3rd Asia Conference on Earthquake Engineering (ACEE)

Bangkok Thailand | December 2010

Case Study for Performance Based Design of 50-Story Building with Ductile
Core Wall System

Dr. Naveed Anwar Thaung Htut Aung Deepak Rayamajhi

CEO Projects Coordinator External Consultant
AIT Consulting AIT Consulting AIT Consulting
Asian Institute of Asian Institute of Asian Institute of
Technology Technology Technology
Bangkok, Thailand Bangkok, Thailand Bangkok, Thailand
[email protected] [email protected] [email protected]

Summary
The case study building is 50-story tower (about 166.8 meters above the ground level) and 3½-story
of below grade parking (extending approximately 13 m below the grade). The tower consists mainly
of residential units, and a terrace and amenity deck. The ground level contains retail and back of the
house space. It is a reinforced concrete building, which is laterally braced by ductile core wall
system together with buckling restrained braces.
The objective of this study is to present the advantages and application of performance based design
approach on tall reinforced concrete core wall buildings. The design procedure follows the
LATBSDC 2008 alternative guidelines in which the design adequacy is checked against the two
distinct levels of earthquake ground motions; serviceable behaviour for frequent earthquakes and
very low probability of collapse under extremely rare earthquakes. Firstly, the preliminary design is
performed against Design Basis Earthquake (DBE) (as defined by ASCE 7-05) in accordance with
the code based design procedures, exceptions that the LATBSDC 2008 allow. Then, building is
designed by performance based approach against the seismic hazard which is likely to happen, so
that predictable and safe performance is achieved. Two levels of performance are checked;
Serviceable/Operational Level performance under 43-year return period earthquake (50%
probability of exceedance in 30 years) and Collapse Prevention Level performance under Maximum
Considered Earthquake (MCE) with 2475-year return period (2% of probability of exceedance in 50
years).
Response spectrum analysis is conducted for Design Basis Earthquake Level and Service Level
response spectra. For the evaluation against collapse during extremely rare events, nonlinear time
history analysis is performed for the site specific ground motion records. Average of seven ground
motion approach is used to check the demands in the primary structural members of the lateral force
resisting system. The case study results show that the performance based design approach ensure
the building to be operational under frequent earthquakes and avoid the total or partial collapse of
the building under extreme earthquakes which are very rare in the lifetime of the building.
Keywords: Maximum Considered Earthquake, Design Basis Earthquake, Service/Frequent
earthquake, Performance Based Design, Ductile behaviour, Brittle behaviour

1. Introduction
The traditional codes are developed for low and medium rise buildings whose responses are
typically dominated by first translational mode. These codes use the global force reduction factors
which cannot predict accurately for the structures with significant inelastic response, especially in
tall buildings. Furthermore, the traditional codes attempt to satisfy implicitly the performance
objectives whereas the performance based design explicitly mentions in such a way that the design
satisfies the performance objectives for different levels of earthquakes [3]. The case study tower is
a residential tower located in Makati City, Philippines. It is built of reinforced concrete and 166.8
meters (50-story) tall above the ground level with 34.5 x 26 meters plan area. Reinforced concrete
bearing walls, gravity columns and post-tensioned flat slabs are utilized for gravity load resisting
 3rd Asia Conference on Earthquake Engineering (ACEE)
Bangkok Thailand | December 2010

system. The lateral load resisting system consists of reinforced concrete bearing wall coupled with
outrigger columns, connected by the buckling restrained braces. The building has 3½-story of
below grade parking, resting on the mat foundation.

Principal
minor dir.

Principal
major dir.

Fig. 1: Building Plan

2. Seismic Performance Objectives

The specific performance objectives for the design of the building for three levels of earthquake
hazards are shown in the following table.

Table 1: Performance Objectives

Level of Earthquake Seismic Performance Objective
Frequent/Service: 50% probability of Serviceability: Structure to remain
exceedance in 30 years (43-year return essentially elastic with minor damage
period), 2.5% damping to structural and non-structural
elements
Design Basis Earthquake (DBE): As Code Level: Moderate structural
defined by ASCE 7, Section 11.4, 5% damage; extensive repairs may be
damping required
Maximum Considered Earthquake Collapse Prevention: Extensive
(MCE): 2% probability of exceedance in structural damage; repairs are required
50 years (2475-year return period), 2 to and may not be economically feasible
3% damping

3. Design Approach
To demonstrate that the design is capable of providing code equivalent seismic performance, a
three- step analysis and design procedure is performed.
Step 1 – Preliminary design phase
Step 2 – Serviceability check
Step 3 – Collapse prevention check at MCE Level

3.1 Preliminary Design Phase

In this phase, elastic response spectrum analysis and design are performed in accordance with the
code based design approach by using appropriate load factors and strength reduction factors against
 3rd Asia Conference on Earthquake Engineering (ACEE)
Bangkok Thailand | December 2010

the gravity loads, wind load and seismic load [7]. Site specific response spectrum for DBE level is
used for the preliminary design phase. Structural components to be remained elastic are designed by
applying the appropriate amplification factors [2].

3.2 Serviceability Check

Primary response characteristics such as story drift, coupling beam and shear wall capacity ratios
are checked against the demands resulting from the response spectrum analysis using service level
response spectrum with 43-year return period (50% of probability of exceedance in 30 years) [1].
3.3 Collapse Prevention Check at MCE Level
Design verification is performed by non-linear response history analysis (NLRHA) against the
MCE level earthquakes with 2475-year return period (2% of probability of exceedance in 50 years)
[1]. The initial design is modified as required in order to meet the acceptance criteria.
Core wall coupling beams, core wall flexural response, slab outrigger beam are checked in
anticipation of non-linear response while core wall shear, diaphragms, basement walls, foundations
and columns are checked to remain essentially elastic during the non-linear response history
analysis.

4. Loading Criteria
4.1 Gravity Load
The minimum loading requirements have been taken from Table 4-1 of ASCE 7-05. Live loads are
reduced where permitted in accordance with Section 4.8 of ASCE 7-05. In addition to the uniform
slab loads, a superimposed dead load is applied along the perimeter of plan to account for the
weight of the cladding system.
4.1 Wind Load
Wind load is determined in accordance with ASCE 7-05. The design wind speed for the case study
building is 200 kph and the exposure type is B.
4.2 Seismic Load

4.2.1 Frequent/Service Level Earthquake

For the performance evaluation at Service Level seismic hazard, the following service level site
specific response spectrum is used.

Fig. 2: Response Spectrum at Service Level Earthquake (2.5% damping)

 3rd Asia Conference on Earthquake Engineering (ACEE)
Bangkok Thailand | December 2010

4.2.2 Maximum Considered Earthquake Level

Seven pairs of site specific ground motions are used to conduct the nonlinear response history
analysis. Average of demands from seven ground motions approach is used for design evaluation at
MCE level. The response spectra of seven ground motions are shown in figures below.

Fig. 3: Response Spectrum for Fault Normal and Fault Parallel Earthquakes at MCE Level

5. Modelling and Analysis Tools

A complete full three-dimensional finite element model is created which includes the tower and the
whole podium.
The modelling and analysis of building for evaluation and design at Service Level earthquake and
DBE level are carried out in ETABS 9.5 computational platform. An elastic model is created with
the specified material properties and appropriate stiffness modifiers for the structural components.
For the MCE level performance evaluation, nonlinear three-dimensional model is created in
PERFORM-3D (Version 4.0.4) computational platform.
5.1 Material Models

5.1.1 Concrete
In nonlinear model, effect of confinements is taken into account for the compressive strength and
ductility of concrete. Mander’s (1994) [6] confinement model is used to determine the confinement
effect. In PERFORM 3D, concrete material is modelled with tri-linear backbone curve. Tensile
strength of concrete is neglected. Material cyclic hysteretic degradation is not considered in the
model.

5.1.2 Reinforcing Steel

In nonlinear model, reinforcing steel material is modelled with tri-linear backbone curve. Yield
strength is taken as 1.15 times nominal strength and the ultimate strength is estimated as 1.5 times
expected strength with approximately 1% of strain hardening.
5.2 Shear Wall
Fibre modelling technique is used to model the flexural behaviour of the core wall. PERFORM-3D
shear wall element is used to model the nonlinear behaviour of shear wall [5].
Basically, two parallel fibre sections are used to model the shear wall. The first fibre section
consists of only uniformly distributed steel (Steel only) and the second fibre section consists of both
concrete and boundary zone steel reinforcement. For the uniformly distributed steel, auto-size fiber
elements are used whereas for latter one, fixed size fibre elements are used. Shear behaviour in the
wall is modelled with elastic material properties.
 3rd Asia Conference on Earthquake Engineering (ACEE)
Bangkok Thailand | December 2010

Fig. 4: Fibre Modelling of Reinforced Concrete Shear Wall

5.3 Coupling Beam

In this building, two types of coupling beams are present. First one is deep beam having span to
depth ratio of 1.9 (span/depth < 4), and second one is slender beam having span to depth ratio of 4.3
span/depth > 4). Since deep beams are dominated by shear behaviour, they are modelled for shear
deformation controlled while the slender beams are modelled for flexural deformation controlled.
The deep coupling beam is modelled with elastic frame section with a nonlinear shear hinge located
at mid span of the element. The capacity of the shear hinge is calculated based on the diagonal
reinforcements. The elastic stiffness of the deep beams is reduced to 0.16EIg. The shear capacity of
diagonal reinforcement is calculated based on formula provided in ACI 318-08. The ultimate point
is taken as the 1.33 times of the yielding capacity.
The slender coupling beam is modelled with two moment hinges placed at the ends of the beam.
The capacity of the moment-curvature hinges are calculated based on the longitudinal
reinforcements provided in the beams. The deformations capacities are taken from ASCE 41-06 for
the flexural coupling beams. The elastic stiffness of the slender beams is reduced to 0.5EIg.
5.4 Columns and Girders
The columns and girders are modelled as elastic frame member. The capacities of these members
are checked against the forces extracted from MCE analysis. The elastic stiffness of the columns
and girders are reduced to 0.7 EIg and 0.5Ig respectively [4].

5.5 Floor Slab

In the tower portion, the floor is modelled as rigid floor diaphragm. The slab is not modelled for the
tower portion. However, equivalent “slab outrigger beams” are modelled in order to study impact of
slab to core and column only. Slab outrigger beams are modelled with nonlinear hinges at both the
ends of the beam. Moment-curvature type of hinge is used to model nonlinearity in the slab-beam.
The moment capacity of the slab beam is calculated based on the reinforcement in the slab.
However, the performance of the moment hinges is not specifically reviewed.
At the podium and basements level, the slabs are modelled without rigid floor diaphragm. Slabs in
the podium and basement are modelled using shell element. The elastic flexural stiffness of the
slabs and equivalent slab-beams are reduced to 0.5EIg [4].
5.6 Support/ Foundation
The base of the reinforced concrete shear wall is modelled as pinned at the location of mat whereas
the columns and basement walls are modelled as fixed support. Furthermore, for this analysis, the
basement walls are also restrained by lateral springs in the lateral direction to take into account the
restraining effect of lateral soil. In order to consider the flexibility of the diaphragm, the stiffness of
the ground floor and below–grade diaphragms are reduced to 0.1 Ag.
 3rd Asia Conference on Earthquake Engineering (ACEE)
Bangkok Thailand | December 2010

5.7 Buckling Restrained Braces

BRBs are used in the main lateral force resisting system in order to enhance the performance of the
building. The intended benefits of using BRB in this building are to reduce the story drifts and
lateral displacement as well as to participate in the outriggering effect on the overturning moment in
the tower. Moreover, BRBs can reduce the base shear in the building by dissipating the energy.
Two different levels of designed forces are used for the BRBs in this building. One is from level
19th - 23rd floor and another from 43 rd - 47th floor. 16 BRBs are used in the building. Eight BRB’s
are located at 19th - 23rd floor and rests are at 43rd - 47th floor, principal minor direction.

6. Analysis Results
6.1 Modal Analysis
The natural periods of the building are 5.75 s and 4.86 s in principal directions with 0.40 and 0.42
modal participating mass ratios.

Mode 1 Mode 2 Mode 3

Fig. 5: Mode Shapes
6.2 Base Shear
The base shear is compared between DBE level response spectrum analysis and average of MCE
level nonlinear response history analysis in the following table. The base shear is calculated above
the podium level and considered the tower portion only. The seismic weight of the tower above the
podium level is 616,900 kN.
Table 2: Base Shear Comparison
Load Cases Base Shear % of Seismic
Weight
(KN)
DBE level (Along principal major dir.) 21,012 3.56
DBE level (Along principal minor dir.) 22,691 3.84
MCE level (Along principal major dir.) 47,892 7.76
MCE level (Along principal minor dir.) 46,462 7.53

6.3 Story Shear and Story Moment

Story shears and story moment distributions are nearly triangular shape, showing the dominance of
first modes in each principal direction. Furthermore, the story shear at the basement level is
generally decreased in most of the time history except some time histories where the story shear has
increased. This may happened due to the irregular distributions of basement walls and supports. The
application of BRBs reduces the moment and shear demands in the core wall, especially in the
principal minor direction.
 3rd Asia Conference on Earthquake Engineering (ACEE)
Bangkok Thailand | December 2010

Fig. 6: Story Shear

Fig. 7: Story Moment

6.4 Story Drifts
In the preliminary investigation, the story drifts are checked at MCE level without using BRBs.
Then, the BRBs are applied in the model and the story drifts are rechecked. One advantage is that
BRBs reduce the story drift of the building in the principal minor direction. The maximum story
drifts envelopes for both principal directions are less than 3% which is acceptable limit against
MCE level earthquakes.

Fig. 8: Story Drift

6.5 Axial Strain Shear Wall
The flexural capacity of shear wall is evaluated in terms of the yielding of steel and crushing of
concrete materials. The strain in steel fibre and concrete fibre are checked against the acceptable
strain limits. The compression strain of MCE analysis is increased by 2 times and compared with
the limit set in the performance criteria.
 3rd Asia Conference on Earthquake Engineering (ACEE)
Bangkok Thailand | December 2010

7. Findings
7.1 Service Level Performance
At service level earthquake, all storey drifts are less than 0.5%. The response of the columns and
coupling beams in shear and moment, shear walls in flexure and shear and buckling restrained
braces in axial direction are within the elastic limit. The capacity of each element at service level is
higher than the corresponding demand in the element. There are a few elements in which the
demand exceeds the capacity such as deep coupling beams shear, however, which is permissible.
7.2 MCE Level Performance
The design base shear (shear calculated above the podium) is approximately 3.5% and 3.8% in each
principal direction, which is higher than the minimum limit of 3%, set by the LATBSDC-2008
guidelines. Furthermore, the dynamic base shear calculated from the average of seven time histories
is approximately two times higher than the design base shear, which is typical in high rise buildings.
From the storey shear and storey moment plots of seven time histories, in average the results
demonstrate that the building is mainly dominated by first fundamental modes in both X and y
direction. The shear and moment demands are reduced in the principal minor direction, especially
in the core wall when the BRBs are applied along the principal minor direction of the building.
Moreover, story drift in the principal minor direction of the building is improved by the utilization
of BRBs.
Flexural deformation capacity of shear wall, evaluated by the axial strain of the fibres, is within the
acceptable limit. All the columns including the outrigger columns remain essentially elastic under
MCE level earthquake. On average, all the BRB’s have ductility demand less than 9, the limit set by
ASCE41-06, and indicates that all the BRBs satisfy the performance criteria.

8. Conclusions
In conclusion, overall response of the building remains elastic, i.e. the building will remain
operational during the service level/frequent earthquakes. Furthermore, the overall response of the
building is within the collapse prevention limit at the MCE level, extreme earthquakes, in which the
demand base shear is two times larger than DBE level base shear. On the other hand, it is uncertain
that the building with traditional code based design can resist the MCE level earthquake without
partial or total collapse since the performance is not checked explicitly.

9. References
[1] Los Angeles Tall Buildings Structural Design Council. 2008, An Alternative Procedure for
Seismic Analysis and Design of Tall Buildings Located Los Angeles Region, Los Angeles Tall
Buildings Structural Design Council
[2] ASCE. 2005, Minimum Design Loads for Building and Other Structures (ASCE 7-05),
American Society of Civil Engineers
[3] Council on Tall Buildings and Urban Habitat. 2008, Recommendations for Seismic Design of
High-rise Buildings, Council on Tall Buildings and Urban Habitat
[4] ASCE. 2006, Seismic Rehabilitation of Existing Buildings (ASCE 41-06), American Society
of Civil Engineers
[5] Computers and Structures, Inc. 2006, Perform 3D, Nonlinear Analysis and Performance
Assessment for 3D Structures User Guide, Version 4, Computers and Structures, Inc.,
Berkeley, CA.
[6] Mander J.B., Priestley M.J.N., Park R.,“Theoretical Stress-Strain Model for Confined
Concrete”, ASCE Journal of Structural Engineering, Vol. 144, No. 8, August 1988
[7] ACI. 2008, Building Code Requirements for Structural Concrete (ACI 318-08) and
Commentary, American Concrete Institute

View publication stats