CTBUH Technical Guides

Performance-Based Seismic
Design for Tall Buildings 2nd Edition
An output of the CTBUH Performance-Based Seismic Design Working Group

Ramin Golesorkhi, Leonard Joseph, Ron Klemencic, David Shook & John Viise
Bibliographic Reference:
Golesorkhi, R., Joseph, L., Klemencic, R., Shook, D. & Viise, J. (2019). Performance-Based Seismic Design for Tall Buildings: An Output of the CTBUH
Performance-Based Seismic Design Working Group. Second Edition. Chicago: Council on Tall Buildings and Urban Habitat.

Principal Authors: Ramin Golesorkhi, Leonard Joseph, Ron Klemencic, David Shook & John Viise
Coordinating Editors: Jason Gabel & Daniel Safarik
Layout: Tansri Muliani & Annan Shehadi

© 2019 Council on Tall Buildings and Urban Habitat

Printed in the USA

The right of the Council on Tall Buildings and Urban Habitat to be identified as author of this work has been asserted by them in accordance with
sections 77 and 78 of the Copyright, Designs and Patents Act 1988.

All rights reserved. No part of this book may be reprinted or reproduced or utilized in any form or by any electronic, mechanical, or other means, now
known or hereafter invented, including photocopying and recording, or in any information storage or retrieval system, without permission in writing
from the publishers.

Trademark notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation
without intent to infringe.

Library of Congress Cataloging-in-Publication Data

A catalog record has been requested for this book

ISBN 978-0-939493-72-2

CTBUH Headquarters
The Monroe Building
104 South Michigan Avenue, Suite 620
Chicago, IL 60603, USA
Phone: +1 312 283 5599
Email: info@ctbuh.org
www.ctbuh.org
www.skyscrapercenter.com

CTBUH Asia Headquarters

Wenyuan Building
College of Architecture and Urban Planning (CAUP)
Tongji University
1239 Si Ping Road, Yangpu District, Shanghai
China 200092
Phone: +86 21 6598 2972
Email: china@ctbuh.org

CTBUH Research Office

Iuav University of Venice
Dorsoduro 2006
30123 Venice, Italy
Phone: +39 41 257 1276
Email: research@ctbuh.org

CTBUH Academic Office

S. R. Crown Hall
Illinois Institute of Technology
3360 South State Street
Chicago, IL 60616
Phone: +1 312 567 3487
Email: academic@ctbuh.org

The information contained in this guide is for educational purposes and obtained by CTBUH from sources believed to be reliable. However, neither
CTBUH or its authors guarantee the accuracy or completeness of any information published herein, and neither CTBUH or its authors shall be
responsible for any errors, omissions, or damages arising out of the use of this information. This work is published with the understanding that
CTBUH and its authors are supplying information but are not attempting to render engineering or other professional services. The recommendations
should not be used to circumvent building codes or other municipal or governmental building requirements. The recommendations are general in
nature and may or may not be applicable to any particular building or any specific circumstances.

Front Cover Image: Wilshire Grand, Los Angeles, under construction in 2015. © Gary Leonard/AC Martin
Opening Chapter Image: 350 Mission St., San Francisco, designed using PBSD principles. © Cesar Rubio
 Principal Authors
Ramin Golesorkhi, Langan Engineering, San Francisco
Leonard Joseph, Thornton Tomasetti, New York City
Ron Klemencic, Magnusson Klemencic Associates, Seattle
David Shook, Skidmore, Owings & Merrill LLP, San Francisco
John Viise, Thornton Tomasetti, Chicago

Contributors
Jeff Dragovich, Engineering Consultant Seattle, Seattle
Neville John Mathias, Skidmore, Owings & Merrill LLP, San Francisco
Ian McFarlane, Magnusson Klemencic Associates, Seattle
Jerome Tobolski, Thornton Tomasetti, Chicago
Kevin Aswegan, Magnusson Klemencic Associates, Seattle

Peer Review Panel

CTBUH Japan Structures Committee
Ian Aiken, SIE Inc., San Francisco
Baiping Dong, LERA Consulting Structural Engineers, New York City
Xiaonian Duan, Foster + Partners, London
John Hooper, Magnusson Klemencic Associates, Seattle
Andre Ly, Bouygues-Construction, Paris
Elena Mele, University of Naples, Naples
Elena Mola, ECSD, Milan
Roman Przepiorka, Arup, Washington D. C.
Rob Smith, Arup, San Francisco
John Tessem, DCI Engineers, Seattle
Shah Vahdani, Applied Geodynamics, Inc., El Cerrito
Tom Xia, DCI Engineers, Seattle
Reid Zimmerman, KPFF Consulting Engineers, Portland

See more on CTBUH committees at ctbuh.org/get-involved.

 Contents

Preface 7
Glossary and Abbreviations 8

1.0 Introduction 10

1.1 Overview of Performance-Based Seismic Design 13

1.2 Goals of PBSD 13
1.3 Historical Development of PBSD Provisions 13
1.4 Seismicity and Urban Growth 14
1.5 PBSD in Contrast with Traditional Methods 14
1.6 Performance Objectives and Seismic Demand 14
1.7 Seismic Performance Rating Systems 16
1.8 Procedure of PBSD 16
1.9 Peer Review 17
1.10 Appropriate Usage of PBSD 19

2.0 Site-Specific Seismic Hazard Assessment 20

2.1 Introduction 22
2.2 Developing Site-Specific Target Response Spectra 22
2.3 Range of Structural Periods For Consideration 23
2.4 Near-Fault Ground Motions 24
2.5 MCER- and Service-Level Earthquake (SLE) 24
2.6 Information Required from Structural Engineer 24
2.7 Time Series Record Selection and Development of 24
Site-Specific Time Series
2.8 Vertical Ground Motions 29

3.0 Design Using Linear Analysis 30

3.1 Modeling and Analysis 32

3.2 Acceptance Criteria 35

4
4.0 Verification of Response Under MCER Using NRHA 36

4.1 Modeling and Analysis 38

4.2 Acceptance Criteria 43

5.0 Basis of Design Example 46

5.1 General 48
5.2 Superstructure 48
5.3 Substructure 48
5.4 Foundation System 48
5.5 Code Analysis and Design Criteria 48
5.6 Structural Analysis and Design 49
5.7 Appendices 53

6.0 Performance-Based Seismic Design Examples 54

Design Example 1: Tall High-Occupancy Office Tower 56

Design Example 2: Supertall Mixed-Use Tower 70
Design Example 3: Tall Flat-Plate Office Tower 84
Design Example 4: Tall Residential Tower with Podium 96

7.0 Conclusion 108

References 112
About the CTBUH 114
About the Authors 115
CTBUH Organizational Members 116

5
Preface
In 2008, the Council on Tall Buildings and Urban Habitat (CTBUH) Seismic
Working Group authored the publication “Recommendations for the Seismic
Design of High-Rise Buildings.” This document and subsequent working group
meetings established that a consensus of practitioners believe the process
of Performance-Based Seismic Design (PBSD) is often more appropriate than
prescriptive code-based approaches for the design of tall buildings in regions
of high seismicity. Given that 75 percent of the tallest buildings completed in
2016 were constructed in seismic regions of the world where some form of
non-prescriptive design and analysis was necessary for building approval (The
Skyscraper Center 2016), it was apparent that publishing the design principles
inherent to the PBSD design process would be useful to an international
audience. As a result, the CTBUH Performance-Based Seismic Design (PBSD)
Working Group was formed with the goal of producing a publication to
introduce PBSD principles to an international audience and provide examples
of its application.

Although the practice and protocol for non-prescriptive design is quite mature
in certain countries (e.g., China and Japan), the methods used for PBSD as
practiced in western regions of the United States are of high interest to other
countries. PBSD guidelines have been reconsidered based on local practice
and implemented into the design of tall buildings in areas including Turkey, the
Philippines, and Russia. The methods used for PBSD have produced innovative
and cost-effective buildings in these regions. As a result, this publication, now
in its second edition, may be an especially helpful reference for practitioners
working internationally, and for jurisdictions looking to develop their own PBSD
guidelines and protocols for a design approval process.

In adopting this methodology for use outside the United States, it is recognized
that local design practices will vary internationally. Technical areas where
significant differences with US practice may occur would include performance
criteria, approvals processes, materials design standards, and definitions of
seismic hazard. Local structural engineers should closely review and resolve
these issues with local municipalities and appropriate approval authorities.

It is the intent of CTBUH to disseminate these methods to an international

audience for the advancement and expansion of PBSD principles. Additionally,
the presentation of several case study examples demonstrate the issues
commonly encountered when using PBSD for the design of tall buildings and
how practicing engineers have resolved them.

Preface | 7
Glossary and Abbreviations
Action: A force, moment, strain, displacement, or other
deformation resulting from the application of design load
combinations.
Arias Intensity: A measure of the strength of ground motion.
Buckling Restrained Brace (BRB): A structural member
designed to exhibit well-controlled, predictable yielding
behavior in tensile, compressive, and cyclic loading. It
typically consists of a steel core plate encased in a mortar-
filled outer steel shell that restrains compression buckling.
Capacity Design: A design approach that configures the
structure to concentrate yielding and inelastic behavior in
specific locations where elements are detailed to reliably
develop such behavior. This ductile behavior allows seismic
demands on other portions of the structure to be reduced to
remain essentially elastic during earthquake response.
Capping Strength: The peak strength attainable by a
structural component under monotonic loading.
CBC: California Building Code.
Conditional Mean Spectrum (CMS): A site-specific mean
acceleration response spectrum conditioned on the
occurrence of a target spectral acceleration at the period
of interest.
Critical Action: Force-controlled actions in which the failure
mode poses severe consequences to structural stability
under gravity and/or lateral loads.
Deformation-Controlled Action1: An action for which reliable
inelastic deformation capacity is achievable without critical
strength decay.
Design Earthquake (DE): Ground shaking defined by 2/3 of
the MCER ground shaking.
Demand-to-Capacity Ratio (DCR): Demand (value of force
or deformation predicted to occur in the specified seismic
event) divided by capacity (value of anticipated upper limit of
member strength or ability to accept deformation).
Deterministic Seismic Hazard Analysis (DSHA): Considers
the occurrence of a particular magnitude earthquake on a
particular fault at a distance from a site.
1
Definition from PEER/TBI.
8 | Glossary and Abbreviations
Equivalent Lateral Force (ELF): A defined pattern of
horizontal static forces applied to all floor levels to
approximate the overall building shear and overturning
moment forces generated by seismic events. Commonly used
in simplified approaches to seismic design.
Expected Strength1: The probable peak strength of a
structural element considering inherent variability in material
strength and strain hardening.
Force-Controlled Action1: An action for which inelastic
deformation capacity is not assured.
Ground Response Analysis: A computational technique
based on wave propagation theory for estimating ground
shaking at a site.
Hazard Level: A probability of exceedance within a defined
time period (or return period) at which ground shaking
intensity is quantified.
Interstory Drift Ratio (IDR): The maximum difference in
lateral displacements at two adjacent floor levels predicted
to occur during a seismic event, divided by the story height
between those levels.
IBC: International Building Code.
LATBSDC: Los Angeles Tall Buildings Structural Design
Council (http://www.tallbuildings.org/). LATBSDC was
formed in 1988 and produces guidelines for the design of tall
buildings in Los Angeles with lateral force resisting systems
that are not recognized by the governing code. LATBSDC is
also a liaison organization to CTBUH.
Lower-bound Strength: The probable minimum strength
that a structural element might develop considering
potential variability in material strength and workmanship.
Maximum Considered Earthquake, Risk Targeted (MCER):
The level of shaking specified by the ASCE 7 standard as a
basis for derivation of design ground motions.
Monotonic Loading: Loading of a structural component in
which the displacement increases monotonically without
unloading or reloading.
NEHRP: National Earthquake Hazards Reduction Program.
Nonlinear Response History Analysis (NRHA): RHA where
geometric and material non-linearity are taken into account
NIST: National Institute of Standards and Technology.
Noncritical Actions: Force-controlled actions where
failure does not result in structural instability or potentially
life-threatening damage.
Peak Strength: The maximum resistance an element will
develop under a specific loading protocol.
Performance-Based Seismic Design (PBSD): A building
seismic design methodology that facilitates the exception
to, enhancement of, or performance verification of,
building code provisions through the explicit evaluation of
performance objectives.
Peak Ground Acceleration (PGA): For a time-series record,
the maximum amplitude of acceleration that is recorded on
the ordinate of the record.
PEER/TBI: Pacific Earthquake Engineering Research / Tall
Buildings Initiative (https://peer.berkeley.edu/research/
building-systems/tall-buildings-initiative). The Pacific
Earthquake Engineering Research Center (PEER) has
responded to the surge of high-rise construction using
new framing systems by leading an initiative to develop
design criteria that will ensure safe and usable tall buildings
following future earthquakes.
Probabilistic Seismic Hazard Analysis (PSHA): Determines
the frequency with which a seismic hazard will occur at
a site, where a ground motion amplitude is larger than a
specified value.
Return Period1: The average time span between shaking
intensity that is equal to or greater than a specified
value, also known as the recurrence interval; the annual
frequency of exceeding a given intensity is equal to the
reciprocal of the return period for that intensity.
Response History Analysis (RHA): Also referred to as Time
History Analysis (THA), this type of analysis can be linear or
nonlinear, and includes a time-step analysis that captures
a solution for all the elements of a structure at each time
step over a ground motion time history record.
Response Spectrum Analysis (RSA): A linear-dynamic
statistical analysis method that measures the contribution
from each natural mode of vibration to indicate the likely
maximum seismic response of an essentially elastic structure.
Risk Category: A categorization of buildings for
determination of earthquake loads based on risk associated
with unacceptable performance.
Service Level Earthquake (SLE): Ground shaking represented
by an elastic 2.5 percent damped acceleration response
spectrum that has a return period of 43 years, approximately
equivalent to a 50 percent exceedance probability in 30 years.
Section Property Modifier: A value, typically less than 1.0,
which is used in computer analyses to reduce a member’s
effective stiffness due to damage in an earthquake. They are
selected based on the intensity of shaking and experimental
testing, and in some instances may be prescribed by the
governing design standard.
Site-Response Analysis: Analysis of wave propagation
through a soil medium used to assess the effect on spectral
shape of local geology.
Specified Strength: The specified minimum design strength
of a structural element.
Soil-Structure Interaction (SSI): Process in which the
response of the soil influences the response of the structure.
Conventional design methods neglect this effect.
Structural Engineer of Record (SEOR): Engineer responsible
for the design and permitting of the structural system.
Uniform Hazard Spectrum: A site-specific acceleration
response spectrum constructed such that the ordinate at each
period has the same exceedance probability or return period.
Glossary and Abbreviations | 9
1.0 Introduction
1.0 Introduction

Performance-Based Seismic Design

(PBSD) is a structural design
methodology that has become more
common in urban centers of the
western United States, especially for
the design of high-rise buildings.
It is a design methodology that
allows for design flexibility and offers
design opportunities to enhance
building performance and encourage
innovation. The most common use
of PBSD in practice is to substantiate
exceptions to specific prescribed code
requirements, such as height limits on
select structural systems. A second use
of PBSD is the ability to demonstrate
higher performance levels for a
structure at various intensities of a
seismic event.

An integral component of PBSD is

Nonlinear Response History Analysis
(NRHA). This advanced method of
analysis has been incorporated into
the design process in regions with
high seismicity, such as in China,
Philippines, Indonesia, Turkey, Japan,
etc. The design considerations
required by the process of PBSD are
extensive and require substantial
knowledge of nonlinear structural
behavior, seismic design, building
performance, and analytical modeling.
These demands have not limited the
design of structures, but instead led
to a number of highly efficient tall 5Figure 1.1: One Rincon Hill, San Francisco, designed using PBSD principles. © Magnusson Klemencic Associates
building designs that would not be
possible following a traditional code-
prescriptive design approach. methodology in the United States will zones. This publication is not intended
lead to a more detailed understanding as a standard such as ASCE 7, or as a
PBSD is currently accepted in of building response in seismic events group of guidelines such as PEER/TBI
numerous urban centers of the and allow for further innovations in and LATBSDC. Instead, this is a bridging
United States such as Los Angeles (see seismic design. document to introduce PBSD methods
Figures 1.1 and 1.2), San Francisco, to an international audience. Structural
Seattle, San Diego, Oakland, and Salt This publication provides structural engineers should look to develop a
Lake City. The current version of the engineers, developers, and project-specific basis of design founded
American Society of Civil Engineers contractors—in the United States on the references provided and engage
loads standard (ASCE 7–16) includes a and internationally—a general their local jurisdictions for appropriate
detailed framework for PBSD, making understanding of the PBSD process steps needed for project approval. The
it possible to use PBSD methods and examples from leading structural PBSD process is regularly evolving and
in all US jurisdictions adopting this engineering firms with a history of the latest standards and guidelines
standard. The broad acceptance of this designing tall buildings in high seismic should be referenced.

12 | Introduction
1.1 Overview of Performance-Based
Seismic Design
Performance-based seismic design is a
highly developed design methodology
that provides greater design flexibility to
structural engineers than that afforded
by prescriptive code-based approaches.
However, the methodology also involves
significantly more effort in the analysis
and design stages, with verification
of building performance required at
multiple seismic hazard levels using
linear and advanced nonlinear analysis
techniques. PBSD uses first principles of
engineering to proportion and detail
structural systems and components to
meet specific performance objectives.
Using PBSD methodology, the focus of
the structural engineer changes from
a prescriptive “check list” approach
of code provisions to requiring the
designer to more fully understand
building performance and the code’s
intent. Developing structural designs
through a more detailed knowledge
of building behavior during a seismic
event often results in solutions that
satisfy the targeted performance levels
5Figure 1.2: Wilshire Grand Center, Los Angeles, designed
using PBSD principles. © Gary Leonard/AC Martin
more efficiently. Although PBSD requires
additional design effort, the benefits
can be significant: reduced construction
costs, improved lease spaces, and
enhanced seismic performance.
1.2 Goals of PBSD
Developers and structural engineers
utilize PBSD for a variety of reasons.
Common goals of PBSD include:
 the ability to make exceptions to
specific code requirements, such
as height limits for select seismic
force-resisting systems;
 the use of seismic force-resisting
systems and innovative designs not
prescribed by code;
 the use of high-strength materials
and mechanical devices not
prescribed by code; and
 the reduction of structural and
non-structural damage through
enhanced seismic performance
objectives at specified levels of
seismic intensity.
A common example of a seismic force-
resisting system not recognized by code
is a core-and-outrigger seismic force-
resisting system. In the United States, this
is not one of the seismic force-resisting
systems recognized in ASCE 7. The use
of PBSD methods facilitates a method
to evaluate and design such seismic
force-resisting systems.
1.3 Historical Development of
PBSD Provisions
Historically significant earthquake
events (e.g., 1971 San Fernando, 1989
Loma Prieta, and 1994 Northridge)
caused significant damage and
down-time to businesses, residences,
and infrastructure in California. As a
result of these seismic events, major
market sectors like the airline industry
in the Los Angeles area and the
computing industry in Silicon Valley
desired to enhance the performance
of their buildings to minimize the
risk of casualties, damage to facilities,
and down-time of their existing and
new facilities should a more frequent
event occur. This demand served as a
catalyst to the engineering community
in the United States to develop design
methods to assess performance of
existing structures and to develop
design methodologies to enhance
the performance of these systems, as
well as ways to quantify the impact
of these enhancements. Performance
of existing structures is quantified by
the development of performance
objectives that are defined for structural
systems and components of the system.
Principles central to PBSD were
developed to rationally and efficiently
guide the design of seismic retrofits to
enhance the performance of existing
structures. These provisions ultimately
resulted in ASCE 41. The current ASCE
41-13 (Seismic Evaluation and Retrofit
of Existing Buildings) outlines a series of
evaluation levels for existing buildings.
Some levels of these types of retrofits
are defined as Tier 1 and 2, which
involve more prescriptive procedures.
The Tier 3 methodology utilizes PBSD
principles and includes performance
objectives that are implied in the code
and illustrated in Figure 1.3.
Current PBSD documents such as
PEER/TBI (The Pacific Earthquake
Engineering Research Center/Tall
Buildings Initiative) and LATBSDC
(Los Angeles Tall Buildings Structural
Design Council) refer to ASCE 41-13 for
one source of acceptance criteria at
performance levels described in Section
1.6 of deformation-controlled elements
such as coupling beams, shear walls,
and moment frames. Although these
acceptance criteria are provided in a
Introduction | 13

Base Shear
Structural Displacement Δ (earthquake intensity)
5Figure 1.3: Structural Performance Objective Illustration. Source: CTBUH based on Civil Engineering Home diagram
seismic evaluation and retrofit building
standard, they are appropriate for use
in PBSD of new buildings. Thus, they are
also useful for setting enhanced seismic
performance goals.
1.4 Seismicity and Urban Growth
Urban centers are growing worldwide,
especially in regions of moderate
to high seismicity. The number of
supertall buildings (300-meter-plus)
completed annually has steadily risen
since 2010, with 10 completing in
2016 alone (CTBUH 2016). Many urban
areas in seismically active regions
have adopted reasonable seismic
building code requirements, but even
sensible prescriptive code designs
often increase building costs and
suppress innovation due to limitations
on the applications of specific seismic
force-resisting systems. PBSD design
methods offer these urban centers the
needed safety, potential for enhanced
performance, and cost-effective
innovations desirable for expanding
urban habitats.
14 | Introduction
1.5 PBSD in Contrast with Traditional
Methods
PBSD using Nonlinear Response History
Analysis (NRHA) stands in contrast
to common analysis and design
methodologies. Equivalent Lateral
Force (ELF) analysis and Response
Spectrum Analysis (RSA) are both
traditional linear-analysis methodologies
that use seismic reduction factors to
account for energy dissipation through
nonlinear behavior. These methods
require engineers to make assumptions
regarding the anticipated level of energy
dissipation. At the system scale, a seismic
reduction factor is applied based on
code-prescribed qualitative grading of
structural systems (“R” value in ASCE 7).
The seismic reduction factor implies
a reasonably well-distributed level of
energy dissipation. This has been shown
to be reasonable for buildings of modest
height and conventional geometry. At
the component scale, property modifiers
are applied to all elements to account
for effective member stiffness. As a
result, critical areas where significant
nonlinear response may develop can
be overlooked, or energy dissipation could
be underestimated. For example, some
practicing engineers have observed lower
levels of energy dissipation in NRHA than
implied by code R values for bearing shear
wall systems applied to tall buildings.
This has resulted in shear wall demands
(revealed by NRHA) above code-prescribed
levels in some cases. An important
advancement in the use of PBSD is the
more accurate consideration of energy
dissipation of individual components
through modeling of nonlinear behavior.
Recently, the FEMA P-695 methodology
allowed for the quantitative evaluation
of system R-factors. Several systems
were evaluated by ASCE 7 committees,
which found that the committee-
developed factor, and associated
design requirements, provide collapse
performance that is generally consistent
with the standard’s stated intent.
Code-prescriptive designs using RSA
have been shown to produce adequate
tall building designs in large-magnitude
earthquakes such as the Tōhoku
earthquake of 2011 and the Chilean
earthquake of 2010. Unfortunately,
damage levels and locations are not
always well predicted by RSA. Although
generally shown to be adequate for the
majority of buildings, many engineers find
that code-prescriptive requirements limit
structural innovation and creativity.
1.6 Performance Objectives and
Seismic Demand
Seismic performance objectives are
an important consideration when
applying PBSD methods, since enhanced
performance objectives can be considered
and code exceptions can be taken.
In both cases, demonstration of code
equivalent performance is often required.
Performance standard definitions
come from ASCE 7 and ASCE 41. ASCE
7 performance objectives are primarily
related to minimum life-safety in a
Design Earthquake (DE) level event
associated with a specified Risk
Category. ASCE 41 identifies a series of
performance objective targets that can
be related to specific levels of seismic
intensities. Descriptions of anticipated
levels of structural and non-structural
damage under each performance
objectives are described in ASCE 41-13
(see Table 1.1).
Three levels of seismic shaking are
considered and defined below:
Risk Targeted Maximum Considered
Earthquake (MCER )
The level of shaking specified by the
ASCE 7 standard as a basis for derivation
of design ground motions. ASCE
7–16 defines this as “The most severe
earthquake effects considered by this
standard determined for the orientation
that results in the largest maximum
response to horizontal ground motions
and with adjustments for targeted risk.”
Design Earthquake (DE)
Ground shaking represented by
an elastic five percent damped
acceleration response spectrum defined
as 2/3 of the MCER earthquake.
Service Level Earthquake (SLE)
Ground shaking represented by an
elastic 2.5 percent damped acceleration
response spectrum that has a return
period of 43 years, approximately
equivalent to a 50 percent exceedance
probability in 30 years.
Figure 1.4 describes relationships
between performance objectives,
earthquake intensities, and Risk
Category in ASCE 41. These can
be used to help developers and
structural engineers relate structural
performance and seismic intensities
when considering performance
objectives, especially enhanced
performance objectives.
These performance objectives have
been written with US practice in mind.
Other international codes (such as
Eurocode 8-3, which is analogous to
ASCE 41–13) may use different objectives
and levels of shaking, although the
general concept remains the same.
Although code-equivalent performance
is the minimum standard for PBSD,
many structural engineers and peer
reviewers require slightly higher than
code-minimum performance in a
tall building. For example, ASCE 7–16
allows the average maximum drift of all
response history evaluations under MCER
loading to be four percent of the story
height, whereas PEER/TBI and LATBSDC
require three percent for Risk Category
II structures. Furthermore, both PEER/
TBI and LATBSDC require verification
that residual drifts meet certain criteria,
but ASCE 7–16 has no verification of
residual drifts. ASCE 7–16 is a minimum
loading standard for all buildings, not
just tall buildings. PEER/TBI and LATBSDC

Damage Control and Building Performance Levels

Target Building Performance Levels

Overall Damage
Structural Components
Nonstructural Components
Comparison with Performance
Intended for Typical Buildings
Designed to Codes or
Standards for New Buildings,
for the Design Earthquake
Collapse Prevention
Level (5-D)
Severe
Little residual stiffness and
strength to resist lateral loads,
but gravity load-bearing columns
and walls function. Large
permanent drifts. Some exits
blocked. Building is near collapse
in aftershocks and should not
continue to be occupied.
Extensive damage. Infills and
unbraced parapets failed or at
incipient failure.
Significantly more damage and
greater life safety risk.
Life Safety
Level (3-C)
Moderate
Some residual strength and
stiffness left in all stories.
Gravity-load-bearing elements
function. No out-of-plane failure
of walls. Some permanent drift.
Damage to partitions. Continued
occupancy might not be likely
before repair. Building might not
be economical to repair.
Falling hazards, such as
parapets, mitigated, but many
architectural, mechanical, and
electrical systems are damaged.
Somewhat more damage and
slightly higher life safety risk.
Immediate Occupancy
Level (1-B)
Light
No permanent drift. Structure
substantially retains original
strength and stiffness.
Continued occupancy likely.
Equipment and contents are
generally secure but might not
operate due to mechanical failure
or lack of utilities. Some cracking
of façades, partitions, and ceilings
as well as structural elements.
Elevators can be restarted. Fire
protection operable.
Less damage and low life safety
risk.
Operational
Level (1-A)
Very Light
No permanent drift. Structure
substantially retains original
strength and stiffness. Minor
cracking of façades, partitions,
and ceilings as well as
structural elements. All systems
important to normal operation
are functional. Continued
occupancy and use highly likely.
Negligible damage occurs.
Power and other utilities
are available, possibly from
standby sources.
Much less damage and very low
life safety risk.

5Table 1.1: Excerpt from ASCE 41-13 for Building Performance Levels (Table C2-3).

Introduction | 15
 Performance
Level
Earthquake Immediate Collapse
Intensity Operational Life Safety
Occupancy Prevention

Service Level
Es St
Earthquake (SLE) se an
nt da
iala Hi rd
nd gh O
Ha Oc cc
cu upa
za pa nc
rd nc ies
Design Level o us y(
Fa Ca (C
Earthquake (DE) cil te ate
itie g or go
s( yI rie
Ca II) sI
te an
go dI
ry I)
IV
Maximum Considered )
Earthquake (MCE)

5Figure 1.4: Performance Levels of Code-Based Buildings at Various Risk Category Levels as described in ASCE 41 © Skidmore, Owings & Merrill LLP

apply only to tall buildings. Although
PEER/TBI and LATBSDC requirements
are more stringent than ASCE 7-16,
they are considered appropriate for tall
building design.
1.7 Seismic Performance Rating
Systems
Recently, guidelines have been produced
to help identify enhanced performance
objectives meeting project and
developer intent. This includes rating
systems published by the United States
Resiliency Council and the Resilience-
Based Earthquake Design Initiative.
The United States Resiliency Council
Rating System (USRC) building
rating system identifies expected
consequences of an earthquake or other
hazards affecting buildings. The rating
considers the performance of a building’s
structure, its mechanical, electrical and
plumbing systems, and architectural
components such as cladding, windows,
partitions, and ceilings. The performance
of these elements affects occupant
safety, the cost and time to carry out
necessary repairs, and when the building
can be re-occupied following an event.
USRC’s approach provides consistency,
usefulness and transparency to
increase free market demand for better
performing buildings. The attention to
building safety and business continuity
will over time improve the building stock
and make our cities and communities
more resilient. Currently, the USRC is
offering building ratings for earthquakes.
Ratings for other hazards are expected to
be developed soon.
The Resilience-Based Earthquake
Design Initiative (REDi™) Rating System,
developed by Arup, proposes a
framework for owners, architects, and
engineers to implement “resilience-
based earthquake design.” It describes
design and planning criteria to enable
owners to resume business operations
and provide livable conditions quickly
after an earthquake, according to their
desired resilience objectives. It also
presents a loss evaluation methodology
for assessing the success of the adopted
design and planning measures in
meeting the resilience objectives.
Recently, Los Angeles has adopted the
REDi™ design methodology for targeting
enhanced seismic performance levels.
1.8 Procedure of PBSD
For the structural design engineer, PBSD
design is typically a two-step process
where the structural engineer first
proportions the building using linear
analysis methods (typically RSA) and
then verifies the design using NRHA.
The design engineer should explicitly
note exceptions taken to the code
provisions or enhanced performance
criteria at the beginning of the linear
design phase, and identify appropriate
steps taken to ensure the NRHA results
will substantiate the exceptions and/or
enhancements to performance levels.

16 | Introduction
Step One: Service-Level Design
and Evaluation
The service-level earthquake (SLE)
evaluation and design step is used
to verify that the proposed structural
system meets the minimum required
strength and stiffness for earthquake
resistance under SLE demands. This
typically corresponds to immediate
occupancy performance. The linear
elastic modal response spectrum
analysis does not utilize the prescriptive
provisions of R, Ωo, ρ, and Cd defined in
ASCE 7. Expected material properties are
typically utilized.
It is noted that explicit verification of
immediate occupancy would require
significant NRHA, using demand-
appropriate ground motions. Instead,
as traditionally done in code-based
designs, performance is sufficiently
implied using linear Response Spectrum
Analysis (RSA) with demand-to-capacity
ratios limited to relatively low values,
such that an essentially elastic response
is anticipated.
If wind demands are expected to control
any aspects of the structural design, it
is during this step that the wind design
should also be completed. Where
wind demands control over the SLE
demands, the structural element should
be designed for these higher demands
and verified during the next step for
MCE demands.
Step Two: Maximum Considered
Earthquake (MCER) Level Evaluation
The MCER level evaluation step is
used to verify the minimum required
strength and stiffness for earthquake
resistance under MCER level demands.
The NRHA and verification does not
incorporate the prescriptive provisions
of R, Ωo, ρ, and Cd of ASCE 7, and
expected material properties are
utilized when determining component
stiffness. The appropriate material-
specific strength reduction factors
typically follow local practice and are
dependent upon the PBSD guidelines
followed; they are typically taken equal
to the values specified in the material
standard (e.g., ACI 318 or ASCE 341) or
equal to unity.
Other methods of design and
verification have been proposed and
successfully utilized on projects, but the
two-step procedure described above
is currently the most common. Some
jurisdictions may require additional
design verifications at the Design
Earthquake (DE) level.
For the geotechnical engineer, site-
specific response spectra for each
level of seismic shaking are to be
developed. Target response spectra
such as conditional mean spectra may
be developed for use as ground motion
selection and scaling targets for ground
motion selection and scaling. Finally,
one or more suites of ground motions
will be selected and modified as agreed
with the design review panel.
Site soil information needed for
development of response spectra, target
spectra, and ground motion selection
and development must be identified
early in the project to ensure that
physical sampling of site soils, such as
boring, obtain needed information.
Artificial earthquake records should not
be used.
1.9 Peer Review
The purpose of a peer review
panel is to provide expertise to the
building permitting process that
might not otherwise be available
to jurisdiction reviewers. The local
jurisdiction often divides the review
responsibility between lateral systems
and gravity systems. In these cases
the local jurisdiction maintains review
responsibility of gravity system code-
conformance review, while the peer
review panel focuses on PBSD of the
lateral system and its effect on the
design of the gravity system. The local
jurisdiction should remain very close
to the peer review process. Typically,
the local jurisdiction is copied on all
official correspondence related to the
peer review process and is invited to
meetings between the peer review
panel and Structural Engineer of
Record (SEOR).
A peer review panel often consists
of a minimum of three members: a
well-established practicing structural
engineer with a high level of experience
and expertise with the type of
seismic force-resisting system being
proposed, a professor/expert with
significant experience in the types
of analytical modeling and systems
being proposed, and a ground motion
specialist (e.g., seismologist) to review
the appropriateness of selected and
developed input ground motions.
Generally, review panel members
are retained by the local jurisdiction
or the developer. The practicing
For the structural
design engineer, PBSD
is typically a two-step
process where the
structural engineer
first proportions the
building using linear
analysis methods
(typically RSA) and
then verifies the
design using NRHA.
Introduction | 17
A structural
designer working
with the developer
and appropriate
approval authorities
should consider
the implications of
significant damage
and high repair costs
for a tall building
system when
establishing PBSD
criteria for a project.
structural engineer on the panel is
often identified as the lead peer review
panel member and is responsible
for maintaining a comment log and
ensuring other members of the panel
attend appropriate meetings. It is
critical that no conflicts of interest occur
between individuals or companies
involved in the process. The peer-review
panel members need not reside in the
jurisdiction, but should be well versed
in the local jurisdiction requirements
and governing building codes.
The PBSD process begins with regular
meetings early in the process that
increase in frequency towards the
end of the design. This process helps
develop confidence in the design
and encourages knowledge-sharing
between the peer review panel and
SEOR. This regular exchange of ideas
has led to a significant enhancement
in the body of knowledge for firms
practicing PBSD, and is often viewed as
an opportunity to advance the broader
practice of structural engineering.
18 | Introduction
When some level of new development
occurs during the design, engineers are
encouraged to produce publications for
dissemination and advancement in the
practice of PBSD.
The scope of the structural design
review services may include, but is not
be limited to, review of the following:
a. Earthquake hazard determination.
b. Site-specific ground motion
characterization.
c. Seismic performance goals.
d. Basis of design, design methodology,
and acceptance criteria.
e. Mathematical modeling and
simulation.
f. Interpretation of analysis results.
g. Member selection and design.
h. Detail concepts and design.
i. Construction documents, including
drawings and specifications.
j. Innovative technology
implementation, such as isolator or
damper testing requirements and
quality control procedures.
k. At the discretion of the jurisdiction,
the scope for services for the
peer review panel may include
the review of other building
aspects, including design for
wind resistance, design of special
foundation or earth-retaining
systems, or the design of critical
non-structural elements.
Review Process
The review process is a simple
framework of regular meetings, a formal
comment log, and a calculation report.
This regular interaction generates
confidence and allows the peer review
panel time to develop comments and
investigate, as needed, SEOR proposals.
An example general schedule of review
panel meetings is as follows:
Concept Design
A peer review panel could be arranged
during this time, but it is not required.
The SEOR focuses on the development
of a structural system fitting with site
demands and architectural needs. The
need for PBSD is identified and the
developer is made aware of implications.
If PBSD is new to the jurisdiction, local
officials must be engaged in a series
of in-depth meetings to explain and
develop an agreeable framework for
permitting and approvals. The project’s
Basis of Design, with all needed design
criteria, is developed by the SEOR.
Schematic Design
Peer review panel candidates are
submitted by SEOR to the local
jurisdiction for review and selection.
Peer review panel members enter into
contract with jurisdiction or developer,
and a Basis of Design document
proposal is submitted by the SEOR to
the peer review panel for review and
comment. The document is not finalized
during this phase of design, but key
issues should be discussed. In some
cases a draft Basis of Design document
is submitted to the local jurisdiction
with the site permit during this phase
of design. Ideally, the peer review panel
submits an initial list of NRHA analysis
results expected of the SEOR.
Typically, the developer and architect
engage in one or more pre-application
meetings during this phase. The SEOR
must notify and engage the local
jurisdiction regarding PBSD at this time,
if PBSD is not already established in
the jurisdiction.
Detailed Design
The proportioning of lateral system
members is finalized, and the NRHA
begins in the middle of or near the end
of this phase of design. The Basis of
Design document should be agreed to
with the peer review panel, although it is
understood that it could be modified as
the design progresses. The seismologist
from the peer review panel is engaged
during this phase for ground-motion
criteria selection to ensure they are
appropriate for the site and represent
the intent of code requirements.
Peer review comments should be
made in sequence with the progress
of design to avoid time-consuming
rework. This requires on-going
communication between the peer
review panel and design engineer.
Construction Documents
Final NRHA is conducted early in this
phase of design, and the majority
of global building assessments,
primary member demands, and key
component designs are submitted
to the peer review panel by the
mid-point of this phase. Meetings with
the peer review panel should occur
regularly, with progress presentations
of the design and results. The peer
review panel may request additional
studies be conducted to investigate or
confirm that building behavior satisfies
the criteria in the Basis of Design
document.
Early construction packages are
common for tall buildings, due to
length of schedule. The design team
and peer review panel should adjust
and prioritize accordingly. It is common
for the peer review panel to issue a
findings letter specific to foundations/
basements prior to the final findings
letter to allow below-grade permitting.
Approval
The end of the peer review process
occurs when the peer review panel and
SEOR agree that the criteria of the Basis
of Design are satisfied. The peer review
panel then typically writes a joint letter
to the local jurisdiction showing their
findings and recommendations for
approval by the local jurisdiction. In
most jurisdictions accepting PBSD, the
local jurisdiction is willing to review its
portion of the documents (i.e., gravity
system components) while the peer
review panel is finishing its review of the
lateral system.
1.10 Appropriate Usage of PBSD
PBSD represents a significant
advancement in the field of structural
engineering. The incorporation of PBSD
into tall building design has produced
significant cost savings to developers,
and a better understanding of how
structures respond to ground shaking.
In some urban centers, the use of
PBSD methods for the rehabilitation
of seismically deficient buildings has
become mandatory. Although ASCE
41-13 includes provisions for PBSD, an
engineer should be cautious about the
wholesale use of this retrofit standard
for new construction, since ASCE 41-13
focuses more on improving seismic
response of existing buildings, rather
than producing designs that are strictly
compliant to current standards. PBSD
is decidedly appropriate for all tall
buildings in regions of moderate to
high seismicity. It can be particularly
beneficial to tall building designers, as
it affords increased flexibility for seismic
force-resisting systems, which may
include non-traditional components and
configurations not recognized by code.
Careful attention must be given to
the transposition of methods and
numerical models used in the United
States to other countries. Differences
in code seismic hazard definitions,
material standards (i.e. ASTM
standards), component detailing, and
local construction practices must
be considered.
Because explicit nonlinear modeling
is used, the numerical model must
reflect the condition expected to be
constructed. For example, all ductile-
reinforced concrete components, such
as link beams, in the United States utilize
ASTM A706 GR.60 rebar. If a project in
South Korea, for example, utilizes the
NAISH (2013) link beam numerical
model, the project would be required to
utilize A706 GR.60 rebar or equivalent.
Thus, seemingly small decisions by the
structural engineer can have significant
impact on the project.
While PBSD does offer many benefits, it
may not be appropriate for all projects.
PBSD projects require additional
effort by the structural design team,
which could affect the design and
procurement process. This is especially
true for projects proposing PBSD in
jurisdictions where an approval process
is not established. Furthermore, if
code-level performance is adequate
and traditional structural systems are
cost-effective, PBSD may produce
limited benefits. This is particularly
important in light of skyrocketing
building heights in urban centers and
increasingly complex building forms
designed by architects. This flexibility
allows the structural engineer to
specify a more project-specific seismic
force-resisting system appropriate for a
particular building, instead of forcing a
prescribed system on a building.
It is important to also consider the
consequence of structural damage
levels in tall building systems. Tall
building component performance is
often designed to satisfy life safety
(see Section 1.3). A structural designer
working with the developer and
appropriate approval authorities
should consider the implications of
significant damage and high repair
costs for a tall building system when
establishing PBSD criteria for a project.
Given the substantial potential
monetary cost and down-time cost
of rehabilitating a tall building after a
seismic event, developers may want
to quantify initial costs and consider
enhanced performance objectives to
limit their risk.
Introduction | 19
2.0 Site-Specific Seismic
Hazard Assessment
2.0 Site-Specific Seismic Hazard Assessment
2.1 Introduction
The seismic design of structures
should include proper evaluation
of seismic hazards. These hazards
include the level of ground shaking
for structural design and liquefaction,
ground deformations, loss of bearing,
and slope stability hazards that
may impact the performance of
foundations. PBSD guidelines typically
recommend establishing site-specific
ground motions, rather than using a
prescriptive code spectrum, scaled
up for the Risk-Targeted Maximum
Considered Event or scaled down for
the Service Level Earthquake (SLE).
Open communication between the
geotechnical and structural engineers
is critical in the development of
site-specific ground motions for tall
buildings. These communications
ideally should be initiated at the outset
of projects. Information regarding
structural periods for the fundamental
and higher modes is imperative for the
development of appropriate ground
motion criteria (e.g., time series) for
structural evaluations and design. If
a detailed Soil-Structure-Interaction
(SSI) analysis is performed, additional
information regarding location, spacing,
and dimensions of structural elements
(e.g., basement walls, basement floors,
mat foundation, deep foundation
elements, etc.) and properties of the
structural elements (e.g., Poisson Ratio,
Young’s/Elastic Modulus, unconfined
compressive strength for concrete
elements, minimum yield strength
for steel elements, moment of inertia,
cross-sectional area, etc.) are needed.
Levels of shaking for design (i.e.,
earthquake-induced forces) are in
general quantified by acceleration
response spectra. A response spectrum
is the maximum response of an elastic
single-degree-of-freedom (SDOF)
simple damped oscillator for a number
22 | Site-Specific Seismic Hazard Assessment
of natural periods and a specified critical
damping ratio. In recent building codes,
starting with ASCE 7-10, the spectra for
design are quantified in the maximum
direction rather than the geometric
mean (geo-mean). The quantification
of the maximum-direction spectra
are generally done by correcting the
SDOF geometric mean spectrum
via correction factors. However, the
correction factors have been developed
by computing the maximum response
of an elastic two-degree-of-freedom
(TDOF) simple damped oscillator and
comparing it to the SDOF oscillator
response. Walker et al. (2010) present
a comprehensive discussion of the
maximum-direction spectra.
2.2 Developing Site-Specific Target
Response Spectra
Generally, the level of shaking
quantified as a Target Response
Spectrum could be determined using
probabilistic seismic hazard analysis
(PSHA) (Cornell 1968 & McGuire 2004),
deterministic seismic hazard analysis
(DSHA), Conditional Mean Spectrum
(Baker and Cornell 2006), and NIST
(2011) or ground response analysis.
All of these methods result in a target
spectrum for ground motion scaling
and matching.
Probabilistic Seismic Hazard Analysis
(PSHA)
In a PSHA, a level of ground shaking is
defined as a probability of exceedance
in a given period of time, typically 50
years. The spectral values are developed
for the same mean annual frequency
of exceedance, which represents
a uniform hazard, hence the term
Uniform Hazard Spectrum (UHS). A
UHS includes earthquake hazard from
all considered sources in the area of
study and does not represent a single
earthquake. A typical level of hazard
defined in US building codes is the
Risk-Targeted Maximum Considered
Earthquake (MCER), which corresponds
to a two percent probability of
exceedance in 50 years; however, in
highly seismic active areas there is
typically a deterministic cap for the
MCER (see Section 2.5).
Deterministic Seismic Hazard Analysis
(DSHA)
A DSHA is a scenario earthquake
approach. It is a relatively simple
approach that considers the occurrence
of an earthquake of a particular
magnitude, typically a maximum
earthquake on a particular fault and
the closest distance to the fault. Unlike
PSHA, DSHA does not explicitly consider
the probability or frequency of the
occurrence of a particular earthquake.
Uncertainty is considered through
the use of standard deviation of the
predictive relationships. The typical
spectral levels considered are the
median or 84th percentiles.
Conditional Mean Spectra (CMS)
CMS is an alternative target spectrum,
determined either using PSHA or
DSHA. It can be used in the selection
and scaling/matching of time series
for nonlinear structural analysis. CMS
provides a methodology such that the
expected mean response spectrum
is conditioned on the occurrence of a
target spectral acceleration value at the
period of interest. Because the Uniform
Hazard Spectrum (UHS) is a summation
of hazards from all sources, it does not
represent a scenario earthquake and
provides higher spectral values than the
CMS at all periods except the period
of interest. Therefore, CMS addresses
some of the conservatism inherent in
using the UHS as the target spectrum,
and can be used as the basis to develop
and select an appropriate suite of time
series for different spectral periods.
Figure 2.1 presents an example of
the PSHA, DSHA, and CMS for three
5Figure 2.1: Example of Site-Specific Spectra. © Langan
conditioning periods for the DSHA
spectral levels. CMS were developed for
periods of 1.1, 2.3, and 5.6 seconds.
Ground Response Analysis
Ground response analysis is a
computational technique based on the
theory of wave propagation through
the soil. For this analysis, an idealized
soil column is shaken by an earthquake
(input) time series at the base layer.
The nonlinear soil behavior is modeled
by an equivalent-linear approach
(SHAKE-91; Idriss and Sun, 1992) and
nonlinear approach (DEEPSOIL; Hashash
et al. 2015 and D-MOD 2000, Matasovic
and Ordonez 2011, etc.). To quantify the
interaction of the structure with the soil,
SSI analyses are appropriate. While these
types of evaluations are not routine and
are not required, they are becoming
more common in the development of
site-specific ground motions for tall and
supertall structures. SSI models, which
use nonlinear ground-response finite-
element or finite-difference models
of the soil and structure, quantify the
stress-strain behavior of soil material
in a more direct fashion. It should be
noted that for a two-layer site condition
(e.g., 40 meters of fill and soft clay over
rock), the use of NGA-West2 attenuation
relationships, which are based on Vs30
values (average shear wave velocity in
the top 30 meters, measured from the
ground surface or below the basement
level, see ASCE 7-16) could result in an
overestimation of long-period spectral
values, which are important for tall
buildings. For these site conditions,
it is suggested to develop ground
motion at the surface of rock/firm soil
depth and perform ground-response
analysis to arrive at more reasonable
ground-surface or basement-level
spectral response values. Two- and
three-dimensional nonlinear SSI
analyses using computer programs—
such as FLAC, Plaxis, LS-DYNA, SASSI,
ADINA, OpenSees, and Midas—are
some of the modeling techniques
used by geotechnical practitioners. In
addition, guidelines for incorporation of
kinematic and inertia interaction effects
are provided in NIST (2012).
2.3 Range of Structural Periods
for Consideration
The use of CMS as a tool for selection
and scaling/matching of time series has
become the preferred method for the
development of site-specific ground
motions for structural evaluations and
design. As presented in Figure 2.1,
CMS is equal to the target spectrum
at the conditioning periods and is less
than the target spectrum for other
periods. The CMS values presented in
Figure 2.1 for the three conditioning
periods are: 0.91g for a 1.1-second
conditioning period, 0.50g for a
2.3-second conditioning period and
0.20g for a 5.6-second conditioning
period. Therefore, it is imperative
that an adequate number of CMS is
developed such that the drop in the
spectral values is not too severe. The
provisions of ASCE 7-16 require that
the envelope of the CMS does not
fall below 75 percent of the target
spectrum. Typically, two conditioning
periods, one representing the
fundamental modes of the structure
and one representing the second or
higher mode vibrations, are considered.
However, three or more conditioning
period of CMS may be required in the
event that the fundamental mode
period and the higher mode period
are too far apart. The determination
of these spectral periods is the
responsibility of the structural engineer
and should be part of the interactive
communication between the structural
Site-Specific Seismic Hazard Assessment | 23

ASCE 7-16 defines
near-fault sites as
those within 10
and 15 kilometers
of faults, capable
of generating
earthquakes with
magnitudes greater
or equal to 6.5 and
7, respectively.
and the geotechnical engineers.
Alternatively, an envelope of two
higher modes of conditioning-period
CMS may be considered, for the
purpose of limiting the number of
structural evaluations. For instance, in
Figure 2.1 for the short period CMS,
one could consider an envelope of the
1.1- and 2.3-second CMS.
2.4 Near-Fault Ground Motions
ASCE 7-16 defines near-fault sites as
those within 10 and 15 kilometers
of faults, capable of generating
earthquakes with magnitudes greater
or equal to 6.5 and 7, respectively.
Recent research suggests that pulse-
type ground motions can occur up to
distances of about 15 to 20 kilometers,
and polarization of ground motions
in the direction normal to the fault
rupture can occur for distances up
to about five kilometers (NIST 2011).
Recordings from earthquakes suggest
24 | Site-Specific Seismic Hazard Assessment
that the motions in the distances less 2. The SLE is defined as a PSHA level
than or equal to five kilometers from having a 50 percent probability of
fault rupture have higher long-period exceedance in 30 years (a 43-year
(i.e., greater than one second) spectral return period) for a 2.5 percent
values due to forward directivity, (i.e., critical damping ratio.
rupture propagating forward towards
the site and in the direction normal to
the fault rupture). These effects have 2.6 Information Required From
been known and demonstrated by Structural Engineer
various investigators (Somerville et al.
1997; Abrahamson 2000; Golesorkhi As previously discussed, open and
& Gouchon 2002). The earlier studies
considered these effects to be
interactive communication between
the geotechnical and structural
broadband; however, recent studies engineers is imperative in the
have shown that these effects tend to development of site-specific ground
be narrowband (Abrahamson 2011 and motions for tall buildings. Information
Spudich et al. 2014). Furthermore, in regarding structural periods for the
the near-fault region, ground motions fundamental and higher modes is
tend to be polarized in the directions required for the development of
normal (larger than average spectral
values in the long periods) and parallel
appropriate time series for structural
evaluations and design. Additional
(smaller than average spectral values in information such as location, spacing,
the long periods) to the fault rupture. and dimensions of structural elements
As such, in the near-fault region, these (basement walls, basement floors, mat
effects should be considered in the foundation, deep foundation elements,
development of site-specific spectra. etc.) and properties of the structural
For distances larger than five kilometers elements (sizes, Poisson Ratio, Young’s/
from the rupture, the polarization of Elastic Modulus, material compressive
the ground motions tend to be random and yield strengths, etc.) are needed if
(Shahi and Baker 2014); the maximum SSI modeling is performed.
values could occur at random directions
to the fault rupture.
2.7 Time Series Record Selection
and Development of Site-Specific
2.5 MCER- and Service-Level Time Series
Earthquake (SLE)
Proper selection and development
For the design of tall buildings, two of a suite of site-specific time series
levels of spectral values should is a critically important step in the
be considered: seismic design of tall buildings. The
intent in the process of selecting time
1. The MCER is defined as a Risk- series is to choose those that have
Targeted (MCE modified by risk a similar magnitude, distance, fault
coefficients) PSHA level of shaking mechanism (e.g., strike-slip, trust, etc.),
having a 2 percent probability of and duration. Because long periods are
exceedance in 50 years (a 2,475-year of special importance to tall buildings,
return period) with a deterministic ground motions should have long
cap at the 84th percentile level of period attributes and signatures, such
the governing fault for a 5 percent as velocity and displacement pulses
critical damping ratio quantified in and long-period energy, which are
the maximum direction. appropriate. NIST (2011) provides
guidelines for the selection and Spectral Matching time series. However, both methods
development of site-specific ground In this approach, the shape of the can change the frequency content
motion time series. response spectrum of the original and/or characteristics of the original
time series is modified to match a time series if spectral matching is not
ASCE 7-16 requires at least 11 pairs target response spectrum (see Figure done properly. The advantage of this
of ground motion for each target 2.2). Time- or frequency-domain approach is that the spectra of the
spectrum. In the past, a suite of seven matching can be used. In general, a matched time series is consistent with
ground motion pairs has been used. time-domain approach is preferred the site-specific target spectrum, and
As part of the Pacific Engineering because it minimizes the degree of as such controls the excessively high
Research Center’s (PEER) lifeline alteration and distortion to the original or low spectral values that may be part
projects, an applied research project
was undertaken to develop the
next generation of ground motion
prediction equations (GMPEs) or next-
generation attenuation relationships
(NGAs). Because of the PEER NGA
and NGA-West2 projects, access to
appropriately processed and quality-
controlled earthquake recordings
have become routine and relatively
simple. The earthquake recordings
from these projects can readily be
accessed and downloaded from the
PEER website. The provision of ASCE
7-16 that at least 11 pairs of ground
motions be used in performance-
based design is appropriate for the
design of tall buildings for each target
design spectrum.

For the development of site-specific

time series, the criteria in ASCE
7-16 could be used. In general, the
average of SaRotD100, defined as the
maximum spectral acceleration when
rotated over all horizontal orientations
of the suite, should not fall below the
target spectrum.

Spectrally Compatible Time Series

Traditionally, two approaches have
been used in developing site-specific
time series; these approaches are
spectral matching and amplitude
scaling. Both approaches have
advantages and disadvantages
associated with them, which are
discussed as follows.

5Figure 2.2: Spectrally Matched Time Series. © Langan

Site-Specific Seismic Hazard Assessment | 25

of the original recording. However, the
disadvantage of this approach is that
it collapses the natural variations of
the recorded motions, and eliminates
the dispersion in the ground motions
that may be of interest. Furthermore,
matching both components to
the target spectrum (e.g., MCER or
CMS) could be overly conservative.
To avoid undue conservatism, the
stronger horizontal component of the
seed motions could be matched to
SaRotD100, and the other horizontal
component to SaRotD50, defined as the
median spectral acceleration rotated
over all horizontal orientations.

Spectral matching can be performed

using computer software such as
RSPMatch developed by Abrahamson,
which is available to the earthquake
engineering community freely online or
other commercially available software
(e.g., ETABS 2015).

Spectral Amplitude Scaling

This is a relatively simple approach in
which a single scalar is used to modify
the spectral values of the original time
series. An example of spectrally-scaled
time series is presented in Figure
2.3. Because a single scalar approach
scales the entire spectrum by a single
value, there could be very large
discrepancies between the scaled
spectrum and target spectrum at any
5Figure 2.3: Spectrally Scaled Time Series. © Langan
period; however, the characteristics of
the time series do not change (except
for amplitude). The advantage of this
approach is that the natural signatures
of the recorded motions in terms of preserved. However, the spectral Hybrid Approach
the shape, relative amplitude and scaling method could result in the The concepts of this approach were
frequency content of acceleration, overestimation of ground motions originally developed by Mazzoni et al.
velocity, and displacement traces in a short period range. Figure 2.3 (2012). In this approach, scalar ratios
and spectrum are not altered, and presents an example of spectrally- of the spectral values are developed
as such the natural dispersion in scaled time series. relative to the target spectrum. For
the ground motions for design is example, the average maximum

26 | Site-Specific Seismic Hazard Assessment

direction spectrum, RotD100, of
the suite of the considered ground-
motion time series is computed.
Then the ratio of the average
RotD100 to the target is developed
for the periods of interest following
a prescribed rule, for example, that
the average RotD100 spectrum
not fall below the target (Ratio =
Average RotD100/Target). Once
these ratios are developed, then
each component of the suite of
records is divided by these ratios.
These modified spectra are used
5Figure 2.4: Unscaled RotD100 and Target Spectra. © Langan
as the target spectra for spectral
matching of the corresponding
time series. The advantages of this
approach over amplitude scaling
are that it: (1) controls and reduces
the impact of a “large” scalar factor
that could result in excessively high
spectral values that may not be
physically real and (2) eliminates
the issue with spectral matching
both components of the time series
to a single spectrum, which in turn
eliminates the spectral variations
that naturally exist between the
two components of the recorded
motions. In effect, this approach 5Figure 2.5: Ratio of Average RotD100 to Target Spectrum. © Langan
preserves the dispersion in spectral
values that are of interest in
structural design, while eliminating
overly large spectral values. Figures
2.4–2.6 present an example of this
hybrid approach.

As indicated in the beginning of

this section, spectral matching
and scaling have traditionally
been used as the two standard
approaches in the development of
site-specific time series. However,
because the hybrid approach
provides an attractive alternative
in the development of site-specific 5Figure 2.6: Comparison of Scaled and Original Spectra. © Langan

Site-Specific Seismic Hazard Assessment | 27

time series in that it mostly preserves
the natural variations (i.e., dispersion)
in the design ground motions and
also controls potentially very large
spectral values, it is becoming more
commonly used by practitioners.
Guidelines for Selection of Time Series
Selection of recorded time series
is an important step in developing
site-specific ground motions. The
intent in this selection process is to
choose time series that have a similar
magnitude and distance as that of the
recommended target spectrum.
Recorded time series may be
downloaded from the PEER NGA-
West2. The attributes that may be
considered in this approach may
include a range of magnitude,
distance, Vs30, and spectral shape.
Computer software such as
QuakeManager (Earthquake Solutions,
The application of
the time series and
the orientation with
which they should be
applied to structures
depends on whether
the site is near-field,
distances less than
five kilometers from
the fault, or far-field.
28 | Site-Specific Seismic Hazard Assessment
2015) may be used to select the
records from a large collection of
recordings. The algorithm used in
QuakeManager calculates the sum
of the squared error (SSE) between
the target spectrum (CMS) and
the time series, using single or
bidirectional parameters such as
DRot100, geometric mean, Square
Root of the Sum of the Squares, etc.
for each component or pair of time
series in the database. The proposed
time series may be selected generally
based on the least SSE; however,
one should also use judgment,
considering magnitude, duration,
pulse characteristics, and other criteria
to include time series from other
earthquakes that may not have had
the lowest SSE.
Other criteria that should be
evaluated in the selection process are
the estimates of duration and Arias
intensity. Relationships by Silva et
al. (1996) and Kempton and Stewart
(2006) may be used to estimate
the durations, D5-75 and D5-95, for the
considered range of earthquake
magnitudes. The empirical
relationship developed by Travasarou
et al. (2003) may be used to estimate
Arias intensity.
Because of the long fundamental
period of tall structures and the
damaging effects of pulse-type
recordings, time series that have
velocity and displacement pulses
should be considered for the suite of
records that will be used for structural
evaluations and design. As such,
relationships by Hayden et al. (2014)
and Shahi and Baker (2011) may be
used to assess the percentage of
pulse-type records that should be
included in the suite. Pulse periods
are generally narrowband; therefore,
variation in the periods of the pulses
should also be taken into account
when selecting time series. In the
selection of pulse-type motions,
consideration should be given to the
selection of time series with pulse
periods that are within ±2.0 seconds
from the target CMS period. Also, the
lowest useable frequencies of the seed
motions should be checked to make
sure that they cover the period range
of importance for structural response.
Application of Time Series and
Considered Orientations
The application of the time series
and the orientation with which they
should be applied to structures
depends on whether the site is
near-field (distances less than five
kilometer from the fault) or far-field.
For near-field sites, the time series
should be rotated in the normal
and parallel directions to the fault
rupture. For far-field sites, the
time series should be applied in
a random direction. In order not
to preferentially apply time series
either conservatively (i.e., strongest
shaking applied in the weak axis
of the structure) or liberally (i.e.,
weakest shaking applied in the strong
axis of the structure), it is useful to
obtain the preferred direction of
the maximum response of the time
series. This can be obtained from
bidirectional acceleration response
(orbital) plots at periods of interest.
An example of such a plot is shown
in Figure 2.7. Assuming the vertical
axis is north–south, the direction
of the maximum response is in the
northwest–southeast direction. These
plots can shed light on the direction
 Spectral Acceleration Sy (in/sec2)
Spectral Acceleration Sx (in/sec2)
5Figure 2.7: Bidirectional Acceleration Response Plot (Calexico – T=5.5 Seconds, Damping Ratio 5%). © Langan
of maximum response as it relates to
the building axes, and subsequently
assess random direction of
application of the time series.
2.8 Vertical Ground Motions
Vertical ground motions are not
commonly required in the design of
tall buildings. However, in the event
that vertical grounds are required,
they can be developed for projects.
Codes generally recommend a
constant value of 2/3 times the
horizontal spectrum as a simplified
approach for the development of the (2016) present some of the more
vertical spectrum. However, for sites in recent work on the development
close distances to the fault, less than of vertical spectra. Vertical time
about 10 kilometerss, the recorded series could be obtained from the
data have shown that vertical spectra PEER website, which could be either
at short periods, usually less than 0.3 spectrally matched or scaled to the
second, are higher than horizontal target spectrum to develop site-
spectra. Over the years, a vertical specific vertical ground motions.
spectrum has been developed by
using ratios of vertical-to-horizontal
spectra and applying them to the
horizontal spectrum. However, ground-
motion prediction equations have
been developed to estimate vertical
spectra. Gulerce and Abrahamson
(2011) and Bozorgnia and Campbell
Site-Specific Seismic Hazard Assessment | 29
3.0 Design Using Linear
Analysis
3.0 Design Using Linear Analysis

The initial proportioning of a building specific performance objectives are TBI and LATBSDC, with key parameters
consists of a complete design targeted at SLE-, DE-, and MCER-level in Tables 3.1 & 3.2.
process whereby all members of the ground shaking, verification at each
seismic force-resisting system are level may be required. The analysis model should include
proportioned. Linear design can be all lateral force-resisting elements,
done using SLE-, DE-, or MCER-level primary gravity system elements, and
earthquake demands. Most engineers 3.1 Modeling and Analysis basements. P-Delta effects should
prefer using SLE demands, with design be included. Slab openings affecting
methods appropriately adjusted For initial design using linear analysis, diaphragm stiffness should be included
for the lower demand level. Some modeling and design methods with semi-rigid diaphragm modeling.
engineers have utilized DE or MCER appropriate for the level of earthquake
level demands based on particular demands should be considered. Materials
building types or preference. When Response spectrum analysis is typically For SLE-level design, expected material
this is done, additional verification used. As mentioned above, for this properties should be utilized for
at DE or SLE may be required to document SLE demands are used for realistic estimates of stiffness. For
substantiate building performance and initial linear design. Material strength projects using ASCE 7 criteria and
code equivalency. SLE-based design is and stiffness assumptions, section specified ASTM material standards,
primarily considered in this document property modifiers, and material unless more detailed justification
and is described in detail in PEER/TBI strength reduction factors appropriate can be produced, expected material
and LATBSDC documents. The intent for SLE demands should be used and properties as shown in Table 3.1
of designing using SLE-level demands differ from DE-level assumptions. can be used. In jurisdictions not
is to inherently satisfy DE performance Material and section property using ASCE 7 and associated ASTM
objectives by verifying performance modifiers appropriate for SLE-level standards, robust testing of local
under SLE and MCER demands. If design are described in detail in PEER/ materials or historical information

Material Expected Strength

Expected Yield Strength, fye , psi Expected Ultimate Strength, fue , psi

Reinforcing Steel A615 Grade 60 70,000 106,000

A615 Grade 75 82,000 114,000
A706 Grade 60 69,000 95,000
A706 Grade 80 85,000 112,000

Structural Steel*** Hot-rolled structural shapes and bars

ASTM A36/A36M 1.5 fy* 1.2 fu**

ASTM A572/A572M Grade 50 1.1 fy* 1.1 fu
ASTM A913/A913M Grade 50, 60, 65 or 70 1.1 fy* 1.1 fu
ASTM A992/A992M 1.1 fy* 1.1 fu

Plates
ASTM A36/A36M 1.3 fy 1.2 fu
ASTM A572/A572M Grade 50, 55 1.1 fy 1.2 fu

Concrete f'ce =1.3f'c†

*fy is used to designate the specified (nominal) yield strength of steel materials in this Guideline. It is equivalent to fy or fyt used in ACI 318 and Fy used in AISC (2006) standards.
**fu is used to designate the specified (nominal) ultimate strength of steel materials in this Guideline. It is equivalent to Fu used in AISC (2006) standards.
***For steel materials not listed, refer to Table A3.1 of ANSI/AISC 341-16
†
f'c = specified compressive strength. Expected strength f'ce is strength expected at approximately one year or longer. Note that the multiplier on f'c may be smaller for high-strength concrete,
and can also be affected by (1) use of fly ash and other additives, and/or (2) local aggregates.

5Table 3.1: Expected Material Strength. Source: PEER/TBI

32 | Design Using Linear Analysis

 Service-Level Linear Models MCER-Level Nonlinear Models
Component
Axial Flexural Shear Axial Flexural Shear
Structural walls (in-plane)
1
1.0EcAg 0.75EcIg 0.4EcAg 1.0EcAg 0.35EcIg 0.2EcAg
Structural walls (out-of-plane) – 0.25EcIg – – 0.25EcIg –
Basement walls (in-plane) 1.0EcAg 1.0EcIg 0.4EcAg 1.0EcAg 0.8EcIg 0.2EcAg
Basement walls (out-of-plane) – 0.25EcIg – – 0.25EcIg –

l  l 
Coupling beams with conventional or 0.07   Ec Ig 0.07   Ec Ig
1.0EcAg h 0.4EcAg 1.0EcAg h 0.4EcAg
diagonal reinforcement
≤ 0.3Ec Ig ≤ 0.3Ec Ig

Composite steel / reinforced concrete l  l 

1.0(EA)trans 0.07   ( EI )trans 1.0EsAsw 1.0(EA)trans 0.07   ( EI )trans 1.0EsAsw
coupling beams 2 h h
Non-PT transfer diaphragms (in-plane only)3 0.5EcAg 0.5EcIg 0.4EcAg 0.25EcAg 0.25EcIg 0.1EcAg
PT transfer diaphragms (in-plane only) 3
0.8EcAg 0.8EcIg 0.4EcAg 0.5EcAg 0.5EcIg 0.2EcAg
Beams 1.0EcAg 0.5EcIg 0.4EcAg 1.0EcAg 0.3EcIg 0.4EcAg
Columns 1.0EcAg 0.7EcIg 0.4EcAg 1.0EcAg 0.7EcIg 0.4EcAg
Mat (in-plane) 0.8EcAg 0.8EcIg 0.8EcAg 0.5EcAg 0.5EcIg 0.5EcAg
Mat (out-of-plane)
4
– 0.8EcIg – – 0.5EcIg –
1
Values are relevant where walls are modeled as line elements. Where walls are modeled using fiber elements, the model should automatically account for cracking of concrete and the
associated effects on member stiffness.
2
(EI)trans is intended to represent the flexural rigidity of the cracked transformed section. It is acceptable to calculate the transformed section properties based on structural mechanics
or to use (EI)trans = EcIg/5 + EsIs per ACI 318.
3
Specified stiffness values for diaphragms are intended to represent expected values. Alternative values may be suitable where bounding analyses are used to estimate bounds of
force transfers at major transfer levels. For diaphragms that are not associated with major force transfers, common practice is to model the diaphragm as being rigid in its plane. Flexural
rigidity of diaphragms out of plane is usually relatively low and is commonly ignored. The exception is where the diaphragm acts as a framing element to engage gravity columns as
outrigger elements, in which case out-of-plane modeling may be required.
4
Specified stiffness values for mat foundations pertain to the general condition of the mat. Where the walls or other vertical members impose sufficiently large forces, including local
force reversals across stacked wall openings, the stiffness values may need to be reduced.

5Table 3.2: Reinforced Concrete Effective Stiffness Values. Source: PEER/TBI

is needed to estimate appropriate account for cracking and damage to In jurisdictions not
expected material parameters. the components, through section
property modifiers with reduced using ASCE 7 and
PEER/TBI recommends the use of effective stiffness of the member. associated ASTM
expected material properties for Property modifiers are based on
analysis-model component stiffness, experimental testing. Since SLE standards, robust
but specified material properties for
component strength capacity.
demands are often considered,
LATBSDC and PEER/TBI have published
testing of local
concrete section property modifiers for materials or historical
LATBSDC recommends expected use in SLE-and MCER-level events. The
material properties for analysis- application of property modifiers can information is
model component stiffness and have a significant impact on member needed to estimate
strength capacity. force levels and should be carefully
considered for each project. Other appropriate expected
Since MCER evaluation using NRHA
is also conducted, either method is
resources that engineers should review
include PEER/TBI (see Table 3.2), ASCE
material parameters.
valid, but the PEER/TBI method is 41-13 Table 10-5 for all concrete
more conservative. elements, and ATC 72-1 Table 4-1 for
link beams. For link beams reinforced
Section Properties with steel wide flanges, AISC 341-10
In linear elastic analyses, section Commentary H4 can be consulted. It
properties need to be reduced to should be noted that there are

Design Using Linear Analysis | 33

inconsistencies between these Category
Action
documents, and engineers should use Critical Ordinary Non-critical
judgment appropriate to their
Connections of braces to beams, columns and walls X
building, anticipated damage to
Axial demand on braces in Eccentric Braced Frames X
components, and experimental testing
to determine appropriate section Column splice forces X
property modifiers. Additionally, if

Structural Steel1
Axial loads on column X
performance objectives, and therefore Moments and shears on moment connections X
damage levels, differ from ASCE 7, the
Compression on vertical boundary elements of steel plate
engineer should reconsider property X
shear walls
modifiers appropriately. Compression on horizontal boundary elements of steel plate
X
shear walls
Damping Forces in members of transfer trusses X
A site-specific response spectrum All other force-controlled actions 2
X
should be used with 2.5 percent
damping applied, as recommended by Shear in beams, columns, and beam-column joints of special
X
moment frames
PEER/TBI.
Shear in columns not part of special moment frames X
Loading and Analysis Axial load in columns of intentional outrigger systems, or in
X
As recommended by PEER/TBI, load columns supporting discontinuous vertical elements

combinations including dead (D), live Combined moment and axial load in gravity columns3 X
(L), and earthquake (E) loads should be Shear and moment in transfer girders X
combined as follows: Shear in structural walls that are part of the primary lateral-
X
force-resisting system
1.0D + Lexp + 1.0 Ex + 0.3 Ey Shear and moment in basement walls X
Reinforced Concrete

Shear in coupling beams without special diagonal

1.0D + Lexp + 1.0 Ey + 0.3 Ex reinforcing4
X

Compression on struts in strut and tie formulations X

Where:
Tension on struts in strut and tie formulations X

In-plane shear in transfer diaphragms X

D = Self-weight plus superimposed
5

dead loads (i.e., finishes, exterior walls, In-plane shear in other diaphragms X
etc.) Force transfer from diaphragms to vertical elements of the
seismic-force-resisting system, including collector forces and X
shear-friction between diaphragms and vertical elements
Lexp = Expected Live Load (PEER/TBI
recommends 0.8L for live loads that In-plane normal forces in diaphragms other than collectors6 X

exceed 100 psf and 0.4L otherwise.
LATBSDC recommends 0.25L.)
Ex = Seismic demands applied in the
X-direction
Ey = Seismic demands applied in the
Y-direction
Since the anticipated level of hysteretic
behavior is assumed to be very low
under SLE demands, no response
modification factors such as R, Ωo, ρ, Cd
Shear in shallow foundation elements, including spread
X
footings and mat foundations
Moment in shallow foundation elements, including spread
X
footings and mat foundations
All other force-controlled actions2 X
1
Structural steel elements designed and detailed to conform to the prescriptive requirements of AISC 341 and AISC 358 need not be
evaluated in accordance with the criteria for force-controlled elements.
2
Other force-controlled items should be categorized considering the criticality of the action to the overall building performance. The
default category is shown as Critical.
3
As an alternative, column flexure combined with axial force can be modeled as a deformation-controlled action if appropriately detailed.
4
Coupling beam shear may be considered an ordinary action only if the consequence of element failure is minimal.
5
Where walls beneath transfer diaphragms are adequate to provide required lateral force resistance in the event of diaphragm failure,
transfer diaphragms may be treated as ordinary force-controlled actions.
6
Diaphragm chord forces fall into this category.
5Table 3.3: Typical Classification of Component Actions. Source: PEER/TBI

34 | Design Using Linear Analysis

prescribed by ASCE 7 for DE demands
are required.
3.2 Acceptance Criteria
Acceptance criteria are often
reported in two basic categories:
global and component behavior.
Since SLE demands are used, each is
adjusted from ASCE 7/ACI 318/AISC
341 requirements appropriately as
described below. Typical adjustments
from these codes are listed below.
Component Behavior
When considering component
behavior, the term “action” is often used.
An example of an “action” is shear in
a link beam or flexure in a column. A
single structural component has three
actions: axial, moment, and shear.
Actions are classified as deformation- or
force-controlled actions. Deformation-
controlled elements respond in
a ductile fashion, being able to
withstand larger deformations without
substantial loss in post-yield strength.
Force-Controlled elements are generally
less ductile and not able to withstand
significant deformations without
brittle failure. For example, flexure in
a moment frame beam would be a
deformation-controlled action, but
shear would be a force-controlled
response. A summary of typical
classification of component actions
from PEER/TBI is provided in Table 3.3.
SLE demands on the components of
the structure are checked against their
capacities φ(Un) (i.e., strength reduction
factor times the nominal strength)
using expected material properties.
As recommended by LATBSDC, the
demand-to-capacity ratios (DCR) should
use Φ = 1.0 and satisfy the following for
risk category II structures:
1. Deformation-Controlled Actions
DCR ≤ 1.50.
2. Force-Controlled Actions
DCR ≤ 0.70.
It is noted that PEER/TBI suggests
the same 1.5 limit for deformation-
controlled actions, but specified
material properties are used. For force-
controlled actions, specified material
strengths, core prescribed Φ values, and
a D/C ratio limit of 1.0 are used.
No distinction is given to critical and The consequence
non-critical actions in SLE design
checks, but are sometimes considered of failure is also
in MCER-level verification. considered in
For SLE design, some engineers will use defining an element
different factors based on experience,
typically lowering the force-controlled
as force-controlled
action DCR of some actions even further or deformation-
than 0.7 to avoid issues when verifying
the design under MCER demands using controlled, which is
NRHA. This is sometimes done when termed “criticality.”
considering wall shear action of a
core-only seismic system.
These D/C ratio limits are based
on experience and do not always
guarantee acceptable performance
in MCER evaluation. If changes to
geometry or effective stiffness are
made during MCER evaluation to
obtain acceptable behavior, the
SLE evaluation may need to be
reconsidered.
Global Behavior
If SLE-level demands are used, drift
should not exceed 0.5 percent of the
story height. If DE-level demands
are used, code specified drift limits
(typically two percent) should be used.
If MCER-level demands are used, drift
should not exceed 1.5 x DE-level drift
limits (typically three percent).
Design Using Linear Analysis | 35
4.0 Verification of Response
Under MCER Using NRHA
4.0 Verification of Response Under MCER Using NRHA

With the seismic force-resisting system
fully proportioned using linear analysis,
verification of performance under MCER
level shaking using NRHA is conducted.
The intent is to verify the design, but
some design modifications can be made
during this step to ensure design criteria
and performance levels are satisfied.
Generally, it is only required to repeat
linear analysis checks if the MCER analysis
and design leads to modifications
to the non-negotiable dimensional
proportioning of the structure.
MCER peak ground acceleration can be
4–8 times higher than SLE. It is expected
that deformation-controlled actions
of components of the structure will
exhibit inelastic response, and as a result,
a nonlinear analysis model that takes
this inelastic response into account is
necessary to appropriately evaluate the
building performance objectives.
With appropriate modeling applied,
global and component acceptance
criteria need to be verified for
conformance with specified
performance objectives.
Component Modeling
All finite elements are composed of
deformation and force-controlled
actions identified in the linear analysis
and design stage (see Table 3.3).
For components with force- and
deformation-controlled actions,
elements are composed of linear and
nonlinear responses. For components
with only force-controlled actions, fully
linear elastic assumptions are utilized.
Although there are a variety of
modeling methods available in a variety
of software packages, commonly
used component modeling methods
are described in the following list.
Further descriptions of these and other
component modeling methods can be
found in ATC-72 and ATC-114.
1. Wall Elements: Shear is a
force-controlled action and
modeled elastically. Axial and
flexural behavior are represented
with a series of nonlinear bar
elements. Three types are typically
included, and are referred to
as “fibers” representing vertical
wall reinforcement, confined
concrete, and unconfined
concrete. Each fiber has an
independent backbone curve.
A graphical example of a fiber
arrangement in a core wall is
shown in Figure 4.1. Here, each
red dot represents a steel fiber
and a concrete fiber. If the area
is shaded red the concrete fiber
is confined, if it is not shaded red
it is unconfined. Fiber properties
are determined by the material
within each zone. Typically two to
four fibers are located in boundary
zones and only two are located
in non-boundary areas between
boundary zones. This allows for
linear extrapolation of results
outside of the fibers if desired.
2. Coupling beams: These elements
are often modeled with “lumped”
plasticity as a mid-span hinge or
as flexural hinges at each end.
Behavior needs to be closely
matched to representative physical
testing, similar with material
strengths, detailing, and span/

4.1 Modeling and Analysis

Modeling of elements is significantly

more detailed with NRHA than RSA,
and appropriate care should be given
to all modeling assumptions. A much
higher level of fidelity in modeling
and results is possible with NRHA.
This leads not only to a more refined
understanding of building behavior,
but also to opportunities for modest
adjustments. For example, modeling
methods in NRHA help identify with
greater clarity the formation of a plastic
hinge zone, or if a more distributed
yielding is to be expected. This allows
for more appropriate detailing in high-
deformation areas, and more efficient
reinforcement layouts elsewhere.
5Figure 4.1: Example Graphical Fiber Element Arrangement. © Skidmore, Owings & Merrill LLP

38 | Verification of Response Under MCER Using NRHA

 Plastic hinges at ends of clear length
Elastic segments
5Figure 4.2: Element with Hinges at Each End (left) and Elements with a Hinge at Mid-Span (right). Source: CTBUH based on ATC-72
depth ratios. Care should be taken
to ensure the physical testing
substantiating the proposed
coupling beam is appropriate
(see Figure 4.2). Refer to ASCE 41
and published physical testing
results. In most cases, specific
nonlinear properties are published.
If only a hysteresis is available in a
publication, engineers may need
to overlay the hysteresis resulting
from a proposed nonlinear
backbone curve to verify their
material model.
3. Gravity Columns: Axial compression
is a force-controlled action and
modeled elastically. If limited
nonlinear behavior is permitted
in the column, a flexural hinge at
each end is appropriate. If flexure is
to remain elastic, linear response is
appropriate and reinforcement can
be verified after analysis. If columns
are modeled with nonlinear
behavior, refer to ATC 114 for
modeling information and ASCE 41
for rotation limits.
4. Slab Equivalent Frames: It is
common to represent floor framing
systems (beam or flat slab) using
an equivalent frame with lumped
plasticity at each end where
yielding is anticipated. This is done
to account for the “micro-outrigger”
effect of flat slabs on columns,
and to verify that framing rotation
Elastic segments
Stiff end zone Rigid-plastic shear hinge
limits are satisfied. In tall buildings,
floor framing systems will often
yield in the upper portion of the
tower, where the core lifts up and
the gravity columns do not. Often
flat-plate post-tensioned slabs are
utilized and can yield significantly.
Refer to ATC 114 for modeling
information and ASCE 41 for
rotation limits.
In some instances, deformation-
controlled actions are modeled
elastically. For example, while
basement walls are categorized as
deformation-controlled for flexure
and force-controlled for shear, their
non-linear behavior is not modeled, as
they are not anticipated to yield.
Nonlinear Behavior Modeling with
Lumped Plasticity
For deformation-controlled elements
with a lumped plasticity, a nonlinear
backbone curve is employed in the
mathematical model using the general
force deformation shape as shown in
Figure 4.3. The backbone curves (black
line) are established based on physical
testing (red line). Based on testing
results performance objectives levels
as defined by ASCE 41 (Immediate
Occupancy/IO, Life Safety/LS, Collapse
Prevention/CP). Generic backbone
curves can be generated using ASCE
41, which defines curves for particular
types of elements. It is important
that any lump plasticity model be
Stiff end zone
(if needed)
Moment strength section
substantiated by physical testing results
that are highly representative of the
planned application.
An example of the calibration of a
coupling-beam shear-hinge model with
a cyclic load deformation response is
shown in Figure 4.4. Note that where
actual test data is used for calibration,
the majority of the effort in the
calibration is in how the loading and
reloading is defined. Further examples
of backbone curves are included in
the design examples included in this
report. For further detail on backbone
construction and specification of energy
dissipation, see ATC 114/ATC 72.
Materials
For modeling force-controlled action
stiffness, expected material strengths
are used (see Section 3.1). For strength
capacity of force-controlled actions see
Section 4.2.
Section Properties
Stiffness properties are discussed in
Section 3.1.
Monitoring of Deformations
The monitoring of deformation-
controlled actions, such as core-wall
vertical strain, may require the definition
of additional monitoring elements if
the deformations are not recorded in
the format the engineer may desire for
comparison with acceptance criteria.
In some software programs these
Verification of Response Under MCER Using NRHA | 39
 F

FC
Fy
 Effective yield strength and
deformation (Fy and δy)
Fr  Effective elastic stiffness, Ke=Fy/δy
 Strength cap and associated
Ke deformation for monotonic loading (Fc
and δc)
 Pre-capping plastic deformation for
δy δc δr δu δ monotonic loading, δp
 Effective post-yield tangent stiffness,
Kp=(Fc-Fy)/δp
δp δpc
 Residual strength. Fr=κFy
5Figure 4.3: Backbone Curve and Parameters (ATC-72) (Red = Tested, Black = Idealized).  Ultimate deformation, δu
Source: CTBUH based on ATC-72

200 890 are called “strain gages” and are often

Test (CB24F) defined at boundary-zone extents and
Vn Hinge the corners of core walls, as shown in
100 445 Figure 4.5. These types of elements are
a convenience to aid in post-processing
of the results. Another example is link
Lateral Load [k] 0 0 Lateral Load [kN] beams, where plastic rotations are
used for comparison with acceptance
criteria, but total rotations may be
-100 445 desired for descriptions of behavior. A
frame element assigned negligible axial
stiffness with pinned ends can be used.
-200 890
-0.12 -0.06 0 0.06 0.12 Diaphragm Modeling
For floor slabs with limited openings
Beam Chord Rotation [rad] and appropriate proportions,
5Figure 4.4: Example of Backbone Curve Fit to Experimental Testing Date of Coupling Beam.
diaphragms can be modeled with a
Source: Naish, 2013 rigid diaphragm assumption. When
diaphragm stiffness is more flexible or
has significant influence, it should be
modeled with 2D shell elements with
appropriate section property modifiers.
Typically, podium- and basement-level
slabs are modeled due to the potential
of transferring forces between the core
and podium-only shear walls, which are
often added due to the increased mass
or eccentricity of podium levels.

It is not recommended to transition

from rigid diaphragms to semi-rigid
diaphragms and then back to rigid
diaphragms in analysis models. This
type of assumption has the potential
to create unrealistic spikes in shear
Frame Element Axial ‘Strain Pinned Frame Analysis Model at
Extension to Capture Gage’ Element to Monitor Centerline of Wall demands. Generally, once a semi-rigid
Strain at End of Wall Total Roation diaphragm is specified (at top of
5Figure 4.5: Monitoring of Deformations. © Skidmore, Owings & Merrill LLP

40 | Verification of Response Under MCER Using NRHA

podium, for example) semi-rigid
diaphragms should continue to be
specified for all lower floors. Although
this will increase computation time, it
can avoid unrealistic shear demands.
Mass Modeling
Seismic mass should be included in
the model for above- and below-grade
elements when NRHA is conducted. In
some software programs, it is lumped
at nodal locations. Having the mass
appropriately distributed throughout
the floor plan can help in appropriate
modeling of torsion and P-Delta effects
(see Figure 4.6).
Accidental torsion is not included
in the model of the structure unless
required by the local jurisdiction or peer
review process, or where the torsional
amplification factor from linear analysis,
Ax, as defined in ASCE 7, exceeds 1.5.
Where the torsional effects are significant
and are to be included, the plan mass
distribution in the mathematical model
should be adjusted to include the effect
(the effective center of mass is moved).
Where semi-rigid diaphragms are
employed, this requires that the applied
nodal mass at each floor be adjusted.
Because the model employs nonlinear
analysis, superposition of a static
torsional moment to reflect accidental
torsion is not appropriate.
If vertical earthquake accelerations
are to be included in the analysis
of cases where large cantilevers or
transfer conditions exist, this should be
identified early, due to its impact on
how the system mass is modeled. The
use of vertical earthquake accelerations
requires a defined vertical mass that is
distributed in plan. The mass should be
located at the connection to the vertical
gravity system to preclude spurious
vertical modes in the floor system.
Damping
The deformation-controlled elements
provide hysteretic damping through
inelastic deformations in the system,
but the analysis should include an
additional amount of viscous damping
to account for the additional inherent
damping provided by the building
as a whole. This inherent damping
comes from a number of sources that
are not explicitly modeled, including
plasticity in elements modeled elastically,
material damping (from molecular-level
interaction), and interfacial damping
(friction at connections and fit-out
elements such as partitions). Where
modal damping is used and the analysis
software allows, a small amount of
stiffness-proportional damping should
be included to attenuate higher mode
response, on the order of 2.4 percent
modal and 0.1 percent stiffness-
proportional Rayleigh damping.
When performing NRHA, the initial
elastic stiffness matrix has often been
used to form the Rayleigh damping
matrix, due to the high computational
cost resulting from the use of the
updated tangential stiffness matrix.
In analysis models utilizing concrete
fibers to model the nonlinear behavior
of RC shear walls, RC columns and/or
RC beams, the initial elastic stiffness is
calculated assuming that all concrete
fibers are initially uncracked. The same
stiffness-proportional damping term is
used after the concrete cracks. This can
lead to unrealistically large amounts of
viscous damping in concrete fibers after
cracking occurs. It is recommended
that in such situations, the stiffness-
proportional damping constant of the
Rayleigh damping matrix associated
with the concrete fibers be reduced
appropriately. For more on this
topic, the following references are
recommended: Chapter 18 of the
Perform 3D User’s Guide; and Finley A
Charney, “Unintended Consequences of
The monitoring
of deformation-
controlled actions,
such as core wall
vertical strain, may
require the definition
of additional
monitoring elements
if the deformations
are not recorded the
way the engineer
may desire for
comparison with
acceptance criteria.
Modeling Damping in Structures,” ASCE
Journal of Structural Engineering, Vol.
134, No. 4, April 2008.
A maximum ratio of three percent
additional proportional damping
is permitted by ASCE 7-16, but the
designer should carefully review the
lateral system and materials used to
determine the damping ratio that is
most appropriate. ATC 114 (see Chapter
3) and ATC 72 Chapter 2 provide a
detailed discussion on both historical
and researched damping ratios for
a variety of structures, and offer the
following equation for estimation of
critical damping ratios based on story
height and material type:
D = α/30 (for N < 30)
D = α/N (for N > 30)
Verification of Response Under MCER Using NRHA | 41
5Figure 4.6: Distribution of Nodal Masses (red dots). © Skidmore, Owings & Merrill LLP
Where:
D = Maximum percent critical
damping
N = Number of stories
α = Coefficient with a recommended
range of 60 to 120. In general, steel
structural systems would tend
toward the lower range (α = 60), and
reinforced concrete systems would
tend toward the upper range (α = 120)
This inherent damping is typically
modeled within analysis software
packages through the use of Rayleigh
damping, which is formulated with the
assumption that the damping matrix is
proportional to the mass and stiffness
matrices. Most commercially available
software packages will self-calculate the
42 | Verification of Response Under MCER Using NRHA
coefficients for Rayleigh damping based
on user input frequencies or periods.
It is critical that the designer select
the appropriate range for the first and
second-mode shapes to capture the
damped behavior of the system. It is
suggested to set the first period as the
dominant first-mode shape and the
second period low enough to ensure all
the desired mode shapes are included.
A value of 20 percent for the first
period is typically sufficient, but if the
structure has a significant amount of
high-frequency response, the designer
should be careful to not set the lower
bound too high, as the corresponding
modal responses below this threshold
will be essentially eliminated, due to
the high damping resulting from the
formulation.
Loading and Analysis
Similar to linear analysis, ASCE 7 would
require full dead loads and expected
live loads. Different from response-
spectrum analysis, an orthogonal pair of
seismic records is applied concurrently
under the “E” term.
1.0D + Lexp + 1.0E
Where:
D = Service dead load
Lexp = Expected live load (see
Section 3.1)
E = Orthogonal pair of MCER time
history records
The orientation of ground motions
is important and is typically random.
In near-field sites this should be
discussed with the peer review panel
for appropriate origination of near field
records in the suite of ground motions.
For sloped sites where there is
significant unbalanced lateral earth
pressure, this should be applied
with a load factor of 1.0, along with
the gravity load. The presence of an
initial unbalanced lateral load on the
lateral system can reduce earthquake
deformation capacity on one side
of an element hysteresis, due to the
non-zero starting point.
The dynamic analysis duration may
require additional time with no
dynamic force input to complete
the free-vibration cycles needed for
determination of residual deformations.
Where the structural model has slab
“outrigger” elements representing
floor framing, care should be taken
when a gravity load is applied.
Because the gravity load is “turned
on” all at once, significant column
shortening can occur at the top of
the structure. The column shortening
can pull down nonlinear outrigger
elements, which affects the nonlinear
hinge moments and rotations. If the
analysis program has a feature to
turn off elements for gravity load,
the outrigger slab system should
be included in this. An alternative is
to adjust the outrigger hinge force
deformation relationship, so that the
dynamic analysis starts at the correct
gravity load in the hinges.
Backstay Effect Studies
Many critical structural components,
such as transfer diaphragms, are
modeled elastically with a section
property modifier as discussed
previously. Transfer diaphragms
often occur at ground level or
elevated podium levels and create a
load-sharing between the central core
and additional basement or podium-
only shear walls. The actual section
properties during a MCER event are not
known, and are likely to vary from the
assumed section property modifiers.
Additionally, the influence of the
mat and soil stiffness can influence
the load distribution and should be
considered. Thus, a bounding study
is often done to evaluate upper and
lower bounding assumptions.
This study is often done from
the perspective of maximizing
the demands on the core and
minimizing the demands of the
transfer diaphragm, known as the
lower-bound study, and conversely
minimizing demands on the core
and maximizing the demands on
the transfer diaphragm (see Figure
4.7), known as the upper-bound
study. This is discussed in detail with
recommendations for upper- and
lower-bound property modifiers
and appropriate considerations of
foundation systems in PEER/TBI.
This bounding study has been shown
to be very influential on projects, and
should not be delayed until the final
stages of verification. Instead, designers
are encouraged to start this evaluation
immediately after running their
base-case analysis model, to obtain
results as soon as possible. This can be
particularly influential if soil springs are
not included in the base-case analysis
model, but are included in the backstay
effect studies.
Numerical Stability
Traditionally, one of the biggest
challenges in performing a nonlinear
time-history analysis involves
numerical stability and how it is
achieved by setting various time-
integration parameters within the
analysis software. It is likely that
the selected analysis software
used allows for various numerical
integration methodologies, which
can be explored through the various
software documentations and is not
the focus of this document. As the
choice of method is contingent on
striking a balance between accuracy
and solution speed, no one method
is favored here, and it is up to the
designer to review the methods
available to choose the best one for
their project.
In general, most stability errors may
be corrected by selecting a smaller
time step, but certain integration
methods may have factors that can
introduce numerical damping to the
solution and help to limit excessive
vibrations in higher frequency modes
that may cause numerical instabilities.
However, these techniques shall be
used sparingly, and caution should
be used to ensure an accurate result
is achieved, without introducing too
much additional numerical damping
just to converge a solution.
4.2 Acceptance Criteria
Component Behavior
Acceptance criteria using NRHA
differs from linear-response spectrum
analysis in several ways. Many code
provisions, including the use of a
strength reduction factor “Φ,” expected
material properties, and DCR limits,
need to be reconsidered for MCER-
level design and described in the
Basis of Design. All code requirements
still apply, except, where specifically
stated in the Basis of Design with
corresponding justification.
Deformation-Controlled Actions
Deformation-controlled actions
are not typically subject to code
force-based DCR limits; instead their
Verification of Response Under MCER Using NRHA | 43

V (modeled with nonlinear properties),
M
5Figure 4.7: Transfer Diaphragm Example. Source: CTBUH drawing based on PEER/TBI
deformations (i.e., rotation or strain
levels) are compared to test data results
and corresponding performance
objectives. This is described further in
ASCE 41, with deformation limits at
each performance level of common
structural elements. Typically, mean
deformations of all ground motions are
used for comparison to deformation
limits stated in the Basis of Design.
For example, the peak rotation of a
particular link beam from all ground
motions of a given suite are averaged
44 | Verification of Response Under MCER Using NRHA
tower core wall
main backstay
diaphragm
foundation
and compared to the stated
deformation limit, for verification
that demands are meeting
performance objectives.
Typically, for code-equivalent
buildings using ASCE 7, the following
limits are used. It is noted that
these are developed from ASCE
7 performance objectives and
associated ASTM standards. They
should be reconsidered when not in
full adherence to these standards.
 Concrete compressive strain,
Unconfined = 0.003
 Concrete compressive strain,
Confined = 0.015 (special
boundary element per ACI 318-14)
 Coupling beam plastic rotation
(radians):
 0.05 for diagonally-reinforced
(ASCE 41-13 Table 10-19)
 0.02 for conventionally-
reinforced (ASCE 41-13 Table
10-19)
 Or as substantiated by physical
testing (e.g., steel coupling
research by Motter)
The unconfined compressive strain
limit of concrete is well understood
to be 0.003. Testing and subsequent
analysis using Perform-3D by Wallace
(2007) has shown that Perform-3D
and perhaps other software may
underestimate compressive strains
by 50 percent of shear wall systems.
Additionally, since unconfined
concrete creates a brittle failure
the limit is divided by 1.5, resulting in
a limit of 0.01.
Some practicing engineers have
found this approach to unconfined
compressive strength overly
conservative, but until substantiating
tests verify analysis predictions of
Perform-3D, some engineers have
utilized this approach.
Force-Controlled Actions
Force-controlled actions use force-
based criteria. Code prescribed
methods of calculating capacity
are used (except for Φ and material
properties in some cases), but their
demands are based on a factored
 Risk Category from ASCE 7-10 Table 1.5-1
I 1.00
II
III
IV
5Table 4.1: Risk Reduction Factors (excerpt from LATBSDC).
average of the time-history results.
The capacity is compared to demands
based on the factored average of peak
shear from all ground motions of a
given suite.
For MCER acceptance criteria, an
additional classification is given to
force-controlled actions: critical and
non-critical actions (see glossary). An
example of a critical action would
be shear capacity of shear walls.
An example of a noncritical force-
controlled action is flexure in
transfer diaphragms.
The calculation of force-controlled
action capacities has different
approaches in PEER/TBI and LATBSDC.
PEER/TBI recommends the use of
specified material properties in
strength calculations, except for
actions specifically protected by a
well-defined yielding mechanism.
LATBSDC permits either specified or
expected material strengths, but with
different factors applied to demands
(see Table 4.1). Engineers should ensure
there is consistency between the
selected method and all other analysis
and design assumptions.
All applicable detailing provisions
of the code apply regardless of
whether it is deformation-controlled
or force-controlled, unless specifically
stated in the Basis of Design with
substantiating justification.
Risk Reduction Factor, κi
1.00
0.80
Value to be established by SPRP
Global Behavior
The following summarizes the typical
global behavior checks and limits for
Risk Category II structures:
 Story drift: three percent taken as
the average peak rotation of time
histories in a given suite, with a
maximum story drift for any given
response of 4.5
 Residual drift: one percent taken
as the average peak drift of time
histories in a given suite, with a
maximum story drift for any given
Where the
response of 1.5 percent structural model
Story drifts are typically checked in has slab outrigger
each principal building direction
independently. Residual drift limits
elements, care
should generally not be an issue for a should be taken
well-behaved system. If the designer
is finding difficulty achieving the drift
when a gravity load
limits provided, the lateral system is applied. Because
in the concept and analysis model
shall be reviewed thoroughly, as poor the gravity load is
residual drift performance is typically
indicative of dynamic instability.
“turned on” all at
once, significant
column shortening
can occur at the top
of the structure.
Verification of Response Under MCER Using NRHA | 45
5.0 Basis of Design
Example
5.0 Basis of Design Example
The purpose of a Basis of Design
(BOD) document or design criteria
document is to state deviations
from governing code requirements,
either exceptions or enhancements,
and describe subsequent methods
justifying these exceptions or
enhancements. Content will often
include descriptions of all structural
systems, description of design
procedure, performance objectives,
analytical modeling methods, and
acceptance criteria. It is not intended
to contain all information used for
the design of the building, but should
be a standalone document with
references to all needed information.
No structural engineering results
should be presented in the Basis of
Design document. Typically, Basis of
Design documents range from 10 to
25 pages in length. The BOD should
be included in the design drawings
for future reference by the building
owner, especially if exceptions to
code provisions are taken.
The Basis of Design document is
generally submitted to the peer review
panel and local governing jurisdictions
involved in building permitting for
review and comment early in the
building design process. In some
jurisdictions, the BOD is submitted
with the architectural building site
permit. Typically, the document can
be updated and revised through the
design process, as appropriate, to
reflect the final design.
Engineers should review Appendix B of
PEER/TBI for additional information.
5.1 General
Describe project location, structural
system types used, and the most
important building considerations.
48 | Basis of Design Example
Describe the primary load path
if multiple systems are used, and,
if so, their intended purposes. If
higher occupancies require higher
performance levels by the governing
building code, specify.
Describe the site in terms of
geographic coordinates and include
a description of site considerations.
Describe the relationship of local and
national building code requirements
to this project.
Describe locations of anticipated
inelastic behavior and any enhanced
seismic devices such as buckling
restrained braces, isolation bearings,
dampers, etc.
Representative design drawings
should be included. This can be
satisfied by placing the BOD on a
drawing sheet as part of the set of
building structural drawings.
5.2 Superstructure
This section will describe the
superstructure, which includes the
ground floor and above. For sloped
sites, this would include all elements
above grade.
If the building is connected to
multiple buildings sharing a common
basement, describe how they are
interconnected (i.e., seismic joints,
common transfer diaphragm, etc.).
Include a schematic diagram showing
their overall configuration.
Lateral System
Describe the lateral system in more
detail, with typical dimensions
and material strengths. Identify
primary transfer diaphragms at- and
above-grade.
Gravity System
Describe the gravity system in more
detail, with typical dimensions and
material strengths. Describe if gravity
system components are intended to
resist seismic actions as part of the
seismic-force resisting system.
5.3 Substructure
Basement Levels
Describe basement levels in more
detail, with typical dimensions and
material strengths. For sloped sites,
describe how the site slopes. Provide
plans/sections that schematically
describe unique considerations.
5.4 Foundation System
Describe the foundation system for the
tower and podium including details,
dimensions, and material strengths.
Geotechnical Investigations and Reports
Reference geotechnical investigations
undertaken by the project geotechnical
engineer and provide a reference to
their report. Specify if site-specific
information is being used in the design
of the building.
5.5 Code Analysis and Design Criteria
Building Codes, Standards, Regulations
and Computer Software
Building Codes, Standards and
Regulations
List all codes progressing from local to
national. Also, list non-code sources of
information that are directly used in
the design. Examples would include
supporting publications of nonlinear
material/component behavior and their
acceptance criteria.
Computer Software
List all software, versions used, and
structural elements designed with them.
Code Exceptions
State the specific section(s) of
governing code that are excepted, if
any. It is best to specifically quote the
appropriate portions of code. Include a
brief justification for how the exception
is justified.
In jurisdictions where PBSD is
accepted, there are typically clear
design steps and criteria that are
expected to be followed. Describe
those requirements here and how they
are satisfied. If multi-step sequencing
is used, describe this sequence in
general and how each step leads to
the next.
Provide a description of the scope
of peer review. This may already be
specified by the governing jurisdiction if
PBSD is already adopted.
Performance Objectives
Provide a table that describes the
intended building performance. The
table may specify actions, as in some
cases a single element may have
different objectives for different actions,
such as shear walls. An example is
found in Table 3.3.
If performance of non-structural
components varies from the
governing building code, specify their
performance, including cladding,
partitions, elevators, exit stairs, etc.
Structural Gravity Load Criteria
Include a summarized version of gravity
loading criteria for typical floors and
conditions. This helps the document
stand alone. Not all gravity load criteria
need to be stated, but the exterior wall
should be included.
Structural Lateral Load Criteria
Provide a summary of seismic- and
wind-code-based load criteria, listing
all key code values. For wind, include
all key parameters, such as exposure
category and basic wind speed. For
seismic, include all parameters used
to calculate code base shear. Include
response modification factors such as
R and Ωo.
Seismic Loads Utilizing Site-Specific
Response Spectra and Ground Motions
Describe the site-specific seismic
information provided by the
geotechnical engineer and a brief
description of their methods. The
geotechnical engineer should help
develop this text, or text should
be adopted from the geotechnical
engineer’s report. Specify the level of
damping assumed in the spectra.
Specify the target spectrum for the
selection and development of ground
motions. If target spectra are other
than MCER, such as conditional mean
spectrum (CMS), are used, then describe
their development. Provide a plot of
spectral acceleration versus period
with code-based and seismic-specific
MCER, DE, and SLE design spectra. Also
provide a similar plot showing MCER
and developed CMS target spectra.
Examples of these two plots are found
in Figure 5.1. Black-and-white should
be used, as they may be placed on a
drawing sheet.
Structural Materials
Provide a list of all typical materials
used, and organize the list by concrete,
reinforcement, and structural steel.
State the grade, yield strength, and
appropriate ASTM designation for
reinforcement and steel. State the
typical application for each.
5.6 Structural Analysis and Design
The following sections provide
appropriate detail for each step in
the structural system design and
verification process. This information
should be conveyed in a logical,
sequential manner.
Typically, the initial design of the
seismic force-resisting system is
based on response spectrum analysis.
Then, subsequent verification is
conducted using NRHA. The specifics
of this process can vary based on
the requirements of the jurisdiction,
project-specific requirements, and
on-going advancements of the PBSD
design process.
Each step in the described process
should include common information
such as performance objectives, design
criteria, analysis model information
(including particulars of nonlinear
component modeling), and acceptance
criteria of global performance and
individual members. This common
format will help convey differences
between each step in the design and
verification process.
The initial member designs using
linear analyses are not always reviewed
by the peer review panel, but could
be very important to the governing
jurisdiction looking to ensure a basic
standard of care was utilized, similar to
code-based buildings. Thus, including
the entire design process in this
document is advantageous.
Step 1: Design Using Linear Analysis
The purpose of this step is to describe
the methods used for design of the
seismic force-resisting system. Often, a
service-level linear analysis is used with
adjusted acceptance criteria to ensure
appropriate performance under MCER
Basis of Design Example | 49
 1.6 Design Criteria
1.5 Site-Specific MCE If a code-based response spectrum
1.4 Site-Specific DE approach is used, then referring to
1.3 ASCE 7 MCE
ASCE 7 DE
code design criteria is sufficient. If
1.2
1.1 SL service-level or MCER demands are
1.0 considered, provide adjusted design
0.9 criteria for the considered level of
0.8 seismic demands. Provide appropriate
Sa (g)

0.7
references for these design criteria.
0.6
0.5
0.4 Specify load combinations used and
0.3 their appropriate references. Include
0.2 the portion of the live load considered.
0.1
0.0
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5 7.0 7.5 8.0 8.5 9.0 State if specified or expected material
Period (s) properties are utilized. This includes
strength and stiffness properties.

1.6
If response-modification parameters
1.5 Site-Specific MCE
1.4
such as R, Ωo, ρ, and Cd are used,
75% Site Specific MCE
1.3 state their value and how they
CMS - Short Period
1.2 are considered. For DE, the design
CMS - Long Period
1.1 adheres to code provisions, but
1.0
for SLE or MCER, linear analysis-
0.9
0.8
adjusted values may be used with
Sa (g)

0.7 substantiating references.

0.6
0.5 Analysis Model
0.4
Describe the analysis model
0.3
0.2 considerations. At a minimum, include
0.1 the following:
0.0
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5 7.0 7.5 8.0 8.5 9.0
 Analysis model software
Period (s)
implementation.
5Figure 5.1: Example of Target Spectra. © Skidmore, Owings & Merrill LLP
 State if basements or foundations
are included.

 State support conditions at walls

demands. Alternatively, DE or MCER
linear analysis could be used.
Describe the intent of this step
and expected performance level;
for example, “minimum life safety
seismic performance at the design-
level earthquake if DE demands
are considered.” Appropriate code
language should be considered.
Performance Objective
Describe the anticipated level of
damage for the considered level of
demands. For example, “the building
is anticipated to have moderate
structural damage when it is
subjected to design-level earthquake
shaking. This damage will require
extensive repairs.”
and columns.
 State if gravity and lateral systems
are explicitly modeled. If the
gravity system is not modeled,
describe provisions for actions
such as P-Delta.
 State if diaphragms are rigid or
semi-rigid.

50 | Basis of Design Example

  State the level of damping
assumed.
 State if accidental torsion is
included.
 State if P-Delta effects are
included.
 State if soil-structure integration is
considered.
 State if basement mass is
included.
 Provide table of all property
modifiers. State references for
selected property modifiers.
 Describe how non-traditional
seismic element such as BRBs,
isolation bearing, damping
devices, etc. are considered
in the analysis model. Provide
appropriate references.
Acceptance Criteria
State the acceptance criteria. If using
DE design, state that code provisions
are followed. If using SLE or MCER,
state adjustments to code revisions.
State how the strength reduction
factor, φ, is applied.
Verification Using Linear Analysis
(If Required)
The purpose of an additional step
using linear analysis is to address
requirements of the governing
jurisdiction, which may not be satisfied
by the traditional two-step approach
described by LATBSDC and PEER-TBI.
For example, if the building’s seismic
force-resisting system is designed
using a non-code methodology, then
the governing jurisdiction may require
a code-based verification (minus
stated exceptions). It is common to
use linear analysis with SLE or even
MCER demands for the initial seismic
force-resisting system design, but the
City of San Francisco, California, for
example, would require an additional
verification be done using code
provisions and DE demands.
If this step is required, follow the
provisions of Step 1, except as a form
of verification.
Step 2: Verification of MCER Using NRHA
The purpose of this step is to verify
that the seismic force-resisting system If a code-based
and its individual elements satisfy code
or specified enhanced performance response spectrum
targets using NRHA. approach is used,
Describe the intent of this step and then referring to
expected performance level; for
example, “adequate safety against code design criteria is
collapse at the MCER-level earthquake.” sufficient. If service-
Performance Objective level or MCER demands
Describe the anticipated level of
damage for the considered level of
are considered, provide
demands. For example, the seismic adjusted design
force resisting-system is anticipated
to provide a low probability of total criteria for the
or partial collapse in an MCER considered level of
earthquake event. Describe any
anticipated damage. seismic demands.
Design Criteria
Provide appropriate
State if any force- modification factors references for these
such as R, Ωo, ρ, and Cd are considered.
Typically, they are not for NRHA. design criteria.
Describe force and deformation-
controlled actions of elements
(see Table 3.3).
Specify load combinations used and
their appropriate references. Include
the portion of live load considered.
State if specified or expected material
properties are utilized. This includes
strength and stiffness properties.
Basis of Design Example | 51

Describe the analysis model
considerations. At a minimum, include
the following:
In many cases, a
bounding analysis is
warranted to ensure
appropriate force
levels are considered
for the verification
and design of transfer
diaphragms and
core walls. Potential
sources of variability
include the effect
of foundations and
assumed element
stiffness property
modifiers.
52 | Basis of Design Example
Analysis Model
 Analysis model software
implementation.
 State if basements or foundations
are included.
 State how transfer diaphragms are
modeled.
 State support conditions at walls
and columns.
 State if gravity and lateral systems
are explicitly modeled. If the
gravity system is not modeled,
describe provisions for actions
such as P-Delta.
 State if diaphragms are rigid or
semi-rigid.
 State the level of damping
assumed.
 State if accidental torsion is
included.
 State if P-Delta effects are
included.
 State if soil-structure integration is
considered.
 State if basement mass is included.
 Provide a table of all property
modifiers. State references for
selected property modifiers.
 Describe how frame, plastic, and
total rotations are obtained.
 Describe how strain
measurements are obtained.
 Describe how non-traditional
seismic element such as BRBs,
isolation bearing, damping
devices, etc. are considered
in the analysis model. Provide
appropriate references.
Component Modeling
Due to the number of methods
available to model nonlinear
components such as shear walls, link
beams, frames, etc. a specific section is
required for component modeling.
All modeled elements should be
included. This includes shear walls, link
beams, moment frames, and gravity
frames. This should include foundations
and non-traditional elements such as
BRBs, isolation bearing, and damping
devices as appropriate.
Shear Walls
Describe if fiber elements are used,
how they are specified (average or
actual reinforcement levels), and if
shear is modeled as linear or non-
linear. Describe the methodology used
for the development of stress-strain
relationships, including unconfined
concrete, confined concrete, and steel
fibers. Provide appropriate references.
Link Beams
Describe the method of modeling the
link beam, such as shear-hinges or
moment-hinges. If the analysis program
does not facilitate fixity of frame
elements to shell elements, describe
how moment fixity is accommodated.
Provide appropriate references.
Moment Frames
Describe the method of modeling
moment-frame beams and columns.
If the analysis program does not
facilitate fixity of frame elements to
shell elements, describe how moment
fixity is accommodated. Provide
appropriate references.
Slab-Frame Elements
When flat slabs are used, it
is common to consider their
influence on building performance
through an equivalent frame.
Describe how these elements are
developed based on the provided
slab reinforcement. Describe how
effective widths are accounted for.
If the analysis program does not
facilitate fixity of frame elements
to shell elements, describe how
moment fixity is accommodated.
Provide appropriate references.
average of the ground motion results.
If distinctions are made between
critical and non-critical elements,
clearly describe their acceptance
criteria. Include critical and non-critical
designations in the performance
objectives table as a separate column.
For deformation-controlled actions,
state their performance criteria, such
as strain limits in shear walls, plastic
rotation limits in link beams, and
plastic rotation limits of moment and
gravity frames.

Backstay Effect
In many cases, a bounding analysis 5.7 Appendices
is warranted to ensure appropriate
force levels are considered for the Include the following documents as
verification and design of transfer appendices:
diaphragms and core walls. Potential
sources of variability include the A. Geotechnical report including
effect of foundations and assumed site-specific hazard analysis
element-stiffness property modifiers. B. Appropriate research papers
To simplify the evaluation of these C. Wind tunnel report (if appropriate)
bounds, two studies are often
conducted. The intent of the first
study is to maximize the force in the
transfer diaphragms; the intent of
the second study is to maximize the
force in the shear-wall core.

Describe each bounding analysis

and provide appropriate references.
Specify how many ground motions
are run for each study and how
results will be considered in member
verification/design.

Acceptance Criteria
State acceptance criteria for global
metrics and force-controlled actions.
State how Φ is considered.

For global metrics, include drift and

residual drift limits of average and
individual records.

For force-controlled actions, state

the multiplier applied to the

Basis of Design Example | 53

 6.0 Performance-Based
Seismic Design Examples
Performance-Based Seismic Design Examples1
Design Example 1: Tall High-Occupancy Office Tower

This design example includes

documentation to demonstrate the
implementation of a performance-based
design approach. A performance-based
design approach should follow an
approved Basis of Design document
that establishes criteria for determining
performance acceptability at both a
service-level earthquake (SLE) and a
maximum considered earthquake (MCER).
The objectives of the performance in these
events are further described in the Basis
of Design, which is not provided with
this document. The examples included
are not intended to present complete
or comprehensive results. The example
presented is representative of the state of
practice at the time of this project’s design,
and therefore the acceptance criteria or
other metrics will vary from current practice
depending on when the project was
designed, project type, jurisdiction, design
guidelines referenced, and peer reviewers.
This example focuses on extracts from PBSD
reports that highlight specific element
types, show effective formats for presenting
results, and display acceptance criteria
evaluations that occurred for this design.

This design example is for a 525-foot

(160-meter), 36-story office tower (see
Figure 6.1). The design consists of a
reinforced concrete central core, with floor
framing consisting of steel floors with
composite steel decking. This example
highlights a building that is designed
as Risk Category III due to the number
occupants, and, therefore, with acceptance
criteria modified accordingly.
5Figure 6.1: Service-Level Analysis Model. © Magnusson Klemencic Associates

Mode Number Mode Description

SLE Model Period 6.1 Serviceability Event Analysis
(Seconds)
and Verification
1 X direction (translation) 4.12
2 Y-direction (translation) 2.45
Evaluation at the service-level earthquake
3 Torsion 2.30 (SLE) is required in order to quantify
5Table 6.1: Service-Level Periods of Vibration. response characteristics that relate to the
serviceability performance of the structure.

1
The following sections present design examples of real buildings designed using a performance-based seismic design approach. These designs generally follow the guidelines
described in this document. However, because the designs were completed for real buildings in various jurisdictions within the United States, there are some `differences in the
design processes specific to each building.

56 | Performance-Based Seismic Design Examples

These characteristics are items such as
story drift, coupling beam demands, and
shear wall demands. The acceptance
criteria for this serviceability-level event
are selected to encourage essentially
elastic behavior of the elements under
consideration.

6.2 Elastic Model Description

The analysis model used for the

service-level verification is a three-
dimensional, linear elastic, 2.5 percent
damped mathematical model. Walls
5Figure 6.2: Example Corner Points for Drift Determination. © Magnusson Klemencic Associates
and slabs are modeled as elastic shell
elements. Columns are modeled as Max SLE Seismic (Site Specific) Story Drift
frame elements. Torsion in the model is 525
525
limited to inherent torsion; accidental 500
500
475
torsion is not included. Cracked section 475
450
450
Y Direction
Y Direction
properties are included in the model 425
425 X Direction
X Direction
as described in the Basis of Design 400
400
375
Drift Limit
375
document. The analysis performed is Drift Limit

0.5% DRIFT ACCEPTANCE CRITERIA

350
350

a modal response spectrum analysis 325

325
300
300
procedure, including a minimum 275
275
Elevation
(ft)

mass participation of 90 percent of 250

250
(ft)
Elevation

225
225
the seismic mass, with the results 200
200
combined using the complete 175
175
150
150
quadratic combination (CQC) method. 125
125
The seismic mass includes the 100
100
75
75
building’s estimated self-weight, the 50
50
superimposed dead load, and any live 25
25
00
load required by ASCE 7 to be included, -25
-25
such as mechanical equipment and 0.0000
0.0000
-50
-50 0.0010
0.0010 0.0020
0.0020 0.0030
0.0030 0.0040
0.0040 0.0050
0.0050 0.0060
0.0060
Drift
Drift
a portion of storage loads. Mass is
5Figure 6.3: Max SLE Seismic (Site-Specific) Story Drift. © Magnusson Klemencic Associates
only assigned above the seismic base
(ground level in this case).

Acceptance Criteria: Story Drift
Story drift is a measure of the building
deformations under the SLE event. By
placing a limitation on building drift,
damage of nonstructural elements
(such as cladding, wall partitions, etc.)
can be limited. The three-dimensional
lateral analysis model includes the
stiffness modification parameters
identified above, which are consistent
with the anticipated behavior at SLE.
For the design example, the
acceptance criteria value for drift at
SLE is 0.5 percent. The full SLE response
spectrum is applied with no scaling
and no accidental torsion. Story drift
is calculated at each corner of the
building by taking the difference in
elastic displacement of adjacent floors
divided by the story height. Story drift
is calculated on a per-corner basis in
order to correctly include the effects
of inherent torsion and the rotational
response. An example of the corner
points considered is identified in Figure
6.2. Many software analysis tools have
the ability to directly output story drift.
The diagram in Figure 6.3 indicates the
maximum story drift recorded for all

Performance-Based Seismic Design Examples | 57

 Coupling Beam Design Method -Seismic
Coupling Beam Design Method - Seismic
four corners considered plotted over the
10
MF Beam building height. The story drift reported
9 Diag Beam
Deep
is substantially less than the acceptance
8 criteria; therefore, the story drift is
considered acceptable.
[Vu/sqrt(f'c)bd]

7
[Vu/sqrt(f'c)bd]

6
Acceptance Criteria: Coupling Beams
5 The concrete coupling beams in the
Stress

4
building are evaluated by comparing
Stress

demands from the SLE analysis to the

3
design strength associated with specified
2 material properties and code-specified
1
phi-factors. The acceptance criteria in
the Basis of Design for coupling beams
0
0.0 1.0 2.0 3.0 4.0 5.0 6.0 is limited to a demand-to-capacity ratio
Coupling Beam Demand
Span // Depth
Span Depth
vs Capacity (DCR) of 1.5 (implying that a small amount
of inelastic behavior is acceptable).
5Figure 6.4: Coupling Beam Stress versus Span/Depth Ratio. © Magnusson Klemencic Associates
These beams are either diagonally or
conventionally reinforced, depending
0.00 0.20 0.40 0.60 0.80 1.00 1.20
on the aspect ratio and shear stress, as
L37-PH
LEV 36
shown in Figure 6.4. Figure 6.5 shows the
LEV 35 DCR for a line of coupling beams in the
LEV 34 structure. Note that the DCRs are less than
LEV 33
the acceptance criteria of 1.5, indicating
LEV 32
LEV 31
acceptable performance. A similar check is
LEV 30 performed for all of the coupling beams in
LEV 29 the structure.
LEV 28
LEV 27
LEV 26
Acceptance Criteria: Shear-Wall Shear
LEV 25 The shear walls in the building are
LEV 24 evaluated by comparing shear demands
LEV 23 from the SLE analysis to the design
LEV 22
Level

strength associated with specified

LEV 21
LEV 20 material properties and code-specified
LEV 19 phi-factors. The acceptance criteria in the
LEV 18 Basis of Design for wall shear is limited to
LEV 17
a DCR of 0.7.
LEV 16
LEV 15
LEV 14
LEV 13
LEV 12
LEV 11
LEV 10
LEV 09
LEV 08
LEV 07
LEV 06
LEV 05
LEV 04
LEV 03
LEV 02
LEV 01
5Figure 6.6: Example Shear Pier Identification.
5Figure 6.5: Coupling Beam Demand-to-Capacity Ratios (SLE). © Magnusson Klemencic Associates © Magnusson Klemencic Associates

58 | Performance-Based Seismic Design Examples

 Service Level Shear Stress Ratios by Wall Pier
600
The core walls are divided into linear
shear piers based on the geometry of
the wall openings. See Figure 6.6 for an 500
example identification of these piers.

See Figure 6.7 for an example of a wall 400

shear stress plot and Figure 6.8 for
demand-to-capacity ratio confirmation.
300
Acceptance Criteria: Shear Wall Flexure
Heigt (ft)
Height(ft)
The shear walls in the building are
evaluated for flexure by comparing 200
the demands from the SLE analysis to
the design strength associated with
specified material properties and 100
code-specified phi-factors. Demands
are calculated by combining gravity and
lateral demands, according to the load 0
combinations in the Basis of Design. For 0.00 1.00 2.00 3.00 4.00 5.00 6.00 7.00 8.00

the example case, gravity demands are

calculated via tabulated load takedown, -100
and lateral demands are calculated via WallStress
Wall StressRatio
Ratio [Vu/(rt(f'c)bd)]
[Vu/(rt(f'c)bd]

pier-force output from the SLE analysis

model. By combining these demands, 5Figure 6.7: Shear Wall Stress Ratio (SLE). © Magnusson Klemencic Associates
each load combination will result in
set of axial-load-plus-biaxial-bending
moments (P-M-M).
Service Level Shear Demand to Capacity Ratios
600

Wall piers are grouped into flexural piers

(typically planar or L-shaped). Each load
500
combination is checked for flexural
limit states and is limited to a DCR of
1.5. It is common to batch-calculate this
400
DCR using software or via spreadsheets
(see Figure 6.9). The numbered points
represent a P-M-M load combination,
300
and the solid lines represent the code-
Heigt (ft)
Height(ft)

based capacity (DCR = 1.0). Therefore,

with the P-M-M load combinations
200
within the capacity curve, all demand-
to-capacity ratios are less than the
acceptance criteria of 1.5.
100

6.3 Nonlinear Response History

0
Analysis and Verification 0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90 1.00

A nonlinear response history analysis

-100
(NRHA) is used to evaluate the design Wall Wall
ShearStress Ratio [Vu/(rt(f'c)bd]Ratio
Demand-to-Capacity

5Figure 6.8: Shear Wall Demand-to-Capacity Ratio. © Magnusson Klemencic Associates

Performance-Based Seismic Design Examples | 59

 P (kip)
160000 shear behavior, diaphragm slabs, and
columns are expected to remain elastic
and are modeled with elastic properties.

The nonlinear model includes the core

walls from the mat foundation to the
roof. Non-rigid diaphragms composed
(Pmax) (Pmax) of manually-meshed shell elements
are used at and below the seismic
y base, as well as at upper levels where
x there are significant core stiffness
changes. At all other elevated slabs,
516 x 238 in rigid diaphragms are used. Stiffness
Code: ACI 318-08 assumptions are bounded through the
Units: English use of two different nonlinear models.
Run axis: Biaxial The “stiff” model uses upper-bound
Run option: Investigation stiffness assumptions on the basement
86
118
85
117
Slenderness: Not considered walls, and flexible diaphragms, and
84
83
116
115 6
38
5
37
Column type: Structural 134
102
101
133
lower-bound stiffness for the mat
Bars: ASTM A615 100
132
99
131 22
54 foundation. The “soft” model uses
Date: 05/09/17 lower-bound stiffness assumptions
Time: 10:16:28
-1200000
(Pmin) (Pmin)
1200000 on the basement walls, and flexible
M (84°) (k-ft) diaphragms and upper-bound stiffness
-20000
for the mat foundation.

STRUCTUREPOINT - spColumn v5.10 (TM). Licensed to: Magnusson Klemencic Associates. A “P-Delta column” is also modeled at
File: i:\rufusblock19\engineers\rpb\131210 core flexural\verif\38-m06.col
the center of the core, representative
Project: Rufus Block 19 of the remaining gravity columns.
Column: 17- M Lev 06 Engineer: rpb Gravity loads not tributary to the core
f'c = 8 ksi fy = 60 ksi Ag = 22488 in^2 242 #6 bars
(or the explicitly-modeled columns) are
Ec = 5098 ksi Es = 29000 ksi As = 106.48 in^2 rho = 0.47%
applied to this P-Delta column. Gravity
fc = 6.8 ksi e_yt = 0.00206897 in/in Xo = 237.00 in Ix = 6.76278e+007 in^4
e_u = 0.003 in/in Yo = 369.00 in Iy = 9.86518e+008 in^4
load tributary to the core is applied via
Beta1 = 0.65 Min clear spacing = 7.10 in Clear cover = N/A point loads throughout the core.
Confinement: Tied
phi(a) = 0.8, phi(b) = 0.9, phi(c) = 0.65
Ground motions are applied at the
level of the top of the mat foundation.
5Figure 6.9: Flexural Pier P-M-M Interaction Evaluation. © Magnusson Klemencic Associates The top of the mat is the assumed
elevation of support. For the “soft”
model, the nodes at all contact
points are fixed in translation (X, Y,
for the MCER event. Each acceptance 6.4 NRHA Model Description Z), and pinned for rotation (vertical
criterion is evaluated, and the design and horizontal translation supports).
is modified as necessary in order to The design example implements For the “stiff” model, the nodes are
meet the appropriate acceptance an NRHA verification model using assigned vertical spring stiffness
criteria. The following describes the CSI Perform-3D. The model includes values, which are developed through
analysis model and corresponding inelastic member properties for consultation between the geotechnical
acceptance criteria evaluations. elements that are anticipated to be and structural engineers. The lateral
loaded beyond their elastic limits. resistance and damping effects of the
These include the coupling beams and soil on the sides of the basement walls
core-wall flexural behavior. Core-wall are neglected.

60 | Performance-Based Seismic Design Examples

Inelastic vertical fiber elements are
used for the core walls at all levels
to capture the nonlinearity of the
moment and axial load interaction.
These elements are comprised of two
materials: reinforcing steel and concrete
(of varying confinement). Coupling
beams are modeled as nonlinear shear
links. Element stiffness, yield, and
degradation characteristics are matched
to experimental testing.

A detailed description of the model

is provided in the following sections,
with figures 6.10–6.14 illustrating and
describing important aspects of the
nonlinear model.
5Figure 6.10: NRHA Model 3D Isometric and Elevation View. © Magnusson Klemencic
Associates

Slab Extent (Not Modeled; Shown

for Graphical Purposes Only)
Coupling Beam Typical Level Modeled as a Rigid
Frame Elements Diaphragm
w/ Inelastic
Shear Hinges

Shear Wall Elements

with Concrete and Steel
Fibers with Inelastic
Material Properties

Extreme Points Slaved to

Diaphragms for Determination
of Story Drifts

5Figure 6.11: Typical Floor Wall and Coupling Beam Modeling. © Magnusson Klemencic 5Figure 6.12: NRHA Model – Typical Floor Slab Modeling. © Magnusson
Associates Klemencic Associates

Accounting for
ramp with slot in
diaphragm
Slab
Micro-Outriggers
Where Apprporiate

P-Delta Column
(at diaphragm
center of mass)

Outrigger Columns
Explicitly Modeled

5Figure 6.13: NRHA Model – Typical Floor Micro-Outriggering Modeling. 5Figure 6.14: Below-Grade Diaphragm Modeling of Ramp Discontinuities.
© Magnusson Klemencic Associates © Magnusson Klemencic Associates

Performance-Based Seismic Design Examples | 61

The design example
is a Risk Category III
Structure; therefore,
the typical rotation
limitations (0.06
for diagonally-
reinforced and 0.04
for conventionally- Perimeter Points
Around Slab
(for Computing Drift)

reinforced coupling
beams) are multiplied
by the ratio of code-
prescribed importance
factors (1.0/1.25).

5Figure 6.15: Perimeter Points of Computing Story Drift. © Magnusson Klemencic Associates
MCE Building Interstory Drift (Stiff Diaphragm Case): X Direction @ Core
(Max Story Drift from 7 Time Histories)
[Spectrally Matched Records]

Acceptance Criteria: Story Drift 550.0

550

Story drift is a measure of the building

deformations under the maximum 450.0
450
considered earthquake event. Drift, avg

Controlling drift is an important factor Chile3

in limiting the probability of collapse 350.0
350
El Salvador3
Height Above
Lower Lev 0G

Drift,avg
to within the intended performance
Diaphragm 250.0

Chile3
Height Above Tokachi-Oki3
ACCEPTANCE CRITERIA: 2.4%

parameters. The three-dimensional Height Above

Lower Lev Lev
0G0G
El Salvador3
(ft)

Lower Tokachi-Oki3
Diaphragm (ft) 250.0
nonlinear response history analysis Diaphragm
(ft)
250
Furukawa3
Furukawa3
Inage3
identified previously is used to Iran3Inage3
determine story drift from seven Northridge3
150.0

350.0

450.0

550.0

150
150.0 Iran3
-50.0

2.4% Max Drift

50.0

time-history records.
Northridge3
0.000

For the design example, the acceptance 50.0

50 2.4% Max Drift

criteria value for drift at MCER is 2.4

percent, taken as the average of 0.000
0.000 0.005
0.005 0.010
0.010 0.015
0.015 0.020
0.020 0.025
0.025 0.030
0.030
-50
-50.0
seven analyses. The design example Interstory Drift
Interstory Drift
5Figure 6.16: MCER Story Drift in X Direction. © Magnusson Klemencic Associates
MCE Building Int
0.005

62 | Performance-Based Seismic Design Examples

 MCE Building Interstory Drift (Stiff Diaphragm Case): Y Direction @ Core
(Max Story Drift from 7 Time Histories)
[Specrtrally Matched Records]
550.0
550

is a Risk Category III structure,

therefore the typical drift limitation 450.0
450 Drift, avg
for a Risk Category II structure (three Chile3
percent) is multiplied by the ratio of
El Salvador3
code-prescribed importance factors 350.0
350
Drift,avg

(1.0/1.25). The story drift is calculated Height Above

Chile3 Tokachi-Oki3

ACCEPTANCE CRITERIA: 2.4%

Height Above El Salvador3
at each corner of the building (see Lower LevLev
Lower 0G0G Furukawa3
Tokachi-Oki3
Diaphragm (ft) 250.0
Diaphragm 250
Figure 6.15) by taking the difference in (ft)
Furukawa3
Inage3 Inage3
displacement at the floor in question Iran3
Iran3
Northridge3
from the displacement of the floor 150.0
150 2.4% Max Drift
Northridge3
below, divided by the story height. The
maximum story drift for each of the 2.4% Max Drift
seven analyses is reported, and then 50.0
50

the average of these seven analyses is

0.000
0.000 0.005
0.005 0.010
0.010 0.015
0.015 0.020
0.020 0.025
0.025 0.030
compared to the acceptance criteria. -50.0
-50
See Figure 6.16 for x-direction results Interstory Drift
Interstory Drift

and Figure 6.17 for y-direction results. 5Figure 6.17: MCER Story Drift in Y Direction. ©Residual
Magnusson
StoryKlemencic
Drift Associates
(X Direction)
The story drift reported is less than the 600
600
acceptance criteria, therefore the story EQ 1: Chile3
drift at MCER is considered acceptable. 500
500
EQ 2: El Salvador3

Residual drift is also evaluated for

ACCEPTANCE CRITERIA (SINGLE EVENT): 1.5%

EQ 3: Tokachi-Oki3

ACCEPTANCE CRITERIA (AVERAGE): 1.0%

400
400
acceptance. Residual drift is a measure EQ 1: Chile3 EQ 4: Furukawa3
of the permanent, inelastic deformation EQ 2: El Salvador3

300
300 EQ 5: Inage3
EQ 3: Tokachi-Oki3
following the maximum considered Height
Height EQ 4: Furukawa3
AboveAbove
earthquake event. For the design L01 L01
EQ 5: Inage3
EQ 6: Iran3
EQ 6: Iran3

example, the acceptance criteria value 200

200 EQ 7: Northridge3
Avg
EQ 7: Northridge3
for drift at MCER is one percent taken as Limit for Avg
Avg
Limit for Max of Any Single Record
the average of seven analyses and 1.5 100
100

percent from any single analysis. See Limit for Avg

Figures 6.18 and 6.19 for verification of 0 Limit for Max

0.0000
0.0000 0.0020
0.0020 0.0040
0.0040 0.0060
0.0060 0.0080
0.0080 0.0100
0.0100 0.0120
0.0120 0.0140
0.0140 0.0160
0.0160 of Any Single
residual drift for acceptance. Record
-100
-100
Story DriftStory Drift
Acceptance Criteria: Amazon - Rufus 2.0 Block 14
5Figure 6.18: Residual Story Drifts (X-Direction).Residual
© Magnusson Klemencic Associates
Story Drift
Coupling Beam Rotations (Y Direction)
For the design example, the 600
600

EQ 1: Chile3
acceptance criteria for coupling
beam rotations are 0.048 radian 500
500 EQ 2: El Salvador3

for diagonally-reinforced coupling EQ 3: Tokachi-Oki3

ACCEPTANCE CRITERIA (SINGLE EVENT): 1.5%

beams and 0.032 radian for 400

400
ACCEPTANCE CRITERIA (AVERAGE): 1.0%

EQ 1: Chile3 EQ 4: Furukawa3
conventionally-reinforced beams, EQ 2: El Salvador3
EQ 3: Tokachi-Oki3
taken as the average of seven Height
Height
300
300
EQ 4: Furukawa3
EQ 5: Inage3

analyses. The design example is a Risk Above

Above
L01
EQ 5: Inage3
EQ 6: Iran3 EQ 6: Iran3
L01
Category III Structure; therefore, the 200
200 EQ 7: Northridge3
Avg EQ 7: Northridge3
typical rotation limitations (0.06 for Limit for Avg
Limit for Max of Any Single Record
100 Avg
diagonally-reinforced and 0.04 for 100

conventionally-reinforced coupling Limit for Avg

00
beams) are multiplied by the ratio 0.0000
0.0000 0.0020
0.0020 0.0040
0.0040 0.0060
0.0060 0.0080
0.0080 0.0100
0.0100 0.0120
0.0120 0.0140
0.0140 0.0160
0.0160 Limit for Max
of Any Single
of code-prescribed importance factors Record
-100
-100
(1.0/1.25). StoryStory
Drift
Drift

5Figure 6.19: Residual Story Drifts (Y-Direction). © Magnusson Klemencic Associates

Performance-Based Seismic Design Examples | 63

 Link Beam Chord Rotations (Avg of 7 Records)
[Spectrally Matched Records, Stiff Diaphragms]
550.0
550
East Wall (Furthest South)

East Wall (2nd from South)

450.0
450

East Wall (2nd from North)

East Wall (Furthest North)

ACCEPTANCE CRITERIA (AVERAGE): 4.8%

350.0
350
East Wall (Furthest South)
Figures 6.20 and 6.21 are graphs West Wall (Furthest South)
East Wall (2nd from South)
Height Above
presenting the coupling beam Height
Lower
Lower
Above
Lev 0101
Level
East Wall (2nd from North)
East Wall (Furthest North)
Diaphragm 250.0250
rotations in each bay of the building. Diaphragm
(ft) (ft)
West Wall (2nd from South)
West Wall (Furthest South)
West Wall (2nd from South)

The chord rotation of the coupling West Wall (2nd from North)
West Wall (2nd from North)
West Wall (Furthest North)
beams is calculated by dividing the 150
150.0 6% Max Rotation
West Wall (Furthest North)
peak hinge displacement by the length
of the coupling beam. The chord 6% Max Rotation
rotations are then compared to the 50.0
50

rotation acceptance criteria. See Figure

0.000
0.0000 0.010
0.010 0.020
0.020 0.030
0.030 0.040
0.040 0.050
0.050 0.060
0.060 0.070
0.070
6.20 for the plot of average coupling -50
-50.0
Link Beam Chord
Link Beam ChordRotation (rad)
Rotation (rad)
beam rotations of all diagonally-
Dispersion
5Figure 6.20: Link-Beam Chord Rotations (Average of
of Link Beam ©
7 Records). Chord Rotations
Magnusson Klemencic Associates
reinforced coupling beams and Figure East Wall (Furthest South Beam)
6.21 for a dispersion of MCER responses [Spectrally Matched Records, Stiff Diaphragms]
from one sample coupling beam. 550.0
550

Acceptance Criteria: Wall Strains

(Compression/Tension) 450.0
450
East Wall, Furthest South,
This section presents tension and MCER, Average
compression strains in the flexural East Wall, Furthest South,
reinforcing at certain locations 350.0
350 MCER, Case 1 (Chile3)

ACCEPTANCE CRITERIA (AVERAGE): 4.8%

East Wall, Furthest South, MCE, Average

along the height of the tower. The East Wall,

East Wall, Furthest South,Furthest
MCE, Case1South,
(Chile3)
MCE , Case 2 (El Salvador3)
Height
HeightAbove
Above R South, MCE, Case2 (El Salvador3)
longitudinal reinforcement strains are Lower
LowerLevel 01
Lev 01
East Wall, Furthest
East Wall,
East Wall, Furthest South,Furthest
MCE, Case3South,
(Tokachi-Oki3)
Diaphragm (ft) 250.0
determined as the average of seven Diaphragm
(ft)
250
MCER,South,
East Wall, Furthest CaseMCE,
3 (Tokachi-Oki3)
Case4 (Furukawa3)

analyses. The core-wall concrete axial East Wall, Furthest South, MCE, Case5 (Inage3)
East Wall, Furthest South,
East Wall, Furthest South, MCE, Case6 (Iran3)
MCER, Case 4 (Furukawa3)
strains are determined as 1.5 times the East Wall, Furthest South, MCE, Case7 (Northridge3)
150.0
150 East Wall, Furthest South,
average of the seven analyses. 6% Max Rotation
Case 5 (Inage3)
East Wall, Furthest South,
MCER, Case 6 (Iran3)
For the design example, the
50
50.0
acceptance criterion for tensile strain East Wall, Furthest South,
MCER, Case 7 (Northridge3)
of the longitudinal reinforcement is
0.000
0.000 0.010
0.010 0.020 0.030
0.030 0.040
0.040 0.050
0.050 0.060
0.060 0.070
0.070 6% Max Rotation
0.04, taken as the average of seven -50.0
-50
analyses. The design example is a Link
Link BeamChord
Beam Chord Rotation
Rotation (rad)
(rad)

Risk Category III Structure; therefore, 5Figure 6.21: Link-Beam Chord Rotations (Dispersions of Sample
Wall Tension Strains Beam). © Magnusson Klemencic Associates
[Spectrally Matched Runs)
the typical tensile rebar strain 600
600
Tensile Strain (x 10^-3):
(0.05) is multiplied by the ratio of Chile3, Pt 1
code-prescribed importance factors 500
500
Tensile Strain (x 10^-3):
Chile3, Pt 4
(1.0/1.25). A strain of 0.04 corresponds KEY PLAN:
Tensile Strain (x 10^-3):
to 16.5 times the expected yield strain. Chile3, Pt 8
400
400
The tension strain is determined from Tensile Strain (x 10^-3):
Tensile Strain (x 10^-3): Chile3, Pt 1

strain gages at each meshed point of Chile3, Pt 12

Tensile Strain (x 10^-3): Chile3, Pt 4
Tensile Strain (x 10^-3): Chile3, Pt 8

the core walls. The maximum strain of Height Height 300

300 Tensile Strain (x 10^-3):
Tensile Strain (x 10^-3): Chile3, Pt 12
Chile3, Pt 16
Above Above Tensile Strain (x 10^-3): Chile3, Pt 16
each strain gage is recorded for each Lev 0G
Lower Lev (ft)
Tensile Strain (x 10^-3): Chile3, Pt 19
Tensile Strain (x 10^-3):
MCER case and the average of seven 0G (ft) 200
200
Tensile Strain (x 10^-3): Chile3, Pt 22
Chile3, Pt 19
Tensile Strain (x 10^-3): Chile3, Pt 26

cases is reported. A plot of this average Tensile Strain (x 10^-3):

Tensile Strain (x 10^-3): Chile3, Pt 30
Chile3, Pt 22
Tensile Strain (x 10^-3): Chile3, Pt 33

tensile strain at each corner point of 100

100
ey
Tensile Strain (x 10^-3):
the core is provided with maximum Chile3, Pt 26

strains well below the 0.04 limit. The 0

Tensile Strain (x 10^-3):
Chile3, Pt 30
dispersion of MCER responses from 0.0
0.0 1.0
1.0 2.0 3.0
3.0 4.0
4.0 5.0
5.0 6.0 7.0
7.0 8.0 9.0
9.0 10.0
10.0
Tensile Strain (x 10^-3):
one sample gage (point #22) is also Chile3, Pt 33
-100
-100
included. See Figures 6.22 and 6.23 for Strain (x10^-3)
Strain (x 10^-3) ey

5Figure 6.22: Wall Tensile Strains; Example of Single Record for Entire Wall. © Magnusson Klemencic Associates

64 | Performance-Based Seismic Design Examples

 Dispersion of Wall Tension Strains
[Spectrally Matched Runs)
600
600
PT #22, Record #1:
Chile3
500
500 PT #22, Record #2: El
KEY PLAN: Salvador3

PT #22, Record #3:

400
400 Tokachi-Oki3

PT #22, Record #4:

PT #22, RECORD #1: Chile3
Furukawa3
PT #22, RECORD #2: El Salvador3
HeightHeight 300
300
PT #22, RECORD #3: Tokachi-Oki3
AboveAbove PT #22, Record #5:
PT #22, RECORD #4: Furukawa3
Lev 0G
Lower Lev Inage3
PT #22, RECORD #5: Inage3
(ft)
0G (ft) 200
200
PT #22, RECORD #6: Iran3
PT #22, Record #6:
PT #22, RECORD #7: Northridge3

For the design

Iran3
Average
ey PT #22, Record #7:
Northridge3
example, the
100
100

Average

00
0.0
0.0 1.0
1.0 2.0
2.0 3.0
3.0 4.0 5.0 6.0 7.0
7.0 8.0
8.0 9.0
9.0 10.0
10.0
ey
acceptance criterion
-100
-100
for tensile strain
of the longitudinal
Strain
Strain(x10^-3)
(x 10^-3)

5Figure 6.23: Wall Tensile Strains; Dispersion of Sample Point #22. © Magnusson Klemencic Associates

reinforcement is 0.04,
Wall Compressive Strains
[Spectrally Matched Runs)
600 Compressive Strain

KEY PLAN:
(x 10^-3): Northridge3, Pt 1

Compressive Strain
taken as the average
500 (x 10^-3): Northridge3, Pt 4
of seven analyses.
Compressive Strain
(x 10^-3): Northridge3, Pt 8
400
Compressive Strain
(x 10^-3): Northridge3, Pt 12
Compressive Strain (x 10^-3): Northridge3, Pt 1
Compressive Strain (x 10^-3): Northridge3, Pt 4

300 Compressive Strain

Compressive Strain (x 10^-3): Northridge3, Pt 8
HeightHeight
(x 10^-3): Northridge3, Pt 16
Compressive Strain (x 10^-3): Northridge3, Pt 12
AboveAbove
Lev 0G Compressive Strain (x 10^-3): Northridge3, Pt 16
Lower Lev
(ft) Compressive Strain
Compressive Strain (x 10^-3): Northridge3, Pt 19
0G (ft) (x 10^-3): Northridge3, Pt 19
200 Compressive Strain (x 10^-3): Northridge3, Pt 22
Compressive Strain (x 10^-3): Northridge3, Pt 26
tensile stra-+in plots.
Compressive Strain
Compressive Strain (x 10^-3): Northridge3, Pt 30
(x 10^-3): Northridge3, Pt 22
100
Compressive Strain (x 10^-3): Northridge3, Pt 33
For the design example, the acceptance
Compressive Strain
(x 10^-3): Northridge3, Pt 26 criterion for concrete axial strain of
00
Compressive Strain the core wall is 0.012, where the strain
(x 10^-3): Northridge3, Pt 30
0.0
0.0 0.3
0.3 0.6
0.6 0.9 1.2 1.5
1.5 1.8 2.1 2.4 2.7
2.7 3.0 3.3 is taken as 1.5 times the average of
Compressive Strain
(x 10^-3): Northridge3, Pt 33
seven analyses. The design example
-100
StrainStrain
(x10^-3)
(x 10^-3) is a Risk Category III Structure;
5Figure 6.24: Wall Compressive Strains; Example of Single Record for Entire Wall. © Magnusson Klemencic Associates therefore, the typical fully-confined
Dispersion of Wall Compressive Strains concrete compression strain limit
[Spectrally Matched Runs)
600
600
(0.015) is multiplied by the ratio of
PT #33, Record #1:
Chile3 code-prescribed importance factors
KEY PLAN:
(1.0/1.25). The compressive strain is
500
500 PT #33, Record #2:
El Salvador3 determined from strain gages at each
meshed point of the core walls. The
PT #33, Record #3:
400
400 Tokachi-Oki3 maximum strain of each strain gage is
PT #33, Record #4:
recorded for each MCER case and the
PT #33, RECORD #1: Chile3

Height 300
300
Furukawa3
PT #33, RECORD #2: El Salvador3 average of seven cases is reported. A
Height
AboveAbove PT #33, RECORD #3: Tokachi-Oki3
PT #33, Record #5:
plot of this average compressive strain
LowerLevLev0G PT #33, RECORD #4: Furukawa3

0G (ft)
(ft) Inage3
PT #33, RECORD #5: Inage3 (with a 1.5 amplification factor) at each
200
200 PT #33, RECORD #6: Iran3

PT #33, Record #6:

PT #33, RECORD #7: Northridge3 corner point of the core is provided in
Average Iran3 Figure 6.24, with a maximum strain for
100
100
PT #33, Record #7: Pt 12 corresponding to approximately
Northridge3
0.0013, well below the 0.012 limit. The
00 Average dispersion of MCER responses from one
0.0
0.0 0.3
0.3 0.6
0.6 0.9 1.2 1.5
1.5 1.8
1.8 2.1 2.4 2.7 3.0
3.0 3.3
3.3
sample gage (point #33) is also included
-100
-100
(see Figure 6.25).
StrainStrain (x 10^-3)
(x10^-3)

5Figure 6.25: Wall Compressive Strains, Dispersion of Sample Point #33. © Magnusson Klemencic Associates

Performance-Based Seismic Design Examples | 65

Acceptance Criteria: Wall Confinement noting the strain limits associated with Item Axial Strain Limit
This section presents the verification of the provided level of confinement.
Unconfined Concrete 0.003
core-wall confinement design based on
Partially Confined Concrete
core-wall compressive strain results. Partial confinement of the core wall (#5 @ 8” vert, 16” horz)
0.004
was provided at select upper levels of 5Table 6.1: Axial Strain Limits
For this design example, the strain the building due to core wall setbacks,
limits (see Table 6.1) for unconfined where strain concentrations are Level Level of Confinement
concrete and partially confined typically expected to occur under high
Level 32-Top Unconfined
concrete were selected based on seismic demands.
Partial Confinement (#5
the Razvi confinement model and Level 29-32
@ 8” vert/16” horz)
recommendations from ATC-72. A strain Table 6.2 is a summary of confinement Level 21-29 Unconfined
limit associated with 80 percent of the for this design example. See Figures
Partial Confinement (#5
peak stress on the descending branch 6.27 and 6.28 for additional wall Level 18-21
@ 8” vert/16” horz)
was selected. See Figure 6.26 of the confinement information. Level 13-18 Unconfined
stress-strain curves for various concrete
Partial Confinement (#5
types for the determination of the Acceptance Criteria: Wall Shear Level 9-13
@ 8” vert/16” horz)
strain limits. This section presents the design of core Full Confinement
Level P6 –9
walls for wall shear. (#5 @ 4” vert/ 8” horz)
The dispersion of compressive strains at *Note: All confinement ties are Fyt=80ksi
a sample location (point #1) is provided, 5Table 6.2: Levels of Confinement



  
  

  
  
   

  
  

  
  








 





UNCONFINED PARTIALLY CONFINED 
STRAIN LIMIT STRAIN LIMIT 
 


 
        
Figure #550: Razvi Confinement Model (Strain Limits) 

Structural Calculations Figure 78 ; Razvi Confinement Model

NLRHA Analysis and- Strain Limits
Verification
Page 66 of 90
Amazon – Rufus 2.0 Block 19, Seattle, WA
5Figure 6.26: Sample Concrete Stress-Strain Curves (Razvi model). © Magnusson Klemencic Associates

66 | Performance-Based Seismic Design Examples

5Figure 6.27: Confinement Summary. © Magnusson Klemencic Associates
For the design example, the core
walls throughout the entire height
of the building are designed for
the mean shear demand from the
seven earthquakes, multiplied by an
amplification factor of 1.5. The shear
design meets the criteria of ACI 318-08,
Section 21.7.4.1, where the capacity (ф
Vn) is defined as follows:
Where:
Acv = net area of concrete section
f'c = nominal concrete strength
f’c,exp = 1.3 x f'c = expected concrete
strength
Рn = ratio of distributed shear
reinforcement
fy,exp = expected strength of
reinforcement
ф = 1.0 = strength reduction factor [in
accordance with the Basis of Design]
See Figure 6.29 for sample graphical
results for the shear-wall horizontal
forces in one individual pier.
Performance-Based Seismic Design Examples | 67
Transfer Diaphragm Design
This section presents the design of a
diaphragm that is transferring lateral
forces from one element to another.
This commonly occurs at the top of a
podium on a high-rise building or at
5Figure 6.28: Dispersion of Compressive Strains with Confinement Limits. © Magnusson Klemencic Associates
5Figure 6.29: Wall Shear Design Example (MCER). © Magnusson Klemencic Associates
68 | Performance-Based Seismic Design Examples
grade level where foundation walls
are engaged.
Design forces are based on the transfer
force output from the nonlinear
analysis model. Section cuts are used
φ φ
φ φ
to determine average forces from the
time-history records. In Figure 6.30, a
simple beam model is used to describe
the force transfer from tower walls to
basement walls. In this case, the tower
walls are the point of load application,
and the basement walls support the
simple beam reactions. The simple
beam model can then be evaluated for
diaphragm shear and flexure.
Foundation Design
In this example, a mat foundation
founded on suitable bearing material
is utilized to support a tower core.
Gravity and mean MCER overturning
demands are determined from the
nonlinear time-history analysis and
applied to a separate foundation
analysis model in order to complete
the design of the foundation
components. Overturning demand
over the building height is shown in
Figure 6.31, and an example of MCER
overturning applied to a foundation
analysis model is shown in Figure 6.32.
Depending on the component action
being designed, the demands from
the time-history analysis are modified
with scale factors. For a component
action that is ductile in the event of
overload, but not intended to yield in
the MCER event, such as flexure of a
mat foundation, it may be scaled above
the average MCER demands. In this
case, a 1.2 scale factor was selected to
provide the desired yielding hierarchy.
For a component action that is brittle,
a higher scale factor may be warranted
in order to protect against brittle failure.
Two-way and one-way shear are scaled
by 1.5, consistent with a critical force-
controlled behavior.


For a component
action that is ductile in
the event of overload
but not intended
to yield in the MCER
event, such as flexure
of a mat foundation,
it may be scaled
above the average
MCER demands.



5Figure 6.30: Simple Beam Model for Transfer Diaphragm. © Magnusson Klemencic Associates


Core OTM Comparison for Mat Design about X Axis (k-ft)
600 Y

1.6 x Wind - Y Direction

500 SLE - Y Direction
1.5 x MCE Average (Shear Design)
Mxx
1.2 x MCE Average (Flexural Design)
X X
400

300
Height (ft)

200 Y

100

MAT DESIGN
VALUES
0

 0.0 500000.0 1000000.0 1500000.0 2000000.0 2500000.0 3000000.0 3500000.0 4000000.0

4500000.0 5000000.0

-100
Overturning Moment (k-ft) 

5Figure 6.31: Foundation Overturning Moment Determination. © Magnusson Klemencic Associates 5Figure 6.32: Example of MCER Overturning Applied
Page
to Foundation Analysis Model. 2
© Magnusson Klemencic
Associates

Performance-Based Seismic Design Examples | 69

Performance-Based Seismic Design Examples
Design Example 2: Supertall Mixed-Use Tower
5Figure 6.33: Isometric 3D Model with Outriggers at Three Levels. © AC Martin
6.5 Project Overview
PBSD modeling approaches,
acceptance criteria, and results
summaries can be used for steel-framed
structures, composite structures,
and mixed-material structures. More
economical and appropriate designs
and details can be based on the
better understanding of member
demands from PBSD. For consistency
with the previous example, another
high-rise building is used to illustrate
this point. Its structural framing
incorporates both composite elements
(concrete-filled steel box columns)
and mixed construction. The tower
has long, narrow floors and a slender
70 | Performance-Based Seismic Design Examples
central concrete core to suit its primary
function as a hotel. Perimeter columns
are composite, rectangular steel box
columns with high-strength concrete
fill to minimize impact on hotel room
windows. Three sets of outrigger braces
provide occupant comfort during
windy conditions and safety during
earthquakes (see Figure 6.33). A rooftop
“sail” feature and a tapered spire (not
shown) are other major steel elements.
PBSD using NRHA through the analysis
program Perform-3D was selected
to show acceptable performance
of the proposed structural system.
Prescriptive seismic code provisions
do not include core-and-outrigger
systems, and prescriptive seismic code
provisions for buildings of this height
would require a dual system (special
core and special moment frames) that
would not satisfy building performance
criteria. Meeting the provisions of the
Los Angeles Tall Buildings Structural
Design Council (LATBSDC) PBSD
guidelines was considered sufficient to
demonstrate acceptability to the Los
Angeles building department and its
peer review panel. Buckling-restrained
brace (BRB) diagonals at all outrigger
levels provide stable cyclic nonlinear
behavior and limit forces generated at
columns, connections, and core walls.
At the lowest set of outriggers, each
diagonal is composed of four individual
BRBs extending through three stories.
The group of four BRBs provides
exceptionally large capacity while
using individual elements in sizes that
have already been tested. The middle
outriggers have an unusual ”X-braced
Vierendeel” configuration, discussed
further in this section. The top three-
story-tall outriggers are pre-loaded by
jacks to address long-term differential
shortening between the concrete core
and concrete-filled steel perimeter box
columns. The outrigger connection
details are complex in order to handle
large forces and deformations, but
were developed with contractor input
to enable practical construction (see
Figures 6.34 and 6.35).
The middle outriggers occur roughly
2/3 of the way up the tower and extend
vertically through six stories, from level
53 to level 59. These outriggers form
“X-braced Vierendeel trusses,” which are
concealed within the walls separating
hotel rooms. To keep the hotel corridor
clear, a three-foot (0.9-meter) deep
steel girder extends from the outrigger
braces to a pinned connection at the
concrete core wall at each floor level,
shown in Figure 6.35. Steel posts share
forces among girders, but some posts
are omitted to minimize the restraint
 59
W.P. 938' - 6"

BOX GIRDER 24"MIN.x16"

W33X318
BR
B W.P.

W8X67
W.P.

W14X82
14 "
"x 10
10 "x
" 14 2
B
BR S-009.19.1 TYP.
58
927' - 0"

W33X318
BR 2
B

W14X82
"
10 14
"x
S-009.19
"x W.P.
14 10
"
B
W.P. BR W.P.
57
915' - 6"

W33X387 W.P. W.P.

"
10

W8X67
W.P. "x
BR 14
B
B
14 BR
"x
10
" 56
904' - 0"
2
W33X318 S-009.19
0" BR
W.P. "x 1

W14X82
B
14 W.P. 14
B "x
BR 10
"
W.P.
55
892' - 6"

W33X387 W.P.

W8X67
W.P.

W14X82
BR " 3
B 10 TYP.
14 "x S-009.19 U.N.O.
"x 14
10 B
BR
W.P. " 54
881' - 0"

9 W33X318 "
10 BR 4
"x
14

W14X82
B S-009.19
B 14

W8X67
BR W.P.
"x
10
"
W.P.
53
869' - 6"

BOX GIRDER 24"MIN.x16" W.P.

5Figure 6.34: Bottom Outrigger Extending Through Three Stories. © Thornton Tomasetti 5Figure 6.35: Middle Outrigger Using Deep Girders to Cross Corridors.
© Brandow & Johnston

of long-term shortening on core-wall
concrete within the outrigger height.
Steel belt trusses at the bottom and
top outrigger levels, visible in Figure
6.36, link the 10 outrigger columns
to all perimeter tower columns. By
engaging all 20 perimeter columns,
the stiffness of the lateral load-resisting
system is maximized, and differential
vertical movements between columns
are minimized. Belt trusses also act as
“virtual (or indirect) outriggers,” reducing
tower deflections in the long direction.
The load path to accomplish this is
described later.
General Steps in the PBSD Process
Step one in the design process
was sizing core walls, columns, and
outriggers for strength-level forces
based on a 1,700-year mean recurrence
interval (MRI), ASCE 7-10 wind
loads based on the building’s high
occupancy, and checking dynamic
properties for occupant comfort using
10-year MRI wind informed by local
wind climate data.
Step two was estimating structural-
component seismic demands using
linear elastic response-spectrum
analysis (RSA) results, scaled up to
provide base shear of at least 85
percent of the minimum base shear
equations in the prescriptive code. In
this case, equivalent lateral force for
the design earthquake level, as used
for conventional analysis and design,
governed over spectrum-based
shear. This was done by considering a
building period of 7.0 seconds in the
transverse (north–south) direction
and 3.5 seconds in the longitudinal
(east–west) direction, and local high
seismicity. Note that in regions of
low seismicity, the minimum base
shear equations may not govern over
the spectrum-based shear value. For
overturning, scaled RSA results were
used in preliminary checks. For shears,
scaled RSA results were tripled because
higher modes are major sources of
story shear (but not overturning) in tall
flexible buildings. That behavior is not
addressed well by code methods that
directly relate overturning and shear
as appropriate for more common,
shorter buildings, and because a
1.5 factor is applied to mean results
for force-controlled shear checks.
Subsequent NRHA results confirmed
the reasonableness of this approach.

Performance-Based Seismic Design Examples | 71

5Figure 6.36: Top Outriggers and Belt Trusses in the 70th-floor Hotel Lobby. © AC Martin
Composite columns Figure 6.37 shows that MCER NRHA
base shears in both directions (the
were studied in right two bars) are more than double
parallel runs, one the code (85 percent of minimum
base shear formula) values (the left
using transformed two bars). This is before applying a 1.5
areas and the factor for shear checks. It also shows
a large reduction in EW base shear
other using fiber for the longitudinal, non-outrigger
modeling. Results direction (blue) when comparing the
middle elastic response spectrum
showed only minor cases to the right, NRHA case. This
reduction is due to yielding of the
differences, so the core wall coupling beams, longitudinal
simpler transformed direction. In contrast, north–south
base shear for the transverse, outrigger
area approach direction (red) doesn’t decrease
was retained. substantially when going from elastic
to nonlinear modeling, since outrigger
BRB yielding doesn’t affect higher
modes that generate shear in the solid
“web” walls of the core. Figure 6.38
shows MCER NRHA overturning in each
direction similar to code. Yielding of
coupling beams (blue) and BRBs (red)
provides large reductions compared
to their elastic MCER response-
spectrum values.
72 | Performance-Based Seismic Design Examples
Step three was checking service-level
demands. As expected, service-level
base shear and base moment demands
are well below code and MCER
demands. Results of the elastic response
spectrum runs showed demands well
below yield throughout the structure,
meeting service-level acceptance
requirements. This was important to
confirm that the service level did not
control design for this project, which
can happen under some conditions.
Step four was developing the NRHA
model, with 11 spectrally-matched
time histories used for demands.
When modeling the core wall, each
coupling beam was a shear hinge
rather than a pair of moment hinges
because, for short-span deep beams,
this provides comparable results with
less computational effort, and wall piers
used distributed concrete and rebar
fibers. BRB properties were initially
estimated and later updated to reflect
testing and analytical data once the BRB
vendor was selected. Steel members
were modeled elastically unless yielding
 TAT5 (Fiber Modeling)

TAT5 (Linear Modeling)

5Figure 6.37: Base Shear Comparisons for ELF-based Design Earthquake Minimum Shear Formulas,
Elastic Response Spectra at Two Damping Values for Service and MCER Events, and MCER NRHA.
© Thornton Tomasetti

5Figure 6.38: Building Overturning Calculated Using Code Equations, RS, and NRHA. 5Figure 6.39: Fiber vs. Linear D/C Results. © Thornton Tomasetti
© Thornton Tomasetti

was anticipated or found. Composite
columns were studied in parallel runs,
one using transformed areas and the
other using fiber modeling. Results
showed only minor differences (see
Figure 6.39), so the simpler transformed-
area approach was retained.
Step five was developing multiple
NRHA models of the tall tower to study
and bracket the effects of a deep and
stiff “basement box” and an adjacent
low-rise podium that shares the same
basement. Three tower models were
used. AC0 was a “stand-alone” tower
model fixed at its base, which ignores
any help from the “basement box.”
AC1 added “dummy frames” in both
directions to simulate stiffness of the
surrounding basement box. AC2, with
“dummy frames” and a rotational spring
at the tower base as determined from
the dynamic sub-grade modulus,
resulted in the largest predicted transfer
forces between the tower and the
basement-box diaphragm floors. It
was found that the results above the
basement were very similar for all three
tower models, so AC0 results were
used for reporting tower performance
results to reviewers. The podium and
its basement levels were designed for
prescriptive seismic code demands,
plus basement seismic forces resulting
from interaction with the tower, rather
than using a PBSD approach.

Performance-Based Seismic Design Examples | 73

 Floor ID
5Figure 6.40: East – West Mean MCER Story Drifts. © Thornton Tomasetti
6.6 Presentation of Core-Wall Element
Demands and Acceptance
For the results of the core wall elements,
including story drift, coupling beam
end rotation, pier compressive and
tensile strains, and pier shear demands,
see Figures 6.40–6.44. Acceptance
criteria in the applicable PBSD
guidelines document were intended
for typical commercial construction
in Risk Category II, but this tower was
in Risk Category III based on its high
occupancy. Higher risk categories are
intended to have lower probabilities
that demand exceeds capacity. To
reflect that probability, the peer review
committee requested that drift and
rotation results summaries like Figures
6.40 and 6.41 present both mean MCER
74 | Performance-Based Seismic Design Examples
5Figure 6.41: Coupling Beam S-54 Plastic Rotation Demands (highest utilized
coupling beam). © Thornton Tomasetti
results (per PBSD guidelines) and mean
MCER-plus-one standard deviation
results. Mean-plus-standard-deviation
results that are still within acceptance
limits means a lower probability of
exceedance. (More recent editions
of the LATBSDC PBSD Guidelines
address risk categories directly
by factoring acceptance criteria.)
Coupling-beam rotation acceptance
limits were recommended values
based on laboratory tests of beams
with rebar detailing, consistent with
current American Concrete Institute
requirements for ductile seismic-
resisting elements, as reported by
Naish at UCLA in 2010. Limit values
vary with the beam aspect ratio,
reinforcing arrangement, and shear
stress. These limits were deemed
more appropriate than ASCE 41 limits,
which were established for beams
in older, existing concrete buildings
with outdated details, such as low-or
no-beam concrete confinement. Note
that acceptance was demonstrated by
mean MCER being below the halfway
point between life safety and collapse
prevention limits, and mean-plus-
standard-deviation results being less
than the collapse prevention limit. Both
figures show “pinch points,” where story
and beam deformations are reduced by
belt trusses working in parallel with the
core wall (top and middle) and by a stiff
“basement box” acting as a lateral prop
(below floor 1).
Strain summaries were used for wall
piers, using model results from “strain
 23 24
(P4-F52) (P4-F52)
5 3 1 7 9

13 14 15 16 17 18 11
12

25 26 31 32
27 28 29 30
6 10
4 8
2

5Figure 6.43: Core Geometry and Representative Wall Strain Gage Locations (coupling beams in
green). © Thornton Tomasetti

F72 OUTRIGGER F72

F70 F70
F66 F66
F64 F64
F62 F62
F60 F60
F58 F58
F56 F56
OUTRIGGER
F54 F54
F52 F52
F50 F50
F48 F48
F46 F46
F43 F43
F41 F41
F39 F39
F37 F37

Core Strains F35 F35

CONFINED
CONFINED
F33 F33

Above Yield F31

F29 OUTRIGGER
F31
F29
F26 F26
STEEL YIELD

STEEL YIELD
F24 F24
F22 F22
F20 F20
F18 F18
F16 F16
F12 F12
F7 F7
F5 F5

CONFINED
CONFINED

F1 <<0.003 OK F1
P2 P2
P BASEMENT P4
0.0030 0.0025 0.0020 0.0015 0.0010 0.0005 0.0000 -0.0005 -0.0010 -0.0015 0.0030 0.0025 0.0020 0.0015 0.0010 0.0005 0.0000 -0.0005 -0.0010 -0.0015

Figure 17 Distribution of strains at Gage Line 6 (Model AC0) Figure 28 Distribution of strains at Gage Line 17 (Model AC0)
5Figure 6.42: Concrete Core Flexural Demand Contours (yielded 5Figure 6.44: Mean MCER Strains at Two Core Locations Show Local Peaks at Potential Hinge Zones
regions shown in red). © Thornton Tomasetti Where Confinement is Provided. © Thornton Tomasetti

gages” located at the ends of each
core pier. Figure 6.42 shows that steel
yielding is concentrated just above
the stiff “basement box” and just above
the stiff lower outrigger, which was
consistent with expectations and the
design approach. Figure 6.43 shows
strain gage locations, and Figure 6.44
shows mean MCER compression and
tension strain results at a core corner
and an interior door jamb. With mean
MCER net tension slightly above rebar
yield and mean MCER compression
well below the concrete crushing
limit, results are clearly acceptable.
Notations show outrigger levels and
highlight the stories with fully confined
concrete to assure ductile behavior at
anticipated hinge zones. Elsewhere,
the number of cross-ties was reduced
and the vertical spacing between sets values, but using BRB diagonals will
of hoops and cross-ties was doubled, provide a natural “cap” on that demand.
compared to the hoop and cross-tie Because PBSD guidelines don’t reflect
arrangement, and the requirement this beneficial behavior, the model
that vertical spacing meet ACI was run two ways: first with MCER
standards for wall boundary-element time histories, then by multiplying
seismic confinement. the resulting shears by 1.5, and 1.5 x
MCER time histories, with shears used
directly. Two different time-history
6.7 Core-Wall Shear Demands and load cases were run to confirm the
Acceptance – Use of BRB Properties capacity-protection design approach:
to Limit Demands if the structure behaved elastically,
the results would be identical; while
Core-wall shear is a critical force- having beneficial yielding would result
controlled action rather than a in reduced demand values for the
deformation-controlled action, 1.5MCER case compared to the 1.5 x
which by PBSD guidelines mean MCER case. Figure 6.45 shows that BRB
MCER demands are multiplied by 1.5. yielding does indeed provide capacity
“Panel-zone” shear at outriggers could protection. It reduces panel-zone shear
be the largest wall shear demand to an acceptable level without adding

Performance-Based Seismic Design Examples | 75

 F72 OUTRIGGER
F70
F66

Two different time- F64

Shear Capacity-1
Shear Capacity-2
F62

history load cases were

F60

10sqrt(fce')
Demands

Demands
F58
OUTRIGGER 10sqrt(fce')
F56 Shear Capacity-1
run to confirm the F54
F52
Shear Capacity-2

capacity protection F50

F48

design approach. If
F46
F43

Floor ID
F41

the structure behaved F39

F37

elastically, the results

F35
F33
F31

would be identical, F29

-14000 F26
(NOT TO
IF 1.5 MCE OUTRIGGER

while having beneficial SCALE) F24

-18000
F22

yielding would result (NOT TO F20

SCALE) F18
F16

in reduced demand F12

F7

values for the 1.5MCER

F5
F1
BASEMENT
0 5000 10000 15000 20000 25000

case compared to the 1.5 x MCE Shear (kips) at Wall T05

1.5 x MCER case. 5Figure 6.45: Wall Shears at 1.5 x MCER Showing “Panel Zone” Effects from Outrigger Force Couples, reducing or
even reversing wall shears. Curves to the left symbolically reflect reversal seen under static loading, using same-
force increments (while the solid curve to right shows “directionless” results from dynamic analyses). The shear
under mean of 1.5 MCER (inner curves) is less than that under 1.5 times mean MCER, showing the force-limiting
effect of BRB yielding. The green shear capacity line is based on rebar provided regardless of concrete-based limits,
to provide perspective. © Thornton Tomasetti

slow-to-build and costly embedded
steel members to the core walls.
The large jumps in shear capacity
(ignoring any concrete-strength-based
upper limits) reflect confinement
reinforcement at the two anticipated
hinge zones.
BRB Diagonal Member Demands and
Acceptance
For steel members, only the outrigger
diagonal BRBs are displacement-
controlled. Peak strain in each BRB for
each MCER time history was tracked.
Mean MCER strains in tension and
compression appear in Figure 6.46 as
multiples of yield strain, drawn to the
same vertical scale. Top and bottom
outriggers experience less strain than
middle outriggers, which have yield
zones that are a smaller proportion of
their total length between work points.
Top and bottom outriggers also show
much larger strains in compression than
in tension. That is due to the kinematics
of core flexure. Like an under-reinforced
beam, as core wall tension-side rebar
yields, tension-side strains are much
larger than compression-block strains.
Outrigger columns on the core tension
side also experience larger outrigger
forces, which strain the attached
outrigger diagonal (in compression)
more than the outrigger diagonal on
the core compression side is strained
(in tension). The mean MCER strain of
12 times yield is well within the tested
capabilities of these BRBs, so the results
are acceptable.
6.8 BRB Force Demands on
Connections
While BRB strains vary widely by
location and direction, BRB forces are
very well defined, with maximum forces
clearly capped. Figure 6.47 shows mean,
mean-plus-one standard deviation
(84th percentile), and peak (largest of
any time history) MCER forces for the
10 diagonals at lower outriggers. For
these four-element diagonals, design
yield force is 8,800 kips. At the 84th
percentile, no force exceeds 10,000
kips (some peak values slightly exceed
10,000 kips). For designing outrigger
chords and connections to columns
and core, the over-strength factors for
BRBs in AISC seismic provisions were
overly conservative for this situation.
AISC provisions assume BRBs are sized
to meet design (reduced) seismic forces
and therefore experience very large
inelastic strains. For this project, axial
forces enveloping at least the 84th-
percentile values, or conservatively peak
values, were used for connection design
after applying the mandatory 1.1 force

76 | Performance-Based Seismic Design Examples

 ϭϮ͘Ϭ

dĞŶƐŝŽŶ
ϭϬ͘Ϭ ŽŵƉƌĞƐƐŝŽŶ

ϴ͘Ϭ
∆LJ
∆Ƶͬ∆

ϲ͘Ϭ

ϱ͘Ϭ

dĞŶƐŝŽŶ
ϰ͘ϱ
ŽŵƉƌĞƐƐŝŽŶ

ϰ͘Ϭ ϰ͘Ϭ

ϯ͘ϱ ϯ͘ϱ

dĞŶƐŝŽŶ
ϯ͘Ϭ ϯ͘Ϭ
ŽŵƉƌĞƐƐŝŽŶ

∆LJ
∆Ƶͬ∆
Ϯ͘ϱ Ϯ͘ϱ

Ϯ͘Ϭ
Ϯ͘Ϭ Ϯ͘Ϭ
∆Ƶͬ∆∆LJ

ϭ͘ϱ ϭ͘ϱ

ϭ͘Ϭ ϭ͘Ϭ

Ϭ͘ϱ Ϭ͘ϱ

Ϭ͘Ϭ
ZϭͲdϬϳE ZϭͲdϬϳ^ ZϭͲdϬϲE ZϭͲdϬϲ^ ZϭͲdϬϱE ZϭͲdϬϱ^ ZϭͲdϬϰE ZϭͲdϬϰ^ ZϭͲdϬϯE ZϭͲdϬϯ^ Ϭ͘Ϭ Ϭ͘Ϭ
ZϯͲdϬϳE ZϯͲdϬϳ^ ZϯͲdϬϲE ZϯͲdϬϲ^ ZϯͲdϬϱE ZϯͲdϬϱ^ ZϯͲdϬϰE ZϯͲdϬϰ^ ZϯͲdϬϯE ZϯͲdϬϯ^

5Figure 6.46: Mean MCER Strains in Tension and Compression. Note asymmetry between tension (smaller) and compression. © Thornton Tomasetti

factor. Connections also considered

simultaneous yield moments Ry x Fy
x Z, where Ry is the ratio of expected 12000
yield stress to the specified minimum 10000
yield stress Fy, Z is the plastic section
modulus for flexure, and resulting 8000
shears from outrigger chords being 6000
deformed as the BRBs stretch or Mean (T) 84th %-ile (T) Peak (T)

compress, as shown in Figure 6.48. 4000 Mean (C) 84th %-ile (C) Peak (C)
BRB Axial Force (kips)

2000
To minimize induced moments, chords
0
were designed as shallow as possible BRB1-T07N BRB1-T07S BRB1-T06N BRB1-T06S BRB1-T05N BRB1-T05S BRB1-T04N BRB1-T04S BRB1-T03N BRB1-T03S
considering simultaneous axial -2000
outrigger forces and transverse floor
-4000
loading. Flange and web thicknesses
of the built-up box chords (bottom -6000
outrigger) and I-shaped chord with
-8000
reduced beam sections (top outrigger)
were established to meet seismically- -10000
compact criteria, along with bracing
-12000
at ductile hinge locations as needed. BRB ID
The resulting forces were resolved in 5Figure 6.47: Forces in Bottom Outrigger BRBs. Note: Forces vary only slightly from mean to peak, by location, or
gusset plates, and then in steel plates by force direction; forces are “capped” by well-defined yield behavior. © Thornton Tomasetti

Performance-Based Seismic Design Examples | 77

cast into the core wall faces, with
reinforcing rod tension ties threaded
to couplers and headed steel studs
for vertical shear force transfer.
Gusset sizes and shapes minimize
stress concentrations, and local
stiffeners avoid buckling.
6.10 Composite Column Demands behavior of outrigger BRBs, the model
and Acceptance was run with 1.5xMCER time histories.
This also greatly simplified checking,
Tower composite columns and belt- as capacity curves were established
truss chord and diagonal members within the Perform-3D analysis
were checked as force-controlled program that closely followed AISC
elements. To reflect the force-capping composite-member capacity formulas

Acceptable connection performance

(only localized yielded zones and no
 0
0,1
67,))72
*866(7
 0
0,1

67,))60$7&+%2
2; &/ &2/
)/$1*( 3/$7(6



buckling of gussets) was checked by

6
:3

 

&/ %0

the BRB fabricator’s engineer using & 7

LY

6

6,'('%2/7('
6+$3(*866(7 ',$3+5$*0
 7<3
)25&251(5 %(7:((1
75$16,7,21 &/($5)25 *866(76 LL
6 &-3:(/'

nonlinear Finite Element Analysis,

LLL %$&.(56

considering simultaneous BRB 6+$3(*866(7

compression, flexure induced by

)25 &251(5
75$16,7,21

frame deformations, shear and axial

forces by steel chords, transverse
loading from seismic acceleration of
7 &
the BRB mass, and transverse loading
from story deformations.


7<3
6

6.9 Use of Pushover Analysis to

Identify Progressive Peak Demands 

in Connections
/$363/,&(7<3 7<3
6
 0,1
7<3

5Figure 6.48: Simultaneous Forces Acting on Bottom Outrigger Connections. © Thornton Tomasetti
At the middle outriggers, interplay
of adjacent BRB diagonals and small Axial
refinements in geometry complicated Force W33x318
Pushover
2125
1200k 1070k
connection force determination. The 1940

W33x318
Graphs
overall building model used simple
To 1.5% 1180k 1115k
centerline work points, which were Drift
245
W33x387
sufficient for general BRB strain and
force determination. But a separate 1190k 1055k
W33x318
detailed model with nonlinear BRB
elements was used to study effects of 243 1170k 1085k
W33x387
local geometry and BRB interactions
through a pushover approach. 1150k 1030k
W33x318
Resulting overall outrigger stiffness 245

findings were used to establish 1120k 1050k

appropriate properties for main 110
W33x318
building model elements to reflect
realistic behaviors. More important,
tracking the progression of forces
through connections revealed 118

temporary peak demands (see Figure

6.49) that were not present in the
1800

final, fully yielded condition. These 1990

peaks were used for the design of

outrigger-to-core pins, gussets, and 5Figure 6.49: Axial Force Pushover Graphs with Maximum BRB forces shown to the right. Note: Maximum forces
embed plates. are not all at maximum drift. © Thornton Tomasetti

78 | Performance-Based Seismic Design Examples

 F71
F71

F69
F69

F65
F65 TAT9
F63
F63 TAT8
F61
F61
TAT7
F59
F59
TAT6
F57
F57
TAT5
F55
F55
TAT4
F53
F53

F51 TAT3
F51

F49
F49
TAT2

F47
F47

F45
F45
P3D_Mx
F42
F42
P3D_My
F40
F40
AISC Method_Mx
F38
F38

Floor ID
AISC Method_My
F36
F36

F34
F34

F32
F32

F30
F30
Axial force (kips)

F28
F28

F25
F25

F23
F23

F21
F21

F19
F19

F17
F17

F15
F15

F11
F11

F6
F6

F2
F2

P1
P1

P3
P3

Mat
Mat
Moment (kip-in)
� 0.00
0.00 0.20
0.20 0.40
0.40 0.60
0.60 0.80
0.80 1.00
1.00

��������������������������������������������������� ����� Column Utilization

���������������������������������������������������

5Figure 6.50: Column Capacity Modeling in Perform-3D Compared to AISC Composite 5Figure 6.51: D/C for Columns at Mean of 1.5MCER. © Thornton Tomasetti
Method 2. © Thornton Tomasetti

or steel member formulas at trusses,
as shown in Figure 6.50. DCRs were
determined for each time step of each
event, with the maximum demand-
to-capacity (D/C) value for each event
used to determine mean 1.5xMCER
D/C column results (see Figure 6.51).
Jumps in D/C values represent changes
in column sizes, locations above or
below outriggers, and/or column
participation as a vertical in the belt
trusses. Looking at one column, Figure
6.39 shows little effect on D/C results
from using a fiber model (one steel
fiber for each box column face plus a
central concrete fill fiber), rather than a
simpler transformed area model.
6.11 Composite Column Steel Plate
Welding Demands and Acceptance
NRHA results were used to help
simplify and economize steel box
column fabrication and erection.
In Figure 6.52, axial demands from
the mean of 1.5MCER time histories
at one of the outrigger columns
shows large compression forces and
relatively small tension forces. Jogs
in the graph reflect load application
from outriggers and/or participation
of the column as a belt-truss vertical.
Figure 6.53 shows net tension steel
stresses for the same column under
1.5MCER. Where net stresses are less
than 25 ksi, welded column splices
at two facing plates of the box were
permitted to use partial penetration
welds with a 1/8” land, or flat surface in
bearing contact, easing erection work.
While the other two plates would
have complete penetration welds,
terminating at backer and runoff tabs
through generous weld access holes.
At higher stress locations, complete
penetration welds were required on
all four box plates. Acceptability of
partial penetration welds, and the
inevitable flaws in all welds, were
studied, considering crack growth
from decades of cyclic wind loading,
At the middle
outriggers, interplay
of adjacent BRB
diagonals and small
refinements in
geometry complicated
connection-force
determination.
followed by potential for fracture in
a major quake. Charpy V-notch (CVN)
testing and minimum values for base
metal, filler metal, and completed
sample welds were established to
provide sufficient toughness to
avoid failure in this scenario. Note
that CVN minimum values stated in
seismic provision recommendations
referenced by building codes may not

Performance-Based Seismic Design Examples | 79

always be sufficient to address these
multiple concerns, particularly where
applied or weld-shrinkage loads act
in the through-thickness “cross-grain”
direction of plate or rolled shapes
or in highly constrained weldments.
Bridge welding codes may provide
useful guidance.
6.12 Steel Belt-Truss Analysis and
Acceptance
Belt truss members experience
loads from several sources, such as
distributing vertical components
from outriggers to adjacent columns,
redistributing gravity loads along the
perimeter, distributing forces from
differential shortening along the
perimeter, and “virtual (or indirect)
outrigger” behavior trying to resist core-
rotation and story drifts from lateral
movements in the east–west direction.
For “virtual (indirect) outrigger” behavior
in particular, the load path involves floor
diaphragms at or near the belt-truss
top and bottom chords: core rotation
at the belt elevation causes relative
lateral displacements of adjacent floors,
but belt trusses attempt to restrain the
relative displacements by inducing
changes in perimeter-column axial
forces. This is similar to, but less efficient
than, conventional outriggers that
directly engage perimeter columns to
restrain core rotation. Thus the terms
“virtual” and “indirect” are applied to this
type of behavior. Since diaphragms are
key elements of the core-belt-column
load path, diaphragm stiffness is an
important determinant of indirect
outrigger effectiveness and the forces
generated. To bracket the likely range
of forces in belt truss members,
sensitivity studies were performed,
including varying diaphragm stiffness at
multiple levels and applying gravity and
lateral loads in various combinations.
Figure 6.54 shows a floor diaphragm
with “dummy” W14 brace members
to simulate in-plane floor stiffness.
Member sizes ranged over a factor of
10 to cover uncracked-to-extensively-
cracked floor slab scenarios. At lower
belt diagonals shown in Figure 6.55,
DCRs for one loading scenario are
shown in Figure 6.56. Other scenarios
caused variations in D/C values, but all
were within an acceptable range. Since
dummy members in the computer
model will share some of the load
in truss chords (horizontals), chord

F71
TAT3 F71
F69
F69
F65
F65
F63
F63
F61
F61
F59
F59
F57 TA-T3
F57
F55
F55
F53 F53
F51 F51
F49 F49
F47 F47
F45 F45
F42 F42
Floor ID

F40 F40
Floor ID

F38 F38
F36 F36
F34 F34
F32 F32
F30 F30
F28 F28
F25 F25
F23 F23
F21 F21
F19 F19
F17 F17
F15 F15
F11 F11
F6 F6

F2 F2

P1 P1

P3 P3

Mat Mat
0.00 0.20 0.40 0.60 0.80 1.00
-20000 -15000 -10000 -5000 0 5000 10000
PT/A (ksi)
Column Axial Force (kips)

5Figure 6.52: Axial Force Envelope at Mean of 1.5MCER. © Thornton Tomasetti 5Figure 6.53: Axial Stress in Steel Box Plates. © Thornton Tomasetti

80 | Performance-Based Seismic Design Examples

member forces would be under-
reported. So, truss chords were checked
using forces from computer-model
section cuts that reflect both actual
and dummy members, as well as hand
calculations considering likely ranges
of forces in adjacent verticals (columns)
and diagonals.

6.13 Rooftop Feature Sail and Spire

Demands and Design Approach

It was recognized that demands acting 5Figure 6.54: Plan of Dummy Members for Diaphragm Sensitivity Studies at Belt Trusses. © Thornton Tomasetti
on the 100-foot-tall (30-meter) steel
braced-frame “sail” and an adjacent
tapered steel spire at the top of the
building would be driven by behavior
of the building below them. For other
“nonstructural” elements in the building,
floor-by-floor accelerations were read F31
B1S3c B1S4c B1S5c B1S6c
from MCER time history runs, and their W B1S3b B1S4b B1S5b
F30
B1S1 B1S2 B1S6b B1S7
dispersion is presented in Figures 6.57 B1S3a B1S4a B1S5a
F29
B1S6a
and 6.58. A simplified design criterion F28
for each direction was defined by the
dashed green lines. It was interesting to
5Figure 6.55: Key Elevation of Lower Belt Truss to Track D/C Results at 1.5MCER. © Thornton Tomasetti
note that the 2/3 MCER value, normally
considered as a Design Basis Earthquake,
is quite similar to code formula values for
flexibly-supported equipment. However,
the “whiplash effect,” visible as a jump 1.00
in accelerations at the tower top, was
South
of particular concern for sail and spire 0.90 North
1.5 x MCE

1.00
design. To establish an initial design,
0.80 South
we developed secondary, roof-level 0.90 North
MCE

response spectra for the two principal 0.70

with 1.5 xwith

plan directions, taken by enveloping 0.80

BT Utilization

0.60
the peaks of reported rooftop spectra. 0.70
Spectra for the non-principal direction 0.50
BT Utilization

are shown in Figure 6.59, showing 0.60

only a small increase going from MCER 0.40
0.50
to 1.5MCER at most periods, thanks 0.30
to yielding of the outrigger BRBs. The 0.40
1.5MCER envelope spectra were used 0.20
0.30
for elastic design of the sail and spire
0.10
framing, because post-earthquake 0.20
visible permanent deformations of these 0.00
0.10
B1-3a

B1-3b

B1-4a

B1-4b

B1-5a

B1-5b

B1-6a

B1-6b
B1-1

B1-2

B1-3c

B1-4c

B1-5c

B1-6c

B1-7

elements was considered unacceptable

by the reviewers and the owner. Other 0.00 BT ID
B1-3a

B1-3b

B1-4a

B1-4b

B1-5a

B1-5b

B1-6a

B1-6b
B1-1

B1-2

B1-3c

B1-4c

B1-5c

B1-6c

B1-7

5Figure 6.56: Lower Belt Truss Diagonal Member DCRs for Mean of 1.5MCER. © Thornton Tomasetti
BT ID

Performance-Based Seismic Design Examples | 81

 ϳϮ ϳϮ

ϲϴ ϲϴ

ϲϰ ϲϰ

ϲϬ 241215'&
#%%'.'4#6+105(14 ϲϬ Et;DĞĂŶͿ
&+#2*4#)/&'5+)0
ϱϲ ϱϲ Et;DĞĂŶн^ŝŐŵĂͿ 241215'&
#%%'.'4#6+105(14
^;DĞĂŶͿ Et;DĞĂŶͲ^ŝŐŵĂͿ &+#2*4#)/&'5+)0
ϱϮ ϱϮ
^;DĞĂŶн^ŝŐŵĂͿ
ϰϴ ϰϴ
^;DĞĂŶͲ^ŝŐŵĂͿ
ϰϰ ϰϰ #EEGNGTCVKQP
'9
ϰϬ ϰϬ
&ůŽŽƌ/

&ůŽŽƌ/
ϯϲ ϯϲ
09
ϯϮ ϯϮ
#EEGNGTCVKQP
05
Ϯϴ Ϯϴ
5'
Ϯϰ Ϯϰ

ϮϬ ϮϬ

ϭϲ 09 ϭϲ

ϭϮ ϭϮ

ϴ 5' ϴ

ϰ ϰ

Ϭ Ϭ

Ͳϰ Ͳϰ
Ϭ͘Ϭ Ϭ͘ϱ ϭ͘Ϭ ϭ͘ϱ Ϯ͘Ϭ Ϯ͘ϱ ϯ͘Ϭ ϯ͘ϱ ϰ͘Ϭ Ϭ͘Ϭ Ϭ͘ϱ ϭ͘Ϭ ϭ͘ϱ Ϯ͘Ϭ
0HDQ0&( WĞĂŬ&ůŽŽƌĐĐĞůĞƌĂƚŝŽŶ;ŐͿ WĞĂŬ&ůŽŽƌĐĐĞůĞƌĂƚŝŽŶ;ŐͿ
5WOOCT[QH2GCM(NQQT#EEGNGTCVKQPU/%' 5WOOCT[QH2GCM(NQQT#EEGNGTCVKQPU/%'
5Figure 6.57: North–South MCER Diaphragm Accelerations. © Thornton Tomasetti 5Figure 6.58: East–West MCER Diaphragm Accelerations. © Thornton Tomasetti
%GPVGT&T5WKVG 2TQLGEV 9KNUJKTG)TCPF6QYGT %GPVGT&T5WKVG 2TQLGEV 9KNUJKTG)TCPF6QYGT
.QU#PIGNGU%# .QU#PIGNGU%#
2 &CVG T 2 &CVG T
( (
$[ -) 2CIG $[ -) 2CIG

North-South Spectrum response-reduction measures, such as

7.00 damped rocking were studied, but found
MCE 5% Mean Spectrum
to require impractically large movements
6.00 1.5MCE 5% Mean Spectrum
since the tower roof level itself would be
displacing several feet. The sail and spire
preliminary framing was then reflected
5.00
1.5 MCE smoothed in the NRHA model, and the results were
spectrum for ETABS
studies?
small increase for MCE --> 1.5 MCE found to be acceptable.
because BRB outrigger yielding
4.00
limits forces at lower modes
MCE smoothed
Sa (g)

spectrum for ETABS

studies?
3.00 6.14 Differential Shortening Effects
on Building Performance and
2.00
Construction

The final topic of study was the effect

1.00
of differential shortening on building
performance. The concrete core will
0.00
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0
shrink and creep over decades of
Period (s) service, with the movement being slow
and gradual. In contrast, the concrete fill

5Figure 6.59: MCER and 1.5MCER North-South Spectra at Tower Rooftop. © Thornton Tomasetti

82 | Performance-Based Seismic Design Examples

of the perimeter columns is well-sealed
against drying, so column creep and
shrinkage will be much smaller. The
outriggers will attempt to enforce
equal shortening at the columns
and core, resulting in transfer forces
relieving core load, adding to column
load and putting BRBs in tension or
compression, depending on their
orientation. Even so, some differences
will occur, as shown in Figure 6.60.
The top outrigger BRBs were jacked in
compression to 50 percent of capacity
at yield, because predicted long-term
differential shortening over the life of
the building could generate enough
tension in top outriggers to approach
yield. This approach ensures that some 5Figure 6.60: Predicted Shortening of Outrigger Columns and Concrete Core at Floor 72 at 50 Years After
reserve to yield is present at all points Construction (shortening subsequent to floor construction). © Thornton Tomasetti
in the building’s service life to resist
wind loading. Ample reserve to yield 0.010

exists at the other outriggers. Standard

0.008 Construction Sequence + Jacking
Long-Term Shortening
Shortening and force transfer values
0.006
were predicted using the Bazant B3
model, modified to reflect rebar or 0.004
steel-plate participation and iterative
building models. In Perform-3D, core
Drift (radians)

0.002

shortening was modeled by artificially

compressing selected stories an 0.000
0 10 20 30 40 50 60
appropriate amount, with resulting
induced forces checked against -0.002

an ETABS model using an artificial

temperature reduction on the core
to simulate the same phenomenon.
Transfer load adjustments were also
made to reflect realistic construction
stages, rather than applying dead load
to the complete model. The effect
of shortening on building behavior
was checked by looking at story
movements under various loading
scenarios. As Figure 6.61 shows,
shortening forces do affect the timing
of peak seismic responses, but do
not worsen the maximum response
values. This confirms that the building
performance will meet acceptance
criteria, whether a large event occurs
-0.004
-0.006
-0.008
Time (s)
5Figure 6.61: MCER Story Drift Comparisons for no Shortening, Construction Sequence & BRB Jacking, and Long-
Term Shortening (Floor 45, north–south direction). © Thornton Tomasetti
soon after opening day or many demonstrate acceptable performance
decades in the future. under PBSD. But with well-planned
models, the necessary NRHA results can
This case history shows that structures be extracted and clearly presented in a
with mixed materials and composite reasonably efficient process.
elements require studies and checks
of numerous design aspects to

Performance-Based Seismic Design Examples | 83

Performance-Based Seismic Design Examples
Design Example 3: Tall Flat-Plate Office Tower
5Figure 6.62: Photos of Completed Building. © Skidmore, Owings & Merrill LLP
6.15 Project Description
The following is a structural system
description for a tower situated in
downtown San Francisco, California (see
Figure 6.62). The occupancy consists of
office above grade and parking below
grade. The 30-story tower is 384 feet
and two inches above grade with three
basement levels below grade and a
total building area of approximately
455,000 square feet. The seismic force-
resisting structural system consists of
reinforced concrete core walls from the
foundation to roof. The gravity system
is a long-span, flat plate, post-tensioned
system. This combination of lateral and
gravity systems is common in high-rise
residential construction, but has been
adapted in an efficient manner for
high-rise office application.
84 | Performance-Based Seismic Design Examples
Superstructure
The superstructure consists of a central
reinforced concrete shear-wall core,
perimeter gravity columns, and two-
way flat-plate slab framing. The tower is
roughly square in plan with dimensions
of 125’–0” x 130’–0”. The typical office
floor-to-floor height is 13’-2”.
Lateral System
The lateral system consists of a centrally
located reinforced concrete shear-wall
core. The shear wall core has an
external plan area of 43’–0” x 52’–6” and
is located around the service area of
the structure, passenger and service
elevators, and back-of-house areas.
The height-to-depth ratio of the core
is approximately 9:1. The shear wall
core extends from foundation to roof.
The shear walls vary in thickness from
24” to 33” and in concrete compressive
strength from 6,000 psi to 8,000 psi. The
shear wall core is interconnected with
the use of ductile diagonally-reinforced
link beams at openings required for
doorways and corridors.
Gravity System
The gravity framing system both
inside and outside the core consists
of a two-way post-tensioned (PT)
flat-plate slab. The slab clears spans
from the core to the perimeter and
has a uniform thickness of 11”. The
long-span flat slab is post-tensioned
to maintain uncracked section under
service gravity loads, and cambered for
a portion of the long-term deflection.
The slab utilizes concrete with a
compressive strength of 5,000 psi. Due
to the long-span condition, a digitally

RC Columns
11“ Thick PT
Flat Slab
Ductile
RC Core
5Figure 6.63: Typical Floor Plan. © Skidmore, Owings & Merrill LLP
mapped camber program was used
to ensure deflections would meet
tenant requirements. The perimeter
vertical gravity columns are typically
composed of conventional reinforced
concrete sections, varying in size from
42” x 42” square to 26” x 26” square.
The columns utilize concrete with
compressive strengths ranging from
6,000 psi to 8,000 psi. The tall columns
at the entry lobby consist of 42” x 42”
composite members utilizing steel
cruciform shapes embedded within
the concrete column. The southwest
corner is double-cantilevered 30
feet with the use of upturned post-
tensioned beams (see Figure 6.63).
Substructure
Vertical elements of the superstructure
continue down through the
115 ft
123 ft
45 ft
60“ wide x 25“ deep
Upturned PT Beam
substructure to the foundation.
Shear walls are 24”–33” thick, with a
compressive strength of 8,000 psi.
The columns are typically 36” x 48”
and consist of 8,000 psi concrete. The
gravity system in the substructure
consists of a 10” thick, conventionally
reinforced, two-way flat-plate slab that
utilizes concrete with a compressive
strength of 5,000 psi.
Foundations
The foundation system consists of a
10’–0” thick, conventionally-reinforced
concrete mat foundation (see Figure
6.64). A perimeter reinforced concrete
foundation wall system consists of
conventional 16”–22” thick cast-in-place
concrete walls. The substrate consists of
dense sands over pre-consolidated clay.
5Figure 6.64: Building Section. © Skidmore, Owings
& Merrill LLP
6.16 Design Criteria
The building was designed under the
San Francisco Building Code SFBC 2010,
which refers to the California Building
Code (2010) and ASCE 7 (2005). The
CBC is adopted from previous IBC
publications with amendments specific
to California.
The tower in this study takes exceptions
to the prescriptive code height limits of
ASCE 7, specifically, the building height
limit prescribed by CBC 2010 and ASCE
7-05 for shear wall-only buildings. This
exception to the code height limit
requires that the structural system
be classified as a non-prescriptive
seismic force-resisting system. The
building is required to meet the seismic
performance intent of the building
Performance-Based Seismic Design Examples | 85
codes and is required to utilize San
Francisco Building Code Administrative
Bulletin AB-083 for seismic design,
“Requirements and Guidelines for the
Seismic Design of New Tall Buildings
using Non-Prescriptive Seismic-Design
Procedures.” AB-083 also requires
review of the design by a Structural
Peer Review Panel (SPRP), which has
to coincide with the requirements
stated in AB-082, “Requirements
and Guidelines for Structural Design
Review Procedures.”
To demonstrate that the building is
designed to have seismic performance
at least equivalent to the intent of a
code-prescriptive building, AB-083
requires that a three-step procedure
be performed by the SEOR with an
independent, objective, peer review.
Step One: Code-Level Seismic Design
The code-level design is used to
identify the exceptions being taken
to the prescriptive requirements of
the SFBC building code and to define
The design criteria
for the MCER-level
evaluation includes
a suite of 10 time-
history pairs of ground
motions, selected
and modified using
a motion-specific
amplitude-scaling
factor for each pair to
achieve compatibility
with the target MCER
shaking spectrum.
86 | Performance-Based Seismic Design Examples
the minimum required strength and
stiffness for earthquake resistance
under DE-level seismic demands. The
design incorporates the prescriptive
provisions of R, Ωo, ρ, and Cd into a
linear elastic modal response-spectrum
analysis. The results of the linear elastic
analysis are used to evaluate force
levels in the lateral system components.
The structural demand in each
element is used to initially design each
member, considering appropriate load
combinations and ultimate strength
equations per ACI 318. The ultimate
shear values for shear walls and link
beams are multiplied by 1.5 to more
closely approximate shear demands
expected under MCER-level evaluation.
Prior efforts in the design of tall
buildings using nonlinear analysis have
demonstrated that code-prescribed
R-values may be non-conservative for
shear design of walls causing design
issues after NRHA is conducted. The
initial prescriptive code-based member
designs are used as inputs to the linear
and nonlinear response-history analysis
evaluations below.
Since the time of this design, the design
firm has adopted different initial design
procedures that better capture MCER
demands and behavior. The factor of
1.5 applied to DE-level shear demands
in walls has since been determined to
be ineffective. This resulted in design
changes based on NRHA, extending
design efforts. This highlights the
importance of initial proportioning as
it relates to the effort needed by the
design engineer conducting PBSD.
Step Two: Service-Level Evaluation
A service-level evaluation is required to
demonstrate essentially elastic seismic
performance for frequent earthquakes.
The Service-Level Earthquake (SLE)
is defined as having a 43-year mean
return period (50 percent probability
of exceedance in 30 years). The design
for the service-level evaluation does
not incorporate the code-prescriptive
provisions of R, Ωo, ρ, and Cd. These
values are all taken to be 1.0. Rather,
the SLE shaking is derived from a
site-specific 2.5 percent damped linear
uniform hazard acceleration-response
spectra. The following acceptance
criteria are evaluated:
a) The DCRs for all lateral system
components shall not exceed 1.5.
b) The story drift shall not exceed
0.5 percent of the story height in
any story.
The capacity of the lateral system
components is defined as the design
strength, which is taken as the
specified strength multiplied by the
corresponding strength reduction
factor Φ in accordance with ACI 318.
The adopted service-level procedure
follows the PEER/TBI 1.0 approach,
which utilized specified strength of
materials and the strength reduction
factor Φ. Whereas, by contrast, LATBSDC
(2014) recommends using expected
material properties and not applying the
strength reduction factor. Commentary
on this difference can be found in
LATBSDC (2014) commentary C.3.4.5.
Step Three: Maximum Considered
Earthquake (MCER)-Level Evaluation
The MCER-level evaluation is intended
to verify that the structure has an
acceptably low probability of collapse
(on the order of 10 percent) under
severe earthquake ground motions. The
MCER-level ground motion is defined as
the lesser of the spectrum, having a two
percent probability of exceedance in 50
years (2,475-year return period) or 150
percent of the median deterministic
spectrum for the governing fault. The
performance objective, design criteria,
modeling and design procedures, and
acceptance criteria are summarized in
the following sections.
Ground Motions
The design criteria for the MCER-level
evaluation includes a suite of 10
time-history pairs of ground motions,
selected and modified using a motion-
specific amplitude scaling factor for
each pair for compatibility with the
target MCER shaking spectrum. The
10 time-history ground motions are
selected from a set of 25 ground
motions based on factors such as
spectral shape, peak velocity, etc.
Each of the 10 time-history pairs are
randomly rotated and used as the
seismic input in the NRHA 3D computer
model. The Perform-3D software is used
to perform NRHAs.
Acceptance Criteria
The acceptance criteria consider the
individual components and global
response measures. The component-
level criteria can be classified
into force-controlled actions and
deformation-controlled actions. Force-
controlled actions are those in which
inelastic deformation capacity cannot
be assured. These include axial forces
in columns and shear in reinforced
concrete shear walls. The acceptance
criteria for these compare the factored
mean value obtained from the suite of
analyses to the strength based on ACI
318, using expected material properties.
As additional measures, the strength-
reduction factor is applied for axial load
in columns, and 1.5 times the mean
response is used to determine shear
demands in walls.
Deformation-controlled actions
are those in which reliable inelastic
deformation capacity is achievable
without critical strength decay.
These include axial load in shear
walls with special confined boundary
zones, and flexural and shear loads
in diagonally-reinforced link beams.
The deformation capacity for shear
walls is taken from the collapse
prevention values published in Table
6-18 and 6-19 of ASCE 41 for nonlinear
response procedures. For link beams,
previous test results for similar beams
(Naish 2010, 2013) are used to derive
acceptance criteria. The deformation
acceptance limit corresponds to a 20
percent strength loss.
At the global level, peak and residual
drifts are evaluated. In each story, the
mean of the peak transient drift ratios
from the suite of analyses shall not
exceed three percent. The maximum
story drift ratio in any one analysis shall
not exceed 4.5 percent. In each story,
the mean of the residual drift ratios
from the suite of analyses shall not
exceed one percent. The maximum
residual story drift ratio in any analysis
shall not exceed 1.5 percent.
6.17 Analytical Model
General
The nonlinear analysis model was 5Figure 6.65: 3D View of Model. © Skidmore, Owings
constructed in Perform-3D, which & Merrill LLP
provides a number of modeling
features, including nonlinear models
of beams, columns, braces, walls, and which uses an event-to-event solution
other structural elements. For beams strategy, which, while effective, can
and columns, it is possible to use be slow for large models. Hence,
fiber-based elements, which provide only members that are likely to affect
flexural and axial nonlinear behavior the dynamic response are modeled
and interaction. It is also possible to in Perform-3D. This usually includes
use pre-defined backbone curves with members of the lateral force-resisting
strength and stiffness degradation, system and any additional members with
similar to what is proposed in ASCE important behavior. For the case of the
41-06, and with the possibility of current tower, this included the following:
defining plasticity-based interaction core shear wall and link beams, gravity
envelopes between axial forces and columns, typical floor slabs as equivalent
biaxial moments. For walls, Perform-3D frames, explicit ground-floor slab, explicit
provides the ability to discretize walls basement-floor and sloped-ramp slabs
into concrete and steel fibers, which and basement walls. Figure 6.65 shows
allows modeling of the flexural behavior a 3D view of the analysis model, while
of the walls, as well as the incorporation Figure 6.66 shows the equivalent frames
of axial-flexure interaction. for typical floor slabs.
Nonlinear analysis for large structures Core Wall
needs high computing efficiency, The proposed element for modeling core
particularly when using Perform-3D, wall and wall segments is a four-node
Performance-Based Seismic Design Examples | 87
 “Shear Wall Element” in Perform-3D. This
engineered element incorporates three
“layers:” a vertical layer used for elastic or
inelastic fiber modeling of axial/bending
behavior, using vertical concrete and
steel fibers; a horizontal layer that is
elastic to model elastic transverse
bending behavior; and a “conventional
shear” diagonal layer to model elastic or
inelastic concrete shear behavior. The
axial/bending behavior of the wall and
wall segments is modeled using inelastic
fiber sections (In Perform-3D: Shear Wall,
Inelastic Section) over the entire height
of the core, using the vertical layer of the
shear wall element. Shear is modeled
elastically by assigning an elastic shear
material to the “conventional shear layer”
of the shear wall element. As presented
earlier, the elastic shear wall material
used to model the shear behavior has a
5Figure 6.66: Typical Floor Slab-Equivalent Frame (left) and Ground-Floor Slab Shell Element reduced shear stiffness to reflect shear
Modeling and Typical Basement Slab-Shell Element Modeling (right). © Skidmore, Owings & Merrill LLP cracking; however, its shear strength is
modeled independently from the axial
51'-9" load effect.
1 1
3 2'-4 8 " 1 2'-7 2 "
5'-8" 5' 5'-116 " 3'-11 4 " 6'-10" 5'-7" 5'-3" 5'-2" 4'-3"
The finite shear wall element mesh
for the wall is designed so that it
1
5'-10"
corresponds to the boundary zone
9'-11 8 "
widths and locations in the wall. In other
words, each wall segment is divided
7'-5"
horizontally into a number of shear wall
elements that guarantees that there is
at least one separate element for each
7'-4"
boundary zone, along with at least one
41' 21'-2" 41' element that covers the non-boundary
zone, depending on wall dimension.
A plan view of the core wall fibers is
13'-1"
shown on Figure 6.67.

Wall Axial Strain Measurement

9'-11"
Wall strains can be measured in two
1
7'-4 8 "
ways in Perform-3D. The first is directly
reading fiber strains within the program.
6'-11" 3'-6" 5' 5' 6'-10" 3'-11" 4'-5" 3'-11" 5' 5'-6 1 "
The other method incorporates using
2
1
1'-8 2 " axial strain gages placed at the desired
extreme fibers. The second method is
Steel fiber in boundary zone element Confined concrete fiber in boundary zone element adopted in this study. Monitoring the
Steel fiber in non-boundary zone element Unconfined concrete fiber in non-boundary zone element extreme fiber strains is necessary to

5Figure 6.67: Core Wall Mesh and Fiber Discretization. © Skidmore, Owings & Merrill LLP

88 | Performance-Based Seismic Design Examples

investigate the possibility of exceeding
steel and concrete strain limits.
Link Beams
The link beams are diagonally reinforced
and are categorized into the following
groups, approximated by span-to-depth
ratios 1, 1.5, and 2. Due to the relatively
low aspect ratio of the coupling beams,
all coupling beams are modeled using
elastic beam elements with an inelastic
shear displacement-based translational
spring, denoted “Shear Hinge,
Displacement Type” in Perform-3D.
The shear-rotation curves contain initial
secant stiffness, based on the effective
stiffness for the elastic portion of the
link beam element, and a modified
backbone curve to account for post-
capping strength deterioration and
cyclic deterioration. According to Naish
2010 and ATC 114/ATC 72, the effective
stiffness values shown on Figure 6.68
are used for the elastic portion of the
link beam element on either side of
the shear hinge (spring). The adjacent
effective stiffness values account for
shear cracking and the added flexibility
due to bond slip/extension. Linear
interpolation between the effective
stiffness values may be used.
Column Modeling
Reinforced concrete columns are
modeled inelastically in Perform-3D
using a lumped plasticity model that
is composed of an elastic column
element that accounts for axial load–
biaxial moment interaction (P-M2-M3
column) and a nonlinear rotation-based
2.0 ≤ Ln/h ≤ 4.0
1.1 ≤ Ln/h ≤ 1.4
Ln/h ≤ 1.1
5Figure 6.68: Effective Stiffness Values for Diagonally Reinforced Link Beam (ATC 72). © Skidmore, Owings & Merrill LLP
plastic hinge element (P-M2-M3 Hinge)
that also accounts for axial load–biaxial
moment interaction.
The elastic strength of columns is
calculated using XTRACT fiber sectional
analysis software using the expected
material strength. The backbone curves
of the columns are generated according
to ASCE 41-06, Supplement 1, Table
6.8 recommendations. All columns are
“Condition i” according to ASCE 41.
Cyclic degradation is not modeled in
columns. Acceptance criteria for collapse
prevention limit state is also selected
according to the Table 6.8, ASCE 41-06.
Equivalent Frame Modeling
Since the intention was to model
explicitly the ground floor slab and the
basement floor slabs in Perform-3D,
using shell elements to accurately
capture the back-stay effect, these slabs
were reasonably meshed to a uniform
and well-discretized mesh scheme.
The basement floors ramp slabs were
also explicitly modeled as sloped slabs.
The model of typical floor slabs in
Perform-3D is an equivalent frame. That
is, and in order to help account for the
outrigger and P-Delta effect, the typical
floor post-tensioned slabs were modeled
as equivalent beams (frame elements)
with an effective width equal to the
tributary slab width. The total stiffness
was reduced to 25 percent of the original
stiffness.
Damping Ratio
Estimating the damping ratio of a tall
building undergoing intense ground
Flexure Shear Axial
0.15 EcIg 0.30 EcAv EcAg
0.15 EcIg 0.10 EcAv EcAg
0.10 EcIg 0.07 EcAv EcAg
shaking needs careful consideration.
The overall damping in a structure
is composed of equivalent viscous
damping within the elastic range,
along with the additional effective
damping resulting from intense
inelastic action. The effective damping
resulting from inelastic action is not
modeled in Perform-3D. Instead, the
backbone-constitutive relationships
for the inelastic components
dissipate the input shaking energy
through inelastic deformations.
Well-established research shows that
the equivalent viscous damping of
a tall concrete building is less than
the conventionally used five percent
damping ratio. A three percent modal
damping plus 0.5 percent Rayleigh
damping was used in accordance with
ATC-72. A damping sensitivity study
was performed to evaluate a lower
2.25 percent modal damping plus
0.25 percent Rayleigh damping and
its effects on interstory drift and link
beam rotations. The results indicated
a small increase in interstory drift
(5–8 percent) and a corresponding
small increase in link beam rotations
(3–20 percent).
6.18 Analysis Results
Modal Characteristics
The primary mode periods for the
MCER nonlinear model were 3.60, 3.43,
and 2.76 seconds in the Y direction,
X direction, and torsion, respectively.
Ninety percent mass participation was
reached around the 22nd mode, while
the first translational mode's mass
participation was about 54 percent.
Interstory Drift Ratio
Figure 6.69 depicts the interstory
drift ratio (IDR) resulting from NRHA
under the effect of the 10 MCER
ground motion pairs along with mean
and mean plus/minus one standard
Performance-Based Seismic Design Examples | 89
 35 35
Mean_X-Dir Mean_Y-Dir
Yarimca_X-Dir Yarimca_Y-Dir
LPGC_X-Dir LPGC_Y-Dir
30 Sylmar_X-Dir 30 Sylmar_Y-Dir
Chi101_X-Dir Chi101_Y-Dir
ChiTCU_X-Dir ChiTCU_Y-Dir
DuzDuz_X-Dir DuzDuz_Y-Dir
DuzKog_X-Dir DuzKog_Y-Dir
25 25
Hotville_X-Dir Hotville_Y-Dir
Abbar_X-Dir Abbar_Y-Dir
PS-10_X-Dir PS-10_Y-Dir
Individual CP Limit Individual CP Limit
20 20
Average CP Limit Average CP Limit
Mean +1 Sigma Mean +1 Sigma
Mean -1 Sigma Mean -1 Sigma

Story 15 Story 15

10 10

5 5

0 0
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0

Peak Interstory Drift Ratio % Peak Interstory Drift Ratio %

5Figure 6.69: Drift Response in the X (left) and Y (right) Directions. © Skidmore, Owings & Merrill LLP

deviation. Note that only mean is used reasonable stability in the event of
for acceptance criteria. aftershock ground motions.

Since the mean The maximum Y-direction drift of 1.5 Core-Wall Shear
response of wall percent occurred at level 25, while Since the mean response of wall
the maximum X-direction drift of 1.2 vertical strains is used to evaluate the
vertical strains is percent was at level 12. The dispersion conformance of wall flexural response
used to evaluate the of drift distribution over the building
height was more pronounced in the
to the collapse prevention limit state,
it is considered appropriate from a
conformance of wall Y-direction. In both directions, drift capacity design perspective to use 1.5
over the height conformed to the MCER times the mean to evaluate the shear
flexural response acceptance criteria, with no noticeable response. This approach is considered
to the collapse drift concentration at any single story. in conformance with LATBSDC (2014)
It is worth mentioning that the drift philosophy of shear response as force-
prevention limit for any of the 10 ground motion pairs controlled action that deserves using
state, it is considered did not exceed even the three percent
MCER acceptance criteria for the
higher than the mean response to
guarantee a low probability of collapse.
appropriate from mean response. The dispersion of the The strength reduction factor specified,
X-direction drifts was more significant Φ, is taken as 1.0 as recommended by
a capacity-design than those in the Y-direction. LATBSDC (2014).
perspective to use
The residual drift results show that both Figure 6.70 displays the 1.5 times the
1.5 times the mean X and Y directions for all 10 ground mean demands of the core walls as a
to evaluate the motions satisfy both the mean and
maximum MCER residual drift limits. As
whole over the building height. It is
worth mentioning that the core-wall
shear response. described by PEER/TBI 1.0, this indicates shear reinforcement based on elastic

90 | Performance-Based Seismic Design Examples

 East Wall
30 DE
29
28 MCE Mean
27 MCE Mean + Sigma
26
DBE design did not initially conform to 25 MCE 1.35*Mean
the ACI 318-08 acceptance criteria using 24
23
MCE 1.5*Mean
Capacity
1.5 times the mean demand at many 22
21 8√f'c
locations, including the ground-floor 20
level and at some floors in the upper 19
18
half of the building, due to higher 17
16
mode effects. Accordingly, the shear 15

Story
14
reinforcement was revised at those 13
floors. The shear capacity profile is 12
11
depicted in Figure 6.70, which shows 10
9
general conformance to ACI 318-08 limit. 8
7
6
It is worth mentioning that the flexural 5
4
ductility ratio demand of the core wall 3
2
was generally around 2, which did not 1
trigger reduction of the ACI 318 shear 0
-1
strength according to FEMA 306, due to -2
-3
high flexural ductility demand. 0 2,000 4,000 6,000 8,000 10,000 12,000 14,000 16,000 18,000
Shear Force (kips)
Core-Wall Tensile Strains 5Figure 6.70: Core-Wall Shear Profile. © Skidmore, Owings & Merrill LLP
Figure 6.71 exhibits the core-wall tensile
strain usage ratios relative to the yield
strain. A red color indicates exceedance
of yield strain. Notably, these strain ratios
were calculated based on axial strain
gages placed at the edges of boundary
zones and the interface of boundary/
non-boundary zones, hence the steel
strains are slightly non-conservative
for the outer edges of boundary zones.
It can be observed that the core wall
yielded under the effect of 7 out of 10
ground motions, with a negligible local
yield in the remaining three ground
Yarimca LPGC Sylmar Chi-101 Chi-TCU Duz-Duz Duz-Kog Hotville Abbar PS-10
motions. Plastic hinge generally occurred
at the intended level 5 to level 10 region.
However, the yielding in some ground
motions spread downward towards the
ground-floor level, especially in the south
wall, where large openings exist at the
ground floor.

Figure 6.71 also shows the tensile-

strain usage ratios for the core wall
under the effect of the 10 ground
motion pairs relative to the collapse
prevention acceptance limit, which
is 0.05 tensile strain. The red color Yarimca LPGC Sylmar Chi-101 Chi-TCU Duz-Duz Duz-Kog Hotville Abbar PS-10
indicates exceedance of 30 percent of Minimum usage ratio for each color: 0.0 0.05 0.1 0.2 0.3

5Figure 6.71: Core Tensile-Strain Usage Ratios (Top: Yielding Limit - Red, Bottom: CP
Limit - Red). © Skidmore, Owings & Merrill LLP

Performance-Based Seismic Design Examples | 91

collapse prevention limit state. It can
be observed that the CP limit state is
satisfied with a large margin of safety.
In general, the mean ductility demand
at the typical floors and ground floor
did not exceed 2 and 2.5, respectively.
This indicates the proper wall
proportioning and reinforcement
was provided. No large concentrated
inelastic demand was observed. The
higher-mode effects, combined with
the reduction of wall thickness from
28 inches to 24 inches in the upper
half of the tower, led to slight yielding
higher up in some ground motions at
elevation 225 feet. However, the mean
ductility demand was as low as 1.1 for
this localized effect. The low ductility
demand in the wall permitted the use
of ACI 318-08 shear strength without
further reduction. It is noteworthy
wall tensile-steel strains are quite low, This represents the shear hardening
buckling of longitudinal reinforcement in of the link beams according to the
compression is not likely. input backbone curves. The link beam
yielding took place almost in all link
A plot of the wall strains is shown in beam locations in the north and south
Figure 6.72 at the location of gauge 1, at walls and for most floors at the east
the southwest corner of the core walls. wall. The link-beam shear dispersion
was generally low compared to other
Link Beam Shear response quantities, reflecting the fact
Figure 6.73 shows the link beam shear- that for any ground motions that cause
stress coefficient (defined as shear stress link-beam yield in shear, the shear value
divided by the square root of expected will be the same following the input
concrete strength) distribution over backbone curve.
the building height for the most
critical northwest corner link beam. It is noteworthy that the estimation of
The mean shear demand was higher core-wall shear demand and link-beam
than the shear capacity by a factor of shear demand at the MCER inelastic
25–33 percent for the yielding beams. level, by amplifying the shear demand
400 400
YARIMCA Compression YARIMCA Tension

that in response to the concentrated 375 LPGC Compression LPGC Tension 375

inelastic strains at ground-floor level, 350

SYLMAR Compression SYLMAR Tension
350
CHICHI Compression CHICHI Tension
although they were marginal, wall
ChiTCU Compression ChiTCU Tension
vertical reinforcement at this level was 325
DuzDuz Compression DuzDuz Tension
325

revised to reduce strain concentration. 300 DuzKog Compression DuzKog Tension 300
Hotville Compression Hotville Tension

Core-Wall Compression Strains 275 Abbar Compression Abbar Tension 275

The compression-strain usage PS-10 Compression PS-10 Tension

250 250
ratios relative to the CP limit state of Mean Compression Mean Tension

unconfined concrete strain (0.002) and 225

Mean+1 Compression Mean+1 Tension
225
Mean-1 Compression Mean-1 Tension
confined concrete strain (0.012) for
200 200
the 10 ground motion records were Height (ft) Height (ft)
on the order of 40–50 percent for 175 175

unconfined concrete and 10–20 percent

150 150
for confined concrete. Perform-3D
tends to significantly underestimate 125 125

compression strains; therefore, a 40–50

100 100
percent reduction of the allowable
confined and confined concrete strains 75 75

from material models was applied,

hence the values 0.012 and 0.002 50 50

above. It is also worth mentioning 25 25

that the compression strains can be
conservatively used to judge yielding 0
-0.4% -0.2% 0.0% 0.0% 0.2% 0.4% 0.6% 0.8% 1.0% 1.2% 1.4%
0
1.6%

of longitudinal steel in compression. -25 -25

The mean compression strain did not
reach steel yield strain for any of the wall -50 -50
South Wall Loc 1 Compression Strain Profile South Wall Loc 1 Tensile Strain Profile
locations. Since the ductility demand of
5Figure 6.72: Shear Wall Vertical Strain Profile. © Skidmore, Owings & Merrill LLP

92 | Performance-Based Seismic Design Examples

of the elastic DE model by 1.5, could
underestimate the actual shear
demands at MCER. An MCER base shear
of 2.0–2.5 times the DE shear was
observed in this study. The significant
shear yielding of link beams reveals
that they attracted much more force
than the 1.5 amplification originally
anticipated in the DE elastic model.

Link-Beam Plastic Chord Rotations

Figure 6.74 exhibits the link-beam
rotation-usage ratios for the 10 ground
motion pairs relative to the collapse-
prevention rotation limit, as described
in the acceptance criteria. The red 5Figure 6.73: Link Beam Shear Stress (in Av√f’c). © Skidmore, Owings & Merrill LLP
color in the plot refers to exceedance
of CP plastic rotation. The link-beam
plastic chord rotations exceeded the
CP limit for the northwest beam in
three ground motions, while those of
southeast beam exceeded the CP limit
in two ground motions. It is crucial to
stress that the ground motions that
caused large demands of these link
beam rotations were counter-intuitively
oriented closer to the north–south
direction (Y-direction), especially the
most critical ground motion, Yarimca.
This can be explained as follows: The
uplift (or compression) forces on the
tension side (or compression flange)
Yarimca LPGC Sylmar Chi-101 Chi-TCU Duz-Duz Duz-Kog Hotville Abbar PS-10
of the wall under a Y-direction ground
motion needs to be transferred
between the different segments of the
compression or tension flange. The
compression or tension flanges here
are the north and south walls. Hence,
the large tension or compression
demand transfer from the exterior
north or south wall pier to the interior
wall pier occurs through shearing the
link beams. This effect is difficult to be
observed through two-dimensional
analysis of core walls.

Figure 6.75 depicts the link-beam plastic

chord rotation distribution over building Yarimca LPGC Sylmar Chi-101 Chi-TCU Duz-Duz Duz-Kog Hotville Abbar PS-10
height for the 10 ground motion Minimum usage ratio for each color: 0.0 0.25 0.5 0.75 1
records, and the mean and mean plus/ 5Figure 6.74: Link Beam Chord-Rotation Usage Ratios to CP Limit State (Top: South Wall, Bottom: North Wall).
© Skidmore, Owings & Merrill LLP

Performance-Based Seismic Design Examples | 93

 minus one standard deviation. The percent. This indicates that the most
peak mean plastic chord rotation for critical mean plastic chord rotation
link beams was around one percent conforms to the collapse prevention
for most link beam locations. However, limit state. Furthermore, the mean-
for the most critical northwest and plus-one-standard-deviation of the
southeast locations, the peak mean above-mentioned critical link beams
link-beam rotation was about two was less than four percent. This indicates
the low probability of collapse for the
most critical link beam. The dispersion
of the link beam distribution over the
building height was generally acceptable,
with a more pronounced dispersion for
In addition to the the northwest and southeast link beam
confinement provided locations. The two most critical ground
motions for link beam rotations were
in all boundary zones, Yarimca-Kogali and ChiChiY101.
additional non-
boundary cross-ties 6.19 Project-Specific Designs
and Detailing
are provided in the
non-boundary zone Link Beam Detailing
To ensure ductile performance and
areas of shear walls accommodate construction methods,
in the primary plastic innovative reinforcement detailing
methods were utilized. These detailing
hinge zone above practices are in addition to the minimum
provisions of the governing codes,
ground level. but assist in assuring performance
of the different components of the
390
377
364
Mean_-ve Rot
seismic system to improve construction
efficiencies (see Figure 6.76).
ACI 318-08 Section 21.9.7 permits the
confinement of the entire link beam
section, as well as traditional confinement
of only the diagonal reinforcement. This
alternate confinement option is done
to alleviate the constructability issues
encountered when confining just the
diagonal reinforcement. Confinement is
provided by overall hoops and vertical
and horizontal cross-ties. Cross-ties are
to “engage” the hoop, side and top and
bottom reinforcement, according to
ACI 318. To improve constructability,
the 90° hook of the cross-tie is replaced
with a “T”-head end anchorage (ASTM
A970-13a). The anchor is to be nine times
the area of the bar and tied directly to the
intersection of the hoop and side/top/
bottom bar, such that the large flat T-head
overlaps and engages the perimeter
reinforcement. Since the 90° hook of
a cross-tie is typically the first rebar to
cause spalling, its replacement with a
T-head anchor is an improvement.
Non-Boundary Zone Ties
In addition to the confinement provided
in all boundary zones, additional
non-boundary cross-ties were provided
390 in the non-boundary zone areas of
377

PS-10_-ve Rot
Mean_+ve Rot

PS-10_+ve Rot
364 shear walls in the primary plastic
351 351
338
Yarimca_-ve Rot Yarimca_+ve Rot
338 hinge zone above ground level. This
325
312
LPGC_-ve Rot

Sylmar_-ve Rot
LPGC_+ve Rot

Sylmar_+ve Rot
325
312
additional non-boundary confinement
299
286
Chi101_-ve Rot Chi101_+ve Rot 299
286
was provided through cross-ties evenly
273
ChiTCU_-ve Rot

DuzDuz_-ve Rot
ChiTCU_+ve Rot

DuzDuz_+ve Rot
273 distributed vertically and horizontally
260 260
247 DuzKog_-ve Rot DuzKog_+ve Rot 247 in non-boundary areas between the
234 234
Height
221
Hotville_-ve Rot

Abbar_-ve Rot
Hotville_+ve Rot

Abbar_+ve Rot
221
Height
foundation and level 10. Although
(ft) 208
195 Mean+1 sigma Mean+1 sigma
208 (ft)
195
this reinforcement was not specifically
182
169
Mean-1 sigma Mean-1 sigma 182
169
accounted for in the design, the detailing
156 156 measure will help avoid vertical rebar
143 143
130 130 buckling in the plastic hinge zone, and
117 117
104 104 improve resistance to shear demand
91 91
78 78 under high overturning conditions.
65 65
52 52
39 39
26 26
13 13
0 0
-8 -7 -6 -5 -4 -3 -2 -1 0 1-1 0 1 2 3 4 5 6 7 8

5Figure 6.75: Link Beam Rotation Profile. © Skidmore, Owings & Merrill LLP

94 | Performance-Based Seismic Design Examples

 USE APPROVED END ANCHOR WHEN
TOP BARS INSUFFICIENT SPACE TO PROVIDE
SEE SCHEDULE 1.25 TIMES BAR DEVELOPMENT
LENGTH
DIAGONAL REINF. COLUMN
VERTICAL CROSSTIES,
SEE SCHEDULE
SEE PLANS FOR SLAB CONCRETE BEAM STIRRUPS B
THICKNESS AND SHEAR WALL SEE SCHEDULE
REINFORCEMENT CL BEAM HORIZONTAL CROSSTIES,
SEE SCHEDULE TOP BARS
SEE SCHEDULE
T/SLAB T/SLAB
EL: SEE PLANS EL: SEE PLAN
"T"
SEE SCHEDULE

SIDE BARS, SEE SCHEDULE

DEPTH

1 1/2" CLEAR
STIRRUPS SEE SCHEDULE FOR ENT SIDE BARS
MIN. LOPM
SIZE AND TYPE DEVE SEE SCHEDULE
IMES
1.25 T LENG
TH
S/2 S
BOTTOM BARS, DIAGONAL BARS,
VERTICALLY, HORIZONTALLY, AND 2" MAX
SEE SCHEDULE SEE SCHEDULE
LONGITUDINALLY CROSS TIES STIRRUP SPACING BOTTOM BARS SEE
SHALL ALTERNATE 135 DEGREE SEE SCHEDULE SCHEDULE
HOOK AND T-HEAD TERMINATOR.
SEE NOTE 1. NOTES:
1. T-HEADED TERMINATOR SHALL BE
DIAGONAL REINF. ROW HRC-120 (SQUARE HEADED).
WIDTH 2. T-HEAD TO BE LOCATED OUTSIDE 6" LINK BEAM LENGTH 6"
DIAGONAL BARS, OF STIRRUP AT THE INTERSECTION
SEE SCHEDULE MATCH THICKNESS SEE ELEVATION
OF SHEAR WALL OF STIRRUP AND LONGITUDINAL
26 BAR.
3. T-HEAD SHOULD BE SECURED TO
SECTION ELEVATION
B ENSURE THE T-HEAD ENGAGES THE
STIRRUP AND LONGITUDINAL BAR. A

25A TYPICAL REINFORCED CONCRETE LINK BEAM DETAILS (ALTERNATE: T-HEAD CROSS TIES)
SCALE: NONE

USE APPROVED END ANCHOR WHEN

TOP BARS INSUFFICIENT SPACE TO PROVIDE
SEE SCHEDULE 1.25 TIMES BAR DEVELOPMENT
LENGTH
DIAGONAL REINF. COLUMN
VERTICAL CROSSTIES,
SEE SCHEDULE
SEE PLANS FOR SLAB CONCRETE BEAM STIRRUPS B
THICKNESS AND SHEAR WALL SEE SCHEDULE
REINFORCEMENT CL BEAM HORIZONTAL CROSSTIES,
SEE SCHEDULE TOP BARS
SEE SCHEDULE
T/SLAB T/SLAB
EL: SEE PLANS EL: SEE PLAN
"T"
SEE SCHEDULE

SIDE BARS, SEE SCHEDULE

DEPTH

1 1/2" CLEAR
STIRRUPS SEE SCHEDULE FOR T
MIN. PMEN SIDE BARS
SIZE AND TYPE VELO SEE SCHEDULE
S DE
TIME TH
1.25 LENG S/2 S
BOTTOM BARS, DIAGONAL BARS,
2" MAX SEE SCHEDULE
CONSECUTIVE CROSS TIES SHALL SEE SCHEDULE
STIRRUP SPACING BOTTOM BARS SEE
HAVE THEIR 90° HOOKS AT SEE SCHEDULE SCHEDULE
OPPOSITE SIDES OF LINK BEAM

DIAGONAL REINF. ROW

WIDTH 6" LINK BEAM LENGTH 6"
DIAGONAL BARS,
SEE SCHEDULE MATCH THICKNESS SEE ELEVATION
OF SHEAR WALL

SECTION ELEVATION
B A

25 TYPICAL REINFORCED CONCRETE LINK BEAM DETAILS

SCALE: NONE

5Figure 6.76: Link Beam Detailing (Top Left: Construction Photo, Top Right: Example T-Head Cross-Tie, Bottom: Typical Detail). © Skidmore, Owings & Merrill LLP

Performance-Based Seismic Design Examples | 95

Performance-Based Seismic Design Examples
Design Example 4: Tall Residential Tower with Podium

118 ft

PT Flat Slab

52 ft
RC Gravity
Column

E RC Link Beam
S
RC Core Wall

33 ft
91 ft
E E
S
E 31 ft

27 ft
5Figure 6.77: Rendering. © Skidmore, Owings 5Figure 6.78: Typical Floor Plan. © Skidmore, Owings & Merrill LLP
& Merrill LLP

6.20 Project Description
The following is a structural system
description for a residential tower
situated in downtown San Francisco,
California (see Figure 6.77). The
occupancy consists of apartment
residential units above grade, and
parking below grade. The 42-story
tower is 420’–0” tall above grade with
six basement levels below grade and
a total building area of approximately
743,500 square feet. The seismic force-
resisting structural system consists of
reinforced concrete core walls from the
foundation to roof. The gravity system is
a long-span, flat-plate, post-tensioned
system. This combination of lateral and
gravity systems is common in high-rise
residential construction.
Superstructure
The superstructure consists of a
central reinforced concrete shear wall
core, perimeter gravity columns, and
two-way flat-plate slab framing (see
Figure 6.77). The tower is roughly square
in plan with dimensions of 91’–0”
x 118’–0” and a larger podium. The
typical residential floor-to-floor height
is 9’–3” (see Figure 6.78). The podium is
exceptionally large for similar buildings
in San Francisco at nine floors, and plays
a significant role in building’s response
to lateral loading.
With the removal of the code-
prescribed moment frame, the
improvement of the typical floor
section is demonstrated in Figures 6.79
and 6.80. By removing the moment
frame, inefficient material is removed,
floor-to-floor height is reduced, and
constructability is improved.
Lateral System
The lateral system consists of a centrally-
located reinforced concrete shear wall
core. The shear wall core has an external
plan area of 33’–0” x 52’–0” and is located
around the service area of the structure,
passenger and service elevators, and
back-of-house areas. The height-to-
depth ratio of the core is 12.1 to 1. This is
a very slender application of a core-only
lateral system approach. The shear wall
core extends from foundation to roof.
At podium levels, additional shear walls
are included to assist the core, due to
the added mass and eccentricity. The
shear walls vary in thickness from 36”
to 24”, and in concrete compressive
strength of 8,000 psi at the core to
6,000 psi at the podium. The shear wall
core is interconnected with the use of
ductile diagonally-reinforced link beams
at openings required for doorways
and corridors.
Core shear wall elevations are shown in
Figure 6.81, where the darker shading
indicates confined boundary zones.
Gravity System
The gravity framing system both inside
and outside the core consists of a two-
way post-tensioned (PT) flat-plate slab.
The slab clear-spans from the core to the

96 | Performance-Based Seismic Design Examples

 5’-0”

ELEC./TEL. CORRIDOR BATH CL BEDROOM

ROOM 1” FINISH

5’-0”

ELEC./TEL. CORRIDOR BATH CL BEDROOM

ROOM 1” FINISH

5Figure 6.79: PBSD Core-Only Approach. © Skidmore, Owings & Merrill LLP

5’-0”

ELEC./TEL. CORRIDOR BATH CL BEDROOM

ROOM 1” FINISH

5Figure 6.80: Code-Prescribed Dual System Approach. © Skidmore, Owings & Merrill LLP

BLOCK 9 TRANSBAY, LLC.

A JOINT VENTURE OF:

perimeter and has a uniform thickness

of seven inches. The long-span5’-0” flat slab Architects:
500 FOLSOM
at Folsom Street and First Street,
San Francisco, CA 94105

is post-tensioned to maintain
ELEC./TEL.
uncracked
CORRIDOR BATH CL BEDROOM
section under service gravity loads. The
SKIDMORE, OWINGS & MERRILL LLP
ONE FRONT STREET, SAN FRANCISCO, CA 94111

ROOM 1” FINISH

slab utilizes concrete with a compressive Fougeron Architecture

521 Francisco Street
San Francisco, CA 94133

strength of 6,000 psi.

p: 415.641.5744
f: 415.282.6434

Structural Engineer:

Substructure SKIDMORE, OWINGS & MERRILL LLP

ONE FRONT STREET, SAN FRANCISCO, CA 94111

The vertical elements of the Consultants:

superstructure continue down through

the substructure to the foundation.
Seal & Signature:

The gravity system in the substructure Issued For:

consists of a 10” thick, conventionally

No.: Description: Date:

reinforced, two-way flat-plate slab that

utilizes concrete with a compressive
strength of 6,000 psi.

Foundations
The foundation system consists of a
10’–0” thick conventional reinforced
concrete mat foundation. A perimeter- Key Plan:

CLEMENTINA ST

reinforced concrete foundation wall

ECKER ST

FIRST ST

system consists of conventional NORTH

FOLSOM ST

cast-in-place concrete walls. The

Sheet Name:

SHEAR WALL
ELEVATIONS

substrate consists of dense sands over Project No.: Sheet No.:

3/28/2016 6:49:43 PM

213094

Z_PPT_CoreShearWalls
Drawn By: Author

pre-consolidated clay. Due to soils that

Checked By: Checker
Scale:
1/16" = 1'-0" c Copyright Skidmore, Owings & Merrill LLP 2015

5Figure 6.81: Shear Wall Elevations: West, North, East, and South (from left to right). © Skidmore, Owings & Merrill LLP

Performance-Based Seismic Design Examples | 97

could potentially “liquefy” in a seismic
event, jet-grouted soil improvement
columns were specified by the
geotechnical engineer.
6.21 Design Criteria
The building was designed under the
San Francisco Building Code SFBC 2013,
which refers to the California Building
Code (2013) and ASCE 7 (2010). The
CBC is generally adopted from IBC
publications, with amendments specific
to California.
The tower in this study takes exceptions
to the prescriptive code limits of ASCE
7; specifically, the building height
requirement by CBC 2010 and ASCE
7-10 for Special Reinforced Concrete
Shear Walls under Bearing Wall Systems
(ASCE 7-10 Table 12.2-1). This exception
to the prescriptive code limits and
design procedures requires that the
structural system be classified as a
non-prescriptive seismic force-resisting
The design team
elected to use spectral
scaling as opposed
to spectral matching,
as spectral matching
can significantly
reduce dispersion in
the ground motion
demands, unless
special provisions in
the matching process
are considered.
98 | Performance-Based Seismic Design Examples
system. The building is required to
meet the seismic performance intent
of the building codes, and is required
to utilize San Francisco's Building Code
Administrative Bulletin AB-083 and
AB-082 for seismic design.
The three-step procedure as outlined by
AB-083 is described in further detail in
Example 3 (see page 84). For this project,
an additional step was included at the
onset of design using linear analysis
and MCER-level response-spectrum
demands. Subsequently, it was
evaluated for SLE and DE demands using
linear analysis and MCER using NRHA.
Step One: MCER-Level Linear Design
Based on issues the design team
encountered on prior projects during
MCER nonlinear evaluation using DE
and SLE-level response-spectrum
demands, a new method was utilized.
The design criteria used for the linear
MCER evaluation does not incorporate
the code-prescribed provisions of R, Ωo,
ρ, and Cd. Instead, the R-factors were
selected based on desired performance.
An R of 2.0 is used in the design of
force-controlled actions, such as shear
in shear walls and transfer diaphragms.
For deformation-controlled-action, two
R values are considered. For elements
where significant energy dissipation is
desired, such as flexure in link beams
and axial/flexure in shear walls at the
plastic hinge zone, an R-value of 7.5
is utilized. For elements with modest
energy dissipation, such as axial/flexure
in shear walls away from the plastic
hinge zone an R of 3.0 is utilized. The
intent is to create a focused plastic hinge
that can concentrate energy dissipation
and reduce demands elsewhere in the
structure. Using an R-value greater than
5.0 can result in elements with capacity
less than DE-level code requirements,
but this is listed as a code exemption
in the design criteria. This is considered
appropriate, since both strain levels are
explicitly monitored in the shear walls
and a push-over analysis is conducted
to verify overall building performance.
Expected material properties are utilized.
Step Two: Code-Level Seismic Evaluation
The DE-level evaluation is used to
verify the minimum required strength
and stiffness for earthquake resistance
under DE-level demands, as prescribed
by the building code corresponding
to life safety performance (see Section
5.6). The design incorporates the
prescriptive provisions of R, Ωo, ρ, and Cd
of ASCE 7-10 into a linear elastic modal
response-spectrum analysis. Specified
material properties are utilized. This step
is required in AB-083.
Steps three and four match steps
one and two of the typical procedure
described in Chapter 5 (see page 46).
Ground Motion Selection
The MCER-level evaluation incorporates
two target response spectra, using
the conditional mean spectrum
(CMS) approach. One CMS response
spectrum is identified for lower periods
dominating building response, and
the second CMS response spectrum is
identified for higher periods, which can
significantly influence building behavior
(see Figure 6.82). Eleven time-history
pairs of ground motions are selected
and spectrally scaled, such that their
spectral average approximates each
target CMS. Thus, a total of 22 ground
motion pairs were identified for
evaluation of MCER performance.
The design team elected to use spectral
scaling as opposed to spectral matching,
as spectral matching can significantly
reduce dispersion in the ground motion
demands, unless special provisions in the
matching process are considered.
A common difficulty of spectral
scaling is unrealistically high spectral
responses away from the target period
of interest in some ground motions.
 2
A-LOLETA
1.8 B-YARIMCA
C-TCU054
D-YERMO
1.6 E-TCU015
F-TCU039
G-PS10
1.4
H-BOTANICAL GARDENS
I-SHIRLEY LIBRARY
1.2 J-WESTSIDE ELEM
K-CALEXICO
MCE
1
CMS-LP
AVERAGE ROTD100
0.8

0.6

0.4

0.2

0
0 1 2 3 4 5 6 7 8 9 10

2
A-CHICHI36
B-DUZCE
1.8
C-LOMAP
D-LANDERS
1.6 E-CHY024
F-TCU065
G-DENALI
1.4 H-CAMPEMEND FFS
I-CAMPEMEND LFS
1.2 J-DARFIELD CNH
K-DARFIELD HORC
MCE
1 CMS-SP
AVERAGE ROTD100
0.8

0.6

0.4

0.2

0
0 1 2 3 4 5 6 7 8 9 10

5Figure 6.82: Spectral Demands of Short and Long Period Suites Compared to MCER and CMS Spectra. 5Figure 6.83: 3D View of P3D Model. © Skidmore,
© Skidmore, Owings & Merrill LLP Owings & Merrill LLP

For example, for ground motions
selected based on long-period CMS
target spectrum, it is likely that
some ground motions may have
unrealistically high spectral demands
at shorter periods and vice-versa. For
this project, that was true for the short
period suite, where some ground
motions had high spectral responses
at long periods.
The ground motions were randomly
rotated and evaluated for potential bias.
6.22 Analytical Model
General
For evaluation of MCER demands
using NRHA, Perform-3D was utilized.
All components of the lateral
force-resisting system were explicitly
modeled, including shear walls, link
beams, transfer diaphragms, and
foundation walls. All components
of the gravity system were modeled
as well, with all columns explicitly
modeled and non-transfer diaphragms
represented with equivalent frames.
The mat foundation and supporting
soil were not represented in the
baseline model, but were included in
the upper bound study.
All shear walls were modeled
using representative fibers for axial
deformation of reinforcement, confined
concrete, and unconfined concrete
using the “Shear Wall Element” of
Perform-3D. Shear of walls was
modeled elastically. All link beams and

Performance-Based Seismic Design Examples | 99

slab-equivalent frames were modeled
with a lumped plasticity. All other
elements including columns, transfer
diaphragms, basement walls, and mat
foundation were elastically modeled,
with appropriate modification to
stiffness to account for anticipated
levels of cracking. Figure 6.83 shows a
3D view of the analysis model, while
Figure 6.84 shows the equivalent
frames for typical floor slabs, the level
9 transfer diaphragm, and the level 1
transfer diaphragm.
Mass Distribution
Since linear elastic analysis was
conducted in ETABS, its ability to
account for mass was explicitly
accommodated through element
properties, and applied surface loads
5Figure 6.84: Typical Tower Floor Slab Equivalent
Frame Modeling (top), Podium Slab Shell Element
Modeling (middle) and Basement Slab Shell Element
Modeling (bottom). © Skidmore, Owings & Merrill LLP
100 | Performance-Based Seismic Design Examples
were utilized for the accounting of
seismic mass. When the ETABS model
is exported to Perform-3D, it is desired
to remove elastic shell elements at
all floors except transfer diaphragms
to improve computational efficiency.
Thus, mass needs to be lumped at
nodes prior to exporting to Perform-
3D. This is done as represented in
Figure 6.85. Since mass is included
at the perimeter, torsion and P-Delta
effects are sufficiently included. Other
techniques utilize a “leaning column”
approach to effectively achieve this
effect and remove some column and
equivalent frame elements for further
computational efficiency. Design
engineers for this project elected to
be more explicit through the inclusion
of all columns and representative
equivalent frames.
5Figure 6.85: Nodal Mass Distribution on Tower Floors. © Skidmore, Owings & Merrill LLP
Core Walls
The core is meshed with elements
representing confined and unconfined
concrete separately. Within each
finite element, two or four fibers are
defined for steel and concrete. Thus, a
single finite element has four or eight
fibers. These fibers are represented
in Figure 6.86. For the monitoring of
axial deformations fiber results can be
used, but they do not represent the
most extreme strain levels, as they
are interior to the shell element. Thus,
additional strain monitoring elements,
termed “strain gages,” are added at the
geometric ends of all boundary zones.
Additionally, at the four corners of the
core, they are placed at the true face
of concrete through the use of stiff
horizontal frame elements located the
face of concrete, as shown in Figure 6.86.
 Frame Element Axial ‘Strain Pinned Frame Analysis Model at
Extension to Capture Gage’ Element to Monitor Centerline of Wall
Strain at End of Wall Total Roation

5Figure 6.86: Core Wall Mesh and Fiber Discretization (left) and Monitoring of Deformations (right). © Skidmore, Owings & Merrill LLP

1.6
PT Slab
1.4 RC Slab
1.2 No Slab
ASCE 41 mod
1 for slip/ext
V/V Node

0.8

0.6

0.4

0.2

0
0      2      4      6      8     10     12     14
Beam Chord Rotation [%]

5Figure 6.87: Backbone Curve Proposed from Testing by Naish et al. (2013) for 5Figure 6.88: Slab Frame Elements at Tower Levels (Red = Moment-Rotation
Various Link Beam and Slab Arrangements. © Skidmore, Owings & Merrill LLP Hinges). © Skidmore, Owings & Merrill LLP

Link Beams
All link beams are diagonally reinforced
and modeled with lumped plasticity
based on physical testing results,
typically the results of testing by Naish et
al. (2013). In the dissertation appendix,
specific Perform-3D modeling inputs are
provided by the authors. Naish published
test results with hysteretic behavior of
link beams, along with adjoining RC and
PT slabs (see Figure 6.87). The results
corresponding to the link beam with the
effect of RC slab results were utilized, as
they were similar to the PT.
Total chord rotations are obtained
using a rotational monitoring
element, analogous to the strain gage
mentioned above, as plastic and total
rotations were of interest to the design
team and peer review panel.
Equivalent Frame Modeling
Equivalent frame modeling as
described in ASCE 41 is conducted
with moment-rotation hinges that
allow for different positive and
negative stiffness and strength
characteristics. This robust approach
was utilized to ensure that particularly
thin flat-slab performance was
represented (see Figure 6.88). Only
equivalent frames spanning between
the core and perimeter columns were
modeled with plastic hinges, as it
was not anticipated that perimeter
equivalent frames would experience
inelastic behavior. This was verified
after NRHA.
Damping Ratio
The overall damping of a structure
is composed of equivalent viscous

Performance-Based Seismic Design Examples | 101

damping within the elastic range, along
with the additional effective damping
resulting from inelastic action. For
elastic equivalent viscous damping, a
two percent modal damping plus 0.5
percent Rayleigh damping was used in
accordance with ATC-72. Based on prior
design experience, this is a typical level
of damping for these types of structures.
Design Quantities
It can be beneficial to the design
team and peer review panel to
create summarized plots describing
reinforcement and strength over the
building height. Two such plots that
help describe certain key behaviors of
this building are described below.
First, the total amount of vertical
reinforcement in the core and podium
walls divided by the area over the
building height is shown in Figure
6.89. This gives a global picture of
vertical shear wall reinforcement. As
can be observed, the reinforcement is
lower near level nine to encourage the
development of a ductile plastic hinge.
A plastic hinge at the ground floor
was desired and explored with rebar
adjustments using NRHA, but one could
not be formed at the ground level.
Also, reinforcement increases near the
top to limit strain levels in very high-
magnitude events, well above MCER,
which caused high link beam rotations
and increased shear wall demands.

All Shear Walls

LINK BEAM SHEAR STRENGTH PER f'c ‐ Vpr/(bh√f’c)
45
44 L45 B1
CORE L44
43
42 B2
41 TOTAL L43
40 L42 B3
39 L41
38 L40 B4
37 L39
L38 B5
36 L37
35 L36 B6
34 L35
33 L34 B7
32 L33
31 L32
30 L31
29 L30
28 L29
27 L28
26 L27
25 L26
L25
24 L24
23 L23
22 L22
21 L21
Floor

20 L20
19 L19
18 L18
17 L17
16 L16
15 L15
14 L14
L13
13
L12
12 L11
11 L10
10 Hinge Zone L09
9 at Top of L08
8 L07
7 Podium L06
6 L05
5 L04
4 L03
3 L02
2
L01
1
0 B01
-1 B02
-2 B03
-3 B04
-4 B05
-5 B06
0.00% 0.50% 1.00% 1.50% 2.00%
0 2 4 6 8 10 12 14
Area Rebar (%) Strength

5Figure 6.89: Ratio of All Vertical Reinforcement in All Walls to Concrete Area. © 5Figure 6.90: Link-Beam Shear Strength Normalized to f’c (expected
Skidmore, Owings & Merrill LLP material properties). © Skidmore, Owings & Merrill LLP

102 | Performance-Based Seismic Design Examples

Second, link-beam shear strength is
plotted for each bay of link beams in
Figure 6.90. The intent is to verify a
generally consistent strength over the
building height, which can encourage
a more consistent behavior. The change
in link beam strength of link beams B3
& B4 may have contributed to a jump
in link beam rotations near level 9. B3
is the link beam on the east side of the
core and B4 is the link beam on the
west side of the core. These link beams
experience the highest rotations, as the
north–south direction of the core is the
direction of least stiffness.
6.23 Analysis Results
For the following analyses, the results
of long-period ground motions will
be provided, as they are the most
representative of the long-period
demands and corresponding
performance. In the following
results, link beam rotation, strain
deformations, and story drift are
described for key components and
locations in the building.
Link Beam Rotations
As stated previously, the link beams
on the east and west side of the core
(B3 & B4 in Figure 6.91) experience
the highest rotational demands. Since
strength degradation, residual strength,
and strength loss beyond 10 percent
rotation are included in the nonlinear
component modeling, behavior at
these high rotations is reasonably
captured in the analysis. In particular,
the redistribution of force resistance
from link beams experiencing strength
loss to shear wall pier can be observed
in the strain results.
As can be observed, link beam
rotations reduce near the top of the
tower. This is due to the large panel
zones at the top of the bay of link
beams, as shown in Figure 6.91.
Shear Wall Strain Demands
Most piers do not experience
particularly high axial compression or
tension strains, except the longer east
and west wall piers. Although relatively
high strains occur at the corners of
the core (see Figure 6.92), the highest
strain levels occur at adjoining link
beams B3 and B4. Thus, the degree
of coupling provided by link beams
is only moderate. This is reasonably
common in residential buildings, as
their low floor-to-floor heights typically
only accommodate relatively shallow
link beams. Conversely, office buildings
with higher floor-to-floor sections and
deeper link beams have a higher degree
of coupling, and thus higher strain
demands at the four corners of the core.
For most conditions around the core,
compressive strains are less than the
unconfined limit, which is taken as
0.003/2 = 0.0015. Although it is well
understood that concrete has an
unconfined compressive strain limit
of 0.003, research by Wallace (2007)
demonstrates that Perform-3D may
underestimate compressive strain by
as much as 50 percent in some cases.
Thus, it was agreed between the design
team and peer review panel that a
compressive limit of 0.0015 be utilized,
with compressive strains amplified by
1.5. The 1.5 factor is incorporated to
ensure that a non-ductile condition,
such as unconfined concrete, is not
exceeded. Special confinement as
prescribed by ACI 318-14 is provided
for all boundary zones near the plastic
hinge area above and below level nine,
as defined by ACI 318. Additionally, the
boundary zone adjoining link beams
B3 and B4 are specially confined over
the entire height. As noted above,
link beam rotations reduce near the
top of the tower; thus, it is expected
and observed that shear wall strain
demands near the top will increase.
This behavior is observed in the east
and west walls, in which yielding
An important
consideration for
tall buildings is the
design of the solid
area of concrete
above and below
stacks of openings.
of reinforcement is the dominant
expectation. Vertical boundary-zone
reinforcement was increased in this
area to limit yielding to be two times
the yield strength that the design
team considers “moderate” yielding.
As previously mentioned, this
behavior could not have been directly
accounted for in traditional response-
spectrum design.
A plastic moderate hinge is developed
at level nine, which is the top of the
podium. This was partially successful
due to the increased “R”-value used
at the plastic hinge zone above and
below level nine. It was not entirely
successful since higher mode effects
still cause for yielding between x1 and
x2 yield in areas above level nine.
Additionally, compressive strain
demands were reviewed to determine
where special and ordinary boundary
zones could be utilized; tensile stains
were reviewed to determine where
ties corresponding to potential bar
buckling needed to be considered (See
Figure 6.93).
These additional provisions would not
have been identified by code-based
prescriptive design and will significantly
improve building performance in an
actual seismic event.
Performance-Based Seismic Design Examples | 103
 B3

B4

B3 ‐LP B3‐SP
L45 L45
L44 L44
L43 L43
L42 L42
L41 L41
L40 L40
L39 L39
L38 L38
L37 L37
L36 L36
L35 L35
L34 L34
L33 L33
L32 L32
L31 A ‐Capemend L31 A ‐ CHICHI36
L30 L30
L29 B ‐Kocaeli L29 B ‐ DUZCE
L28 L28
L27 C‐TCU054 L27 C ‐ LOMAP
L26 L26
L25 L25
L24 D ‐Landers L24 D ‐ LANDERS
L23 L23
L22 E ‐ TCU015 L22 E ‐ CHYO24
ELEVATION

ELEVATION
L21 L21
L20 F ‐ TCU039 L20 F ‐ TCU065
L19 L19
L18 G ‐ Denali L18
L17 L17
G ‐ DENALI
L16 L16
L15 H ‐ Darfield CBG L15 H ‐ CAPEMEND FFS
L14 L14
L13 I ‐ Darfield SHLC L13 I ‐ CAPEMEND LFS
L12 L12
L11
L10
J ‐ Sierra.Mex CIW L11
L10
J ‐ DARFIELD CNH
L09 L09
L08
K ‐ Sierra.Mex CXO L08
K ‐ DARFIELD HORC
L07 L07
L06 Applicable Limit L06 Applicable Limit
L05 L05
L04 Mean L04 Mean
L03 L03
L02 L02
L01 L01
B01 B01
B02 B02
B03 B03
B04 B04
B05 B05
B06 B06

‐13% ‐11% ‐9% ‐7% ‐5% ‐3% ‐1% 1% 3% 5% 7% 9% 11% 13% ‐13% ‐11% ‐9% ‐7% ‐5% ‐3% ‐1% 1% 3% 5% 7% 9% 11% 13%
TOTAL LINK BEAM ROTATIONS (RADIANS) TOTAL LINK BEAM ROTATIONS (RADIANS)

B4 ‐LP B4 ‐SP
L45 L45
L44 L44
L43 L43
L42 L42
L41 L41
L40 L40
L39 L39
L38 L38
L37 L37
L36 L36
L35 L35
L34 L34
L33 L33
L32 L32
L31 A ‐Capemend L31 A ‐ CHICHI36
L30 L30
L29 B ‐Kocaeli L29 B ‐ DUZCE
L28 L28
L27 C‐TCU054 L27 C ‐ LOMAP
L26 L26
L25 L25
L24 D ‐Landers L24 D ‐ LANDERS
L23 L23
L22 E ‐ TCU015 E ‐ CHYO24
ELEVATION

L22
ELEVATION

L21 L21
L20 F ‐ TCU039 L20 F ‐ TCU065
L19 L19
L18 L18
L17
G ‐ Denali L17
G ‐ DENALI
L16 L16
L15 H ‐ Darfield CBG L15 H ‐ CAPEMEND FFS
L14 L14
L13 I ‐ Darfield SHLC L13 I ‐ CAPEMEND LFS
L12 L12
L11
L10
J ‐ Sierra.Mex CIW L11 J ‐ DARFIELD CNH
L10
L09 L09
L08
K ‐ Sierra.Mex CXO L08
K ‐ DARFIELD HORC
L07 L07
L06 Applicable Limit L06 Applicable Limit
L05 L05
L04 Mean L04 Mean
L03 L03
L02 L02
L01 L01
B01 B01
B02 B02
B03 B03
B04 B04
B05 B05
B06 B06

‐13% ‐11% ‐9% ‐7% ‐5% ‐3% ‐1% 1% 3% 5% 7% 9% 11% 13% ‐13% ‐11% ‐9% ‐7% ‐5% ‐3% ‐1% 1% 3% 5% 7% 9% 11% 13%
TOTAL LINK BEAM ROTATIONS (RADIANS) TOTAL LINK BEAM ROTATIONS (RADIANS)

5Figure 6.91: Link Beam Rotations. © Skidmore, Owings & Merrill LLP

104 | Performance-Based Seismic Design Examples

 Pier 2

East Wall
Outer

(ACI 318-14 18.10.6.5 - No Yielding)

Ordinary Boundary - 8"

1x yield ≤ εt < 2x yield

εc ≤ 0.001
Pier 3
Outer
EAST WALL - PIER 2 - OUTER BOUNDARY

L43 A - CHICHI36
L42
L41
L40
L39 B - DUZCE
L38
L37
L36 C - LOMAP
L35
L34
L33
(ACI 318-14 18.10.6.5 - No Yielding)

L32 D - LANDERS

Panel Zone
εt ≤ 1x yield
L31

εc ≤ 0.001
(ACI 318-14 18.10.6.5 - Yielding)
L30
L29
L28 E - CHYO24
L27
Ordinary Boundary - 8"

1x yield ≤ εt < 2x yield

Ordinary Boundary - 6"

2x yield ≤ εt < 10x yield

L26
L25

0.002 > εc > 0.001

L24 F - TCU065
L23
L22
εc ≤ 0.001

L21 G - DENALI
L20
L19
L18
L17 H - CAPEMEND FFS
L16
L15
L14
L13 I - CAPEMEND LFS
L12
L11
L10 J - DARFIELD CNH
L09
L08
L07
L06 K - DARFIELD HORC
L05
L04
L03
L02 C1-SP Mean
L01
B01 Applicable Limit
B02 UNCONFINED
Panel Zone
εt ≤ 1x yield

B03
εc ≤ 0.001

Hinge 2x yield ≤ εt < 10x yield

B04
(ACI 318-14 18.10.6.5 - Yielding)

B05 Applicable Limit Compressive Strain (

max(lw & Mu,CS/4Vu,CS

0.002 < εc ≤ 0.006

CONFINED Unconfined Limit = 0.

(ACI 318-14 18.10.6.4)

-1.5% -1.0% -0.5% 0.0% 0.5% 1.0% 1.5%

Ordinary Boundary - 6"

2x yield ≤ εt < 10x yield

Ordinary Limit = 0.00

Special Boundary
0.002 > εc > 0.001

EAST WALL - PIER 3 - OUTER BOUNDARY 1x yield < εt ≤ 2x yield Special Limit = 0.013
8" Tied Panel Zone

*Reduction in limit by 2 pe
L43 A - CHICHI36
εc ≤ 0.001
L42 **Reduction in limit by 1.5
L41
L40
L39 B - DUZCE Tensile Strain (εt) Lim
L38
L37
L36
Unrestrained bar limit
C - LOMAP
L35 8" spacing bar of rest
Plastic

L34
Zone

L33 6" spacing bar of rest

L32 D - LANDERS
L31
L30
L29
L28 E - CHYO24 lw
L27
L26 εc: Tensile Strain
L25 F - TCU065 εt: Compressive Strain
L24
L23
L22
L21 G - DENALI
L20
L19
Hinge 2x yield ≤ εt < 10x yield

L18 Compressive Strain (εt) Limits

L17
max(lw & Mu,CS/4Vu,CS

H - CAPEMEND FFS
0.002 < εc ≤ 0.006

L16 Unconfined Limit = 0.003/2*/1.5** = 0.001

(ACI 318-14 18.10.6.4)

L15
L14 Ordinary Limit = 0.004/2* = 0.002
Special Boundary

L13 I - CAPEMEND LFS

1x yield < εt ≤ 2x yield

L12 Special Limit = 0.013/2* = 0.006

8" Tied Panel Zone

L11
L10 *Reduction in limit by 2 per Wallace, 2007
L09 J - DARFIELD CNH
L08
εc ≤ 0.001

**Reduction in limit by 1.5 due to force-controlled action

L07
L06 K - DARFIELD HORC
L05
L04
L03
Tensile Strain (εt) Limits
L02 C1-SP Mean Unrestrained bar limit = 1x yield
L01 8" spacing bar of restraint = 2x yield
Plastic

Zone

B01
B02
Applicable Limit 6" spacing bar of restraint = 10x yield
UNCONFINED
B03
B04
Wallace, 2007: Modelling Issues for Tall Reinforced Conc
Applicable Limit
B05
-1.5% -1.0% -0.5% 0.0% 0.5% 1.0% 1.5% CONFINED lw Wall Buildings, The Structural Design of Tall and Special
16:615-632.
5Figure 6.92: Tension & Compression Strain at Core. © Skidmore, Owings & Merrill LLP 5Figure 6.93: Proposed Mapping of NLTH analysis results to ACI 318
Chapter 18.10.6.4.2 provisions. © Skidmore, Owings & Merrill LLP

Performance-Based Seismic Design Examples | 105

 NORTH-WEST NORTH-EAST
CORNER CORNER

SOUTH-WEST SOUTH-EAST
CORNER CORNER
y

South-West Corner X-direction South-West Corner Y-direction

L43 L43
L42 A - CHICHI36 A - CHICHI36
L42
L41 L41
L40 L40
L39 B - DUZCE L39 B - DUZCE
L38 L38
L37 L37
L36 L36
L35 C - LOMAP L35 C - LOMAP
L34 L34
L33 L33
L32 L32
L31 D - LANDERS L31 D - LANDERS
L30 L30
L29 L29
L28 E - CHYO24 L28 E - CHYO24
L27 L27
L26 L26
L25 L25
L24 F - TCU065 L24 F - TCU065
L23 L23
Elevation
Elevation

L22 L22
L21 L21
L20 G - DENALI L20 G - DENALI
L19 L19
L18 L18
L17 L17
L16 H - CAPEMEND L16 H - CAPEMEND
L15 FFS L15 FFS
L14 L14
L13 L13
L12 I - CAPEMEND L12 I - CAPEMEND
L11 LFS L11 LFS
L10 L10
L09 L09 J - DARFIELD
J - DARFIELD
L08 L08 CNH
CNH L07
L07
L06 L06
L05 K - DARFIELD L05 K - DARFIELD
L04 HORC L04 HORC
L03 L03
L02 L02
Mean Mean
L01 L01
B01 B01
B02 Limit B02 Limit
B03 B03
B04 B04
0.0% 1.0% 2.0% 3.0% 4.0% 5.0% 0.0% 1.0% 2.0% 3.0% 4.0% 5.0%

Interstory Drift Ratio (%) Interstory Drift Ratio (%)

5Figure 6.94: Drifts. © Skidmore, Owings & Merrill LLP

106 | Performance-Based Seismic Design Examples

Drifts
Drift levels were in conformance with
the acceptance criteria (see Figure
6.94). One minor exceedance of the
4.5 percent limit on an individual
motion was observed in one ground
motion. This ground motion is
roughly 175 percent MCER at the
building-dominant period.
6.24 Project-Specific Designs
and Detailing
Panel Design Conditions
An important consideration for tall
buildings is the design of the solid
area of concrete above and below
stacks of openings. This issue has been
addressed in NIST GCR 14-917-25
Appendix C “Recommendations for
Seismic Design of Reinforced Concrete
Wall Buildings Based on Studies of the
2010 Maule, Chile Earthquake.” The
recommendations in this report are to
take the design force of these panel
zones as the capacity of adjoining
boundary zones.
5Figure 6.95: Panel Zones Above and Below Stacks of Openings. © Skidmore, Owings & Merrill LLP
For this project, the average panel-zone
shear (see Figure 6.95) multiplied by
1.5 was utilized with expected material
properties, or with peak shear from
ground motions with high link-beam
rotations used.
The above-mentioned guidelines
suggest using the boundary-zone
tensile capacity as a design force, but
this is with the assumption of response-
spectrum analysis, where actual
boundary-zone demands are less clear.
For this project, the practitioners set an
envelope of 1.5 times average panel-
zone shear, with expected material
properties. Peak shear was from ground
motions, and high link-beam rotations
were used. In one case, the peak
shear of a single ground motion was
greater than the average peak shear of
the suite times 1.5. Conservatively, as
recommended in the above-mentioned
design guides, special confinement was
provided for the entire panel zone. Also,
two-way reinforcement was designed
instead of diagonal reinforcement, since
essentially elastic behavior is expected.
Performance-Based Seismic Design Examples | 107
7.0 Conclusion
Conclusion

In the past decade, PBSD has
dramatically increased both the design
flexibility and economic viability of tall
buildings located in the western region
of the United States. Additionally,
it has advanced the knowledge of
structural analysis and design among
practicing engineers and the broader
construction community. With PBSD
principles being implemented in ASCE
7-16, these methods will soon find
acceptance outside of select western
municipalities, opening further
opportunities across the United States.
The practice is also likely to continue
gaining acceptance outside of the
world’s most seismically active regions,
as its challenges and benefits are more
widely disseminated.
within tall buildings, alongside the
ability to provide structural damping
as a consequence of dynamic motion.
It can provide damping for both wind
and seismic events, and also control
the forces within the attached columns.
The outrigger system falls outside of a
standard lateral force-resisting system,
and is thus an ideal candidate for
the use of PBSD. Indeed, some of the
knowledge gained during its analysis
formed the basis of the original CTBUH
seismic guidelines (2008). For further
information see Smith and Willford
(2007) and Smith (2016).
While conventional structural systems
provide sufficient strength and rigidity
for tall buildings, conventional energy-
dissipating elements such as link beams,
braced frames, and moment frame
connections all sacrificially experience
permanent damage in seismic events to
mitigate seismic forces. These elements
require substantial repair or even
replacement after seismic events.

Future Considerations

Future PBSD projects should

begin adopting enhanced seismic
performance objectives, using rating
systems such as those proposed
by USRC and REDi. Additionally,
the enhanced goals demonstrated
by the PBSD methodology should
be incorporated into the strategic
resiliency goals of urban communities
in seismically active areas, which
would greatly reduce downtime after a
seismic event.

Perhaps most importantly, new

seismic force-resisting components
and systems need to be developed
and tested for incorporation into
PBSD projects prior to explicit code
adoption. Examples of this include the
damped outrigger system and the Pin-
Fuse® seismic system. These systems
have been proposed by practicing
engineers to significantly improve
performance and reduce repairs after a
seismic event.

In brief, the damped outrigger

system is a method that combines
the structural efficiency of outriggers 5Figure 7.1 Damped Outrigger System. © Arup

110 | Conclusion
Pin-Fuse joints provide significant
energy dissipation, but do so without
significant damage to the elements. By
emulating the pivotal movements of
a human shoulder joint, more natural
movements can be accommodated.
This joint uses materials that allow
engineers to define a certain coefficient
of friction, so the joint remains fixed,
then slips during high loads. As soon
as the load gets large enough, the
joints start to move. By using this joint
to provide localized flexibility, the
5Figure 7.2 Pin-Fuse Seismic Systems. © Skidmore, Owings & Merrill LLP
earthquake forces can be dissipated and
damage to the structure minimized.
These two examples highlight innovative
systems developed by practicing
engineers to increase resiliency.
Unfortunately, conventional design
codes have limited their use. The design
methods employed by engineers should
demand a high level of technical rigor,
while offering innovative solutions for
our increasingly dense urban centers.
PBSD is successful on both fronts.
It is the intent of this document to
demonstrate the benefits and technical
challenges of PBSD, as locations outside
the United States look to develop
non-prescriptive design methods.
Individuals seeking to develop
methods similar to those described
in this publication are encouraged to
review the recommended resources
herein and contact the Performance-
Based Seismic Design Working Group
of CTBUH.
Conclusion | 111
References
General References for PBSD
ATC 72-1: Modeling and Acceptance Criteria
for Seismic Design and Analysis of Tall
Buildings, Applied Technology Council and
Pacific Earthquake Engineer Research Center,
October 2010. ATC is developing a new
guideline, ATC-114 Guidelines for Nonlinear
Structural Analysis and Design of Buildings,
which is expected in 2017. https://www.
atcouncil.org/.
LATBSDC: An Alternative Procedure for
Seismic Analysis and Design of Tall Buildings
Located in the Los Angeles Region-A
Consensus Document, Los Angeles Tall
Buildings Structural Design Council, 2014
Edition. http://www.tallbuildings.org/.
PEER/TBI: Guidelines for Performance-Based
Seismic Design of Tall Buildings, Pacific
Earthquake Engineering Research Center-Tall
Building Initiative, Report No. 2017/06, May
2017. https://peer.berkeley.edu/research/
building-systems/tall-buildings-initiative.
Relevant Design Standard References
ACI 318-08: Prior version of ACI 318-14.
ACI 318-14: Building Code Requirements
for Structural Concrete and Commentary,
American Concrete Institute, 2014.
ASCE 7-10: Minimum Design Loads for
Buildings and Other Structures, American
Society of Civil Engineers, 2010.
ASCE 7-16: Minimum Design Loads for
Buildings and Other Structures, American
Society of Civil Engineers, 2016. This version
will include NRHA chapter that incorporates
many of the PBSD concepts described in this
and other documents.
ASCE 41-06: Prior version of ASCE 41-13.
ASCE 41-13: Seismic Evaluation and Retrofit
of Existing Buildings, American Society of
Civil Engineers, 2013.
112 | References
AISC 360-10: Specification for Structural
Steel Buildings, American Institute of Steel
Construction 14th Edition, 2010.
AISC 341-10: Seismic Provisions for Structural
Steel Buildings, American Institute of Steel
Construction 14th Edition, 2010.
Select Journal Paper References
Motter (2014). Large-Scale Testing Of Steel-
Reinforced Concrete (SRC) Coupling Beams
Embedded Into Reinforced Concrete Structural
Walls. PhD. University of California Los Angeles.
Naish (2013). 2013A: Reinforced Concrete
Coupling Beams – Part 1: Testing. American
Concrete Institute Structural Journal.
Naish (2013). 2013B: Reinforced Concrete
Coupling Beams – Part 2: Modeling. American
Concrete Institute Structural Journal.
Naish (2010). Testing and Modeling of Reinforced
Concrete Coupling Beams. PhD. University of
California Los Angeles.
Wallace (2007). Modelling Issues For Tall
Reinforced Concrete Core Wall Buildings. The
Structural Design of Tall and Special Buildings
16: 615–32.
City-Specific References
AB-82, Requirement and Guidelines for
Structural Design Review Procedures, San
Francisco Administrative Bulletin (AB)-082.
AB-83, Requirements and Guidelines for the
Seismic Design of New Tall Buildings using
Non-Prescriptive Seismic-Design Procedures,
San Francisco Administrative Bulletin (AB)-083.
Relevant Seismic Design and Analysis
Guidelines
NEHRP Seismic Design Technical Briefs:
http://www.nehrp.gov/.
No. 1 - Seismic Design of Reinforced
Concrete Special Moment Frames: A Guide
for Practicing Engineers Second Edition
(NIST GCR 16-917-40)
No. 2 - Seismic Design of Steel Special
Moment Frames: A Guide for Practicing
Engineers (NIST GCR 16-917-41)
No. 3 - Seismic Design of Cast-in-Place
Concrete Diaphragms, Chords, and
Collectors: A Guide for Practicing Engineers
(NIST GCR 10-917-4)
No. 4 - Nonlinear Structural Analysis For
Seismic Design: A Guide for Practicing
Engineers (NIST GCR 10-917-5)
No. 5 - Seismic Design of Composite Steel
Deck and Concrete-filled Diaphragms: A
Guide for Practicing Engineers (NIST GCR
11-917-10)
No. 6 - Seismic Design of Cast-in-Place
Concrete Special Structural Walls and
Coupling Beams: A Guide for Practicing
Engineers (NIST GCR 11-917-11REV-1)
No. 7 - Seismic Design of Reinforced
Concrete Mat Foundations: A Guide for
Practicing Engineers (NIST GCR 12-917-22)
No. 8 - Seismic Design of Steel Special
Concentrically Braced Frame Systems: A
Guide for Practicing Engineers (NIST GCR
13-917-24)
No. 10 - Seismic Design of Wood Light-
Frame Structural Diaphragm Systems: A
Guide for Practicing Engineers (NIST GCR
14-917-32)
No. 11 - Seismic Design of Steel Buckling-
Restrained Braced Frames: A Guide for
Practicing Engineers (NIST GCR 15-917-34)
Relevant Geotechnical Engineering
References
Abrahamson, N. A. (2011). Update of the
Abrahamson (2000) Directivity Model for
Strike-Slip Earthquakes Appendix K, Report on the
Analysis of the Shoreline Fault Zone, Central Coast
California. Pacific Gas and Electric Company.
Abrahamson, N. (2000). Effects of Rupture
Directivity on Probabilistic Seismic Hazard
Analysis. In: Proceedings of Sixth International
Conference on Seismic Zonation. Palm
Springs, California.
Baker, J. W. and Cornell, C. A. (2006). Spectral
Shape, Epsilon and Record Selection.
Earthquake Engineering & Structural Dynamics
35(9): 1077–95.
Baker, J. W. and Cornell, C. A. (2005). A Vector-
valued Ground Motion Intensity Measure
Consisting of Spectral Acceleration and
Epsilon. Earthquake Engineering & Structural
Dynamics 34(10): 1193–217.
Bozorgnia, Y. and Campbell, K. W. (2016).
Vertical Ground Motion Model for PGA, PGV,
and Linear Response Spectra Using the NGA
West-2 Database. Earthquake Spectra, 32(2).
Cornell, C. A. (1968). Engineering Seismic Risk
Analysis. Bulletin of the Seismological Society of
America 58(5).
Earthquake Solutions (2015). QuakeManager
Software.
Golesorkhi, R. (2002). Near-Source Effects and
Correlations to Recent Recorded Data. In:
Proceedings of Seventh US National Conference
on Earthquake Engineering. Boston.
Gulerce, Z. and Abrahamson, N. (2011). Site-
Specific Design Spectra for Vertical Ground
Motion. Earthquake Spectra 27(4).
Hashash, Y., Musgrove, M., Groholski, D.,
Phillips, C. and Park, D. (2015). DEEPSOIL 6.0,
User Manual and Tutorial. [online] Available at:
http://deepsoil.cee.illinois.edu/.
Hayden, C., Bray, J. and Abrahamson, N. (2014).
Selection of Near-Fault Pulse Motions. ASCE
Journal of Geotechnical and Geoenvironmental
Engineering 140(7).
Idriss, I. M. and Sun, J. I. (1992). “User’s
manual for SHAKE91 a Computer Program
for Conducting Equivalent Linear Seismic
Response Analyses of Horizontally Layered
Soil Deposits – program modified based
on the original SHAKE program published
in December 1972 by Schnabel, Lysmer
and Seed.” Center for Geotechnical Modeling
Department of Civil & Environmental
Engineering, University of California, Davis.
Kempton, J. (2006). Prediction Equations for
Significant Duration of Earthquake Ground
Motions Considering Site and Near-Source
Effects. Earthquake Spectra 22(4).
Matasovic, N. and Ordonez, G. (2011).
D-MOD 2000 A Computer Program Package
for Seismic Response Analysis of Horizontally
Layered Soil Deposits, Earthfill Dams and Solid
Waste Landfill. GeoMotions, LLC.
Mazzoni, S., Hachem, M. and Sinclair, M.
(2012). An Improved Approach for Ground
Motion Suite Selection and Modification
for Use in Response History Analysis. In:
Proceedings of the Fifteenth World Conference
on Earthquake Engineering. Lisbon.
McGuire, R. (2004). Seismic Hazard and Risk
Analysis Monograph MNO-10. Earthquake
Engineering Research Institute.
NIST (2012). “Soil-Structure Interaction for
Building Structures.” National Institute of
Standards and Technology, NEHRP Consultants
Joint Venture, A partnership of the Applied
Technology Council and the Consortium
of Universities for Research in Earthquake
Engineering, NIST GCR 12-917-21.
NIST (2011). “Selecting and Scaling
Earthquake Ground Motions for Performing
Response-History Analyses.” National
Institute of Standards and Technology, NEHRP
Consultants Joint Venture, A partnership
of the Applied Technology Council and the
Consortium of Universities for Research in
Earthquake Engineering, NIST GCR 11-917-15.
Silva, W., Abrahamson, N., Toro, G. and
Costantino, C. (1996). Description and
Validation of the Stochastic Ground Motion
Model. Brookhaven National Laboratory.
Shahi, S. (2014). NGA-West2 Models for
Ground Motion Directivity. Earthquake
Spectra 30(3).
Shahi, S. and Baker, J. (2017). Regression
Models for Predicting the Probability of
Near-fault Earthquake Ground Motion Pulses,
and Their Period. In: Proceedings of 11th
International Conference on Applications of
Statistics and Probability in Civil Engineering.
Zurich: .8.
Somerville, P., Smith, N., Graves, R. and
Abrahamson, N. (1997). Modification of
Empirical Ground Motion Attenuation
Relations to Include Amplitude and Duration
Effects of Rupture Directivity. Seismological
Research Letters 68: 199–222.
Spudich, P., Rowshandel, B., Shahi, S., Baker,
J. and Chiou, B. (2014). Comparison of
NGA-West2 Directivity Models. Earthquake
Spectra 30(3).
Travasarou, T., Bray, J. and Abrahamson, N.
(2003). Empirical attenuation relationship
for Arias Intensity. Earthquake Engineering
Structural Dynamics 32: 1133–55.
Walker, M., Golesorkhi, R. and Hachem,
M. (2017). From Recording to Time
History Analysis: A Primer on Maximum
Direction Ground Motion. In: Proceedings of
Convention of Structural Engineers Association
of California.
Enhanced Performance Objectives
Publications.arup.com. (n.d.). REDi: Resilience-
based Earthquake Design Initiative. [online]
Available at: https://www.arup.com/
perspectives/publications/research/section/
redi-rating-system.
Usrc.org. (n.d.). USRC: United States Resiliency
Council Rating System. [online] Available at:
http://www.usrc.org/.
References | 113
About the CTBUH

The Council on Tall Buildings and Urban Habitat (CTBUH) is the world’s leading
resource for professionals focused on the inception, design, construction, and
operation of tall buildings and future cities. Founded in 1969 and headquartered at
Chicago’s historic Monroe Building, the CTBUH is a not-for-profit organization with
an Asia Headquarters office at Tongji University, Shanghai; a Research Office at Iuav
University, Venice, Italy; and an Academic Office at the Illinois Institute
of Technology, Chicago. CTBUH facilitates the exchange of the latest knowledge
available on tall buildings around the world through publications, research, events,
working groups, web resources, and its extensive network of international
representatives. The Council’s research department is spearheading the
investigation of the next generation of tall buildings by aiding original research on
sustainability and key development issues. The Council’s free database on tall
buildings, The Skyscraper Center, is updated daily with detailed information,
images, data, and news. The CTBUH also developed the international standards for
measuring tall building height and is recognized as the arbiter for bestowing such
designations as “The World’s Tallest Building.”

www.ctbuh.org
www.skyscrapercenter.com

About CTBUH Technical Guides

Each of these guides is the product of a CTBUH Working

Group formed specifically to address focused topics in
the industry. The intention of each guide is the same – to
provide working knowledge to the typical building owner or
professional who wants a better understanding of available
options for improving tall buildings, and what affects their
design. The object of the series is to provide a toolkit for the
creation of better-performing tall buildings, and to spread
the understanding of the considerations that need to be 2019, 136 pages 2018, 204 pages 2018, 148 pages
made in designing tall. Other guides in this series include: ISBN: 978-0-939493-68-5 ISBN: 978-0-939493-64-7 ISBN: 978-0-939493-63-0

2017, 224 pages 2017, 116 pages 2017, 104 pages 2014, 240 pages 2013, 48 pages 2012, 184 pages
ISBN: 978-0-939493-57-9 ISBN: 978-0-939493-56-2 ISBN: 978-1-864707-28-1 ISBN: 978-186470-593-5 ISBN: 978-0-415714-59-4 ISBN: 978-0-415-50958-9

114 | About the CTBUH

About the Authors
Ramin Golesorkhi
Langan Engineering
Ramin Golesorkhi, PhD, PE, GE
leads Langan’s Seismic Engineering
Services. He has nearly three decades
of worldwide experience in earthquake and foundation
engineering, with projects throughout the United States,
Latin America, Southeast Asia, India and Middle East.
Ramin develops seismic and geotechnical design criteria
appropriate for tall, industrial, residential, private and
government office buildings, hospitals and healthcare
facilities, bridges, elevated freeways and viaducts, structures
with energy dissipation systems, and tunnels.
Ron Klemencic
Magnusson Klemencic Associates
Ron Klemencic is Chairman and C.E.O.
of Magnusson Klemencic Associates, an
award-winning, 200-person structural
and civil engineering firm headquartered
in Seattle, Washington, with a branch office in Chicago. A
past five-year Chairman of the Council on Tall Buildings and
Urban Habitat, Ron's focus is complex high-rise and mixed-
use designs. He has worked on projects in 24 states and 24
countries, with developments up to 650,000 m² (7 million ft²),
and is sought out by developers, architects, and contractors
for his creativity, “big picture” approach, and unique ability to
consistently produce cost-effective, innovative designs.
John Viise
Thornton Tomasetti
For 24 years, John Viise has been
providing structural services for high-rise
and special-use structures throughout
the world. He continues to apply his
engineering expertise on cutting-edge, international projects
and has led efforts for performance-based design on a number
of projects. John’s interdisciplinary background in structural
and mechanical engineering makes him uniquely qualified for
the wind and seismic engineering challenges of tall building
design. He is at the forefront of technical design, through active
participation in industry research and knowledge-sharing.
Leonard Joseph
Thornton Tomasetti
With more than 40 years of experience,
Leonard Joseph has analyzed, designed
and reviewed high-rise buildings, sports
facilities, hangars, hotels, historic buildings, manufacturing
facilities, and parking garages. He works with a wide variety
of materials, including structural steel, reinforced concrete,
precast and post-tensioned concrete, masonry, wood, and
light-gage framing. For buildings around the world, Leonard
deals with seismic, wind and other environmental hazards,
and incorporates local construction practices into his designs.
David Shook
Skidmore, Owings & Merrill LLP
David Shook is an Associate Director
of structural engineering with
Skidmore, Owings & Merrill LLP, in
San Francisco. His work has spanned structural topics as
diverse as performance-based seismic design, integration
of multidisciplinary data and optimization, mass timber
in high-rise construction, and deep sustainability. His
design experience includes the design of several large-
scale iconic international and domestic towers, as well
as collaborations with students and artists. David is a
frequent lecturer at universities and conferences, both
domestically and internationally.
About the Authors | 115
CTBUH Organizational Members (as of 20 December 2019)

Supporting Contributors CityGroup (CTG) Design CCD / Cheng Chung Design Longman Lindsey
COIMA CCL M Moser Associates
AECOM Country Garden Cerami & Associates Maeda Corporation
alinea Consulting EID Architecture Cermak Peterka Petersen Manntech
Arcadis Enclos Corp. China Architecture Design & Research Group MAURER SE
Autodesk Envision Engineering Consultant China State Construction Overseas Development Metal Yapi
BuroHappold Engineering Epstein Chongqing Jinke Design Research Institute MicroShade A/S
CCDI Group Fender Katsalidis Civil & Structural Engineering Consultants (Pvt) Ltd. Moelven
CITIC HEYE Investment Co., Ltd. Front Inc. Code Consultants, Inc. Mori Building Company
Dassault Systèmes Halfen United States Conrad Gargett Mott MacDonald Group
Dow Chemical Company Hanking Group Cosentini Associates MVRDV
EFC Engineering Consulting Co., Ltd. Hill International Cottee Parker Architects Nabih Youssef & Associates
Illinois Institute of Technology Investa Property Group Cotter Consulting Inc. National Fire Protection Association
IUAV University of Venice Jensen Hughes Council on Tall Buildings and Urban Habitat NIKKEN SEKKEI LTD
Kingdom Real Estate Development JLL Cox Architecture Norman Disney & Young
Kohn Pedersen Fox Associates Larsen & Toubro CoxGomyl NORR Group Consultants International Limited
KONE LeMessurier Craft Holdings Limited O'Donnell & Naccarato
Multiplex LERA Consulting Structural Engineers CS Group Construction Specialties Company OLYMPIQUE Facade Access Consulting
Otis Elevator Company LWK & Partners Cubic Architects Omrania
Ping An Real Estate Co., Ltd. Magnusson Klemencic Associates Daewoo E&C Ornamental Metal Institute of New York
PS-Co. McNAMARA • SALVIA Davy Sukamta & Partners Structural Engineers Palafox Associates
Schindler Mirvac Group DCA Architects PAN Partners
Shanghai Tower Construction & Development Nishkian Menninger Consulting and Structural DCI Engineers Pavarini McGovern
Shenzhen Parkland Group Co., Ltd. Engineers Decibel Architecture Peikko
Siemens Smart Infrastructure OJB Landscape Architecture Deerns Pepper Construction Company
Skidmore, Owings & Merrill Outokumpu DIALOG Perkins and Will
Sufrin Group PDW Architects Dong Yang Structural Engineers Plus Architecture
Sun Hung Kai Properties Limited Pei Cobb Freed & Partners EG Portman Architects
Taipei Financial Center Corp. Permasteelisa Group Elenberg Fraser Procore Technologies
Tongji University Pickard Chilton Architects Elevating Studio Pte. Ltd. Profica
Turner Construction Company PLP Architecture Enstruct Group Pty Ltd Qingdao Conson Hai Tian Center Construction
Wentworth House Partnership Limited PNB Merdeka Ventures Sdn. Berhad Environmental Systems Design R.G. Vanderweil Engineers
WSP PT. Gistama Intisemesta EPEXYL S.A. Raftery CRE, LLC
YUAN LIH Construction Co., Ltd. Quadrangle Architects Eric Parry Architects RAW Design
Yuanda Group (CNYD) Ramboll FINE DNC Real Estate Management (UK) Limited
Rothoblaas Fletcher Priest Architects Related Midwest
SAMOO Architects and Engineers FM Global Rhode Partners
Patrons Schuco Forster Engineering Consultants Co., Ltd. Rise Management Consulting Ltd
Severud Associates Consulting Engineers Foster + Partners RJC Engineers
BG&E Shanghai Construction (Group) General FXCollaborative Robert A.M. Stern Architects
BMT Shenzhen Aube Architectural Engineering Design GEI Consultants Rogers Stirk Harbour + Partners
Brandston Partnership, Inc. Sika Services AG GERB Vibration Control Systems (USA/Germany) Ronald Lu & Partners
Centralcon Group Studio Gang GGLO Ronesans Holding
Dar Al-Handasah (Shair & Partners) Syska Hennessy Group Gilsanz Murray Steficek Royal HaskoningDHV
DeSimone Consulting Engineers Tata Realty Global Wind Technology Services Sanni, Ojo & Partners
East China Architectural Design & Research Institute TAV Construction Glumac Sauerbruch Hutton Gesellschaft von Architekten
Emaar Properties, PJSC Tongji Architectural Design Group gmp • Architekten von Gerkan, Marg und Partner SECURISTYLE
Gensler UNStudio GOA (Group of Architects Co., Ltd) SETEC TPI
HOK, Inc. V & A Waterfront Goettsch Partners Shimizu Corporation
Hongkong Land Walter P. Moore and Associates Gradient Wind Engineering Inc. SHoP Architects
ISA Architecture WATG Urban Graziani + Corazza Architects Siderise
Jaeger Kahlen Partners Architects Ltd. Webber Design Pty Ltd Grimshaw Architects SilverEdge Systems Software, Inc.
KLCC Property Holdings Berhad Webcor Builders Guangzhou Jianke Citiexpo Co.,Ltd SimpsonHaugh
Kuraray America, Inc. Willow Guangzhou Yuexiu City Construction Jones Lang Stanley D. Lindsey & Associates
Langan Engineering WME Engineering Consultants LaSalle Property Management Co., Ltd. Stantec Ltd.
Meinhardt Group International Woods Bagot Hariri Pontarini Architects Steel Institute of New York
NBBJ Wordsearch 添惠达 HASSELL Stein Ltd.
Pace Development Corporation Plc. Yitian Design Group Co., Ltd. Hathaway Dinwiddie Stora Enso Wood Products Oy Ltd
Pelli Clarke Pelli Architects Zaha Hadid Architects Heller Manus Architects Studco Australia Pty Ltd
POHL Group Henning Larsen Architects SuperTEC
Priedemann Facade Experts Hilti AG Surface Design
Rene Lagos Engineers Contributors Hitachi, Ltd SWA Group
Rider Levett Bucknall HKA Elevator Consulting Swinerton Builders
Rowan Williams Davies & Irwin AkzoNobel HKS Architects Taisei Corporation
SL Green Realty Corp. Aliaxis HOK Architects Corporation Takenaka Corporation
Softwood Lumber Board Alimak Housing and Development Board Technal Middle East
Studio Libeskind Allford Hall Monaghan Morris HPP Architects Tengyuan Design Institute Co., Ltd
The Durst Organization Altitude Facade Access Consulting Humphreys & Partners Architects, L.P. Terracon
Thornton Tomasetti Alvine Engineering Hutchinson Builders Tetra Tech
thyssenkrupp Elevator AMSYSCO ICD Property The Harman Group
Tishman Speyer Andrew Lee King Fun & Associates Architects Ltd. IDOM UK Ltd. The Pakubuwono Development
Windtech Consultants ArcelorMittal Inhabit Group Vetrocare
Archilier Architecture International Code Council Vidaris, Inc.
architectsAlliance Interpane GmbH Voice Architecture Lab
Donors Architectural Design & Research Institute of Irwinconsult VS-A Group
Tsinghua University Israeli Association of Construction and Werner Sobek Group
A&H Tuned Mass Dampers Architectus Infrastructure Engineers Weston Williamson + Partners
Adrian Smith + Gordon Gill Architecture Armstrong Ceiling Solutions ITT Enidine wh-p Ingenieure
Aedas Arney Fender Katsalidis JAHN WilkinsonEyre
AKF Group ASHTROM GROUP LTD Jaros, Baum & Bolles WOHA Architects
Al Ghurair Construction Barker Mohandas, LLC Jotun WTM Engineers International
Architects Hawaii, Ltd. bKL Architecture JQZ WZMH Architects
Architectural Design & Research Institute of South Bonacci Group KEO International Consultants Y. A. Yashar Architects
China University of Technology Boundary Layer Wind Tunnel Laboratory KHP Konig und Heunisch Planungsgesellschaft
Arup Bouygues Batiment International Killa Design Participants/Academic & Media Institutes
Aurecon Broad Sustainable Building Co. Kinemetrics Inc.
BALA Engineers Broadway Malyan Kinetica There are an additional 311 members of the Council
Bates Smart Brunkeberg Systems Kobi Karp at the Participant/Academic Institute/Media Institute
Beijing Fortune Lighting System Engineering Calatrava International Koltay Facades level. Please see online for the full member list.
Bjarke Ingels Group Canary Wharf Group KS Ingenieure ZT GmbH http://members.ctbuh.org
Bosa Properties Inc. Canderel Management LCI Australia Pty Ltd.
Carazo Architecture Careys Civil Engineering LCL Builds Limited
China Construction Steel Structure Corp. Ltd. Cary Kopczynski & Company Lendlease Corporation
China International Marine Containers (Group) Ltd. CB Engineers Liberty OneSteel

Supporting Contributors are those who contribute $10,000; Patrons: $6,000; Donors: $3,000; Contributors: $1,500; Participants: $750; Academic & Media Institutes: $500.

116 | CTBUH Organizational Members