Automation in Construction 22 (2012) 554–566

Contents lists available at SciVerse ScienceDirect

Automation in Construction
journal homepage: www.elsevier.com/locate/autcon

Non-linear time domain analysis of base isolated multi-storey building under site
specific bi-directional seismic loading
A.B.M. Saiful Islam a, Raja Rizwan Hussain b,⁎, Mohammed Jameel a, Mohd Zamin Jumaat a
a
Department of Civil Engineering, University of Malaya, Kuala Lumpur, Malaysia
b
CoE-CRT, Department of Civil Engineering, College of Engg., King Saud University, Riyadh, Saudi Arabia

a r t i c l e i n f o a b s t r a c t

Article history: The incorporation of base isolation in building construction in the region of medium risk seismicity is now an
Accepted 28 November 2011 important issue. Thorough investigation is needed for buildings located in those regions, to be incorporated
Available online 30 December 2011 with base isolator and then dynamic analysis to carry over. This research provides incorporation of Lead Rub-
ber Bearing and High Damping Rubber Bearing as base isolators in addition to focussing on the changes of
Keywords:
structural parameters for isolating effects in those vicinities. Nonlinear models of Lead Rubber Bearing and
Non-linear dynamic
Time domain
High Damping Rubber Bearing have been built up. The design of base isolators for building construction is
Seismic isolation covered along with structural feasibility. Linear static, free vibration and nonlinear dynamic time domain an-
Bi-directional earthquake alyses are performed for both isolated and non-isolated buildings under site specific bi-directional earth-
Multi-storey structures quake. The automated Newmark-beta time integration approach has been adopted for solution in time
Medium risk seismicity domain. The nonlinearities, arising due to base isolated bearings and seismic forces are duly considered.
The study reveals that for medium rise building construction, isolation can significantly reduce seismic re-
sponse in soft to medium stiff soil. The reduction of overturning base moment due to isolation indicates
that the building becomes more stable compared to the fixed base structure. Modelled non-linear bearings
have been found to be suitable to cope with the precise nonlinearities. The building experiences more flexi-
bility even when using the same structural element configuration. In addition, the flexibility of the structure
envisages some sort of savings due to reduced structural responses through incorporation of the isolator. In
seismic vulnerable areas where the main concern is the mitigation of the seismic instability with the support
of critical components, the study shows the effectiveness of the base isolation system in terms of lessening
structural responses under seismic loading.
© 2011 Elsevier B.V. All rights reserved.

1. Introduction responses of multi-storey buildings for near fault motion isolated by

LRB. Dall'Asta and Ragni [15,16] have covered experimental tests, an-
Seismic isolation is the separation of the structure from its base to alytical model and nonlinear dynamic behaviour of HDRB. Bhuyan
negotiate the destructive movement of the ground by providing flex- [17] has developed a rheological model of high damping rubber bear-
ibility and energy dissipation capability through the insertion of isola- ing for seismic analysis identifying nonlinear viscosity. Analysis pro-
tors between the foundation and the building structure [1]. Unlike the cedures to investigate the dynamic structural behaviour with the
conventional design approach, which is based on an increased resis- isolation have also been discussed [18,19]. Although it is a relatively
tance (strengthening) of the structures, the seismic isolation concept recent technology, seismic isolation for multi-storey buildings has
is aimed at a significant reduction of dynamic loads induced by the also been well evaluated and reviewed [20–35]. Base isolator with
earthquake at the base of the structures themselves [2]. Invention of hardening behaviour under increasing load has been developed for
lead rubber bearing (1970's) and high damping rubber (early medium-rise buildings (up to four storey) and sites with moderate
1980's) sets forth a new dimension for the design of base isolated earthquake risk [36]. Nonlinear seismic response evaluation was per-
structure [3–8]. A significant amount of both past and recent research formed by Balkaya and Kalkan [37]. Resonant behaviour of base-
in the area of base isolation has focused on the use of elastomeric isolated high-rise buildings under long-period ground motions was
bearings, such as high damping rubber bearing and lead rubber bear- studied by Ariga et al. [38] and long period building responses by
ings [9–12]. Jangid [13] and Providakis [14] investigated seismic Olsen et al. [39]. Ebisawa et al. [40], Dicleli and Buddaram [41], Cas-
ciati and Hamdaoui [42], Di Egidio and Contento [43] have also
given effort in the progress of isolated system. Komodromos et al.
⁎ Corresponding author. [27], Sharma and Janggid [44], and Kilar and Koren [45] focused on
E-mail address: [email protected] (R.R. Hussain). the seismic behaviour and responses through dynamic analyses of

0926-5805/$ – see front matter © 2011 Elsevier B.V. All rights reserved.
doi:10.1016/j.autcon.2011.11.017
 A.B.M.S. Islam et al. / Automation in Construction 22 (2012) 554–566 555

isolated buildings. Wilkinson and Hiley [46] have presented a non- bearing (LRB) and High damping rubber bearing (HDRB) are used as
linear response history model for the seismic analysis of high-rise isolation devices. The structural system is excited in x and y directions
framed buildings. by two components of earthquake ground motions at East–west and
Though the implementation of isolator is going to be very familiar North–south directions respectively. Nonlinear dynamic analysis has
all over the world, there is a lack of proper research to implement the been carried out for both FB and BI structure using SAP2000 [48].
device practically for local buildings in a medium risk seismic region
like Dhaka, Bangladesh as per the local requirements. So, a thorough 2.1. Design of isolators
study regarding the feasibility of implementing isolator in this locality
is a must to go task. Besides, bi-directional earthquake has rarely been The HDRB and LRB isolators are designed as per the procedure
considered as an issue. The time domain method is relatively more mentioned by Kelly [49], Kelly et al. [50], and Naeim [51]. In this re-
time consuming, lengthy and costly. The frequency domain method, search, a computer code ISODNG09, has been generated to iteratively
on the other hand, is relatively more rapid, concise, and economical. design both the isolators. The total seismic weight, dimensions, layer
However, the time domain method has been employed for consider- thickness and number of layers of bearings are considered as initial
ing non-linearity present in the structural systems. input. The isolator parameters such as post elastic stiffness, high ini-
Time domain analyses, following bi-directional earthquake history, tial stiffness, yield strength, post yield stiffness ratio and effective
have been carried out. Site specific seismic record in two orthogonal di- damping are computed using the above code. These parameters are
rections has been selected to evaluate the bi-directional behaviour. then defined in SAP2000. The bearings are linked at the bottom of
Combined model of High Damping Rubber bearing (HDRB) and Lead each column. The detailed sequential procedure for the design of
Rubber Bearing (LRB) has been adopted here to explore the feasibility. both isolators is shown in a flow chart (Fig. 2). The higher shear strain
Preliminary exploration for determining the suitability of isolator incor- limit for HDRB results in smaller plan size compared to LRB. Due to
poration has been done with equivalent static analysis. Then dynamic the large vertical stiffness of HDRB it can carry heavy loads from the
analysis has been performed to satisfy the structural limitation execut- structure [6]. For the present study, interior columns are isolated
ing different comparative contributions. The analysis and design of the using HDRB and exterior columns are supported by LRB. Dynamic
isolators for a 10-storey sample residential building in Dhaka using analysis of the three dimensional building has been carried out con-
SAP 2000 [47] were performed. Design parameters of the isolator for sidering the associated nonlinearities.
this building and several buildings varying number of storey have
been evaluated. Static analysis and free vibration analysis were also
2.2. LRB isolator
performed along with dynamic analyses. Finally, the acceleration exci-
tation behaviours for fixed and isolated buildings were discussed with
LRB is formed by force-fitting the lead plug into a preformed hole
the displacement patterns at different levels as well. Base shear and
in the low damping elastomeric bearing as shown in Fig. 3. The steel
overturning moments were also compared to and for certain cases.
plates force the lead plug to deform in shear. Performance of LRB is
Every comparison was enforced mentioning the maximum and mini-
maintained during repeated strong earthquakes with proper durabil-
mum values on structural excitation.
ity and reliability. LRB produces the required amount of damping and
has higher initial stiffness. The behaviour of LRB is influenced by yield
2. Mathematical formulation capacity of the lead core, horizontal stiffness of the lead core and hor-
izontal stiffness of the elastomer. The non-linear behaviour of the iso-
An ideal model for a multi storey building is shown in Fig. 1. The lator is modelled considering the approach suggested by Nagarajaiah
building has been considered as moment resisting concrete frame et al. [52]. The hysteresis loop for LRB is shown in Fig. 5a.
structure. The superstructure is configured as a linear elastic system.
The base and floor between which the isolator is placed are assumed 2.3. HDRB isolator
to be infinitely rigid. The nonlinearities arising due to base isolator
bearings and seismic forces are properly considered. The superstruc- The HDRB isolator consists of thin layers of high damping rubber
ture and base are modelled by a consistent mass approach having and steel plates built in alternate layers as shown in Fig. 4. Horizontal
six degrees of freedom at each node. The base isolator carries the ver- stiffness of bearing is controlled by low shear modulus of elastomer
tical load without undergoing vertical deformation. Lead rubber while steel plates provide high vertical stiffness as well as prevents

Fig. 1. Structural model of multistory building: a) FB and b) BI.

556 A.B.M.S. Islam et al. / Automation in Construction 22 (2012) 554–566

Define Material Define seismic input Define Set assumed

Properties and loading on bearing Bearing bearing
types dimensions

Calculate bearing properties for assumed dimensions

LRB HDRB

Choose Characteristic Strength Choose Yield Displacement

Find Shear modulus from Rubber Properties Choose Shear modulus adjusting strain

Calculate stiffness, Yield Force, Yield Displacement Calculate stiffness, Yield Force

Check buckling and strain

Check buckling and strain

Change Bearing Dimensions, if necessary

Change Bearing Dimensions, if necessary

Adjust Shear Modulus and Stiffness

Calculate seismic performance for DBE and MCE

Choose Damping ratio and Isolator Period Define Isolator Period & Thickness

Calculate Hysteresis area, Bearing Force and Choose Damping ratio or Calculate
Damping Coefficient
Calculate Hysteresis area, Bearing
Calculate Spectral displacement and Force and Damping Coefficient
Spectral Acceleration
Calculate strain and adjust Shear modulus
Check for Displacement

Calculate Spectral displacement and

Spectral Acceleration

Calculate load capacity under maximum displacements

Check for Displacement

Adjust bearing dimensions at step “Set assumed bearing Get Design Properties
dimensions” and iterate again, if necessary

Fig. 2. Design flow chart of isolator properties.

a b Stiffening Plates

Rubber Layers

Lead Plug

Side Plate

Fig. 3. Lead rubber bearing a) geometry and b) deformation due to loading.

 A.B.M.S. Islam et al. / Automation in Construction 22 (2012) 554–566 557

a b

Side Plate Rubber Layers

Fig. 4. High damping rubber bearing: a) geometry and b) deformation due to loading.

bulging of rubber. High vertical stiffness of the bearing has no effect where, S = soil structure interaction, T = time period of structure and
on the horizontal stiffness. The damping in the bearing is increased W = effective weight of structure
by adding extra-fine carbon block, oils or resins and other suitable
2
fillers. The dominant feature of HDRB system is the parallel action of qz ¼ C C C I C Z V b ð3Þ
spring and viscous damping. The damping in the isolator is neither
viscous nor hysteretic, but somewhat in between. HDRB isolator has where, qz = sustained wind pressure at height ‘z’ kN/m 2, CC = velocity
lower horizontal stiffness which causes higher natural period of the to pressure conversion = 47.2 × 10 − 6, CI = structure importance coef-
structure. The stiffness and energy dissipation characteristics for ficient, CZ = combined height and exposure coefficient, Vb = basic
HDRB are highly nonlinear and depend on shear strain as shown wind speed at km/h,
(Fig. 5b). The force-deformation behaviour of the HDRB isolator is
considered as nonlinear force displacement hysteresis. The hysteresis P z ¼ C G C p qz ð4Þ
loop area is obtained from the shear strain corresponding to shear
modulus and damping. Pz = design wind pressure at height ‘z’ kN/m 2, CG = gust coefficient
and Cp = pressure coefficient.
2.4. Static analysis
2.5. Nonlinear dynamic analysis
The isolators are designed considering earthquake and wind loads to
be static. The procedure mentioned in Bangladesh standard BNBC, 1993 Non-linear dynamic analysis has been carried out considering a
[53] to compute static loads due to earthquake and wind is considered. typical bi-directional ground motion recorded at Dhaka, Bangladesh.
Lateral loads for the building located in Dhaka, Bangladesh are deter- Finite element analysis package SAP2000 is found to be more appro-
mined considering Z (seismic zone factor), R (response modification priate for the current study. The governing equations of motion are
factor), C (coefficient for soil profile) and I (importance factor). The obtained considering equilibrium of all forces at each degree of free-
base shear due to earthquake and wind can be calculated using Eqs. (1) dom. The equations of motion for super structure and isolated base
and (3) respectively. are written in Eq. (5).
h in o
½Mfy_ þ y€ b g þ ½C fy_ g þ ½K fyg ¼ −½M  T g y€ g ð5Þ
ZIC
Base shear ¼ W ð1Þ
R where, [M], [K] and [C] are the mass, damping and stiffness matrices
of the superstructure respectively; {y} is displacement of super struc-
1:25S ture; {yb} and {ÿg} are base displacement and acceleration relative to
C¼ ð2Þ
T 2=3 the ground; [Tg] is the earthquake influence coefficient matrix.

a) LRB b) HDRB
F Fmax F F max
Fy αKr Ku=k r
Qd
Ku Keff
Ku

Δ
Δy Δmax
Δmax Δ

EDC
EDC

Fig. 5. Idealized non-linear force-displacement curve of bearing.

558 A.B.M.S. Islam et al. / Automation in Construction 22 (2012) 554–566

All nonlinearities are restricted to the base isolator elements only. 3. Numerical study
The above dynamic equilibrium equations considering the super
structure as elastic and base isolator as nonlinear can be written as: A 10 storey building situated at Dhaka, Bangladesh is considered for
the present study. The building is in seismic zone 2 and soft to medium
stiff soil as per Bangladesh national building code (BNBC 1993). It is a
My€ ðt Þ þ C y_ ðt Þ þ K L yðt Þ þ r N ðt Þ ¼ r ðt Þ ð6Þ moment resisting framed structure consisting of 4 bays having spacing
7.62 m c/c in both directions as represented in Fig. 6. The material
and geometric properties considered are f'c = 28 MPa, fy= 414 MPa,
where KL is the stiffness matrix for the linear elastic super structure; C
dead load (excluding self-weight) = 4.8 kPa, live load= 2.4 kPa, slab
is the proportional damping matrix; M is the diagonal mass matrix; rN
thickness = 150 mm, all exterior corner columns 750 mm× 750 mm,
is the vector of forces from nonlinear degrees of freedom in the isola-
all exterior middle columns 950 mm× 950 mm, all interior columns
tor elements; y, y,_ and ÿ are the relative displacement, velocity and
1000 mm× 1000 mm. LRB and HDRB have been assigned for exterior
acceleration with respect to ground; r is the vector of applied loads.
and interior columns respectively. Shear modulus, ultimate elongation,
The effective stiffness at nonlinear degrees of freedom is arbitrary,
elastic modulus and material constant of rubber have been selected as
but varies between zero and the maximum stiffness of that degree of
400 kPa, 650%, 1350 kPa and 0.87 respectively. 40 mm thick steel plates
freedom. The equilibrium equation can be rewritten as
are attached at two sides of each bearing. Non-linear unloading stiffness
is 9.39 kN/mm for LRB and 28.87 kN/mm for HDRB. Whereas the corre-
My€ ðt Þ þ C y_ ðt Þ þ K L yðt Þ þ r N ðt Þ ¼ r ðt Þ−½r N ðt Þ−K N yðt Þ ð7Þ sponding Yield strengths are 155.93 kN and 586.69 kN.
The vertical loads from static analysis are used to design the base
isolators (Table 1). The earthquake loads on the bearing obtained
where K = KL + KN; KL = stiffness matrix of all linear elements, KN = from the dynamic analysis of base isolated (BI) building are used to
stiffness matrix for all of the nonlinear degrees of freedom. check the roll-out condition of base isolators. The designed diameter
of LRB and HDRB is 800 and 950 mm respectively. Their cross section
is given in Fig. 7. The HDRB isolator is defined by plan size and rubber
2.6. Solution technique layer configuration. In addition to plan size and rubber later the LRB
isolator has lead core.
Fast nonlinear analysis (FNA) technique suggested by Wilson [54] For static analysis of the fixed based (FB) building seismic loads
has been considered for solution of equilibrium equations. The meth- are calculated as per BNBC 1993. For isolated building response mod-
od is extremely efficient as it is designed for structural systems which ification factor and importance coefficient is RI = 2.0 and I = 1.0 re-
are primarily linear-elastic, but have limited number of predefined spectively [49]. Table 1 shows results of static analysis. Since there
nonlinear elements. For the FNA method, all nonlinearities are re- is a lack of earthquake record of Dhaka, in the present study the
stricted to the isolator elements. The site specific time history load Natore earthquake record is used to generate acceleration time histo-
is applied quasi-statically with high damping. The FNA considers a ry for the Dhaka earthquake [7,55]. The particulars for the Natore
ramp type of time history function which increases linearly from earthquake are Station ID: ALTUS S/N 2928, Channel 1, 6th Jan 2009
zero to one over a length of time. The nonlinear modal equations 16:04:03 (GMT), Magnitude 4.0. The acceleration time history for
are solved iteratively in each time step. The program considers that the Dhaka earthquake has been illustrated in Fig. 8. Non-linear dy-
the analysis results vary during a time step. The iterations are carried namic time domain analysis is performed using time history of the
out until the solution converges. If convergence cannot be achieved, Dhaka EQ for both x and y directions. The local building construction
the program divides the time step into smaller sub steps and tries codes BNBC used in this research are comparable and up to mark with
again. the national as well as international standards. Each building model

4span @ 7.62m both ways

Y
Z

Fig. 6. Plan view of the multi storied building (FB or BI).

 A.B.M.S. Islam et al. / Automation in Construction 22 (2012) 554–566 559

Table 1 4. Results and discussion

Static analysis results (FB building).

Parameter Value 4.1. Static analysis

Design base shear (EQ load) 3936 kN
Design base shear (Wind load) 2829 kN The results of static analysis of FB and BI building are shown in
Design base moment (EQ load) 89,523 kN-m Table 1 and Table 2 respectively. The design base shear of earthquake
Design base shear (wind load) 48,547 kN-m loading is greater as compared to that of wind loading. Therefore, the
Maximum top story displacement (EQ load) 58.9 mm
building and isolators are designed for seismic load and checked for
Maximum Top story Displacement (Wind load) 35.6 mm
Base displacement (EQ and wind load) 0 wind loads. Lateral load due to wind is approximately 3% of building
Total weight of building 127,754 kN weight. It satisfies one of the conditions for base isolation i.e. lateral
Governing axial load on interior column 7215 kN load due to wind should be less than 10% of building weight [49].
Governing axial load on exterior column 4546 kN The design base shear and moment for FB building is 3936 kN and
89523 kN-m respectively. The maximum displacement at top story
for BI building is 88.5 mm. In case of BI building the isolators absorbs
and damping system configuration is analysed for 30 s durational the seismic load and displaces by 72.8 mm. The lateral loads for wind
earthquake. The Newmark-beta time integration approach has been cause a relative displacement of 1.6 mm at top story. Due to wind
adopted for solution in time domain. The time step used for numeri- loads, the isolator displaces by 52.2 mm using the axial loads on inte-
cal integration is 0.005 s. rior and exterior columns for which the isolators are designed. Table 3
shows the designed results of LRB and HDRB base isolators. The diam-
eter of LRB isolator is 800 mm. The HDRB isolator is slightly larger
having a diameter of 950 mm. All other parameters such as layer
thickness, number of layers, height and shape of both isolators are
12 mm same.

40mm 4.2. Free vibration analysis

Natural time periods or natural frequencies are the important

240mm
characteristics of a structure. It can be used to analyse the results
10mm obtained by dynamic analysis. To evaluate natural frequencies, free vi-
bration analysis of both FB and BI buildings has been carried out. The
16 layers each of 10mm nonlinearities of base isolators have been considered appropriately for
160mm BI buildings. Tables 4–7 show the first 15 natural time periods, natural
frequencies and modal accelerations of FB and BI buildings respectively.
The first time period of the FB building is 0.9132 s whereas the BI build-
ing is having 2.847 s. The natural frequency associated with the building
on seismic isolator is far lower than the corresponding frequency for the
conventional fixed based foundation. The frequency is shifted to
0.35 Hz, which is in the target range of 0.3–0.5 Hz [2]. The first mode
maximum acceleration has been reduced from 29.77 cm/s2 to
40mm 9.768 cm/s 2 which is only one third of the former value. Results also
show that the first global modes for non-isolated foundations are main-
Fig. 7. Vertical section of HDRB (diameter 950 mm) and LRB (diameter 800 mm). ly rocking modes, whereas in the case of seismic isolation they are

Fig. 8. Time history for Dhaka EQ.

560 A.B.M.S. Islam et al. / Automation in Construction 22 (2012) 554–566

Table 2 Table 5
Static analysis results (BI building). Free vibration analysis result (BI building).

Top story Isolator Total structure Mode Period Frequency Cicular frequency
displacement displacement drift no.
s Cyc/s Rad/s
(mm) (mm) (mm)
1 2.847 0.351 2.207
Displacement (EQ load) 88.5 72.8 15.7
2 2.847 0.351 2.207
Displacement (wind load) 53.8 52.2 1.6
3 2.837 0.353 2.215
4 0.478 2.090 13.135
5 0.478 2.090 13.135
Table 3 6 0.416 2.407 15.121
Dimensions of designed isolators. 7 0.214 4.680 29.406
8 0.214 4.680 29.406
Bearing dimensions LRB HDR 9 0.194 5.151 32.365
10 0.169 5.913 37.153
Plan dimension (mm) 800 950
11 0.160 6.258 39.321
Layer thickness (mm) 10 10
12 0.160 6.258 39.321
No. of layers 16 16
13 0.145 6.875 43.194
Lead core size (mm) 150 –
14 0.141 7.084 44.511
Shape Circular Circular
15 0.135 7.383 46.387
Total height (mm) 240 240

associated to pure translation movements. It is also noted that the shift sudden shoot up occurs at 10 s. The maximum base shear peak occurs
of natural vibration period of an isolated system points out that base at 2365 kN in the negative direction. In the case of the y-direction,
isolation provides more flexible isolated system. both negative and positive peaks occur between 12 and 13 s. The os-
As the structural time period is less than 1.0 unit, seismic base cillation then reduces to 15–25% of the earthquake ground excitation.
shear is larger than base shear for wind and this wind induced lateral It is also noticeable that the maximum positive base shear value is
force is less than 10% of seismic weight of the building. Therefore, the 1730 kN. It occurs first, whereas the negative base shear occurs 1 s
incorporation of isolator is feasible for the structure. later with a 5% lower value. The analysis represents the maximum lat-
eral base shear force as V = 2365 kN in the x-direction and 1730 kN in
the y-direction. Here the y-directional value is 17% lower than that of
4.3. Dynamic analysis
the x-direction.

Dynamic analysis in time domain has been performed for FB and

4.3.1.2. Overturning moment time history. Overturning moment is an
BI structures. Responses of both types of structures are obtained in
important design criterion for buildings subjected to seismic loading.
x and y directions. The time histories obtained are maximum base
It is the product of base shear and the lever arm of lateral forces.
shear, overturning moment and displacements. Then, the results of
Fig. 10 shows the time histories of overturning moment for the FB
FB and BI structures are compared with each other. It demonstrates
building. In comparison with the base shear response, the overturn-
the advantages of the BI structure over the FB structure.
ing moment represents a quite similar trend of oscillation. In the x di-
rection the maximum positive base moment value is 19,740 kN-m. It
4.3.1. Responses of the FB structure occurs first whereas the maximum negative value occurs 1 s later
Table 8 describes the obtained result from the dynamic analysis of with a 62% higher magnitude. In the y-direction, minor oscillations
conventional fixed base building structure. Parameter wise compari- continue in same pattern up to 11 s and then the excitations are am-
son has been shown in the subsequent sections. plified. Within 11 to 12 s both negative and positive peaks transpire.
It is also noticeable that the positive peak base moment value
4.3.1.1. Base shear time history. Time histories of base shear in x and y 24,200 kN-m, crops up first while the negative is 5% lower occurring
directions for the FB building subjected to Dhaka earthquake are 1–1.5 s later. In the y-direction the maximum overturning moment
shown in Fig. 9. In the x-direction, the response reflects minor oscilla- is 24,800 kN. It occurs in the positive direction. The analysis predicts
tions and it continues more or less in same pattern up to 10 s until a a maximum moment of 32,120 kN-m in the x-direction and

Table 4 Table 6
Free vibration analysis result (FB building). Modal acceleration for varying Period (FB building).

Mode Period Frequency Cicular frequency Period Acceleration Acceleration

no.
s Cyc/s Rad/s s x-Direction (cm/s2) y-Direction (cm/s2)

1 0.913201 1.095 6.8804 0.913201 29.77 8.935

2 0.913201 1.095 6.8804 0.913201 29.77 8.935
3 0.820971 1.2181 7.6534 0.820971 33.044 9.917
4 0.305778 3.2703 20.548 0.305778 46.06 13.824
5 0.305778 3.2703 20.548 0.305778 46.06 13.824
6 0.277169 3.6079 22.669 0.277169 46.06 13.824
7 0.169141 5.9122 37.148 0.169141 46.06 13.824
8 0.169141 5.9122 37.148 0.169141 46.06 13.824
9 0.156279 6.3988 40.205 0.156279 46.06 13.824
10 0.112683 8.8745 55.76 0.112683 46.06 13.824
11 0.109486 9.1335 57.388 0.109486 45.87 13.767
12 0.109486 9.1335 57.388 0.109486 45.87 13.767
13 0.106621 9.379 58.93 0.106621 44.814 13.45
14 0.106621 9.379 58.93 0.106621 44.814 13.45
15 0.100209 9.9791 62.701 0.100209 42.449 12.74
 A.B.M.S. Islam et al. / Automation in Construction 22 (2012) 554–566 561

Table 7 4.3.1.4. Displacement response. The structure displaces laterally due to

Modal acceleration for varying period (BI building). earthquake excitation. Fig. 12 shows the displacement time histories
Period Acceleration Acceleration at the top level of the FB building under the selected Dhaka earth-
quake. It is observed that the top floor displacement oscillates be-
s x-Direction (cm/s2) y-Direction (cm/s2)
tween +2.9 and −4.3 cm in the x-direction. In the y-direction the
2.847212 9.768 2.932
displacement fluctuates between −4.9 and +3.4 cm in the 12 to
2.847212 9.138 2.743
2.836714 9.285 2.787 13 second zones. This phenomenon is expected as displacement is
0.478372 46.05 13.821 proportional to the lateral force. It is clearly evident from the FB
0.478372 44.588 13.382 case that with support the lateral displacement is zero. Maximum dis-
0.41553 45.13 13.545 placement of the top floor resembles total structural drift which is
0.213669 46.059 13.823
0.213669 44.803 13.446
4.3 cm in the x-direction and 4.9 cm in the y-direction respectively.
0.194135 45.23 13.575 Displacement in the y-direction is 14% higher compared to the dis-
0.169118 45.96 13.794 placement in the x-direction. The structure deforms from zero at
0.15979 45.931 13.785 the base to maximum sway at the top floor in a parabolic pattern as
0.15979 45.834 13.756
shown in Fig. 1a.
0.145465 45.93 13.785
0.14116 45.961 13.794
0.135452 45.864 13.765 4.3.1.5. Statistical analysis. Table 9 shows the results of statistical anal-
ysis of the FB building. The maximum, minimum, mean, standard de-
viation and root mean square values of base shear, overturning
Table 8 moment, floor acceleration and floor displacement respectively are
Dynamic analysis results (FB buildings).
evaluated. The maximum base shear and overturning moment values
Parameter Time domain analysis value are 2365 kN and 32,120 kN-m respectively. It occurs in the negative
Design base shear (kN) in the x-direction 2365 x-direction. The values fluctuate with a standard deviation of 284.65
Design base shear (kN) in the y-direction 1730 and 3706.58 respectively. The FB building has a maximum displace-
Design base moment (kN-m) in the x-direction 32,120 ment of 4.9 mm at the top
Design base moment (kN-m) in the y-direction 24,800
Top story displacement in x-direction (cm) 4.3
Top story displacement in the y-direction (cm) 4.9 4.3.2. Responses of BI structure
The bearings designed are linked at the bottom of the respective
columns at base level of the building to ensure all the properties in
the spring. The structure with isolators is then analysed again for
time domain. The evidences in Tables 10–11 have been attained
24,800 kN-m in the y-direction. The self-weight of the structure plays
from dynamic analysis of the BI building.
a vital role in stabilizing the building against the overturning mo-
ment. The overturning moment is larger in the y-direction because
it results from maximum base shear in the x-direction. 4.3.2.1. Base shear time history. The time histories for base shear of the
BI building are illustrated in Fig. 13. The analysis executes the collapse
4.3.1.3. Acceleration response. Floor acceleration plays a vital role for of the structure to occurred at V = 363 kN in the x-direction and
structural analysis as it outputs the building inertial load for different 651 kN in the y-direction. Peak base shear reduces drastically com-
levels. Once the mass is known for a level the inertia load can be de- pared to the FB structure which fulfils the demand of isolating the su-
termined through simple linear calculation. In contrast to the relative perstructure. It is clear that for the isolated building base shear
displacement response, the floor accelerations are found to be more reduces by up to 43% of the corresponding base shear of fixed value.
sensitive to impact occurrences. Acceleration histories for the FB It predicts a good amount of structural savings and economic assis-
building are demonstrated in Fig. 11. The transitional larger oscilla- tances as well.
tion for the top floor acceleration is from the 11 to 15 second region
of the excitation history in the x-direction. In the y-direction the vac- 4.3.2.2. Overturning moment time history. Accordingly in Fig. 14 over-
illation pattern is also effective for the similar transition area but turning moment corresponding to the BI building is presented. The
more peaks are larger here and the average of this region is high as breakdown forecasts the ultimate moment capacity M = 12,510 kN-m
well. The peak floor acceleration at the top is 0.215 g in the x-direc- in the x-direction and 11,260 kN-m in the y-direction to prevent col-
tion which is about 40% greater and 0.122 g in the y-direction lapse. Overturning base moment reduces by up to 35% of the fixed foun-
which is about 25% greater than the input Dhaka earthquake ground dational base moment. So the building becomes more stabilized
excitation. This shows that with increasing height the acceleration is compared to the FB structure. The building experiences more flexibility
also increased due to the seismic impact on the building structure. even when using the same structural element configuration. Allowance

Fig. 9. Base shear in x and y directions (fixed based case).

562 A.B.M.S. Islam et al. / Automation in Construction 22 (2012) 554–566

Fig. 10. Base moment in x and y directions (fixed based case).

Fig. 11. Acceleration time history in x and y directions (fixed based case).

Fig. 12. Displacement time history in x and y directions (fixed based case).

of translational movements of support abruptly changes the pattern of 0.159 g in the x-direction which is about 5% greater and 0.095 g in
whole structure deformation like in Fig. 1b. the y-direction which is about 5% lower than the input ground excita-
tion. For the isolation flexibility the structure experiences a mention-
4.3.2.3. Acceleration response. Acceleration time histories for the seis- able amount of acceleration at the base also that is valued as 0.12 g in
mically isolated building are given in Fig. 15 for the top and base of the x-direction and 0.08 g in the y-direction. For the isolated building
the structure. The peak floor accelerations in this case at the top are peak, acceleration at the top floor reduces by up to 74% of the

Table 9
Statistical analysis results of time domain responses (FB building).

structural parameter Direction Fixed

Maximum Minimum Mean Standard deviation Root mean square

Base shear (kN) x 1506 2365 − 0.03957 284.6573 284.5625

y 1730 1649 0.268015 262.4023 262.315
Base moment (kN-m) x 19,740 32,120 − 4.19639 3706.582 3705.349
y 24,800 23,060 − 6.27736 3937.748 3936.441
Top point acceleration (cm/s2) x 129 211 − 0.00454 20.91838 20.91141
y 120 118 − 0.00148 20.56363 20.55678
Base acceleration x – –
y – –
Top point displacement (cm) x 2.9 4.3 − 0.00057 0.545929 0.545748
y 3.4 4.9 0.00459 0.721892 0.721666
Base displacement (cm) x – –
y – –
 A.B.M.S. Islam et al. / Automation in Construction 22 (2012) 554–566 563

Table 10 analysis. The peak top displacement reduces by up to 35% of the

Displacement output in dynamic analysis (BI building).
fixed foundational lateral movement. Magnitudes of displacement
Isolator displacement Total structure history indicate that the base isolator absorbs most of the displace-
(mm) drift (mm) ment of the superstructure.
x-Direction (static analysis) 72.8 15.7
y-Direction (static analysis) 72.8 15.7 4.3.2.5. Statistical investigation. Table 12 shows the statistical study of
x-Direction (time domain analysis) 16.9 8.1 the BI structure for nonlinear dynamic behaviour. It includes base
y-Direction (time domain analysis) 29.8 12.2
shear, overturning moment, floor acceleration and floor displacement
with their maximum and minimum values and mean, standard devi-
ation, root mean square values as well.
Table 11
Base Shear and base moment in dynamic analysis (BI building).
The acceleration responses reduce significantly in the base-
isolated building compared to the original building as expected. The
Parameter Time domain analysis value base shears in each direction are decreased compared to the fixed
Design base shear (kN) in the x direction 363 building. On the other hand, as regards the building response in
Design base shear (kN) in the y direction 651 terms of displacements, the maximum horizontal displacements eval-
Design Base Moment (kN-m) in the x direction 12,510
uated at the foundation level were well below the expected static de-
Design base moment (kN-m) in the y direction 9523
sign displacement of isolators.
Peak base shear reduces significantly compared to the FB structure
corresponding top point acceleration of the fixed one. The isolated as expected, which fulfils the demand of isolating superstructure. The
building exerts a good amount of acceleration at the base. But for reduction of overturning base moment due to isolation indicates that
the fixed one, at base, there is a null acceleration excitation and the building becomes more stabilized compare to the FB structure.
displacement.
4.3.2.6. Influence of nonlinear automated simulation. Consideration of
4.3.2.4. Displacement response. Time histories of displacement for the nonlinearity of base isolation exerts actual behaviour of the isolating
BI building are shown in Fig. 16 for the top and base of the structure. elements and their upshot on the structural responses. The selected
In case of isolation the isolator itself moves laterally. So, there is also automated simulation includes the precise modelling and analysis in
mentionable lateral movement in the joint/support at base level. It is a consistent manner. Automated fast nonlinear analysis is found to
observed that the peak displacement of the top floor is 2.5 cm in the be efficient requiring very less time but it offers precise solution. Iso-
x-direction and 4.2 cm in the y-direction whereas the values at the lating strategy predicts well amount of structural savings and also
base level are 0.8 cm and 1.7 cm respectively. Here total structure economic assistances as well. The building experiences additional
drifts come to 1.7 cm in the x-direction and 2.5 cm in the y-direction flexibility even using same structural elements configuration. Allow-
significantly lower than that in the FB case. In case of isolation the iso- ance of transitional movements of support abruptly changes the pat-
lator itself moves laterally while displacement is zero for the FB build- tern of whole structure deformation. As a concluding remark, for
ing. So, total structural drift is lesser than that of the fixed based case. applications like the medium rise building structures in soft to medi-
For the isolated structure, the displacement of both the superstruc- um stiff soil of Dhaka, where the main concern is the mitigation of the
ture and isolator increases as the superstructure becomes more flex- seismic excitation at the supports of critical components, a base isola-
ible. This trend is true for static, free vibration and time domain tion is to be viably recommended.

Fig. 13. Base shear in x and y directions (isolated based case).

Fig. 14. Base moment in x and y directions (isolated based case).

564 A.B.M.S. Islam et al. / Automation in Construction 22 (2012) 554–566

Fig. 15. Acceleration time history in x and y directions (isolated based case).

Fig. 16. Displacement time history in x and y directions (isolated based case).

Table 12
Statistical analysis results of time domain responses (BI Building).

Structural parameter Direction Isolated

Maximum Minimum Mean Standard deviation Root mean square

Base shear (kN) x 363 306 0.7575748 69.668 69.648909

y 546 651 − 0.85059 131.0317 130.9908
Base moment (kN-m) x 12,510 11,260 18.24944 2655.256 2654.434
y 9523 8041 9.170651 1648.975 1648.451
Top point acceleration (cm/s2) x 94 156 − 0.00272 16.43835 16.43287
y 93 73 − 0.01663 15.31261 15.30752
Base acceleration (cm/s2) x 87 117 0.006199 14.13679 14.13208
y 71 76 − 0.00627 13.34395 13.3395
Top point displacement (cm) x 2.5 2.0 − 0.00337 0.428622 0.428492
y 4.2 3.3 0.007431 0.847827 0.847577
Base displacement (cm) x 0.8 0.9 − 0.00108 0.170408 0.170355
y 1.7 1.3 0.003522 0.319779 0.319692
 A.B.M.S. Islam et al. / Automation in Construction 22 (2012) 554–566 565

5. Conclusion Malaya funded the research work through the High Impact Research
Grant D000002-16001.
Nonlinear time domain analyses under bi-directional earthquake
history have been carried out. The performances of multi storey struc- References
tures isolated with the bearing systems are evaluated by Fast Nonlinear
Analysis. The responses due to structural changes are accurately formu- [1] M. Ismail, J. Rodellar, F. Ikhouane, An innovative isolation device for aseismic de-
sign, Engineering Structures 32 (2010) 1168–1183.
lated to estimate responses of base-isolated structures. The results from [2] I. Micheli, S. Cardini, A. Colaiuda, P. Turroni, Investigation upon the dynamic
the analyses, static, free vibration and time domain are obtained. The structural response of a nuclear plant on aseismic isolating devices, Nuclear Engi-
summarized findings are as follows. neering and Design 228 (2004) 319–343.
[3] Rizwan Hussain Raja, Ishida Tetsuya, Computer-aided oxygen transport model of
mass and energy simulation for corrosion of reinforced steel, Automation in Con-
1. The major effect of base isolation on the seismic response is by far
struction 20 (5) (2011) 559–570.
the radical reduction of horizontal accelerations. It is observed [4] A.B.M.S. Islam, M. Jameel, M.Z. Jumaat, Seismic Base Isolation for Buildings in Re-
that, on soft to medium stiff soil, the reduction of response peak gions of Low to Moderate Seismicity: A Practical Alternative Design, Practice Pe-
riodical on Structural Design and Construction, ASCE, 2010, doi:10.1061/(ASCE)
accelerations at the support level is not far from ten times for iso-
SC.1943-5576.0000093, Published online at 22 December.
lators with respect to the case without base isolation. This com- [5] A.B.M.S. Islam, M. Jameel, M.A. Uddin, S.I. Ahmad. Simplified design guidelines for seis-
parison was performed for foundations on soft to medium stiff mic base isolation in multi-storey buildings for Bangladesh National Building Code
soil, but similar conclusions could be drawn for any type of soil. (BNBC), International Journal of the Physical Sciences 6 (23) (2011) 5467–5486.
[6] A.B.M.S. Islam, S.I. Ahmad, T.M. Al-Hussaini, Effect of isolation on buildings in
2. As regards the building response in terms of displacements, the Dhaka, BES, 3 rd International Earthquake Symposium, Bangladesh, Dhaka, 2010,
maximum horizontal displacements evaluated at the foundation pp. 465–472.
level were well below the expected static design displacement [7] R.R. Hussain, A.B.M.S. Islam, S.I. Ahmad, Base Isolators as Earthquake Protection
Devices in Buildings, First ed VDM Publishing House Ltd. Benoit Novel, Simulta-
of isolators. neously published in USA & U.K, 2010.
3. Extensive sensitivity studies to find the influence of various im- [8] A.B.M.S. Islam, M. Jameel, M.Z. Jumaat, Seismic isolation in buildings to be a prac-
portant structural parameters of both isolator and superstructure tical reality: behaviour of structure and installation technique, Journal of Engi-
neering and Technology Research 3 (4) (2011) 97–117.
on the behaviour of isolated structure are possible by the SAP [9] A.B.M.S. Islam, M. Jameel, M.Z. Jumaat, Study on optimal isolation system and dy-
module. namic structural responses in multi-story buildings, International Journal of the
4. Both the displacement and acceleration responses of a super- Physical Sciences 6 (9) (2011) 2219–2228.
[10] A.B.M.S. Islam, S.I. Ahmad, M. Jameel, M.Z. Jumaat, Study on corollary of seismic
structure without an isolator are much more sensitive while the
base isolation system on buildings with soft story, International Journal of the
displacement and acceleration responses of a superstructure Physical Sciences 6 (11) (2011) 2654–2661.
with an isolator reduce significantly. [11] F. Kozin, H.G. Natke, System identification techniques, Structural Safety 3 (1986)
269–316.
5. The base isolator is more effective to mitigate displacement than
[12] R. Jangid, T. Datta, Performance of base isolation systems for asymmetric building
acceleration. subject to random excitation, Engineering Structures 17 (1995) 443–454.
6. Peak base shear reduces significantly compared to the FB struc- [13] R.S. Jangid, Optimum lead-rubber isolation bearings for near-fault motions, Engi-
ture which fulfils the demand of isolating superstructure. It pre- neering Structures 29 (2007) 2503–2513.
[14] C.P. Providakis, Effect of LRB isolators and supplemental viscous dampers on seis-
dicts well the amount of structural savings and also provides mic isolated buildings under near-fault excitations, Engineering Structures 30
economic assistance. (2008) 1187–1198.
7. There is a reliable reduction of overturning base moment than the [15] A. Dall'Asta, L. Ragni, Experimental tests and analytical model of high damping
rubber dissipating devices, Engineering Structures 28 (2006) 1874–1884.
fixed foundational base moment. So the building becomes more [16] A. Dall'Asta, L. Ragni, Nonlinear behaviour of dynamic systems with high damping
stabilized compared to the FB structure. The building experiences rubber devices, Engineering Structures 30 (2008) 3610–3618.
more flexibility even when using the same structural element [17] A.R. Bhuiyan, Y. Okui, H. Mitamura, T. Imai, A rheology model of high damping
rubber bearings for seismic analysis: identification of nonlinear viscosity, Interna-
configuration. Allowance of transitional movements of support tional Journal of Solids and Structures 46 (2009) 1778–1792.
abruptly changes the pattern of whole structure deformation. [18] F. Casciati, Stochastic dynamics of hysteretic media, Structural Safety 6 (1989)
8. For applications like the medium rise building structures in 259–269.
[19] R. Clough, J. Penzien, Dynamics of Structures, 2nd ed McGraw-Hill, 1993.
Dhaka, where the main concern is the mitigation of the seismic
[20] A.J. Carr. An introduction to seismic isolation : by R.I. Skinner, W.H. Robinson and
excitation at the supports of critical components, a base isolation G.H. McVerry, Wiley, Chichester, 1993, ISBN 0 471 93433 X, Structural Safety
is to be viably recommended. 1994; 14: 219–20.
[21] S.L. Dimova, A. Elenas, Seismic intensity parameters for fragility analysis of struc-
9. This investigation was based on free-field excitations in accor-
tures with energy dissipating devices, Structural Safety 24 (2002) 1–28.
dance with the site specific bilateral EQ data. For applications on [22] V.A. Matsagar, R.S. Jangid, Influence of isolator characteristics on the response of
buildings on soft soils where more significant long period excita- base-isolated structures, Engineering Structures 26 (2004) 1735–1749.
tions are to be taken into account, the design of the base isolation [23] W. Hong, H. Kim, Performance of a multi-story structure with a resilient-friction
base isolation system, Computers and Structures 82 (2004) 2271–2283.
needs particular care, in order to avoid resonance effects. [24] A. Baratta, I. Corbi. Optimal design of base-isolators in multi-storey buildings,
10. In this case, the most effective choice appears that of HDRB and Computers & Structures 82 (2004) 2199–209.
LRB bearings, as resulting in a lower isolation frequency and [25] V.K. Agarwal, J.M. Niedzwecki, J.W. Van de Lindt, Earthquake induced pounding in
friction varying base isolated buildings, Engineering Structures 29 (2007)
then in lower peak accelerations, but the isolation choice should 2825–2832.
generally be based on the best compromise between the reduc- [26] P. Komodromos, S. Stiemer, Seismic isolation for earthquake resistant structures,
tion of floor accelerations and the amplification of building Applied Mechanics Reviews 54 (2001) B112.
[27] P. Komodromos, P. Polycarpou, L. Papaloizou, M. Phocas, Response of seismically
rigid-body displacements. isolated buildings considering poundings, Earthquake Engineering and Structural
11. To accurately determine the collapse loads and acceleration– Dynamics 36 (2007) 1605–1622.
displacement behaviours of the structures, in the nonlinear time [28] P. Komodromos, Simulation of the earthquake-induced pounding of seismically
isolated buildings, Computers and Structures 86 (2008) 618–626.
history analysis P − δ effects need to be considered as does the
[29] C. Loh, C. Lin, C. Huang, Time domain identification of frames under earthquake
more true sketch be accessed. loadings, Journal of Engineering Mechanics 126 (1999) 693.
[30] L.-Y. Lu, G.-L. Lin, Predictive control of smart isolation system for precision equip-
ment subjected to near-fault earthquakes, Engineering Structures 30 (2008)
Acknowledgement 3045–3064.
[31] E.M. Güneyisi, G. Altay, Seismic fragility assessment of effectiveness of viscous
The authors gratefully acknowledge the effort by the Bangladesh dampers in R/C buildings under scenario earthquakes, Structural Safety 30
(2008) 461–480.
Earthquake Society and King Saud University, Deanship of Scientific [32] C. Bucher, Probability-based optimal design of friction-based seismic isolation de-
Research, College of Engineering Research Center. The University of vices, Structural Safety 31 (2009) 500–507.
566 A.B.M.S. Islam et al. / Automation in Construction 22 (2012) 554–566

[33] C.C. Spyrakos, I.A. Koutromanos, C.A. Maniatakis, Seismic response of base- [44] A. Sharma, R. Jangid, Behaviour of base-isolated structures with high initial isolator
isolated buildings including soil-structure interaction, Soil Dynamics and Earth- stiffness, International Journal of Applied Science, Engineering and Technology 5
quake Engineering 29 (2009) 658–668. (2009).
[34] K. Goda, C.S. Lee, H.P. Hong, Lifecycle cost–benefit analysis of isolated buildings, [45] V. Kilar, D. Koren, Seismic behaviour of asymmetric base isolated structures with
Structural Safety 32 (2010) 52–63. various distributions of isolators, Engineering Structures 31 (2009) 910–921.
[35] P.C. Polycarpou, P. Komodromos, Earthquake-induced poundings of a seismically [46] S. Wilkinson, R. Hiley, A non-linear response history model for the seismic anal-
isolated building with adjacent structures, Engineering Structures 32 (2010) ysis of high-rise framed buildings, Computers and Structures 84 (2006) 318–329.
1937–1951. [47] CSI Computer & Structures Inc, SAP2000. Linear and Nonlinear Static and Dynam-
[36] A. Pocanschi, M.C. Phocas, Earthquake isolator with progressive nonlinear ic Analysis of Three-dimensional Structures, Computer & Structures, Inc, Berkeley
deformability, Engineering Structures 29 (2007) 2586–2592. (CA), 2004.
[37] C. Balkaya, E. Kalkan, Nonlinear seismic response evaluation of tunnel form build- [48] A. Habibullah, SAP 2000. Static and Dynamic Finite Element Analysis of Struc-
ing structures, Computers and Structures 81 (2003) 153–165. tures, Computers and Structures Inc, Berkeley, California, 2005.
[38] T. Ariga, Y. Kanno, I. Takewaki, Resonant behaviour of base-isolated high-rise [49] T.E. Kelly, Base Isolation of Structures: Design Guidelines, Holmes Consulting
buildings under long-period ground motions, The Structural Design of Tall and Group Ltd, Auckland, 2001.
Special Buildings 15 (2006) 325–338. [50] T.E. Kelly, W.H. Robinson, R.I. Skinner, Seismic Isolation for Designers and Struc-
[39] A. Olsen, B. Aagaard, T. Heaton, Long-period building response to earthquakes in tural Engineers, Robinson seismic Ltd., Wellington, 2006.
the San Francisco Bay Area, Bulletin of the Seismological Society of America 98 [51] F. Naeim, J. Kelly, Design of Seismic Isolated Structures: From Theory to Practice,
(2008) 1047. Wiley, 1999.
[40] K. Ebisawa, K. Ando, K. Shibata, Progress of a research program on seismic base [52] S. Nagarajaiah, A.M. Reinhorn, M.C. Constantinou, Nonlinear dynamic analysis of
isolation of nuclear components, Nuclear Engineering and Design 198 (2000) 3-D base-isolated structures, Journal of Structural Engineering 117 (1991)
61–74. 2035–2054.
[41] M. Dicleli, S. Buddaram, Comprehensive evaluation of equivalent linear analysis [53] Bangladesh National Building Code, BNBC, Bangladesh: Housing and Building Re-
method for seismic-isolated structures represented by sdof systems, Engineering search Institute, Bangladesh Standard and Testing Institute, 1993.
Structures 29 (2007) 1653–1663. [54] E. Wilson, Three-dimensional Static and Dynamic Analysis of Structures, 3rd ed
[42] F. Casciati, K. Hamdaoui, Modelling the uncertainty in the response of a base iso- Computers & Structures Inc, Berkeley, 2002.
lator, Probabilistic Engineering Mechanics 23 (2008) 427–437. [55] A.B.M.S. Islam, M. Jameel, M.A. Rahman, M.Z. Jumaat, Earthquake time history for
[43] A. Di Egidio, A. Contento, Seismic response of a non-symmetric rigid block on a Dhaka, Bangladesh as competent seismic record, International Journal of the
constrained oscillating base, Engineering Structures 32 (2010) 3028–3039. Physical Sciences 6 (16) (2011) 3921–3926.