See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/275459922

Performance-Based Seismic Design Principles and Structural Analysis of

Shanghai Tower

Article in Advances in Structural Engineering · April 2014

DOI: 10.1260/1369-4332.17.4.513

CITATIONS READS

16 6,297

4 authors, including:

Huanjun Jiang
Tongji University
147 PUBLICATIONS 1,611 CITATIONS

SEE PROFILE

All content following this page was uploaded by Huanjun Jiang on 05 September 2017.

The user has requested enhancement of the downloaded file.

 Performance-Based Seismic Design Principles and
Structural Analysis of Shanghai Tower

by

H.J. Jiang, X.L. Lu, X.J. Liu and L.S. He

Reprinted from

Advances in Structural Engineering

Volume 17 No. 4 2014

MULTI-SCIENCE PUBLISHING CO. LTD.

5 Wates Way, Brentwood, Essex CM15 9TB, United Kingdom
 Performance-Based Seismic Design Principles and
Structural Analysis of Shanghai Tower

H.J. Jiang1, 2,*, X.L. Lu1, 2, X.J. Liu1 and L.S. He1
1Research Institute of Structural Engineering and Disaster Reduction, Tongji University, Shanghai, China
2State Key Laboratory of Disaster Reduction in Civil Engineering, Tongji University, Shanghai, China

Abstract: Shanghai Tower is a 632 m high super tall building adopting innovative
steel-concrete hybrid mega frame-core tube-outrigger lateral structural system. Since
both of the height and irregularity of Shanghai Tower are far beyond the current
Chinese design code provision, non-prescriptive performance-based seismic design
(PBSD) approach is required to employ in the seismic design of this super tall
building. The general PBSD methodologies of code-exceeding tall buildings are
briefly introduced at first. According to PBSD principles of code-exceeding tall
buildings, the seismic performance objectives are selected for Shanghai Tower. The
design criteria for structural components and systems consistent with the performance
objectives are established. The elasto-plastic time-history analysis of the structure
under frequent, basic and rare earthquakes, respectively, is carried out with the aid of
ABAQUS program. Seven sets of ground motions are selected as the input motions.
The detailed numerical simulation results, such as the structural dynamic
characteristics, the displacement responses, the base shear, the overturning moment,
and the damage conditions, are presented. The structural analysis results indicate that
the structure can meet the predetermined performance objectives with relatively large
safety margin.

Key words: performance-based seismic design, performance objective, elasto-plastic time-history analysis, code-
exceeding tall building.

1. INTRODUCTION during earthquakes (Zou et al. 2008). In recent decades,

Conventional methods for seismic design are mainly
based on the strength to achieve the single performance
objective of life safety. From the past experience,
although the buildings designed according to current
codes performed well during earthquakes from the life
safety perspective, the structural and non-structural
elements were seriously damaged, leading to
unexpectedly high economic losses. One major
drawback of this traditional design approach is that it
does not directly address structural inelastic seismic
responses and thus cannot effectively deal with the
damage loss due to structural and non-structural failure
seismic design has been in the process of fundamental
change, with the emphasis changing from “strength” to
“performance” around the world. The need to improve
seismic performance of the structures through the
development of performance-oriented procedures and
guidelines has been repeatedly highlighted (Chandler
and Lam 2001). Up to now, the philosophy of
performance-based seismic design (PBSD) has been
sufficiently developed (SEAOC 1995, 1996; FEMA
2000). PBSD has been regarded as a more efficient and
applicable tool of designing structures, compared to the
traditional seismic design method. It has become a

*Corresponding author. Email address: jhj73@tongji.edu.cn; Fax: +86-21-65982668; Tel: +86-21-65986151.

Advances in Structural Engineering Vol. 17 No. 4 2014 513

 Performance-Based Seismic Design Principles and Structural Analysis of Shanghai Tower
modern approach to earthquake-resistant design
(Berahman 2013). In USA, a series of guidelines and
reports on performance-based approach for seismic
design and analysis of tall buildings have been
published since 2007 (SEAONC 2007; LATBSDC
2008; TBI Guidelines Working Group 2010). In current
Chinese design codes (GB50011 2010; JGJ3-2010), the
PBSD method is provided and recommended for the
buildings with stricter requirements than normal
buildings.
In mainland China in recent two decades, many tall
buildings have been constructed as a result of rapid
economic growth and urbanization. Owing to the
limited availability of land, the preference for
centralized services and the diverse needs of the owners,
users and society for commercial or aesthetic purpose,
the height of tall buildings has grown taller, and the
configuration as well as structural system has become
more complex, resulting in a large number of code-
exceeding tall buildings. The uniqueness in these
structures beyond the scope of current design codes
brings new challenges to engineers since the structural
behaviour of complex tall building is difficult to predict
and evaluate. The current design codes typically provide
minimum requirements for the design of code-
compliant structures to ensure life safety and prevent
structural collapse. As the code-exceeding buildings are
concerned, the procedures and requirements of such
non-prescriptive design have not been well defined.
However, in engineering practice the PBSD approach is
highly recommended for the seismic design of code-
exceeding tall buildings. For each project, a seismic
peer review panel shall be convened and provide an
independent objective, and technical review of those
aspects of the structural design for the building that
relates to seismic performance.
In this study, incorporating with the experience
gained from previous practice and current design codes
applied in Mainland China, the general methodologies
of PBSD analysis and design of code-exceeding tall
buildings are summarized briefly. Then following this
method, the PBSD principles and performance
objectives of Shanghai Tower are introduced. The 124-
story Shanghai Tower has a total architectural height of
632m while the total structural height is 580 m, adopting
the steel-concrete hybrid mega frame-core tube-
outrigger structural system. Since the total structural
height and irregularity exceed the limit specified in the
Chinese seismic design code, Shanghai Tower is
classified as a code-exceeding tall building. According
to the advice provided by the seismic peer review panel,
514
the PBSD method should be employed. To evaluate the
seismic performance and verify its performance
objectives, comprehensive numerical analysis is carried
out with the aid of ABAQUS program. The seismic
responses are obtained by elasto-plastic time-history
analysis.
2. GENERAL METHODOLOGIES FOR PBSD
OF CODE-EXCEEDING TALL BUILDINGS
Compared with traditional seismic design, the
significant characteristic of PBSD approach for code-
exceeding tall buildings is that this design procedure is
clear and transparent for designers and owners to
understand the seismic performance and inherent risks
of the structures under various levels of ground motions
during the expected life cycle and then select multiple
performance objectives accordingly. The most
important task for seismic design of code-exceeding
tall buildings is to demonstrate that the desired seismic
safety and performance objectives can be assured by
taking effective measures to counteract the negative
impacts exerted by the code-exceeding conditions. The
key structural components and potential weak positions
related to the code-exceeding conditions should be
identified and consequently additionally strengthened
so that they no longer fail first or suffer severe damage.
Good understanding of structural behaviour under the
earthquake is prerequisite to accomplish this task.
Good engineering practice and judgement are vital in
some cases. Sufficient evidence for the rationality of
the structural solutions and realization of the pre-
defined seismic performance objectives should be
provided by comprehensive analytical studies and/or
testing.
The design criteria should be established
corresponding to the desired performance objectives.
These minimum acceptance criteria ascertain that the
performance objectives could be accomplished. In
particular, the identified key components and potential
weak positions, and the corresponding limit values of
responses should be addressed so as to enhance their
seismic performance. The criteria are usually set in
terms of limit values of stress, load-carrying capacity,
deformation such as strain, plastic rotation, inter-story
drift ratio, etc. They can also be quantified by reliable
damage index, such as displacement ductility and
hysteretic energy.
The PBSD procedure consists of two design phases.
In the first phase, after the preliminary design is
completed with the basic configuration and structural
layout selected, the code-exceeding conditions are
Advances in Structural Engineering Vol. 17 No. 4 2014
H.J. Jiang, X.L. Lu, X.J. Liu and L.S. He

identified, and the seismic performance objectives are
determined accordingly. Furthermore, the key structural
components which are crucial to the seismic safety of
overall structure are identified and laid particular
emphasis. The design criteria are established to achieve
the desired performance objectives. Different
performance requirements should be proposed for
different types of structural components. The seismic
effects under the frequent earthquake and the effects of
other actions are determined on the basis of linear-
elastic behaviour. The dimensions and reinforcement of
structural members are derived by using the
conventional strength-based design code. The general
method for determining the seismic effects is the modal
response spectrum analysis using elastic design spectra.
In the second phase, the seismic performance of the
target building is evaluated by comprehensive
numerical analysis. Nonlinear analysis should be
properly substantiated with respect to the seismic input,
the constitutive model used, the method of interpreting
the results of the analysis and the requirements to be
met. The earthquake responses, plastic mechanism,
distribution of damage, etc., are estimated against the
preset allowable limit. For tall buildings which greatly
exceed the height limit or have very complex or unique
as well as innovative structural system without design
experience and referential bases, structural testing on
the joint, member, or full structural model is highly
recommended to conduct in order to study the structural
behaviour and check the seismic performance directly.
If the pre-defined seismic performance objectives could
not be satisfied, design iteration should be done until
satisfied. The flowchart of the general PBSD procedure
is shown in Figure 1.

Start

Define initial input parameters and perform preliminary design

(Structural layout, structural member sizes, site soil condition, material properties, etc.)

Identify code-exceeding conditions and select seismic performance objectives

(Earthquake design level, performance level, importance of the building)

Establish design criteria

(Identify key structural components and potential weak positions)

Perform elastic analysis for non-seismic actions and frequent earthquake with elastic design spectra

Design components for required strength following current strength-based seismic design code

Evaluate seismic performance under different earthquake level Alter design parameters
(Numerical analysis, structural testing)

No
Performance objectives are satisfied ?

Yes

End

Figure 1. Flowchart of performance-based seismic design procedure

Advances in Structural Engineering Vol. 17 No. 4 2014 515

 Performance-Based Seismic Design Principles and Structural Analysis of Shanghai Tower
3. STRUCTURAL DESCRIPTION OF
SHANGHAI TOWER
Shanghai Tower is a multi-functional skyscraper located
in Lujiazui Financial and Trade District in Shanghai,
China, which is mainly used for office, hotel,
commerce, tourism and other purposes (as shown in
Figure 2). Due to the requirement of building function,
the building is divided into 9 zones along the vertical
direction (as shown in Figure 3). The structural plan
layout of typical floors in each zone is shown in Figure
4. The steel-concrete composite superstructure resists
lateral loads with a central reinforced concrete shear
wall core interconnected with the composite mega-
frame through six two-story high outrigger trusses.
Gravity loads are resisted by steel-concrete composite
floor system. Considering the requirement of
architectural function, from zone 5 the four corners of
core are removed and the left core is cruciform. The
mega-frame consists of eight mega-columns, four
corner columns, and eight circular two-story high belt
trusses distributed rough evenly in lower eight zones.
The stories containing the belt trusses are regarded as
stiffened stories. The structural plan layout of 19F
(normal story) and 22 F (stiffened story) is shown in
Figures 5 and 6, respectively. The eight mega-columns
rise to the top of zone 8 while the four corner columns
Figure 2. Perspective view of Shanghai Tower (Thornton
Tomasetti Inc. 2009)
516
Z9:Sightseeing
/facility
Z8:10 floors
Boutique office
Z7:15 floors
Hotel/Service
Z6:14 floors
office
Z5:14 floors
office
Z4:13 floors
office
Z3:13 floors
office
Z2:12 floors
office
Z1:Podium
Underground
Figure 3. Vertical zoning of Shanghai Tower
end at the top of zone 5. The mega-columns vary in
cross-section from 5.3m by 3.7 m at the bottom to 2.4 m
by 1.9 m at the top. The belt trusses also act as transfer
members transmitting the vertical loads on the floors
between two adjacent stiffened stories to mega-
columns. In each zone two parallel circular trusses are
set to enhance the torsional resistance. All the trusses
are composed of H-shaped steel members. The amount
and location of outrigger trusses are optimized. Six
outrigger trusses are determined to be distributed in
zones 2, 4, 5, 6, 7, 8, respectively. The layout of all main
components in the stiffened stories is shown in Figure 7.
Twenty-one one-story high radial trusses are installed in
the stiffened stories to transmit the vertical loads from
outer curtain walls and floors to belt trusses. The plan
layout of radial trusses is shown in Figure 8.
According to the current Chinese code, the following
items are identified as code-exceeding: (1) the total
structural height of 580 m exceeds the limit of 190 m
stipulated for steel-concrete composite frame-core tube
structures; (2) the elevation irregularity including
several stiffened and transfer stories exceeds the code
Advances in Structural Engineering Vol. 17 No. 4 2014
H.J. Jiang, X.L. Lu, X.J. Liu and L.S. He

(a) Zone 1−4 (b) Zone 5

(c) Zone 6 (d) Zone 7

Figure 6. Plan layout of 22F

Outrigger truss

Core wall
(e) Zone 8 (f) Zone 9

Figure 4. Structural plan layout of typical floors in each zone

Mega-column
Belt truss Corner column

Figure 7. Main components in stiffened stories

limit; (3) the overhanging length of 14 m of the radial

truss in the stiffened stories exceeds the limit of 4m and
10% of the total span.

4. PERFORMANCE OBJECTIVES AND

Figure 5. Plan layout of 19F
DESIGN CRITERIA OF SHANGHAI TOWER
Seismic performance objectives are defined as the
coupling of expected performance levels with expected
levels of seismic ground motions. Due to the fact that
both of the height and irregularity of Shanghai Tower are
far beyond the code specification, the enhanced seismic

Advances in Structural Engineering Vol. 17 No. 4 2014 517

 Performance-Based Seismic Design Principles and Structural Analysis of Shanghai Tower
Figure 8. Plan layout of radial trusses
performance objectives, stricter than those for ordinary
buildings, are adopted according to the advice of the
seismic peer review panel, expressed as follows: fully
operational under frequent earthquakes, operational
under basic earthquakes, and life safety under rare
earthquakes. The following stories are identified as the
potential weak stories: the lower stories located in zones
1 and 2, the stiffened stories, and the stories one story
lower or higher than the stiffened stories. The following
structural components are identified as key components
crucial to the seismic safety of the overall structure: belt
trusses, outrigger trusses, mega-columns, the core walls
of the above-mentioned weak stories, and the joints
connecting the above members. The flowing structural
components are identified as common vertical
components: the core walls of other stories and the
components of the top crown. The left are regarded as
energy-dissipation components.
The design criteria for structural components and
systems consistent with the above performance
objectives are established, which set individual
requirements on inter-story drift ratio, loading-carrying
capacity, and damage state of different types of
structural components under each earthquake level.
Under frequent earthquakes, all structural components
perform elastically. The inter-story drift ratio should not
be larger than 1/500. Under basic earthquakes, the
structure performs rough elastically. The key
components should keep elastic. The common vertical
components should be prevented from yielding. The
yielding of energy-dissipation components is allowed.
518
The inter-story drift ratio should not be larger than
1/200. Under rare earthquakes, the structure is
moderately damaged. The key components should be
prevented from yielding. As the common vertical
components are concerned, although the yielding is
allowed, the total external load effect should be less than
the ultimate load-carrying capacity, and shear failure is
not allowed. The local collapse as well as overall
collapse should be prevented. The inter-story drift ratio
should not be larger than 1/100.
The seismic strengths of structural components are
checked against the demands under the basic earthquake
and rare earthquake. Under the basic earthquake, the
seismic effects can be determined by the modal response
spectrum analysis using elastic design spectra based on
the fact that the structure keeps rough elastic. Under the
rare earthquake, the seismic effects can be determined
by the simplified equivalent linear method through the
increasing of the damping ratio and the decreasing of the
stiffness of the structure. The design should be modified
until the performance requirements are satisfied. To
check the requirement of keeping elastic for the
structural component, the following equation is applied:
1.2SGE + 1.3βE1SEk < R/ γ RE (1)
where SGE is the effect of the representative value of
gravity load; SEK is the effect of the standard value of
frequent earthquake; βE1 is the ratio of the PGA of the
considered earthquake level to that of the frequent
earthquake; R is the design value of the load-carrying
capacity; and γ RE is the seismic modification coefficient
of load-carrying capacity. The following equation is
applied to check the requirement of unyielding:
SGE + βE1SEk < Rk (2)
where Rk is the standard value of the load-carrying
capacity.
5. ELASTO-PLASTIC TIME-HISTORY
ANALYSIS OF SHANGHAI TOWER
5.1. Numerical Model
5.1.1. Element type
The numerical model of Shanghai Tower is constructed
with the aid of commercial software ABAQUS. The
quadrilateral shell element denoted by SR4 is used to
model mega-columns, coupling beams, and core walls.
The beam element denoted by B31 is used to model the
steel plates which are embedded in the mega-columns
and core walls. The steel bars in mega-columns,
coupling beams, and core walls are modelled by steel
Advances in Structural Engineering Vol. 17 No. 4 2014
H.J. Jiang, X.L. Lu, X.J. Liu and L.S. He

Tensile damage factor-strain relationship

3.0 1.0

2.5

Tensile damage factor

Tensile stress(MPa)
2.0

1.5 0.5

1.0
Tensile stress_strain relationship
0.5

0.0 0.0
0.0000 0.0003 0.0006 0.0009 0.0012 0.0015
Tensile strain

Figure 10. Tensile stress-strain and damage-strain relationships of

concrete

Compressive damage factor-strain relationship

40 1.0

35

Compressive damage factor

Compressive stress(MPa)

z 30
y
25
x
20 0.5

15
Figure 9. Global numerical model
10 Compressive stress-strain
relationship
5
rebar layer. The steel components such as steel beams 0 0.0
and columns are modelled by the beam element denoted 0.000 0.004 0.008 0.012 0.016
Compressive strain
by B31. Secondary components such as floor beams and
steel columns which are used just for withstanding Figure 11. Compressive stress-strain and damage-strain
vertical loads are ignored in the global numerical model. relationships of concrete
Beam-core wall connections are modelled as fixed. The
assumption of rigid floor is adopted. The global
numerical model is shown in Figure 9.
σt

5.1.2. Constitutive model of material σ t0

The plastic damage model based on continuum damage

mechanics, developed by Lee and Fenves (1998), is
adopted for concrete. Two damage patterns, tensile
cracking and compressive crushing, are considered.
Tensile damage factor dt and compressive damage ε
factor dc are used to measure the damage level of two
damage patterns and decrease the stiffness
accordingly. The stress-strain and damage-strain
relationships of concrete with the strength grade C60
are shown in Figures 10 and 11. The stress-strain
hysteretic models for the concrete and the steel are Figure 12. Stress-strain hysteretic model for concrete (Hibbitt
shown in Figures 12 and 13. Bilinear isotropic et al. 2004)

Advances in Structural Engineering Vol. 17 No. 4 2014 519

 Performance-Based Seismic Design Principles and Structural Analysis of Shanghai Tower

σ 5.3. Earthquake Ground Motions

To evaluate the seismic performance of the structure by
nonlinear time-history analysis, appropriate earthquake
ground motions should be selected as the seismic
excitation. Seven sets of three-dimensional earthquake
records including five sets of natural motions and two
sets of artificial motions are selected according to current
Chinese seismic design code. The site soil condition of
the selected motions is similar to that of the construction
ε
site which is classified as Class IV (soft soil). Five sets
of natural earthquake records are from 1971 San
Fernando Earthquake, 1968 Borrego Mountain
Earthquake, and 1985 Mexico Earthquake. The
comparison between the design spectrum and the
spectrum of horizontal component of individual motions
is shown in Figure 15 (PGA = 100 gal). The mean
spectrum is consistent with the design spectrum
Figure 13. Stress-strain hysteretic model for steel (Hibbitt et al. 2004) specified in the design code.
Three levels of seismic hazard, minor or frequent
earthquakes with the exceeding probability of 63.2% in
kinematic hardening model is adopted for the steel. 50 years (50 year return period), moderate or basic
The Bauschinger effect is taken into account, and there earthquakes with the exceeding probability of 10% in 50
is no stiffness degradation between different cycles. years (475 year return period), and strong or rare
The post-yield stiffness is taken as 1/100 of the elastic earthquakes with the exceeding probability of 2% in 50
stiffness in consideration of the material hardening years (2475 year return period), are considered here
effect. according to current Chinese seismic design code. The
seismic protection intensity of Shanghai is seven. The
5.2. Dynamic Characteristics peak ground acceleration (PGA) of frequent earthquakes,
Prior to the nonlinear time-history analysis, the basic earthquakes, and rare earthquakes with intensity 7
modal analysis is conducted to get the natural is 35, 100, and 200 gal, respectively. The three
vibration characteristics of the structure. The first components are scaled with the PGA ratio of 1: 0.85:
50 vibration modes of the structure at initial elastic 0.65 between the primary horizontal component,
state are obtained by Lanczos Method. The secondary horizontal component, and vertical
characteristics of the first 6 modes are listed in component according to Chinese code.
Table 1. Figure 14 shows the mode shape of the first
three modes. 5.4. Earthquake Responses
5.4.1. Displacement responses
The responses of floor displacement in Y direction are
Table 1. Characteristics of the first 6 vibration slightly larger than those in X direction. The maximum
modes roof displacements and maximum inter-story drift ratios
in Y direction are shown in Table 2. The envelops of
Number of Vibration Characteristic
floor displacements and inter-story drift ratios are
mode period (s) of vibration
shown in Figures 16 and 17, respectively.
1 8.85 The first-order translation The maximum inter-story drift ratios under the three
in X direction
levels of ground motions are 1/582, 1/248, and 1/134,
2 8.33 The first-order translation
in Y direction satisfying the requirements of above performance
3 3.28 The first-order torsion objectives. Under frequent earthquakes, the maximum
4 3.14 The second-order translation inter-story drift ratio occurs at the 110th floor which is at
in X direction the middle part of Zone 8. Under basic earthquakes, the
5 3.10 The second-order translation maximum inter-story drift ratio occurs at the 109th floor
in Y direction
6 1.62 The second-order torsion
which is at the middle part of Zone 8. Under rare
earthquakes, the maximum inter-story drift ratio occurs at

520 Advances in Structural Engineering Vol. 17 No. 4 2014

H.J. Jiang, X.L. Lu, X.J. Liu and L.S. He

z z z

z z z
x
y x
y x y x
y x y
y x

(a) The first mode (b) The second mode (c) The third mode
(translation in Y) (translation in X) (torsion)

Figure 14. Mode shape for the first three modes

MEX006 US724 generally, the displacement responses are the largest under
4 MEX007 US725 the motions of MEX006-008, especially when the motions
US1214 US256
of MEX007 are inputted in the principal direction.
Spectrum acceleration/m.s−2

US1215 US257
SHW3 US334
3
S79010 US335 5.4.2. Base shear and overturning moment
S79011
Mean spectrum
The time history of base shear and overturning moment
2 Design spectrum responses under the ground motions of MEX006-008
with the principal component MEX007 acting in Y
direction is shown in Figures 18 and 19, respectively. It
1
is shown that when the base shear of the structure
reaches the maximum, the core walls carry 60% of the
0 total shear fore of the structure, while 30% of
0 2 4 6 8 10 the overturning moment at the base is resisted by the
Period/s core walls when the base overturning moment of the
Figure 15. Comparison of spectra of selected earthquakes with structure reaches the maximum.
design spectrum
5.4.3. Damage conditions
Under frequent earthquakes all structural components
the 92nd floor which is at the middle part of Zone 7. The remain elastic and intact. Under basic earthquakes, in
inter-story drift ratios decrease significantly in the stiffened the mega-columns only slight tensile damage of
stories since the stiffness there is much larger than that of concrete occurs at the height of connecting with
the adjacent two stories. As shown in Figures 16 and 17, outrigger trusses, no compressive damage of concrete
there is large difference between the displacement occurs, and the mega-columns keep roughly elastic.
responses excited by different ground motions. In Similarly, in the core walls slight tensile damage of

Advances in Structural Engineering Vol. 17 No. 4 2014 521

 Performance-Based Seismic Design Principles and Structural Analysis of Shanghai Tower

Table 2. Roof displacements and inter-story drift ratios in Y direction

Location of
maximum
Level of seismic Earthquake Roof displacement Maximum inter- inter-story drift
hazard record (mm) story drift ratio ratio (floor)
Frequent earthquakes MEX006-008 489 1/582 110
US334-336 207 1/753 111
US1213-15 255 1/810 111
US256-258 206 1/941 111
US724-726 286 1/872 110
S79010-12 389 1/620 111
SHW3 481 1/763 110
Basic earthquakes MEX006-008 1323 1/258 92
US334-336 538 1/257 109
US1213-15 779 1/272 109
US256-258 632 1/315 110
US724-726 743 1/319 109
S79010-12 984 1/248 109
SHW3 1216 1/298 109
Rare earthquakes MEX006-008 2457 1/134 92
US334-336 964 1/164 109
US1213-15 1495 1/152 109
US256-258 1207 1/169 109
US724-726 1364 1/168 108
S79010-12 1741 1/164 93
SHW3 2053 1/173 109

120 120 120

100 100 100

80 80 80
Floor
Floor

Floor

60 60 60
US256-258 US256-258 US256-258
US334-336 US334-336 US334-336
40 40 40
S79010-12 S79010-12 S79010-12
US1213-15 US1213-15 US1213-15

20 US724-726 20 US724-726 20 US724-726

MEX006-008 MEX006-008 MEX006-008
SHW3 SHW3 SHW3
0 0 0
0 250 500 0 250 500 0 250 500
Displacement/mm Displacement/mm Displacement/mm

(a) Under frequent earthquakes (b) Under basic earthquakes (c) Under rare earthquakes

Figure 16. Envelops of floor displacement response in Y direction

concrete occurs at the stiffened stories and the adjacent
stories one story lower or higher than the stiffened
stories, no compressive damage of concrete occurs, and
the core walls keep roughly elastic. The plastic
deformation occurs in some coupling beams in the
middle part and upper part of the core walls. All the
components in the outrigger trusses and belt trusses
remain at the stage of elasticity.
Under the rare earthquake of MEX006-008, the
structure suffers the most severe damage. The concrete
damage state of mega-columns and core walls
subjected to MEX006-008 is shown from Figures 20 to
22, and the Mises stress nephograms of steel members
in main components are shown in Figure 23. The
yielding strength of the steel is 345 MPa. In the mega-
columns very slight compressive damage of concrete

522 Advances in Structural Engineering Vol. 17 No. 4 2014

H.J. Jiang, X.L. Lu, X.J. Liu and L.S. He

120 120 120

100 100 100

80 80 80
Floor

Floor

Floor
60 60 60

40 US256-258 40 US256-258 40 US256-258

US334-336 US334-336 US334-336
S79010-12 S79010-12 S79010-12
US1213-15 US1213-15 US1213-15
20 20 US724-726 20
US724-726 US724-726
MEX006-008 MEX006-008 MEX006-008
SHW3 SHW3 SHW3
0 0 0
0 1/1000 1/500 0 1/400 1/200 0 1/200 1/100
_ _ _
Inter story drift ratio Inter story drift ratio Inter story drift ratio

(a) Under frequent earthquakes (b) Under basic earthquakes (c) Under rare earthquakes

Figure 17. Envelops of inter-story drift ratio in Y direction

4 4
Base shear force (105 kN)

Base shear force (105 kN)

2 2

0 0

−2 −2
Total shear force Total shear force
Mega-columns Mega-columns
Core walls Core walls
−4 −4
0 10 20 30 40 50 0 10 20 30 40 50
Time (s) Time (s)
(a) X direction (b) Y direction

Figure 18. Time history of base shear

1.0 1.0
Overturning moment (108 kN.m)

Overturning moment (108 kN.m)

0.5 0.5

0.0 0.0

−0.5 −0.5
Total overturning moment Total overturning moment
Mega-columns Mega-columns
Core walls Core walls
−1.0 −1.0
0 10 20 30 40 50 0 10 20 30 40 50
Time (s) Time (s)
(a) X direction (b) Y direction

Figure 19. Time history of base overturning moment

Advances in Structural Engineering Vol. 17 No. 4 2014 523

 Performance-Based Seismic Design Principles and Structural Analysis of Shanghai Tower

appears, and severe tensile damage of concrete occurs,
but the steel bars and steel plates keep elastic. The
outrigger trusses and belt trusses remain elastic. Both
severe tensile and compressive concrete damage occurs
in considerable coupling beams. The steel bars and
steel plates embedded in the core walls are still in the
elastic stage. In core walls severe tensile damage of
concrete occurs at the stiffened stories and several
stories adjacent to the stiffened stories. Severe
compressive damage of concrete occurs in very few

DAMAGAC
DAMAGAC Max: 0.095 SNFG, (fraction = −1.0)
SNFG, (fraction = −1.0) (Avg: 75%)
(Avg: 75%)
0.095
0.095 0.087
0.087 0.079
0.079 0.071
0.071 0.064
0.063 0.056
0.055 0.048
0.048 0.040
0.040 0.032
0.032 0.024
0.024 0.016
0.016 0.008
0.008 0.000
0.000 Max: 0.095
Max: 0.095
Elem: SHZX−1.1845
Elem: SHZX−1.9662
Node: 11240
Node: 15444

max: 0.095

z y
z y

x
x

(a) Compressive damage in Zone 1−4 (b) Compressive damage in Zone 5−8

DAMAGET DAMAGET DAMAGET

SNFG, (fraction = SNFG, (fraction = SNFG, (fraction =
(Avg: 75%) (Avg: 75%) (Avg: 75%)
0.896 0.967 0.969
0.822 0.887 0.889
0.747 0.806 0.808
0.672 0.725 0.727
0.598 0.645 0.646
0.523 0.564 0.565
0.448 0.484 0.485
0.373 0.403 0.404
0.299 0.322 0.323
0.224 0.242 0.242
0.149 0.161 0.162
0.075 0.081 0.081
0.000 0.000 0.000

y
z x

z z z

x y x y x y

(c) Tensile damage in Zone 1−3 (d) Tensile damage in Zone 4−6 (e) Tensile damage in Zone 7−8

Figure 20. Damage state of mega-columns under rare earthquake

524 Advances in Structural Engineering Vol. 17 No. 4 2014

H.J. Jiang, X.L. Lu, X.J. Liu and L.S. He

dc dt
0.989 0.984
dc dt 0.907 0.902
0.991 0.985 0.824 0.820
0.908 0.903 0.742 0.738
0.826 0.821 0.659 0.656
0.743 0.739 0.577 0.574
0.661 0.657 0.494 0.492
0.578 0.575 0.412 0.410
0.496 0.493 0.330 0.328
0.413 0.410
0.330 0.328 0.247 0.246
0.248 0.246 0.165 0.164
0.165 0.164 0.082 0.082
0.083 0.082 0.000 0.000
0.000 0.000

(a) Compressive damage in (b) Tensile damage in (c) Compressive damage (d) Tensile damage in
Zone 1−4 Zone 1−4 in Zone 5−9 Zone 5−9

Figure 21. Damage state of core walls and coupling beams under rare earthquake

dc dc dt
dt 0.989
0.991 0.984
0.985 0.907
0.908 0.902
0.903
0.826 0.821 0.824 0.820
0.743 0.742 0.738
0.739
0.661 0.657 0.659 0.656
0.578 0.575 0.577 0.574
0.496 0.493 0.494 0.492
0.413 0.410 0.412 0.410
0.330 0.328 0.330 0.328
0.248 0.246 0.247 0.246
0.165 0.164 0.165 0.164
0.083 0.082 0.082 0.082
0.000 0.000 0.000 0.000

z x

(a) Compressive damage in (b) Tensile damage in (c) Compressive damage in (d) Tensile damage in
Zone 1−4 Zone 1−4 Zone 5−9 Zone 5−9

Figure 22. Damage state of web walls and coupling beams under rare earthquake

walls. The damage in the web walls is more severe than requirements of the predefined performance objectives.
that in the flange walls. The damage conditions under Furthermore, there is relatively large safety margin
the three levels of ground motions meet the under rare earthquakes.

Advances in Structural Engineering Vol. 17 No. 4 2014 525

 Performance-Based Seismic Design Principles and Structural Analysis of Shanghai Tower

S, Mises S, Mises S, Mises

Multiple section points Multiple section points Multiple section points
(Avg: 75%) (Avg: 75%) (Avg: 75%)
208.18E + 03 292.07E + 03 254.29E + 03
190.83E + 03 267.73E + 03 233.10E + 03
173.49E + 03 243.39E + 03 211.91E + 03
156.14E + 03 219.05E + 03 190.72E + 03
138.79E + 03 194.71E + 03 169.53E + 03
121.44E + 03 170.38E + 03 148.34E + 03
104.10E + 03 146.04E + 03 127.15E + 03
86.75E + 03 121.70E + 03 105.96E + 03
69.40E + 03 97.36E + 03 84.77E + 03
52.05E + 03 73.02E + 03 63.58E + 03
34.70E + 03 48.68E + 03 42.38E + 03
17.36E + 03 24.34E + 03 21.19E + 03
7.89E + 00 3.27E + 00 2.73E + 00 Max: 254.29E + 03
Max: 208.18E + 03 Max: 292.07E + 03 Max: 254.29E + 03
Elem: SHZX−1.24676 Elem: SHZX−1.26318 Elem: SHZX−1.13598
Node: 15444 Node: 6241 Node: 1408

Max: 208.18E + 03

z y z y z y

x x Max: 292.07E + 03 x

(a) Steel plates in mega-columns (b) Outrigger trusses (c) Belt trusses

S, Mises S, Mises
Multiple section points Multiple section points
(Avg: 75%) (Avg: 75%)
292.31E + 03 199.74E + 03
267.99E + 03 183.31E + 03
243.67E + 03 166.89E + 03
219.34E + 03 150.46E + 03
195.02E + 03 134.04E + 03
170.70E + 03 Max: 292.31E + 03 117.61E + 03
146.38E + 03 101.18E + 03
122.05E + 03 84.76E + 03
97.73E + 03 68.33E + 03
73.41E + 03 51.91E + 03
49.09E + 03 35.48E + 03
24.76E + 03 19.05E + 03
440.85E + 00 2.63E + 03
Max: 292.31E + 03 Max: 199.74E + 03
Elem: SHZX−1.50954 Elem: SHZX−1.52142
Node: 14705 Node: 2291

z y
z y

x Max: 199.74E + 03
x

(d) Steel plates in core walls (e) Steel plates at bottom of core walls

Figure 23. Mises stress nephogram of steel members in main components (kN/m2)

6. CONCLUSIONS which is a typical code-exceeding tall building are

PBSD approach has been highly recommended and
become necessary to demonstrate the seismic
performance of code-exceeding tall buildings at least
equivalent to code intent of safety. In this paper, the
general methodologies for PBSD of code-exceeding
tall buildings in Mainland China are summarized
briefly. Accordingly, the performance objectives and
corresponding design criteria of Shanghai Tower
introduced. To verify the prospective performance
objectives of Shanghai Tower, nonlinear time-history
analysis is carried out to evaluate the seismic
performance of the structure. Based on the numerical
analysis, the following conclusions can be drawn:
(1) The maximum inter-story drift ratios of the
structure under frequent earthquakes, basic
earthquakes and rare earthquakes are less than

526 Advances in Structural Engineering Vol. 17 No. 4 2014

H.J. Jiang, X.L. Lu, X.J. Liu and L.S. He
the allowable limits. The belt trusses and
outrigger trusses are very effective to reduce the
lateral displacement of the structure.
(2) The mega-frame carries about 70% of the total
overturning moment and 40% of the total base
shear force of the structure. A favourable dual
structural system with additional seismic
defence is realized in this structure, which is
very helpful to ensure the safety of the structure
under rare earthquakes.
(3) The damage conditions of the structure under
frequent earthquakes, basic earthquakes and rare
earthquakes meet the requirements of the
predefined performance objectives.
(4) In general, the designed structure system is an
efficient solution to resist earthquakes and the
pre-selected seismic performance objectives, i.e.,
fully operational under frequent earthquakes,
operational under basic earthquakes, and life
safety under rare earthquakes, could be
accomplished with considerable safety margin.
ACKNOWLEDGMENTS
The authors are grateful for the support from the
National Key Technology R&D Program under Grant
No. 2012BAJ13B02, the owner of Shanghai Tower,
Architectural Design and Research Institute of Tongji
University (Group) Co., Ltd., Gensler, and the
structural design consultant Thornton Tomasetti, Inc.
REFERENCES
Applied Technology Council (1996). Seismic Evaluation and
Retrofit of Concrete Buildings, ATC-40, Report No. SSC 96-01,
Redwood City, USA.
Berahman, F. (2013). “Performance-based seismic evaluation of
the Icon Hotel in Dubai, United Arab Emirates”, The
Advances in Structural Engineering Vol. 17 No. 4 2014
Structural Design of Tall and Special Buildings, Vol. 22, No.
3, pp. 300–326.
Chandler, A.M. and Lam, N.T.K. (2001). “Performance-based
design in earthquake engineering: a multi-disciplinary review”,
Engineering Structures, Vol. 23, No. 12, pp. 1525–1543.
FEMA (2000). Prestandard and Commentary for the Seismic
Rehabilitation of Buildings, FEMA-356, Federal Emergency
Management Agency, Washington, DC, USA.
Hibbitt, D., Karlsson, B. and Sorensen, P. (2004). ABAQUS Analysis
User’s Manual, Pawtucket, USA.
Lee, J. and Fenves, G.L. (1998). “Plastic-damage model for cyclic
loading of concrete structure”, Journal of Engineering
Mechanics, ASCE, Vol. 124, No. 8, pp. 892–900.
Los Angeles Tall Building Structural Design Council (2008). An
Alternative Procedure for Seismic Analysis and Design of Tall
Buildings Located in the Los Angeles Region, Los Angeles, USA.
GB50011-2010 (2010a). Code for Seismic Design of Buildings,
China Architecture and Building Press, Beijing, China.
JGJ3-2010 (2010b). Technical Specification for Concrete Structures
of Tall Building, China Architecture and Building Press, Beijing,
China.
Structural Engineers Association of California (1995). Performance
Based Seismic Engineering of Buildings, Vision 2000,
Sacramento, USA.
Structural Engineers Association of Northern California (2007).
Recommended Administrative Bulletin for San Francisco, San
Francisco, USA.
TBI Guidelines Working Group (2010). Guidelines for
Performance-Based Seismic Design of Tall Buildings, Pacific
Earthquake Engineering Research Center, Berkeley, USA.
Thornton Tomasetti Inc. (2009). Report of Shanghai Tower Submitted to
Code-exceeding Tall Building Peer Review Panel, Shanghai, China.
Zou, X.K., Chan, C.M., Li, G. and Wang, Q. (2008). “Multiobjective
optimization for performance-based design of reinforced concrete
frames”, Journal of Structural Engineering, ASCE, Vol. 134, No.
9, pp. 1569–1581.
527
View publication stats