Professional Documents
Culture Documents
th
12 Structural Engineering Convention-An International Event
(SEC 2020)
Abstract
The technique of coupled building strategy is more viable to protect the adjacent buildings of dissimilar in characteristics against seismic
hazards. The philosophy of coupled building control in which adjacent buildings are allowed to exerts counter-acting forces one upon another.
This study includes the seismic performance of various Coupled Buildings with respect to Uncoupled Buildings for which two models of
Coupled Building are considered that is, first one is the two shear type adjacent buildings connected in-line with MR dampers and second one is
same as first with taller building is isolated by Resilient-Friction Base Isolator at its foundation. The seismic response analysis of RC coupled
buildings are studied in terms of peak responses subjected to unidirectional excitation due to Kobe 1995 earthquake. The governing equation of
motion of various coupled buildings is solved numerically by Newmark’s step-by-step method. The dynamic behaviour of semiactive MR
damper and R-FBI base isolation system has been predicted by modified Bouc-Wen model and Wen’s model respectively. This study employed
the Lyapunov direct approach as a control algorithm for the stability analysis and design of semiactive MR controller. The responses of various
coupled buildings and uncoupled buildings are simulated through MATLAB computing software. This study outlined that model of Coupled
Building-2 performs more effective in controlling the seismic responses whereas model of Coupled Building-1 perform well not only in reducing
the responses but also avoid pounding from the adjacent buildings. Further, there is significant reduction in responses of taller building isolated
by base isolation system whereas marginal reduction takes place in shorter building which is not isolated.
Keywords: Seismic performance, responses, Uncoupled Building, Coupled Building, Semiactive MR Damper, R-FBI base isolator, Pounding
1. Introduction
The issue of seismic hazard mitigation of adjacent semiactive control has attracted a great deal of attention in
connected buildings has been investigated over the years recent years as it operates with few watts of power and
through various strategies of Coupled Buildings. But still continues to work as a passive device when the control
from the early decades, significant damages of civil algorithm fails [1]. The Spencer et al [2] has proposed a
structures have been caused by seismic hazards which have model to predict the dynamic behaviour of MR damper,
been underscored to rethink about the existing structural referred as phenomenological model, that can effectively
control philosophy. Therefore, the structural community has predicts the response over wide range of operating
started to establish the control systems through which conditions. The comparative study [3] has been carried out
counteracting control force produces and get exerted over for different base isolation systems under uniform shear
the seismically excited coupled buildings. In recent past, the beam in terms of superstructure acceleration and bearing
dissipation technique using dampers is combined with the displacement. Moreover, an updated review [4] on
base isolation technique, called hybrid control and this behaviour of base isolated buildings against seismic
hybrid control strategy not only applied to the individual excitation has been studied through which it is stated that
buildings but also applied effectively to the adjacent sliding base isolation system is quite effective in reducing
buildings. Moreover, seismic response mitigation of the large levels of superstructure acceleration to a maximum
individual building has been studied by many researchers acceleration transmission which is limited to maximum
and same strategy has been applied coupled buildings. A frictional force.
equation of motion (Eq. 1) may be represented in the form where, H () is Heaviside step function, when the function H
of state-space equation is () is greater than zero, command voltage supplied to the
MR driver is maximum (V= Vmax) otherwise, the command
z t = A z t + Bd f d t + Bb f b t + E ug t (2)
voltage set to zero (V= 0).
where, z is the state variable, A is the system matrix For predicting the MR damper force accurately,
composed of structural mass, stiffness and damping, Bd and there are several models are developed, from which a
Bb are the distribution matrices of damper and bearing force modified Bouc-Wen model [2] as shown in Fig. 2 (b) is used
respectively and E is the matrix of excitation force and are to describe the dynamic behaviour of MR damper.
explicitly given as χd
k0 ud
u u O I O c1
z ; z ; A= -1 ;B =
-M -1C d M -1 D p
, c0
u u -M K
O O Bouc-Wen Fd
Bb = -1 E=
M B p -r
Orifice
Piston Seal and
k Bearing
where, [I] and [O] are the identity and null matrices, 1
respectively; vector ż (t) is the state variable of structural
system which contains relative velocity and acceleration
response with respect to ground. Wires to
This study employed the Lyapunov direct approach as a Accumulator Coil MR fluid electromagnet
control algorithm for the stability analysis and design of MR
controller. This theory requires the use of Lyapunov
function, denoted by L ({Z}), which must be a positive Fig. 2 (a). Cross section of Small scale MR damper
definite function of state of the system {Z}. According to
fundamental approach of Lyapunov theory, if the rate of The equation of damper force predicted by this model is
change of Lyapunov function L ({Z }) is negative semi- f d c1 x k1 (ud x0 ) (7)
definite function, the origin is stable in the sense of z u x ( z ) z
( n 1) n
(u x ) z A (u x ) (8)
d d d d
Lyapunov. Thus, in determining the control law, goal is to
choose a control input, which will result in making L as 1
x 0 z c0ud k0 ( ud x ) (9)
negative as possible. In this approach, a Lyapunov function c0 c1
is chosen in the form as where, ud is the damper displacement with respect to ground;
x is an internal pseudo-displacement of damper; z is the
S. M. Dumne .et.al.
hysteretic displacement of damper that accounts for history friction force then base starts to slide and rubber core
dependence of response; k1 is the accumulator stiffness; c0 is deforms as a result resistance setup.
introduced to control the viscous damping of damper at
large velocities, c1 is the viscous damping used to produce
non-linear roll-off in the force-velocity loop at low ub
velocities; k0 is introduced to control the stiffness of damper
at large velocities; x0 is the initial displacement of linear kb
spring k1; 0 is the evolutionary coefficient and , β, n and Ad
are the damper parameters that controls the shape of
hysteresis loop and dot () represents the differentiation with
cb mb
respect to the time. The model parameters are depends on ub
command voltage, c0 ,c1 , 0 which are expressed
respectively as
c c c U
0 0a 0b
(10) Fig. 3 (b). Schematic
Fig. 3 (a). Resilient–friction diagram of RFBI
base isolator
c1 c1a c1bU (11) The Resilient-Friction Base Isolator (R-FBI) describe a
0 oa obU (12) bearing force (fb) is given by
where, U is the output of first order filter and is given by the f b cb vb kbub f r (15)
following equation
where, cb and kb are the damping and stiffness of base
(U V )
U (13) isolator respectively, vb and ub are the velocity and
An Eq. (13) is necessary to model the dynamics involved in displacement of bearing system respectively, fr is the friction
reaching rheological equilibrium and in driving the force produced at the interface of sliding system. The value
electromagnet in the MR damper. A small time lag exists of frictional force (fr) can be obtained by two approaches
between the command signals and damper force due to referred as, conventional model and hysteretic model. In this
inductance in coil of electromagnet. This time lag is study, hysteretic model is considered to compute the
modeled by first-order filter equation between the maximum frictional force in which hysteretic displacement (z) is
commands voltage applied (Vmax) and output of first-order evaluated using Wen’s equation [18]. The frictional force
filter (U) using time constant (1/η) of first order filter. mobilized at the interface of system is
The damper displacement {ud} and velocity ud of fr = fs z (16)
interconnecting in-line semiactive MR damper is given by where, fs is the limiting frictional force which is expressed
the following expressions is by f s= μ Mt g. where, Mt is the total mass of building
including mass of isolation floor, g is the gravitational
ud mni u mni u m ni and
acceleration, μ is the friction coefficient of sliding system
that depends on an instantaneous velocity of base floor is
ud m ni u m ni u m ni (14) expressed by Eq. (17). The friction coefficient (μ) of sliding
system with Teflon-steel bearing [19] is modeled and
where, m and n are the degrees-of-freedom for described by an equation as
superstructure of taller and shorter buildings respectively, i
max exp a vb (17)
is the counter ranging, (i = 1, 2, 3,n) and subscript in
bracket indicate the damper position at the level of where, μmax is the maximum friction coefficient at large
connectivity. velocity of sliding (after leveling off), μmin is the minimum
friction coefficient at small velocity of sliding, ∆μ is the
2.2 Computation of Bearing Force difference of maximum and minimum friction coefficient
respectively at large and small velocity at the interface of
The frictional force mobilized at the interface of system, and its value is assumed to be independent of
sliding system is assumed to be non-linear function of relative velocity (μ=0) at the sliding interface which leads
displacement and velocity at interface. Furthermore, to coulomb-friction idealization, a is the calibration
modeling of bearing force depends on type of base isolation coefficient or constant for a given bearing pressure and
system used. The sliding base isolation system which is interface condition is taken as 20 sec/m and z is the
considered in this study is described below hysteretic displacement evaluated by the Wen’s model [18]
Resilient-Friction Base Isolator satisfying the nonlinear first order differential equation as
The resilient-friction base isolator (R-FBI) as shown in n 1 n
qz b vb zb zb vb zb Avb (18)
Fig. 3(a) provides an isolation effects through parallel action
of friction, damping and restoring spring [5] along with its where, q is the yield displacement of bearing, β and τ are the
schematic diagram as shown in Fig. 3(b). This system is strengthening coefficient of lead plug that controls the shape
designed such that vertical load is carried only by the sliding and size of hysteresis loop, n and A are the integer constant
ring; which are very rigid in vertical direction. As soon as that controls the smoothness of transition from elastic to
ground motion exceeds certain level, lateral load exceeds the plastic state. The parameters β, τ, n and A are so selected so
as to provide a rigid-plastic shape (typical Coulomb-friction 0=0.18 m, and k1=0.0087 kN/m. Further, for sliding base
behaviour). isolator, the parameters [16] are taken as Tb=4s, b= 0.1 and
Table-1. Natural modes and frequencies (ωn) of taller and μmax= 0.04. This model is subjected to unidirectional
shorter Buildings excitation due to a real Kobe N00E 1995 earthquake having
PGA= 0.834g (JMA). The maximum command voltage
supplied to the current driver of MR damper is 6V and base
Natural Natural Frequency (rad/sec) shear (Bsy) is normalized by weight of respective building.
mode
The peak response parameters of interest are, top floor
Taller Building Shorter Building displacement (uf), top floor acceleration (af), normalized
1 12.94 15.98 storey shear (Ssy/W), storey drift (ur) and normalized base
2 38.54 47.40 shear (Bsy/W). The natural modes and natural frequencies of
3 63.28 77.20 uncoupled taller and shorter buildings are shown in Table-I.
4 86.60 104.38
Fig. 4 illustrate the time varying response of top floor
5 107.99 128.00 displacement, top floor acceleration and base shear of taller
building for both Uncoupled and Coupled Buildings under
unidirectional earthquake motion. From the plotted graph, it
The natural period of isolation system is controlled by is observed that significant reduction in responses of taller
selecting an appropriate radius of curvature for concave building of Coupled Building 2 with respect responses of
surface. The stiffness or sliding curvature (kb) and damping uncoupled taller building.
(cb) of R-FBI system are selected so as to obtain the desired
value of isolation period (Tb) and damping ratio (ξb) of the
system using following expression 2 Uncoupled taller Building
Coupled Building-1
1
Mt Coupled Building-2
Tb 2 and
Bsy/W
0
kb -1
cb -2
b
Uncoupled taller Building
2
(19) Coupled Building-1
2 M t b Coupled Building-2
af (g)
0
where, Mt and Wt are the total mass and weight of building -2
including isolation floor, respectively, kb, cb and b are the 10 Uncoupled taller Building
stiffness, damping and natural frequency of isolation bearing 5
Coupled Building-1
uf (cm )
Coupled Building-2
respectively. 0
-5
-10
3. Numerical Study 0 5 10 15 20 25
T (second)
Fig. 4. Time varying top floor response of taller building under Kobe
Earthquake (Vmax= 6V; b= 0.1, Tb= 4s, max= 0.04)
3
3
Further, time varying response of top floor
2
5
the percentage reduction in responses of Coupled Buildings
Normalized Bearing force x10
with respect to response of uncoupled taller and shorter
buildings. The observations implies that significant 0
reduction in responses of taller building of Coupled
Buildings 1and 2 are observed that is, 5.5% and 92.53%
respectively with respect to the uncoupled taller building -5
whereas marginal reduction in responses of shorter building
which is about 8% for both Coupled Building models. -12 -8 -4 0 4 8 12
However, it is also observed that model of Coupled Bearing displacement (cm)
Building-1 is not performed effectively against seismic Fig. 8. Force-deformation behaviour of R-FBI system
0
-1
taller and shorter Buildings. The seismic performance of
-2
Uncoupled shorter Building Coupled Buildings is studied by plotting time varying
2 Coupled Building-1
Coupled Building-2 response and values of peak responses of Uncoupled
af (g )
0
Building and various models of Coupled Buildings. The
-2 concluding remarks are, taller building of Coupled Building-
16 Uncoupled shorter Building
Coupled Building-1
2 which is isolated at its base performs very effective
uf (c m )
8
0
Coupled Building-2
whereas shorter building of same model works marginal in
-8 reducing the seismic responses with respect to the respective
-16
0 5 10 15 20 25
uncoupled taller Building. Similarly, Coupled Building-1
T (second)
Fig.5. Time varying top floor response of shorter building under responded lesser in reducing the seismic responses but
Kobe earthquake (Vmax= 6V; b= 0.1, Tb= 4s, max= 0.04) perform well in avoiding impact from the adjacent
buildings. Hence, Coupled Building-2 involves MR dampers
and base isolation as a result, it perform more effective than
The hysteretic behaviour of top floor in-line MR Coupled Building-1 linked only with MR dampers.
damper used for the coupled building 1 and 2 under Kobe
earthquake having damper and isolator parameters are, V= 6
References