Soil Dynamics and Earthquake Engineering 149 (2021) 106827

Contents lists available at ScienceDirect

Soil Dynamics and Earthquake Engineering

journal homepage: www.elsevier.com/locate/soildyn

Application of controlled-rocking isolation with shape memory alloys for an

overpass bridge
Rajesh Rele a, *, Ranjan Balmukund a, S. Bhattacharya b, Liang Cui b, Stergios A. Mitoulis b
a
R.R Consulting Engineers, Mumbai, India
b
Department of Civil and Environmental Engineering, University of Surrey, Guildford, UK

A R T I C L E I N F O A B S T R A C T

Keywords: Resilience of bridges in seismic zones can be realised by taking the advantage of rocking isolation which aims at
Bridges reducing the permanent drifts after a seismic event. The seismic forces at the base of the bridge can be reduced by
Pier allowing uplift in the foundation when subjected to ground shaking. Conventional monolithic connection of
Rocking isolation
bridge pier to the foundation often leads to severe damages (or even collapse) during high magnitude earth­
Elastomeric pad
Resilient
quakes. In this context, this article proposes a novel seismically resilient pier footing which rocks on elastomeric
Shape memory alloy pads and external restrainers (provided by shape memory alloy bars). Seismic performance of a typical existing
overpass motorway bridge is compared with the proposed rocking isolation concept. The proposed technique
shows good re-centring capability during earthquakes with negligible residual drifts. Furthermore, it is also
observed that forces in the pier and size of pier footing are reduced as compared with the reference bridge
considered in this study.

1. Introduction
The concept of ductility is used in conventional design of bridge piers
wherein the pier reinforcement is detailed to develop flexural plastic
hinges at the base and at the top. This is to accommodate certain degree
of inelastic behaviour. While bridges designed in this manner may be
safe against complete collapse (satisfied ULS: Ultimate Limit State), they
may suffer damages (violating SLS: Serviceability Limit State criteria)
during severe earthquake excitations. Fig. 1 shows observed damage
patterns in bridge piers in three different earthquakes and they are ex­
amples of ULS failure. During the 1960 Chilean earthquake, few tall
slender structures such as water tanks escaped damage and this made
Housner [1] to identify the dynamic stability of rocking structures. The
response of a structure free to rock on its foundation offers a means of
base isolation which help to reduce the damage [2]. Beck and Skinner
[3] studied the rocking response of a step bridge pier, a system later used
in design of South Rangitikei Railway Bridge in New Zealand in 1981.
The rocking isolation in the form of structural rocking or geotech­
nical rocking of the bridge pier experience far less damage when sub­
jected to highly intensive seismic ground motions [5]. The rocking
isolation also has additional advantage of pier being re-centred to its
initial position due to the increased period of vibration owing to the
flexibility of the resilient pier. Pollino and Bruneau [6] studied a seismic
retrofit technique that allows bridge steel truss piers to uplift and rock
on their foundation and thus leave the bridge with no residual dis­
placements following an earthquake. The concept of damage avoidance
design philosophy by means of rocking was studied experimentally by
Solberg [7] and Rodgers [8]. Many researchers [9-16] have proposed to
place central tendon in the bridge pier along with external dissipaters in
order to minimise the damage during seismic events. The concept of
dissipative controlled rocking which is a combination of free rocking,
post-tensioning and dissipative devices was introduced by Liu and
Palermo [17] in order to have least damage in the pier. Agalianos et al.
[18] proposed two rocking isolation techniques in which the first one
allows the pier to rock on the foundation while the piers are not
monolithically connected to the foundation but are designed to uplift
and rock under seismic motions. It was suggested that a recess shall be
provided in footing to avoid sliding of pier and promote only rocking
phenomenon. The second concept promotes rocking of the pier and
foundation assembly by full mobilization of the soil bearing capacity.
Bridge upgrades based on traditional retrofitting methods are chal­
lenging, costly, time-consuming and cause extensive disruptions. Bridge

* Corresponding author. R.R Consulting Engineers, 801 Vighnaharta Tower, Shivaji Path, Behind M.H School, Thane (W), 400602, Maharashtra, India.
E-mail addresses: rajesh_rele@rediffmail.com (R. Rele), rbalmukund@rediffmail.com (R. Balmukund), s.bhattacharya@surrey.ac.uk (S. Bhattacharya), l.cui@
surrey.ac.uk (L. Cui), s.mitoulis@surrey.ac.uk (S.A. Mitoulis).

https://doi.org/10.1016/j.soildyn.2021.106827
Received 22 July 2020; Received in revised form 14 May 2021; Accepted 20 May 2021
Available online 28 June 2021
0267-7261/© 2021 Elsevier Ltd. All rights reserved.
R. Rele et al. Soil Dynamics and Earthquake Engineering 149 (2021) 106827

design and retrofitting philosophies urgently need a shift towards Table 1

accelerated constructions, minimal damages and rapid restorations Performance levels, corresponding damage state and drift limits (ATC 40 [27]
[19]. Energy dissipation can also be enhanced by using easily replace­ and FEMA 356 [28]).
able external steel bars which are designed to yield and offer minimum Performance level Damage state Drift limits
damage to the bridge piers [20]. The design philosophy in the current Fully operational, No damage <0.2%
codes [21–25] is to allow damages under extreme loading conditions Immediate occupancy
which lead to high repair costs and significant downtimes. Operational Repairable <0.5%
Performance-based earthquake engineering (PBEE) evaluates the Life safe Irreparable <1.5%
Near collapse, Limited safety Severe
behaviour of a structure under a certain seismic action by achieving <2.5%
Collapse Failure >2.5%
stated performance objective related to a level of damage to the struc­
ture. The concept of performance-based design is well documented in
SEAOC [26]. Table 1 shows the performance levels in terms of per­
Table 2
centage drift (a ratio of pier horizontal displacement to its height) as
Damage/Limit state of bridge components (Dutta and Mander [29]).
mentioned in ATC 40 [27] and FEMA 356 [28]. Five different damage
states (Table 2) for bridge pier have also been recommended by Dutta Damage state Description Drift limits
and Mander [29]. Almost no First yield 0.005
The researchers in the past have used external prestress tendon and Slight Cracking, spalling 0.007
external dampers to stimulate the rocking isolation in order to have a Moderate Loss of anchorage 0.015
Extensive Incipient column collapse 0.025
low damage design which can lead to high maintenance cost in the long Complete Column collapse 0.05
run. Piers rocking on soil have also been explored in the past by delib­
erately under designing the foundation so that it rocks on soil under­
neath and the energy dissipation occurs in soil itself [30]. Such type of to have a practical aspect so that pads can be designed for suitable sizes
rocking isolation causes excessive settlements while the sinking of as per requirement of forces when the pier rocks on it. The novelty of this
footing in yielding soil also increases the risk of failure. The concept of research is its rocking mechanism on pads and restrainers as opposed to
pier rocking on neoprene pad of size covering the entire area of footing most of the literature which mentions rocking on soil surface [32,33].
of a reinforced concrete column was studied by means of centrifuge The additional element used in the proposed system which is an
experiment and shake table test [31]. The results showed that the re­ extension of author’s work [34] is the shape memory alloy bar which
sidual displacement and force demands were minimum as compared to controls the large horizontal displacements of pier as well as fluctuation
conventional pier design. The proposed rocking system avoids the of axial force in the piers [35]. The proposed bridge is integral which
plastic impact of two concrete surfaces as well as yielding of soil as it does not have expansion joint between the deck and the piers. This re­
rocks on elastomeric pads. These pads are placed in discretised manner duces substantially the risk of having the span failure of the bridge deck

Fig. 1. (a) Flexural-shear failure at pier midheight in the 1995 Kobe Earthquake (b) Failure of flexural plastic hinges in bridge piers in the 1994 Northridge
earthquake (c) Shear Failure of flared column in the 1994 Northridge earthquake (d) Collapse of the Hanshin expressway in the 1995 Kobe earthquake [4].

2
R. Rele et al.
[36]. Thus, the major failures in bridge pier which are pier yielding,
deck unseating and pounding of decks can be avoided in the proposed
system.
This paper proposes a controlled rocking resilient foundation which
uses elastomeric pad incorporated beneath the footing of bridge piers
and shape memory alloy bars (SMA) as external restrainers, see Fig. 2b.
The main aim of this paper is to evaluate the seismic performance (in
both longitudinal and transverse direction of bridge) of a conventional
system (Fig. 2a) and compare with the proposed system (Fig. 2b). The
proposed technique is described with its assumptions, advantages and
retrofit strategy. The reference overpass bridge which is located in
Greece is described showing its important features. This is then followed
by detailed description of development of finite element model created
for the reference bridge and the proposed bridge with rocking isolation
system. The seismic behaviour is assessed by performing advanced
analysis such as non-linear dynamic time history and pushover in order
to predict the response as closely as possible. The modelling of pad and
shape memory alloy bars are then described and followed finally by the
conclusions obtained from this research study.
2. Proposed novel rocking isolation technique
The proposed rocking bridge pier foundation is shown in Fig. 2 (b). It
rests on firm concrete sub-base with elastomeric pads placed first on a
rough surface and then the footing along with pier is erected. The recess
made in the base acts like stoppers to restrict horizontal sliding and thus
allow only rocking of the foundation without any translational move­
ments. After the superstructure and substructure loads are activated, the
pad deforms and then the shape memory alloy (SMA) bars of required
length are installed.
2.1. Assumptions
Horizontal translation of footing and pads are not allowed in order to
avoid the walking off phenomena [37]. This is achieved by providing
stoppers around the periphery of the footing. In practice, a cushion in
the form of rubber pad is attached vertically alongside of the stopper in
order to avoid damage to footing due to pounding of the two concrete
surfaces [18]. The footing simply rests on the elastomeric pads placed
beneath it. There is no connection between the footing and pads. The
uplift of footing is allowed within permissible limits. Pads must be
correctly chosen to satisfy buckling criteria given in the appropriate
Fig. 2. Studied Configuration (a) Conventional system (b) Proposed rocking system.
3
Soil Dynamics and Earthquake Engineering 149 (2021) 106827
codes. Since pier footing rest on elastomeric pads, it is recommended to
keep the pads on stable foundation such as hard rock, stiff soil or pile cap
to avoid settlement of foundation while rocking.
2.2. Advantages of the new technique as compared to conventional
seismic design philosophy
The conventional pier design relies on pier yielding and the energy
dissipation occurs in the lowermost portion of pier by formation of
plastic hinge. In the proposed technique, by allowing uplift of footing
the pier yielding is avoided. The pads and restrainers act as energy
dissipating elements instead of pier as in the conventional seismic design
philosophy. The post-earthquake serviceability in the proposed tech­
nique is enhanced as the residual displacements in pier are almost
negligible as the system rocks on the pads. Furthermore, the seismic
demands for the proposed pier foundation are less as compared to the
conventional pier. The uplift of foundation can also be controlled by
means of external restrainers and thus this novel system has application
where bridges are expected to undergo relatively high seismic
excitations.
2.3. Purpose of elastomeric pads
Pads isolate the base of footing and thus increase the flexibility of the
system [38]. Time period of structure can be controlled by varying the
sizes of pad and thus the seismic demands on pier can be reduced. The
provision of pads enhances recentring of pier and thus negligible per­
manent drifts can be achieved [34]. Plastic impacts between two con­
crete surfaces are avoided by placing pads below the footing, which is
not the case in conventional concrete to concrete rocking [18].
2.4. Purpose of external restrainers (SMA bars)
The main purpose of restrainers is to reduce the pier horizontal
displacements in case of bridges that are subjected to very high seismic
excitations. The SMA bars will also act as sacrificial element in case of
very high magnitude of earthquakes. The additional purpose of these
bars is that they control the excessive foundation uplifts when piers are
subjected to large ground accelerations.
R. Rele et al.
2.5. Retrofit strategy for the proposed rocking isolation system
After significant earthquakes it is possible that the pads may need
replacement. In order to replace the pads in the near future, flat jacks are
used. The diameter of flat jack shall be chosen according to the axial load
in the pad. The bridge is lifted by 10–15 mm so that pads can be
replaced. The pads in the centre of footing are replaced first after the
pads at periphery are removed. A typical layout for future replacement
of pads with jack locations are shown in Fig. 3.
3. Description of reference bridge
An actual overpass bridge (A01-TE20) of the Attiki Odos Motorway
(Athens, Greece) is used as reference bridge forming the basis of the
developed numerical models. Besides its simplicity, the selected bridge
system is similar to 30-% of the bridges considered for metropolitan
motorways in general [39,40]. The selected reference bridge is a
three-span bridge with semi-integral prestressed concrete voided slab
supported on two reinforced concrete circular piers of 2-m in diameter
and 8.8 m in height. (Fig. 4). The end of the bridge is supported on
abutment by four elastomeric bearings. Each bearing is in the size of 0.3
m x0.5 m (longitudinal by transverse) in plan and has elastomeric height
of 63 mm. The pier footings are founded on stiff clay with undrained
shear strength of 150 kPa. The footing of pier of this reference bridge is
in square shape of size 8 m by 8 m, while the abutment consists of
retaining wall of 9 m in height and 1.5 m in thickness. The grade of
concrete is C-25 and the grade of steel is S500.
3.1. The reference bridge with new rocking isolation technique
With the new rocking isolation technique, the foundation of the
reference bridge has pier footing of size 4 m by 4 m rocking on elasto­
meric pads and external restrainers (Fig. 5) resting on the same stiff clay.
The horizontal sliding of footing is avoided by means of stoppers to
prevent the walking off phenomenon. For future replacement of
damaged bearings, the jack locations are shown in Fig. 5c. A detailed
information about pads replacement procedure is given in Appendix B.
The diameter of flat jack is based on the capacity of jack as mentioned in
flat jack manual [41].
Fig. 3. A typical plan showing location of jack for future replacement of pads.
4
Soil Dynamics and Earthquake Engineering 149 (2021) 106827
4. Numerical modelling of the two analysed bridge systems
The proposed rocking bridge and the reference bridge are modelled
in finite element computer programme [42] as shown in Figs. 6 and 7
respectively. A nonlinear stage construction case for the proposed
rocking bridge is defined in which elastomeric pads are deformed first
due to self-weight of pier, footing and dead load of the superstructure.
The nonlinear time history analysis was performed using the stiffness at
end of this nonlinear stage construction case. The superstructure was
modelled using frame elements as per the cross-section of the reference
bridge. The dead load mass of the superstructure was activated during
the construction stage analysis performed in the computer programme.
To simulate post-yield behaviour of piers, a concentrated plastic hinge is
assigned to the frame element of the pier. Deformation beyond the
elastic limit occurs only within the hinges modelled at the top and the
bottom of pier for fixed and rocking pier. Inelastic behaviour is obtained
through integration of the plastic strain and plastic curvature which
occurs within the pre-defined hinge length [4]. To capture the coupled
axial and bending behaviour, hinge is assigned to the top and bottom of
the piers (with the input hinge model being the moment-curvature graph
obtained from the section designer of the computer programme (CSi
Bridge) which adopts Caltrans seismic design criteria [43]. The
confinement of reinforced concrete sections has been taken into account
using the confined model [44] to represent the stress-strain behaviour of
the concrete core.
5. Elastomeric pads supporting the pier footing
A total of 13 pads (Fig. 5 c) with five at each end of footing and three
at centre of footing are modelled using frictional isolators available in
the chosen computer programme. The elastomeric bearings at abutment
ends are also modelled using these isolators and are connected to the
deck by spring element. The frictional isolators have coupled friction
properties for shear deformations and carries only compression. This
particular non-linear link element was chosen since the footing simply
rests on bearings and thus compression is only allowed in the link
element while friction between pad and footing is modelled by setting
coefficient of friction as 0.8. The frictional isolator model is based on the
hysteretic behaviour proposed by Wen [45] and recommended for base
isolation by Nagarajaiah [46]. The size of bearing is 455 mm × 455 mm
with height of 155 mm which consists of four rubber layers of 35 mm in
thickness and three steel shims of 5 mm in thickness, which was selected
after the design of pad was made as per Eurocode [47]. This particular
size and the number of bearings were chosen to control large vertical
initial compression of pad under effects of permanent loads. A detailed
design of the pads is given in Appendix A. The gap elements are used
between the stopper and side of footing to simulate the cushion in the
form of rubber pad which is attached vertically alongside of the stopper.
The vertical (kv) and horizontal stiffness (kh) of elastomeric pads is
evaluated using following equations [48]:
Ec . A
kv = ∑ ; (1)
ti
G .A
kh = ∑ ; (2)
ti
′
Ec. B
Ec = ; (3)
Ec′ + B
(4)
′
Ec = 6.73 GS2
where, A= area of pad; G is the shear modulus of the bearing with a
value of 0.7 MPa; B is the bulk modulus as 2000 MPa, Ec = Instanta­
neous compression modulus; E’c= Effective compression modulus with
incompressibility effect and S is the shape factor which for square pad is
R. Rele et al. Soil Dynamics and Earthquake Engineering 149 (2021) 106827

Fig. 4. Details of Overpass Bridge of the Attiki Odos motorway used as reference bridge.

Fig. 5. The reference bridge with new rocking isolation technique (a) Elevation (b) Detail Y (c) Footing Plan.

5
R. Rele et al. Soil Dynamics and Earthquake Engineering 149 (2021) 106827

Fig. 6. Finite element model of the reference bridge.

Fig. 7. Finite element model of the proposed bridge.

a/4t where “a" is the size of rubber pad and t is the thickness of each pad.
The initial vertical deformation of pad with superstructure load is found
to be 12.2 mm and the stress in the pad is found to be 4.45 MPa, which
falls within the permissible stress of 10–25 MPa mentioned in Eurocode
[47].
6. Modelling of shape memory alloy bars as external restrainers
SMAs are unique materials with a paramount potential for various
application in bridges. The novelty of this material lies in its ability of
undergoing large deformations and return to its undeformed shape
through stress removal (super elasticity) or heating (shape memory ef­
fect). In particular, Ni–Ti alloys have distinct thermo mechanical prop­
erties including super elasticity, shape memory effect and hysteresis
damping.
The parameters to model shape memory alloy bars are shown in
Table 3. These properties are adapted from Desroches and Delemont
[49]. By using these stress-strain values, an analytical model with

6
R. Rele et al. Soil Dynamics and Earthquake Engineering 149 (2021) 106827

Table 3
Constitutive material properties for NiTi based SMA bar.
Parameters
Austenite to martensite starting stress (1)
Austenite to martensite finishing stress (2)
Martensite to austenite starting stress (4)
Martensite to austenite finishing stress (3)
Yield strain
Recoverable pseudo elastic strain
32 mm composed nickel and titanium (Ni–Ti) alloy were chosen on each
side of the footing and were connected from mid height of footing by stiff
Value arms. The SMA bars are activated after the initial deformation of pads
(13.9 mm) that occurs due to dead load of the bridge.
523 MPa
588 MPa
241 MPa 7. Modelling of soil
225 MPa
1.1% The load settlement curve for footing on stiff clay is obtained from
6.2%
Plaxis 3D [51] where a surface displacement at top of footing kept on
soil (stiff clay) of 20 m depth is applied as shown in Fig. 10. The bottom
boundary is fixed in all directions for loading purpose. This programme
uses plastic calculation and automatic load stepping to simulate the
indentation of the footing. The material model chosen was Mohr
Coulomb and with undrained condition as this model is ideally
well-defined for clayey soil. The input soil parameters adopted are
shown in Table 4. The elastic modulus and shear strength of soil was
varied with depth by the Plaxis programme. The load deformation curve
(Fig. 11) obtained from Plaxis is then assigned to the non-linear area
springs to represent soil at the base slab of the developed numerical
model with the values of curves being negative to represent only
compression behaviour of the soil.

8. Selection of earthquake excitations

Fig. 8. Analytical model of SMA restrainer of 32 mm diameter and unit length
used in CSi Bridge.
idealised stress-strain curve as shown in Fig. 8 is created in CSi Bridge to
model the SMA bar of 32 mm diameter with unit length. The stress-strain
relationship is characterised by an elastic region, a long horizontal
plateau, followed by a significant increase in stiffness which demon­
strates a high-level energy dissipation and a super elastic hysteresis.
From the free end of stiff-arm, the SMA bars in the form of link elements
are connected to the centre of the base slab as shown in Fig. 7. The SMA
bars are modelled using a double link element. The first element which is
multi linear plastic element (PLE) (Fig. 9a) uses Pivot model [50] to
define the hysteresis loop. The second element is a multi linear elastic
element (MLE) (Fig. 9b) which is used to shift the hysteresis loop away
from the origin. Three numbers of shape memory alloy bars of diameter
The reference bridge and the proposed bridge with new rocking
isolation technique were analysed for seven real accelerograms (Table 5)
compatible to ground Type C- 1dependent elastic spectra of the Euro­
code. The Type C also represents for stiff clay as mentioned in Eurocode
[52] and thus chosen for this study. The peak ground accelerations
(PGA) selected was 0.60 g to represent high seismic excitation by scaling
[53] the real accelerograms. The response spectra of the analysed
accelerograms are shown in Fig. 12. The acceleration time histories are
shown in Figs. 13 (a) to 13 (g) for 0.6 g and are matched to ground Type
C-dependent Eurocode 8-1 elastic spectra using Seismo Match pro­
gramme [54]. These time histories are imposed at the base of the model,
which is free to move along x-x axis which is the longitudinal direction
of the bridge. The duration of all the time history analyses is 90 s to
allow for the pier models to balance after the ground motions for
assessing their post-earthquake condition. The horizontal seismic mo­
tion is applied in the longitudinal direction as well as in the transverse
direction of both the bridges considered in this study.
9. Results and discussion
The effective fundamental period for proposed rocking bridge was

Fig. 9. Modelling of SMA bar (a) Multi linear elastic link element (b) Multi linear plastic link element.

7
R. Rele et al.
Fig. 10. Soil model in Plaxis 3D.
Table 4
Soil parameters.
Parameters
Shear strength, Su
Young’s modulus, E
Bulk unit weight, γ
Fig. 11. Load settlement curve of footing on stiff clay.
1.49 s while for reference bridge was 1.07 s. The fundamental period of
vibration of the proposed system is found to be about 40% more than the
conventional one since the elastomeric pads provided beneath the pier
footing enhanced the flexibility of the rocking bridge system. The first
three modes of both bridges are shown in Figs. 14 and 15. The modal
participating mass ratio is 0.92 for the first mode of the proposed
rocking bridge while is 0.86 for the reference bridge. Both the systems
have frame action in the longitudinal direction while cantilever action in
the transverse direction. The elastomeric bearings (at abutment ends)
are free to rotate and thus larger mass participation is observed in
Soil Dynamics and Earthquake Engineering 149 (2021) 106827
second mode shape of rotation. The mean values of responses of pier,
which are calculated based on the seven acceleration time histories for
PGA of 0.6 g for the reference bridge and the proposed rocking bridge in
Values both the directions of the bridge are shown in Tables 6 and 7.
150 kPa
9000 kPa 9.1. Comparison of bending moments, shear forces and axial forces in
18kN/mᶟ
pier
The bending moment and shear force in the piers of rocking bridge
are reduced by 45% and 16% respectively as compared to the conven­
tional system of fixed base pier when seismic is considered in the lon­
gitudinal direction of bridge. For transverse direction seismic, the
bending moment and shear force are both reduced by approximately
25% compared to the reference bridge considered in the study. The
variations of axial load for both the bridges were negligible when dy­
namic analysis was performed in both the directions of the bridges.
9.2. Comparison of residual and maximum drifts
The comparison of residual and maximum drift for both the bridges
in transverse and longitudinal direction are shown in Tables 8 and 9
respectively. The proposed rocking bridge reduces the residual drift by
81% in transverse direction and 16% percent in the longitudinal direc­
tion with respect to conventional pier as adopted in the reference bridge.
The maximum drift of the pier with new rocking isolation scheme is
increased by 9% in transverse direction and 50% percent in longitudinal
direction due to the flexibility of pads provided beneath the footing of
the proposed pier. The pier drift time history of the bridges is shown in
Fig. 16 for Kobe earthquake.
9.3. Comparison of pushover curve and ductility demand
Fig. 17 shows the results of displacement controlled non-linear static
pushover analyis of both the bridges in longitudinal direction and
transverse directions. The lower lateral force for yielding and failure of
rocking bridge as compared to the conventional bridge indicates that
rocking bridge is subjected to less damage than the conventional one. In
the transverse direction, both the bridges are stiff as the yield
displacement is approximately 1.5 times more that in longitudinal di­
rection. Based on the results from Table 6, a summary of the average
ductility demand is shown in Table 10. The ductility demand of
8
R. Rele et al. Soil Dynamics and Earthquake Engineering 149 (2021) 106827

Table 5
Characteristics of selected earthquake records.
Earthquake Country Date of event Station Magnitude PGA(g) Rrup(km) Fault type

Hollister United States 20/02/1988 Hollister array 3 5.45 0.52 13.11 Strike Slip
Imperial United States 15/10/1979 Imperial Valley 6.54 0.46 23.85 Strike Slip
Kobe Japan 16/01/1995 Kobe University 6.9 0.54 0.9 Strike Slip
Kocaeli Turkey 17/08/1999 Aydin 7.51 0.77 349.5 Strike Slip
Kozani Greece 13/05/1995 Kozani 6.5 0.14 79.4 Normal
Loma Prieta United States 17/10/1989 Apeel10-Skyline 6.93 0.5 41.88 Reverse Oblique
Northridge United States 17/01/1994 Northridge-17645 Saticoy 5.28 0.62 11.14 Reverse

Rrup = Closest distance to fault rupture.

adopted in this study. The footing uplift (at corner node) is observed
more in case of transverse excitation than the longitudinal one as seen in
Fig. 19 as frame action is available only in longitudinal direction while
cantilever action prevails in the transverse direction. Figs. 20 and 21
show the deflection contours in Uz direction in longitudinal and trans­
verse direction respectively.

9.6. Behaviour of SMA bar

Fig. 22 shows the superelastic behaviour of SMA bars in which the

Fig. 12. Response spectra of accelerograms compatible to ground Type C
dependent Euro code 8-1 elastic spectra (PGA 0.6 g).
proposed rocking bridge is 26% more than the conventional monolithic
pier construction adopted in current bridge design practice.
9.4. Comparison of acceleration displacement response spectra (ADRS)
The ADRS plot aims at evaluating the seismic performance point
where the capacity of the structure matches the demand during an
earthquake. If the performance point occurs within the central portion of
the capacity curve or closer to the point where elastic range of structure
ends then it is an indication that the structure would suffer less damage
during an earthquake [27]. Although, the existing bridge is located in
less seismic zone (Zone II-0.16 g) of Greece, still the highest seismic
coefficient as per the Greek code is considered to plot the accleration
displacement response spectra. The seismic coefficient considered is 0.4
(against a value of 0.36) for high seismic zone in Greece. The structural
behaviour type B (which is in close approximation of the existing site)
was chosen to plot the accleration displacement response spectra. From
the Fig. 18a and b, the spectral accleration (0.336 g) of the existing
bridge is 10% more than the rocking bridge (0.311 g) and thus damages
in the existing bridge would be more than the rocking bridge. However,
the spectral displacement of rocking bridge is 27% more than the con­
ventional one as the rocking bridge is subjected to more displacement as
also observed from the time history analysis performed.
9.5. Stresses in elastomeric pads and uplift of footing of rocking bridge
The maximum vertical stresses in pads under footing of rocking pier
for all seven-time histories are shown in Table 11. The average
maximum stress in pad is 12.58 MPa which is within the permissible
limit of 25 MPa as mentioned in Eurocode. Thus, the chosen configu­
ration of pad and its size suits the geometry of the reference bridge
process of yielding and successive increase in stiffness allows SMA bars
to dissipate large amount of energy with superelastic hysteresis. These
bars also reduce the pier displacement and excessive footing uplift for
bridges subjected to high seismic excitations. Also, it can be seen that
when the seismic excitation is in longitudinal direction the maximum
stress reached in the SMA bars is 579 MPa which is within the permis­
sible value of 588 MPa as mentioned in Table 3. This is not true when the
seismic excitation is in transverse direction where the maximum stress in
SMA bar reaches a value of 798 MPa, which is an indication that higher
diameter of bar would be required in transverse direction of bridge to
resist the seismic forces. Note that the bars only act in tension since they
have been designed for the same as no response in compression zone is
observed.
9.7. Comparison of moment rotation curves
Generally, loops passing always through the axis of origin is good
indication of self-centring capacity of the system. It has been observed
from the moment rotation curves plotted (Fig. 23) for the pier consid­
ered in this study that the proposed pier footing has good re-centring
capacity while the fixed pier footing did not show such a behaviour
where some residual rotation is observed. The proposed pier showed low
energy dissipation compared to the conventional fixed pier and this is an
acknowledge effect in rocking isolation of piers [55].
9.8. Cost and sustainability of the proposed system
The forces in pier such as bending moment and shear are reduced
since the isolated base rocks on elastomeric pads and leads to release of
forces as it loses contact with the pad while rocking. Thus, section of pier
can be reduced which is not the case when pier is designed using current
seismic design philosophy. For high seismic zones, there is a possibility
that a greater number of SMA bars will be required, especially in the
transverse direction. However, considering the features of the proposed
system such as negligible residual drifts and re-centring capability, the
cost factor can be overlooked for bridges subjected to severe seismic
excitations. In order to protect the proposed novel foundation from
environmental effects, a precast concrete box can be cast along all sides
so that the pads and restrainer do not suffer from climatic changes and
are intact and protected from any hazardous conditions. A solid steel/
concrete cover lid can be placed just beneath the base of pier having cut
out of the shape of pier so that entire system is well protected from se­
vere environmental effects.

9
R. Rele et al. Soil Dynamics and Earthquake Engineering 149 (2021) 106827

Fig. 13. Acceleration time histories matched to ground Type C-dependent Euro code 8-1 elastic spectra for 0.6 g (a) Imperial (b) Hollister (c) Kobe (d) Kocaeli (e)
Kozani (f)Loma Prieta (g) Northridge.

Fig. 14. Mode shapes of the reference bridge.

10
R. Rele et al. Soil Dynamics and Earthquake Engineering 149 (2021) 106827

Fig. 15. Mode shapes of the rocking bridge.

rocking isolation system based on numerical modelling. However, an

Table 6
Average of the maximum values from the seven-time history analysis in longi­
tudinal direction.
Parameter
Horizontal movement at pier top
(mm)
Residual drift (%) (P1/P2)
Axial forces (kN)in pier (max/
min)
Shear force (kN) in pier
Bending Moment (kN-m) in pier
Table 7
Average of the maximum values from the seven-time history analysis in trans­
verse direction.
Parameter
Horizontal movement at pier top
(mm)
Residual drift (%) (P1/P2)
Axial forces (kN)in pier (max/
min)
Shear force (kN) in pier
Bending Moment (kN-m) in pier
10. Conclusions
experimental verification would add more value to this proposed
research so that this concept of novel rocking isolation can be imple­
mented practically in the field of bridge engineering. Based on nonlinear
PGA 0.6 g(Longitudinal direction) dynamic time history analysis (considering seven earthquake records as
pier (Rocking pier (Reference per recommendations of Eurocode 8 [52]) and non-linear static push­
bridge) bridge) over analysis performed on the developed numerical models, the
128 85 following conclusions are drawn:
0.016/0.014 0.019/0.022
1. The bending moments are considerably reduced by 45% in the lon­
− 15848/-7174 − 16828/-5253
gitudinal direction and 25% in the transverse direction in the pro­
12741 15123 posed rocking pier as compared to the conventional fixed base pier
42007 76144 system.
2. The proposed rocking pier on pad with external restrainers has
excellent recentring capability since residual drifts are almost
negligible for high seismic excitation of 0.6 g and thus is a viable
solution for areas subjected to frequent seismic of high intensity.
3. The proposed rocking pier exhibit around 26% more ductile behavior
PGA 0.6 g(Transverse direction) than the conventional monolithic pier construction adopted in cur­
rent bridge design practice.
pier (Rocking pier (Reference
bridge) bridge)
4. SMA bars showed super elastic hysteresis controlling the pier hori­
zontal displacements and uplift of footing in the proposed rocking
299 274
bridge model. This is advantageous where bridges are located in
0.02/0.016 0.048/0.048 higher seismic zones
− 12016/-10056 − 14417/-9125 5. The proposed rocking bridge has better seismic performance than the
conventional as seen from acceleration displacement response
3107 4136 spectra where the spectral acceleration is less than for the conven­
28727 38528
tional one. The time period of the system was also enhanced which
makes rocking bridge more flexible and subjected to less damage
than the conventional bridge.

This paper proposes a novel controlled rocking isolation technique
for an overpass bridge. This new technique uses elastomeric pads at the
base of footing and shape memory alloy bars as restrainers to promote
controlled rocking of the pier. In this paper, the seismic performance of
conventional fixed base pier system is compared with the proposed new
11. Author Statement
Rajesh Rele: Conceptualization, Methodology, Software Validation,
Formal analysis, Writing - Original Draft Visualization.
Ranjan Balmukund: Supervision.

11
R. Rele et al. Soil Dynamics and Earthquake Engineering 149 (2021) 106827

Table 8
Residual and maximum drift comparison for both bridges in transverse direction.
Residual drift (%) Maximum drift (%)

(Rocking Bridge) (Reference Bridge) (Rocking Bridge) (Reference Bridge)

EQ-Transverse Pier P1 Pier P2 Pier P1 Pier P2 Pier P1 Pier P2 Pier P1 Pier P2

h =8.8 m h =8.8 m h =8.8 m h =8.8 m h =8.8 m h =8.8 m h =8.8 m h =8.8 m

Imperial 0.017 0.014 0.087 0.087 4.818 4.818 3.852 3.852

Hollister 0.004 0.004 0.022 0.022 2.398 2.398 2.523 2.523
Kocaeli 0.000 0.000 0.010 0.010 2.557 2.557 2.500 2.500
Kozani 0.009 0.009 0.001 0.001 2.864 2.864 2.477 2.477
Kobe 0.005 0.005 0.032 0.032 4.091 4.091 3.011 3.011
Loma Prieta 0.024 0.024 0.045 0.045 4.011 3.909 2.943 2.943
Northridge 0.006 0.005 0.141 0.141 3.034 3.034 4.511 4.511
Mean 0.009 0.009 0.048 0.048 3.396 3.382 3.117 3.117

Table 9
Residual and maximum drift comparison for both bridges in longitudinal direction.
Residual drift (%) Maximum drift (%)

(Rocking Bridge) (Reference Bridge) (Rocking Bridge) (Reference Bridge)

EQ-Longitudinal Pier P1 Pier P2 Pier P1 Pier P2 Pier P1 Pier P2 Pier P1 Pier P2

h =8.8 m h =8.8 m h =8.8 m h =8.8 m h =8.8 m h =8.8 m h =8.8 m h =8.8 m

Imperial 0.014 0.004 0.018 0.027 1.727 1.727 1.170 1.170

Hollister 0.015 0.007 0.007 0.005 1.193 1.205 0.841 0.841
Kocaeli 0.012 0.002 0.010 0.022 1.148 1.159 0.773 0.773
Kozani 0.007 0.003 0.020 0.009 1.193 1.205 0.807 0.807
Kobe 0.022 0.011 0.048 0.059 1.693 1.705 1.114 1.114
Loma Prieta 0.017 0.007 0.008 0.002 1.716 1.716 0.943 0.943
Northridge 0.012 0.023 0.020 0.030 1.500 1.500 1.182 1.182
Mean 0.014 0.008 0.019 0.022 1.453 1.460 0.976 0.976

Fig. 16. Drift time history for Kobe EQ (a) Transverse direction (b) Longitudinal direction.

S. Bhattacharya: Supervision. Declaration of competing interest

Liang Cui: Supervision, Resources Data Curation.
Stergios A. Mitoulis: Writing - Review & Editing, Resources Data The authors declare that they have no known competing financial
Curation. interests or personal relationships that could have appeared to influence
the work reported in this paper.

12
R. Rele et al. Soil Dynamics and Earthquake Engineering 149 (2021) 106827

Fig. 17. Pushover curve comparison for both the bridges (a) Longitudinal direction (b) Transverse direction.

Table 10
Average ductility of the pier from dynamic THA.
Parameter Rocking Pier Fixed base pier

Peak Drift (%) 1.46 0.976

Yield Drift (%) 0.51 0.43
Ductility Demand, μd 2.86 2.26

Fig. 18. Acceleration Displacement Response Spectra (a) Reference bridge (b) Rocking bridge.

Table 11
Maximum stresses in pad of the rocking bridge.
Maximum Stresses (MPa) in pad

EQ EQ-Longitudinal EQ-Transverse

Imperial 14.08 15
Hollister 11.43 11.41
Kocaeli 10.61 11.42
Kozani 11.14 12
Kobe 13.46 13.66
Loma Prieta 14.01 12
Northridge 12.97 12.57
Mean 12.53 12.58

13
R. Rele et al. Soil Dynamics and Earthquake Engineering 149 (2021) 106827

Fig. 19. Time history of footing uplift for rocking bridge pier footing for Kobe Earthquake 0.6 g in both directions of excitation.

Fig. 20. Deflected shape at t=6.30s for Kobe earthquake in longitudinal direction of excitation (a) Full bridge (b) at Pier P1

14
R. Rele et al. Soil Dynamics and Earthquake Engineering 149 (2021) 106827

Fig. 21. Deflected shape at t=8.72s for Kobe earthquake in longitudinal direction of excitation (a) Full bridge (b) at Pier P1

Fig. 22. Stress-strain curve for SMA bar in the rocking bridge for Kobe EQ (a) Longitudinal excitation (b) Transverse excitation.

15
R. Rele et al. Soil Dynamics and Earthquake Engineering 149 (2021) 106827

Fig. 23. Moment rotation curve (a) Reference Bridge (b) Rocking Bridge.

Appendix A

In this section, basic design check as per Eurocode EN 1337-Part 3 2005 [47] of the elastomeric pads under dead loads and live load in the form of
braking effect are shown as under:

Fig. A.1. Sectional elevational of pad showing steel layers.

Maximum vertical load on Pad under dead load, Fz = 830 kN (Pad 1); Fz =1400 kN (Pad2).
Pad Size 455 x 455 x155 mm; 4 rubber layers of 35 mm and 3 steel layers of 5 mm.

A.1 Check on buckling stability

/ ′
′ 2Ga S
Fz A < ;
3. Te
1.29 MPa <4.92 MPa.

A.2 Check on limitation of distortion

εcd + εqd + εad < 7

1.5Fz
∈ cd =
G.Ar.S
Vx
∈ qd =
Te
′ ′
a 2 αa + b 2 αb
∈ αd = ∑
2 tγi

Table A. 1
Check on limitation of distortion

Case Fz vx αa αb εcd εqd εαd (εcd+εqd+εαd<7)

N mm rad rad

DEAD LOAD 940000 0 0 3.10 0 0.052 3.15

BRAKING 970000 0 0 3.20 0 0.003 3.20

16
R. Rele et al. Soil Dynamics and Earthquake Engineering 149 (2021) 106827

A.3 Check on Rotational stability

Kr=Rotation factor =3

′ ′
a αa + b αb
Vz >
Kr
Fz
Vz =
Kv

Table A. 2
Check on rotational stability

Case αa αb Vz (a’αa + b’αb)/Kr

rad rad mm

DEAD LOAD 5.80E-04 0 17 0.0860

BRAKING 0.0014 0 17 0.2077

A.4 Steel plate thickness

2.6 Fz ti
ts >
Ar f y

ts = 2.6 x 830x1000x35/(455 x 455 x235) = 1.55 < 5 mm; ok.

Appendix B

Pads replacement procedure: In order to replace the pads, flat jacks are used. The flat jacks are placed for lifting of the bridge structure along with
footing as shown here in Fig B1. All jacks are to be connected to one manifold for equal lifting of all jacks. New pads must be kept ready for replacement
with old pads before lifting of bridge begins. The pads (weight about 82 kg) on periphery can be easily removed once the jacks are placed and the
system is lifted by 10–15 mm with help of hydraulic jacks. Bridge structure is now standing on flat jacks and old pad at centre of footing now needs to
be replaced. The central pad can be removed by pushing the pad with help of small horizontal jack pressing against the thrust block as shown in Fig B2.
The ram of jack gets extended by mechanical means which helps in displacing the central pad as shown in Fig. B3. As the extended capacity of jack is
over, additional precast concrete block is placed against the old thrust block for new location of the horizontal jack as shown in Fig.B4 to finally
remove the old pad. A new thrust block is cast at the other end as shown in Fig B5. New pad can now be placed at the location by pushing with the help
of the same horizontal jack pressing against the thrust block as shown in Fig.B6. Once the new pad is placed in the central position as shown in Fig.B7,
the thrust block can be demolished. The other pads in periphery are now placed at their locations. If the all positions of pads are satisfactory, then jacks
may be lowered so that footing now rests on the new pads.

Fig. B.1. The system being lifted with help of jacks.

17
R. Rele et al. Soil Dynamics and Earthquake Engineering 149 (2021) 106827

Fig. B.2. Horizontal thrust block and jack placed.

Fig. B.3. Ram of jack extended after pushing by mechanical means.

Fig. B.4. Central bearing being removed.

Fig. B.5. Placing of new thrust block at other end.

Fig. B.6. Central bearing being pushed to its position.

18
R. Rele et al. Soil Dynamics and Earthquake Engineering 149 (2021) 106827

Fig. B.7. Central bearing finally being placed in position.

References [25] EN 1998-2. Eurocode 8: design of structures for earthquake resistance–Part 2:

Bridges. Brussels: European Committee for Standardization; 2012.
[26] Vision SEAOC. Performance based seismic engineering Of Buildings, vols. I and ii:
[1] Housner GW. The behavior of Inverted pendulum structures during earthquakes.
conceptual framework. Sacramento (CA): Structural Engineers Association of
Bull Seismol Soc Am 1963;53(2):403–17.
California; 1995.
[2] Priestley MJN, Evison RJ, Carr AJ. Seismic response of structures free to rock on
[27] ATC 40. Seismic evaluation and retrofit of existing concrete buildings. Redwood
their foundations. Bull N Z Natl Soc Earthq Eng 1978;11(3):141–50.
City (Ca): Applied Technology Council; 1996.
[3] Beck JL, Skinner RI. The seismic response of a reinforced concrete bridge pier
[28] FEMA 356. Federal emergency management agency. 2008. Washington, [DC].
designed to step. Int J Earthq Eng Struct Dynam 1974;2:343–58.
[29] Dutta A, Mander JB. Seismic fragility analysis of highway bridges. In: Proc., center-
[4] Priestley M, Seible F, Calvi G. Seismic design and retrofit of bridges. New York:
to-center project workshop on earthquake engineering in transportation systems;
Wiley; 1996.
1999. p. 311–25. Tokyo.
[5] Mashal M, White S, Palermo. Concepts and developments for accelerated bridge
[30] Anastasopoulos I, Kontoroupi Th. Simplified approximate method for analysis of
construction and dissipative controlled rocking. In: 15th world Conference on
rocking systems accounting for soil elasticity and foundation uplifting. Soil Dynam
earthquake engineering; 2012 [Lisboa].
Earthq Eng 2014;56:28–43.
[6] Pollino M, Bruneau M. Seismic retrofit of bridge steel truss piers using a controlled
[31] Espinoza A, Mahin S, Jeremic B, Kutter B, Ugalde J. Rocking of bridge piers
rocking approach. J Bridge Eng 2007;12(5):600–10.
subjected to multi-directional earthquake loading. New York, USA: Fifth National
[7] Solberg K, Mashiko N, Mander JB, Dhakal RP. Performance of a damage-protected
Seismic Conference on Bridges & Highways; 2006.
highway bridge pier subjected to bidirectional earthquake attack. J Struct Eng
[32] Anastasopoulos I, Loli M, Georgarakos T, Drosos V. Shaking table testing of
2009;135(5):469–78.
rocking-isolated pier on sand. J Earthq Eng 2013;17:1–32.
[8] Rodgers G, Mander J, Chase J, Dhakal R. Beyond ductility: parametric testing of a
[33] Hung HH, Liu KY, Hui TH, Chang KC. An experimental study on the rocking
jointed rocking beam-column connection designed for damage avoidance. J Struct
response of bridge piers with spread foundations. Earthq Eng Struct Dynam 2011;
Eng 2015;10:C4015006. 1061/(ASCE) ST.1943-541X.0001318.
40:749–69.
[9] Palermo A, Pampanin S, Calvi GM. Use of ‘controlled rocking’ in the seismic design
[34] Rele RR, Dammala PK, Bhattacharya S, Balmukund R, Mitoulis S. Seismic
of bridges. In: Proc., 13th world conf. on earthquake engineering; 2004.
behaviour of rocking bridge pier supported by elastomeric pads on pile foundation.
Vancouver, Canada, Paper 4006.
Soil Dynam Earthq Eng 2019;124:98–120.
[10] Guo T, Cao Z, Xu Z, Lu S. Cyclic load tests on self-centering concrete pier with
[35] Titirla M, Zarkadoulas N, Mitoulis SA, Mylonakis G. Rocking isolation of bridge
external dissipators and enhanced durability. J Struct Eng 2015;10:1–15. 1061/
piers on elastomeric pads. In: 16th world conference on earthquake engineering
(ASCE) ST.1943-541X.0001357.
(16WCEE); 2017 [Santiago, Chile].
[11] Marriott D, Pampanin S, Palermo A. Quasi-static and pseudo-dynamic testing of
[36] Cardone D, Perrone G, Sofia S. Seismic response of simply supported deck bridges
unbonded post-tensioned rocking bridge piers with external replaceable
with auxiliary super elastic devices. J Procedia Eng 2011;14:2315–22.
dissipaters. Earthq Eng Struct Dynam 2009;38:331–54.
[37] Makris N, Vassiliou MF. The dynamics of the rocking frame. Seismic assessment,
[12] Trono W, Jen G, Panagiotou M, Matthew S, Claudia PO. Seismic Response of a
behavior and retrofit of heritage building and mouments. Comput Methods Appl
damage-resistant reentering post tensioned -HYFRC bridge column. J Bridge Eng
Sci 2015;37. https://doi.org/10.1007/978-3-319-16130-3_2.
2014;20(7):1–13.
[38] Motaref S, Saiidi M, Sanders D. Shake table studies of energy-dissipating segmental
[13] Dawood H, Mohamed G, Hewes J. Behavior of segmental precast posttensioned
bridge columns. J Bridge Eng 2014;19:186–99.
bridge piers under lateral loads. J Bridge Eng 2012;17(5):735–46.
[39] Anastasopoulos I, Kontoroupi Th. Simplified approximate method for analysis of
[14] Zatar W, Mutsuyoshi H. Reduced residual displacements of partially prestressed
rocking systems accounting for soil elasticity and foundation uplifting. Soil Dynam
concrete bridge piers. In: 12th world Conference on earthquake engineering; 2000
Earthq Eng 2014;56:28–43.
[Lisboa].
[40] Anastasopoulos I, Sakellariadis L, Agalianos A. Seismic analysis of motorway
[15] Gabriele G, Jose IR, Milena M, Vervelidis A. Seismic Behavior of posttensioned self-
bridges accounting for key structural components and nonlinear soil-structure
centering precast concrete dual shell columns. J Struct Eng 2012;4(141):1–11.
interaction. Soil Dynam Earthq Eng 2015;78:127–41.
[16] Hamid NH, Mander JB. Damage avoidance for buildings. KSCE J Civ Eng 2014;18
[41] Freyssinet flat jacks, reference manual, sustainable technology. France; 2016.
(2):541–8.
[42] Computers and structures, inc. Berkeley CA. Computers and Structures, Inc. CSI
[17] Liu R, Palermo A. Low damage design and seismic isolation: what’s the difference?
analysis and reference manual for CSI BRIDGE; 2017.
New Zealand society for earthquake engineering conference. 2015.
[43] Caltrans Bridge design specifications manual. California Department of
[18] Agalianos A, Psychari A, Vassiliou MF, Stojadinovic B, Anastasopoulos I.
Transportation; 1990.
Comparative assessment of two rocking isolation techniques for a motorway
[44] Mander J, Priestley M, Park R. Theoretical stress-strain model for confined
overpass bridge. Front Built Environ 2017. https://doi.org/10.3389/
concrete. J Struct Eng 1988;8(1804):1804–26. 10.1061/(ASCE)0733-9445(1988)
fbuilt.2017.00047.
114.
[19] Saiidi SM, Tazarv M, Nakashoji B, Varela S, Kavianipour F. Resilient and
[45] Wen YK. Method for random vibration of hysteretic systems. J Eng Mech Div ASCE
sustainable bridges of the future. Innovations on Bridges and soil-bridge interaction”
1976;102. N0.EM2.
IBSBI Athens. 2014 [Greece].
[46] Nagarajaiah S, Reinhorn AM, Constantinou MC. 3D-Basis:Nonlinear dynamic
[20] Mitoulis SA, Rodriguez JR. Seismic performance of novel resilient hinges for
Analysis of three-dimensional; base isolated structures:Part II ,techincal repeort
columns and application on irregular bridges. ASCE J Bridge Eng 2017;22(2).
NCEER91-0005. Buffalo: National Center for Earthquake Engineering Research,
[21] JRA Japanese Road Association. Specification for highway bridges, Part V. Seismic
State University of New York at Buffalo; 1991 [N.Y].
Design; 1996.
[47] EN 1337-3. Structural bearings. Part 3—elastomeric bearings. Brussels: European
[22] AASHTO. Standard design specification for the design of highway bridges. 16th Ed.
Committee for Standardization; 2005.
Washington, D.C.: American Association Of State Highway And Transportation
[48] Naeim F, Kelly JM. Design of seismic isolated structures: from Theory to Practice.
Officials; 1996.
Wiley; 1999.
[23] Greek EAK. Seismic code. Athens: Organization of Seismic Planning and Protection;
[49] Desroches R, Delemont M. Seismic retrofit of simply supported bridges using shape
2000.
memory alloys. Eng Struct 2002;24:325–32.
[24] IRC. 6 standard specifications and code of practice for road bridges: section: II, load
[50] Dowell RK, Seible FS, Wilson EL. Pivot hysteretic model for reinforced concrete
and stresses. Delhi, India: Indian Road Congress; 2014.
members. ACI Struct J 1998;95:607–17.

19
R. Rele et al. Soil Dynamics and Earthquake Engineering 149 (2021) 106827

[51] PLAXIS 3D. Reference manual. version 8. 2016. [54] SeismoMatch. Reference manual. Pavia, Italy: Seismsoft; 2016.
[52] EN 1998-1. Eurocode 8: design of structures for earthquake resistance–Part 1: [55] Rodgers G, Mander J, Chase J, Dhakal R. Beyond ductility: parametric testing of a
general Rules, seismic actions and rules for buildings. Brussel: European jointed rocking beam-column connection designed for damage avoidance. J Struct
Committee for Standardization; 2005. Eng 2015;10:C4015006. 1061/(ASCE) ST.1943-541X.0001318.
[53] Kalkan E, Chopra AK. Practical guidelines to select and scale earthquake records
for nonlinear response history analysis of structures. United States Geological
Survey; 2010.

20