This indicates that there exists an
optimum value of coefcient of friction for which the
base shear in the tank attains the minimum value. On
the other hand, the bearing displacement for both tanks
decreases with the increase of coefcient of friction.
This is expected as the friction coefcient of the sliding
system increases it transmit higher earthquake acceler-
ation (due to increase of limiting frictional force) into the
tanks resulting in higher base shear. Since the limiting
frictional force is increased with the increase of friction
coefcient, as a result, less sliding takes place in the
isolation system leading to reduction in the bearing dis-
placement. Further, the sloshing displacement mildly
decreases with the increase of friction coefcient of the
sliding system.
Fig. 8. Effects of friction coefcient on the peak resultant response
of liquid storage tanks isolated by the FPS system (T
b
2 s and
m 0).
Fig. 9. Effects of friction coefcient on the peak resultant response
of liquid storage tanks isolated by the R-FBI system (T
b
4 s, x
b
0.1 and m 0).

6.4. Effects of isolation damping
The peak resultant seismic response of slender and
broad tanks isolated by the R-FBI system is plotted
against the isolation damping, x
b
in Fig. 10. The gure
indicates that the base shear initially decreases with
increases of damping and attains the minimum value
thereafter, it increases with increase of isolation damp-
ing. This implies that there exists an optimum value of
isolation damping at which there is a minimum value of
base shear. On the other hand, the sloshing displacement
and the base displacement decreases with the increase
of isolation damping. Thus, increase in the isolation
damping can reduce the displacement response of the
tank but under certain conditions the high damping may
produce more earthquake forces into the system.
The effects of friction coefcient and damping on the
peak base shear of the tank are similar to the well-known
phenomenon that high damping in isolation systems can
reduce the displacements but it may transmit higher
earthquake forces [19,20]. It happens due to the fact that
the high bearing damping transmits more acceleration
into the system for the earthquake energy at the higher
frequencies. Thus, one should select the optimum value
of coefcient of friction and damping in the sliding sys-
tems for effective design of isolated liquid storage tanks.
Fig. 10. Effects of damping ratio of the R-FBI system on the peak
resultant seismic response of the liquid storage tanks (T
b
4 s,
m
max
0.04 and m 0).
These optimum parameters will be dependent on the
properties of the tank, isolation system and the character-
istics of the earthquake ground motion.
7. Conclusions
The response of liquid storage tanks supported on the
sliding systems subjected to bi-directional earthquake
ground motions is investigated. The bi-directional inter-
action and velocity dependence of frictional forces of
the sliding system in two orthogonal directions are duly
incorporated in the governing equations of motion of the
system. The response of the isolated system under the
recorded earthquake ground motions is analysed to
investigate the performance of sliding systems for seis-
mic isolation of tanks. From the trends of the results of
the present study the following conclusions may be
drawn:
1. The sliding systems are found to be quite effective in
reducing the base shear and impulsive displacement
of the liquid storage tanks. However, the sloshing dis-
placement of the tank is not greatly inuenced due to
isolation by the sliding systems.
2. The dependence of the friction coefcient on the rela-
tive velocity of the system has no noticeable effects
on the peak response of the isolated liquid storage
tanks. Therefore, these effects may be ignored in
determining the peak response of the system.
3. The bi-directional interaction of frictional forces has
signicant effects on the response of isolated tanks.
If these effects are ignored then the sliding displace-
ment will be underestimated which can be crucial
from the design point of view.
4. The effectiveness of isolation systems for tanks
increases with the increase of the exibility of the
sliding systems.
5. There exists an optimum value of friction coefcient
and damping for which the base shear in the liquid
storage tanks attains the minimum value under earth-
quake ground motion. However, the sloshing and
bearing displacement decreases with the increase in
the friction coefcient and damping of the sliding sys-
tem.
References
[1] Buckle IG, Mayes RL. Seismic isolation: history, application and
performancea world overview. Earthquake Spectra
1990;6(2):161202.
[2] Kelly JM. Seismic base isolation: review and bibliography. Soil
Dynamics and Earthquake Engineering 1986;5:20216.
[3] Jangid RS. Response of pure-friction sliding structures to bi-
directional harmonic ground motion. Engineering Structures
1997;19(2):97104.
[4] Mostaghel N, Khodaverdian M. Dynamics of resilient-friction
base isolator (R-FBI). Earthquake Engineering and Structural
Dynamics 1987;15(3):37990.
[5] Zayas VA, Low SS, Mahin SA. Simple pendulum technique for
achieving seismic isolation. Earthquake Spectra 1990;6(2):317
33.
[6] Gueraud R, Noel-leroux J-P, Livolant M, Michalopoulos AP.
Seismic isolation using sliding elastomer bearing pads. Nuclear
Engineering and Design 1985;84:36377.
[7] Jangid RS, Londhe YB. Effectiveness of elliptical rolling rods
for base-isolation. Journal of Structural Engineering, ASCE
1998;124(4):46972.
[8] Fan FG, Ahmadi G. Multi-storey base-isolated buildings under a
harmonic ground motionpart II: a comparison of various sys-
tems. Nuclear Engineering and Design 1990;123:1426.
[9] Jangid RS, Datta TK. Performance of base isolation systems for
asymmetric building to random excitation. Engineering Structures
1995;17(6):44354.
[10] Kim NS, Lee DG. Pseudo-dynamic test for evaluation of seismic
performance of base-isolated liquid storage tanks. Engineering
Structures 1995;17(3):198208.
[11] Malhotra PK. New methods for seismic isolation of liquid-storage
tanks. Earthquake Engineering and Structural Dynamics
1997;26:83947.
[12] Malhotra PK. Method for seismic base isolation of liquid storage
tanks. Journal of Structural Engineering, ASCE
1997;123(1):1136.
[13] Chalhoub MS, Kelly JM. Theoretical and experimental studies
of cylindrical water tanks in base isolated structures. Report no.
UCB/EERC-88/07, Berkeley, 1988.
[14] Shrimali MK, Jangid RS. Earthquake response of base-isolated
liquid storage tanks. Proceedings of the National Symposium on
Advances in Structural Dynamics and Design, VETOMAC-1,
Chennai, 2001:45966.
[15] Housner GW. Dynamic behaviour of water tanks. Bulletin of
Seismological Society of America 1963;53:3817.
[16] Rosenblueth E, Newmark NM. Fundamentals of earthquake
engineering. Englewood Cliffs, NJ: Prentice Hall, 1971.
[17] Haroun MA. Vibration studies and test of liquid storage tanks.
Earthquake Engineering and Structural Dynamics 1983;11:179
206.
[18] Constantinou MC, Mokha AS, Reinhorn AM. Teon bearing in
base isolation II: modeling. Journal of Structural Engineering,
ASCE 1990;116(2):45574.
[19] Inaudi J, Kelly JM. Optimum damping in linear isolation systems.
Earthquake Engineering and Structural Dynamics 1993;22:583
98.
[20] Jangid RS. Optimum damping in a non-linear base isolation sys-
tem. Journal of Sound and Vibration 1996;189(4):47787.

Seismic Response of Liquid Storage Tanks Isolated by Sliding Bearings

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Seismic Response of Liquid Storage Tanks Isolated by Sliding Bearings

Uploaded by

Copyright:

Available Formats

Engineering Structures 24 (2002) 909921

0) the resultant of the frictional forces mobilized at the

0.1 and m 0).

You might also like