Vous êtes sur la page 1sur 14

Seismic responses of

secondary systems in baseisolated structures


Fa-Gung Fan and Goodarz Ahmadi
Department of Mechamcal and Aeronautical Engineering, Clarkx'on Umversity, Potsdam,
NY 136799, USA
(Recetved November 1989)

This paper presents the results of a series of numerical simulation


studies on seismic responses of secondary systems in base-isolated
structures including equipment-structure interactions. A three-storey
building is used as the primary structure, while the equipment is
modelled as a single-degree-of-freedom linear system. A number of
base isolation systems such as the Laminated Rubber Bearing, the
Pure Friction, the Resilient-Friction, and the Electricite de France
system are considered. Several earthquake records including the
N00W component of El Centro 1940, the $16E components of
Pacoima Dam 1971, and the N90W component of Mexico City 1985
earthquakes are used as ground accelerations. Acceleration and
deflection response spectra of the secondary system under different
conditions are evaluated and the effects of equipment- structure interactions are studied. It is shown that the use of base isolation provides considerable protection for structural contents. However, peak
responses of the secondary systems vary substantially depending on
the base isolation system used. Among the base isolation systems
considered, the Laminated Rubber Bearing system appears to be
remarkably effective in reducing peak responses of secondary
systems under a variety of conditions.

Keywords: base-isolated structures, seismic response

The goal of aseismic design is to protect the structure,


as well as the structural content. During an earthquake,
a fixed-base shear frame structure filters the generally
broad-band ground excitation into narrow-band
responses at various elevations. The equipment conrained in the structure is then excited by these floor
motions and interacts with the primary structure. Due to
tuning or near tuning, certain secondary systems may be
damaged significantly even in a low intensity earthquake. There have been numerous cases where nonstructural components experienced major damage while
the structure itself survived the earthquake attack. In
earthquake resisting design of structures containing
critical and/or expensive equipment such as nuclear
power plants, hospitals, computer centres, and telecommunication buildings, protection of secondary systems is
as important as the structure itself.
In contrast to the conventional strengthening methodology, an alternative aseismic design strategy is to isolate
the structure from the ground excitations during earthquakes. In the past two decades, several base isolation
systems for stiff and compact structures were proposed,

and their performances were studied. Kelly 1,2 provided


extensive reviews on earlier and recent developments on
the subject. Numerous reported numerical simulations
and shaking table experiments have shown that the peak
transmitted accelerations and the deflections generated
in the structures are dramatically reduced by using properly designed base isolation systems~'2. However, the
consequences of using various base isolation devices on
responses of secondary systems are not fully understood. In particular, the presence of frictional elements
m the base isolation may have adverse effects on certain
secondary and non-structural components. Recent
studies by Su et al. 3 and Fan et al. 4.5 showed that high
frequency components are generated in the acceleration
responses of a structure with a frictional base isolation
system. These high frequency motions could be damaging to stiff equipment and non-structural attachments.
Numerous studies on responses of secondary systems
m fixed-base structures have been performed. The current design procedure for aseismic design of light secondary systems favours the floor response spectra
approach 6,7. Several theoretical and numerical

0141-0296/92/01035-14
1992 Butterworth-Hememann Ltd

Eng. Struct. 1992, Vol. 14, No 1 35

Setsmic responses of base-molated structures. F,-G. Fan and G Ahmad/

schemes for evaluating peak responses of secondary


systems Including primary -secondary interactions have
been developed 8-2 A number of shaking table experiments have also been conducted to verify these
procedures 2j-26 A state-of-the-art review on response
analysis of secondary systems has been provided by
Chen and Soong 27 On the other hand, studies on secondary system responses for base-isolated structures are
rather scarce. A series of shaking table tests on the
influence of the rubber bearing base isolation systems on
earthquake responses of light secondary equipment was
reported by Kelly 2j 28 Buckle et al. 29 studied floor
response spectra for a fast breeder reactor, and Kelly
and Tsai ~ described responses of secondary systems
which are attached to various floors of a five storey
base-isolated building. Certain results concerning
responses of non-structural components in base-isolated
structures were reported by Ikonomou 3~ and Wu and
Seidenstlcker 32. Recently, Fan and Ahmad133 conducted
a comparatwe study of the floor response spectra for a
multi-storey budding with various frictional and rubber
bearing type base isolation devices It was shown that
the features of these floor spectra vary significantly
depending on the type of base isolation systems used.
In this work, seismic responses of secondary systems
m base-isolated structures under various earthquake
ground excitations are studied The interactions between
the secondary system and the primary structure are
included in the analysis. Several base isolation systems
including the Laminated Rubber Bearing, the Pure Friction, the Resilient-Friction, and the Electricite de France
systems are considered. The N00W component of El
Centro 1940, the S16E component of Pacoima Dam
1971, and the N90W component of Mexico City 1985

,
/
/
/
/
/
/

earthquakes are used as ground excitations. The peak


acceleration and deflection responses of different secondary systems under various conditions are evaluated and
the results are presented as response spectra curves_ Particular attention IS given to analysing the effects of the
mass ratio of the secondary system The influences of
the frequency and damping ratio of the secondary system
are also studied The results are compared with the floor
response spectra and the significance of the primarysecondary interactions and mass ratio for near tuned
conditions are discussed.

Base isolation systems


This section provides brief descriptions of the base isolation systems considered in this study. Laminated Rubber
Bearing (LRB) is the most common base isolation
system 12,34 This system has been used in a number of
structures worldwide including several buildings m
Cahforma and Utah. A laminated rubber bearing consists of alternating layers of rubber and steel with the
rubber being vulcanized to the steel plates. The bearing
is rather flexible in the horizontal direction but quite stiff
in the vertical direction The LRB provides protection
against earthquakes by shifting the fundamental frequency of vibration to a much lower value and away
from the energy containing range of the earthquake
ground motion. Figure l a shows a schematic diagram
for the mechanical behaviour of the Laminated Rubber
Bearing system. This diagram also represents the GERB
helical spring and visco-damplng base Isolation system
described elsewhere 35.
A Pure-Friction (P-F) isolator offers resistance to
motion and dissipates energy only through horizontal

,[

coZl

"1

/
/
/
/
/
/

Xg

.I

'

B-

v I

c7'

"V

'

/
/

xg

co -r]
xe

1 Schematic d=agrams of isolation systems (a), LRB system, (b), P-F system, (c), R-FBI
system, (d), EDF system

Figure

36

Eng. Struct. 1992, Vol. 14, No 1

Seismic responses of base-isolated structures: F.-G. Fan and G. A h m a d i

friction force 36. A schematic diagram for the P-F base


isolator is shown in Figure lb. Use of a layer of sand in
the foundation of a building which behaves essentially as
a Pure-Friction base isolation system has been
described 37. A Resilient-Friction Base Isolation system
(R-FBI) is composed of several layers of Teflon coated
friction plates with a contral core of rubber or steel reinforced laminated rubber 3s. The core provides the
resilient force for the system, while energy is dissipated
by the friction forces. The mechanical behavlour of the
R-FBI system is illustrated m the schematic diagram
shown in Figure lc. The Alexisismon base isolator of
Ikonomou 39 can also be represented by this diagram.
A system which was designed mainly for base isolation of nuclear power plants in regions of high seismicity
was developed under the auspices of Electricite de
France (EDF) 4 An EDF base isolator unit consists of
a laminated (steel-reinforced) neoprene pad topped by a
lead-bronze plate which is m frictional contact with a
steel plate anchored to the base raft of the structure. The
behaviour of the EDF base isolator is shown schematically in Figure ld.

Governing equations
A three-storey building with a secondary system
attached to its top floor as shown in Figure 2 is used as
the structural and equipment model throughout this
study. In this section, the equations governing the
motions of the structure and various base isolation
systems are presented. The criteria used for motion transitions for different frictional base isolators are also
described.

R-FBI, P-F, and LRB systems For frictional base isolation systems the equations of motion for sliding and
nonsliding conditions are different. For the R-FBI
system, the governing equations during the sliding phase
are given as

Orb

c~xl + k~xl -- .Xg


mb

(1)

X3 ~ C

m3

. 1 1 3 r d floor

3, k3
m2

X2

x1

2nd floor

o2, k2
ml

--1 1st floor

/ Cl, kl

~/7
///////~

Figure2

rnb ~

- m,(2~,%z + oJ2z)[e]
= (2~'oOot + ~O2oS+ I~g s~n(s)

Baseraft
Isolation system

x#
V//////
~//////////::////'/~Foundation

Schemat=cdiagram of primary and secondarysystems

~b

ClX 1

+_kjxi)\ [m]
mb

'+2~'(

[1 ]

(2)

l+m~zm3/+to~(m'~zl+m3/
c3

--

kj

k~

X2 "{- C 3 X3 - - - - X 2 + - - X 3
m3
m3
m3
m3

(3)

Here s is the relative displacement between the base of


the structure and the ground, x, is the relative
displacements of each floor with respect to the base, z
is the relative deflection of the secondary system, Jog is
the horizontal ground acceleration, # is the friction coefficient, g is the acceleration of gravity, and [m], [c]
and [k] represent mass, damping, and stiffness
matrices, respectively. The natural circular frequency of
the bearing w0, and its effective damping ratio ~'0 are
defined as
2~.oO~o_ M
Co

w02 - /Co
M

(4)

where M is the total mass of the structure, Co and ko are


the damping and the horizontal stiffness of the bearing,
respectively. The parameter % is the ratio of base mass
to the total mass of structure, and is defined as

Ub =

2~'oWoS + w~s + I~g sgn(~)


+

[m]lx} + [c]lxl + [k]lx}

mb
--

M = m b + m~ E

m,

(5)

t=l

where m, is the mass of the i th floor, and m b is the


mass of the base. In equation (2), [ 1 1 = [ 1 1 1 } r,
[ e ] = I 0 0 1 1 r, and S'~h(S) is a function which Is equal
to + 1 when s is positive, and - 1 when s is negative.
The value of s ~ ( 0 ) is undefined but it is bounded
between - 1 and +1.
The deflection of the secondary system is governed by
equation (3) wRh ~,, ~', being the natural frequency and
the effective damping ratio of the secondary system,
respectively. Clearly, equations ( 1 ) - ( 3 ) are coupled
and must be solved slmultaneusly. The parameter controlling the interaction between the primary and the
secondary system is the mass ratio "/(=ms/m3). This
parameter has significant effects on the acceleration and
deflecUon responses of equipment. For the nonsliding
phase
s = s = 0

(6)

and the deflection of the secondary system is governed


by equation (3) which is coupled to equation (2). The
criteria for transition from the nonsliding phase to the
sliding phase and vice versa, and for reversing the

Eng. Struct. 1992, Vol. 14, No 1

37

Seismic responses o f base-isolated structures: F.-G. Fan and G. A h m a d t

sliding direction without sticking are summarized m the


following. Addaional details have been described 4's_
Under the assumption of the Coulumb friction law, the
nonsliding phase continues as long as

system are described in this section. During the sliding


phase, the equations of motion of the EDF isolator are
given as
A

(ll)

2~'0oJ0Y + w6Y = Izg sgn(, - y)


]OtbJCg + W 2 S -- ( C t X 1 +

klxT)lMI < #g

(7)

As soon as the condition


IOtbX e + W o S -- (ClX I +

klxl)lMI

= Izg

1(

~tg

OtbXg + o~2S

C~XlMkl)X~

clxl + klXl

Olb

mb

x~

(12)

(8)

is met the sliding phase starts and equatton (1) governs


the slip displacement of the isolator At the moment of
transition, the shding direction is given by

s'~(0) - -

2~-0o~0y + ~2y

where y is the deflection of the rubber and s is the total


relative displacement (sum of y and the slip displacement
in the friction plate) between the base raft of the structure and the ground. In the nonshding phase, the EDF
system behaves as a LRB system. The equations of
motion of the isolator then become

(9)
s = ~

That is, for g~'n(0)= +1 (or ---1), the direction of


sliding of the base mass is in positive (or negative)
direcuon.
The sliding of the structure in one direction may be
terminated by a transition to a stick condition, or by a
reversal of the sliding direction. During the sliding
phase, whenever s becomes zero, the criterion

s =y

(13)

Y + 2~'0oJ0y + 6%~2v clxl + klxl _


0/b

Jog

(14)

mb

The deflections of the structure for both phases of


motion are governed by

[m]lXl + [c][x] + [k]lxl


M

Isl _> 2 # g - -

(10)

- m,(2~,o~s:~ +

~02z)lel

mb

= (2~'o~ooy + oJ~y
is checked for determining the subsequent behavlour
The inequality given by equation (10) is the necessary
and sufficient condition for the reversal of the sliding
direcuon 4'5. That is, if the inequality holds, equations
( 1 ) - (3) continue to govern the subsequent sliding phase
motion with sg-"h(s) taking the sign of S at the moment of
transition. If the criterion given by equation (10) is not
satisfied, the structure will stick to its foundation and
equations (2), (3) and (6) apply. Under exceptional conditlons, S and ~- might become simultaneously zero. Such
a case corresponds to a smooth transition to the stick
condition. Equations ( 1 ) - (10) are quite general and also
cover the motions of structures with Pure Friction and
Laminated Rubber Bearing base isolation systems.
When the rubber element is absent (e.g. ~'0 = co = 0,
o~o = k0 = 0), these equations govern the motion of a
structure with a P-F base isolation system. Similarly,
when the frictional dement is absent (i.e. # = 0), equations (1)-(3) reduce to the governing equation of
motion for a structure with a LRB base isolation system.
In this latter case, no stick duration occurs and equation
(6) and the criteria for motion transitions, equations
(7) - (10) become irrelevant
It should be emphasized that here the LRB is modelled
as a linear spring-viscous damper system. Under high
strain or for high damping rubbe&~significant strainhardening occurs which has to be accounted for. Such
behaviour may be appropriately represented by a
nonlinear hysteretic model.

EDF system
The governing equations of motion for a three-storey
building with an EDF base isolator and a secondary

38

Eng. Struct. 1992, Vol. 14, No 1

Olb

ClX 1 ~_ klXl.~[m]{1
mb

(15)

The criterion for the nonsllding phase to continue is:


IOtb(~; +5g) -- (Cl.~l + kjxl)/MI < tzg

(16)

AS soon as the condition


Itb('~ + "~s) -- (CtXl + ktxt)/MI = I~g

(17)

is satisfied, sliding commences and equations (11), (12)


and (15) coupled with equation (3) govern the subsequent motion in the sliding phase. The direction of the
sliding is then given by
s'~(0) =

/zg
-1 ( tbff + Xg)

clXaMklX~)

(18)

at the moment of transiaon.


It should be noted that a nonsliding phase is bound to
occur at the end of each sliding phase for the EDF
system. This is because when the base raft changes the
direction of its motion, the massless bearing has to stick
to it due to the absence of inertial resistance.

Structural

model and solution techniques

In this section, the structural model is described, and the


procedure for numerical evaluation is discussed. As
noted previously, the three-storey building shown in
Figure 2 is used as the primary structure. It is assumed
that the masses of different floors and the base raft are
identical (ml = m2 = m3 = mb = m) and the stiffness

Seismic responses of base-iso/ated structures: F.-G. Fan and G. A h m a d i


Table I

Values of p a r a m e t e r s used f o r v a n o u s base ~solators

Base isolation s y s t e m

Natural
penod
T O (s)

Damping
ratio
~'o

Friction
coeff,cmnt
~(~,~ h,)

Larminated Rubber Bearing


(LRB)
Pure-Frict=on (P-F)
Resdent-Fr.ction (R-FBI)
Electnc=te de France (EDF)

2
4
1

0.08
0.08
0.08

0 1
0 05
0.2

Responses to E! Centro earthquake

and damping of columns of various storeys are also


equal. The corresponding mass, damping, and stiffness
matrices are given as
100

[m] = m

I :1 I
I
0 1

[c] = c

00

[k] = k

2
-1

-1
2

-1

-1

-1

-1

-1

0 1
-1 ,

1
,

(19)

where k and c are the stiffness and the damping coefficient, respectively. For a structure with a fundamental
natural period of 0.3 s and the corresponding damping
ratio of 0.02, the circular natural frequencies and the
damping ratios for various modes become,
01 = 20.944 r a d s -~,

o2 = 58.685 r a d s -~

o3 = 84.802 rad s- l
~'1 =

0.02,

~'2 = 0.056,

(20)
h = 0.081

and the peak deflection, z I , = , responses of the secondary system under a variety of different conditions are
evaluated. The resulting peak responses presented as
acceleration and deflection response spectra are discussed in the following sections.

(21)

The corresponding natural frequencies become 3.33,


9.34, and 13.5 Hz. This primary structural model was
used earlier 4'5.
The secondary system is modelled as a single-degreeof-freedom oscillator with a natural frequency of
f~ = % / 2 x and a damping ratio L. A value of 0.01 for
L is assumed throughout this study except in the section where the effects of variation in damping ratio are
studied. The values of natural periods T0, damping
ratios ~'0, and friction coefficients # for the isolators
studied are summarized in Table 1. Based on the fourthorder Runge-Kutta scheme, a double precision
FORTRAN routine for numerical evaluation of the
equations of motion is developed. A time step of
At = 0.0005 s is used in the continuous phases of motion
away from the transitions. A successive time-stepcutting procedure is used to approach each of these transitions with an accuracy of 10 -1 m s-: in evaluating
the stick-slip criteria and an accuracy of 10 -l m s -~ in
evaluation the zeros of ~ (or s - ) and y for the EDF
system).

The N00W component of the E1 Centro 1940 earthquake


accelerogram is used as the ground acceleration in this
section, and peak responses of the secondary system for
different base-isolated structures and the fixed-base one
are evaluated. A secondary system with a mass ratio of
7 = 0.01 is assumed to be attached to the top floor of the
primary structure. The first 20 s of the response time
histories are used in these analyses. As noted
elsewhere 27, the peak response of a secondary system
may occur at much later time when complared with the
time of peak ground excitation due to the beating
phenomenon. To verify that the first 20 s of the response
time histories are sufficient for determining the peak
responses of secondary systems for E1 Centro 1940
earthquake, longer time durations for several cases were
used. For all the cases studied, peak responses of the
secondary system occurred before 20 s.
Figure 3 shows the acceleration response spectra for
the secondary system attached to a structure with different base isolation systems. The results for a fixedbase structure are also shown for comparison. It is
observed that the acceleration response spectra for the
fixed-base structure contains a sharp peak with an
amplitude of about 10 g at the frequency of about
3.33 Hz. This peak is due to tuning of the secondary
system to the fundamental frequency of the primary
structure. For frequencies more than 5 Hz, the
amplitude of the response spectrum varies between 1 to
> g. Figure 3 shows that the use of base isolation
systems eliminates the resonance peak observed for the
fixed-base structure, and reduces the peak spectral
amplitude by a factor of more than five. The exception
is the P-F system which leads to peak acceleration
responses comparable to that for the fixed-base structure
for the frequency range of 5 < f~ < 12 Hz.

F-B

i.

~:',:,',,

,, ,~, ,

"',

P-F

I h ~",: ; ' \ /

'~.. j , , \

F-I,'L.'/r'
F I / ; ,, / ~ , . "'~

F,,t;/:..,

"

"

~-,

\ "v,...
~~

.,_.._.

E':: ~

-.

", ". R-FBI

.....

-. i - . ............

10 -1
o

Responses analyses
Using the formulations described in previous sections,
the peak absolute acceleration, (:g+ ~' + Jr3 + ~) Im,~,

10

15

20

Frequency, fs (Hz)

Figure 3

Peak acceleration response of secondary systems for El


Centro 1940 earthquake

Eng. S t r u c t .

1992,

Vol. 14, No 1

39

Setsmtc responses o f base-isolated structures" F.-G. Fan and G. Ahmadt

Figure 3 shows that the trend of variations of peak acceleranon of a light secondary system for the R-FBI
system is similar to that of the P-F system; however, its
magnitude is lower, by a factor of approximately two
Except for a sharp peak at f, = 1 Hz which corresponds
to the natural frequency of the Isolator, the spectral
amplitude for the EDF system varies between 0.3 to 1 g.
It is also noted that the acceleration spectrum for the
LRB system has a peak at f, = 0.5 Hz corresponding to
the natural frequency of the bearing used Away from
this peak, the spectral amphtude ~s about 0 15 to 0.3 g
which is the lowest among the isolation systems considered
Figure 4 shows the equipment deflection response
spectra for various base isolated structures and the fixedbase one, It is observed that the response spectrum of a
secondary system which Is attached to the fixed-base
structure contains a resonance peak at ~ts natural frequency of 3.33 Hz. Beyond this frequency the spectral
amplitude decreased rapidly Figure 4 also shows that,
the use of a base isolation system eliminates the
resonance peak and significantly reduces the peak
deflections of the secondary systems with f > 2 Hz. In
particular, for f~ > 2 Hz, the LRB system leads to the
lowest equipment deflection response spectra among the
isolators considered. For f, < 2 Hz, the LRB and the
EDF systems generate peak responses which are higher
than those for the fixed-base structure due to the tuning
with the natural frequency of the isolator. It should be
noted, that the natural frequencies of important nonstructural components are usually higher than those of
the structure. Thus, the use of base isolation systems
significantly reduces the peak deflection generated in the
secondary systems for the practical range of frequencies
Based on the results presented in this section, ~t may
be concluded that, for the E1 Centro earthquake ground
excitation, use of base isolation systems provides considerable protection for the secondary systems and the
structural contents. Among the isolators considered, the
LRB system leads to the lowest peak acceleration and
peak deflection responses in the secondary systems. The
frictional systems, however, are less effective m certain
high frequency ranges

~'~

10-1 ~

10-7

F-B

LR

"--

10-3
0

I
1

)
2

/R-FBI

.-

I
3

EDF

J
4

)
5

J
6

....

"~l
7

I
8

~ 9

10

Frequency, fs (Hz)
Figure 4 Peak d e f l e c t i o n r e s p o n s e s o f s e c o n d a r y s y s t e m s f o r El
Centro 1940 earthquake

40

Eng. Struct. 1992, Vol. 14, No 1

Effects of mass ratio for E1 Centro earthquake


The influence of mass ratxo, ~, on acceleranon
responses of the secondary system for the E1 Centro
1940 earthquake are shown in F~gure 5 Fzgure 5a
displays the equipment acceleration response spectra for
several values of 3' for the fixed-base structure It is
observed that for 3, = 0 0, which corresponds to the
non-interaction case, the spectrum has a resonance peak
at the tuning frequency of f,---3_33 Hz with an
amplitude of about 18 g. Away from th~s peak, the spectral amplitude decreases rapidly and approaches zero for
small frequencies, and to a value of about 2 g for large
.f, For ",/= 0.001, it is observed that the acceleranon
spectrum coincides with that for 3' = 0 0 except for the
tuned or near tuned conditions. The peak spectral
amplitude for "y = 0 001 is lower than that for "7 = 0.0
by about 2 g. When 3' IS increased to a value of 0 01, the
response spectrum of the secondary system deviates
significantly from the corresponding floor response
spectrum for the frequency range near the fundamental
frequency of the primary structure The peak spectral
amplitude for "7 = 0.01
is about
10 g.
For
8 < f < 11 Hz, small differences between the
response spectra for 3, = 0 0 and 3' = 0.01 are also
noticed. This is due to tuning of the secondary system
to the frequency of the second mode of vibration of the
primary structure. Away from these two frequency
ranges, the peak acceleration response is identical to that
of the non-interaction case. As 1' increases to 0.1, it is
observed that the peak at fs = 3 33 Hz drops to about
4 g. Furthermore, for large frequencies the peak acceleranon experienced by the secondary system becomes
slightly larger than the amphtude of the corresponding
floor response spectrum
Figure 5b shows the eqmpment acceleration response
spectra for several values of 3' for a structure w~th the
LRB base isolation system. It is observed that the LRB
system, generally, leads to low-amplitude acceleration
spectra Furthermore, the spectra for 3' = 0 0, 0.001,
and 0.01 are almost identical for the entire range o f f
considered. Fo~ 2/= 0 1, the acceleration responses
somewhat deviate from those for 3' = 0 0 a t f = 0.5 Hz
and f, --- 6 Hz For other frequency ranges, the peak
acceleranon responses become indistinguishable from
the floor response spectra
The acceleration response spectra for the secondary
system for various values of-), in a structure with the P-F
system are shown in Fzgure 5c. It is observed that these
spectra contain a number of peaks in the frequency range
of 3 to 12 Hz. The peak acceleranon response
approaches zero for small f , and tends to a value of
about 0 6 g for largef~ For 3' = 0.001, the spectrum is
almost identical to that for 3/-- 0 0 for the entire range
of frequency considered. For "y = 0.01, certain differences, particularly for the sharp peaks are observed.
As q, increases to 0.1, the shape of the spectrum
becomes qmte different from the corresponding floor
response spectrum. The interaction between the primary
and the secondary system eliminates the sharp peaks,
and significantly reduces the amplitude of the response
spectrum For large J,, the spectral amplitudes for
3~ = 0.1 are slightly lower than those for the noninteraction ('y--O) case
Figure 5d shows the peak acceleration responses of

Smsmic responses of base-isolated structures: F.-G. Fan and G. A h m a d i

18-

7=00
....................... -t = o.ool
.........
3, = O.Ol
. . . . . . . . . . . . 3-=01

14
43

i
1 0

(3-

2
0

0
a

10

15

20

Frequency, f, (Hz)

10

15

20

Frequency, fs (Hz)

5,
4
t-

r-

3
4
2

O-

0
C

10
15
Frequency, fs (Hz)

20

10

15

Frequency, f, (Hz)

20

0
e

5
10
15
Frequency, f, (Hz)

20

Figure 5 Peak acceleration responses of secondary systems for various mass ratms for El Centre 1940 earthquake. (a), F-B, (b), LRB,
(c), P-F, (d), R-FBI, (e), EDF

the secondary system for several mass ratios for a structure with the R-FBI system. Features of these spectra are
similar to those noted for the P-F system. The spectra
for 7 = 0.0, 0.001 and 0.01 contain a number of peaks
in the frequency range of 3 to 15 Hz, and the interactions between the primary and the secondary systems
reduce the spectral amplitudes. For 7 = 0.1, the sharp
peaks are eliminated due to the interaction effects. A
comparison of Figures 5e and 5d shows that the
magnitude of peak acceleration experienced by a nonstructural component in a structure with the R-FBI
system is much lower than that for the P-F System.
The equipment acceleraUon response spectra for a
structure with the EDF base isolation system are shown
in Figure 5e. The spectra contain two sharp peaks, one
is at the natural frequency of the isolator ~ = 1 Hz),
and the other is at a frequency of about 6 Hz. As mass
ratio increases, the magnitude of the maximum acceleration near the sharp peaks decreases. Away from these
two peaks, the effects of primary-secondary interactions are relatively insignificant.
The influences of 3" on peak deflection responses of
the secondary system for various base isolation systems
and for the fixed-base structure are also studied and the
results are presented in Figure 6. Figure 6a shows the
equipment deflection response spectra for the fixed-base
structure. It is observed that the floor response spectra
(i.e., 3' = 0 . 0 ) contains a resonance peak at f~ =
3.33 Hz corresponding to the natural frequency of the
primary structure. Figure 6a also shows that the deflection response spectra for 3' = 0.001 and 0.01 are almost
~dentical to the non-interaction one, except for the frequency range near 3.33 Hz and 9.5 Hz. (The second fre-

quency corresponds to the second mode of vibration of


the primary structure). For a large mass ratio such as
3' = 0.1, the resonance peak at f~ = 3 33 Hz is totally
ehminated.
Figure 6b shows the equipment deflection response
spectra for a structure with the LRB system. It is observed that the spectra have resonance peaks with
magnitudes of about 1 m or more forf~ = 0.5 Hz. This
is due to the tuning between the secondary system and
the base-isolated primary structure which essentially
vibrates at the natural frequency of the LRB isolator.
Beyond this frequency, amplitudes of spectra decrease
rapidly. Figure 6b also shows that response spectra for
finite mass ratios are slightly different from the corresponding floor response spectra (3" = 0.0) for f~ near
0.5 Hz and 6 Hz. Away from these two frequency
ranges, the peak responses are essentially identical for
different values of 3".
The deflection response spectra for secondary systems
with different 3, for a structure which is base-isolated
with the P-F system are displayed in Figure 6c. It is
observed that the spectrum for 3, = 0.001 is identical to
the one 3' = 0.0 for the entire range of frequency considered. As 3" increases to 0.01, effects of interaction
become noticeable for f~ > 2 Hz. Figure 6c shows that
for 3' = 0 . 1 , the equipment peak deflections are
significantly reduced due to the primary-secondary
interactions.
Figure 6d presents the equipment deflection response
spectra for a base-isolated structure with the R-FBI
system. While their amplitudes are lower, the general
features of these response spectra are similar to those
shown in Figure 6c for the P-F system. Peak deflection

Eng. Struct.

1 9 9 2 , Vol. 14, No 1

41

Seismic responses of base-tsolated structures: F.-G. Fan and G. Ahmadt

-r=oo
...................... -~ = 0.001
........
"r = 0.01
...........
-r=Ol

10 -1

10-1 1
"o

-~

10 -2

10-2 I

10 -3
0

!10_1
C

I
1

I
2

=
3

I
4

I
5

" ' " ' : ' ~

10-7

= = I
6 7 8

10-:

=
9 10

Frequency,r, (Hz)

"'" ~
5 6 7

I
8

9 10

Frequency, fs (Hz)

10 -1

==
g.

Frequency, f, (Hz)

!1o_,

10-2 I

10-3/
0

10 -3
10

Frequency, fs (Hz)

lO-2I
lO -31

10

9 10

Frequency, f, (Hz)

Figure 6

Peak deflect=on responses of secondary systems for various mass ratios for El Centro 1940 earthquake (a), F-B, (b), LRB, (c),
P-F, (d), R-FBI, (e), EDF

responses decrease as f~ increases, and the interactions


between the secondary and the primary systems become
significant for 3" greater than 0.01
The deflection response spectra for secondary systems
in a structure with the EDF base isolation system are
shown m Figure 6e There are two peaks, one at
f~ = 1 Hz corresponding to the natural frequency of the
isolator, and the other at f~ = 6 Hz. It is observed that
the peak deflecuon responses for 3' -< 0.01 are almost
identical to those for 3' = 0.0 except for frequencies
close to the two peaks. For 3" = 0.1 , the influences of
interaction are clearly noticeable forf~ close to 3.33 Hz
and for f~ > 4 Hz. Furthermore, the peak at the frequency of 6 Hz is completely eliminated due to primarysecondary interactions.
The results presented in this section show that, for a
secondary system attached to a fixed-base structure, the
primary tuning occurs when the natural frequency of the
secondary system coincides with the fundamental
natural frequency of the structure. While, for a secondary system attached to a structure with a LRB or an
EDF base isolation system, tuning occurs when the
natural frequency of the secondary system is close to that
of the isolator. For a light secondary system, the effects
of primary-secondary interaction are not significant
except for the tuned and the near tuned conditions. For
mass ratios greater than 1%, the interaction effects
reduce the peak response of the secondary system for the
entire range of frequencies; the amounts of reduction are
most significant near the tuning frequencies. For frictional systems, such as the P-F and the R-FBI systems,
the nature of tuning is not qmte clear. The friction force
generates high frequency components in the acceleration

42

Eng. Struct. 1992, Vol 14, No 1

transmitted to various floors, and the vibration energy is


scattered over a broad range of frequencies. The effect
of mass ratio in reducing peak responses is similar to
those for the fixed-base structure and the LRB-type base
isolators. The results also indicate that the floor
response spectra approach which neglects the primarysecondary interactions provides conservative estimates
for peak responses of equipment and non-structural
components.

Effects of damping of secondary system


To study the effects of damping coefficient, peak
acceleration and peak deflection responses of the secondary system for various values of ~', for the base~solated structure with the LRB and the R-FBI systems
are evaluated. The mass ratio is kept fixed at 3' = 0.01,
and the N00W component of the El Centro 1940 earthquake is used as the seismic excitation The resulting
response spectra are described in this section and are
compared w~th those for the fixed-base structure.
Figure 7a shows the acceleration response spectra of
the secondary system for the fixed-base structure. It is
observed that as ~', increases, peak acceleration
responses decrease. The effect of increasing damping is
most significant close to the tuning frequency of
3.33 Hz. At the tuned condition, the peak acceleration
response is reduced by a factor of five when ~', increases from 0.01 to 0.2. For f, > 15 Hz, the damping
has no observable effect on the peak acceleration
response of the secondary system.
The equipment acceleration response spectra for the
LRB system for various values of ~', are displayed in

Seismic responses of base-isolated structures: F.-G. Fan and G. Ahmadi


101

03

~', = 0.01
.......................

r-

_o

~, = 0.02
~', = 0.05

............

8ca

~', = 0.1

(g
EL

~', =0.2

10 -1

10

15

20

15

20

Frequency, fs (Hz)

=AlI

10

A
03
C

.o

I1.

EL

10 -1

b
Figure 7

I0

15

10 -1

2o

Frequency, f, (Hz)

0
c

10
Frequency, f, (Hz)

Effects of damping ratios of secondary systems on the peak acceleration responses. (a), F-B, (b), LRB, (c), R-FBI

Figure 7b. This figure shows that amplitudes of


resonance peaks (at f~ = 0.5 Hz and f~ = 6 Hz) are
reduced by increasing J',. As ~', increases from 0.01 to
0.2, the peak response at f, = 0.5 Hz decreases by a
factor of about four. The effect of variation in damping
becomes negligible for f, > 12 Hz.
Figure 7c shows the effects of L on the acceleration
response spectra of light non-structural components for
the R-FBI system. The damping of the secondary system
affects the acceleration responses significantly for
f, < 15 Hz. As ~', increases, amplitudes of the sharp
peaks are reduced and they become smoother. For
> 15 Hz, the effects of J', become insignificant.
The equipment deflection spectra for various ~', for
the base-isolated structure with the LRB and the R-FBI
systems, as well as, for the fixed-based structure are
displayed in Figure 8. It is observed that, as ~',
increases, the deflection spectra become smoother, and
their magnitude decrease for the entire range off, considered, In particular, the effect of damping is most
significant at the tuned condition.
The results presented shows that peak responses of
secondary systems can be significantly reduced by
increasing their damping coefficient. Therefore, use of
additional damping mechanism for protecting sensitive

secondary systems may be recommended. The presented


results also show that, for a secondary system attached
to a fixed-base structure or a structure with a LRB
system, effects of damping ~'s are important only at
tuned conditions. However, for a secondary system
attached to a structure with a R-FBI system, effects of
~-, are equally significant for a rather broad range of f~.

Responses to Pacoima Dam earthquake


In this section, the S 16E component of the Pacoima Dam
1971 earthquake accclerogram with a peak acceleration
of 1.17 g is used as the ground acceleration. A mass
ratio of -y = 0.01 is considered, and the first 20 s of the
response time histories of secondary systems are used
for evaluating the peak responses. Time histories of
longer duration were also used to verify that the first 20 s
is sufficient to determining the peak responses of secondary systems.
Figure 9 shows the peak accelerauons experienced by
different secondary systems which are attached to the
third floor of various base-isolated structures and the
fixed-base one. It is observed that the spectra for the
fixed-base structure contains an extremely high
resonance peak (about 50 g) at f~ = 3.33 Hz. Beyond

Eng. Struct.

1992,

Vol.

14, No 1

43

Seismic responses o f base-tsolated structures: F.-G. Fan and G. Ahmad/

~', = 0 01

.......................

............

.~ 10_1
r-

{'~ = 0.02

_, = 0.05

_~

~'s = 0 1

-o

~'~ = 0 2

a_

10 -2

10 -3
0

10 -1

~.~ 10 -1
t-

r-

_o

x N'%

'4--

10 -3
0

10-3

Ftgure 8

10

Frequency, f, (Hz)

\ ", . . o o

g. lO-2

~o_ 10 -2

"-g'-8

10

Frequency, fs (Hz)

Effects of damping ratios of secondary systems on the peak deflect=on responses (a), F-B, (b), LRB, (c), R-FBI

this frequency, the amphtude of the spectrum varies


between 4 to 7 g. Figure 9 also shows that the resonance
peak observed for the fixed-base structure are entirely
eliminated by the use of a base isolation system. Peak
accelerations of the secondary systems for various baseisolated structures are about 10 to 25 times lower than
that for the fixed-base structure. It also observed that the
equipment spectral amplitudes for the P-F system vary
between 2 to 5 g in the frequency range of 4 to 10 Hz.
The corresponding peak accelerations for the R-FBI
system are lower than that for the P-F system by a factor
of about two. For f, > 15Hz, the acceleration
responses for the P-F system reduce to about 0.9 g, and
those for the R-FBI system reduce to about 0.6 g. The
LRB and the EDF systems, generally, lead to the lowest
peak responses of about 0.4 to 1 g for f, > 2 Hz for
secondary systems.
Figure 10 shows the eqmpment deflection response
spectra for various base-isolated structures. It is noticed
that the floor response spectrum of the fixed-base structure has a resonance peak at a frequency of about 3.33 Hz
with an amplitude of about 0.8 m. This figure also
shows that all base isolation systems considered here are
highly effective in eliminating the resonance peak at the
fundamental natural frequency of the structure. For fs

44

10

Frequency, fs (Hz)

Eng. Struct. 1992, Vol

14, No 1

greater than 1.8 Hz, peak deflections of the secondary


system are significantly reduced for the base-isolated
structure. In the frequency range off, > 2 Hz, the LRB
system leads to the lowest spectral amplitudes for secondary systems. For frequencies lower than 2 Hz, the RFBI system generates the lowest peak deflecuons. Comparison of Figures 4 and 10, shows that the equipment
deflection spectra corresponding to frictional base isolation systems do not change to an appreciable extent.
Thus, it appears that for structures with a frictional base
isolator, the maximum deflections generated in secondary systems are relauvely insensitive to an increase in
the intensity of ground excitation.
The results presented for a h]gh intensity ground excitation show that the base isolation systems are highly
effective in protecting the structural content and the nonstructural components from severe earthquakes. In this
case, all base isolation systems reduce the peak
responses of secondary systems over the entire range of
frequencies greater than 2 Hz.

Effects of mass ratio for Pacotma Dam earthquake


In this secUon, the effects of mass ratio -y on peak
responses of secondary systems for the Pacoima Dam

Seismic responses o f base-isolated s t r u c t u r e s : F.-G. Fan and G. A h m a d i


10

.:,f,

F-B

.... ,jr

"' " "

I,~:,: / ~, ~,

,,

",

- ~,

',----,

"-./
L,, ~" .............
" ................

tO

i ~.,]-'-~'/

~- -,-- ~ " " - " . -

__

~"

LRB

/
EDF

O.

I0 -I

10

15

20

Frequency, fs (Hz)

Figure 9 Peak acceleratmn responses of secondary systems for


Pacoima Dam 1971 earthquake

.,F B
A
t-

\ \'L--,'~",.-~-.-, \ - ~ - ,

.9

,
"\

'-

'(L "'",

a.

10 -2

/\

LRB

~- -

EDF "~ ~10 -3

10

Frequency, fs (Hz)

F~gure 10 Peak deflection responses of secondary systems for


Pacmma Dam 1971 earthquake

80

i:t

3,=00
70

.......................

60
O_

(J

3, = 0.001
3, = 0.01

............

50

1971 earthquake are studied. The resulting peak acceleration and deflection responses are presented in Figures
11 and 12, respectively. Figure 11a displays the equipment acceleration response spectra for different values
of 3' for the fixed-base structure. It is observed that the
non-interaction (3" = 0.0) spectrum has a very high
resonance peak (about 70 g) at f, = 3.33 Hz. The spectrum for 3, = 0.001 is essentially identical to the one for
3" = 0.0 except for the tuned and the near tuned conditions. For this mass ratio, the primary-secondary
interaction reduces the magnitude of the resonance peak
to about 60 g. When 3, increases to 0.01, the amplitude
of the resonance peak drops to about 30 g. For other frequency ranges, no significant effect is observed except
for f, close to 9.5 Hz which corresponds to the second
mode of vibration of the primary structure. By further
increasing 3' to 0.1, it is observed that the two tuning
peaks (at f~ = 3 33 Hz and 9.5 Hz) are totally
eliminated.
Figure 11b shows the eqmpment acceleration response
spectra for several values of 3' for the R-FBI system. It
is observed that, for light secondary systems, these spectra contain a number of sharp peaks for 3 < f, <
15 Hz. As 3" increases, the primary-secondary interactions reduce the peak acceleration responses over a wide
range of frequencies. For 3' = 0.1, this figure also shows
that the sharp peaks are essentially eliminated.
Figure 12a shows the effects of mass ratio on the
secondary system deflection response spectra for the
fixed-base structure. It is observed that the noninteraction case contains a resonance peak at
f, = 3.33 Hz with an amplitude of about 1 m. Furthermore, for a hght secondary system, the effect of interaction is not s|gnificant except for the tuned and the near
tuned frequencies. For 3' = 0.1, the resonance peak is
eliminated and the spectral magnitudes are shghtly
reduced for .f, > 2 Hz.
The effects of 3" on the equipment deflection response
spectra for a structure with the R-FBI system are shown
m Ftgure 12b. This figure shows that peak deflection
responses decrease with an increase in 3' for f, > 2 Hz.

3,=0.1

40
30

O.

20
10
0

J1'-f

0
a

10
Frequency, fs (Hz)

15

20

10

15

20

Frequency, fs (Hz)

Figure 11 Peak acceleration responses of secondary systems for various mass ratios for Pacoima Dam 1 971 earthquake (a), F-B, (b),
R-FBI

Eng. Struct.

1992,

Vol.

14, No 1

45

Seismic responses of base-isolated structures: F.-G. Fan and G. Ahmad/

1t

3' = 0.0
....................... 3'=0001
A

-?=001

. . . . . . . . . . . .

r"

10 -1

=o.1

_8

R-"0

~_ 10-2

10-1
a.

10 -2

I
1

I
2

I
3

I
4

J
5

I
6

J
8

I
9

10 -3
10

I
1

I
2

I
3

I
4

fs (Hz)

Frequency,

F~gure 12

]
7

I
5

I
6

I
7

I
8

I
9

10

Frequency, fs (Hz)

Peak deflect=on responses of secondary systems for vanous mass rat=os for Pacolma Dam 1971 earthquake. (a), F-B, (b), R-FBI

However, the effects of primary-secondary mteractaon


are insignificant for a light secondary system with
3' -< 0.01.

Responses to Mexico Ctty earthquake


Peak responses of a secondary system with a mass ratio
of 0.01 attached to a base-isolated structure subjected to
the Mexico City 1985 earthquake ground excitation are
studied in this section. The peak ground acceleration of
the N90W component of Mexico City earthquake was
about 0.17 g and it contained considerable energy at the
frequency of about 0.5 Hz. The first 80 s of responses
are used for evaluating the equipment acceleration and
deflection response spectra.
Figure 13 shows the equipment acceleration response
spectra for the three storey structure with various base
isolation systems and for the fixed-base one. It is observed that, for the fixed-base structure, the response
spectrum contains two sharp peaks, one at f~ --- 0.5 Hz
with an amplitude of about 1.7 g, and the other at
fs = 3.3 Hz with an amphtude of about 1.2 g. Away

from these two peaks, the spectral amplitudes of the


secondary system are about 0.3 g. Under this earthquake

ground motion, the acceleration response spectrum for


the LRB system shows a high resonance peak at
f~ = 0.5 Hz with an amplitude of about 10 g. Beyond
this frequency, the peak acceleration remains a constant
at about 0 8 g. The high peak observed in the spectrum
is due to the peculiar frequency content of Mexico City
1985 earthquake which coincides with the natural frequency of the LRB base isolator used and the assumption
of linear behaviour of the isolator. As noted before, at
high strain, the stiffness and damping of the rubber increases which could reduce the peak responses
significantly.
Figure 13 also shows that the acceleraUon floor spectra for a structure with the P-F, and the R-FBI systems
are sharply peaked at the frequency of about 6 Hz. For
4 < ~ < 15 Hz, the spectral amplitudes for the P-F and
the R-FBI systems are higher than those for the fixedbase structure. Under this ground excitation, the EDF
system seems to perform reasonably well when compared with other base isolation systems. While ~t

I"

10

i
/k..--LRB

Iq

iI

I~I ~ L R B
:./P-F

e-

_o

--~ 10-1
',', V,7\,~

tO

F-B

'~-I, ~,.-

P-F
O-

"-':-:::::::2 ::Z

10-2

",
R-FBI

EDF
10 -1

10

15

10-3

20

I
1

Frequency, f, (Hz)

Figure 13 Peak acceleratmn r e s p o n s e s of s e c o n d a r y s y s t e m s for


M e x i c o City 1 9 8 5 earthquake

46

~_~

Eng. Struct. 1992, Vol. 14, No 1

I
2

..... .~;-~,,

\ _ ...~
I
3

I--"
4

~ ~-~-~..
I
5

I
6

3
7

""
10

Frequency, fs (Hz)

Figure 14 Peak d e f l e c t i o n r e s p o n s e s of s e c o n d a r y s y s t e m s for


M e x i c o City 1 9 8 5

earthquake

Seismic responses of base-isolated structures: F.-G. Fan and G. Ahmadi

eliminates the sharp peak at the natural frequency of


structure, it does not lead to spectral amplitudes higher
than those of the fixed-base ones for f~ > 2 Hz.
The secondary system deflection response spectra for
the Mexico City earthquake are presented in Figure 14.
The peak deflections are quite high at low frequencies,
and decrease with an increase in f~. The LRB system
leads to spectral amplitudes which are higher than those
for the fixed-base structure over the entire range except
for frequencies around 3.33 Hz. Forf~ > 4 Hz, the P-F
and the R-FBI systems also generate peak accelerations
which are higher than the fixed-base ones. Figure 14
further shows that peak deflections generated in the
secondary system are lower for the EDF system when
compared with other base isolation systems considered
for this long period earthquake excitation.
The results presented in this section shows that, for
ground excitation with considerable energy at low frequencies, the base isolation systems, generally, do not
function properly. In particular, they do not provide
effective protection for the secondary systems. These
observations further emphasize the known fact that base
isolation systems should not be used in regions which
may experience earthquakes with low-frequency energy
content.

(7)

For a structure with a frictional base isolation


system, peak responses of secondary systems can
be significantly reduced by increasing their effective damping coefficients
(8) For regions which have the potential of generating
earthquakes with low-frequency energy contents,
the use of base isolation systems should be avoided

In the present study, idealized models for various base


isolators and structure are used. The nonlinear behaviour of rubber at high strain, the velocity dependence of
friction coefficient, the soil-structure interaction~ and
the effect of torsion and rocking of structure are
neglected. Nevertheless, the presented results could provide a basis for understanding the behaviour of secondary systems in various base-isolated structures. Further
refinement of the model to include these additional
effects is left for future studies.

Acknowledgment
This work has been supported by the National Center for
Earthquake Engineering Research, State University of
New York at Buffalo under grants no. NCEER
8 7 - 2 0 0 7 and 88-2012A.

Conclusions
A study of peak responses of secondary systems attached
to a multi-storey structure with various base isolation
systems under different ground excitations has been
carried out The primary-secondary interactions are
included in the analysis and the resulting equipment
response spectra for different base-isolated structures
are compared with those for the fixed-base one under a
variety of conditions. Effects of damping of the secondary system on their peak responses for a structure with
a frictional base isolation system are also studied. Based
on the presented results, the following conclusions may
be drawn
(1)
(2)

(3)
(4)

(5)

Peak seismic responses of secondary systems are


significantly reduced by using a properly designed
base isolation system
Use of base isolation systems eliminates the
resonance peaks of the equipment response spectra
which occurs at the natural frequency of the fixedbase structure
Mass ratio significantly affects peak responses of
secondary systems for the tuned and near tuned
conditions
The primary-secondary interactions, generally,
reduce the peak response of a secondary system.
Thus, the floor response spectra provide conservative estimates for peak responses for tuned or
nearly tuned conditions
Among the base isolation systems considered, the
linear LRB system leads to the lowest peak
responses for secondary systems m most practical
cases

(6) The frictional base isolation systems generate high


frequency components in the acceleration
responses of the structure which could be damaging
to stiff nonstructural components

References
1 Kelly, J M 'Aselsmlc base isolation revmw and blbhography,' Soil
Dyn Earthquake Engng., 1986, 5, 202-216
2 Kelly, J M 'Aseismlc base Isolation,' Shock Vtbr Dtg, 1982, 14,
17-25
3 Su, L,, Ahmad=, G and Tadjbakhsh, I G 'A comparaave study of
performances of various base tsolaaon systems part I: shear beam
structures,' Earthquake Engn8 Struct Dyn , 1989, 18, 1 1 - 3 2
4 Fan, F -G , Ahmadl, G , and Tadjbakhsh, I G 'Multi-story base~solated buddmgs under a harmonic ground motion - - part I, a comparison of performances of various systems Part II sensltw=ty
analysis" Nuclear Engng Design 1990, 123, 1 - 2 6
5 Fan, F -G , Ahmadl, G , Mostaghel, N , and Tadjbakhsh, I G 'Performance analysts of aselsmic base isolation systems for a multistorey budding,' Sod Dyn. Earthquake Engng_, 1991, 10, 152-171
6 U S Nuclear Regulatory Commission, Standard Review Plan,
NUREG-751087, November 1975
7 U S Nuclear Regulatory Commission, Regulatory Guide, Revision
1_222, February 1978
8 Sackman, J L and Kelly, 1 M 'Seismic analysis of internal equipment and components m structures,' Engng Struct_, 1979, l,
179-190
9 Gupta, A K and Tembulkar, J M. 'Dynarmc decouphng of secondary systems,' Nuclear Engng and Design., 1984, 81, 359-373
l0 Foss, J W and Ntkolakopoulou, G A 'Electromc data processmg
eqmpment braced to resist setsmlc loading,' Proc 7th WCEE, Istanbul, 1980, 477-480
II Hernried, A G andSackman, J L 'Response of segondary systems
in structures subjected to transient excltaUon,' Earthquake Engng
Struct Dyn_, 1984, 12, 737-748
12 Sackman, J_ L and Kelly, J. M. 'Equipment response spectra for
nuclear power plants,' Nucl Engn8 Destgn. 1980, 57, 277-299
13 Singh, M P 'Generation of selsn'nc floor spectra,' J Engng
Mech. Dzv , ASCE, 1975, 101, 593-607
14 Slngh, M P 'Seismic design input for secondary structure,' J_
Struct Dw , ASCE, 1980, 106, 505-517
15 Der Kmreghtan, A., Sackman, J L and Nour-Onud, B 'Dynarmc
analysis of hght equipment m structures response to stochasuc mput,' J. Engng Mech Dw, ASCE, 1983. 109, 9 0 - 1 1 0
16 Der Kiureghian, A Asfura, A , Sackman, J L and Kelly, J_ M.
'Seismic response of multiply supported piping system,' Proc 7th
lnt Conf on Struct. Mech Reactor Technol, paper No. K7/7,
Chmago, II1 , pp 495-502, August 1983

Eng. Struct. 1992, Vol. 14, No 1

47

Seismic responses o f base-~solated structures: F.-G. Fan and G. Ahmadt


17 Gupta, A K and Jaw J W 'Seismic response o1 nonclasslcally
damped systems,' Nucl_ Engng Design. 1986, 91, 1 5 3 - 1 5 9
18 Gupta, A- K and Jaw, J. W 'Coupled response spectrum analysis of
secondary systems using uncoupled model properties.' Nucl Engng
Destgn, 1986 92, 6 1 - 6 8
19 Igusn, T and Der Kmreghmn, A "Dynamic characterlzauon of twodegree-of-freedom eqmpment structure systems.' J Engng Mech
DIv, ASCE, 1985, 111, 1 - 1 9
20 Lee. M C and Penzten, J 'Stochasnc selslmc analyszs of nuclear
power plants and piping systems subjected to multxple support excltatmns,' VCB/EERC-80/119, University of Cahfornm, Berkeley, 1980
21 Kelly, J M 'The influence of base isolation on the seismic response
of hght secondary equipment,' Rep No UCB/EERC-81/I17,
UmversJty of Cahfornla, Berkeley, 1982
22 Fisher, E G. and Dause, W M "Combined analysts and test of
earthquake resistant circuit breakers," Earthquake Engng Struct
D y n , 1976, 5, 2 3 1 - 2 4 3
23 Saemer, S and Godden, W 'Shaking table tests of piping systems
with energy absorbing restramters," Rep UCB/EERC-80/34,
Umverslty of Cahforma, Berkeley, 1982
24 Schneider, S , Lee, H and Godden, W 'Behawor of a piping system
under seismic excitation,' Rep U C B / E E R C - 8 2 / 0 3 , Umverslty of
Cahfornla, Berkeley. 1982
25 Dermlzakls, S N and Mahm, S A 'Development of substructurmg
techmques for on-hne computer controlled setsmJc performance
testing,' Rep UCB/EERC-85/04, Umverslty of Cahfornta,
Berkeley, 1982
26 Manolls, G D , Juhn, G and Remhorn A M "Experimental investtgat~on of primary-secondary system interaction,' Rep
N C E E R - 8 8 - 0 0 1 9 , State Umversity of New York at Buffalo,
Buffalo. NY, May 1988
27 Chen, Y . and Soong. T T ~ S t a t e - o f - - t h e - a r t review seismic
response of secondary systems. Engng Struct , 1988, 10, 2 1 8 - 2 2 8
28 Kelly, J M 'The influence of base ~solatlon on the seismic response
of hght secondary eqmpment," Rep E P R I - N P - 2 9 1 9 , Electrtc
Power Research Instttute, Palo Alto, CA 1983
29 Buckle, I G , Kelly, T E , and Jones, L R 'Basic concepts of
seismic ~solatlon and their application to nuclear structures,' Setsmtc

engmeermg recent advances m destgn, analysts,

48

Eng. Struct. 1992, Vol. 14, No 1

testing and

quahfication methods, P V P - V o l 127, pp 4 2 9 - 4 3 7 1987


30

Kelly, J M , and Tsal, H C "Seismic response of hght mternal


equlprnent m base-isolated structures,' Earthquake Engng Struct
Dyn, 1985, 13, 7 1 1 - 7 3 2
31 Ikonomou, A S 'Parasitic response of eqmpment m base isolated
structure mounted on shdmg bearing,' Setsmtc engmeermg, recent

advames mdestgn, analysts, testing and quahficat:on rnethod~,


32

P V P - V o l 127, ASME, pp 4 0 5 - 4 1 1 , 1987


Wu, Y and Seldensacker, R W 'Sensmvlty studies of a seismically
~solated system to low lrequency amphficatlon, m Setamtt engmeer-

rag, Recent advances m de~tgn, analvst~, testing and quahficatton


method,~ P V P - V o l 127, ASME, pp 4 6 1 - 4 6 7 , 1987
33

34

35

36

37
38

39

40

41

Fan, F G and Ahmadl, G 'Floor response spectra base isolated


multi-storey structures" Earthquake Engng Dvn 1990, 19,
377-388
Kelly, J M , and Hodder, S B 'Experimental study of lead and
elastomenc dampers for base isolation systems m laminated
neopreme bearings," Bullenn of New Zealand National Soc for Earthquake Engng , 1982, 15, 5 3 - 6 7
Huffmann, G R Full base isolaUon for earthquake protection by
hehcal springs and wscodampers," Nuclear Engng Design, 1985, 114,
331-338
Mostaghel, N and Tanbakuchl, J 'Response of shdmg structures to
earthquake support morion,' Earthquake Engng Struet Dyn, 1983,
ll, 729-748
Lt, L 'Base lsolauon measure for aselsmlc budding m china," Proc
8WCEE, San Franosco, CA, 2 1 - 2 8 July, 1984, Vol Vl, 7 9 1 - 7 9 8
Mostaghel,
N
and
Khodaverdmn,
M
'Dynamics of
resihent-friction base isolator ( R - F B I ) , ' Earthquake Engng Struct
D),_n . 1987, 15, 3 7 9 - 3 9 0
Ikonomou, A S 'Alexlstsmon seismic isolation for translatmnal and
rotatmnal seismic input,' Proc 8WCEE, San Franosco, CA. 2 1 - 2 8
July, 1984, Vol V, 9 7 5 - 9 8 2
Gueraud, R , Noel-Leroux,
J -P , Llvolant, M , and
Mlchalopoulos, A P 'Seismic isolatmn using shdmg elastomer bearIng pads,' Nuclear Engng and Design, 1985, 84, 3 6 3 - 3 7 7
Tajlnan, F F and Kelly, J M 'Testing of seismic lsolatmn bearmgs
for advanced hquld metal reactor p n s m , ' in Selsmw, shock, and
vtbratton :solation -1988, P V P - V o l 147, ASME. 9 5 - 9 9

Vous aimerez peut-être aussi